UNITED STATES � ANTI-DUMPING MEASURES ON
STAINLESS STEEL PLATE IN COILS AND STAINLESS
STEEL SHEET AND STRIP
FROM KOREA

Report of the Panel

(Continued)

II. RESPONSES OF THE UNITED STATES TO QUESTIONS POSED BY KOREA

A. UNPAID SALES

Q.1. In its oral statement, the United States describes that it classified POSCO's US sales through POSAM as constructed-export-price ("CEP") sales, while it classified POSCO's direct sales to US customers as export-price ("EP") sales.⁸³ The United States then describes that it made (or did not make) the following adjustments for "bad debt" expenses:

For the CEP sales, the United States deducted the allocated US "bad debt" expense from the price to the first unaffiliated customer.⁸⁴

For the CEP sales, the United States did not add the allocated US "bad debt" expense to the normal value that was compared to the constructed export price.⁸⁵

For the EP sales, the United States did not deduct the allocated US "bad debt" expense from the sales price to the first unaffiliated customer.⁸⁶

The United States did not address the fourth circumstance, however. In other words, the United States did not state whether, for those sales classified as EP sales, the allocated US "bad debt" expense was added to the normal value that was compared to the export price. Would the United States please confirm whether the allocated US "bad debt" expense was added to the normal value that was compared to the export price for the sales classified as "EP" sales?

74. As the United States explained in its first submission and oral statement, the circumstance of sale adjustment for direct selling expenses was made by deducting from normal value the direct selling expenses for domestic sales in Korea and adding to normal value the US direct selling expenses. The United States further explained that the allocated portion of bad debt was treated as a direct selling expense. Therefore, for POSCO's sales directly to independent US buyers - i.e., EP sales - the addition to normal value included an allocated portion of the US bad debt expense.⁸⁷

Q.2. If the allocated US "bad debt" expense was added to the normal value that was compared to the export price for EP sales, is it the position of the United States that this adjustment was authorized by Article 2.3 of the Anti-Dumping Agreement?

75. As the United States explained in its first submission and oral statement, the circumstance of sales adjustment was made to allow for differences in conditions and terms of sale, consistent with Article 2.4 of the AD Agreement, not as a deduction to construct export price, consistent with Article 2.3. of the AD Agreement. Therefore, in the comparison of EP sales and normal value, the addition to normal value that was made as part of the circumstance of sale adjustment, and which included an allocated portion of bad debt, was an Article 2.4 allowance.⁸⁸

Q.3. Is it the US position that there are no limits on adjustments that may be made under Article 2.3 of the Anti-Dumping Agreement? If there are limits, what are they?

76. Korea has not asserted a claim under Article 2.3 of the AD Agreement, therefore, this issue is not before the Panel. Nevertheless, it is the view of the United States that deductions to construct export price must be consistent with the object and purpose of Article 2.3, which is to construct the export price between the exporter and the associated importer. Beginning with the price to the first independent buyer and deducting all expenses related to the resale by the associated importer is a reasonable methodology for constructing export price, which is consistent with the object and purpose of Article 2.3.

Q.4. The United States contends that, if "the seller agrees to sell on credit ... [it] accepts a credit expense, including the risk of non-payment."⁸⁹ It also states that, in this case, "POSCO agreed to sell to this US customer on credit and in doing so accepted the risk of non-payment as a condition of sale."⁹⁰ The United States contends that POSCO's acceptance of this risk "is sufficient to warrant including bad debt in an Article 2.4 adjustment."⁹¹ In light of these assertions,

a. What evidence was there in the Plate and Sheet cases regarding the risk of non-payment at the time that POSCO made its US sales to the ABC Company (and, according to the United States, agreed to accept the risk of non-payment)?

77. As the United States explained in its first submission and oral statement, the risk of bad debt is inherent in an agreement to sell on credit rather than demand payment on delivery. The fact that these sales were made on credit establishes that POSCO accepted the risk of non-payment at the time of sale.

b. Was there any evidence in the Plate and Sheet cases demonstrating that, at the time POSCO made its sales to the ABC Company, there was a difference in the risk of non-payment for sales to customers in the United States and for sales to customers in Korea?

