|
|
|
espa�ol - fran�ais - portugu�s |
Search
|
|
|
|
espa�ol - fran�ais - portugu�s |
|
Search
|
UNITED
STATES � ANTI-DUMPING MEASURES ON
Report of the Panel (Continued) 43 Let me start with Article 2.4.2. As we have discussed previously, Article 2.4.2 of the Anti-Dumping Agreement requires that dumping margins be calculated "on the basis of a comparison of a weighted average normal value with a weighted average of prices of all comparable export transactions or by a comparison of normal value and export prices on a transaction-by-transaction basis." This language clearly indicates that only one average normal value and one average export price is contemplated for each comparison.
44. In its questions during the first meeting, the Panel asked us to explain how this interpretation of Article 2.4.2 can be reconciled with the common practice of calculating separate dumping margins for individual products or different levels of trade. While we have set forth our answer at length in our written submissions, let me try to summarize our position here: The key is in the phrase "comparable transactions." Article 2.4.2 does not require one average of every single home-market sale and another average of every single export sale. Rather, it requires a single average for all comparable transactions. Thus, to properly understand Article 2.4.2, it is necessary to analyze the factors that affect whether transactions are comparable.
45. As we have noted before, the word "comparable" means "capable of being compared." In the context of Article 2.4.2, this means that transactions that can be compared under the provisions of the Anti-Dumping Agreement are "comparable" transactions. Transactions that cannot be compared under the provisions of the Anti-Dumping Agreement are not "comparable" transactions.
46. There are, of course, a number of substantive limitations on the transactions that may be compared under the Anti-Dumping Agreement - which we have listed in our submissions. Significantly, there is no provision in the Anti-Dumping Agreement that says that sales made when the exchange rate is at one level cannot be compared to sales when the exchange rate was at another level. The only provision of the Agreement that addresses exchange rates is Article 2.4.1. And, the United States has specifically recognized that Article 2.4.1 does not "establish[] a limit on which transactions may be considered 'comparable' within the meaning of ... Article 2.4.2."25
47. Korea's position is also consistent with the past Panel decision in the Cotton Yarn case. In that case, the Panel specifically concluded that "[t]he exchange rate in itself is not a difference affecting price comparability." That Panel reasoned that an exchange rate "is a mere instrument for translating into a common currency prices that have previously been rendered comparable" under other rules.26 The Panel explained that determinations of price comparability occur at a different, earlier stage of a dumping analysis than do the actual comparisons of export price to normal value.27 The Cotton Yarn Panel stated that the exchange rate's function arose "subsequently" to determinations of price comparability.28
48. There is another problem with the multiple-averaging methodology adopted by the DOC: That methodology is inconsistent with the framework established by Article 2.4.1 of the Agreement for dealing with currency movements.
49. As mentioned, Article 2.4.1 is the only provision of the Anti-Dumping Agreement that addresses exchange rates or the permissible modifications to the dumping calculation methodology to account for exchange rate movements. Significantly, Article 2.4.1 sets forth special rules that apply to situations in which the exporting country's currency has been appreciating. However, Article 2.4.1 does not permit any adjustment to the dumping calculations to account for a depreciation of the exporting country's currency. Consequently, because the multiple-averaging methodology was adopted to account for the depreciation of the Korean won, it was inconsistent with the framework established by Article 2.4.1.
50. Finally, as we have noted previously, the multiple-averaging methodology was particularly unfair in the unique circumstances of the SSPC and SSSS cases. The US industry's claims for relief in these cases were predicated on the claim that anti-dumping orders were needed to protect the US industry from an increase in imports after the devaluation of the Korean won.29 Because these cases were predicated on the effects of the devaluation, a fair analysis of whether POSCO was truly engaged in dumping necessarily should have focused on pricing data after the devaluation. Yet, it was precisely that data that the multiple-averaging methodology effectively "walled off" from the DOC's price comparisons.
51. The result was a dumping determination based solely on pre-devaluation data in a case that was predicated on post-devaluation imports. Such a result is plainly unfair, and it therefore violates the "fair comparison" requirement of Article 2.4.
52. The United States agrees that Article 2.4.2 requires the calculation of a single average normal value and a single average export price for each "group of comparable transactions."30 It also concedes that changes in exchange rates by themselves do not "establish[] a limit on which transactions may be considered 'comparable' within the meaning of ... Article 2.4.2."31
53. Instead, the United States offers four arguments to justify the adoption of the multiple-averaging methodology. First, it contends that the word "comparable" in Article 2.4.2 is so ambiguous that the investigating authorities are free to interpret it in essentially any manner they see fit. Second, it contends that the multiple-averaging methodology was properly adopted to ensure that the comparisons were made "in respect of sales at as nearly as possible the same time." Third, it argues that Korea's arguments about multiple averaging are an impermissible attack on the practice of "zeroing." And, fourth, it contends that the multiple averaging methodology was needed to prevent the exchange rate shifts from "disguising" dumping. Let me address each of these arguments in turn.
