List of abbreviations 444 table of cases


Increase in world market share - Article 6.3(d) of the SCM Agreement



Download 0.89 Mb.
Page12/18
Date conversion14.06.2018
Size0.89 Mb.
1   ...   8   9   10   11   12   13   14   15   ...   18

2. Increase in world market share - Article 6.3(d) of the SCM Agreement
Questions to the United States
75. Could the United States explain further the textual basis of its argument that "Article 6.3(d) is not concerned with absolute market share and whether or not in any given year a member's market share would have been lower if subsidies were removed"? (Rebuttal Submission of the United States, para. 401)
Questions to Brazil
76. What is the view of Brazil on the argument of the United States that an inquiry under Article 6.3(d) of the SCM Agreement requires two distinct elements: first, a demonstration of an increase in the world market share of a Member as compared to the average share it had during the previous period of three years, and, second, a demonstration that this increase in world market share compared to the average share the Member had during the previous period of three years is part of a consistent trend over a period when subsidies have been granted? (Rebuttal Submission of the United States, para. 399)
113. The United States respectfully refers the Panel to its general comments above regarding Brazil's claims under Articles 5(c) and 6.3(d) of the SCM Agreement as well to the U.S. arguments on this issue in its prior submissions.612

114. In its response to this question Brazil asserts that the second element of Article 6.3(d) – "this increase follows a consistent trend over a period when subsidies have been granted" – can be met simply by showing an increase in U.S. world market share from MY 1998 to MY 2005. Brazil's approach – if adopted – effectively write out of Article 6.3(d) both the reference to "this" increase and to a "consistent" trend.

115. As the United States explained in its answer to Question 75, Article 6.3(d) of the SCM Agreement is concerned with increase in world market share.613 An increase is – by definition – "the action, process, or fact of making or becoming greater."614 Thus, what is at issue in Article 6.3(d) is movement ("becoming greater"), not something static. The key question, then, is what movement is relevant for purposes of that provision The text provides the answer. The movement that is relevant is that in getting from point A – "the average share [the allegedly subsidizing Member] had during the previous period of three years," – to point B, the level in the year that is the subject of the claim. If upward movement – i.e., "increase" or "growth" – is involved in getting from the former to the latter, and the upward movement is shown to be "the effect" of a challenged subsidy – the first element of Article 6.3(d) is satisfied.
116. An "increase" is also at issue in the second element of Article 6.3(d) – the requirement that "this increase" be shown to "follow[] a consistent trend over a period when subsidies have been granted." Again, the text makes clear which particular increase is at issue – it is "this increase"   i.e.
, the same increase that was relevant for purposes of the first element. Thus, under the second element, the complaining party must show that "this increase" – the "increase in the world market share it had during the previous period of three years" – "follows a consistent trend over a period when subsidies have been granted."

117. Brazil's argues that "the United States attempts to read the separate element of a three-year average into the element of a 'consistent trend,'" but ignores the fact that Article 6.3(d) specifically refers to "this" increase, relating back to the reference to the increase over the previous three-year average. Brazil offers no reason why the drafters would have referred back to "this increase" if they had intended that a completely different increase – one defined by the complaining party – was to be the relevant increase for purposes of the second element of Article 6.3(d).

118. Instead, Brazil confuses two entirely separate questions. One question – just discussed – is what the relevant comparison is in assessing whether there has been an "increase" (i.e., whether the proper points of comparison are the share of world production in a particular year versus the average share over the previous three-year period, as the text suggests, or rather two years selected by the complaining party, as Brazil urges). A second, and separate, question is whether an increase must be shown in every year in order to show that "this increase follows a consistent trend over a period when subsidies have been granted" (emphasis added). These are distinct issues. Contrary to Brazil's assertions, the United States has never suggested that for a "consistent trend" to be shown there must be an increase in every single year that subsidies have been granted. Nor is this a necessary implication of the U.S. interpretation of Article 6.3(d).