78. As discussed above, when a seller sells on credit there is an inherent risk of non-payment. That risk may differ between the two markets and thus have different effects on prices in the two markets. Accordingly, the United States makes an adjustment to eliminate such differences. Because it is the only practical means of making due allowance for any difference in bad debt risk, we base the allowance on the company's actual experience in each of the two markets during the period of investigation. That is, we rely on the actual bad debt expenses the company recognizes with respect to each of the two markets being compared. The evidence placed on the record in the cases at issue showed that, during the period of investigation, POSCO actually recognized greater bad debt expenses, as a proportion of sales, in the US market than in the Korean market.

c. If there was no difference in the risk of non-payment for sales in Korea and the United States at the time POSCO made its sales, what is the basis for claiming that there was a difference in the conditions and terms of sale?

79. As discussed above, the evidence placed on the record in the cases at issue showed that, during the period of investigation, POSCO actually recognized greater bad debt expenses, as a proportion of sales, in the US market than in the Korean market.

Q.5. Does the actual bad debt experience that occurs after a sale is made provide an accurate measure of the risk of bad debt before the sale was made?

80. As discussed above, because it is the only practical means of making a due allowance for that difference, we base the allowance on the company's actual bad debt experience in the two markets during the period of investigation. That is, we rely on the actual bad debt expenses the company recognizes with respect to each of the two markets being compared. Furthermore, in accordance with GAAP, companies normally account for bad debt through a reserve accounting method that is based on the company's experience. Therefore, the bad debt expenses reflected in a company's accounting records for a particular market provides a reasonable measure of the price effect of the risk of bad debt in that market.

Q.6. The United States asserts that the risk of non-payment is equivalent to the risk that a warranty expense will be incurred. The past decisions by the United States have recognized that warranty expenses may fluctuate from year to year and that, where there are fluctuations in warranty expenses, the use of a historical average is appropriate to avoid distortions. The standard questionnaire used by the Department of Commerce therefore requests information on the historical warranty expense experience for sales in the home market and the United States. Did the Department of Commerce request information on POSCO's historical bad debt experience in the home market or the United States in the Plate and Sheet cases? Did the United States evaluate whether the bad debt experience during the investigation periods in those cases was consistent with POSCO's historical experience?

81. As discussed above, companies normally account for bad debt using a reserve accounting method that is based on the company's historical experience. Under a reserve accounting method, the bad debt expense recognized in a given period reflects that experience. Therefore, unlike warranty expense, the bad debt expense reported by an exporter normally reflects the company's historical experience and no additional historical evidence is necessary.

82. The United States learned at verification that POSAM did not use a bad debt reserve. Therefore, the actual amount of bad debt recognized during the period was the only reasonable basis for the adjustment. Furthermore, POSCO only argued that the sales to ABC company should be excluded or, in the alternative, that the bad debt should be allocated over a broader category of sales. POSCO never argued that the amount of bad debt recognized during the period should be adjusted to reflect experience and never provided any data establishing that such an adjustment was appropriate and which could be used as a basis for making such an adjustment.

Q.7. The United States has contended that, under US law, adjustments for differences in "conditions and terms of sale" are referred to as "circumstance-of-sale" adjustments.⁹² The Department of Commerce's regulations state that circumstance-of-sale adjustments will normally be made only for "direct selling expenses" (or for expenses that the seller assumes on behalf of the buyer).⁹³ In these circumstances, what is the significance of the finding by the Department of Commerce that the "bad debt" expense on POSCO's US sales was a "direct" selling expense?

83. As explained previously, "direct" selling expenses are those related to the conditions and terms of sale, which are the basis for the circumstance of sale adjustment to normal value. Because bad debt is a direct selling expense, it is included in the circumstance of sale adjustment.

Q.8. What evidence is there on the records of the Plate and Sheet investigations to indicate that POSCO (or any of its affiliates) experienced actual bad debt expenses on any sales of Plate or Sheet to the United States prior to the sale to the ABC Company?

84. See the United States' response to question 3 (Treatment of Unpaid Sales) from the Panel.

Q.9. Is it the position of the United States that investigating authorities may include any export sales in the calculation of export price, no matter how aberrational they are and no matter how much their inclusion distorts the calculation of the dumping margins? Is it the US position that there are no limits at all on the investigating authorities' discretion to include export sales in the price comparison, including the "fair comparison" requirement of Article 2.4?

85. See the United States' response to question 9 (Treatment of Unpaid Sales) from the Panel.

B. MULTIPLE AVERAGING

Q.1. In employing its "multiple-averaging" comparison methodology (i.e., comparing sub-period averages to sub-period averages and combining the results) in the Plate and Sheet cases, did the Department of Commerce treat the dumping margins for sub-periods that had negative margins as if they were zero margins? In other words, did the United States employ the practice known as "zeroing"?

86. Yes. The United States measured the amount of dumping as the amount by which "the export price of the product exported from one country to another is less than the comparable price, in the ordinary course of trade, for the like product when destined for consumption in the exporting country," in accordance with Article 2.1 of the AD Agreement. After measuring that difference, the United States did not offset the amount calculated under Article 2.1 by an amount by which some products may have been sold at greater than the normal value.

87. The United States first notes, as did third party Japan,⁹⁴ that this issue is not properly before the Panel, and should not be addressed by the Panel. Korea never mentioned this issue in its request for panel review, nor did it do so during consultations. Indeed, Korea has never directly asserted that the practice it refers to as "zeroing" is not authorized by the AD Agreement. Consequently, this question is simply irrelevant for purposes of determining whether the United States properly used multiple averaging in the cases before the Panel.

88. In any event, the practice referred to as "zeroing" is not inconsistent with the AD Agreement. Consequently, should the Panel conclude that the issue is properly before it, the Panel should uphold the action of the United States with respect to this issue. An interpretation of Article 2.4.2 which prohibited "zeroing," if taken to its logical conclusion, would distort many of the requirements of Article 2.4 for a fair comparison and the making of due allowances for differences which affect price comparability.

89. All that Article 2.4.2 requires is that, in making comparisons between the export price and the normal value of the like product in an investigation, each comparison shall be made either on a weighted-average-to-weighted-average basis or a transaction-to-transaction basis. This requirement of comparing weighted-average-to-weighted-average figures or transaction-to-transaction figures is explicitly made subject to the requirements of Article 2.4. Thus, it is clear that the considerations discussed in Article 2.4 may be relevant in determining whether sales are comparable, and thus should be included in the same weighted-average.

90. The considerations of time, levels of trade and physical characteristics addressed in Article 2.4 illustrate that if adjustments are not made to eliminate such differences, then comparisons normally must be made on a level-of-trade basis, a product-specific basis (to account for differences in physical characteristics) and a time-period basis. Thus, there may be several to several thousand export prices and normal values which are compared within an investigation for a respondent company, depending on, inter alia, the variety of products, levels of trade, and selling conditions over time. In the cases before the Panel, the United States not only created different averages based on whether the transactions occurred before or after the devaluation, but also based on the physical characteristics of the products. Thus, these cases involved a multitude of average normal values and average export prices. This multitude of comparisons resulted in the calculation of varying dumping amounts for some US sales, and no dumping for other US sales.

91. It is worth recognizing that Article 2.4.2 was newly introduced with the Uruguay Round of negotiations to address a specific concern of certain Members with respect to the conduct of investigations. Previously, the practice of some Members, including the EC and the United States, was to compare individual export price transactions to weighted-average normal values. Article 2.4.2 was included in the AD Agreement to provide that, except in the case of targeted dumping, margin calculations in an investigation would be made on a consistent basis, i.e., weight-average to weight-average or transaction to transaction.⁹⁵ Thus, the intent was to eliminate transaction-to-average comparisons in investigations, not to alter the manner in which authorities calculated overall margins after all appropriate comparisons were made.