54. I will turn first to the US arguments regarding the meaning of "comparable transactions" in Article 2.4.2. The US argues that its interpretation of the word "comparable" is entitled to deference.32 As we have noted previously, this claim for deference is inconsistent with the provisions of Article 17.6(ii) of the Anti-Dumping Agreement.33
55. In any event, the US claim for deference is particularly strained in this context: There is no US interpretation of the word "comparable" to which the Panel could defer. Korea has submitted that the word "comparable" means, in essence, "capable of being compared."34 The United States has neither disputed that interpretation nor offered any interpretation of its own. The meaning of the word "comparable" is simply not in dispute.
56. Let me turn now to the US arguments regarding the timing requirement of Article 2.4. In its first oral statement, the United States for the first time suggested that the multiple-averaging methodology was adopted to implement the requirement of Article 2.4 that sales be compared at "as nearly as possible the same time." In its second submission, this has become the principal US defense for its methodology. This post hoc defense is, however, without merit.
57. The multiple-averaging methodology in these cases did not purport to limit the comparisons to sales made at the same time. Indeed, that methodology did not take the amount of time between the sales into account at all. It did not say that the only sales that could be compared were the sales made within the same month or week or day. Instead, it said that all sales prior to October 31 could be included in one comparison and that all sales after November 1 could be included in another comparison. In the Plate case, for example, this meant that sales as much as ten months apart (from January 1 to October 31) might be included in the same comparison, but that sales one day apart (from October 31 to November 1) could not. Such a methodology might limit the comparisons to sales made under similar exchange rates. However, it simply does not limit the comparisons to sales at the same time. Consequently, the timing requirement of Article 2.4 cannot justify the multiple-averaging methodology.
58. Moreover, the US arguments regarding the timing requirement of Article 2.4 ignore the nature of a comparison that is based on an average normal value and export price. An average-to-average methodology does not compare the individual transactions that are included in the calculation of the average. Rather, the comparison is made only after the average has been calculated. When a series of home-market sales is averaged, the result is an average home-market sale made, on average, at the mid-point of the period. When a series of export sales is averaged, the result is an average export sale made, on average, at the mid-point of the period. As long as the sales in both markets are spread in a similar manner throughout the period, the averaging process will result in an average home-market sale and an average export sale made, on average, at the same time. Thus, the averaging process necessarily takes care of the timing requirement of Article 2.4 (unless the sales in the two markets are weighted disproportionately in different parts of the period).
59. The United States has also suggested that Korea's arguments about multiple averaging are a surreptitious attack on the practice of "zeroing" - which, the United States, contends, is explicitly permitted by the Anti-Dumping Agreement.35 This argument is, however, off target.
60. It is true, of course, that a finding that the Anti-Dumping Agreement does not permit "zeroing" as a general matter would require that the DOC's multiple-averaging methodology be overturned - because "zeroing" was an essential part of that methodology. Korea has, therefore, challenged the practice of zeroing as applied through the multiple-averaging methodology in this case. Moreover, as described in Korea's previous submissions, it was the zeroing aspect of the multiple-averaging methodology that led to the inflation of the dumping margins and the unfair bias in the DOC calculations.
61. Of course, the issue of "zeroing" is not critical to Korea's position - because the provisions of Article 2.4.2 requiring a comparison of single averages are independent of any zeroing issues. Nevertheless, we note for the record that Korea does not believe that zeroing is permitted as a general matter in dumping calculations under the Anti-Dumping Agreement.
62. Finally, let me turn to the real basis for the US decision - which was the DOC's sense that the depreciation of the Korean won would "disguise" dumping margins unless a multiple-averaging methodology were adopted.
63. As an initial matter, I would note that the negotiating history underlying the averaging provisions suggests that the Panel should be wary about accepting notions of "disguised" margins to overrule the actual language of the provisions. As noted in our second submission, the provisions of Article 2.4.2 represent a carefully negotiated bargain among the WTO Members on issues that were highly contentious during the Uruguay Round negotiations.36 By arguing that the single-average methodology which is required by those provisions should be disregarded because it "disguises" dumping margins, the United States is essentially asking the Panel to undo the careful bargain reflected in Article 2.4.2. The Panel plainly should not accept that invitation.
