Questions to both parties 78. Could both parties comment on the statements of Canada that "(a)t issue is whether these programmes....threaten to cause serious prejudice simply by virtue of their existence" and that "(c)ertain subsidy programs, by their very nature, give rise to a constant likelihood of support that results in a permanent threat of serious prejudice"? (Third Party Submission of Canada, paras. 9-10) 172. As set out in Brazil's answer to this question, Brazil is not asserting that the existence of the marketing loan and counter-cyclical subsidy programs, in and of themselves, causes a threat of serious prejudice.
173. Brazil notes the U.S. assertion in its response that a challenge to the very "existence" of such programs would be "extraordinary" and "remarkable." In effect, what the United States asserts is that "per se" challenges – i.e., challenges to the very existence – of subsidy programs under Articles 5 and 6.3 of the SCM Agreement are, as a practical matter, simply not viable. Yet, the United States' preliminary ruling arguments in this case seek to require the categorization of Article 6.3 claims in strict "per se" and "as applied" terms. According to the United States, "as applied" claims could only be asserted against the application of subsidies, i.e., the payment of subsidies for a specified period of time.301 And the only implementation obligation regarding such subsidies that cause adverse effects, according to the United States for such subsidies, is to remove the effects of such historical subsidy payments (i.e., the "application").302On the United States view, a Member can take no action when recurring payments are found to cause adverse effects because the "effects" of past payments are wiped away by new payments that, it contends, cannot be regarded as "measures taken to comply", even though they have an extremely close connection to the DSB's recommendations and rulings regarding past payments. Brazil addresses this issue further in its comments on the United States' response to question 45, above.
174. In its response to question 78, the United States has now confirmed its views of the practical impossibility ("extraordinary" and "remarkable") of establishing a "per se" violation of Articles 5 and 6.3 of the SCM Agreement. Yet, in doing so, the United States has also confirmed that its arguments reduce the obligation in Article 7.8 of the SCM Agreement to inutility where a Member grants annual recurring payments to support the production of an annual crop. Under its strict segmentation of serious prejudice claims into "per se" and "as applied" categories, this remedy becomes meaningless. Indeed, acceptance of such a strict categorization would prevent Members suffering serious prejudice from securing any prospective relief for the future application/payment of recurring subsidies under such programs. This is because, as the United States has argued, only "per se" challenges can result in any obligation to change the statutory or regulatory provisions mandating subsidies.303 Members seeking relief by challenging, "as applied", the application of those statutory and regulatory provisions would, in the case of recurring annual payments, have to continually institute new WTO proceedings, but would never have any right to take "countermeasures" because there would never be any compliance measures.304 As Brazil has demonstrated, such an interpretation is neither justified nor mandated by the text, context, and object and purpose of Articles 5 and 6 of the SCM Agreement.305
79. Could the parties state their views on the analysis of the ordinary meaning of the term "threat" in paras. 15-28 of the Third Party Submission of Canada?
175. The United States' response reaffirms its attempt to import the "imminent and clearly foreseen" standard of threat in Article 15.7 into Part III of the SCM Agreement.306 Brazil has explained in its response to this question and in its First Written Submission why this is contrary to the text, context, and object and purpose of Articles 5 and 6 of the SCM Agreement. In this comment, Brazil addresses briefly a number of points raised by the U.S. response.
176. The United States is incorrect in asserting that footnote 5/Article 6 and Article 15.7 of the SCM Agreement, together with Article 4.1 of the SafeguardsAgreement "serve a similar function."307 Brazil explained in detail in its response to this question308 that the "imminent and clearly foreseen" threat standards in Article 15.7 and Article 4.1 serve a distinct function, i.e., regulating unilateral action by investigating authorities of imminent surges of imports.309 By contrast, Part III of the SCMAgreement deals with the far broader object of protecting against the effectsofsubsidies in a multilateral dispute settlement context. Such effects may or may not involve the imminent shipment of goods. Thus, as Brazil has argued, a threat standard such as "significant likelihood" is consistent with both the slower pace of WTO dispute settlement procedures and its the associated remedies, as well as the time it takes for the effects of certain types of subsidies to materialize.
177. The United States' response distinguishes between a temporal ("imminent") and a probabilityofoccurrence ("clearly foreseen") component of threat under Article 15.7. The United States asserts that Canada's (and Brazil's) approach ignores the temporal element altogether.310 This is not correct. Rather, Brazil asserts that the notion of "imminent" is appropriate given the ability of investigating authorities (under Article 17 of the SCMAgreement) to take rapid action against subsidized imports. Yet, such a standard is not appropriate in a multilateral proceeding. In effect, the United States is requesting the compliance Panel to adopt a "temporal" standard that will force WTO Members to wait just until subsidies are about to cause serious prejudice before asserting their multilateral rights. Brazil has explained why application of the particular "temporal" element ("imminent") of Part V into Part III of the SCMAgreement is inconsistent with rules of treaty interpretation.
178. However, Brazil emphasizes again that it does not suggest that there be no temporal element to a threat standard in Part III of the SCMAgreement. Rather, Brazil's "significant likelihood" standard incorporates the notion that the threat must be real, not be based on allegation, conjecture or remote possibility. In sum, the temporal element should be tailored to the particular object and purpose and remedies and procedures applicable to multilateral disputes settlement proceedings applicable to Part III of the SCMAgreement, not blindly imported from the different context in Part V of the SCM Agreement.
Questions to the United States 80. How does the United States address the argument of Japan that in view of the different purposes of Parts III and V of the SCM Agreement the standard for determining threat of material injury in Article 15.7 of the SCM Agreement is an inappropriate standard for determining the existence of a threat of serious prejudice under Part III of the SCM Agreement? (Third Party Submission of Japan, paras. 8-12.) '
179. Brazil believes that Japan has correctly identified significant differences between Parts III and V of the SCMAgreement that demonstrate the inappropriateness of importing the "imminent and clearly foreseen" standard of Article 15.7 into Part III of the SCMAgreement. Brazil therefore disagrees with the primary point made by the United States in this response, i.e., that there is no difference between the function, object, and purpose of Article 15.7 and Part III of the SCMAgreement.
180. The United States suggests that "there is no basis for understanding the same terms in [Article 5/footnote 13 and Article 15.7] to have different meanings based on the an assumption that national authorities are necessarily prone to 'misusing' their discretion whereas WTO panels are not."311 This is a false argument. Neither Brazil nor Japan "assumes" that investigating authorities will misuse their discretion. Instead, Brazil emphasizes that there are considerably different "temporal" considerations for such authorities compared to a WTO panel.
181. While the United States champions the "temporal" element of "imminent", the U.S. argument ignores the vastly different temporal periods involved in a rapid unilateral provisional remedy based on Part V and years of WTO dispute settlement proceeding under Part III of the SCMAgreement.
182. It is entirely appropriate that investigating authorities armed with rapid provisional remedies to control imports be compelled to make a finding that such imports are "imminent." The purpose of the rapid provisional remedies under Articles 17 and 15.7 of the SCMAgreement ensure that the domestic industry need not suffer material injury while waiting for remedies to take effect. Those rapid remedies available in the unilateral trade remedy context explain why the drafters used an "imminent" temporal standard of threat.
183. But the situation in a challenge under Part III of the SCMAgreement is totally different. Part III is not focused on controlling subsidized imports but rather on the adverse effects of any subsidy. Even the expedited dispute settlement procedures potentially available (but seldom used) under Article 7 of the SCMAgreement require at least 18 months of litigation time before a ruling and implementation can take place. No provisional "self-help" remedies are available under these rules. If "imminent" threat were the standard, then a WTO panel could be confronted with facts demonstrating that the threat was "clearly foreseen" but was not "imminent" because the effect of the subsidies would not take place for another year or two.312 (By contrast, a domestic industry could petition investigating authorities only months before major shipments of subsidized goods were scheduled to be shipped and obtain relief under procedures sanctioned by Articles 15.7 and 17 of the SCMAgreement.) The complaining WTO Member would, in effect, be left without a remedy for future serious prejudice that was real, but not "imminent."
184. The United States asserts that both Article 15.7 and footnote 13/Article 5 strike a balance between disciplining and permitting subsidization.313 Yet, if the real balance between disciplining subsidies and permitting their use is to be maintained, the concept of threat must be flexible enough to allow the disciplining of the effects of subsides – in practical terms – well before subsidized products are about to hit the market. Avoiding real and foreseen future adverse effects in the form of material injury and serious prejudice is the object of both Article 15.7 and Part III of the SCMAgreement. However, the entirely different temporal aspects of these two different remedies and procedures compel the use of different threat standards to maintain that balance. Otherwise, the U.S. interpretation creates far greater an opportunity for subsidizing Members to cause serious prejudice without any effective multilateral relief. This appears to be the main objective of the Japanese arguments and is one that Brazil supports.
81. How does the United States respond to the argument of Australia that "it is beside the point for the United States to argue that the programmes under consideration are due to expire in late 2007"? (Third Participant Oral Statement of Australia, para. 13) 185. The United States' response repeats its erroneous assertions that there is no threat of serious prejudice because the FSRI Act of 2002 will allegedly expire by the end of marketing year 2007, i.e., 31 July 2008.314
186. Brazil disagrees that the alleged expiration of the FSRI Act of 2002 in the end of July 2008 is a relevant fact before the compliance Panel.315 The "matter" before the compliance Panel with respect to Brazil's threat of serious prejudice claim is whether marketing loan and counter-cyclical subsidies for MY 2006 and MY 2007 cause a threat of serious prejudice. MY 2006 is the relevant year to consider whether threat of serious prejudice presently exists. It will not end until 31 July 2007, i.e., after the compliance Panel issues its determination. MY 2007 will not end for another 16 months, i.e., on 31 July 2008. The FSRI Act of 2002 remains in effect until the end of MY 2007.316 Brazil's threat of serious prejudice claims in this proceeding involve the determination of whether there is a threat of serious prejudice today, not at the end of MY 2007 or beyond.
187. Yet, having asserted (incorrectly) that the expiration of the FSRI Act of 2002 is a relevant fact by suggesting that marketing loans and counter-cyclical subsidies may also completely expire, the United States is in no position to then assert that it is entirely irrelevant for the compliance Panel to consider whether the marketing loan and counter-cyclical subsidies will continue, in some form or another, in the next U.S. Farm Bill.317 It is undisputed that there will be some new Farm Bill that will be enacted at some time in the future after these proceedings have finished.318 The marketing loan program for upland cotton has been in existence since 1986 and, since, has been a key component of U.S. farm legislation. The current USDA proposal calls for a continuation of both the marketing loan and the counter-cyclical payment programs with certain changes.319 Thus, Australia is certainly correct when it states that "there is no guarantee that [the marketing loan and counter-cyclical programs] will not be rolled over or maintained in another form with adverse effect."320 The compliance Panel cannot, of course, determine the exact scope of the marketing loan and counter-cyclical provisions that will exist in the next U.S. farm bill.
188. Finally, whether the provisions of any new Farm Bill fully remove the serious prejudice caused by the FSRI Act of 2002, or the ongoing threat thereof, cannot be determined at this time. However, the United States is incorrect to suggest that the mere expiration of the FSRI Act of 2002 in the future would represent a "withdrawal" of the subsidy.321 If the marketing loan program and the counter-cyclical payment programs remain in some form in the new Farm Bill – or are replaced by programs with similar effects – then these subsidies will not have been withdrawn, within the meaning of Article 7.8 of the SCMAgreement.322
82. Could the United States comment on the projections of marketing loan and counter-cyclical payments in Table 26 of Brazil's First Written Submission and on the projections of prices and subsidy payments in Table 27 of Brazil's First Written Submission? Could the United States explain how the data in these Tables support its argument that producers are likely to expect low or no marketing loan payments in MY 2007? (Rebuttal Submission of the United States, para. 418) 189. In its answer to this question, the United States focuses on projections of prices323 to derive projected subsidies and downplays projections of subsidyoutlays itself (what it terms "aggregate figures"324), arguing that it is "difficult to fit these projections"325 with the projections of prices. This is highly misleading and contrary to basic accounting principals used and espoused by USDA.326 Projections of prices and subsidy outlays are, in fact, entirely consistent, as explained below.
190. Consider the following simplified example, as articulated by Professor Sumner in Exhibit Bra-659.
But, let us take the case of a farmer with the following view of the future AWP. He considers that there is a 40% chance of an AWP of 42 cents, a 20% chance of an AWP of 52 cents and a 40% chance of an AWP of 62 cents. The expected value is exactly 52 cents. He expects no marketing loan payout, but the value of the program is 4 cents per pound to him. That is 0.4(52-42) = 4 cents. Only if the farmer was absolutely sure the AWP could not be below the loan rate would the program be of no value.327
191. Even if the Adjusted World Price ("AWP") is expected to be $0.52 per pound, or higher, it does not mean that the marketing loan program provides no benefit. As noted by the Appellate Body, even if U.S. upland cotton farmers expected prices to be above the adjusted world price at the time of planting, they "were also aware that if actual prices were ultimately lower, they would be 'insulated' by government support, including not only marketing loan program payments but also counter-cyclical payments, which were based on a target price of $0.724 per pound."328
192. The distinction between the prices farmers expect and the marketing loan subsidies they expect is akin to the difference between deterministic and stochastic projections of marketing loan subsidy outlays.329 Like farmers, USDA also ascribes value to the marketing loan program even when prices are expected to be at or above the loan rate. For example, USDA's deterministic projection of marketing loan subsidies, which is based on price projections, is $0 in MY 2009, while its stochastic projection is $468 million.330 193. USDA's deterministic projection of marketing loan subsidies in MY 2007 is $238 million, indicating that USDA expects the AWP to be below the loan rate.331 However, based on the probably that prices will be even lower, USDA's stochastic projection of marketing loan subsidies is $798 million. USDA's latest projection that upland cotton production will be 20.7 million bales in MY 2007332 suggest per-unit expected marketing loan subsidies of 8 cents per pound333, not 2 cents per pound, as put forward by the United States.334
194. In sum, by relying on expected prices, instead of expected subsidies, the United States is repeating a mistake it has made throughout this proceeding. Only stochastic estimations of marketing loan subsidies capture the likelihood that prices will be lower and payments higher. Relying on the difference between the expected AWP and the loan rate does not capture this effect. As explained by USDA, "deterministic projections, by their nature, tend to underestimate outlays."335
195. Brazil notes that the United States does cite to FAPRI's July 2006 stochastic projections of marketing loan subsidies. However, this projection of marketing loan payments of 1.9 cents per pound336 is significantly below USDA's own projections of 8 cents per pound. There is a very good reason why USDA estimates are more reliable than the July 2006 FAPRI projections – they are more recent. USDA estimates are found in USDA's Commodity Estimates Book for the FY 2008 President's Budget, which was released in February 2007 – eight months later than the FAPRI baseline that the United States relies on.337 196. Furthermore, while the complete 2007 FAPRI baseline has not been released338, the parts that have been released confirm that FAPRI's projections of upland cotton prices in July 2006 were too high. For example, FAPRI now projects the MY 2007 U.S. farm price for upland cotton to be $0.518 per pound.339 Last July, it predicted the same price to be $0.534 per pound.340 197. Current futures prices also indicate that large marketing loan and counter-cyclical payments can be expected due to expected low prices in MY 2007. The following chart details daily closing prices of the December 2007 contract for upland cotton.
Figure 6 – Prices for the December 2007 Upland Cotton Futures Contract341
198. The average January to March closing price of the December 2007 futures contract was $0.588 per pound.342 The price of the December 2007 futures contract has since fallen to less than $0.57 per pound. Brazil recalls that marketing loan subsidies are based off of the adjusted world price, which is, on average, 17.3 cents below the average January to March futures price.343 A simple linear regression of futures prices and actual marketing loan payments indicates that, based on the current futures price of $0.588 per pound for MY 2007, farmers can expect a marketing loan subsidy of 12.3 cents per pound344, or about $1.2 billion.345 This would represent 17 percent of the total projected cost of producing a pound of cotton in MY 2007.346
199. In the end, it is impossible to know how large marketing loan subsidies will be in MY 2007. Brazil notes that the average futures price of $0.588 per pound during the planting period of MY 2007 is similar to futures prices in the planting period of MY 2001, 2003 and 2006.347 Actual marketing loan subsidies in these years varied widely, ranging from $2.6 billion in MY 2001, to $757 million (projected) in MY 2006 and to just $184 million in MY 2003.348 The inability of futures prices to predict actual subsidies was even more apparent in MY 2004 and 2005. The average futures price in MY 2004 was $0.674 per pound, while in MY 2005 it was $0.529 per pound. Yet, in MY 2004, marketing loan subsidies amounted to $1.8 billion and were 50 percent greater than marketing loan subsidies of $1.2 billion in MY 2005.349 The following chart shows the progression of nearby futures prices over the past four years:
Figure 7 – Nearby Futures Prices350
200. In sum, every farmer knows that expectations based on futures prices, just like USDA and FAPRI projections are just that – projections – and that actual prices and actual subsidies vary widely. Despite the U.S. focus on "expectations," the underlying reality throughout the life-time of the FSRI Act of 2002 has been self-sustaining low upland cotton prices and enormous price-contingent marketing loan and counter-cyclical subsidies. The evidence in the record fully supports the conclusion that such subsidies and their price suppressing effects will continue throughout the lifetime of the FSRI Act of 2002.
D. Export credit guarantees
1. Outstanding export credit guarantees Questions to the United States 91. In paragraph 342 of its First Written Submission, Brazil indicates that the total amount of guarantees under the GSM 102, GSM 103 and SCGP programs outstanding on 1 July 2005 amounted to $8.5 billion. (c) Pleaseindicatewhatproportionofthatamountconcernsexportsofscheduledproducts, andinparticularrice(pleasedistinguish, ineachcase, betweenprincipalandinterest).
201. Although the United States asserts that it does not maintain data concerning outstanding ECGs on a commodity-specific basis, FAS' monthly summaries of GSM 102 program activity is recorded on a commodity-specific basis (at least with respect to select commodities, such as rice).351
202. However, if the United States does not maintain data regarding ECGs issued and outstanding on a commodity-specific basis, this fact is telling, and confirms that there is no factual basis for the United States to characterize the measure subject to Brazil's claims of inconsistency with Articles 10.1 and 8 of the Agreement on Agriculture as "the pig meat and poultry meat GSM 102 guarantees".352 Neither the amended GSM 102 program in its totality, nor the individual amendments, set out terms or conditions that differ as between different eligible products.353 In its response to the compliance Panel's question, the United States insists that it does not even maintain data on GSM 102 ECGs issued and outstanding on a product-specific basis.