92. As discussed above, Articles 2.4 and 2.4.2 provide for a fair comparison between export price and normal value and further provide that such comparisons in investigations should normally be on a weight-average-to-weight-average basis or on a transaction-to-transaction basis. The "zeroing" practice is not covered by Articles 2.4 and 2.4.2 because it arises at a step subsequent to the comparison of export price and normal value. The "zeroing" took place at the stage when the margins from each comparison of comparable weighted-average normal values and export prices were combined into an overall average rate of dumping for the entire investigation.⁹⁶ The lack of guidance on the "zeroing" practice is confirmed by the fact that Article 2.4.2 explicitly permits transaction-to-transaction comparisons without providing a methodology for combining margins calculated pursuant to that methodology.

93. When this stage is reached, the individual, product-specific differences between normal value and export price may be positive or negative. If positive, they represent the aggregate amount of dumping duties that the importing country is permitted to collect for that product or group of transactions. If negative, they represent the amount by which the export price exceeded the normal value; however, the AD Agreement imposes no liability on the importing country to make payments to the importer or anyone else involved in the transaction for not dumping the merchandise in question. The negative difference between normal value and export price simply means there is no dumping; i.e., the dumping margin for that product or group of transactions is zero. Thus, for such products with no dumping margins, the amount of dumping duties which the importing country is permitted to collect is zero.

94. Equally important, when the investigating authority calculates an overall, average rate of dumping, neither Article 2.4.2 nor any other Article of the AD Agreement requires that more credit be given for negative dumping amounts than if the dumping duties were to be collected on a product-specific basis. Nevertheless, that would be the result if Article 2.4.2 were interpreted to prohibit zeroing. This problem may be illustrated with the following example:

	Aggregate Domestic Market Value in Currency Units (CU)	Aggregate Importing Country Value in CU	Aggregate Dumping Amount Calculated in CU	Product-specific Aggregate Dumping Duties Which May be Collected in CU
Product 1	5500	5000	500	500
Product 2	1800	2000	-200	0
Product 3	3300	3000	300	300
Product 4	4500	5000	-500	0
Product 5	2200	2000	200	200
TOTAL	17300	17000	300	1000

95. Based on the above figures, the overall, average rate of dumping is 5.88 percent (1000/17000). Moreover, the application of that dumping margin to the total import value (5.88 percent * 17000) would result in the collection of 1000 CU in dumping duties - no more and no less than the importing country is permitted to collect on a product-specific basis.

96. By contrast, if this calculation were to be performed by offsetting aggregate dumping amounts by negative amounts, the overall, average rate of dumping would be 1.76 percent (300/17000). Even if we were to ignore the fact that this is a de minimis margin, the application of that margin of dumping to the total import value (1.76 percent * 17000) would result in the collection of only 300 CU in dumping duties. Stated another way, there would be an additional 700 CU in dumping which the importing country would not be permitted to remedy. Moreover, because, as noted above, this methodology would result in the calculation of a de minimis dumping margin, the importing country would actually be unable to place any dumping duties on these products, despite the fact that the majority of the products (on a value and volume basis) were dumped at an average rate of 10 percent (1000/10000).⁹⁷

97. Similarly, Article 2.4.2 does not require positive margins to be offset by negative margins because it would fail to give meaning to the requirements of Article 2.4, which, as noted above, contemplate that comparisons be made at least on a product-specific basis in order to account for differences in physical characteristics. This failure may be observed utilizing the above example by noting that the difference between the total aggregate home market prices and the total aggregate importing country prices (17300-17000) is 300 CU. In other words, a methodology in which positive margins had to be offset by negative margins necessarily obscures the results of the product-specific comparisons and is equivalent to simply aggregating normal values and export prices regardless of comparability. In other words, if zeroing is not permitted by Article 2.4.2, then the limitation of averaging to "comparable" transactions provided for in Article 2.4.2 has no meaning.

98. For these reasons, the United States believes that treating negative differences between normal values and export prices, calculated on a product-specific basis, as simply zero dumping is a permissible interpretation of Articles 2.4 and 2.4.2.

