152. The Philippines argued that Brazil did not have an adequate basis for treating the per hectare production costs of hybrid and tall Philippine coconut trees as "equal". Moreover, even if the per hectare production cost of hybrid and tall trees were deemed to be equal, Brazil had no basis for treating the per fruit production cost as equal by disregarding the contrast in production costs of new and old trees, and the actual difference in the average annual fruit yields of the two types of trees. The Philippines contended that Brazil's calculation was distorted by the use of the per hectare production cost of new hybrid trees, while using the per hectare fruit yield of old or senile tall trees which had much lower fruit yields, involved practically no maintenance costs, and comprised the overwhelming majority of coconut trees in the Philippines. The Philippines further asserted that Brazil ignored the difference in fruit yields when it divided the per hectare production cost of hybrid trees by the lower per hectare fruit yield of tall trees (3,910), instead of dividing that per hectare cost by the higher per hectare fruit yield of hybrid trees (15,000), despite the fact that the latter yield was mentioned in the "Cost Worksheet" from which Brazil derived the actual production cost of hybrid trees. In the Philippines' view, this mixing of figures for new hybrid and old tall trees reflected an unfair manipulation of data. Thus, the calculation must be deemed inconsistent with Article VI:3 of GATT 1994, which requires a rational and reliable determination of the amount of an alleged subsidy.

153. Brazil responded to the Philippines' objections to the information and assumptions underlying the calculation by noting that, with respect to "Step 1" of the calculation, Brazil did, in fact, rely on information supplied by the Philippines, the "Cost Worksheet - Coconut Production; Cost of Planting and Maintenance per Hectare of Coconut" that described the stages, timing, and costs of producing coconuts. The assumption that the trees would be productive in their 8th year and would be productive over 15 years (for amortization purposes) was derived from information supplied by the petitioners and used as best information. Regarding the consideration of costs for hybrid trees when the bulk of the Philippine trees were tall trees, Brazil noted that the cost information provided by the Philippines contained estimated costs for one farm producing hybrid trees on one acre of land. Thus, the only information the Philippines supplied on costs was for hybrid, not tall, trees. Nevertheless, Brazil attempted to check whether such information reasonably reflected costs throughout the Philippines, by considering a comparison of costs done by the Brazilian Institute for Agricultural Research comparing the costs of the two types of trees. Based on that comparison, Brazil calculated a less than 3 per cent difference in costs between the two tree types. Regarding the consideration of production cost of new trees when most of the Philippine trees were old, Brazil asserted that it had again based its calculation on information supplied by the Philippines. Brazil calculated the average yield in the Philippines by dividing the total number of trees by total production as supplied by the Philippines in response to the supplemental questionnaire. Brazil noted that the Philippine government had not submitted any information on profit that could be used in the calculation, but one Philippine exporter had submitted information that indicated a profit significantly higher than 8 per cent. Therefore, in Brazil's view, the use of an 8 per cent profit figure was more favourable to the Philippines than other record evidence would have been.

154. The Philippines argued that Brazil had not identified any factor which would indicate that the Philippine and Brazilian crops were comparable, so as to justify treatment of the production costs of Philippine hybrid and tall trees as the same based on an alleged 3 per cent difference between the production costs of two Brazilian crops, which the Philippines asserted were in any event unspecified. Likewise, Brazil had failed to explain why its subsidy calculation divided the per hectare production cost of hybrid trees by the per hectare fruit yield of tall trees (i.e., 3,910) despite the fact that the substantially higher per hectare fruit yield of hybrid trees (i.e., 15,000) was stated in the very cost worksheet Brazil used to calculate per hectare production cost. This glaring error, by itself, in the Philippines' view rendered Brazil's subsidy calculation fundamentally indeterminate and thus unreliable.

155. Brazil maintained that it had used the average yield of all trees in the Philippines, not only the yield for tall trees, and moreover, that the yield figure was based on the information supplied by the Philippines. Brazil also noted that the Philippines never supplied any production cost information for tall trees.

156. The Philippines argued that it was misleading for Brazil to state that the Philippines supplied information only on costs for hybrid, not tall, trees. Brazil's questionnaires had inquired about production costs that included original planting expenses. Because only hybrid trees were being freshly planted in the Philippines, and all the replanting programmes examined by Brazil involved hybrid trees, the Philippines, in response to Brazil's questions, provided the production cost of hybrid trees. However, the Philippines explained that those types of costs were neither currently nor recently incurred by the tall trees, which were 40 years old or older. Thus, it was improper for Brazil to impute to extremely old tall trees the original replanting and other costs of new hybrid trees. Indeed, during the investigation the Philippines had stressed that tall trees should not be deemed benefited by replanting programmes that only involved hybrid trees. Moreover, because hybrid trees constituted only a tiny minority of Philippine coconut trees, it would be manifestly unreasonable to assume that Philippine exporters of desiccated coconut obtained their raw material solely or mostly from hybrid trees. The Philippines suggested that Brazil should have limited its subsidy calculation to the portion of the Philippine coconut tree population affected by alleged upstream subsidies -- hybrid trees. Since the only data Brazil had concerning a Philippine exporter showed that the exporter purchased only the output of tall coconut trees, Brazil had no basis for concluding that Philippine exports were the product of coconut fruit from subsidized hybrid trees.

157. The Philippines contended that, in order to impose a countervailing duty under Article VI of GATT 1994, Brazil must show through positive evidence that the alleged subsidization caused material injury to the relevant "domestic industry", which is in turn determined by the definition of the "like product". However, in the Philippines' view, the definition of the "like product" in Brazil's determination was ambiguous and contradictory. The Philippines asserted that the Ordinance variously defined the domestic "like product" to include: (1) all desiccated coconut, whether destined for the industrial market (where it is processed as a raw material) or for the retail market (where it is consumed as a finished product); (2) only desiccated coconut destined for industrial use; (3) only desiccated coconut destined for industrial use, and coconut fruit; and (4) all desiccated coconut destined for the industrial and retail markets, and coconut fruit. As a consequence of this failure to define the like product clearly and consistently, the Philippines argued that the scope of the relevant domestic industry varied throughout the Ordinance, specifically in the sections analysing apparent consumption, capacity and employment, industrial demand, and unit cost/price. By alternately narrowing and broadening the range of data to be considered, the shifting definition of the domestic industry made the evaluation of the alleged injury unclear.

158. In support of this contention, the Philippines noted that the Ordinance at one point referred to alleged injury suffered by coconut fruit producers who were supposedly unable to participate in a 19 per cent increase in industrial demand for desiccated coconut, and elsewhere stated that the coconut fruit producers themselves cut down the supply of fruit to desiccated coconut producers because an alternative market offered higher prices for the fruit. Thus, the Philippines questioned whether Brazil had considered coconut fruit production part of the domestic industry. In addition, the Philippines contended that the unclear definitions of the like product and the domestic industry undermined the reliability of the data and analysis regarding material injury and causation set forth in the Ordinance.

159. Brazil took issue with the Philippines' position that it had failed to make a clear determination of like product. ⁵⁰ Brazil asserted that it had clearly determined that the like product was domestically produced desiccated coconut, and that the like product was not subdivided based on the market in which desiccated coconut was sold. Brazil also maintained that it had clearly determined that the domestic industry consisted of the domestic producers of the like product, desiccated coconut, with the exclusion of one domestic producer, found to be a significant importer of the subject product. Brazil had included all other domestic producers in the definition of domestic industry, two of whom (the co-petitioners) accounted for on average 49 per cent of the national production (and 52 per cent of the production of the domestic industry) during the period of the injury investigation. Brazil noted that, in its consideration of injury, it used information from the co-petitioners, the only producers to respond to its questionnaires, where information for the entire domestic industry was not available.

160. Brazil asserted that Article VI does not refer to the concept of like product in the context of countervailing duties or injury, or specifically address the definition of domestic industry. Nevertheless, Brazil agreed that Article VI requires a determination of like product in order to define the domestic industry. Brazil suggested that guidance as to the meaning of the term like product in relation to Article VI of GATT 1947 can be found in the First Report of the Group of Experts:

In addition, Brazil referred to the Tokyo Round SCM Code, which defines the term "like product" to mean "a product which is identical, i.e., alike in all respects to the product under consideration or in the absence of such product, another product which although not alike in all respects, has characteristics closely resembling those of the product under consideration". Under both Article VI and the Tokyo Round SCM Code, therefore, Brazil argued that like product is defined on the basis of physical characteristics. In Brazil's view, the Philippines misread the Ordinance in claiming that Brazil had defined like product four different ways. The portion of the Ordinance cited by the Philippines to support this claim, Brazil argued, was the analysis of physical characteristics and the conditions of competition Brazil had relied upon in order to define the like product.

161. In examining the issue of like product, Brazil asserted that it had found that the domestic product was not alike in all respects to the imported product. There were differences in the amounts of sugar and fat contained in the products and in the size and dryness of the flakes between the two. Nevertheless, both were similar in that they were dried, grated coconut with physical characteristics that permitted overlapping uses. Therefore, Brazil determined that the product most "like" the imported product in terms of physical characteristics was the domestic desiccated coconut, but also concluded that competition between imported desiccated coconut and Brazilian coconut was a significant condition of the Brazilian market that had to be considered in the injury determination. Brazil noted that in its discussion of the like product in the Ordinance, item B, para. 4 was the only place in which it stated what product was "similar" or "like" the imported product. In Brazil's view, this confirmed its position that Item B discussed the analysis of like product in four paragraphs but only provided one definition of like product, and that definition of like product was based on physical characteristics as required by Article VI -- to the extent that Article VI could be interpreted as establishing a "like product" requirement.

162. The Philippines argued that the ambivalence of Brazil's like product determination was demonstrated by what it termed Brazil's apparent confusion about the effect of considering only desiccated coconut for retail sale in evaluating apparent consumption data that included non-competing desiccated coconut for retail sale. In the Philippines' view, the Ordinance states that this would make the share of imports more accentuated, while Brazil argued in its first submission that the share of imports was understated by this comparison. The Philippines argued that Brazil had also argued in its first submission that the like product included desiccated coconut for both the retail and non-retail markets, but continued to describe the data on apparent consumption as overstated due to the inclusion of desiccated coconut for retail sale. In the Philippines' view the inclusion of desiccated coconut for retail sale in the apparent consumption data would simply make the data fit the definition of like product if the like product included desiccated coconut for both retail and non-retail sale.

163. The Philippines referred to several factors, acknowledged in the Ordinance, which in its view undermined the finding of material injury. The Philippines noted that, after an 8 per cent decrease in 1991, the domestic price of Brazilian desiccated coconut increased by 23 per cent in 1992-93, and by 5 per cent in 1993-94. In addition, the operating profits of the Brazilian co-petitioners averaged 25 per cent for the 1991 to 1994 period. One co-petitioner's gross margin was constant from 1990 to 1993, indicating that its sales were not affected by price pressures and its operating margins were also reasonably stable. The other co-petitioner's operating margins were also stable. The production of the co-petitioners increased by 30 per cent from 1991-94, and one co-petitioner expanded its production capacity.

164. In the Philippines' view, such increases in production, stability in profits and operating margins, and substantial price increases belied the existence of any material injury, and instead indicated that the co-petitioners themselves had been thriving. Moreover, given that the co-petitioners were actually able to increase their production significantly from 1991-94, the Philippines considered it obvious that imports of desiccated coconut from the Philippines did not cause the co-petitioners to reduce their prior level of production. Rather, the Philippines considered that the co-petitioners were apparently complaining that the lower-priced imports allegedly limited their ability to increase their production at projected rates. However, the Ordinance did not identify any data supporting such projected increases, and did not even provide figures for the anticipated increase in production rates. As a result, there was nothing on record to show that the co-petitioners' growth and profit projections were reasonable and realistic, rather than merely speculative. The Philippines asserted that these growth projections seemed to be based on information unreliable on its face. The co-petitioners' data on installed capacity were aggregated for desiccated coconut and other products (such as coconut fruit and coconut milk). Consequently, it was possible that any allegedly unsatisfactory growth in production was attributable to those other products, not to desiccated coconut.

165. Brazil maintained that it looked at relevant factors that indicated that the domestic industry was injured. Brazil found that national production (including production by the importing company) fell by 45 per cent in the period 1989 to 1994, while production of the domestic industry (excluding production by the importing company) declined by 31 per cent during the period 1991 to 1994. There was a decrease in capacity utilization throughout the period. The level of employment in the industry declined 13 per cent between 1991 and 1994. Finally, the domestic industry's share of apparent consumption dropped from 63.9 per cent in 1991 to 37.7 per cent in 1994. Brazil also argued that the fact that the production of the two co-petitioners increased during that period did not indicate no injury to the domestic industry. Injury was to the industry as a whole not to individual producers. Therefore, the fact that one of the co-petitioners obtained a greater share of a declining national production did not indicate a lack of injury to the domestic industry as a whole.

166. Thus, Brazil argued, there was substantial evidence that the domestic industry was injured. Brazil noted that the requirement of an objective examination of positive evidence was found in Article 6:1 of the Tokyo Round SCM Code but not in Article VI of GATT 1994. Brazil argued that it nonetheless considered all relevant economic factors, some of which indicated injury, such as declining production, employment, and capacity utilization, others of which, such as operating results, did not. That not all indicators were negative did not detract from its finding. Brazil argued that previous Panel decisions recognized that a finding of injury may be based on an objective examination of positive evidence even where not all factors are negative. ⁵² In Brazil's view, the Philippines' arguments sought to have the Panel substitute its judgement for that of the investigating authorities as to which factors were more important indicators of injury in this case, which Brazil asserted was beyond the authority of the Panel.

167. The Philippines argued that it was misleading for Brazil to rely on the employment trends and capacity utilization data of the two co-petitioners alone as being those of the industry. More importantly, in the Philippines' view, Brazil did not explain how such data could be a reliable basis for drawing any conclusions about the desiccated coconut industry, when Brazil conceded in the Ordinance that the data could not be disaggregated by product. In the Philippines' view, the data were indeterminate, and therefore could not support conclusions about capacity utilization and employment levels in the desiccated coconut industry.

168. The Philippines considered that, even if the domestic industry were deemed to have suffered some material injury, Brazil had failed to show that such injury was caused by the alleged subsidies on desiccated coconut imports from the Philippines. The Philippines asserted that, pursuant to Article VI:6(a) of GATT 1994, an alleged injury must be "the effect of the ... subsid[y]" in order to justify any countervailing action, and Brazil was therefore required to analyze the allegedly subsidized imports' volume, price effect and impact on the domestic market. The Philippines argued that Brazil relied on data referring to different time periods for different aspects of its analysis, inconsistent data, and random presentation of prices, quantities and ratios, which left its analysis unclear and unsupported.

169. Brazil argued that the consideration of the volume, price effect, and impact on the domestic market, was a requirement of the Tokyo Round SCM Code, not Article VI, which did not contain any guidance on the elements of an injury determination. Nonetheless, Brazil asserted that it had properly considered those factors in making its determination.

170. Brazil also maintained that it did not rely on data for different periods in its evaluation of injury, as argued by the Philippines. Brazil acknowledged that data regarding longer periods, from 1989 through 1994, were also considered where available, but asserted that data for the period of investigation, 1991 through 1994, were considered in all cases. The only exception was where Brazil looked at the imports authorized in the first months of 1995, which Brazil asserted was a reasonable examination aimed at using the most up to date data possible.

171. The Philippines argued that Brazil's evaluation of the volume of imports was unclear. For instance, the Philippines noted that the Ordinance referred to an 89.93 per cent increase in imports of all "coconut products" from 1991-94, but did not indicate what part of that increase was desiccated coconut (as distinguished from other types of coconut products), and did not break down desiccated coconut imports by country, by month, or by year. In the Philippines' view, it was impossible to determine from this data the proportion of the total increase in imports accounted for by desiccated coconut, and the proportion accounted for by imports of desiccated coconut from the investigated countries, as distinguished from those not under investigation. While the Ordinance stated that the investigated countries accounted for 80 per cent of imports from 1989-94, in the Philippines' view, this percentage was meaningless because the Ordinance failed to establish the share of the total increase in coconut product imports accounted for by desiccated coconut imports. Moreover, the Philippines asserted that it was improper to use a 1989-94 time frame in calculating the investigated countries' share in desiccated coconut imports, and a 1991-94 time frame in calculating the increase of total coconut imports, because the difference in time periods precluded any accurate comparison.

172. Similarly, the Philippines contended, the Ordinance failed to distinguish desiccated coconut from other coconut products in the analysis of apparent consumption. The Philippines asserted that Brazil had determined that the like product was limited to desiccated coconut destined for the industrial market, excluding desiccated coconut for retail sale, but the Ordinance acknowledged that data on apparent consumption were overstated by the inclusion of desiccated coconut for retail sale, noting that it was not possible to obtain production data by market destination. In the Philippines' view, it was thus impossible to determine whether the like product, desiccated coconut for industrial sale, had an increasing or decreasing share in apparent consumption. As a result, even if import share in domestic consumption increased in absolute terms, it was not possible to determine whether import share also increased relative to the domestic like product. This indeterminate information on apparent consumption thus obscured any causal relationship between the share of imports and the share of the domestic like product in apparent consumption. Moreover, the Philippines noted that, if the like product excluded desiccated coconut for retail sale, then imports for retail sale should have been excluded from the import volume data. However, the Ordinance did not identify the proportion of total imports accounted for by imports for retail sale. On the other hand, if the like product included all desiccated coconut, regardless of the market in which it was sold, then the apparent consumption data were not overstated.

173. Brazil asserted that, in determining the effects of imports, it had cumulated the imports from the five countries under investigation, and found that imports of the subject product increased substantially during the period of investigation. Brazil asserted that DTIC Opinion 006/95 made clear that the only products considered in the analysis were desiccated coconut and coconut milk, the products subject to the investigation. Moreover, Table 10 in DTIC Opinion 006/95 showed both that cumulated imports of desiccated coconut, and the imports from the Philippines alone, increased significantly. The Ordinance also indicated, based on import licenses granted for the period January through May 1995, that imports were likely to increase. Thus, Brazil argued that the data showed both an increase during the period of investigation and in the imminent future of imports in absolute terms. Brazil asserted that it had found that the countries accounting for most of the increase in imports of desiccated coconut were the Philippines, the Ivory Coast, Sri Lanka, and Indonesia (all subject to investigation) which together accounted for 81.2 per cent of the total imports from 1989 to 1995. Brazil also argued that the information showed that imports increased significantly in relative terms. Brazil had found that apparent consumption overall increased 17 per cent in the period 1991 through 1994, while imports represented 35.8 per cent of the consumption in 1991 and 53.8 per cent in 1994. Thus, Brazil argued that the information supported its determination that the increase in subject imports in both absolute and relative terms was significant.

174. The Philippines argued that Brazil's consideration of price data was fundamentally flawed by its failure to examine actual import and domestic prices. Instead of considering actual import prices over a defined time period, the Philippines asserted that Brazil had relied on a single average price for imports from each country, calculated from CIF import prices, without reference to any time period. Referring to DTIC Opinion 004/95, the Philippines argued that there was no breakdown of prices by month and by year, that the only time frame mentioned was the period from May 1993 to April 1994, and that, with the exception of the data for Sri Lanka, the CIF prices cited in the Ordinance differed from those reported in DTIC Opinion 004/95.

175. The Philippines observed that Brazil had constructed domestic desiccated coconut prices based on the average coconut fruit price, a conversion ratio of fruit to processed product, a markup for additional processing costs, and a 15 per cent profit rate, virtually twice as much as the 8 per cent rate used to construct the unsubsidized Philippine price in the subsidy analysis, resulting in domestic prices that in the Philippines' view were artificially higher than import prices. The Philippines argued that there was no justification for the use of a constructed domestic price in lieu of actual prices. Because a countervailing duty could be validly imposed only if the investigated imports directly competed with the domestic product, the imposition of a duty implies that there was in fact such competition, in which case the actual selling prices of both the imports and the domestic product in the Brazilian market should have been available. Yet, such prices were not obtained, and the failure to do so was not explained. In the Philippines' view, it was not possible to determine any price trends or price comparisons between imported and domestic prices in the absence of actual prices.

176. The Philippines also asserted that the price and import volume data available actually indicated that there was no depression of domestic prices. In the Philippines' view, if lower priced imports had been affecting domestic prices, there would have been an inverse relationship between domestic price levels and import volume, as domestic producers would have had to cut prices to prevent their market share from being eroded by cheaper imports. However, on the contrary, the Philippines asserted that domestic prices and import quantities moved in the same general direction - both domestic prices and import volume decreased in 1991, and in each of the next two years. Thus, according to the Philippines, the data indicated that desiccated coconut imports did not depress domestic desiccated coconut prices.

177. Brazil argued that the information supported its conclusion that there had been price undercutting by the cumulated imports, varying from 70 per cent for the Philippines to 104 per cent for Indonesia. In order to determine the price undercutting, Brazil had relied on a constructed domestic price in making the comparison. Brazil argued that it had constructed the domestic prices because there were no domestic prices unaffected by the subsidized imports, because the volume of subsidized imports was such that the domestic price could not remain unaffected, and because the imported product was at a different stage of production from the Brazilian desiccated coconut. Brazil asserted that if the purpose of analysing the prices was to determine the effect the subsidies had on the domestic prices, it was reasonable to attempt to determine what the price would be absent those subsidies. Constructing a price based on the cost of production plus the normal profit, was, in Brazil's view, a reasonable approach to measuring that effect.

178. Therefore, Brazilian prices were not comparable without adjustment. Brazil argued that the Philippines' complaint that the margin of profit used to construct Brazilian prices was 15 per cent while an 8 per cent profit margin was used to construct Philippine prices was a false complaint. Brazil asserted that the 15 percent margin reflected commercial reality in Brazil. Moreover, since Brazil never compared the unsubsidized Philippine price calculated for the subsidy analysis with the constructed domestic price, the Philippine price that Brazil constructed for purposes of determining the level of subsidization had no effect on the injury determination. Brazil asserted that the Brazilian domestic price was then compared to the actual CIF prices of the imports, adjusted to reflect the same level of trade as the Brazilian prices. Brazil asserted that it calculated the import price by adjusting CIF prices from import documents for freight and other expenses associated with bringing the product to the Brazilian market. This resulted, in Brazil's view, in an appropriate comparison at the same level of trade.

179. Brazil asserted that it had also found that average domestic prices had decreased in the first two years of the period of the injury investigation, then increased in the later part of the period of the injury investigation. However, Brazil asserted that it determined that such increases were accounted for by the initiation of the investigation and the implementation of the Plan Real, a currency stabilization plan that led to significant price increases. Therefore, Brazil concluded that these increases did not indicate a lack of price effect by the imports. In addition, Brazil asserted that the price information had been accorded less significance in its determination than other factors.

180. Brazil asserted that, as indicated in the Ordinance, it had placed the most importance on the volume of imports, the decline in production, and the increasing share of apparent consumption accounted for by the imports, in reaching its determination that the subsidized imports caused injury. In Brazil's view, this analysis was consistent with the requirements of both Article VI of GATT 1994 and the Tokyo Round SCM Code. Brazil again noted that Article VI provided no guidance as to what factors to consider in determining injury, and asserted that therefore, a consideration of reasonable factors was sufficient to meet the requirements of Article VI. Brazil argued that its analysis considered all relevant economic factors in making the determination.

181. The Philippines asserted that Brazil's proffered explanations underlying its consideration of constructed prices in evaluating price effects should be disregarded by the Panel because they were ex post facto rationalizations. In addition, the Philippines asserted that Brazil's explanations were in any event flawed. First, by assuming that there were no Brazilian prices unaffected by the subsidized imports, Brazil began its price effect analysis having already assumed such a price effect. Second, Brazil's proffered explanation that prices had to be constructed because the domestic and imported products were at different stages of production, and were thus not comparable without adjustment, was without merit, given that Brazil's formula for constructing the domestic price contained no adjustment for differing stages of production. In addition, the Philippines argued that Brazil had failed to explain or justify its calculation of a single average import price per country. The Philippines had pointed out the difference between virtually all the CIF prices referred to in the Ordinance and those referred to in DTIC Opinion 004/95. Brazil's argument that DTIC Opinion 006/95 was the appropriate reference did not address this concern, since DTIC Opinion 006/95 referred to the same set of import documents as DTIC Opinion 004/95. However, in DTIC Opinion 006/95, different CIF prices were derived from those documents, without any explanation. Thus, in the Philippines' view, Brazil's calculation of import prices was highly suspect, and therefore the evaluation of price effects was not based on positive evidence.

182. Brazil noted that the data contained in DTIC Opinion 004/95 was not necessarily final information used as a basis for the determination in the investigation and differed from the data contained in DTIC Opinion 006/95. Brazil noted that, as stated in the final Ordinance, after the meeting with the Technical Consultative Counsel at which DTIC Opinion 004/95 was discussed, it was decided to collect additional injury information.

183. The Philippines also argued that even DTIC Opinion 006/95 does not state any determination concerning the cause of price increases, but merely notes that price increases "may have" been occasioned by the two factors referred to by Brazil. In the Philippines' view, this was insufficient to constitute a finding on the question. Moreover, Brazil's argument failed to explain how price increases in 1993-94 could be accounted for by the countervailing investigation initiated on 21 June 1994.

184. In the view of the Philippines, the finding of causation ultimately hinged on the mere juxtaposition of the following two factors: the decrease of the market share of the domestic producers of desiccated coconut, which in turn supposedly decreased domestic production due to the lesser demand, and the increase in import volume and in import participation in domestic consumption. However, the Philippines argued that there was no reliable evidence that desiccated coconut imports caused the difficulties allegedly experienced by the Brazilian desiccated coconut producers. Instead, the Philippines argued that the evidence on record indicated that the problems of the domestic producers were caused by other factors wholly peculiar to Brazil's coconut industry.

185. The Philippines argued that, contrary to the requirements of Article VI of GATT 1994, Brazil had ignored those other factors, and had merely listed them without assessing their impact or making any findings to ensure that their adverse effects on the domestic industry were not blamed on the allegedly subsidized imports. Among the factors Brazil assertedly disregarded were Brazil's unusually high domestic raw material and production costs. In this regard, the Philippines cited DTIC Opinion 004/95, referring to Brazil's unusually high production cost. The Philippines maintained that the co-petitioners had admitted that high costs and social benefits made Brazil's production cost higher than that of imported products. The Philippines asserted that, in light of the admissions of the co-petitioners regarding the significance of high domestic production costs, the failure to evaluate the extent to which those costs caused the alleged injury was improper. Moreover, the Philippines argued the cost of coconut fruit became prohibitive for producers of desiccated coconut when an alternative market, the "in natura" market, offered coconut fruit prices higher than desiccated coconut producers. Yet, the Philippines asserted that Brazil had failed to assess how these prohibitive raw material costs affected domestic prices during the period under investigation.

186. The Philippines asserted that the Ordinance listed several other non-import factors that adversely affected the production of coconut fruit, and as a result, the production of desiccated coconut. However, Brazil had not explained why any one or combination of those other factors could not have caused the alleged decrease in the production of coconut fruit and desiccated coconut. Among the other factors the Philippines asserted Brazil had admitted affected the coconut industry were: (1) a considerable number of farmers had opted to change from giant or hybrid trees to dwarf trees; (2) many farmers had converted their farms to non-coconut crops such as sugarcane, especially in Brazil's Northeastern region, where Brazil's coconut fruit production was most heavily concentrated; (3) all coconut farmers chose not to replant senile trees, whose percentage of all of Brazil's coconut trees was not identified; (4) farmers sold coconut lands to non-coconut industries that were widely expanding in Brazil's Northeastern region; (5) the high prices offered by the "in natura" market to fruit producers caused a decrease in the offer of fruit to the processors of desiccated coconut; and (6) a drought lasted for years during the period under review. In the Philippines' view, the drought alone could account for much of the decrease in desiccated coconut production due to the resulting fruit shortage, which in turn caused, rather than was caused by, a corresponding increase in imports to remedy that shortage.

187. Brazil argued that Article VI does not require an examination of other factors in determining whether there is a causal connection between the injury and the subsidized imports. Under Article VI, it was sufficient that the subsidized imports were found to cause injury. However, Article 6:4 of the Tokyo Round SCM Code does require the consideration of other factors, and Brazil asserted that it had examined other factors that might be causing injury but found that the imported product was nevertheless causing material injury to the domestic industry. Referring to the Panel report in Salmon, Brazil argued that past panel practice indicated that the consideration of other factors need not be extensive. Brazil found that, as stated during the investigation, given the importance of the raw material in the total cost of fabrication of the desiccated coconut and the effect of the subsidies on the imported desiccated coconut, the subsidized imports were by themselves the cause of the injury to the domestic industry.

188. The Philippines argued that Brazil erred in taking the position that Article VI does not require an examination of other factors in determining whether there is a causal connection between the injury and the subsidized imports. The Philippines questioned how investigating authorities could be sure that the injury was caused by subsidized imports, rather than by other factors, other than by considering the latter. Referring to the Panel report in Grain Corn, the Philippines argued that it was not enough for an investigating country to have "acknowledged the existence of factors other than subsidized imports having an effect" on the domestic industry, if that country "made no effort to ensure that the injuries caused by other factors were not attributed to the subsidized imports". ⁵³ The Philippines also took issue with Brazil's statement that the analysis of other factors need not be extensive, arguing that while the public summary of that analysis could be brief, the analysis itself should be extensive. Moreover, the Philippines asserted, the explanation of the analysis should be more extensive if the investigation showed that there were significant other factors present that could have caused injury to the domestic industry. A mere listing of the "other factors" was insufficient, in light of the Panel report in Grain Corn. For example, the Philippines asserted that the drought and the availability of an alternative market for coconut fruit, the "in natura" market, could very well account for all three elements which Brazil claimed supported its determination, the volume of imports, declines in domestic production, and the increasing share of apparent consumption accounted for by the imports. Indeed, information on the record indicated a correlation between import volume and drought-induced fruit shortages.

189. Even assuming that the investigated programmes constituted subsidies on the production of coconut fruit, the Philippines asserted that they were not countervailable because they fully complied with the developing country and de minimis exemptions under Article 6 of the Agreement on Agriculture.

190. The Philippines referred to Article 6.2 of the Agreement on Agriculture as recognizing that "government measures of assistance, whether direct or indirect, to encourage agricultural and rural development are an integral part of the development programmes of developing countries ...", and that

The Philippines maintained that, as evidenced in the information furnished to Brazil, the Philippine programmes, which in the Philippines' view involve investment subsidies, e.g, replanting, and input subsidies, e.g, fertilizers and affordable interest rates, complied with Article 6. These forms of assistance were generally available, to the extent funds permitted, to the agricultural sector, including the most disadvantaged sectors such as coconut farming. The Philippines argued that in recognition of the special needs of developing countries, these programmes are exempt from any reduction commitments in the Agreement on Agriculture.

191. The Philippines also asserted that the de minimis provision of Article 6.4 of the Agreement on Agriculture further exempts from reduction commitments even Philippine support measures which do not meet the criteria in Article 6.2. Article 6.4 of the Agreement on Agriculture provides:

The Philippines argued that Brazil failed to estimate the amount of the alleged subsidy and to show that the Philippines provided domestic support up to, or near, the corresponding de minimis amount.

192. Thus, the Philippines argued, since the Philippine programmes complied with Article 6 of the Agreement on Agriculture, Brazil could not impose countervailing measures on imports of Philippine desiccated coconut without demonstrating that such imports caused material injury to the relevant Brazilian industry. However, the Philippines maintained that Brazil had failed to demonstrate that the domestic industry suffered material injury, or that any injury that allegedly occurred could be attributed to Philippine desiccated coconut imports, and not to other causes. Thus, Brazil's countervailing duty on Philippine imports of desiccated coconuts was inconsistent with Brazil's obligations under the Agreement on Agriculture.

193. Brazil asserted that the limitations on actionability provided for in Article 13 of the Agreement on Agriculture did not apply in this case. First, the countervailing duties imposed by Brazil were not subject to the limitations of Article 13, which applies to countervailing duties covered both by Article VI of GATT 1994 and by Part V of the Agreement on Subsidies and Countervailing Measures. Brazil asserted that the Philippines had conceded that the SCM Agreement did not apply to this dispute by virtue of Article 32.3. Therefore, Brazil argued that the Agreement on Agriculture similarly did not apply, and the countervailing duties in this case were not subject to any of the limitations on actionability provided in Article 13.

194. Brazil also maintained that the Philippines' arguments that its subsidy programmes were not countervailable because they complied with developing country and de minimis exemptions under Article 6 of the Agreement on Agriculture were superfluous. Brazil asserted that the Philippines had recognized that the Agreement on Agriculture merely required that Brazil make a finding of material injury to the relevant Brazilian industry before imposing countervailing duties on these products. Brazil read the Philippines' argument to claim that its programmes were covered by Article 13(b)(i), which requires Brazil to make an injury finding. ⁵⁴ Thus even if the Philippines were correct in arguing that the Agreement on Agriculture is applicable, and that the Philippine programmes complied with that Agreement, Brazil had made the requisite injury finding, consistent with the requirements of the Agreement on Agriculture.

195. Brazil also contended that, even assuming, arguendo, that Article 13(b)(i) were applicable in this case, the Philippines had failed to demonstrate that its programmes met the requirements of that provision. Because Article 13 of the Agreement on Agriculture is an exception to the SCM Agreement, in Brazil's view the Philippines bore the burden of demonstrating that its programmes qualified for that exception. Brazil considered that the Philippines' argument, that its programmes fell within Article 13(b) because they provide domestic support within de minimis levels and in conformity with Article 6, was insufficient. In Brazil's view, the Philippines provided no specific evidence that its programmes complied with the requirements of Article 6.2. ⁵⁵ Brazil asserted that, lacking any specific factual support, the conclusory statements of the Philippines failed to meet the Philippines' burden of demonstrating compliance with Article 6.2. Moreover, the conclusory statements themselves failed to state the required conclusions: the Philippines did not state that the subsidies were generally available to "low-income or resource-poor producers," but that they were generally available to the agricultural sector, "including the most disadvantaged sectors such as coconut farming". ⁵⁶ In Brazil's view this made clear that the programmes were not limited to "low-income or resource-poor producers". Moreover, Brazil argued that the Philippines' statement that the programmes were generally available, whether to the agricultural sector or to low-income or resource-poor producers, was inaccurate. Most of the investigated programmes were available only to those growing coconuts, and others (e.g., the Agrarian Reform Programme) were available only for a limited number of products. Other low-income and resource-poor producers did not receive the subsidies in question. Thus, Brazil asserted that the subsidies did not fully conform to the requirements of Article 6.2.

196. Brazil also argued that the Philippines had failed to demonstrate that its subsidies were de minimis. Instead, the Philippines had simply asserted that Brazil had failed to show that the Philippine programmes exceeded de minimis levels. Brazil considered that the Philippines mistakenly assumed that the burden was on Brazil to demonstrate that the Philippines did not qualify for the exception. In Brazil's view, the Philippine argument highlighted the fundamental flaw of the position that the terms of the Agriculture Agreement applied to Brazilian activity that pre-dated the Agreement, by requiring Brazil to have applied the terms of an agreement that was not even in force at the time that Brazil initiated the investigation that resulted in the imposition of duties.

V. ARGUMENTS PRESENTED BY THIRD PARTIES

197. Canada argued that, notwithstanding Article 32.3 of the SCM Agreement, Articles I, II, and VI of GATT 1994 were applicable to existing countervailing measures, even if those measures were first enacted or imposed before the entry into force of the WTO. In Canada's view, there was nothing retroactive in requiring WTO Members to continue to maintain only those tariff and non-tariff measures that were consistent with their GATT 1994 obligations. Moreover, the inapplicability of the SCM Agreement to an investigation did not prevent the application of Article VI of GATT 1994 to existing countervailing measures imposed as a result of such investigations. In Canada's view, the possible applicability of the Tokyo Round SCM Code, and the potential inapplicability of the SCM Agreement, were not germane to the question of whether the Brazilian measure in question could be justified as a countervailing duty under Article VI.

198. Canada argued that the obligations of Articles I and II of GATT 1994 applied to all measures of a WTO Member, including measures that were initiated or imposed before the entry into force of the WTO for that Member. Accordingly, the exceptions to those Articles, in particular Article VI of GATT 1994, were equally available to a Member complained against to justify a measure, provided that it met the criteria set out in Article VI.

199. Canada noted that under the GATT 1947, Contracting Parties enjoyed the protection of Articles VI and XVI of GATT 1947, and the benefits of GATT 1947 dispute settlement, while the signatories to the Tokyo Round SCM Code had additional rights and obligations that they could enforce under the dispute settlement mechanism of that Code. Contracting Parties which were also signatories to the Tokyo Round SCM Code had the choice of proceeding under the Code or the GATT, and the choice of forum depended on the nature of the asserted inconsistency, the nature of applicable obligations, and the remedies available.

200. With the entry into force of the WTO Agreements, the various Agreements that have superseded the Tokyo Round Codes are integral to the WTO and apply to all Members, with the DSB as the single forum for enforcement of rights and obligations under any of the Agreements. However, the fact that a WTO Agreement, in this case the SCM Agreement, may not be applicable to the specific circumstances of a case did not preclude the application of GATT 1994, in this case Article VI, to the same circumstances. In Canada's view, the text of Article 32.2 of the SCM Agreement expressly limited the application of that Agreement only, and not the application of Article VI of GATT 1994. Moreover, there was nothing in GATT 1994 itself that limited its application to measures or practices imposed or enacted after the entry into force of the WTO for the Members concerned. Canada maintained that nothing in the SCM Agreement or GATT 1994 prevented the application of GATT 1994 obligations to existing measures. In Canada's view, the Panel has jurisdiction to examine all existing measures for consistency with obligations that have not been expressly excluded from its jurisdiction. The inapplicability of the SCM Agreement or the applicability of the Tokyo Round SCM Code should not preclude a panel from reviewing the existing measures of a Member for consistency with that Member's on-going GATT 1994 obligations. The fact that enactment or imposition of a measure antedated the entry into force of the WTO should not preclude review of the measure as currently applied in accordance with WTO norms due to concerns over retroactivity or the legal distinctiveness of GATT 1994 and GATT 1947.

201. Canada also argued that the existence of potential recourse to the Tokyo Round SCM Committee for dispute settlement under the Tokyo Round SCM Code was not dispositive of the issue - the availability of one forum and legal instrument did not foreclose the availability of other fora or instruments to WTO Members, and in particular did not foreclose reliance on basic GATT obligations by WTO Members. Similarly, the possibility of domestic review proceedings in Brazil was not dispositive of the issue. Nothing in the Tokyo Round SCM Code, the SCM Agreement, or GATT 1994 required the Philippines to challenge existing measures through domestic review mechanisms to the exclusion of the WTO dispute settlement process.

202. Canada also agreed that Brazil's failure to consult with the Philippines constituted a violation of Brazil's obligations under Article 4.3 of the DSU. In Canada's view, whatever Brazil's view of the appropriate terms of consultations, the refusal to consult without first establishing those terms constituted a refusal to consult in violation of Brazil's obligations. Entering into consultations where a request is made was a basic procedural obligation of every WTO Member. Brazil should have presented its substantive or procedural objections to the Philippines during consultations, and at the meetings of the DSB, and ultimately to a panel, rather than frustrating the consultative process.

203. On the question of the law applicable to this dispute, the EC took the position that the Philippines had brought its complaint to the wrong forum under the wrong law. The EC argued that the principle of non-retroactivity of treaty obligations, embodied in Article 28 of the Vienna Convention, precluded the application of Article VI of GATT 1994 to the Brazilian countervailing measures in question. This interpretation was, in the EC's view, consistent with the transitional provisions of Article 32.3 of the SCM Agreement, as well as the Tokyo Round SCM Committee's Decision on Consequences of Termination. In the EC's view, the principle of non-retroactivity required that the determination made by Brazil be challenged only under the provisions of the law in force between the parties at the time the application was filed, i.e., GATT 1947 and the Tokyo Round SCM Code. The EC noted that the Philippines continued to have recourse to dispute settlement under the Tokyo Round SCM Code, and in addition, if it considered the collection of countervailing duties after 1 January 1995 inconsistent with GATT 1994, it could seek a review under Brazilian law, which review would be conducted in accordance with the SCM Agreement and GATT 1994. Finally, the EC argued that Article 13 of the Agriculture Agreement applies only with respect to investigations covered by both Article VI of GATT 1994 and the SCM Agreement. Pursuant to Article 32.3 of the SCM Agreement, that Agreement did not apply to the Brazilian countervailing duty. Consequently, regardless of the Panel's decision whether Article VI applied to this dispute, Article 13 of the Agreement on Agriculture did not apply, and the Philippines' claim relating to that provision should be deemed inadmissible. The EC also supported the Brazilian request for a preliminary ruling on the question of the admissibility of the Philippines' claims, particularly since Brazil's objections covered all the claims raised by the Philippines.

204. The EC rejected the Philippines' argument that Article 30.3 of the Vienna Convention required the application of GATT 1994 to this dispute. The EC argued that there was no conflict between the Tokyo Round SCM Code and GATT 1994 with respect to the Brazilian investigation, since Article VI of GATT 1994 did not apply to that investigation. Consequently, there was no need to resort to Article 30.3 to determine the applicable law, as there was no conflict between successive treaties dealing with the same subject matter.

205. The EC supported Brazil's view that Article 28 of the Vienna Convention precluded the application of Article VI of GATT 1994 to this dispute, since the investigation was initiated pursuant to an application made prior to the entry into force of the WTO Agreements, including the SCM Agreement. In the EC's view, the operative act in this case was the application for countervailing duty, filed in 1994. The investigation, and the resulting duties, all flowed from that act, and must all be judged in light of the law in effect at the time of that act. The EC noted that the WTO Agreement did not contain any express provision requiring the retroactive application of Article VI of GATT 1994, nor was there any other basis on which such an intention could be established. To the contrary, the terms of the SCM Agreement, and the decisions of the Tokyo Round SCM Committee supported the conclusion that no retroactive application was intended.

206. The EC also noted that, unlike the Tokyo Round SCM Code and GATT 1947, which were distinct agreements with their own dispute settlement procedures and differing membership, Article VI of GATT 1994 and the SCM Agreement are part of the same agreement, the WTO Agreement, bind all Members, and are covered by the same dispute settlement arrangements. The EC asserted that the title and wording of Article 10 of the SCM Agreement evidence the intention of the parties that Article VI of GATT 1994 should apply only in conjunction with the more detailed rules contained in the SCM Agreement. Moreover, the EC asserted that this same intention is reflected in Article 32.1 of the SCM Agreement, and was further confirmed by the language of Article 13 of the Agreement on Agriculture. Article 13 of the Agreement on Agriculture exempted from the imposition of countervailing duties those countervailing duties covered by Article VI of GATT and Part V of the SCM Agreement. The EC argued that this indicated that the drafters of the WTO Agreement did not envisage the autonomous application of Article VI of GATT 1994. The EC maintained that, had such an eventuality been contemplated, it would have been anomalous not to extend the exemption to the imposition of countervailing duties which are covered only by Article VI of GATT 1994.

TO CONTINUE WITH BRAZIL - MEASURES AFFECTING DESICCATED COCONUT

⁵⁰ Brazil reiterated its position that the injury arguments, including the like product arguments raised by the Philippines, were beyond the scope of the Panel's terms of reference.

⁵¹ L/978 (adopted 13 May 1959), BISD 8S/145, para. 12.

⁵² In this regard, Brazil referred to the reports of the Panels in Cotton Yarn, para. 524, Audiocassettes, para. 422, and Salmon, para. 305.

⁵³ Grain Corn, para. 5.2.8.

⁵⁴ In Brazil's view, the Philippines was not arguing that the Philippine programmes were non-actionable under Article 13(a) Agreement on Agriculture, as meeting the requirements of Annex II of the Agreement on Agriculture, the so-called "green box".

⁵⁵ Brazil asserted that the Philippines had not provided this information during the investigation, either.

⁵⁶ Brazil noted that this argument contradicted the Philippines' claim that the subsidies to the coconut growers were funded by the coconut levy. If the programmes were generally available to the agricultural sector, there would be no possibility that a levy on the coconut producers entirely funded the programmes.