4.169 In this context, the Community submitted a study by Ames and Reiter. ⁷⁵ They found that an R² (the figure that roughly indicates how well the regression explains the dependent variable; a perfect match would mean an R² of 1) in excess of 0.5 could be obtained by selecting an economic time series and regressing it against two to six randomly selected time series. If one looks at the data of most regressions undertaken in the Japanese study, the R² is lower than that. Thus, most of these regressions do not produce better results than if shochu consumption had been regressed on production of cucumbers. Therefore, the Community argued, in naive time series analysis one can basically show anything. One can econometrically filter out a lot of these spurious relationships, but for every type of correction one looses a degree of freedom in the data. The more complex the analysis is, the more data points you need; 20 data points are hardly enough to correct for the following problems that appear simultaneously:

4.171 Canada argued further that the evidence that Japan cites does not refute Canada's allegation and demonstration that consumer choice between shochu and whisky is price responsive. A key pillar in Japan's defense of its claim that whisky and shochu are not directly competitive or substitutable products is a consumer survey purportedly canvassing consumer opinions regarding the alcoholic beverages participants would substitute if a particular beverage was not available. Far from demonstrating that whisky and shochu are not directly competitive or substitutable products, the survey shows that there is a sizeable percentage of Japanese consumers of shochu who would consume whisky if shochu was not available. The survey does not address the fundamental question of the impact of relative price differentials between shochu and whisky on consumer choice. Japan purported to show graphically that declines in sales for both shochu and formerly Second Grade whisky immediately following the imposition of tax rate changes in 1989 indicate that there is no cross-price elasticity between whisky and shochu and thus that the two beverages are not directly competitive or substitutable. The graph yields no such conclusion because a competitive relationship need not manifest itself as an instantaneous shift in cross-price elasticity. Indeed, when considered over the period 1989 through 1994, the graph demonstrates a marked transfer of market share from formerly Second Grade whisky to shochu, reflecting the substantial price differentials during this period of time between formerly Second Grade whisky and shochu. In fact, when the variable "price" is considered in assessing consumer preference between whisky and shochu, the results provide clear evidence that there is significant cross-price elasticity between whisky and shochu and thus that whisky and shochu are directly competitive or substitutable products.

4.172 In response to Japan's denial of a directly competitive or substitutable relationship between shochu and other distilled spirits (as well as of cross-price elasticity between shochu and other distilled spirits), the Community, Canada and the United States submitted the results of a study commissioned by the Liquor Committee of the European Business Community in Tokyo and carried out in February 1996 by ASI Market Research (an independent research institution). Unlike Japan's statistical analysis, the ASI study is not based on the historical analysis of the correlation between price and consumption trends, but on the contemporary reactions of a representative sample of shochu drinkers to a series of 36 different combinations of price levels for shochu and five brown spirits (Scotch Japanese whisky, cognac, Japanese brandy and North American whisky. The same exercise has been carried out with respect to three types of white spirits (gin, vodka and rum). The market research firm drew a sample of 400 drinkers (250 in Tokyo and 150 in Osaka). They selected drinkers aged 20 to 59 who had drunk shochu in the past three months. This sample was drawn for comparability with the sample in Japan’s study, and excluded persons associated with the alcoholic beverages industry. The survey-takers showed each person pictures of six types of alcoholic beverages and lists of prices for each, and asked which type of beverage the person would most like to buy given the prices specified. They then proposed a slightly different set of prices, and so on, working through three scenarios of relative prices and products. For the complainants, the ASI study confirms that there is a high degree of cross-price elasticity between shochu and brown spirits as well as between shochu and white spirits, as attested by the results of the following scenarios covered by the study:

4.173 Japan responded that there is a fundamental difference between the choices a real consumer would make and the choices a respondent of the survey is allowed to make. In the real world, consumers may choose beer or sake if the price of shochu increases. Indeed, the Institute for Social Studies found that 35 per cent and 30 per cent of respondents would choose beer and sake, respectively, if shochu were not available. As long as the demand shifts from shochu to beer or sake in response to a rise in the price of shochu, consumption of whisky could not be affected. However, respondents of the ASI survey were not allowed to choose beer, sake or wine. Their choice was confined within the alternatives of whisky, brandy and shochu, or of rum, vodka, gin and shochu. By eliminating more plausible alternatives, the survey forces consumers to consider these products as substitutes. Then Japan presented the example of a survey which would force respondents to choose between a hamburger and ice cream. One who chooses a hamburger if both cost 1 dollar may shift to ice cream if the price of the former rises to 1 dollar 50 cents. This does not prove that a higher price of hamburgers will lead to increased consumption of ice cream in reality. In the real world, consumers are likely to shift to such items as hot dogs, fried chickens, and fish and chips. According to Japan, the ASI's method deprives consumers of real choices and creates an artificial rivalry between products which may not be substitutable. It cannot determine whether or not the goods selected for a survey are competing with each other. At best, it only shows which one of the surveyed goods is relatively, and potentially, competitive with shochu. Moreover, in Japan's view, the Community admits that the differences between whisky/brandy and shochu are so large that the two cannot be regarded as "like products". On the other hand, the ASI report, after comparing the results of a brown spirits survey and a white spirits survey, concluded that "there is a suggestion that, in the white spirit segment, Shochu is not seen as so much of a competitor (i.e., substitutable product) in the eyes of the consumers". The ASI found rum, vodka, and gin are less of a competitor with shochu than whisky and brandy. This, in Japan's view, implies that even from the EC standpoints, rum, vodka and gin cannot be regarded as "like products" with shochu. Second, Japan submitted that the issue before the Panel is not the competition between domestic whisky and domestic shochu, but rather the competition between imported whisky and domestic shochu since Article III concerns itself with national treatment for imported products and not the treatment between two different products as such. The ASI report shows that the only product which will be substantially affected by a rise in the shochu price is domestic whisky, and that it is hard to find a discernible change in the consumption of imported whisky/brandy. The impact of a shochu price change on imported whisky/brandy is not discernible even though other major competitive products are eliminated from the survey. Consequently, Japan argued that it is reasonable to conclude that there is no directly competitive or substitutable relationship between imported whisky/brandy and domestic shochu.

4.174 For Canada, the results of the survey confirm the "macro-economic" evidence of changes in market share between shochu and whisky, reflecting relative price differentials between these two products. At the "micro-economic" level, the survey clearly shows that consumer preferences between whisky and shochu respond to the relative price differentials between the two alcoholic beverages. In response to Japan's argument that protective effects should be found if competitive conditions between domestic products and imports are distorted in favour of domestic production, Canada argued that the 1987 Panel Report set out factors that "were sufficient evidence of fiscal distortions of the competitive relationship between imported distilled liquors and domestic shochu" in favour of domestic shochu production: whether shochu and whisky are "mutually substitutable"; whether specific tax rates on shochu are considerably lower than on imported whisky; and whether shochu in Japan is almost exclusively supplied from domestic production. For Canada, the evidence it had submitted to this Panel more than meets these factors and accordingly shows that competitive conditions between domestic products and imports are distorted in favour of domestic production. Thus, applying Japan's own interpretation of the 1992 Malt Beverages panel report, the conclusion necessarily follows that the Liquor Tax Law, having a "protective effect", affords protection to Japanese production of shochu.

4.175 In support of its claim that there were no protective effects, Japan submitted that shochu is not inherently Japanese. Shochu does not have its origin in Japan. Its origin is believed to be in Southeast Asia. The product was not consumed by the Japanese until the 14th or the 15th century when it was first introduced from Southeast Asia. The oldest documented evidence of shochu consumption in Japan dates back only to 1559. Even today, consumption of this type of liquor is not an exclusively Japanese phenomenon; on the contrary, the product is widely produced and consumed in Southeast to East Asia. For example, Singapore and Malaysia produce samsoo or samsu, Korea is known for soju production, and some shaojiu in China falls under the shochu definition of the Law. These products have the following common three features: First, they use grains or potatoes as the base material, which is readily available at low cost in this part of the world. Second, they have a relatively low alcoholic strength. The maximum alcoholic content is about 35 per cent. Third, they are consumed directly after distillation. They do not normally undergo further post-distillation processing. These features constitute essential elements of the definition of shochu under the Liquor Tax Law. For Japan, the definition is nothing more than an attempt to correctly capture these characteristics. Moreover, as more advanced technologies for distillation or post-distillation processing (e.g., aging, purification, flavouring) became available, and as more expensive materials (e.g., grapes, malts) became affordable, this simple, low-cost liquor emerged as a distinct category. The Asian consumers of this product have recognized its identity, as illustrated in their government classifications of alcoholic beverages. The identity of the product has not been artificially created by Japanese law; it exists in many Asian markets and the largest manufacturer of shochu is not Japan but the Republic of Korea. Foreign production of this magnitude facilitates international trade in this product. In 1994, for example, Japan imported 11,244 kl of shochu. The record of the first half of 1995 was 7,465 kl, and the annual volume could be as high as 14,930 kl. These figures are far larger than the 1994 annual importation by Canada of rum, gin and vodka combined (7,471 kl). The share of imports in the Japanese shochu market is increasing. Despite the disadvantage of being an inexpensive, low-margin commodity, the share rose from 0.4 per cent in 1987 to 2.4 per cent in the first half of 1995. It is noteworthy that the share of imports in the US vodka and rum markets was nil until 1975, and grew gradually to 11.6 per cent for vodka and 7.0 per cent for rum by 1993. Were vodka or rum inherently American in 1975, asked Japan?

4.176 In response to Japan's argument that one of the reasons why the Liquor Tax Law does not have a protectionist effect is because shochu is not an "inherently Japanese product" but, rather, a liquor which is "widely produced and consumed in Southeast to East Asia", the Community responded that whether or not shochu is produced in other countries and, if so, in how many countries, is of no relevance. The only relevant fact is that imports of shochu represent an insignificant fraction of the total sales of shochu in Japan (1.7 per cent in 1994). In contrast, during the same year, imports from third countries accounted for 27 per cent of the total sales of whisky, 29 per cent of the total sales of brandy, 18 per cent of the total sales of spirits and 78 per cent of the total sales of authentic liqueurs. Moreover, sales of domestically produced shochu account for almost 80 per cent of the total sales of domestically produced distilled spirits and authentic liqueurs. Thus, it is indisputable that by giving a privileged tax treatment to shochu, Japan affords protection to its domestic production of distilled spirits. In the Community's view, Japan's argument is based on a misreading of the 1987 Panel Report. The 1987 Panel Report did not hold that the lack of production of the less taxed product in other countries was a necessary condition for a finding of protectionist effect. The 1987 Panel Report records Japan's argument to the effect that "a certain volume [of shochu] was imported from abroad including the EC countries". This fact was not disputed by the other parties. The Panel, therefore, was aware that shochu was produced in an indeterminate number of other countries. Moreover, whether the production of shochu in other countries was significant was an issue not raised by any of the parties to the dispute. In the Community's view, the Panel, therefore, had no grounds whatsoever to infer that most of the world production of shochu was confined to Japan. In view of these contextual elements, the statement in the 1987 Panel Report that "shochu was almost exclusively produced in Japan" can only be read as meaning that the shochu sold in Japan was almost exclusively produced in that country. The Community submitted that in any event, aside from learned etymological considerations, Japan provided very little evidence in the way of proving that shochu or something resembling shochu is in fact produced in any significant quantities in other Asian countries:

4.177 Japan responded that the Community (as well as Canada and the United States) seemed to contradict its initial arguments. For example, the Community initially stated "the following criteria may be relevant in order to determine whether a difference in taxation is ‘protective’ of the domestic production: ... whether the less taxed product is produced in other countries ... ". In its second submission, it argued "the Community notes that whether or not shochu is produced in other countries and, if so, in how many countries, is of no relevance". Arguing that the complaining parties were unanimous in their first written submission in referring to "production outside Japan" rather than to "the level of imports", Japan also submitted that Canada underpinned its complaint of protection by "the almost exclusive production in Japan of shochu" and that the United States argued "the creation of inherently domestic products" was the evidence of protective effect. In Japan's view, the Community rationalizes the Panel's recognition of shochu production outside Japan, and stated that the panel's finding to the effect "that shochu is almost exclusively produced in Japan" ought to be understood as a reference to the import ratio. Japan cited the relevant passage of the report: " ... the fact that shochu was almost exclusively produced in Japan and that the lower taxation of shochu did ‘afford protection to domestic production’ rather than to the production of a product produced in many countries (say, butter) in relation to another product (say, oleomargarine)". ⁷⁷ For Japan, it is obvious that the contrast the panel made between a product "produced in many countries" and a product "almost exclusively produced in Japan" relates not to the import ratio but to production abroad. Japan also referred the Panel to the paragraph 5.73 of the 1992 Malt Beverages panel report which, in Japan's view, did not review the import ratio but focused its analysis on production abroad. Therefore, for Japan, "protectiveness" ought to be judged not by the import ratio but by foreign production. Japan argued that the import ratio depends on the marketing strategy, the exchange rate and other factors, and the level of the ratio has nothing to do with whether or not the product is inherent in a particular country. The import ratio of shochu in the first half of 1995, for example, was 2.4 per cent in Japan. This figure is equal to the import ratio of vodka in the United States in 1980. The import ratio of beer into Canada was 2.9 per cent in 1994, while that of Japan in the same year was 4.2 per cent. The level of the import ratio of shochu in Japan is comparable to those of other international products in open markets.

4.178 Japan reiterated that shochu's origin is in Southeast Asia. The word "shochu" is derived from Chinese. Its largest producer is either the Republic of Korea or the People's Republic of China. There is nothing inherently Japanese in materials or production methods. Germany's korn falls under the shochu definition. Shochu is also produced in Singapore, Malaysia, Vietnam, the United States, Canada and France. Over 90 per cent of the materials used in the production of shochu A are imported, and more than half the rice (smashed rice) used to produce shochu B is imported. Shochu is indeed an international product. As for the identity of Chinese "shaojiu" with shochu, the fact is, there are two categories of shaojiu with high and low alcohol content, and, out of 371 shaojiu brands on display at China's fifth national shaojiu exhibition of 1989, 135 brands, or 36 per cent of the total, belonged to the low alcohol category. The volume of shaojiu production is about 8 times larger than that of Japan's shochu production; 36 per cent of the volume dwarfs Japan's shochu production by 3 to 1 and the UK Scotch production by 2 to 1. In response to the Community's doubt on the production quantity and the identity of samsu or samsoo of Malaysia and Singapore, Japan argued that the product is given an independent tax category, and thus, the production quantity must be sufficiently large to justify such categorization. On the issue of its identity, Japan noted that the product consists of "alcohol and water" according to the label of the samsu bottle. Judging from the contents as well as the low level of alcohol strength, Japan stated that it was confident that it falls under the shochu definition. Finally, the expression of "likeur" on the bottle of Malaysia's samsu means "liquor" and not "liqueur", according to its enquiry in Malaysia. Japan also stated that the liquor tax legislation of the Republic of Korea defines soju in two sub-categories of diluted soju, which is equivalent to shochu A, and "distilled soju, which is equivalent to shochu B, in a manner similar to Japan’s definition. Essentially, shochu and soju are identical products. Japan submitted that since the relationship between shochu and whisky has been discussed, reference should be made to domestic production of whisky as well. In Japan's view, the United States employs an evenhanded criterion of "whether the categorization creates an inherently domestic or foreign product". Moreover, the European Commission has argued before the ECJ, that, if a country domestically produces both of the two categories at issue, an origin-neutral taxation will not go against the national treatment obligation, and that the relationship between the two is only a matter of harmonization. ⁷⁸ Japan informed the Panel that it is the world's fifth largest manufacturer of whisky. In Japan's view, the Liquor Tax Law applies an origin-neutral tax to shochu which is produced in large quantities overseas, and to other distilled liquors which are domestically produced in large quantities. The Liquor Tax Law cannot have protective effect.

4.179 The United States recalled that imports of shochu were nonexistent in 1962 when the Liquor Tax Law was amended to insert the present categorization of distilled spirits. This categorization was designed to perpetuate the market situation of 1962 when all imports into Japan were subject to import quotas. Shochu could still be characterized as a Japanese product benefiting from discriminatory low tax rates. Indeed, the United States argued, when a system of tax classifications had been designed during such a period of absolute protection, the categorization of products based on their status during the quota period had to be re-evaluated after balance-of-payments quotas had been lifted.

4.180 Japan argued that according to the Japanese customs statistics, the import volume of Scotch whisky into Japan remained roughly equal from 23,473 kl in 1987 to 23,705 kl in 1994. In contrast, the import volume of Bourbon whisky grew 90 per cent from 5,998 kl in 1987 to 11,178 kl in 1994. Canadian whisky showed similar growth in the same period from 1,046 kl to 2,195 kl, recording a growth of 110 per cent. It is puzzling why Scotch whisky alone was not successful in exploiting such favourable conditions as the substantial decline in tax and the appreciation of the yen's value. This, in Japan's view, is not a case of failure but success in their pricing policy. Although the volume of imports remained constant, the value of the Scotch imports more than doubled from 82 million pounds sterling in 1987 to 199 million pounds sterling in 1994. Moreover, according to "The Scotch whisky Industry Review" by Alan S. Gray, the rate of gross margin in the price of Scotch whisky increased over the period from 47 per cent to 60 per cent. With such increases in revenue and in margin ratios, the amount of margin in the exportation of Scotch whisky to Japan must have grown more than threefold. The Scotch whisky industry did exploit the decline in the tax and the higher value of the yen, through a strategy emphasizing the value of imports as opposed to the volume, and successfully scooped gross profits three times the amount of 1987 out of the same volume of trade. In other words, all of the Scotch industry, the Bourbon industry and the Canadian whisky industry have been enormously successful in their trade with Japan since 1987.

4.181 Canada responded that the Superfund panel report and the 1987 Panel Report have firmly rejected the notion that the trade effects of a measure are relevant in determining the consistency of the measure with Article III:2. On the presumption of "an adverse impact", the Superfund panel report, the 1987 Panel Report and the 1992 Malt Beverages enunciated the principle that overall increases in market share for imports of the products in issue does not constitute a rebuttal. Indeed, the Superfund case articulates the principle that a finding of fiscal distortion in the competitive relationship between imported and domestic products constitutes an "irrefutable presumption" of nullification and impairment of benefits. Given the principle enunciated in the Superfund case, it is within the context of the fiscal distortion of the competitive relationship between imported and domestic alcoholic beverages that the Panel, and ultimately the DSB, must evaluate the consistency of the Liquor Tax Law with Article III:2, second sentence.

V. INTERIM REVIEW

5. 1 On 28 May 1996, Japan, United States and Canada requested the Panel to review, in accordance with Article 15.2 of the DSU, precise aspects of the interim report that had been issued to the parties on 20 May 1996; Japan and the United States requested the Panel to hold a meeting for that purpose. The Panel met with the parties on 6 June 1996 to hear their arguments concerning the interim report. The Panel carefully reviewed the arguments presented by the parties.

5.2 In approaching the interim review stage, the Panel drew guidance from Article 15.2 DSU which states that "a party may submit a written request for the panel to review precise aspects of the interim report prior to circulation of the final report to the Members". Whilst the Panel was willing to approach the interim review stage with the broadest possible interpretation of Article 15.2 DSU, it was of the view that the purpose of the review meeting is not to provide the parties with an opportunity to introduce new legal issues and evidence, or to enter into a debate with the Panel. In the view of the Panel, the purpose of the interim review stage is to consider specific and particular aspects of the interim report. Consequently, the Panel addressed the entire range of such arguments presented by the parties which it considered to be sufficiently specific and detailed.

5.3 The United States submitted to the Panel and the parties at the review meeting copies of press reports relating to the interim report. After a brief discussion on the need to maintain confidentiality, the Panel appealed and all parties to the dispute agreed, on the utmost importance of maintaining confidentiality so as to preserve the credibility and integrity of the dispute settlement process.

5.4 With respect to the legal status of adopted panel reports, the United States argued that nothing in GATT 1994 modified the status they had enjoyed under GATT 1947 and that they thus should not be considered as subsequent practice in the sense of Article 31 of the Vienna Convention on the Law of the Treaties (VCLT). The Panel drew the attention of the United States to paragraph 6.10 of the report. In order to clarify its position, the Panel introduced some drafting modifications in the panel report.

5.5 In respect of the discussion of Article III:2 in the interim report, both Japan and the United States argued that the Panel should not have rejected their approach according to which the benchmark to evaluate whether domestic legislation is in breach of the obligations contained in Article III:2 is the aim-and-effect test that they felt had its basis in the phrase "so as to afford protection" in Article III:1. The Panel took note of the arguments of the United States and Japan which had been the subject of detailed and serious considerations throughout its deliberations, but for the reasons spelled out in paragraph 6.11ff. the Panel decided not to take any further action in this respect.

5.6 Japan argued that with respect to Article III:2, the Panel’s overall approach would lead to findings of violations of Article III:2 by virtually all tax distinctions. The Panel could not subscribe to Japan’s position. The Panel reiterated that its task was circumscribed by its terms of reference which required it to review the consistency of the Japanese taxation system with respect to certain alcoholic beverages vis-à-vis Japan’s obligation under Article III:2. The Panel consequently limited its conclusions to the subject-matter circumscribed by its terms of reference.

5.7 In respect of the Panel’s discussion on "like products", Japan argued that under the Harmonized System (HS) nomenclature shochu and vodka no longer appear under the same heading. The Panel took note of the statement and, whilst not sharing the legal conclusion by Japan, the Panel proceeded to make certain drafting changes in paragraph 6.22 in order to clarify its point of view.

5.8 Japan argued that the tax/price ratio for domestic shochu A is higher than that for imported vodka, and, consequently, shochu A should be excluded from the Panel’s finding in paragraph 6.27 that Japan violated its obligation under Article III:2, first sentence. The Panel did not share this opinion but felt that it should further explain its position. The additional discussion by the Panel of this point is reflected in paragraph 6.25.

5.9 In respect of the distinction between "like products" on the one hand, and, "directly competitive or substitutable products" on the other, the United States argued that the Panel did not offer any clear distinguishing criteria between the two categories. In response, the Panel expanded its analysis of this distinction in paragraph 6.23.

5.10 The United States argued that the Panel did not offer any useful criteria concerning the interpretation of the term "dissimilar taxation" that the Panel uses in the report. More particularly, the United States argued that the Interpretative Note ad Article III, second paragraph, contained language that could be considered as a necessary condition in order to establish a violation of Article III:2, second sentence, but that it was questionable whether the same language could be considered as sufficient for the same purpose. The Panel added language in paragraph 6.33 to address this argument.

5.11 Japan argued that the complainants did not offer any evidence on liqueurs and that, consequently, liqueurs should be excluded from the findings in paragraph 6.33 that Japan violated its obligations under Article III:2, second sentence. The Panel was not persuaded by the arguments advanced by Japan but added language in paragraph 6.28 in order to clarify the Panel’s position.

5.12 The United States argued that the Panel’s analysis of the phrase "directly competitive or substitutable products" established a requirement to show adverse trade effects as a condition for establishing a violation of Article III. The Panel added language in paragraph 6.33 to make it clear that it follows the reasoning and the conclusions of previous panel reports on this issue and that the Panel felt that there is no need to examine trade effects in the context of Article III, since Article III deals with conditions of competition. A factual determination of whether two products are directly competitive or substitutable is a necessary precondition in order to apply the legal test of dissimilar taxation. In the Panel’s view, this determination takes place in the marketplace and does not mean at all that Article III has been made subject to an effects test.

5.13 Japan requested the Panel to suggest specific ways to bring its measures into compliance with its obligations under Article III:2. The Panel recalled its recommendation in paragraph 7.2, which is consistent with Article 19 DSU, that Japan bring its measures into compliance with the provisions of Article III:2.

5.14 With regard to some other issues raised by the United States, the Panel recalled that the only panel report that contains an analysis of "like products" similar to that of 1992 Malt Beverages is an unadopted panel report that had followed the same reasoning. The Panel also recalled its findings in paragraph 6.21 where it stated that a product’s description in a tariff binding is an "important criterion for confirming likeness" and that "this does not mean that the determination of whether products are ‘like’ should be based exclusively on the definition of products for tariff bindings".

5.15 Japan, the United States and Canada made a number of suggestions concerning language changes that the Panel accepted and introduced in its final report.

5.16 In respect of the interim report’s descriptive section, Japan suggested further changes which the Panel took into account in re-examining that part of the report. The Panel revised the descriptive section of the final report where it accepted the need for these changes.

VI. FINDINGS

6. 1 The Community requests the Panel to find that vodka, gin, (white) rum, genever and shochu are like products and that Japan, by taxing the other four products in excess of shochu violates Article III:2, first sentence. In the event that the Panel does not find the aforementioned products to be like products, the Community requests the Panel to find that they are directly competitive and substitutable products and that Japan, by taxing vodka, gin, (white) rum and genever higher than shochu has failed to observe its obligations under Article III:2, second sentence. The Community further requests the Panel to find that whisky, brandy, liqueurs and shochu are directly competitive and substitutable products and that Japan, by taxing the first three products higher than the latter, violates its obligations under Article III:2, second sentence.

6.2 Canada requests the Panel to find that whisky, brandy, other distilled alcoholic beverages, and liqueurs on the one hand, and shochu, on the other, are directly competitive and substitutable products and that Japan by taxing the former higher than the latter violates its obligations under Article III:2, second sentence.

6.3 The United States requests the Panel to find that white and brown spirits are like products in the sense of Article III:2, first sentence, and, therefore, that the difference in tax treatment by Japan between shochu and vodka, gin, rum and other white spirits, as well as whisky, brandy and other brown spirits, is inconsistent with Article III:2, first sentence. If the Panel is not able to make such a finding, the United States request, in the alternative, that the Panel find that all white spirits are like products in terms of Article III:2, first sentence, and that all distilled spirits are directly competitive and substitutable products in terms of Article III:2, second sentence. In the latter case, the United States requests the Panel to find that the difference in taxation by Japan under its Liquor Tax Law in favour of shochu materially alters the conditions of competition between shochu and other distilled spirits and that Japan thus violates its obligations under Article III:2, second sentence. The United States further claims that the small-volume producer exemption from excise taxes provided under Japan's Taxation Special Measures Law is limited to Japanese producers and that Japan thus fails to respect its obligations under Article III:2, first sentence.

6.4 Japan requests the Panel to find that its taxation system does not violate Article III. Japan claims that the purpose of the tax classification under the Liquor Tax Law is not to afford protection and does not have the effect of protecting domestic production. Japan further argues that spirits, whisky/brandy and liqueurs are not "like products" to either category of shochu, within the meaning of Article III:2, first sentence, nor are they "directly competitive or substitutable products" to shochu, within the meaning of Article III:2, second sentence. Finally, Japan requests the Panel to reject the claim by the United States with respect to its Taxation Special Measures Law because it lies outside the terms of reference of the Panel.

6.5 The Panel first turned to the United States' claim with respect to the Japanese Taxation Special Measures Law. The Panel noted that Japan argued that the claim of the United States is not part of the terms of reference of the Panel. The Panel further noted that its terms of reference, following from Articles 7 and 11 DSU, are circumscribed in WT/DS8/6, WT/DS10/6 and WT/DS11/3. The Panel noted that no mention of the Japanese Taxation Special Measures Law is included in WT/DS8/6, WT/DS10/6 and WT/DS11/3. The Panel concluded that its terms of reference do not permit it to entertain the claim of the United States with respect to the Japanese Taxation Special Measures Law and it proceeded, therefore, to examine the other claims.

6.6 The Panel noted that the complainants are essentially claiming that the Japanese Liquor Tax Law is inconsistent with GATT Article III:2 (hereinafter "Article III:2"). Article III:2 reads:

6.7 The Panel understood the dispute among the parties over the appropriate legal analysis to be applied in this case required it to interpret the wording of Article III:2. The Panel recalled that Article 3:2 DSU states:

TO CONTINUE WITH JAPAN - TAXES ON ALCOHOLIC BEVERAGES

⁷⁵ Ames, E. and S. Reiter, "Distributions of Correlation Coefficients in Economic Times Series", Journal of the American Statistical Association, 56, 1961, pp. 637-56.

⁷⁶ As mentioned before, in Japan's view, the Community's criticism of the Japanese study applies equally to the Community's study. Last September, the Commission submitted a report on excise tax to the Council and the European Parliament and convened a conference in Lisbon last November to discuss the report. In the report, the Commission refers to the Bossard study and states that "[g]iven the importance attached in the Directive to this particular issue [of competition between categories of alcoholic beverages], the Commission engaged an independent firm of consultants to assess the extent of such competition" (COM (95) 285 final, 13.09.1995). Did the Commission submit a defective report to the Council and the European Parliament, asked Japan? Did the Commission convene a conference of manufacturers and consumers to discuss an incomplete analysis?

⁷⁷ 1987 Panel Report, para. 5.11.

⁷⁸ The European Commission v. The United Kingdom ("Wine and Beer"), Case 170/78, 12 July 1983, [1983] ECR p. 2265.

⁷⁹ The asterisk in Article III:2 refers to the Interpretative Note ad Article III, Paragraph 2 that is quoted infra.

⁸⁰ The asterisk in Article III:1 refers to the Interpretative Note ad Article III, Paragraph 1 that is not quoted because it refers to an unrelated issue.

⁸¹ See, for example, the panel report on "Japan - Customs Duties, Taxes and Labelling Practices on Imported Wines and Alcoholic Beverages", adopted on 10 November 1987, BISD 34S/83 (hereinafter "the 1987 Panel Report"); see also the panel report on "EC - Imposition of Anti-dumping Duties on Imports of Cotton Yarn From Brazil", ADP/137, adopted on 30 October 1995, paras. 540ff.; see also the Appellate Body report on "United States - Standards for Reformulated and Conventional Gasoline", WT/DS2/AB/R, adopted on 20 May 1996.