Q.2. Article 2.4 specifically refers to a number of "differences" for which adjustments are permitted because they affect "price comparability." This list includes differences in physical characteristics and differences in levels of trade. Where does Article 2.4 indicate that differences in the timing of sales constitute "differences affecting price comparability"?

99. Article 2.4 specifically states that comparisons shall be made in respect of sales made at as nearly as possible the same time. Thus, sales which are not made at as nearly as possible the same time are not to be compared; that is, they are not "comparable" within the meaning of Article 2.4. This requirement of Article 2.4 could be read to mean that comparisons must be limited to sales made on the same day, if possible. However, the United States believes that this provision can also reasonably be read to mean that multiple-day averages are permitted, provided that investigating authorities adopt approaches which limit the maximum amount of time which could occur between sales in the two averages being compared.

100. In most cases, where no evidence has been presented to contradict this conclusion, the United States determines that sales made within the same year are comparable.⁹⁸ Under this reading, in general, normal value transactions made during one year, would not be considered comparable to export transactions made in a different year. Similarly, averaging together sales made during a five-year period of investigation would not achieve comparability.

101. Although the United States normally applies a one-year standard for determining that sales have been made at the same time, it applies a shorter period when circumstances warrant. For example, in situations of high inflation, the United States uses one month averages in order to prevent the inflation from distorting the comparison. Similarly, in the present case the United States constructed two averaging periods in order to prevent the 50 percent devaluation from distorting the comparison.

C. DOUBLE CONVERSION OF LOCAL SALES

Q.1. If the prices for the "local sales" were fixed in dollars, as POSCO contended, would it have been appropriate for the Department of Commerce to base its normal-value calculations on won amounts shown on the invoices that did not correspond to the amounts actually paid? Would it have been appropriate in such circumstances to base the normal-value calculations on the US dollar prices set forth on the invoices?

102. The principal issue is whether the currency of the transaction is Korean won or US dollars. If the currency of the transaction were US dollars, then it would be appropriate to base normal value calculations on the US dollar amount and no currency conversion would be required or appropriate. If the evidence, however, indicates that the transaction was made in Korean won, then it is appropriate under the AD Agreement to convert the sales into US dollars through the application of the currency conversion rules prescribed in Article 2.4.1, as was done in these investigations.

Q.2. When the United States reviewed the local sales during the verifications in the Plate and Sheet cases, did it confirm that the dollar prices shown on the invoices matched the dollar prices reported by POSCO?

103. In both investigations, the United States verified that both the dollar and won amounts reported to the United States reflected the amounts on the invoices.⁹⁹

Q.3. The United States admits that POSCO told it during the course of both the Plate and Sheet investigations that the prices for its "local sales" were fixed in dollars and not in won.¹⁰⁰ The United States also admits that, in the verification in the Sheet case, the Department of Commerce obtained evidence confirming that the amount the customer actually paid for these "local sales" was based on the dollar prices shown on the invoices, and not on the won amounts shown on the invoices.¹⁰¹  Was there any evidence indicating that any Korean customer who purchased Plate or Sheet in a "local sale" actually paid the won amount shown on the invoice? If there was no evidence that any Korean customer who purchased Plate or Sheet in a "local sale" actually paid the won amount shown on the invoice, what evidence was there to refute POSCO's testimony that the prices for these sales were fixed in dollars, and not in won?

104. POSCO reported these sales in won based on won prices on its invoices and recorded in its accounting records. Although POSCO later claimed that the prices were fixed in dollars, POSCO did not present, and the United States did not discover, sufficient evidence to substantiate this claim. To the contrary, the weight of the evidence provided by POSCO indicated that these were won transactions, as POSCO originally reported. As explained in response to question 15 from the Panel, there was no evidence in SSPC to suggest that the won amount paid was anything other than the amount reported by POSCO. The verification in SSSS provided some evidence that the won amounts on the invoices did not reflect the amounts actually paid but the evidence was very limited and came late in the proceeding.

Q.4. Is it the normal practice of the Department of Commerce to verify that the amounts shown on the invoices to home-market customers correspond to the amounts actually paid by those customers? Is there any evidence that the Department of Commerce departed from that practice in the Plate and Sheet investigations?

105. The United States verifies information to the extent practicable. The United States' normal practice is to spot check the accuracy and adequacy of a company's response by, inter alia, conducting sales traces of the company's documentation pertaining to selected sales, as was done in these cases. In these cases, the United States conducted three comprehensive verifications of POSCO's response in each investigation (one for home market sales, one for home market costs, and one for constructed export price sales in the United States). For local sales, POSCO stated that it was paid on a revolving account basis and thus there was no way to tie the payment directly to a particular sale.¹⁰² Thus, further verification was not practicable. Accordingly, there is no evidence to indicate that the United States departed from its normal practices in these cases. However, even if the United States had conducted no verification in these cases, it would not have violated any of the requirements of the AD Agreement.

Q.5. Were the Plate and Sheet investigations the first anti-dumping investigations where some of the home-market sales were dollar-denominated local sales? Does the United States agree or disagree with the statement on para 4.72 of Korea's First Submission that "neither the United States nor the petitioners in the [Plate and Sheet] investigations cited a single case before the investigations at issue where the United States treated a home-market sale priced in dollars as if it had been priced in the local currency"?

106. In this question, Korea attempts to create the false impression that the United States treated a sale actually priced in dollars, as if it had been priced in Korean won. As discussed above, these are simply not the facts. The United States' actions were fully consistent with a long and well-established practice to use the currency in which the revenue or expense was earned or incurred.

Q.6. Does the United States agree that the records of the Plate and Sheet cases show that POSCO's "internal exchange rate" was the rate published by the Korean Exchange Bank? If not, please describe the evidence to the contrary in the record.

107. See the United States' response to Panel Question # 7.

Q.7. Is it the US position that the fact that the currency conversion was made for accounting purposes in a home-market sale is determinative of, or relevant to, whether that sale was valued in dollars or won? If so, please explain the basis for that position.

108. No. The United States did not base its factual determination on whether a currency conversion was made for "accounting purposes." The United States weighed all of the evidence and concluded that these were won sales based upon the undisputed fact that the sales were invoiced and paid in won and were reported by POSCO as won sales.

83 See US Oral Statement, paras. 9 and 10.

84 See US Oral Statement, para. 9.

85 See US Oral Statement, para. 11.

86 See US Oral Statement, para. 9.

87 First Submission of the United States, paragraphs 82-84 and 87; Oral Statement of the United States at the First Panel Meeting, paragraph 12.

88 US First Submission, at para. 8; US First Oral Statement, at para. 9-12.

89 US Oral Statement, para. 15 (emphasis added).

90 US Oral Statement, para. 17 (emphasis added).

91 US Oral Statement, para. 17.

92 See US Oral Statement, para. 11.

93 See 19 C.F.R. � 410(b).

94 See, Third Party Oral Statement by Japan.

95 See generally, The GATT Uruguay Round: A Negotiating History (1986-1992): Antidumping, Stewart, T., Ed. (1993, Kluwer) at 155-61, (US Ex. 23).

96 See e.g., EC - Imposition of Anti-dumping Duties on Imports of Cotton Yarn from Brazil, ADP/137, (4 July 1995) paras. 500-501 (finding the practice of �zeroing� not to be inconsistent with the Anti-dumping Code).

97 The 10000 CU denominator is the aggregate value of the imports for which there were positive dumping duties.

98 Technically, the United States uses the four most recently completed fiscal quarters prior to the filing of the petition. See, 19 C.F.R. � 351.204(b).

99 See 6 April 1999 Sales Verification Report for POSCO in SSSS, at 14. (ROK Ex. 19). See also 5 January 1999 Sales Verification Report for POSCO in SSPC, at 4-5. (ROK Ex. 6).

100 See US Oral Statement, paras. 36 and 38.

101 See US Oral Statement, para. 39.

102 See, e.g., 5 January 1999 Sales Verification Report for POSCO in SSPC, at 14 (ROK Ex. 6), and Exhibit 39 (US Ex. 49).

To continue with ANNEX 2-5

Return to Table of Contents