119. While there may not be a need to show an increase in every year, where – as here – a complaining party can only point to one615 isolated year in which there has been an increase over the previous three-year average (and even that slight increase is shown to be both part of the ordinary fluctuation of world market share and caused by factors other than the challenged measures) there is clearly no basis whatsoever for an assertion of a "consistent trend."

120. Indeed, it is Brazil's interpretation that is inconsistent with the "consistent trend" language in Article 6.3(d). Brazil has argued that a complaining party should be able to satisfy the second element of Article 6.3(d) by simply selecting two end-points in the period that subsidies have been granted and drawing an upward-slanting line between the two points. However, this would simply show an increase. It would not necessarily show a "trend," and it would most certainly not show a "consistent" trend. To the contrary, it would render the term "consistent" entirely meaningless because – under Brazil's theory – there would only ever be two end-points and one line connecting them. This line could never be "consistent" with anything else because there would simply never be anything else with which to be "consistent." Brazil would – in effect – simply write "consistent" out of Article 6.3(d) altogether. This is not an permissible interpretation of Article 6.3(d).

77. In this connection, could Brazil respond to the argument of the United States that Brazil has not shown that either of these elements are met with respect to the marketing loan and counter-cyclical payment programs"? (Rebuttal Submission of the United States, paras. 399-403)
121. The United States respectfully refers the Panel to its general comments above regarding Brazil's claims under Articles 5(c) and 6.3(d) of the SCM Agreement, to the U.S. comments regarding Brazil's answer to Question 76, as well to the U.S. arguments on this issue in its prior submissions.616
C. Claim of Brazil regarding threat of serious prejudice
Questions to both parties
78. Could both parties comment on the statements of Canada that "(a)t issue is whether these programmes....threaten to cause serious prejudice simply by virtue of their existence" and that "(c)ertain subsidy programs, by their very nature, give rise to a constant likelihood of support that results in a permanent threat of serious prejudice"? (Third Party Submission of Canada, paras. 9-10)
122. The United States respectfully refers the Panel to its general comments above regarding Brazil's "threat of" serious prejudice claims under Articles 5(c) and 6.3(c) of the SCM Agreement as well to the U.S. arguments on this issue in its prior submissions.617

79. Could the parties state their views on the analysis of the ordinary meaning of the term "threat" in paras. 15-28 of the Third Party Submission of Canada?

123. The United States respectfully refers the Panel to its general comments above regarding Brazil's "threat of" serious prejudice claims under Articles 5(c) and 6.3(c) of the SCM Agreement as well to the U.S. arguments on this issue in its prior submissions.618

Questions to the United States
80. How does the United States address the argument of Japan that in view of the different purposes of Parts III and V of the SCM Agreement the standard for determining threat of material injury in Article 15.7 of the SCM Agreement is an inappropriate standard for determining the existence of a threat of serious prejudice under Part III of the SCM Agreement? (Third Party Submission of Japan, paras. 8-12.)
81. How does the United States respond to the argument of Australia that "it is beside the point for the United States to argue that the programmes under consideration are due to expire in late 2007"? (Third Participant Oral Statement of Australia, para. 13)
82. Could the United States comment on the projections of marketing loan and counter-cyclical payments in Table 26 of Brazil's First Written Submission and on the projections of prices and subsidy payments in Table 27 of Brazil's First Written Submission? Could the United States explain how the data in these Tables support its argument that producers are likely to expect low or no marketing loan payments in MY 2007? (Rebuttal Submission of the United States, para. 418)
Questions to Brazil

83. How does Brazil address the argument of the United States that footnote 13 of the SCM Agreement "does not indicate that where a panel finds that a Member is causing present serious prejudice through the use of a subsidy, the panel automatically also finds that the Member is threatening to cause serious prejudice in the future through the use of the same subsidy"? (Rebuttal Submission of the United States, footnote 624)

84. Could Brazil confirm that its claim of threat of serious prejudice is submitted on a contingent basis i.e., that it does not request the Panel to make a finding on this claim if the Panel make a finding of present serious prejudice? How is the contingent character of this "threat of serious prejudice" claim reflected in Brazil's request for the establishment of a panel?
85. Could Brazil explain its request that the Panel "make factual findings with respect to its 'threat of serious prejudice' claim to allow the Appellate Body to complete the analysis, in case it were to disagree with the compliance panel's interpretation"? (First Written Submission of Brazil, para. 241) What are the precise "factual findings" which Brazil requests the Panel to make in this regard?
86. How does Brazil address the argument of the United States that the definition of "threat" of injurious effects in Article 15.7 of the SCM Agreement and Article 4.1(b) of the Agreement on Safeguards "in terms of their close proximity in time and their high probability of occurring" reflects the ordinary meaning of the word "threat" and that, as such, Article 15.7 of the SCM Agreement and Article 4.1(b) of the Agreement on Safeguards provide "useful contextual guidance" for the interpretation of "threat" of serious prejudice?
124. The United States respectfully refers the Panel to its general comments above regarding Brazil's "threat of" serious prejudice claims under Articles 5(c) and 6.3(c) of the SCM Agreement as well to the U.S. arguments on this issue in its prior submissions.619 In addition, the United States submits the following observations regarding Brazil's answer to this question.

125. Article 3.2 of the DSU provides that "the dispute settlement system of the WTO . . . serves to . . . clarify the existing provisions of those agreements in accordance with customary rules of interpretation of public international law." One of these "customary rules" is that treaty terms are to be interpreted "in accordance with the ordinary meaning to be given to the terms of the treaty in their context and in the light of its object and purpose."620 In its response to this question and Question 88, Brazil urges the Panel to ignore the U.S. application of this rule in interpreting the term "threat" in footnote 13. Brazil provides no basis for why Article 3.2 of the DSU should not be applied here.

126. As United States has explained, the ordinary meaning of the term "threat" includes both concepts of probability of occurrence and close proximity in time.621 This is reflected in Articles 15.7 of the SCM Agreement, Article 3.7 of the Agreement on Implementation of Article VI of the General Agreement on Tariffs and Trade ("AD Agreement"), and Article 4.1(b) of the Agreement on Safeguards in describing the meaning of the term "threat." By application of customary rules of treaty interpretation under Article 3.2 of the DSU, the ordinary meaning of "threat" of certain injurious effects as reflected in these other provisions provides important contextual guidance in interpreting "threat" of serious prejudice in Article 5(c)/footnote 13 of the SCM Agreement.
127. There is no merit to Brazil's arguments that the Panel should artificially strip out the concept of close proximity in time from the ordinary meaning of "threat" in footnote 13 and should disregard the important contextual confirmation provided in the definition of "threat" in Articles 15.7 of the SCM Agreement, Article 3.7 of the AD Agreement, and Article 4.1(b) of the Agreement on Safeguards.

128. Brazil alleges, for example, that "Article 4.1(b) and 2 of the Agreement on Safeguards as well as Article 15.7 of the SCM Agreement deal primarily with the threat posed by a potential surge of imports" but that "[b]y contrast, Articles 5(c) and 6.3as well as footnote 13 of the SCM Agreement deal with the likelihood of adverse effects – not subsidized imports – occurring sometime in future."622 Brazil suggests that this is one reason why the concept of close proximity in time should be artificially stripped out of the ordinary meaning of "threat" for purposes of footnote 13. This argument is flawed.

129. First, it is wrong in its characterization of Article 15.7 of the SCM Agreement. Neither that provision – nor, for that matter its counterpart in Article 3.7 of the AD Agreement – "deal primarily with the threat posed by a potential surge of imports," as Brazil suggests. To the contrary, Article 15.7 expressly states that it is concerned with the "threat of material injury," not a "surge in imports." A "surge in imports" is not even required for a showing of material injury. To the contrary, Article 15.7 identifies as only one of five factors that administering authorities "should" consider ("inter alia") in making a "threat of material injury" determination whether there has been "a significant rate of increase of subsidized imports into the domestic market indicating the likelihood of substantially increased importation." The chausette to Article 15.7 even makes clear that "no one of [the five] factors by itself can necessarily give decisive guidance." In other words, Article 15.7 of the SCM Agreement could hardly be more clear that it does not "deal primarily with the threat posed by a potential surge of imports." Brazil's arguments premised on that assertion are, thus, irrelevant.

130. Second, Brazil ignores the fact that "injury" is itself classified as an "adverse effect" under Article 5 of the SCM Agreement. Indeed, Article 5 of the SCM Agreement makes clear in footnote 11 that "injury to the domestic industry" therein is "used here in the same sense as it is used in Part V." Thus, it is illogical for Brazil to assert that the focus in Article 5 on adverse effects somehow places a "threat of serious prejudice" inquiry into a fundamentally different category than a "threat of material injury" inquiry; one that precludes the Panel from consulting the ordinary meaning of "threat" in the latter context as important contextual guidance in interpreting the same word in footnote 13.

131. An inquiry into whether subsidized imports into the domestic market of a Member threaten to cause material injury is no less an inquiry into effects than an inquiry as to whether subsidies threaten to cause serious prejudice to the interests of another Member. The fact that one looks to the effects of subsidized imports while the other looks to the effects of subsidies is not material; the United States has never suggested that footnote 13 and Article 15.7 are identical provisions. And there is certainly no requirement that two provisions be identical in order for one to provide contextual guidance as to another.
132. Brazil also alleges that the Panel should read the concept of close proximity in time or "imminence" out of the ordinary meaning of the term "threat" in footnote 13 because the multilateral remedy available under Article 5 is somehow a "slow" remedy while the remedy available under Part V of the SCM Agreement is a "quick" remedy. Once again, the factual predicate of Brazil's argument is flawed. A remedy in WTO dispute settlement is not necessarily any more "slow" or "fast" than the process of applying duties under the SCM Agreement, AD Agreement, or Safeguards Agreement. Indeed, the pace of a particular WTO, CVD, AD, or safeguard proceeding necessarily depends on the particular circumstances of each proceeding.

133. Investigations by different Members' authorities require differing periods of time. And different Members may have different appeal mechanisms that affect the amount of time to obtain a remedy. In short, any number of factors could affect whether or not the domestic remedy available under the SCM Agreement, AD Agreement, and Safeguards Agreements is a slow or fast remedy. The same is true for the multilateral remedy available through WTO dispute settlement (e.g., the amount of time required for a remedy could depend on the complexity of the dispute brought by the complaining party, whether any appeals are made, the reasonable period of time for implementation of any adverse findings and other such factors). Brazil's generic assertions about the "slowness" or "quickness" of the different remedies is, thus, speculative at best.

134. In any event, Brazil neglects one simple fact – the drafters chose to use the same term ("threat") in all four contexts. If the remedies available thereunder were so different that same word could not even have the same ordinary meaning in the four contexts, the drafters could very well have chosen to use different words. They could very easily have inserted, for example, Brazil's preferred "significant likelihood" standard in footnote 13 to differentiate it from "threat" in Articles 15.7 of the SCM Agreement, Article 3.7 of the AD Agreement, and Article 4.1(b) of the Agreement on Safeguards. But they did not do so. They chose to use "threat" in all four contexts and – for clarity – even expressly reflected the ordinary meaning of that term in certain of the provisions. Brazil has no basis unilaterally to change the text now.
87. Could Brazil comment on the argument of the United States that the standard of "significant likelihood" is without support in the text of the SCM Agreement or in the GATT/WTO dispute settlement reports cited by Brazil? (Rebuttal Submission of the United States, paras. 406, 410,413)
135. The United States respectfully refers the Panel to its general comments above regarding Brazil's "threat of" serious prejudice claims under Articles 5(c) and 6.3(c) of the SCM Agreement as well to the U.S. arguments on this issue in its prior submissions.623

88. Does Brazil agree or disagree with the proposition advanced by the United States that "[a] panel may consider the ordinary meaning of a term as reflected in a particular provision to interpret the same term in another provision (especially of the same agreement) without the need for an express cross-reference." (Rebuttal Submission of the United States, para. 411, footnote 635)

136. The United States respectfully refers the Panel to its general comments above regarding Brazil's "threat of" serious prejudice claims under Articles 5(c) and 6.3(c) of the SCM Agreement, to the U.S. comments regarding Brazil's answer to Question 86, as well to the U.S. arguments on this issue in its prior submissions.624
89. Brazil argues that marketing loan and counter-cyclical payments for upland cotton are designed "in such a manner that payments would be made in consistently large amounts". (First Written Submission of Brazil, para. 270)
(a) Could Brazil in this regard discuss the data in Table 27 of Brazil's First Written Submission that show an increase in the projected farm price and AWP over the period MY 2006 - 2010 and a decline in projected marketing loan payments?
(b) How does Brazil respond to the argument of the United States that "by MY 2008, the projection is of no marketing loan payment at all"? (Rebuttal Submission of the United States, para. 419).
137. The United States respectfully refers the Panel to its general comments above regarding Brazil's "threat of" serious prejudice claims under Articles 5(c) and 6.3(c) of the SCM Agreement as well to the U.S. arguments on this issue in its prior submissions.625

138. In addition, the United States notes that, in its response to this question, Brazil attempts to explain away the evidence – submitted by Brazil itself – undermining its assertions that marketing loan payments are "designed 'in such a manner that payments would be made in consistently large amounts.'" Table 27 submitted by Brazil in its own first written shows the AWP increasing over the period MY 2006-2010 and projected marketing loan payments declining over the same period. Brazil's response is, however, inexplicable.

139. Brazil delves into a discussion of "deterministic" versus "stochastic" projections of government outlays in its response to the Panel's question. However, according to Brazil's own notations, Table 27 presents, inter alia, the AWP and payments under the marketing loan and counter-cyclical payments projected by FAPRI. The total amount of government outlays shown in Table 27 were not projected by FAPRI; to the contrary, the government outlays were estimated by Brazil.626 In any event, as USDA points out in the very document Brazil cites for the proposition that "deterministic projections, by their nature, tend to underestimate outlays":

The Congressional Budget Office (CBO) and the Food and Agricultural Policy Research Institute (FAPRI) have used a stochastic approach to estimate outlays and cost legislative proposals for some time. CBO adopted a stochastic approach for crop sector analysis in September 1995 while working on the 1995/96 budget reconciliation/farm bill debate. FAPRI adopted a stochastic approach just prior to the 2002 farm bill debate.627

140. Indeed, the July 2006 FAPRI baseline from which the data in Table 27 is taken is simply an update of the "full" 2006 baseline published earlier. The earlier baseline clarifies that "[i]n recent years, we have reported average values from the stochastic analysis for government costs and farm income and traditional point estimates for all other variables."628 Thus, the simple answer that Brazil fails to provide is that – at the time FAPRI issued its July 2006 baseline update FAPRI projected that in MY 2007 the marketing loan payment would be less than one-half of a cent and thereafter would amount to nothing at all. The very fact that the projection was made – projections that were considered sufficiently reliable for Brazil to include in its written submission and use in its modeling exercise – confirm that marketing loan program was not set up pursuant to some master "design" to result in large payments.

141. What the marketing loan program under the FSRI Act of 2002 was designed to do, moreover, was expire after MY 2007. Thus, regardless of whether projections are "deterministic" or "stochastic," whether they are of $1 or $1 billion, the fact is none of them take into account the fact that, as currently scheduled, no marketing loan payments will be made under the FSRI Act of 2002 in MY 2008. That fact alone renders Brazil's claims moot to the extent that they extend to any hypothetical measures in that year.

D. export credit guarantees


1   ...   8   9   10   11   12   13   14   15   ...   18


The database is protected by copyright ©hestories.info 2017
send message

    Main page