64. Moreover, the idea of "disguised" dumping margins in the Plate and Sheet cases is highly biased - because it overemphasizes the higher dumping margins that result from an over-valued currency, and ignores the lower dumping margins that result from an under-valued currency. Throughout all of the exchange rate shifts in these cases, the prices for POSCO's Korean and US sales changed little.37 However, the results of a comparison of those prices using the shifting exchange rates changed dramatically. Prior to November 1997, when the won in hindsight was over-valued, the Korean prices seemed higher than the US prices. By contrast, in November and December 1997, when the won was apparently under-valued, the Korean prices seemed lower than the US prices.
65. The multiple-averaging methodology is biased, therefore, because it calculates one dumping margin for the period when the won was over-valued and a separate margin for the period when the won was under-valued - and then effectively ignores the latter through "zeroing." A fair comparison would not allow such a bias. Instead, a fair analysis would have given both periods equal weight, to avoid the distortions inherent in an analysis that examined only a period in which the exchange rate was biased in one direction.
66. I shall now turn to the third and final subject giving rise to the US errors in dispute: The double-conversion of local sales from dollars to won and back to dollars at a different exchange rate announced by a different bank on a different date.
67. POSCO made certain sales in Korea which are known as "local sales." The orders for these sales were fixed in dollars. In fact, the order sheets for these sales showed only a dollar amount and not a Korean won amount. The invoices for these sales also showed a dollar amount. This dollar amount was the same as the dollar amount shown on the order sheet. It is also true that the invoices showed a Korean-won equivalent of the dollar price as an accounting convenience (since the amounts had to be recorded in the Korean accounting records of both POSCO and the customer in Korean won). This Korean-won equivalent shown on the invoices was determined by converting the dollar price from the order sheet into won using the official Korean Exchange Bank rate on the date of the invoice.
68. In its final determinations, the DOC chose to ignore the dollar prices that were reported by POSCO. Instead, it chose to begin with POSCO's converted won data - which had been calculated by POSCO by applying the Korean Exchange Bank's exchange rate on the date of invoice to the dollar price. The DOC then converted those amounts back to dollars using a different exchange rate on a different date - specifically, the DOC's official exchange rate based on the New York Federal Reserve's exchange rate for the date of the US sale. This methodology was virtually guaranteed to distort the normal value of the local sales. In fact, Korea has provided an example where there was a distortion of more than 70%.38
69. As Korea has argued throughout this proceeding, Article 2.4.1 of the Anti-Dumping Agreement only permits currency conversions to be made when they are "required."39 And, although the United States objected to that argument at first, it now appears to concede the point.40 Thus, the only question for the Panel is whether the second conversion from won to dollars was "required." If this conversion was not "required," then the US treatment of POSCO's local sales must be found inconsistent with WTO disciplines.
70. This is another straightforward issue: POSCO reported US dollar prices. There is no dispute that the reported US dollar prices matched the amounts shown on the order sheets and the invoices to the customers. There is no dispute that the payment was based on those US dollar prices. Consequently, a currency conversion was not "required." The DOC's decision to base its calculations on converted-won amounts and then convert them back to US dollars using a different exchange rate violated Article 2.4.1.
71. As we have noted previously, the DOC's methodology was also unfair. It penalized POSCO by inflating the normal value and hence the dumping margins for comparisons involving local sales. And, it did so simply because the exchange rates on the date of the home-market sales (which had been used to convert the dollar prices for these sales into won) were different from the exchange rates on the date of the US sales (which the DOC subsequently used to convert the converted-won amounts back into dollars).
72 As I mentioned previously, the "fair comparison" requirement of the first sentence of Article 2.4 requires, at a minimum, that exporters not be punished for events beyond their control. The DOC's methodology plainly does not meet that standard, because it punished POSCO for a change in exchange rates that was obviously beyond POSCO's control. Again, the issue here is very straightforward.
73. The United States has offered a number of explanations for the DOC's decision to use the converted won amounts, and not the reported dollar prices, as the starting point for its calculations. However, as I noted at the outset, the issue is not whether the United States can now come up with creative post hoc explanations to justify the DOC's decision. In order to assess the validity of the DOC's methodology, it is necessary to examine the reasons the DOC gave for ignoring the reported dollar prices.
74. In the final determination in Plate, the DOC gave three reasons for using the converted won amounts: (1) POSCO recorded the sales in its accounting system in won, (2) the exchange rate used by POSCO to convert the dollar prices into the won amount recorded in its accounting system differed from the exchange rate used by the DOC, and (3) POSCO was paid in won.41 In fact, however, none of these factors explains why the double-conversion was "required." They fail, therefore, to provide any basis for the Panel to find the double-conversion consistent with Article 2.4.1.
75. Let us start first with the purported difference in exchange rates. To begin with, the differences were much smaller than the DOC suggested. As demonstrated in Korea's submissions, the supposed differences result principally from the fact (which the United States admits) that the DOC mistakenly based its comparison on the wrong exchange rate. Under a proper comparison, the supposed differences essentially vanish.42 Thus, the factual underpinnings for the DOC's justification have no merit.
76. But even if the DOC had been correct, what difference would it have made? The exchange rate POSCO used to record won amounts in its accounting system had no bearing on the accuracy of the reported US dollar prices - and it did not prevent the DOC from using the reported dollar prices and avoiding the currency conversions altogether. Moreover, there was no basis for concluding that POSCO had manipulated the won amounts recorded in its accounting system for any purpose. After all, POSCO did not just make up the exchange rates it used. Instead, it used official exchange rates set by the Korean Exchange Bank. In the end, the DOC's decision reduces to a kind of economic chauvinism, which penalized POSCO because it used official Korean rates that differed slightly from rates published by the New York Federal Reserve. That result is patently unfair, and it cannot possibly be reconciled with the requirements of the Anti-Dumping Agreement.
77. The second justification offered by the DOC was that the sales were recorded in POSCO's accounting system in won. That justification is, however, equally irrelevant. It is a basic accounting principle that all of POSCO's sales and costs worldwide, whatever currency they are denominated in, must be recorded in POSCO's books in the same currency.43 For example, POSCO's US sales were recorded in its accounting system in won. Yet, the DOC did not use those converted won amounts to determine the export price. Why, then, should it matter that the local sales were also recorded in POSCO's accounting records in won? Indeed, even the United States has now backed away from any reliance on this factor: It now says that "The United States did not base its factual determination on whether a currency conversion was made for 'accounting purposes.'"44
78. This brings us to the DOC's third and final justification - the fact that the sales were paid in Korean won. Again, this factor has no rational nexus to the DOC's decision to base its calculations on the won amounts shown on POSCO's invoices. Most fundamentally, the won amounts of the payments did not match the won amounts on the invoices. It is plainly absurd to use one won amount as the price for the sale, simply because the customer paid a different won amount - and especially where both won amounts simply reflected the fixed dollar price converted using official exchange rates.
79. Therefore, none of the three reasons given in the final determinations provide any rational basis to conclude that the reported won amounts for the local sales had to be used as the basis for the DOC's calculations and that the subsequent re-conversion of those won amounts into dollars was "required."
80. The United States has once more attempted to muddy the waters by invoking a whole series of alternative justifications for the DOC's determinations. I could simply dismiss them all as irrelevant post hoc rationalizations. However, I feel constrained at least briefly to explain why these justifications are wrong as well as irrelevant.
81. First, the United States claims that the DOC's decision to choose the converted won data over the actual dollar prices was a "factual determination" entitled to extreme deference.45 That assertion is plainly wrong, because there were no facts in dispute. Please allow me a moment to review the key, undisputed facts:
All of these facts were verified by the DOC. Based on those facts, the DOC's verification report in Sheet concluded that "local sales are dollar denominated."46 At no point in any of its determinations did the DOC point to any factual disputes between it and POSCO on this issue.
82. Second, the United States argues that the DOC could not have used the dollar prices of the local sales, because POSCO did not use the methodology for currency conversions specified in Article 2.4.1.47 This argument is, frankly, backwards. The dollar price was the price that was actually fixed by the sales agreement between POSCO and the customer. POSCO did not make any currency conversions to produce the dollar price. Rather, the only currency conversions made by POSCO converted the actual dollar prices into won. The DOC then used that converted won amount as the basis for its calculations and effectively adopted POSCO's exchange-rate conversions. The only way the DOC could have avoided using POSCO's conversions would have been to use the original unconverted dollar prices. Thus, this new US argument does not justify the decision to use converted won data instead of the original dollar prices. To the contrary, it shows why the DOC should have used the dollar prices.
83. Third, the United States now argues that "the customer is charged (invoiced) in won."48 This argument directly contradicts the final determinations themselves. The final determinations clearly and repeatedly stated that the local sales were invoiced in dollars. They noted that the shipping invoice for some of the local sales (i.e., those sold through POSTEEL in "Home-Market Channel 2") "also shows the won price," but the use of the word "also" reinforces that the won amount was secondary to the dollar price on the invoices.49 In any event, the mere appearance of converted won amounts on invoices does not justify the DOC's choice to use the won amounts instead of the dollar prices - since the dollar prices were on the invoices as well.
84. Fourth, the United States now alleges that POSCO failed to provide it enough information quickly enough to demonstrate that the local sales were denominated in dollars, particularly in the Plate case.50 Again, this is another of the post hoc rationalizations that is noticeably absent from the final determinations. The DOC itself did not determine that POSCO's factual submissions on this issue were insufficient. And it is clear from the undisputed facts that such a claim would have been baseless:
POSCO's factual submissions plainly put the DOC on notice that the prices for these sales were fixed in dollars and not in won. If the DOC had further factual questions, it could have requested additional information from POSCO during the investigation - and, indeed, it was required to do so by Articles 2.4, 6.1, and 6.9 of the Anti-Dumping Agreement. Having failed to seek information it claims to have needed, the United States cannot blame POSCO for the alleged deficiencies in the record.
25 US First Submission, para. 142.
26 See EC - Imposition of Anti-Dumping Duties on Imports of
Cotton Yarn from Brazil, Report of the Panel, ADP/137, adopted on 30 Oct.
1995, para. 494.
27 See Cotton Yarn , para. 501; accord United States -
Imposition of Anti-Dumping Duties on Imports of Fresh and Chilled Atlantic
Salmon from Norway, Report of the Panel, ADP/87, adopted on 27 Apr. 1994,
para. 480.
28 See Cotton Yarn , para. 494.
29 See Korea�s First Submission, paras. 3.48 and 4.61.
30 See US Second Submission, para. 42.
31 US First Submission, para. 142.
32 See US Second Submission, paras. 44-45.
33 See Korea�s First Oral Statement, para. 77; Korea�s
Second Submission, paras. 15-18.
34 See Korea�s Responses to Panel Questions, at 4
(Response B.1) (quoting Webster�s Third New International Dictionary).
35 See US Responses, para. 97.
36 See Korea�s Second Submission, paras. 119-21,
citing Terence P. Stewart, ed., II The GATT Uruguay Round: A Negotiating
History, 1986-1992, at 1537-43 (1993) (ROK Ex. 77).
37 See POSCO Rebuttal Brief (SSPC), at 20-21 (ROK Ex. 9);
POSCO Case Brief (SSSS), at 21-22 (ROK Ex. 20).
38 See Korea�s First Submission, para. 3.58.
39 See Korea�s First Submission, paras. 4.66 - 4.69;
Korea�s Second Submission, paras. 161-65.
40 See US Responses, paras. 63, 102.
41 See SSPC Final Determination (ROK Ex. 11) at 15456. In
Sheet, the final determination also mentioned two other factors, but they are
clearly irrelevant and the United States has not raised them as justifications
before this Panel. See SSSS Final Determination, at 30678 (ROK Ex. 24)
("local sales are the only sales made in the home market that are expressly
linked to a dollar value" and "the vast majority of the costs incurred for home
market and US sales are denominated and paid by POSCO in won").
42 See Korea�s First Submission, paras. 3.60, 4.77 -
4.88; Korea�s Second Submission, paras. 145-49.
43 See Korea�s First Submission, paras. 3.49 - 3.54.
44 See US Responses, para. 108.
45 See US First Submission, para. 175.
46 See SSSS Sales Verification Report, at 14 (ROK Ex.
19); Korea�s Responses to Panel Questions, at 21-23 (Responses D.2 - D.3).
47 See US First Submission, para. 178; US Second
Submission, para. 65.
48 See US Second Submission, para. 64.
49 See SSPC Final Determination, at 15456 (ROK Ex. 11)
(emphasis added). Although the DOC noted the won information on the invoices, it
did not rely on that in its reasoning. See id.; see also SSSS
Final Determination, at 30678 (ROK Ex. 24) (citing only the three factors
mentioned above as the basis of the decision in Plate).
50 See US Responses, paras. 35-59.
51 See US Responses, paras. 48, 58 (admitting that, in
both cases, POSCO told the DOC that the amounts due in won were "based on the US
dollar invoiced price").
52 See Korea�s Responses to Panel Questions, at 20-21,
22-23 (Responses D.1, D.3) (and ROK Exhibits cited therein).
53 See SSPC Sales Verification Report, at Ex. 6 (ROK Ex.
6), Ex. 23-24 (ROK Ex. 84); SSSS Sales Verification Report, at Ex. 17, 20 (ROK
Ex. 46, 86).
54 See US Responses, para. 55 (admitting that POSCO so
informed the DOC on November 23, 1998).
55 See SSSS Sales Verification Report, at 14 (ROK Ex.
19); Korea�s Responses to Panel Questions, at 21-23 (Responses D.2 - D.3).
|
|
