1. Significant price suppression - Article 6.3(c) of the SCM Agreement Questions to both parties 51. The parties disagree on whether or not the marketing loan and counter-cyclical payments have more than minimal effects on production of upland cotton. Could each party explain how its approach to the analysis of the impact of these payments on production of upland cotton is supported by the provisions of Articles 5 and 6 of the SCM Agreement and by any other relevant WTO provisions?
66. At the outset, Brazil notes that, in response to a strictly legal question, the United States embarks on a lengthy summary of its various factual arguments regarding alleged "expectations" of U.S. farmers and the alleged absence of anything more than minimal effects from the billions of dollars of marketing loans and counter-cyclical subsidies provided to U.S. upland cotton producers. These U.S. arguments would certainly come as a surprise to the National Cotton Council, one of whose primary functions is to lobby the U.S. Congress for huge marketing loan and counter-cyclical subsidies in order to maintain the commercial viability of U.S. upland cotton production.61 These arguments would also surprise the typical U.S. cotton farmer, embodied by Mr. Stephen Houston, Sr. who was recently quoted as stating that "[market] prices don't have anything to do with what we're doing … [w]e're just looking at the government payments."62 And these arguments certainly would have surprised the original panel which, when confronted with the same U.S. arguments, found that there was "no doubt that marketing loan payments stimulate production and exports and result in lower world market prices."63 Finally, these arguments would surprise the 58 U.S. Senators that recently wrote to President Bush expressing concerns about reductions in U.S. agricultural subsidies. The Senators states that a deal in the Doha Development Agenda would "reduce net farm income through steep cuts in farm programs," indicating their belief that government subsidies play a fundamental role in allowing U.S. farmers, including U.S. upland cotton farmers, to generate sufficient profits from growing crops.64
67. Brazil has responded to these arguments in numerous previous submissions.65 Therefore, Brazil will not repeat the overwhelming evidence demonstrating that U.S. upland cotton farmers have received, do receive, and expect to receive huge marketing loan and counter-cyclical subsidies supporting the production of upland cotton. Brazil's arguments and evidence demonstrating that, but for marketing loan and counter-cyclical subsidies, U.S. upland cotton acreage, production, exports and stocks would be much lower and world market prices for upland cotton would be much higher are summarized in Brazil's answer to question 6966 and found throughout Brazil's many submissions.
68. Brazil explained that the most basic facts concerning the U.S. subsidies at issue and the conditions of competition in the world market for upland cotton suffice to establish a causal link between the subsidies and significant price suppression in the world market.67 These facts are as follows: It is undisputed that U.S. upland cotton is "like" Brazilian upland cotton as well as the upland cotton produced by many other countries.68 It is undisputed that there is a closely integrated world market for upland cotton and that heavily subsidized U.S. upland cotton competes directly in that world market with upland cotton from Brazilian and other third country producers.69 It is undisputed that the price of U.S. upland cotton, like that of other upland cotton producers, is reflected in a world market price, the A-Index.70 Finally, it is undisputed that the United States maintains a 40 percent world export market share and accounts for 20 percent of total world production resulting in a substantial proportionate influence on the world market price of upland cotton.71
69. These undisputed facts, combined with the fact that massive price-contingent U.S. marketing loan and counter-cyclical subsidies represent a 40 percent ad valorem subsidization rate over the lifetime of the FSRI Act of 2002, alone are sufficient for the compliance Panel to find that these subsidies cause significant price suppression in the world market for upland cotton. This is all the more true in light of the consistent inability of the average U.S. upland cotton producer to cover its costs of production without these two subsidies.
70. In these comments, Brazil focuses on the general theme of the U.S. response, i.e., "expectations" allegedly held by U.S. upland cotton farmers at the time of planting. Brazil agrees that an analysis of U.S. upland cotton producers' expectations, at the time of planting, about market prices and subsidy revenue is a useful supplement to the basic facts identified above. Indeed, Brazil conducted just such analyses in its First Written Submission72, Rebuttal Submission73 and Oral Statement.74 However Brazil considers that an assessment of the basic facts noted above is sufficient for the compliance Panel to find that the U.S. marketing loan and counter-cyclical subsidies cause significant price suppression in the world market for upland cotton, in violation of Articles 5(c) and 6.3(c) of the SCM Agreement.
71. The United States emphasizes that it is "important to look at the way that the marketing loan and counter-cyclical payment programs actually operate and interact with production decisions."75 Brazil agrees with this general proposition. U.S. upland cotton farmers, like any other businesses, function to generate revenue and make a profit. One of the questions before this compliance Panel is from what source typical U.S. upland cotton farmers expect to receive revenue sufficient to cover production costs and generate profits from their business. Do market returns alone provide sufficient revenue, or are U.S. marketing loan and counter-cyclical subsidies necessary to achieve the current levels of U.S. upland cotton supply? Both an assessment of actual costs and returns and expected costs and returns are relevant in considering these questions.
72. Expectations of U.S. upland cotton farmers, almost all of whom grow on upland cotton base acres benefiting from upland cotton counter-cyclical payments, are usefully summed up by Figure 8 of Brazil's First Written Submission, showing a consistent high stream of total revenue.76
Figure 1 – Support to Upland Cotton77
73. This figure demonstrates the critical impact of marketing loans and counter-cyclical subsidies. U.S. upland cotton farmers can expect high total revenue even if prices turn out to be low, for instance due to a collapse of Chinese demand or expanded Indian exports. Indeed total revenue will be high regardless of whether prices plunge for demand- or supply-related reasons. Every U.S. upland cotton farmer knows that, year-in and year-out, no matter what happens, he or she will receive a guaranteed revenue for the upland cotton he or she produces. During the lifetime of the FSRI Act of 2002, marketing loan and counter-cyclical subsidies accounted for an average of 40 percent of the market value of upland cotton.
74. The United States now claims that it is "not remarkable" that marketing loans and counter-cyclical payments have this revenue-stabilizing effect.78 Yet, labelling these payments as "income support" and suggesting that this is a normal and expected effect of price-contingent subsidies disregards the critical role of these subsidies for U.S. upland cotton production. Indeed, the U.S. arguments boil down to the following: the planting decisions of the typical U.S. upland cotton farmer have no, or at most minimal connection, to the two price-contingent subsidies that provide revenue worth an average of 40 percent of the market revenue of the crop produced. That argument is totally inconsistent with, to quote the United States, "the way that the marketing loan and counter-cyclical payment programs actually operate and interact with production decisions."79
75. Contrary to the U.S. arguments, which attempt to minimize the revenue-stabilizing effect of these subsidies80, Brazil's "expectation" arguments have always heavily relied on the fact that U.S. upland cotton farmers know, at the time of planting, that they will receive guaranteed and profit-making revenue by virtue of the price-contingent subsidies, no matter how market prices develop in the months ahead. The Appellate Body recognized this when finding that "although farmers had expected higher prices in making their planting decisions, they were also aware that if actual prices were ultimately lower, they would be 'insulated' by … marketing loan program payments [and]also counter-cyclical payments, which were based on a target upland cotton price of 72.4 cents per pound."81 76. Brazil, of course, does not assert that marketing loan and counter-cyclical subsidies are the only consideration for U.S. upland cotton farmers. Certain other factors, such as the current record high corn prices, will influence the planting decisions of upland cotton farmers. This is because even the generous U.S. marketing loan and counter-cyclical subsidies for upland cotton are no match for the returns that a farmer can expect from growing corn in MY 2007.82 Against this background, the remarkable fact is not, as the United States asserts83, that U.S. upland cotton acreage for MY 2007 is down. The remarkable fact is that despite vastly most attractive corn prices, the significant majority of U.S. upland cotton acreage stays in the production of upland cotton.
77. The original panel and Appellate Body found that marketing loan and counter-cyclical subsidies numb U.S. farmer's reactions to market forces – not totally deaden them.84 The considerable weight of the evidence shows that, over the life-time of the FSRI Act of 2002, for many farmers, these government subsidies made the difference between earning a long-term profit and having to leave the business of growing upland cotton.85 This is the background to the planting decisions made by thousands of U.S. upland cotton farmers like Mr. Houston. The result of these guaranteed and price-contingent subsidies is a U.S. upland cotton supply that is much larger than it would otherwise be. This is what USDA economists86 and the original panel found87, what the Appellate Body affirmed88, and what the evidence on record before this compliance Panel demonstrates.
78. In its response to the compliance Panel's narrow legal question, the United States also repeats a number of arguments, at paragraph 72, that Brazil previously responded to in detail. Brazil briefly recalls those responses and cross-references to its own argumentation refuting these points.
79. First, the United States claims that Brazil improperly focuses on actual prices and payments received by U.S. farmers at the time of harvest.89 Of course, actual prices in an isolated marketing year cannot be relevant to planting decisions taken before their realization. However, the collective knowledge of historic actual prices – and large subsidies resulting therefrom – constitute highly relevant evidence showing that, year-in and year-out, U.S. upland cotton farmers know that their revenue will be supported no matter what happens to prices. A typical U.S. farmer planting upland cotton in the spring of 2006, knew that when prices collapsed in MY 2004 against expectations, he received price-contingent subsidies accounting for 60 percent of market revenue.90 This experience is highly relevant to the farmer's planting decision in MY 2006. In any event, Brazil has also provided detailed evidence showing that the typical U.S. upland cotton farmer expected to receive marketing loan and counter-cyclical subsidies every year under the FSRI Act of 2002.91
80. Brazil also notes Professor Sumner's explanation that farmers' price expectations are fundamentally unobservable and that "it is impossible to know precisely what individual growers expect."92 However, given past experience with the price volatility of upland cotton and the price-contingent U.S. subsidies at issue in this dispute, it is appropriate to assume that farmers learn from their experience. Thus, an examination of historical payments is important. The United States asserts that "farmers cannot rewind time."93 But farmers certainly can – and no doubt do – learn from previous time periods. In trying to predict the future, farmers will rely on their past receipt of large marketing loan and counter-cyclical subsidies.
81. Moreover, Brazil demonstrated, relying on the Appellate Body's findings94 and much other evidence95, that farmers' expectations take account of their experience of past very high counter-cyclical and marketing loan subsidies, and the knowledge that, whatever happens to prices, their revenue will be protected by U.S. subsidies at profitable levels.96 The United States has not and cannot explain how such subsidies today suddenly no longer affect planting decisions and how they no longer have significant supply-enhancing and world market price-suppressing effect.
82. Second, the United States claims that Brazil has ignored net revenue and other conditions regarding competing crops vis-à-vis upland cotton.97 This is simply false. Among other evidence, Brazil demonstrated in figures 1-12 of its Oral Statement98 and its Comments on the U.S. Oral Statements99 that it was upland cotton marketing loan and counter-cyclical subsidies that allowed U.S. upland cotton farmers to expect higher profits from planting upland cotton than from switching to substitute crops, such as soybeans or corn.100
83. Third, the United States claims that Brazil has focused on total production, as opposed to plantings, and ignored other factors affecting production that are outside of a farmer's control.101 That is also false. Brazil has recognized that factors such as weather and yields do impact total production and total exports.102 Yet, Brazil has demonstrated that, in MY 2005, but for the two challenged subsidies, U.S. production and exports would actually have been 18 and 25 percent lower, respectively.103 The subsidy-induced extra acreage benefits from good weather or increased yields and significantly contributes to higher U.S. production and higher U.S. exports.104 Brazil also noted that the original panel conducted its analysis at a time when a significant percentage of U.S. upland cotton was already grown from biotech varieties105, and that also during the reference period assessed by the original panel yields had increased.106
84. Fourth, the United States claims that Brazil improperly focuses on total costs and not variable costs. In fact, Brazil's evidence addresses both types of costs. Brazil refers the compliance Panel to its comments on the U.S. response to question 59, below. Indeed, it is the United States that improperly downplays the crucial importance of covering total costs in light of the enormous long-term deficit of $12.4 billion in losses that U.S. upland cotton farmers would have accumulated over the past seven marketing years based only on market revenue.107 85. Fifth, the United States claims that Brazil fails to take into account demand from China in the world market for upland cotton.108 Brazil acknowledged that China plays an important role in the discovery of the A-Index world market prices, primarily from the demand side.109 But Brazil also emphasized that the United States – with a commanding 40 percent world market share of exports – plays a crucial role on the supply side of that price discovery process.110 Indeed, Brazil demonstrated that while Chinese demand skyrocketed in MY 2004 and MY 2005, prices actually fell due, in significant part, to record high U.S. production and exports.111 Thus, Brazil demonstrated that the impact of Chinese demand on the world market prices does not break the genuine and substantial relationship of cause and effect between the challenged subsidies and significant price suppression in the world market.
86. Finally, the United States' response to question 51 summarizes a number of arguments that it elaborates on in more detail in response to other questions. Brazil refers the compliance Panel to its comments on the U.S. response to question 53 with respect to additional planting decision and expectations related arguments112, questions 56/57 with respect to studies on the effects of counter-cyclical payments113, and question 59, with respect to the U.S. cost of production.114 Brazil also notes that the United States again repeats its criticism of Professor Sumner's simulation model that it advanced in its First Written Submissions115, but that the United States has not even attempted to rebut any of the arguments, explanations and rebuttals submitted by Brazil and Professor Sumner in their many submissions since.
52. In its Third Party Submission New Zealand observes: "Marketing loan payments are amber box measures, the category in which are included the non-prohibited measures with the most trade distorting effect on production and trade." (para. 5.19)
Do thepartiesconsider that thefactthat under theAgreementonAgricultureasubsidyisincludedinthe "amberbox" isrelevanttotheanalysisofthesubsidy'sconsistencywithArticles5and6oftheSCMAgreement? 87. Brazil refers the compliance Panel to its own response to this question.
Questions to the United States 53. The United States argues that Brazil has not provided evidence of "actual production inducing" effects of marketing loan and counter-cyclical payments and that Brazil "purports to demonstrate indirect production effects through its claim that the US planting, production, and exports are not responsive to prices". (Opening Statement of the United States at the meeting of the Panel with the parties, paras. 62 and 69, emphasis in original)
a) CouldtheUnitedStatesexplainfurtherthedistinctionbetweenwhatitterms "actualproductioninducingeffects "and "indirectproductioneffects"?CouldtheUnitedStatesalsoelaborateonhowthisdistinctionislegallyrelevantinthecontextofArticles5and6oftheSCMAgreement?
88. Brazil agrees with the United States that there is no legally relevant distinction between "actual" and "indirect" production effects recognized in Articles 5 and 6 of the SCMAgreement.116 That should have been the extent of the U.S. answer. Unfortunately, the United States went on to provide an extensive and unresponsive recap of its factual arguments. Brazil briefly responds to these unresponsive assertions and cross-references to Brazil's earlier submissions where these arguments have been addressed.
89. Brazil's comments on the U.S. response to question 51 summarize a number of Brazil's arguments concerning "expectations" of farmers and how the marketing loan and counter-cyclical subsidies "numb" those farmers response to market forces.117 In an attempt to counter this core finding by the original panel, the United States now argues that "it is unrealistic to suggest that there should be a one-to-one link between upland cotton planting, production, and export and upland cotton prices."118 Yet, Brazil's claims never required such a "one-to-one" link and the original panel never based its findings on such a precise link. Nor does Brazil claim that U.S. producers are incapable of any response to market price signals.119 Rather, Brazil's claim is that the updated evidence continues to demonstrate the correctness of the original panel's finding concerning the "numbness" of U.S. producers to market prices. That is, U.S. upland cotton producers do not react to market price signals and/or do not react to the extent that would be expected absent the price-contingent U.S. subsidies at issue in this dispute.
90. Brazil demonstrated that, as in the earlier MY 1999-2002 reference period, Brazilian and other non-subsidized low-cost producers continued during MY 2002-2006 to react to lower expected world market prices by lowering their acreage and decreasing production.120 In fact, in every year between MY 2002-2006, total foreign harvested acreage moved in the same direction as futures prices.121 By contrast, the planting decisions of subsidized, high-cost U.S. upland cotton producers, in general, did not correlate with changes in cotton market prices.122 For example, between MY 2002-2006, U.S. harvested acreage moved in opposite directions to those expected based on price movements in three out of five years – MY 2003, MY 2005, and MY 2006. The United States attempts to explain this by claiming that, in MY 2003 and MY 2006, this is because "U.S. harvested acreage reacts more conservatively than foreign acreage to increasing futures prices …."123 Yet, it is the lack of a correlation to prices that illustrates how U.S. producers are "numbed" from reacting to market forces.
91. More importantly, the degree and magnitude to which U.S. producers respond to market signals is also muted compared to their non-U.S. counterparts. In the past 5 years (MY 2002-2006), the change in U.S. harvested acreage from the previous year averaged 7 percent. Annual changes in Brazil, by contrast were 36 percent.124 These facts constitute evidence sustaining Brazil's point and that of Mr. Houston125 – that expected upland cotton market prices do not play an important role in the planting decisions of many U.S. upland cotton farmers.126 92. The United States repeats several times the unsupported assertion that "[s]ince the [FSRI Act of 2002] came into effect, U.S. export behavior has been entirely consistent with that of foreign counterparts."127 This is incorrect. A recent USDA analysis of the cotton market found that "[o]n the export side of the world trade equation, the United States has accounted for the lion's share of global gains, with 11 million bales of increased exports since 1999/2000."128 Brazil recalls that record U.S. export volumes were achieved in MY 2001, 2002, 2003, 2004 and 2005. The 40 percent U.S. share of world exports is more than three times higher than the share of Uzbekistan, the second largest exporter.129
93. The dramatic 55 percent increase in absolute U.S. export volumes between MY 2002 and 2005130 inevitably came at the expense of producers in non-subsidizing countries. Australia's average export market share of 12 percent between MY 1999-2001 was cut in half to 6 percent by MY 2005.131 Similarly, West and Central Africa's share of the world market declined to a ten-year low in MY 2005. The UN Food and Agricultural Organization recently summed up what has been occurring:
Subsidies maintain cotton production at otherwise unprofitable levels in industrialized countries, reducing the opportunities for developing countries to export to subsidizing country markets and displacing their exports to third countries.132
94. Of course, exports of upland cotton are the consequence of planting decisions and other factors, including subsidies, yields, domestic demand, etc. The relevant question is, therefore, what would U.S. exports be but for marketing loan and counter-cyclical subsidies. As discussed above, U.S. upland cotton planting are higher than they would be but for the U.S. subsidies at issue. Brazil has demonstrated that butfor the two U.S. subsidies at issue, U.S. exports would have been 25 percent lower in MY 2005.133
95. But even accepting, arguendo, the U.S. argument that U.S. and foreign exports did react in roughly the same manner as non-subsidized foreign exports, the United States ignores the reason why this may have occurred. Many of these non-subsidized foreign producers have far lower costs of production than U.S. producers.134 Brazil recalls ICAC's finding that it costs U.S. upland cotton producers $0.67 to grow a pound of upland cotton, while it costs producers in West and Central Africa $0.45 and producers in Brazil just $0.38.135 Such lower costs are essential to the survival of Brazilian and African producers, because they do not have the benefit of 40 percent ad valorem subsidies for the same basic commodity product as U.S. producers. These lower costs allow them to withstand certain levels of lower prices and still cover their total costs. U.S. producers are only able to compete with foreign producers in MY 2005 because marketing loan and counter-cyclical subsidies allowed them to sustain high levels of planted acreage. Brazil has demonstrated that but for these two subsidies, U.S. producers would have reacted to lower expected prices in MY 2005 (combined with higher costs) by reducing acreage, reducing production, and reducing exports.136 And with lower U.S. exports available to meet world demand, world market prices would be higher.
96. Finally, against this backdrop, the United States claims that the elimination of Step 2 payments caused a major shift in U.S. exports.137 While complete data for MY 2006 is, of course, not available, USDA projects that U.S. exports in MY 2006 will be at their second highest level in history. USDA further projects that the U.S. world market share of exports will be 36 percent in MY 2006. While this does constitute a decline in the U.S. world market share of exports, the United States still holds a commanding market share. As Brazil explained at the Panel meeting, the decline in the U.S. share of world exports in MY 2006 is largely due to the rise of Indian exports to China and the surge of U.S. exports just prior to the termination of the Step 2 program at the end of MY 2005.138 Further, with over 9 million bales of U.S. upland cotton sitting in U.S. warehouses, USDA and FAPRI project that the United States will increase its exports in MY 2007. In fact, FAPRI's 2007 baseline projects the U.S. share of the world export market to rise to 43 percent in MY 2007 and remain above 40 percent through MY 2011.139
b) WhatistheresponseoftheUnitedStatestotheargumentthatthefactthat "U.S. uplandcottonproducersknowthattheiroverallrevenuewillalwaysbeprotectedbymarketingloansandcounter-cyclicalpayments ... playsamajorroleintheirplantingdecisions"?(RebuttalofBrazil, para. 185; seealsoThirdPartySubmissionofNewZealand, paras5.20-5.21)
97. The United States asserts that the "mere" fact that income support programs support the income of agricultural producers "does not, however, mean that the payments have 'major' effects on planting or cause any of the adverse effects listed in Articles 5 and 6 of the SCMAgreement."140 Brazil agrees that different forms of income support may have different effects on production and that Articles 5 and 6 do not contain a presumption of adverse effects from income support. As the complaining Member, Brazil must demonstrate that the particular price-contingent subsidies at issue in this dispute cause serious prejudice to its interests. Contrary to the U.S. assertion, all of the evidence submitted by Brazil in this proceeding is sufficient to more than discharge this burden.
98. In commenting on the U.S. response, Brazil recalls its 2 April response to question 69, its comments on the U.S. response to questions 51 and 53(a), above, and 53(c), below, and the many arguments and evidence provided in its earlier submissions.
99. Brazil does not consider it "mere" or unimportant facts that, during the lifetime of the FSRI Act of 2002, marketing loan and counter-cyclical subsidies represented an average 40 percent ad valorem subsidization rate, covered an average 29 percent of total costs of production141, and were received by 96 percent of U.S. upland cotton producers holding upland cotton base acres.142 Nor is it "mere" fact that, but for these subsidies, U.S. production would be 15 to 16 percent lower143 and U.S. exports 21 to 22 percent lower.144 Nor is it "mere" fact that the original panel found that the subsidy payments are mandatory, i.e., they must be made by the U.S. Secretary of Agriculture without exceptions, and that they are contingent upon world and U.S. prices, i.e., they function to guarantee U.S. producers revenue regardless of how low prices fall. As both the Appellate Body and the original panel found145, the knowledge of U.S. upland cotton producers that these subsidies will be received at times of low prices is an important factor affecting their planting decisions and, consequently, affecting U.S. supply of upland cotton to the world market.
100. While it is possible that some forms and low levels of income support may not cause serious prejudice under particular conditions of competition, this is certainly not the case with respect to the two subsidies at issue: U.S. marketing loan and counter-cyclical subsidies for upland cotton. These particular mandatory, price-contingent subsidies cause serious prejudice in view of their nature and magnitude, and in view of the product they support and the particular conditions of competition in which that product is traded in international markets.146 c) InitsOpeningStatementatthemeetingofthePanelwiththeParties, Brazilobserved: "...we have demonstrated that these subsidies stabilized cotton producers' revenue despite wildly fluctuating market prices, thereby insulating and numbing acreage response to market price signals. These subsidies also cover the huge long-term gaps between market returns and total costs of production. Both effects are closely interrelated." (para.55)
101. The passage cited by the compliance Panel ties together two important themes in Brazil's causation arguments and two important effects of the U.S. marketing loan and counter-cyclical subsidies. The planting decisions analysis presented in the paragraphs that follow this portion of Brazil's Oral Statement shows how marketing loan and counter-cyclical subsidies (i) numb farmers' planting decisions and (ii) cover their total costs of production, resulting in higher levels of planting and supply than would exist but for the subsidies.
102. In its response, the United States accepts that marketing loan and counter-cyclical subsidies stabilize upland cotton producers' revenue, i.e., constitute price-contingent subsidies, but submits that Brazil's demonstration to that effect is not "remarkable".147 Yet, the conciseness of this fact and the simplicity of its demonstration does not make it any less important. It is crucial that every U.S. producer growing upland cotton on upland cotton base knows that he or she will receive a profitable return regardless of market prices at the time of harvest. This knowledge is the key aspect of revenue "expectations" at the time of planting.148
103. While the United States presents a number of arguments concerning the alleged absence of any effects on production of marketing loan subsidies (responses to which are largely set out in Brazil's comments to U.S. response to question 51), the United States fails to demonstrate anyfundamental change in the conditions of competition in the U.S. or world upland cotton market between the reference year of MY 2002 and the reference year of MY 2005 (or MY 2006). The same basic facts continue to exist. Therefore, and because the U.S. arguments raised are without merit, there is no basis for the compliance Panel to deviate from the findings of the original panel, as upheld by the Appellate Body. The FSRI Act of 2002 mandates the payment of marketing loan subsidies for every pound of upland cotton grown in the United States, withoutlimitation, if the adjusted world price falls below $0.52 per pound.149 Farmers know that the more upland cotton they produce, the more they will receive in marketing loan subsidies. Perversely, the operation of this subsidy means that when U.S. supply surges and world prices fall as a result, more marketing loan subsidies are made, perpetuating the cycle of high levels of production.150 In view of this fact, there is little wonder that the original panel found that "[w]e have nodoubt that the payments stimulate production and exports and result in lower world market prices than would prevail in their absence."151 Moreover, the original panel found that "the text of the measure indicates that the payments are mandatory, where certain market conditions prevail."152 Brazil notes that marketing loan subsidies in MY 2005 were 41 percent larger than in MY 2002.153
104. The United States repeats its argument that the actual payment of marketing loan and counter-cyclical subsidies is irrelevant and that only expectations matter.154 As Brazil has repeatedly demonstrated155, actual payments of these subsidies provide U.S. producers with payments worth 40 percent of the market value of a highly price-sensitive commodity.156 The experience of U.S. upland cotton farmers, who received very large subsidies over the years due to actual prices below the trigger prices, is a crucial fact that forms part of the background against which U.S. farmers' build their expectations about future revenue from the market and the subsidy programs.157 This remains unchanged from the facts before and the findings of the original panel and the Appellate Body.158 The U.S. expectation analysis narrowly and incorrectly assumes that farmers expect to receive subsidies only on the basis of a firmly expected price and disregard the possibility (i) that actual prices might deviate from this projection and (ii) that U.S. price-contingent subsidies provide a revenue guarantee against such changes in volatile upland cotton prices.159 105. With respect to the U.S. arguments regarding the absence of effects from counter-cyclical subsidies160, Brazil refers the compliance Panel to its rebuttal of similar arguments made in response to questions 56/57 below.
106. Next, the United States raises several new arguments (or new variations on older arguments) and presents new evidence in its response to question 53(c). In particular, it argues that "in every single marketing year since the FSRI Act came into effect, it would have been economically rational for upland cotton farmers to plant upland cotton rather than allowing their land to sit idle because the farmers expected that, even without any payment from the government, market revenue for upland cotton was more than sufficient to cover variable costs."161
107. This assertion is fundamentally flawed for a number of reasons. First, Brazil has never suggested that all or even most upland cotton acreage would be left idle without marketing loan and counter-cyclical subsidies. The counter-factual before this compliance Panel concerns the question whether, in the absence of marketing loan and counter-cyclical subsidies, upland cotton acreage would decline. Whether that land would be left idle or would be switched to the production of alternative crops that continue to benefit from U.S. subsidies, such as corn, soybeans, wheat, and rice, is not relevant for this assessment.
108. Further, contrary to the U.S. assertion, it is not economically rational to plant upland cotton year after year unless producers are able to cover their total costs, not merely variable costs. The original panel recognized this fact when it found that "[w]e believe that the existence of this gap between upland cotton producers' total production costs and market revenue, on the one hand, and the effect of the subsidies, on the other hand, was to sustain a higher level of output than would have occurred in the absence of the United States subsidies at issue."162 Brazil refers the compliance Panel to its 2 April response to question 71 for a discussion of why total costs, in addition to variable costs, are highly relevant to planting decisions.163
109. Moreover, a significant portion of upland cotton farmers are not able to cover their variable production costs with market revenue alone, nor do many upland cotton farmers expect to be able to do so at the time of planting.164 The portion of upland cotton farmers that are not able to cover their variable costs of production depend on which costs are considered variable.165 As explained in Brazil's Oral Statement, whether or not a cost is variable depends on the time period being considered.166
110. However, in addition to the time period under consideration, the categorization of cost items depends on the nature of the counterfactual examination. In this dispute, the compliance Panel is tasked to examine the effect of permanently removing a major source of upland cotton producers' income.167 This question is fundamentally different from examining the effects of annual fluctuations in market price levels. The counterfactual involving a permanent removal of the subsidies calls for the inclusion of costs in a planting decisions analysis that might not always be included when assessing planting decisions based solely on limited annual price fluctuations. For instance, if a drop in prices is expected to be isolated (i.e., only lasting a single year), a farmer would be more willing to continue producing even if he or she is unable to cover fully the opportunity cost of labor and land and special machineries. However, if the farmer knows that, due to the termination of a government program, a source of revenue representing an average 40 percent ad valorem subsidization rate will permanently be lost, no rational farmer would continue planting upland cotton unless he or she can cover operating, land and labor costs. Instead, the farmer would shift production to other crops or potentially lease his or her land to other farm operators. In other words, a producer's supply response would be greater if the reduction in subsidies was fully anticipated and permanent.168
111. Next, the United States claims that "in each of the years under FSRI Act, shifts in U.S. planted acreage were consistent with what one would expect if there were no marketing loan payments or counter-cyclical payments and U.S. farmers were basing their planting decisions solely on market prices signals and other considerations relevant to planting and production."169 This assertion is based on a flawed U.S. analysis of upland cotton planting decisions. The new U.S. analysis appears to have been engineered to make upland cotton appear more attractive relative to corn and soybeans than it actually was.170
112. The first major flaw in the United States analysis is the manner in which the United States calculates expected market prices (i.e., the farm price). It derives that price by applying a "basis" (or discount) to the January-March average closing price of the harvest period futures contract. For example, the United States discounts the futures price for soybeans and corn by $0.14 and futures prices for upland cotton by $0.05 a pound.171 The United States cites a USDA website in support of its "basis" for corn, soybeans and upland cotton.172 However, the website applies only to soybeans, corn and wheat, but not upland cotton.173 Thus, the United States provides no explanation or any data to justify why its basis of $0.05 per pound is appropriate.
113. While trumpeting FAPRI's analysis in its repeated critique of Professor Sumner's model, the United States rejects Brazil's use of projected prices published in FAPRI's baseline projections. However, the United States does not explain why FAPRI's projected prices should not be used. Nor does the United States explain why the compliance Panel should find that prices calculated by the United States for purposes of this litigation – and not supported by the source cited by the United States – should be adopted by the compliance Panel as superior to FAPRI's published price projections. (Brazil notes that USDA is prohibited by U.S. law from publishing its own price projections for upland cotton.)
114. As noted, Brazil's alternative crop expectations exercise relies on prices projected by FAPRI and published in its baseline during the planting period.174 The differences between these FAPRI-projected prices and the prices constructed by the United States are very significant for most marketing years, and especially for those crucial to the compliance Panel's assessment. The table below shows the expected prices for upland cotton under the two approaches: