Reestimated Guaranteed Loan Subsidy based on Original Budget Year Estimate.
Exhibit Bra- 192
Net Lifetime Reestimates of Guaranteed Loan Subsidy by Cohort.
Exhibit Bra- 193
US Department of Agriculture Office of Inspector General Great Plains Region Audit Report, Commodity Credit Corporation’s Financial Statements for Fiscal Year 2001, Audit Report N° 06401-4-KC, February 2002
Exhibit Bra- 194
US Department of Agriculture Office of Inspector General Financial and IT Operations Audit Report, Audit Report N° 06401-14-FM, Commodity Credit Corporation’s Financial Statements for Fiscal Year 2000, June 2001
Exhibit Bra- 195
US Department of Agriculture Office of Inspector General Financial and IT Operations Audit Report, US Department of Agriculture Consolidated Financial Statements for Fiscal Year 1999, Report N° 50401-35-FM, February 2000
Vincent Whittaker, “The Quick Buck, International Finance and Forfaiting,” 23 Thomas Jefferson Law Review (Spring 2001)
Exhibit Bra- 198
Trade and Forfaiting Review, “Argentina Trade Finance to the Rescue” Volume 6, Issue 9 July/August 2003
Exhibit Bra- 199
COMMENTS OF THE UNITED STATES ON NEW MATERIAL
IN BRAZIL’S REBUTTAL FILINGS AND ANSWER OF THE
UNITED STATES TO THE ADDITIONAL
QUESTION FROM THE PANEL
27 August 2003
TABLE OF CONTENTS
I. Introduction 43 II. Brazil’s AMS Calculation Is Flawed and, Had It Consistently Reflected a
Price Gap Calculation, Would Demonstrate No Peace Clause Breach 45 III. The US Level of Support Argument Does Take Into Account All
Product-Specific Support That Challenged US Measures Grant 47 IV. Brazil’s New Green Box Arguments Are in Error 48V. Answer to Additional Question 67bis from the Panel 50
VI. Payments With Respect to Base Period Production of Certain Commodities
But Not Others Are Not Inherently Product-Specific Support 53 VII. Brazil’s New Arguments Relating to Crop Insurance Do Not Demonstrate
that Crop Insurance Payments Are Product-Specific Support 55 VIII. Brazil May Not Act Unilaterally on Procedural Matters 57 IX. Conclusion 58
I. Introduction 1. The United States thanks the Panel for its prompt reply to the US request of 25 August 2003, granting an opportunity to comment on new material in Brazil’s rebuttal submission and Brazil’s comments on the US responses to questions. We also thank the Panel for its additional question. We present both the US comments on new material in Brazil’s submissions and our answer to that additional question below.
2. As the United States notes in these comments and answer, Brazil’s Peace Clause argument depends on three issues:
First, Brazil relies on budgetary outlays that reflect prevailing market prices that could not have been “decided” by the United States. Brazil ignores the fact that “support” does not mean “budgetary outlays”; in fact, Annex 3 recognizes that an “Aggregate Measurement of Support” for price-based support either shall576 or can577 be calculated using a price gap methodology, which does not rely on budgetary outlays.
Second, Brazil conflates “non-product-specific support” with “support to a specific commodity” by attempting to allocate certain payments to upland cotton. To do so, however, Brazil relies on a reading of the definition of “non-product-specific support” in Article 1(a) that ignores the most relevant context for this term – that is, the (immediately preceding) definition of product-specific support in that same article. Indeed, Brazil’s approach would appear to render the concept of “non-product specific support” so narrow that it becomes almost, if not completely, meaningless.
Third, Brazil mischaracterizes US direct payments and production flexibility contract payments as non-green box support. Brazil has not established that these measures fail to conform to the policy-specific criteria in Annex 2. In fact, Brazil has not even established – pursuant to Brazil’s own reading of the first sentence of Annex 2, paragraph 1, as a stand-alone obligation – that these payments have more than “minimal trade-distorting effects or effects on production.”578
3. The weakness of Brazil’s interpretation that “support” in the Peace Clause means “budgetary outlays” can be seen in this example. Even if US measures were exactly the same in every year of the implementation period as they were in the 1992 marketing year (that is, the same deficiency target price, same marketing loan rate, same acreage reduction percentage, same normal flex acres with planting flexibility579, etc.), under Brazil’s interpretation, US measures would have breached the Peace Clause in each and every year in which outlays increased due to external factors, for example, whenever market prices dipped below the 1992 level.
Would 1999-2002 US measures identical in every respect to those in 1992 “grant support in excess of that decided during the 1992 marketing year”?
In other words, if a Member had decided its support during 1992 for the period through 2004 and never changed its decision, could the Member be deemed to grant support in excess of the level decided during 1992 just because outlays increased, for example, because market prices changed?580
We believe the answer must be “No” because market prices are not “decided” by a Member (as paragraphs 8 and 10 of Annex 3 recognize). And yet, the situation in this dispute is analogous: the United States has changed its measures to reduce the product-specific level of support (by eliminating deficiency payments) since 1992, and yet Brazil claims that the Peace Clause has been breached simply because lower market prices resulted in increased price-based outlays.
4. Market prices are beyond the control of the United States, and therefore the United States cannot “decide” them. Removing the effect of market prices beyond the control of the United States from the measure of support demonstrates that US measures do not and did not grant support in excess of that decided during the 1992 marketing year. In fact, whether gauged (as the United States believes is compelled by the Peace Clause) via the rate of support expressed by US measures581, or via the AMS for upland cotton (calculated through a price gap methodology), or via the erroneous calculations of Brazil’s expert (but limited to product-specific support), the result is exactly the same: in no marketing year from 1999 through 2002 have US measures breached the Peace Clause.
II. Brazil’s AMS Calculation Is Flawed and, Had It Consistently Reflected a Price Gap Calculation, Would Demonstrate No Peace Clause Breach 5. We recall that Brazil has argued that budgetary outlays are the only measurement of “support” for purposes of the Peace Clause proviso comparison, without any foundation in the Peace Clause text and despite the context provided by Annex 3, which explicitly indicates that Members have agreed “support” can be measured without using budgetary outlays. Brazil itself concedes that US measures do not decide support on the basis of budgetary outlays:
Brazil acknowledges that the United States could not possibl[y] determine its expenditures as they would depend to a certain extent on market prices that were also influenced by factors outside the control of the US Government.582
The United States agrees with this statement by Brazil and believes that this statement demonstrates that Brazil’s approach to the Peace Clause comparison is not based on the text nor is it realistic. Instead, in order to hope to succeed, Brazil’s claims require Brazil to use budgetary outlays and so to take into account low prevailing market prices. An Aggregate Measurement of Support calculation using a price gap methodology – that is, that eliminates the effect of market prices and reflects instead the eligible production and applied administered price decided by a Member – reveals that in no year from 1999-2002 have US measures breached the Peace Clause.
6. Brazil has argued that “there are only two types of methodologies that would allow an expression in monetary terms of a decision (or decisions) taken by the United States in MY 1992 regarding its level of support to upland cotton. The first is budgetary expenditures. The second is the calculation of AMS for a particular commodity.”583 In both its table of expenditures584 and its AMS table585, Brazil has attempted to allocate non-product-specific support to a specific commodity. There is no basis in Annex 3 to do so. Annex 3, paragraph 1, explicitly requires an AMS to be calculated “on a product-specific basis for each basic agricultural product” and separately requires that non-product-specific support be calculated and “totalled into one non-product-specific AMS in total monetary terms.” The point bears emphasis: “for each basic agricultural product,” Annex 3 states that an AMS “shall be calculated on a product-specific basis.” Similarly, were “support to a specific commodity” (upland cotton) to be calculated using an Aggregate Measurement of Support, it must be calculated “on a product-specific basis.”
7. As a result, both Brazil’s expenditure table and its AMS table run counter to the terms of Annex 3. Were the Panel to calculate an AMS for upland cotton for marketing years 1992 and 1999-2002, the United States has set forth a calculation consistent with Annex 3 in its rebuttal submission.586 By using a price-gap methodology for both deficiency payments and marketing loan payments587, the upland cotton AMS in 1992 is far higher than in any marketing year from 1999 to 2002, reflecting the US decision to move away from the high support levels of product-specific deficiency payments.
8. In fact, we note that the AMS data presented in paragraph 115 of the US rebuttal submission understates the AMS for marketing year 1992. For example, the United States reduced the price gap calculation for 1992 basic deficiency payments by an adjustment factor (approximately .875) to replicate the calculation used in G/AG/AGST/USA, p. 18. Without the adjustment, which is not called for by paragraphs 10 and 11 of Annex 3, the 1992 deficiency payment calculation would have been $858 million, rather than $755 million as reported in paragraph 115.588 (In case of interest, we also present below the deficiency payment calculation in more detail, reflecting more accurate data, which would increase the deficiency payment calculation slightly, to $867 million.)589 This calculation, moreover, uses the actual payment acreage (that is, acres planted for harvest or participating in the 50/92 programme on which payment was received) to calculate the “eligible production.” Using instead the base acreage minus the 10 per cent acreage reduction figure and the 15 per cent normal flex acres (14.9 million effective base acres590x .75 = 11.175 million acres) and multiplying by the programme yield (602 pounds per acre), the “quantity of production eligible to receive the administered price”591 is 6,727 million pounds, yielding a price gap deficiency payment calculation of $1,009 million. Thus, the figure in paragraph 115 of the US rebuttal reflected a conservative approach that understated the support resulting from a price gap calculation.
9. In this regard, the United States notes Brazil’s argument with respect to the 1995 Statement of Administrative Action, which explained that Peace Clause protection would apply “unless the AMS for the particular commodity exceeds the level decided in the 1992 marketing year.”592 We agree with Brazil that this reference to “AMS” is “non-textual” because the Peace Clause uses the term “support decided” and not “AMS.”593 However, to the extent that the Panel were to examine “the AMS for the particular commodity” – that is, the upland cotton AMS – the United States has demonstrated that in no year from 1999-2002 does that AMS exceed the 1992 level.
III. The US Level of Support Argument Does Take Into Account All Product-Specific Support That Challenged US Measures Grant 10. Brazil has argued that “the United States ‘72.9 methodology’ does not – and cannot account for cottonseed payments, Step 2 payments, storage payments and interest rate subsidies,” which the United States has identified as product-specific support.594 Brazil then alleges that the US methodology “would sanction the cover-up of hundreds of millions – if not billions – of dollars of expenditures.”595 Over-heated rhetoric aside, Brazil’s argument is simply erroneous.
11. Brazil argues that the United States has not accounted for Step 2 payments. The United States directs the Panel’s attention to the US rebuttal submission, paragraphs 111 and 113, and the US first written submission, paragraph 111. The United States has noted that, because the availability of Step 2 payments is contingent on certain price conditions existing during the marketing year, the level of support decided must relate to the payment parameters. While these have changed slightly with the 2002 Act, these minor adjustments do not alter the revenue ensured for producers by the marketing loan rate of 52 cents per pound because Step 2 merely provides an alternative avenue of providing support (through processors rather than directly to producers). In addition, these minor adjustments cannot overcome the greater than 20 cents per pound difference in product-specific support between marketing years 1992 and 1999-2002.
12. Brazil argues that the United States has not accounted for cottonseed payments. The United States directs the Panel’s attention to the US rebuttal submission, paragraph 111 fn. 136, 137 and paragraph 113. While the United States maintains that these measures are not within the Panel’s terms of reference596, we note that cottonseed payments for the 1999, 2000, and 2002 crops ranged in value between 0.6 to 2.3 cents per pound (factoring expenditures – the way these measures were decided – over production). Thus, given the greater than 20 cents per pound difference in product-specific support between marketing years 1992 and 1999-2002, cottonseed payments too do not materially affect the comparison between marketing year 1992 and any other year.
