97. The Philippines objected to the consideration by the Panel of the corrected translations of the two texts. The Philippines considered that there were substantive differences between the initial and corrected translations. The Philippines argued that it lacked the resources to check the corrected translation, that to do so would delay the dispute settlement process, and that to accept the corrected translations at this late date would be unfair to the Philippines.

98. At the meeting, the Panel took note of the corrected translations submitted by Brazil. It informed the Philippines that if it believed the corrected translation inaccurately reflected the original determination it should so inform the Panel, and the Panel would make any such ruling as might be required. The Panel indicated that if the Philippines considered that it needed more time as a result of the submission of the corrected translations, the Panel would grant such additional time.

99. The Philippines subsequently submitted a letter objecting to the "acceptance" by the Panel of the corrected translations. The Philippines objected to Brazil's failure to provide advance warning of the pending corrections, and inferred that the corrections were submitted in response to arguments made during the oral presentation of the Philippines and third parties and/or to address some of the concerns expressed by the Panel in its questions to Brazil. The Philippines indicated that it lacked the resources to engage a professional translator, but pointed out three instances in which it considered that Brazil had gone beyond translation corrections to change the substance of the documents in question.

100. The Philippines observed that the duty to accord sympathetic consideration to and afford adequate opportunity for consultations regarding any representations made by another Member concerning measures affecting the operation of any of the WTO Agreements was one of the most important procedural obligations of the Members of the WTO, citing Article 4:1 of the DSU.

101. The Philippines maintained that the reasons cited by Brazil for its refusal to consult, that the Philippines did not invoke the applicable law, and should invoke its rights under the Tokyo Round SCM Code instead, were arguments which Brazil could have advanced in the consultations, but did not justify a refusal to consult. The obligation to consult would be meaningless if WTO Members could refuse to consult when they did not consider the legal claims of the WTO Member requesting consultations to be justified. Moreover, agreeing to consult under the GATT 1994 would in no way have prejudged Brazil�s position on the applicable law in the present Panel proceedings since Article 4:6 of the DSU makes explicit that "consultations shall be ... without prejudice to the rights of any Member in any further proceeding". The Philippines asserted that, under international law and in GATT practice, it is recognized that a legal finding may be sought for the purpose of obtaining satisfaction or a guarantee of non-repetition. Therefore the Philippines submitted that a finding on this issue was necessary to strengthen the consultative process in general and, in particular, to reduce the likelihood of Brazil's resort to similar arguments to avoid consultations in the future.

102. Brazil disputed the Philippines' contention that it "refused" to consult. Brazil noted that it had offered to consult under the Tokyo Round SCM Code three times. In addition, Brazil submitted a list of the informal consultations that were held, noting that these consultations were considered "informal" at the request of the Philippines.

103. Brazil argued that the Philippines consistently failed to meet its obligation to supply necessary information within the applicable time periods, despite numerous extensions, and Brazil therefore properly based its subsidy decision on the best information available. Moreover, Brazil asserted that in making its determination it in fact relied mainly on information the Philippines did submit. Brazil maintained that Article VI of GATT 1994 does not contain any administrative provision specifically addressing information gathering. Logically, however, it must contemplate an investigation into the existence and nature of the subsidies prior to imposition of a countervailing duty, since it requires a finding of subsidization and injury. Any such investigation presumes the cooperation of the party being investigated in providing the necessary information to enable the investigating authorities to make those determinations. Brazil argued that this interpretation is supported by past practice with respect to Article VI of GATT 1947, which contained identical language. Brazil referred to the Second Report of the Group of Experts, which addressed this issue:

In Brazil's view, the last sentence of this section contemplates the use of the best information available. Brazil also noted that Article 2:9 of the Tokyo Round SCM Code specifically permits the use of best information available:

Brazil noted that it discussed the Tokyo Round SCM Code because the obligations of that Code were the obligations in force on Brazil during the investigation.

104. Brazil argued that the Philippines failed to submit requested information that would have enabled Brazil to determine, on a programme-specific basis, the per unit level of subsidization. Among the information not provided was information as to the annual disbursements of the programmes (including how many growers benefitted), the actual costs of administering the programmes, any specific eligibility criteria, and other requested information.

105. The Philippines asserted that it was the only country that fully cooperated with Brazil during the investigation, making good faith efforts to submit extensive information in meetings and documentation. The Philippines observed that, because it had always believed there were no subsidies on Philippine coconut fruit, it had simply responded to Brazil's questions without unduly trying to discern Brazil's possible purposes for asking those questions or to facilitate purported subsidy calculations.

106. The Philippines considered that Brazil's reliance on best information available in fact reflected that Brazil had not fully considered the Philippines' responses, and had only selectively relied on those responses. In this regard, the Philippines noted that Brazil had relied on the "Cost Worksheet" provided by the Philippines to construct the per hectare production cost of hybrid coconut trees, but ignored the per hectare fruit yield reflected in that same worksheet in calculating the per fruit production cost. In the Philippines' view, there were several clear instances when Brazil simply disregarded relevant Philippine responses. The Philippines also argued that Brazil had improperly treated some of the information provided by the Philippines as unsupported assertions. For example, Brazil considered as evidence of levy allocations a list that showed how records of the Marcos regime accounted for the levy funds, but deemed unsupported assertions the contents of the book written by a knowledgeable coconut industry official discussing the unreliability of that list. ³⁶ Indeed, the Philippines asserted that Brazil neither mentioned nor evaluated the book, 20 Million Farmers are Victims of Levy Racket, and did not even acknowledge that the Philippines submitted material assailing the reliability of the list. In the Philippines' view, this represented an improper one-sided treatment of the Philippines' submissions.

107. Moreover, the Philippines argued, referring to the report of the Panel in Salmon, that, before an investigating country can resort to the use of best information available, "the first question to be asked [is] whether the information requested ... was of the type that would make it possible to calculate the amount of a subsidy ... if this information had been requested, and had not been provided, then subsidy findings could be made `on the basis of the facts available'". ³⁷ In the Philippines view, Brazil never asked the types of questions required for, inter alia, the downstream subsidy analysis called for by the Panel report in Pork. The questionnaire issued by Brazil sought information pertaining to programmes for the upstream product, coconut fruit. Thus, in the Philippines' view, Brazil had no basis for using best information available in regard to the consideration of subsidies to the downstream product, desiccated coconut.

108. Brazil noted that, with respect to the various programmes investigated, the following information was requested and not provided: the official reports of the officials responsible for administering the programme, "administrative norms and regulations, duly-updated, documents that prove the actual coverage of the programmes covered by legislation, commercial conditions of the benefits conceded, and businesses in questions and our cultivators of coconuts benefitted in the period of investigation," supporting documentation to confirm the information provided in response to the first questionnaire, the actual expenditures for the programme during the period of investigation, detailed information requested about each component of the programme, for example the number of farmers benefitted by the "intercropping," the number of hectares occupied by nurseries, and the actual costs of the nurseries. In response to a question from the Panel, Brazil specified that it had requested copies of official reports relative to each of the programmes; the source for any data provided; the official annual reports of the government entities in charge of the programmes cited; for each identified programme, the commercial conditions of the benefits conceded, businesses covered by the programmes, and cultivators of coconuts that benefitted in the period of investigation; information needed to quantify subsidies received during the period of investigation, for example, with respect to the Presidential Decree No. 582/74 programme, the amount received by each farmer, how many farmers benefitted from loans under the programme, the interest rates at which the loans were granted, percentages of borrowers for each interest rate, loan amortization conditions, whether the farmers get (or could get) private loans in the absence of the programme, and if so, at what interest rate and amortization conditions. Brazil recognized that the Philippines provided information during the investigation, but maintained that the Philippines did not provide sufficient information to enable Brazil to make its determination without reliance on best information available. Brazil reiterated that it had relied where possible on the information provided by the Philippines.

109. The Philippines maintained that, of the seven Philippine programmes Brazil assertedly concluded involved subsidies, five involve the redistribution to the coconut farmers of a levy previously collected from them, one comprised a land reform programme that generally applied to landless Philippine citizens; and the last consisted of investment incentives for which traditional coconut products were not eligible, but which the Ordinance speculated might be available to such products in the future. In the Philippines' view, these programmes cannot be characterized as subsidies.

110. The Philippines took the position that industry programmes funded by direct collections from their beneficiaries are not subsidies, referring to the Interpretative Notes to Article XVI:3 of both GATT 1947 and GATT 1994, which state in part that price stabilization schemes can be characterized as subsidies only if "wholly or partly financed out of government funds in addition to funds collected from producers in respect of the product concerned". Accordingly, the five groups of replanting and livelihood programmes involving redistribution to farmers of a levy collected from them from 1973 to 1982 were not subsidies. The Philippines argued that Brazil did not deny that levy-funded programmes are not subsidies, but instead had surmised that programmes after 1984 must have been funded by the Philippine Government because a list of allocations of those funds purported to show that the funds had already been completely redistributed as of 1984. However, the Philippines maintained that Brazil had ignored evidence showing that only a small portion of the levy funds were actually redistributed to the farmers. ³⁸ Moreover, the Philippines argued that the prior misuse of the levy funds had both factual and legal implications for purposes of determining whether the Philippine replanting and livelihood programmes were levy-funded. As a factual matter, any levy funds that were diverted from the coconut industry cannot, in the Philippines' view, be deemed to have subsidized those programmes. As a legal matter, the Philippines was of the view that any coconut industry programmes after February 1986, the departure of former Philippine President Ferdinand E. Marcos, could be pursued in conjunction with the Philippine Government's efforts to recover the levy funds and to rectify their prior misuse. ³⁹ In particular, the Philippines believed it was proper for the Government to finance post-February 1986 programmes with levy funds that the Government had so far traced, such as those in the United Coconut Planters Bank, or recovered, with some limited advances, especially for rehabilitation programmes that alleviate the plight of farmers after natural calamities. Thus, in the Philippines' view, there was no factual and legal basis for characterizing the levy redistribution programmes as subsidies.

111. The Philippines also maintained that the Agrarian Reform Programme, which was of general application to all poor Philippine farmers, and was not specific to coconut lands, which were not even included in the programme until 1988, was not a subsidy. In response to a question from the Panel whether the Philippines considered that a subsidy must be specific in order to be countervailable under Article VI of GATT 1994, the Philippines responded in the affirmative, that it had been the practice in countervailing cases by some GATT Contracting Parties even before Article VI of GATT 1994, but such a specificity requirement was now expressly articulated only in Article 2 of the SCM Agreement.

112. The Philippines further contended that Brazil had not found that the programme's compensation scheme was concessionary to the farmers. The Philippine position was that the Agrarian Reform Programme could not be considered a subsidy since all the farmers had to pay for the lands distributed. The Philippines noted in this regard that the Ordinance stated no basis for determining the existence of a subsidy, and instead focused on the purported lack of information about the land area covered by the programme for purposes of calculating the amount of the alleged subsidy. Brazil had acknowledged that the Philippines had provided information about land valuation and payment forms, but had made no explicit determination that the land reform programme's compensation scheme had the features of a subsidy. Consequently, in the Philippines' view, Brazil had not had any basis for addressing the total land area covered by the programme. The Philippines asserted that if the compensation scheme was not found unduly concessionary, it could not be deemed a subsidy, and it did not matter how much coconut land was included in the programme. The Philippines also challenged Brazil's characterization of data on the total land area covered by the programme as "non-official," and therefore insufficient to enable calculation of the amount of the subsidy. The Philippines noted that the source of the data was the Philippine Coconut Authority, the government agency principally in charge of the coconut industry. The Philippines argued that Brazil should have accepted this information, and if it had, it would have found the programme's effects on coconut lands too insubstantial even if there had been a determination that the programme's compensation scheme was unduly concessionary.

113. Finally, the Philippines asserted that Brazil erred in treating as an actual subsidy the potential grant of future investment incentives under the Omnibus Investment Code of the Philippines, despite recognizing that coconut products were not eligible for such benefits. Article VI:6(a) allows countervailing duties to offset only subsidies that have actually been "bestowed", while grants under the Omnibus Investment Code were only potential. Moreover, the investment incentives were in fact available only to products that were "new" in the sense of being qualitatively different from the non-eligible traditional coconut products, which included desiccated coconut.

114. Brazil maintained that the information submitted by the Philippines was insufficient to demonstrate that the investigated programmes were not countervailable subsidies. Brazil argued that, even assuming arguendo that programmes aimed at increasing coconut production funded by a levy on the coconut producers themselves did not constitute subsides, documents submitted by the Philippine government in the countervailing duty proceeding contradicted the claim that the programmes were levy-funded. According to the documentation provided by the Philippines concerning the collection and disbursement of the coconut levy, which terminated in August 1982, the amount collected was fully disbursed between August 1973 and June 1984. With the exception of Presidential Decree 582, all the programmes under investigation were initiated after June 1984. Brazil asserted that the Philippines never explained how this levy could have financed any programme during the period of investigation (1993-1994), much less programmes initiated after 1984, when it was fully disbursed by June 1984. In addition, Brazil argued that other information provided by the Philippines also contradicted the statements that all programmes were financed by the levy. For example, the information provided by the Philippines indicated that, for the Farm Assistance and Livelihood Project, P 88.7 million of the total P 113.6 million was funded by government sources other than the levy. Given these contradictions that were never explained, Brazil considered that it was fully justified in its finding that the programmes at issue were not financed by the levy, especially in light of the fact that the levy collections had been fully disbursed by June 1984.

115. Brazil considered extraordinary the Philippines' argument that Brazil should not have relied on the submitted document showing that the levy funds were fully disbursed by June 1984 because the accounting in it was false and used to hide the misappropriation of these funds by the Marcos government. In Brazil's view, this argument suggested that Brazil had violated its obligations because it relied on information submitted by the government of the exporting country. Moreover, Brazil asserted that this was the only document submitted that provided any evidence, beyond unsupported assertions, on how the funds were allocated. Finally, even assuming the Philippines was correct that the Marcos regime had misappropriated the funds and did not pass them on to the coconut growers, Brazil maintained that this did not explain how, under the current government, levy funds that had already been misappropriated by 1984 could be disbursed to the growers in the 1990's. In Brazil's view, the Philippines' line of reasoning supported Brazil's finding that the programmes were not funded by the coconut levy.

116. With respect to the Agrarian Reform Programme, Brazil asserted that it had considered this programme to confer a subsidy because the information provided by the Philippines in response to Brazil's questions was inadequate to determine whether or not a subsidy existed. Brazil had noted that the Philippines had provided data as to the assessment and forms of payment, but asserted that the Philippines had not submitted official or substantiated information on the acreage covered and on the expenses and objectives reached. Brazil maintained that a respondent which did not supply that information could not have expected a no subsidy determination. Brazil noted that the record indicated that persons buying land under the programme could obtain government loans to pay for the purchase at 6 per cent interest, and that therefore those loans were not at commercial interest rates. Moreover, Brazil maintained that the Philippines had not presented any official information on the costs of the programmes, on the total amount loaned out, or on the acreage covered. Therefore, Brazil had been unable to determine from the information provided that the administrators of the programme were covering their costs for administering the programme. Brazil stated that all this information had been requested, and that therefore, on the basis of best information available Brazil had found that this programme provided a subsidy but was unable to determine on a programme-specific basis the effect or amount of the subsidy.

117. In response to a question by the Panel, Brazil stated that it did not consider that the Omnibus Investment Code provided a subsidy to either the coconut growers or the producers/exporters of desiccated coconut.

118. The Philippines asserted that Brazil's statement cast serious doubt on the fairness and reliability of the Ordinance as a public statement of Brazil's subsidy determinations. In the Philippines' view, the Ordinance clearly states that Brazil had identified a set of Philippine programmes that were not subject to action, and considered the other programmes, including the Omnibus Investment Code, to have granted subsidies. The Philippines maintained that, if as Brazil now asserted, no subsidy finding could be inferred from the Ordinance's discussion of the Omnibus Investment Code, then the Ordinance should have explicitly identified the Omnibus Investment Code with the other programmes not subject to action. Instead, the Omnibus Investment Code was listed together with the other programmes Brazil had found granted subsidies and thus actionable.

119. The Philippines contended that, even if the Philippine programmes were deemed to constitute subsidies on the production of coconut fruit, Brazil had no legal and factual basis for imputing those subsidies to the production of desiccated coconut. To justify a countervailing duty on desiccated coconut, Brazil was required to determine the existence and extent of a subsidy on desiccated coconut, and could not simply impute the subsidies for coconut fruit to desiccated coconut. Article VI:3 of GATT 1994 prohibits the amount of a countervailing duty from exceeding the amount of the corresponding subsidy allegedly bestowed directly or indirectly on the production of the product that is the object of the countervailing investigation. The Pork Panel had considered the imposition by the United States of a countervailing duty on imports of pork based on the subsidies provided to producers of live swine. The Panel, in considering the United States' analysis of the amount of the subsidy on pork, found the United States' decision, which imputed the subsidy on live swine to pork, based on the close inter-relationship between the two products, to be inconsistent with Article VI of GATT 1947, since Article VI:3 mandated that a countervailing duty be based on a subsidy to the specific product under investigation. ⁴⁰ The Panel ruled that a subsidy determination must be predicated on an examination of all relevant facts, ⁴¹ and where the alleged subsidies are provided to a separate industry producing the upstream product, operating at arm's length from the industry producing the downstream product subject to the investigation, the investigating authorities must at a minimum perform an analysis of the price effect of factors relating to the price paid for the upstream product by producers of the downstream product. ⁴² Based on the Panel decision in Pork, the Philippines argued that any subsidies on coconut fruit production could not simply be imputed to the production of desiccated coconut, because these are two different products that belong to two different industries. The Philippines maintained that Brazil failed to do any analysis of the specific effects of the programmes at issue on desiccated coconut.

120. The Philippines asserted that Brazil acknowledged that coconut fruit is an upstream product, whereas desiccated coconut is a downstream product, and that the two products were produced by separate industries operating at arms' length. In the Philippines' view, Brazil's countervailing duty on desiccated coconut was thus subject to the requirements established by the Panel in Pork. The Philippines read Pork to require, in this situation, that the investigating authorities conduct a "price effect" analysis that examines, at a minimum, whether the alleged subsidies to the raw material led to a decrease in the level of prices for that raw material paid by the processed product's producers below the level they would have had to pay for the raw material from other commercially available sources of supply. The Philippines also considered that Pork required two further relevant factors to be considered: (a) whether the raw material was internationally traded, since it is less likely that subsidies will cause the domestic price of such material to decline by the full amount of the subsidies if the producers of the raw material can export at international prices; and (b) the per unit cost of producing the additional output of the raw material that the subsidies may have caused, since the extent to which such additional output affects the price for the raw material will depend in part on the cost of producing the output.

121. The Philippines maintained that Brazil did not conduct an examination of any of the factors set forth in Pork, nor of any other factors showing a price effect of the alleged subsidies on desiccated coconut. In this regard, the Philippines noted that other relevant considerations could affect the international market. For instance, international market conditions could depress the price of the downstream product so as to in turn depress the price of the upstream product even in the absence of subsidies. The Philippines argued that an investigation of subsidies on downstream products would have to consider such factors in order to avoid attributing such price depression to subsidies on the upstream product in a case where international market conditions appeared to act in conjunction with the subsidies to affect the price of the upstream product in the exporting country.

122. The Philippines argued that, because it was improper for Brazil to have directly imputed subsidies on coconut fruit to desiccated coconut production, Brazil had no basis for resorting to a "constructed value" methodology to calculate the amount of the coconut fruit subsidy. The Philippines argued that, in accordance with Article VI:1(b)(ii) of GATT 1994, a constructed value methodology can sometimes be allowed in an anti-dumping context where a country "has a complete or substantially complete monopoly of its trade and where all domestic prices are fixed by the State". ⁴³ However, the Philippines maintained that this approach was wholly inapplicable in the investigation at hand. First, the methodology was envisioned for an anti-dumping investigation and made little sense in a countervailing duty investigation. Second, Brazil made no effort to establish the predicate for the method's application; Brazil did not examine, and made no findings on, whether the Philippines had a complete or substantially complete monopoly of the trade in desiccated coconut or whether all domestic prices are fixed by the State. In fact, the Philippines noted that the evidence was to the contrary, as the Philippine domestic price for desiccated coconut is governed by the international price for the product, and there was no government intervention in market pricing. Thus, the Philippines contended that Brazil's constructed value methodology had no legal basis in GATT 1994 and was invalid as a basis for imposing countervailing duties under Article VI of GATT 1994.

123. The Philippines also argued that the constructed value approach violated Article VI:3 of GATT 1994 because it lacked any mechanism for adjusting the purported subsidy amount to ensure its proportionality to the number and extent of the alleged upstream subsidies. Such a constructed value calculation would yield the same subsidy amount no matter how many of the programmes were considered to be subsidies, and regardless of the funding levels of such programmes. Ostensibly, the subsidy amount would remain unchanged even if only one out of the seven programme categories were deemed to be a subsidy, and even if the funding levels of all the programmes were much lower. In the Philippines' view, this methodological inflexibility rendered the constructed value approach inconsistent with Article VI:3 of GATT 1994, under which the amount of a subsidy must be determined rationally, not indiscriminately.

124. Brazil took the position that its calculation of the level of subsidization on desiccated coconut was fully consistent with its obligations under either Article VI:3 of GATT 1994 or the Tokyo Round SCM Code. Brazil noted that neither Article VI nor the Tokyo Round SCM Code contained any guidance on the method of calculation of the amount of the subsidy. Brazil contended that, as long as its approach reasonably calculated the subsidy bestowed on the exported product, it was consistent with Article VI:3. Brazil agreed with the report of the Pork Panel that a "price effects" test was a reasonable method of determining whether a subsidy to an upstream product was bestowed on the exported downstream product, observing in addition that there may be other reasonable methods but it was unnecessary to consider them in this case.

125. Brazil agreed that the situation in this case was similar to the situation in the Pork case. Brazil had determined that coconut fruit and desiccated coconut are separate products and that the coconut fruit is the raw material for the desiccated coconut. The Philippines had submitted information that showed that coconut fruit production and desiccated coconut production are two separate industries that operate at arm's length. Thus, the situation in this case comported with the requirements set out in Pork that where separate industries are operating at arm's length the downstream industry cannot be considered subsidized unless the subsidy bestowed on the coconut fruit (the upstream product) has had a price effect on the desiccated coconut (the downstream product). Brazil maintained that in this case, a price effects test was a reasonable method available for Brazil to measure the subsidies that were indirectly bestowed on desiccated coconut, since the Philippines did not provide much of the information Brazil had requested. Had the Philippines provided the information requested, Brazil asserted that it could have determined the programme-specific subsidies provided to coconuts and then conducted a more detailed analysis, based on respondent's information, to determine how much of the programme-specific subsidies on coconuts flowed downstream to the production of desiccated coconut.

126. Brazil asserted that its analysis was consistent with the approach advocated by the Pork Panel. Brazil asserted that in order to determine the price effect, it had used a constructed unsubsidized price for coconut fruit and calculated an unsubsidized price for desiccated coconut. Brazil then compared this price for unsubsidized desiccated coconut to a subsidized price for desiccated coconut, calculated using the price in the Philippines of the subsidized fruit. The difference between the subsidized and unsubsidized prices for desiccated coconut determined the price effect of the subsidies and measured the subsidies indirectly bestowed on desiccated coconut. Only after finding that there were subsidies to the upstream product, coconut fruit, that were passed through to the downstream producers in the form of reduced input prices, did Brazil look at whether these indirect subsidies benefitted desiccated coconut production. Brazil determined that they did and that the subsidy was bestowed on the imported product in the form of reduced prices. Brazil asserted that it had considered all relevant facts, within the constraints of an analysis based on the best information available.

127. Brazil argued that its construction of an unsubsidized price was reasonable and logical. Brazil asserted that the Philippines arguments concerning the use of "constructed value" in the dumping context were completely irrelevant to this dispute. Brazil maintained that it had not conducted a "constructed value" analysis, but rather had attempted to measure the price effect of the subsidies using the best information available in light of the Philippines' failure to provide necessary information.

128. The Philippines asserted that the Pork Panel treated consideration of other commercially available sources of supply as a means of determining whether factors other than the alleged upstream subsidies may have affected the price at which upstream producers sold the raw material to downstream producers. In the Philippines' view, unless such other factors were considered, there could be no reliable finding that the price of the raw material was determined solely or principally by the alleged upstream subsidies. The Philippines noted other relevant considerations pertaining to the conditions of international market demand and competition that, while not enumerated in Pork, could be considered. For example, international conditions could depress the price of the downstream product in a way that in turn also depressed the raw material's price in the exporting country even if there were no subsidies.

129. In this regard, the Philippines referred to the Panel decision in Canada - Countervailing duties on Grain Corn from the United States, SCM/140 (adopted 26 March 1992) ("Grain Corn"), BISD 39S/411, para. 5.2.7. The Philippines argued that, although Grain Corn involved prices in the importing country, it illustrated the possible effects of international conditions on domestic prices:

Grain Corn, para. 5.2.9. Just as it could affect domestic prices in an importing country, the Philippines argued that an international price decline could also depress domestic prices in an exporting country.

130. The Philippines also argued that, where alleged subsidies and international conditions appeared to be concurrent causes of low domestic prices in an exporting country, the investigating country bore the further burden of carefully analysing the international conditions so as to avoid attributing to alleged upstream subsidies any price depression, or degrees of any such price depression, actually caused by international conditions. The Philippines maintained that if the international conditions would, by themselves, have been sufficient to depress raw material prices, e.g., by depressing the prices of the downstream product derived therefrom, even in the absence of upstream subsidies, then the depression of the raw material price cannot be deemed a benefit "bestowed" by the upstream subsidies.

131. The Philippines maintained that Brazil's price effect analysis entailed an a priori disregard of all factors other than the alleged upstream subsidies. In other words, Brazil's comparison between the subsidized fruit price and the constructed unsubsidized fruit price yielded a price differential that Brazil simply treated as the price effect, uninfluenced by any other factors. Thus, in the Philippines' view, Brazil had ruled out by definition the relevance of other factors in its price effect analysis, contrary to the guidance of Pork.

132. Brazil asserted that the Pork Panel's use of the phrase "other commercially available sources of supply" indicated one possible means to assess what the price of the subsidized upstream product, in this case, the fruit, would have been in the absence of the subsidies, but that this was not necessarily the only means available. Brazil maintained that it had considered international trading in the product. However, because the five largest world suppliers of coconuts were under investigation for subsidies, Brazil did not consider the international price to reflect an unsubsidized price. Thus, that Philippine coconut producers could sell in the international market did not mean they would be able to sell at a higher, unsubsidized price. Moreover, Brazil argued that the Philippines had provided no information on coconut prices in countries that did not subsidize coconut fruit production, and that even if an international price existed, there might not be commercially viable access to the upstream product by the producer of the downstream product. Thus, Brazil had considered the constructed price as the most reliable unsubsidized coconut price for comparison purposes.

133. In Brazil's view, the phrase "other commercially available sources of supply" in the Pork Panel report was intended to allow flexibility to determine the appropriate analysis on a case-by-case basis. Brazil asserted that such flexibility was imperative in cases, such as this one, where there was no unsubsidized commercially available source for the upstream product. Brazil noted that there were a number of reasons not to consider international prices in this case. First, Brazil determined in its investigation that the five largest producers of coconuts were subsidizing their coconut production. ⁴⁴ The Philippines provided no information about coconut fruit prices in countries not found to subsidize coconut fruit production. Thus, if the major suppliers of coconut fruit received subsidies, the international price would show the effects of those subsidies. Brazil asserted that subsidies to a product in one country can affect the world market price for a product in at least three different possible ways. If subsidies are bestowed on a product by Country X are substantial and Country X is an exporter of that product, the subsidized prices can, and would be expected to, lower the world market price for that product. If subsidies are bestowed on a product by Country X, which accounts for a significant portion of the world production of that product, the subsidized prices can lower the world market price for that product. If subsidies are bestowed on a product by a number of countries whose production of the product accounts for a significant portion of the world production, the subsidized prices can lower the world market price for that product. Brazil argued that the third possibility was the case during the Brazilian investigation of desiccated coconut from the Philippines. Brazil also noted that an international price must be adjusted to include import duties and other expenses. ⁴⁵ Finally, although an international price may exist, there may not be commercially viable access to the international product because of import restrictions, unreliable suppliers, shipment delays, quality differences, or other reasons. Thus, in this case, Brazil had determined to rely, for comparison purposes, on a constructed price for the Philippines, based on cost information provided by the Philippines, as the most reliable unsubsidized coconut price.

134. Brazil also argued that, to determine the competitive benefit or price effect that the subsidization of the input product (coconut fruit) has on the price of the downstream product (desiccated coconut) the price of the subsidized input must be compared to the price of an unsubsidized input. In other words, the investigating country must determine what it would have cost a producer to purchase the input product in the absence of subsidies. Without such a benchmark, no determination of a price effect or competitive benefit can be made. The issue of subsidies was relevant to the issue of the use of an international price benchmark. A respondent would always argue that a lower subsidized benchmark price should be used for comparison purposes because that would lower any subsidy rate determined by the investigating country. On the other hand, the investigating country's producers would be hurt by the use of a lower subsidized benchmark price especially if the lower benchmark price erroneously resulted in a finding of no pass-through of the subsidies. Moreover, once the subsidies were eliminated, the international prices would rise to the unsubsidized price and this unsubsidized international price could once again be used as an unsubsidized benchmark.

135. The Philippines considered as ex post facto Brazil's assertion that it considered international trading, but did not consider the international price to reflect an unsubsidized price because the five largest world suppliers of coconuts were under investigation. The Philippines noted that Brazil did not cite to either the Ordinance or to DTIC Opinion 006/95 in proffering this explanation, and urged the Panel to disregard it. The Philippines also observed that, in theory, subsidies to a product in one country could conceivably affect the world market for that product, and the international price thereof. However, a subsidy determination could not be based on such a mere theoretical possibility. Rather, the investigating authorities must establish the existence of all factors that could possibly affect the world market for the product under investigation. The Philippines asserted that the Ordinance did not contain an assessment, let alone a determination, regarding the existence of such factors. Thus, Brazil's subsidy determination was flawed by its disregard of international pricing factors, and Brazil's belated rationalizations for such disregard must be rejected as both ex post facto and without merit.

136. Moreover, the Philippines contended that Brazil's explanation was without merit. The Philippines contended that the mere involvement of several of the largest suppliers of coconut products in the investigation did not have any automatic effect, either upward or downward, on the international price at which coconut producers could sell their product. In the Philippines' view, Brazil should have investigated and analyzed the actual price effects of other international trading factors, rather than a priori ruling out the possibility of any price effects. ⁴⁶ In addition, the Philippines argued that Brazil did not even assert that it considered the other relevant factor identified in Pork, the per unit cost of producing the additional output of raw material that the subsidies may have caused.

137. Brazil also maintained that, in a comparison of the actual price of coconuts to the constructed unsubsidized price of coconuts, the difference necessarily reflected the price effect of the subsidies on the raw material. Brazil argued that it did not presume a full pass-through of the subsidies on coconuts to desiccated coconut. The extent of the pass-through calculated consisted only of the difference between the subsidized and the constructed unsubsidized price of coconut fruit. To the extent that the actual subsidized price did not reflect the full amount of the subsidies bestowed on coconut production, those subsidies were not considered to be passed through to the coconut processors. ⁴⁷

138. The Philippines asserted that Brazil's determination was undercut by evidence presented by the Philippines showing that Philippine coconut farmers sell coconut fruit to producers of products other than desiccated coconut, such as coconut oil, copra, fatty chemicals and foodnuts, and that desiccated coconut represents only a small percentage of Philippine exports of coconut products, as coconut exports are dominated by coconut oil. The Philippines considered that the Pork Panel had rejected an imputed subsidy analysis even when the downstream product (pork) "constitute[d] the primary product" of the upstream product (swine), ⁴⁸ and asserted that there was even less legal or factual support for the use of such an analysis in this case, because desiccated coconut is not even the "primary product" of coconut fruit in the Philippines.

139. Likewise, the Philippines asserted that Brazil's determination was undermined by evidence showing that the export price of Philippine coconut products other than coconut oil, including desiccated coconut, and the Philippine domestic price of coconut fruit "husked nuts", the raw material for desiccated coconut, closely tracked the international price of coconut oil, which in turn followed the price trends in the worldwide market for oils and fats, where Philippine coconut oil has only a 5 per cent share. The Philippines argued that Brazil had wrongly ignored information on the dependence of coconut prices on world supply and demand for coconut oil. The Philippines maintained that this evidence, the "Coconut Industry Kit-Series of 1993", prepared by a coconut industry association before the initiation of Brazil's investigation, and the document prepared by the Philippines' National Economic Development Authority, a formal paper for the 1994 Proceedings of the World Conference on Lauric Oils, had been submitted during the investigation, and was not generated specifically in response to Brazil's questionnaires. Thus, Brazil had no basis for completely ignoring these documents. Moreover, the Philippines noted that these documents were not addressed by Brazil in either the Ordinance or DTIC Opinion 006/95, and thus Brazil's explanation for disregarding them was an ex post facto rationalization that the Panel should not consider.

140. The Philippines also argued that Brazil improperly disregarded the formal testimony of a high ranking Philippine government official during the 13 June 1995 meeting, at which he explained the relationship between the price of husked nuts and the international price of soybean oil. In the Philippines's view, even oral evidence, as long as subsequently reduced into writing, is admissible in countervailing investigations.

141. Brazil contended that the Philippines' claim that the subsidies could have no price effect because the price of coconuts is dependent on the price of coconut oil was not supported by evidence submitted in the course of the investigation. Therefore, Brazil had no basis to conclude that coconut prices were unaffected by the subsidies. In this regard, Brazil objected to the Philippines' submission to the Panel of information on the relationship between soybean and coconut oil prices ⁴⁹ that had not been submitted during the investigation.

142. The Philippines maintained that Brazil had set artificially high evidentiary hurdles on the international trade issue in an attempt to obscure the fact that it had failed to address that issue at all in its determination. In this regard, the Philippines asserted that the graphs and tables it submitted to the Panel were intended only to further exemplify the type of information that Brazil should have sought and examined as part of the "other factors" analysis required by Pork.

143. Brazil argued that it did consider the information provided by the Philippines about the price relationship between the various oils and desiccated coconut, but concluded that the information did not support the Philippines' claim. For example, Brazil asserted that the data in the "Coconut Industry Kit" referred to by the Philippines did not demonstrate a correlation in prices of coconut fruit and oil. Moreover, Brazil argued that some of the information was uncorroborated statements by the Philippine government, which Brazil was not obliged to rely upon. Brazil also maintained that it was not clear that information on the relationship between the price of coconuts and the price of coconut oil was relevant. Brazil posited that, if subsidies were necessary to keep the coconut producers in operation given the low price of coconut oil, this still meant that the coconut producers benefitted from the subsidies and that downstream purchasers benefitted from the lower prices that could be charged because of the subsidies, asserting that, if prices drop below a profitable level in the absence of subsidies, the supply of coconut will decline and the users of coconuts will have to pay increasingly higher prices for coconuts in the fact of dwindling supply. Moreover, Brazil considered the relevance of the fact that there are many competing uses for coconuts unclear. Brazil posited that it could mean that desiccated coconut processors did not have the market power to force a pass-through of the subsidy or some portion thereof to themselves alone. It did not mean that the interplay of all supply and demand factors did not result in at least some portion of the subsidy being passed through to all consumers of coconuts, including the desiccated coconut processors.

144. In the Philippines's view, Brazil had adopted the type of imputation analysis that was rejected by the Panel in Pork. In Pork, the issue was the propriety of the United States' imputation of upstream subsidies (whose actual value was unchallenged) to the downstream product using a conversion factor based on the ratio of the hog carcass to the weight of the live hog. In this case, the Philippines argued that Brazil constructed the value of the upstream subsidies, and then improperly imputed those subsidies to desiccated coconut using a conversion factor of 7.5 coconuts per kilogram of desiccated coconut. Although Brazil sought to distinguish its methodology from an imputation approach, claiming that by constructing the "unsubsidized" cost and price of coconut fruit it had measured the price effect of the subsidies on coconuts and limited the extent of the pass-through to the difference between the subsidized and the unsubsidized price of the coconuts, in the Philippines' view, this methodology still constituted the sort of upstream-downstream imputation rejected by the Panel in Pork, because Brazil failed to consider other factors, aside from the effect of upstream subsidies on the raw material's price, to complete the downstream price effect analysis.

145. Brazil considered that the Philippines appeared to allege that the "improper imputation methodology" in Pork was the use of a conversion factor. In Brazil's view, the conversion factor was not the problem in the Pork calculation, the problem was that the United States merely applied a subsidy to swine directly to pork production without determining whether that subsidy passed through. Brazil asserted that it was obvious that, if the subsidy is passed through to a downstream product, a conversion factor will have to be used to convert the upstream product into the downstream product. Brazil noted that presumably the Philippines intended to argue that the methodology found improper in Pork was the one step pass-through without analysis, which in Pork was done by taking the subsidy to swine and multiplying by the conversion factor to arrive at the subsidy for pork. However, Brazil maintained that it did not conduct such a one step pass-through without analysis. Brazil first considered the price difference between subsidized and unsubsidized coconuts to determine whether there was a price effect from the subsidies. Then, finding that there was a price effect, Brazil compared the costs/prices of desiccated coconut based on subsidized coconut costs with the costs/prices of desiccated coconut based on unsubsidized coconut costs to determine the effect of the subsidy on the exported product. To determine the cost of producing one kilogram of desiccated coconut, it was necessary to determine how much coconut goes into one kilogram of desiccated coconut, requiring a conversion factor.

146. The Philippines asserted that even if Brazil's "constructed value" approach were deemed appropriate, the Panel should rule that Brazil's calculation was erroneous and lacked adequate explanation. The Philippines described the steps followed by Brazil in its constructed value calculations as follows: (1) based on a 1993 Philippine "Cost Worksheet" for hybrid coconut fruit, Brazil calculated the annual per hectare production cost by assuming that each tree would be productive in the 8th year after planting, that annual costs would be fully amortized within 15 years from the 8th year at a 12 per cent annual rate, and that the same 12 per cent rate could be used to discount the annual cost, to arrive at a present value of US$ 821.41 per hectare; (2) while recognizing that the prevailing variety of coconut tree in the Philippines was tall trees, Brazil treated the per hectare production costs of hybrid and tall trees as equal, citing a Brazilian study showing a difference of less than three percent. Brazil divided the per hectare production cost of new hybrid coconut trees (US$ 821.41 per hectare) by the per hectare annual fruit yield of old tall coconut trees (3,910 per year), added an 8 per cent profit margin and US$ 0.015 for freight cost, to arrive at a constructed unsubsidized price of US$ 0.242 per fruit. (3) Brazil then applied a conversion ratio of 7.5 nuts per kilogram of desiccated coconut, and added an 8 per cent profit margin, to arrive at a constructed unsubsidized price for desiccated coconut of US$ 2.348 per kilogram. (4) Based on the "coconut fruit price effectively paid" by a Philippine exporter, US$ 0.051, and presumably also using at least the 7.5 nuts per kilogram conversion ratio, Brazil arrived at a subsidized price for desiccated coconut of US$ 0.800 per kilogram. (5) Brazil then deducted the subsidized price of desiccated coconut (US$ 0.800 per kilogram) from the unsubsidized price of desiccated coconut (US$ 2.348 per kilogram) to arrive at a subsidy amount of US$ 1.548 per kilogram. (6) Finally, Brazil divided the subsidy amount for desiccated coconut (US$ 1.548 per kilogram) by the "average weighted CIF export price" of desiccated coconut exported to Brazil (US$ 1.274 per kilogram), thereby deriving a 121.5 per cent countervailing duty.

147. Brazil asserted that there were several problems with the Philippines' description of Brazil's calculation. Brazil objected to the Philippines' insinuations that it had overstated the constructed fruit price. Brazil noted that the Ordinance described how Brazil calculated the price, and that DTIC Opinion 006/95 provided the exact figures and a more detailed description which showed the exact numbers involved to result in the US$ 2.348 figure. The processing cost that the Philippines questioned was provided by a Philippine exporter. Moreover, Brazil noted that in questioning the calculations of the price for the subsidized desiccated coconut, the Philippines failed to account for the processing costs used by Brazil.

148. In the Philippines' view, Brazil's calculation was defective in several respects. First, it was inherently inconsistent to use the actual production cost of allegedly subsidized hybrid coconut trees to derive an unsubsidized price. To the extent that the Brazilian countervailing duty purported to have been imposed on existing Philippine subsidies, and subsidies typically reduce production cost, then the actual production cost of hybrid coconut trees must have presumably been reduced by the alleged Philippine subsidies. By using the actual production cost to construct a price that is higher than an actual price paid by a Philippine exporter, the decision suggested some variant of a dumping charge - that is, that the Philippines exported desiccated coconut at a price that was lower than it would otherwise have been had the actual production cost of coconut fruit been passed on in the price to desiccated coconut producers, and exports of desiccated coconut.

149. The Philippines contended that Brazil had either (a) considered the actual production cost of hybrid trees unaffected by the alleged subsidies, in which event the countervailing investigation should have ended without finding any subsidy; or (b) considered the production cost reduced by the alleged subsidies, in which case it was, in the Philippines' view, incomprehensible how a subsidy-reduced production cost could be the basis for constructing an unsubsidized fruit price that was lower than the actual fruit price. The Philippines maintained that Brazil had resorted to an inappropriate anti-dumping methodology in this countervailing duty investigation, resulting in figures that were meaningless and could not support the imposition of countervailing duties.

150. Brazil asserted that it relied on the cost information submitted by the Philippines, which contained estimated costs. Brazil assumed those costs were estimated rather than actual because they reflected production costs absent the subsidies. Brazil noted that it could have chosen to rely on Brazilian prices, rather than seeking to construct a Philippine price, which would have resulted in a much higher rate of subsidy being found, since Brazilian coconut prices were higher than the cost information submitted by the Philippines.

151. The Philippines asserted that Brazil had presented no credible reason for considering that the Philippines' estimate of the cost for hybrid trees, relied on in the construction of the per hectare production cost of coconut fruit, represented unsubsidized costs. The Philippines had been asked for, and had provided, estimates about the actual cost and price of coconut fruit. Brazil's assertion that it could have relied on Brazilian coconut prices in constructing the cost of production of Philippine coconut fruit was unjustified, in the Philippines view, and was an ex post facto rationalization that was not mentioned in Brazil's determination. In any event, the Philippines argued that Brazil conceded that its production costs were unusually high by international standards, and those costs would therefore not have been an appropriate basis for calculating Philippine production costs.

TO CONTINUE WITH BRAZIL - MEASURES AFFECTING DESICCATED COCONUT

³⁵ L/1141 (adopted 27 May 1960), BISD 9S/194, para. 35. Brazil did not concede that any precedent from GATT 1947 was binding on the WTO. Nor, in Brazil's view, does such precedent fall within the scope of any source of material for treaty interpretation recognized by the Vienna Convention. Nevertheless, Brazil argued that in facing a new issue, past practice or interpretation of similar language might help guide considerations.

³⁶ In this regard, the Philippines noted Brazil's current legislation, Decree No. 1751 (19 December 1995), which, although not in effect during the investigation, recognizes the admissibility of secondary evidence, which is subject to Brazil's verification but can be disregarded only if found to be false or misleading.

³⁷ Salmon, para. 250.

³⁸ The Philippines argued that Brazil neither mentioned nor addressed the book submitted by the Philippines entitled, 20 Million Farmers are Victims of Levy Racket (1992), by retired Brigadier General Virgilio M. David, who was Philippine Deputy Military Supervisor of the Coconut Industry from 1974-77, and is currently the Administrator of the Philippine agency, the Philippine Coconut Authority, that principally oversees the industry.

³⁹ In this regard the Philippines requested the understanding of the Panel in regard to the Philippines' restrained discussion of factual and legal issues about the involvement of the family and close associates of the late President Marcos in the improper diversion of the levy funds. These issues were currently under litigation in Philippine courts, which are the appropriate fora for resolving those issues. Consequently, in this dispute the Philippines limited itself to pointing out Brazil's failure to give these matters due consideration; the Philippines did not seek the Panel's definitive resolution of issues pending before Philippine courts.

⁴⁰ Pork, paras. 4.6 and 4.8.

⁴¹ Pork, para. 4.8. The Panel noted that such "practices" were "reflected in Part I of the [Tokyo Round] SCM Code". Id. Thus, the Philippines argued that the pertinent principles articulated by Pork would apply to the instant matter even if it were considered under that Code.

⁴² Pork, paras. 4.9 and 4.10.

⁴³ Note 2 to Article VI:1(b)(ii) of GATT 1994. The Philippines noted that same limitation applied under the corresponding provision in GATT 1947, and Article 15:1 and Article 15:2(b) of the Tokyo Round SCM Code.

⁴⁴ Brazil stated that it was the practice of other countries conducting an upstream subsidy analysis to refuse to use potentially subsidized prices as a benchmark against which to compare the price in question to determine whether the subsidy was passed-through. See, e.g., Final Affirmative Countervailing Duty Determination on Steel Wheels From Brazil, 54 Federal Register 15523 (U.S. Department of Commerce 18 April 1989).

⁴⁵ Brazil argued that the fact that Brazilian coconut prices were above the price paid by the Philippine processors was one indication that the prices available to those processors were affected by subsidies.

⁴⁶ In this regard, the Philippines argued that, since Brazil's questionnaires did not request information concerning relevant facts about international trading, Brazil had no basis for relying on best information available.

⁴⁷ As an example, Brazil asserted that if the actual subsidized price of coconuts were 25 pesos, the calculated "unsubsidized price" were 40 pesos, and the subsidies equalled 20 pesos, Brazil would have considered only 15 pesos of the subsidy to be passed through to the Philippine coconut processors.

⁴⁸ Pork, paras. 2.8.a, 4.9, and 4.10.

⁴⁹ Attachments 1 and 2 to the first submission of the Philippines.