Annex H-1 Executive Summary of the Opening Statement of Brazil
Annex H-2 Executive Summary of the Closing Statement of Brazil
Annex H-3 Executive Summary of the Opening Statement of the United States
Annex H-4 Executive Summary of the Closing Statement of the United States
STATEMENT OF BRAZIL AT THE SECOND SUBSTANTIVE
MEETING OF THE PANEL WITH THE PARTIES
The United States Has Invented Threshold Burdens for Serious Prejudice Challenges that Do Not Exist in the Text of Articles 5 and 6.3 of the SCM Agreement
1. The United States’ has raised a number of threshold burdens that an Article 6.3 complainant allegedly must meet to establish a claim which are not based in the text of the serious prejudice provisions of the SCM Agreement.
2. First, the United States incorrectly asserts that under Article 6.3, Brazil must show an ad valorem subsidy rate and the amount for each of the challenged US subsidies. The only textual basis the United States provides is Annex IV of the SCM Agreement, which has expired with the text of Article 6.1(a) that contained the now-expired presumption of serious prejudice from a 5 per cent ad valorem subsidization. The United States further relies on countervailing duty measure procedures and interpretations of Brazil and the EC. But these allocation methodologies are irrelevant to Article 6.3 claims because unlike the expired Article 6.1(a) or Part V of the SCM Agreement, the focus of Articles 5(c) and 6.3 is on an examination of the effects of subsidies that are provided either directly or indirectly to producers of a product, such as cotton. In any event, Brazil has demonstrated a collective subsidization rate averaging 95 per cent and subsidies in the amount of $12.9 billion.
3. Second, the United States makes the sweeping argument that there is a legal prohibition on bringing adverse effects claims against subsidies that it alleges cannot be “expensed” or allocated to future years. The United States argues that all subsidies to cotton are “recurring” and therefore, as a matter of law, “cannot be said to be causing serious prejudice” except in the year in which they are provided. In fact, there is no textual basis in Part III or Article 6.3 of the SCM Agreement (or Article XVI:3 of GATT 1994) for distinguishing between the adverse effects of “recurring or non-recurring” subsidies. Nor is there any basis for “expensing” subsidies in one year or another year, as is often done in a countervailing duty investigation. Because Article 5 requires Members to prevent effects, a breach of Article 5 does not necessarily arise when a subsidy is granted, but only when actionable adverse effects occur.
4. The US argument is also inconsistent with the object and purpose of the SCM Agreement which is to protect Members from any subsidy causing serious prejudice. Under the US interpretation, a Member can permanently avoid any liability under Article 6.3 simply by carefully constructing the form of the payment as a recurring annual subsidy. Brazil has also presented evidence of continuing effects from subsidies provided in MY 1999-2002.
Brazil Has Established the Existence of Price Suppression in the US, World, Brazilian, and Other Markets Where Brazilian Producers Export 5. The United States now asserts that even though there may be evidence that New York futures market and A-Index prices as well as prices in other countries are suppressed by a “generalized effect,” the “in the same market” language in Article 6.3(c) requires that US exports be present in the same geographical markets in which Brazilian cotton is present.
6. Brazil has established, consistent with the requirements of Article 6.3(c), that the effects of the US subsidies were to suppress prices in the “same market” in which Brazilian producers marketed their “like” cotton – i.e., in the world’s, Brazil and in third countries. First, Brazil demonstrated the impact of US overproduction on the US and the “world market” prices (A-Index and New York futures market prices). Brazil then demonstrated that prices in Brazil and the countries to which Brazilian exporters shipped their cotton between MY 1999-2002 were suppressed and heavily influenced by US subsidies. The effects of those subsidies are communicated world-wide via a global price discovery mechanism. The parties agree that US, Brazilian and other countries’ cotton are “like products”. Throughout the world, prices for this fungible, price-sensitive commodity are determined by reference to the New York futures market and A-Index prices. Thus, Brazil established that the effect of the US subsidies is significant price suppression in the United States and Brazil, and in countries to which Brazilian producers exported their cotton. Further, the evidence shows that US subsidized cotton was present and contributed to the suppression of prices in 37 of the countries in which Brazilian producers marketed their cotton.
Brazil Has Established the Causal Link between the US Subsidies and Significant Price Suppression, Increased US World Market Shares, and the Inequitable US Share of World Trade 7. Brazil has properly analyzed both US revenues and costs using USDA’s own data and conclusively demonstrated that the US industry producing cotton is heavily dependent on all US subsidies to cover total costs over the short and long run. This finding provides a key economic rationale for the large production-enhancing effects from the US cotton subsidies found by USDA, as well as by US and international economists.
8. Over the long term, even the United States agrees that producers must recover all of their costs and make a profit to stay in business. USDA’s own cost and planted acreage data shows that US producers’ long-term costs from MY 1997-2002 were $12.5 billion greater than their market revenue received. The United States argues that off-farm income should have been included in Brazil’s revenue calculations. This US approach is conceptually as well as legally wrong. The relevant question is whether the “US cotton industry” is profitable from market revenue, not whether this industry is kept alive by cross-financing from other (non-subsidy) sources, such as social security payments.
9. US cotton producers would have suffered a cumulative loss of $332.79 per acre of cotton during MY 1997-2002 if they did not receive contract payments. But USDA’s own data in the Environmental Working Group database demonstrates that almost all US producers of cotton did receive contract payments. And as a result, they made cumulative 6-year “profit” of $106 per acre by MY 2002, allowing them to plant significant acreage to cotton in MY 2002 and 2003.
10. Finally, the United States criticizes Brazil’s comparison of costs of production among various countries. While Brazil agrees that the ICAC “data must be used carefully”, the problems with ICAC’s data do not render them unusable. Comparing ICAC “Variable Cash Costs” – which the United States does not challenge – demonstrates that it is much cheaper to produce a kilogram of cotton in Brazil than the United States.
11. The United States has argued that “in no year from marketing year 1999-2001 would the marketing loan rate be expected to have much of an impact, if any, on producer planting decisions,” and that it did not affect producer decisions in MY 2002. This new US argument contradicts numerous USDA studies. In assessing the credibility of the new US argument that marketing loans provided no production incentives, the Panel should consider that USDA’s own economists Westcott and Price found considerable effects of the marketing loan programme on US cotton production. The results of the Westcott and Price study are neither unique nor unexpected. Numerous other economists have found similar results. Looking at futures prices also reveals that US producers are unresponsive to price changes at planting time.
12. Moreover, the basic US assumption is that if the futures price (minus five cents) is above the marketing loan rate of 52 cents per pound, then “economic logic” demands that the marketing loan can have no impact on planting decisions.
13. The United States’ analysis of the marketing loan programme is based on a completely irrelevant comparison between the “expected cash price” and the marketing loan rate. Cotton marketing loan payments are based on the difference between the loan rate (52 cents) and the adjusted world price (“AWP”) not the price received by US farmers. The AWP is typically far lower than the price received by U.S. farmers. The average spread between the December contract futures price during the period January-March and the adjusted world price of the following marketing year was 18.5 cents per pound. Thus, even using the US “futures price” methodology at planting time, for every year between MY 1999-2002, there was the expectation at planting time that significant marketing loan payments would be made.
14. Even if the expected AWP is above the loan rate, this does not mean that farmers expect a zero marketing loan payment. If the expected AWP lies above the loan rate, farmers would still expect, with a certain likelihood, that the actual AWP could be below the loan rate because the expectations about the AWP is a probability distribution. Thus, they would still expect a positive marketing loan payment.
15. Brazil emphasizes that the US futures price approach suffers from several significant shortcomings. Both farmers’ decisions about planting and marketing of cotton are more complex. Planting decisions take place between January-March and the marketing of cotton takes place during the whole marketing year. Thus, just the February quote of the December futures contract does not properly address the complexity of farmers decisions.
16. The evidence is that the US cotton industry is sceptical of relying too heavily on present futures market prices as an accurate guide to future prices. Farmers have seen such volatility in the past as well as today. Therefore, any cotton farmer planting in the MY 1996-2002 period who actually relied on futures prices would know that the futures market is far from constituting a perfect predictor of future prices. Thus, as Professor Sumner correctly stated, “it is impossible to know what precisely individual farmers expect;” price expectations are “fundamentally unobservable”.
17. The record supports a finding by the Panel that more than $4 billion in contract payments were provided to current producers of cotton in MY 1999-2002. Brazil has demonstrated that the publications listed in the US review literature were largely irrelevant because they are not cotton-specific, as they do not address the re-coupling of production due to the base acreage updating for direct and CCP payments, or the huge target price CCP payments provided to cotton producers. Brazil has also shown that the US subsidies do not meet the criteria of paragraph 6 of Annex 2 of the Agreement on Agriculture, and are therefore not “decoupled”.
18. The record supports a finding by the Panel that the “other effects” apart from increased rental costs include significant production effects tied to upland cotton. First, the large majority of current cotton producers receive much higher per-acre payments for cotton than for other programme crops. The Panel must ask why much higher per acre direct and CCP payments are made to cotton base acreage if these payments are totally de-connected from current production? If that were the intention, as the United States argues and USDA presumes, then all contract payments in the same state or county would provide the same per acre benefit. The reason for the higher payments, of course, is that cotton is a high-cost crop and that cotton farmers insisted they were not receiving enough payments during MY 1999-2001 “to make ends meet”.
19. Second, the United States has now admitted that two million additional cotton base acres were added to the total contract “base” acreage. This means that in MY 2002, an additional $227 million in payments were made to farmers producing cotton during MY 1998-2001. This is not “decoupling” payments from production, but re-coupling to reward farmers for increasing their recent production. And the prospect of future updates will keep many farmers planting cotton in order to protect and even increase future bases. Even USDA economists agree that this creates a link to current production.
20. Third, the 72.4 cent target price triggers CCP payments when cotton prices are lower – not corn, or soybeans prices – but cotton. Why is that? Because the NCC argued and Congress agreed that given the high costs of producing cotton in the United States, current and future cotton farmers will need high payments when prices decline. There would be no reason to set a “target price” to protect against low cotton prices if Congress expected that most farmers with upland cotton base acreage would start planting apple trees.
21. Fourth, Brazil presented the Panel with information from USDA’s own electronic payment data showing that during MY 2000-2002 at least 71.3 to 76.9 per cent of total so-called “de-coupled” cotton base acreage payments were paid to producers of cotton. The data further shows that, in MY 2002, these producers of cotton received 85 per cent of their contract payments from cotton base acreage.
The US Subsidies Increase Exports, in Violation of Article 6.3(d) of the SCM Agreement
22. Brazil has demonstrated that, in violation of Article 6.3(d) of the SCM Agreement, the effect of the US subsidies played a significant role in the increase of the US world market share in MY 2001-2003 over its previous three-year average, following a consistent trend since MY 1996.
23. The US argues that “[u]nder Brazil’s reading, a Member would be free to provide subsidies that increased the share of its own domestic consumption that its producers supplied without any disciplines under Article 6.3(d)”. But this argument ignores the fact that Article 6.3(a) disciplines subsidies that increase domestic production in the market of the subsidizing Member. Further, Article 6.3(b) addresses any export displacement or impedance effects of subsidies in third country markets.
24. Second, the United States now argues explicitly that Article 6.3 has superseded Article XVI:3, second sentence. Assuming arguendo that the United States is correct, the effect of the US interpretation of “world market share” as meaning “world market share of consumption” would be to eliminate any WTO disciplines on production-enhancing subsidies that increase a Member’s world market share of exports. As Brazil has pointed out, this would be contrary to the fact that the language and scope of both Article XVI:3, second sentence and the text of Article 6.3(d) are very closely related.
25. Finally, the entire concept of a “world market share of consumption” is flawed for the purposes of Article 6.3(d) as it results in double counting. The United States argues that “the US share of the world market for upland cotton should be defined as US consumption plus US exports over world consumption”. However, the ordinary meaning in a trade remedy context of “domestic consumption” is total domestic “shipments” (i.e., net use from production or stocks) plus imports minus exports. Total “world consumption” is the sum of each country’s domestic shipments plus imports minus exports. But the US methodology addresses as “consumption” both imports and exports and thus, double counts.
CCC Export Credit Guarantee Programmes
26. The United States considers that Article 10.2 exempts export credit guarantees from the disciplines included in Article 10.1. In Article 10.2, the negotiators reached a good faith agreement to work toward specific disciplines on export credits. That need for a good faith commitment to negotiate explains the difference between the Draft Final Act and the final version of Article 10.2. Given the “magnitude” of those programmes, the United States argues that no Member could possibly have intended for its agricultural export credit programmes to be subject to Article 10.1. But, among others, the EC and Canada, both massive users of export credits, have told the Panel that they consider export credits to be subject to Article 10.1 if they meet the definition of an “export subsidy”. The United States did not think it needed to account for the CCC programmes in its reduction commitments, since it did not consider them to be export subsidies.
27. The United States says that it has offered “uncontroverted evidence” that for 12 of 13 scheduled products, US exports under the CCC export credit guarantee programmes did not exceed the United States’ reduction commitment levels. The correct question, however, is whether total US exports of a scheduled product exceed the quantitative reduction commitments, which Brazil has demonstrated. It is for the United States, under Article 10.3, to prove that those excess quantities did not receive export subsidies.
28. Whomever bears the burden, Brazil has demonstrated that the CCC export credit guarantee programmes confer “benefits” per se, and also constitute export subsidies within the meaning of item (j). The United States argues that even if the CCC programmes constitute export subsidies, because “the quantities were within the applicable US export subsidy reduction commitments[,] they would conform fully to the provisions of Part V of the Agreement on Agriculture”. The United States is in error.
29. With respect to unscheduled products, Brazil has established both actual circumvention and the threat of circumvention. Brazil’s Exhibits 73 and 299 and Exhibit US-41list the billions of dollars of CCC guarantee support that have been provided for exports of unscheduled products during fiscal years 1992-2003, thereby circumventing the US commitment not to provide export subsidies. The mere availability of CCC guarantees for unscheduled products threatens circumvention, since Article 10.1 prohibits any export subsidy for such products.
30. With respect to scheduled products, Brazil has demonstrated actual circumvention for US rice exports benefiting from CCC guarantees that have exceeded the US quantitative export subsidy reduction commitment. In its 18 November submission, the United States argues that because CCC has not disbursed the minimum amounts (at least $5.5 billion in guarantees each year, plus an additional annual amount of at least $1 billion in direct credits or guarantees for exports to “emerging markets”) there is no threat of circumvention. The United States misunderstands the test set out by the Appellate Body in US – FSC. The lack of a legal mechanism that stems, or otherwise controls, the flow of CCC guarantees threatens circumvention. There is no limit on the amount of CCC guarantees and CCC’s is exempt from the standard requirement of new Congressional budget authority for new guarantees.
CLOSING STATEMENT OF BRAZIL AT THE
SECOND SUBSTANTIVE MEETING OF THE PANEL
WITH THE PARTIES
1. The United States enjoys No Peace Clause Protection
1. The record demonstrates that the United States enjoys no peace clause immunity for its upland cotton subsidies. Under any of three methodologies – a “budgetary outlay/expenditure”, an “aggregate measure of support”, or a “rate of support” methodology – the level of support provided in MY 1999-2002 exceeds the level of support decided in MY 1992. The United States has acknowledged that all challenged US subsidies, except PFC and direct payments, are non-green box (trade and production-distorting) subsidies. Brazil and all third parties agree that direct payments under the 2002 FSRI Act are non-green box because of the updating of a fixed base period contrary to Annex 2, paragraphs 6(a) and (b) of the Agreement on Agriculture. Further, PFC and direct payments are non-green box support because of the prohibition on fruits and vegetables contrary to Annex 2, paragraph 6(b). Moreover, payments for cotton base acreage are higher than those for other crop base acreage. The weight of evidence shows that all the challenged US subsidies are “support to” upland cotton within the meaning of Article 13(b)(ii) because they were received by current producers or by users and exporters of US upland cotton. Brazil’s “14/16ths” methodology for estimating the amount of the four contract payments is reasonable and supported by the EWG database and considerable circumstantial evidence.
2. Brazil has established the Elements to Support its Significant Price Suppression Claims under Article 6.3(c) 2. Upland cotton is a basic, widely-traded commodity. Both Brazil and the United States agree that Brazilian upland cotton and other upland cottons are “like” subsidized US upland cotton. Because of this widespread interchangeability among world cottons, increases in world cotton supply by major cotton-producing countries have a major impact on discovery or establishment of world prices reflected in the A-Index and the New York cotton futures exchange.
3. The United States has made a great deal about what it terms some “fundamental” issues about the nature and amount of subsidies. Brazil has demonstrated the absence of any textual basis for incorporating various countervailing duty principles from Part V of the SCM Agreement into Part III and resurrecting Annex IV from the dead. However, Brazil used USDA’s own data to show both the amount and rate of subsidization for each of the subsidies. To make up for US acreage and yield information the US has hidden from Brazil and the Panel for 16 months, Brazil has demonstrated through the EWG database and other circumstantial evidence that its “14/16th” methodology is reasonable. This methodology allocates payments only to current producers of upland cotton and does not “double count” payments provided to other producers of crops. And since the peace clause phase of this proceeding, Brazil has demonstrated that all the US subsidies are “tied” to the production of upland cotton and are “support to” upland cotton.
4. The US government has poured $12.9 billion over the past four years into a number of subsidy programmes specifically targeted at US upland cotton. No other US commodity has a Step 2 programme and no other US commodity received subsidies as high as 136 per cent ad valorem. Even the so-called decoupled contract payments for “historical” cotton base acreage are much higher than for any other crop except rice. The subsidies provide a specific “target price support” of 72.4 cents per pound for upland cotton – not for other crops. And Congress insisted that USDA has no discretion to limit any of the required payments, all of which are mandatory and place no limit on the amount of upland cotton that could be produced with the support of these subsidy programmes.
5. The United States has agreed that all of the challenged US subsidies are “specific” except crop insurance. But USDA’s own evidence showed that this programme is also specific since it is targeted at the industry growing crops, not livestock and thus covers only half of the value of US agricultural commodities and 38 per cent of farmland.
6. While Brazil continues to wait for farm-specific acreage and yield information from the United States, the incomplete Environmental Working Group data based on USDA farm-specific data show almost $3 billion in contract payments paid to upland cotton producers in MY 2000-2002 alone. And the great bulk of the other evidence shows that US upland cotton farmers are dependent upon, need and, in fact, receive such payments to “make ends meet” and “to survive”.
7. Having established the fungible nature of the product and the existence and specificity of the subsidies, Brazil must link the effects of the subsidies to significant price suppression. The first important fact is that the United States is by far the world’s largest exporter, with a world market share of 41.6 per cent, and the second largest producer of upland cotton in the world, with a 20 per cent share. The US subsidization rate of 95 per cent provided by the second largest producer and largest exporter creates the potential of causing serious prejudice to the interests of other Members, including Brazil. It is useful to recall the size of these subsidies compared to the 5 per cent ad valorem rate establishing a presumption of serious prejudice under Article 6.1(a) and compared to the amount, if any, of subsidies received by US competitors.
8. But what was the impact of the large US subsidies on production and world supplies of cotton? One answer to this question is found in the difference between market revenue and the US producers’ total costs. While claiming that only variable costs are important in the short term, the United States admits that in the long-term, US producers have to make a profit to stay in business. Using only USDA’s data for the period MY 1997-2002, the average US upland cotton producer received market revenue that was $872 dollars per acre less than its total costs. This means the cost/revenue gap for all upland cotton farmers between MY 1997-2002 was $12.5 billion.
9. The United States has attempted to leave you with the impression that its upland cotton producers do not rely or need any subsidies between MY 1997-2002 to make up this $12.5 billion gap. In assessing the credibility of these claims, consider that during this same 6-year period, US cotton producers received $16 billion in US subsidies and ended up with a 6-year “profit” of $127 per acre. The US claims that the PFC, market loss assistance, direct payment and CCP payments were not support to cotton. But without those 4 payments, US cotton producers would have lost $333 per acre between MY 1997-2002. The US further claims that the marketing loan payments in MY 1999-2002 made no difference to producers’ planting decisions. But this argument ignores the impact of the subsidies in those producers’ costs. By MY 2002, the average US producer would have been faced with a 3-year loss of $372 per acre if they had not received marketing loan payments during MY 1999-2001. This evidence confirms the conclusion of the Chief USDA economist that by making marketing loan payments “you don’t get cutbacks in production”. Clearly, the marketing loan programme kept many producers from reducing their planted acreage between MY 1999-2002.
10. Indeed, US producers planted between 14.2 – 15.5 million acres of upland cotton between MY 1999-2002 as prices fell to record lows. The combined revenue from all the US subsidies and market prices allowed producers to earn a long-term “profit” of $17.67 per year over the 6-year period. What is most amazing is that after having received record low prices for their cotton in MY 2001 and with futures prices at the time of planting suggesting market prices would remain at record low levels, US producers still planted 14.2 million acres of upland cotton – a similar amount of acreage that was planted when prices were much higher in MY 1996-1998. However, even 135 per cent ad valorem subsidies in MY 2001 were not sufficient to provide a profit to the highest-cost and lowest-yield US producers. This explains why US planted acreage declined to 1996-98 levels in MY 2002.
11. Having established that the US subsidies prevented production cutbacks, the Panel has to estimate how much of a cutback would have been made without US subsidies. USDA economists Westcott and Price estimate a 20 per cent cutback in MY 2001 from only the marketing loan programme. Professor Sumner estimates an average production cutback of 28.7 per cent or a total of 19.8 million bales between MY 1999-2002 from eliminating all subsidies.
12. The Panel then must estimate the effect of these estimated US production cutbacks on world prices. First, Brazil demonstrated the impact of US overproduction on the US and the “world market” prices (A-Index and New York futures market prices). Brazil then demonstrated that prices in Brazil and the countries to which Brazilian exporters shipped their cotton between MY 1999-2002 were also suppressed and heavily influenced by US subsidies. The effects of those subsidies are communicated world-wide via a global price discovery mechanism. Throughout the world, prices for this fungible, price-sensitive commodity are determined by reference to the New York futures market and A-Index prices. There is a world market for upland cotton. Subsidized US cotton and Brazilian cotton compete in this world market, i.e., “in the same market”, as used in Article 6.3(c) of the SCM Agreement. Brazil established that the effect of the US subsidies is significant price suppression in that world market.
13. There are numerous studies from a number of economists finding clear and identifiable amounts of price suppression ranging from 10-33 per cent for the US price and 10-26 per cent of the world A-Index price. Professor Sumner responded to the US critiques of these studies by showing that they are not biased and correcting for some shortcomings are consistent with his results. The United States also claims these studies are useless for this dispute because they did not use “futures prices”, but then admits that USDA and FAPRI models also use “lagged prices” because it is not possible to use futures prices in models to judge farmers’ revenue expectations. Brazil also demonstrated that, using the US futures methodology, farmers expected significant revenue from marketing loan programmes in MY 1999-2002.
14. Finally, the Panel should judge the “significance” of the price suppression by the extent of the impact on Brazilian producers. But even judged in relation to objective levels, any of the price suppression estimated in the various econometric studies is sufficient to establish “significance”.