13. With respect to storage payments and interest rate subsidies, we note that these are US Government estimates of support provided through activities relating to operating the upland cotton marketing loan programme.597 This support is already captured, however, in the level of support expressed by the marketing loan rate. Were these costs not borne by the United States, the costs to the producer would reduce the guaranteed revenue below the loan rate. In fact, Annex 3 of the Agreement on Agriculture explains that, for purposes of market price support calculated using a price gap, “[b]udgetary payments made to maintain this gap, such as buying in or storage costs, shall not be included in the AMS.” Similarly, where the support provided by marketing loans is measured using a price gap methodology (the only appropriate AMS calculation for purposes of the Peace Clause)598, “payments made to maintain this gap,” such as storage payments and interest rate subsidies, should not be counted separately.
IV. Brazil’s New Green Box Arguments Are in Error 14. In its rebuttal submission and comments on US answers to questions from the Panel, Brazil advances two novel arguments. First, Brazil for the first time responds to the US argument that Brazil’s interpretation of paragraph 6(b) would create an inconsistency between that provision and the fundamental requirement of the first sentence of Annex 2, paragraph 1.599 Second, Brazil argues that the US interpretation of Annex 2, paragraph 6(b), would render paragraph 6(e) of that Annex a nullity. Neither of these arguments withstands scrutiny.
15. First, Brazil misunderstands the US argument that Brazil’s reading of paragraph 6(b) creates an inconsistency between that paragraph and the fundamental requirement of the first sentence of Annex 2, paragraph 1, and therefore its arguments go astray. The United States has noted that if payments under a decoupled income support measure were reduced or eliminated if a recipient were to produce any commodity, then the amount of payments would be (on Brazil’s reading) linked to the type of production and therefore inconsistent with paragraph 6(b), even though such a measure would meet the fundamental requirement of Annex 2. Brazil does not contest that such a measure would meet that fundamental requirement but instead argues that “requiring no production, i.e., on all base acres is not a ‘type of production’” because “[t]he notion of ‘type of production’ in paragraph 6(b) is necessarily linked to the amount of payment to some ‘type’ of commodity that is ‘produced’ and not to a production requirement itself.”600
16. With respect, if one were to credit this argument, then Brazil would appear to have misunderstood its own objection to US direct payments and production flexibility contract payments. That is, in the US example, payments are reduced or eliminated if a recipient produces any type of commodity. Similarly, Brazil’s objection to US green box payments is that payments are reduced or eliminated if a recipient produces certain types of commodities. Thus, in the former example, the amount of payment is “based on” (in the sense of being reduced by) “the type” of production undertaken by the producer – for example, production of upland cotton, fruits, vegetables, or wild rice – just as in Brazil’s argument on US green box payments, the amount of payment is “based on” (in the sense of being reduced by) “the type” of production undertaken by the producer – that is, fruit, vegetable, or wild rice production. Brazil’s objection to US green box payments under paragraph 6(b) would therefore apply with equal force to the US example,601 again, posing an inconsistency between Brazil’s interpretation of paragraph 6(b) and the fundamental requirement of Annex 2.602
17. Brazil also argues that the US interpretation of paragraph 6(b) “would render Annex 2, paragraph 6(e)[,] a nullity” because the “US interpretation of paragraph 6(b) as not requiring the production of ‘certain crops’ is the same as 6(e)’s prohibition on not requiring production of ‘any crops.’”603 Brazil’s own re-phrasing of the US argument, however, points to the distinction between the obligations contained in these two provisions. Paragraph 6(e) establishes that under a green box measure: “No production shall be required in order to receive such payments.” Thus, there can be no production requirement “in order to receive such payments,” but the provision is silent with respect to the amount of such payments at any particular time and any links to the “type or volume of production.” That is, were paragraph 6(e) alone part of Annex 2, a Member could arguably link the amount of payments to requirements on the “type or volume of production” so long as payment eligibility were not contingent on production.
18. Paragraph 6(b) forecloses that option by prohibiting a green box measure from linking the “amount of such payments in any given year” to “the type or volume of production.” That is, not only may a green box measure not require production, but the measure may not require a particular “type or volume of production” in order to obtain a payment amount. As the United States has noted, both direct payments and production flexibility contract payments meet that test because no “type or volume of production” is required to receive payments. For example, with respect to the fruits, vegetables, and wild rice planting flexibility issue, a payment recipient need not undertake any “type or volume of production” in order to receive the full “amount of payments” to which the farm’s base acres are entitled. Rather, the recipient need only desist from planting certain commodities. Thus, Brazil’s objection is nothing more than a statement that, under US green box measures, the amount of payments is linked to production not undertaken by the producer.
19. Finally, we note that Brazil’s reading of paragraph 6(b) could prevent Members from imposing on decoupled income support payment recipients any conditions relating to the type of production – for example, the planting of illegal crops or production of unapproved biotech varieties or environmentally damaging production. As a practical matter, no Member could accept not being able to impose any such conditions on payment recipients. The result of Brazil’s reading, then, would be to read decoupled income support out of Annex 2. This may be a favourable result from the Brazilian perspective, but the Panel should not adopt an interpretation of paragraph 6(b) not required by the text, not consistent with its context (in particular, the fundamental requirement of Annex 2), and with such potentially far-reaching results.
V. Answer to Additional Question 67bis from the Panel 67bis. Please state the annual amount granted by the US government in each of the 1999, 2000, 2001 and 2002 marketing years (as applicable) to US upland cotton producers, per pound and in total expenditures, under each of the following programmes: production flexibility contract payments, market loss assistance payments, direct payments and counter-cyclical payments.
20. The Panel’s question would require ascertaining for each programme the amount of upland cotton produced by recipients of payments under the programme. However, the United States does not maintain and cannot calculate this information – that is, it does not maintain information on the amount of expenditures made under the cited programmes to US upland cotton producers. This is because the payments do not relate to, and do not depend on, what crop, if any, is actually produced. Instead, each of these programmes makes payments with respect to pastproduction on base acreage in a fixed and defined base period, not with respect to whether one is currently a producer.
21. Thus, the United States did track total expenditures with respect to base acres of wheat, corn, barley, grain sorghum, oats, upland cotton, and rice under the expired production flexibility contract payments and market loss assistance payments and does track total expenditure with respect to base acres of wheat, corn, barley, grain sorghum, oats, upland cotton, rice, peanuts, soybeans, sunflower seed, canola, rapeseed, safflower, flaxseed, mustard seed, crambe, and sesame seed under the direct payments and counter-cyclical payments.604 However, the fact that a recipient at one time produced one of these crops says nothing about what crops the recipient is currently producing, if any. In other words, payments made on the basis of past production of upland cotton do not tell anything about whether the recipient is currently producing cotton, corn, livestock, hay, or any other crop or is not producing at all. As a result, it is not possible to derive from these payments whether the payment is being received by an upland cotton producer.
22. The Panel’s question points to a fundamental difficulty with Brazil’s approach. Brazil would have the Panel allocate “support to a specific commodity” – upland cotton – on the basis that certain of these measures determine payment amounts (for base acres) based on current or recent market prices for that commodity. However, how could the payment be “support to a specific commodity” (support “provided for an agricultural product in favour of the producer of the basic agricultural product”) if there need be no production of upland cotton in order to receive payment?
23. Brazil attempts to avoid this result by arguing that various US payments (direct, counter-cyclical, production flexibility contract, and market loss assistance payments) are not non-product-specific support because they are not payments to “producers in general.” The United States has addressed this erroneous interpretation in detail in its rebuttal submission. In short, Brazil’s reading requires ignoring the definition of product-specific support in Article 1(a) (that is “support . . . provided for an agricultural product in favour of the producers of the basic agricultural product”), which Brazil has not interpreted, in over 450 pages of submissions and statements, even once.605 In fact, Brazil’s reading of the definition of non-product-specific support (“support provided in favour of agricultural producers in general”) reads the phrase “in general” as meaning “in a body; universally; without exception.” However, this dictionary definition is considered “obsolete”606 and so would hardly be the “ordinary meaning” of the term.
24. As Brazil has conceded, moreover, payments made with respect to upland cotton base acres are not necessarily in favor of upland cotton producers since those acres may not be planted to upland cotton – indeed, may not be planted at all. We note that Brazil has adjusted its entire AMS calculation to reflect its belated realization that, under its own theory, “only the portion of . . . payments [on “upland cotton” base acres] that actually benefits acres planted to upland cotton can be considered support to upland cotton.”607 But Brazil’s adjustment is not enough. Brazil simply takes the ratio of actual upland cotton acreage to “upland cotton” base acreage under a given programme. However, there is no reason why upland cotton acreage need be planted on “upland cotton” base acreage. Consider this example:
One farm could have 100 base acres of upland cotton and currently plant those 100 acres to corn; direct and counter-cyclical payments would be made on those 100 “upland cotton” base acres that actually are planted to corn.
Another farm could have 100 base acres of corn and currently plant those 100 acres to upland cotton; direct and counter-cyclical payments would be made on those 100 “corn” base acres that actually are planted to upland cotton.
Brazil’s approach (dividing upland cotton planted by upland cotton base acres) would simply say that all of the direct and counter-cyclical payments on “upland cotton” base acres are “support to upland cotton” because there are 100 “upland cotton” base acres on which payments were made and 100 acres currently planted to upland cotton, even though these are found on completely separate farms.
Thus, Brazil’s ratio does not identify, even on Brazil’s own terms, the alleged support to upland cotton (that is, “payments that actually benefit acres planted to upland cotton”) under these programmes.608
25. Brazil’s own approach would require Brazil to match up payments for upland cotton base acres with the amount of upland cotton production on those base acres, but Brazil has not done so.609 At best, Brazil speculates as to the likelihood of a person with cotton base acres actually producing upland cotton on those base acres, and even that speculation is flawed.610 However, such an approach amounts to little more than speculation and, even if Brazil’s erroneous interpretation were used, does not meet Brazil’s burden of establishing a prima facie case.
26. In addition, under Brazil’s own approach, the payments made in relation to corn base acres would be support for corn even if planted to upland cotton. However, Brazil’s approach would appear to result in double counting the support – the same payment would be support to corn (because it was related to corn base acres) and support to upland cotton (because cotton was produced on base acres eligible for payments). In other words, Brazil is trying to have it both ways:
First, Brazil argues that payments made based on production on base acres during a base period is support to the crop that was produced during that base period, regardless of what is actually produced currently (that is, payments made for upland cotton base acreage is support to upland cotton even if the producer is now growing corn on that acreage).
Second, Brazil argues that payments made under these programmes are support to the crop that is currently being produced, even if the crop being produced is different from the base crop (that is, payments made for corn base acreage is support to upland cotton if upland cotton is being produced on the corn base).
27. Furthermore, because payments under the cited programmes are made with respect to historic acres and yields during a base period, it is not possible to calculate the “annual amount granted by the US government . . . to US upland cotton producers, per pound.” Counter-cyclical payments, for example, determine the payment rate for base period production as the difference between a target price and the sum of the direct payment rate plus the higher of the market price or the loan rate. However, the per pound payment rate for upland cotton base acres applies only for base period production (base acres x payment yields), not current production. Thus, to express these payments per pound begs the question: “Per pound of what?” Any production figure used – whether base period production or production in any year from 1999 through 2002 – results in a highly artificial per pound rate since (as noted above) these payments will be or were (as the case may be) received by a recipient regardless of whether he or she produced any upland cotton production.
28. We are able to provide to the Panel the total outlays under the cited programmes with respect to upland cotton base acreage: