(c) Channels of Distribution

29. While recognizing that channels of distribution are revealing for a market structure, the Panel wrongly dismissed Korea's distinctions regarding on-premise consumption, and thereby erred in its assessment of the evidence.

30. The essence of Korea's argument was that most of the volume of diluted soju and of western-style drinks was sold and consumed in different types of outlets. This was borne out by the Nielsen Study which clearly shows that, except in the case of Japanese restaurants and café/western-style restaurants, there was no overlap for on-premise consumption. Before the Panel, the United States responded to this by noting that that their embassy personnel knew of nine "traditional Korean-style restaurants" in Seoul serving both whisky and soju. Korea argues that the Panel should not have dismissed Korea's evidence about differences in places of consumption on the basis of evidence concerning nine restaurants, provided by the United States' embassy personnel.

31. The Panel also applied "double standards" to the evidence Korea and the complainants supplied on this issue. While Korea presented a market survey covering 320 restaurants that showed that there are different channels of distribution for the drinks in dispute, the Panel accepted the anecdotal evidence produced by the United States about only nine Korean restaurants.

32. Korea considers that the large, undisputed price differences between diluted soju and the imported beverages are key elements of evidence, and that the larger the price difference between two products, the less influence a change in the price of one will have on the demand for the other. In the present case, there is no price overlap between diluted soju, including its premium version, and any of the western-style drinks.

33. Korea believes that the only evidence on consumer responsiveness to changes in prices which was submitted by the complainants was the Dodwell Study. But Korea raised "fundamental objections" about the Dodwell Study before the Panel, and the Study is so flawed that it should have been rejected. The Panel erred in failing to recognize the weaknesses in the Study. Korea notes that the Panel considered the Dodwell Study "helpful evidence" ³⁴ sufficient to raise a presumption of a directly competitive or substitutable relationship, and rejected the "hard evidence" Korea had submitted in rebuttal. The Panel, therefore, wrongly allocated the burden of proof and also applied a "double standard" since it was lenient with the complainants' evidence, but strict with Korea's rebuttal evidence.

34. Korea contends that the evidence of the large price differences between diluted soju and most of the imported beverages is sufficient to rebut the complainants' claims about the existence of a directly competitive or substitutable relationship between the imported and domestic beverages. However, the Panel essentially disregarded the evidence and did not address Korea's argument that the absolute price differences were so great that behavioural changes were unlikely.

35. When stating that premium diluted soju was a "fast growing category" ³⁵, the Panel neglected Korea's evidence. Korea had emphasized during the second meeting with the Panel that premium soju production was declining, apparently as a result of Korean consumers' unwillingness to pay more for an up-market version of diluted soju.

36. According to the Panel, cognac is a directly competitive or substitutable product for standard diluted soju even though, before any tax is applied, the products differ in price by a factor of 20.³⁶ Admittedly, for some of the western-style beverages, the price differential from soju is smaller, and even negative (e.g. distilled soju as compared to standard whisky). However, the Panel did not distinguish between types of product and concluded broadly that "the price differences [were] not so large as to refute the other evidence".³⁷ Korea believes that in the case of consumer products, to say that an actual price difference of a factor of 10 or 20 is insufficient to refute hypothetical evidence on competition, such as the Dodwell Study, flies in the face of common sense and shows that the Panel wrongly applied Article III:2.

37. Korea observes that, although virtually no evidence on tequila was submitted by either the complainants or the third party, the Panel found that Korea had violated Article III:2 with respect to this beverage. The United States identified tequila as one of the products covered by the measures at issue, and tequila was included in the Dodwell Study presented by the European Communities. Mexico also made certain descriptive comments concerning the physical characteristics, tariff classification and patterns of consumption of tequila and mescal. The Panel included tequila in its examination because evidence was presented with respect to it ³⁸, although it excluded mescal which "was mentioned without positive evidence" being provided.³⁹ The only additional elements concerning tequila were statements made by the complainants that tequila is drunk with spicy food in Mexico, that tequila is becoming popular in Japan⁴⁰ and that tequila was included in the Dodwell Study. Korea considers the evidence on tequila to be insufficient to give rise to a presumption of a directly competitive or substitutable relationship with soju.

38. The Panel made no attempt to analyze what the Dodwell Study actually said. The Dodwell Study shows that consumers responded inconsistently to a possible price change for tequila. In fact, it even appears from the Dodwell Study that demand for tequila may not change if its price were lowered. The Panel, nonetheless, concluded that there was evidence that consumers were sensitive to relative price changes of soju and tequila.

39. Korea submits that, contrary to Article 11 of the DSU, the Panel failed to apply the standard of review appropriate to an Article III:2 dispute. Korea maintains that, in this case, the Panel simply did not have sufficient evidence to enable it to conduct an "objective assessment" and, instead, relied on speculation. The Panel also failed to accord due deference to Korea's description of its own market. Korea believes that, when faced with conflicting descriptions of a foreign market, a panel should be very careful in making assertions about what this market is like and should certainly not engage in speculation about its possible future development. Where there was disagreement between the parties about the Korean market, the Panel should have accepted Korea's description, unless the complainants brought compelling evidence to the contrary.

40. Despite its "strong misgivings" about the Panel Report, Korea states that it does not assert that the Panel acted in bad faith. However, Korea believes that the matters it has raised under Article 11 of the DSU are, nonetheless, serious enough to merit reversal of the Panel's conclusions.

41. Finally, Korea claims that the Panel failed to fulfil its obligation under Article 12.7 of the DSU. Korea considers much of the Panel's reasoning to be obscure, making it very difficult to determine the evidence the Panel relied upon in reaching its conclusions and the weight it gave to different evidence and arguments. In addition, the Panel Report is also "unacceptably vague". The Panel relies upon open-ended concepts, such as "potential" competition in the "near term", "potential" end-uses and the "nature" of competition, to support its conclusions which can be stretched to cover any outcome. Furthermore, certain evidence, such as the Sofres Study, was simply ignored without the Panel giving reasons therefor. The inadequate reasoning, in Korea's view, also prevented the Panel from making an objective assessment under Article 11 of the DSU.

42. The European Communities submits that Korea's appeal is grounded on the erroneous premise that the term "directly competitive or substitutable" must be interpreted "strictly". That proposition finds no support in the GATT, in its drafting history or in previous panel reports. As noted by the Panel, the drafting history of Article III:2 suggests that the drafters had in mind a rather broad notion of "directly competitive or substitutable" products, that could include apples and oranges.⁴¹ Furthermore, the Korean argument that Article III:2, second sentence, must be interpreted "strictly" is equally applicable to virtually any GATT provision.

43. The European Communities believes that the Panel's finding that there is "present direct competition" between the imported beverages and soju⁴² would be sufficient to conclude that those products are "directly competitive or substitutable". The additional finding of "a strong potentially direct competitive relationship" provides further support for that conclusion but is not indispensable.

44. In any event, the Panel's analysis of the evidence of "potential" competition is consistent with the wording of Article III:2, its object and purpose, as well as previous Appellate Body and panel reports. Korea relies on the fact that neither Article III:2 nor the Ad Article mention "potential" competition. But nor do they mention "actual" competition. In the European Communities' view, potential competition is "competition", both in the ordinary economic sense and within the meaning of the Ad Article. The use in the Ad Article of the words "competitive" (rather than "competing") and "substitutable" (instead of "substitute") is a further indication that the drafters envisaged the application of Article III:2 in the case of both "actual" and "potential" competition. The French and Spanish texts also support this view.

45. To the European Communities, the relevance of potential competition flows directly from the fact that Article III does not protect export volumes but expectations of an equal competitive relationship. The prohibition against protective taxation applies even if there are no imports of "directly competitive and substitutable" products. Korea's insistence on the existence of actual competition is, therefore, inconsistent with the proper interpretation of Article III:2. Korea's "but for" test is an admission that potential competition is relevant in some circumstances, but that test is too restrictive and finds no support in Article III:2, second sentence.

46. A determination of whether two products are "directly competitive or substitutable" must be made on a case-by-case basis and in respect of the market of the Member applying the contested tax measures. Nevertheless, other markets may provide a strong indication of the nature of the competitive relationship between the products in the market at issue. This may be particularly true in cases where there is either very little or no actual competition on the market at issue. In the present case, the Panel made very limited use of evidence drawn from third country markets. The Panel looked at evidence from the Japanese market to corroborate findings made concerning the Korean market. Although the Panel could have looked at other markets in addition to the Japanese market, that was not necessary.

47. The European Communities contends that Korea's interpretation of Article III:2 diminishes the rights of Members. Furthermore, the principle in dubio mitius is a supplementary method of interpretation that applies only where there is a genuine ambiguity. That is not the case here. Finally, the Panel's interpretation promotes "predictability".

48. The European Communities views the Panel's decision on how to group the products as a methodological one made for analytical purposes only. It does not involve any interpretation of Article III:2 and does not, therefore, raise any "question of law" which could form the subject of an appeal, unless the Panel failed to make an "objective assessment" of the facts.

49. Contrary to Korea's assertions, the Panel did not find that distilled soju and diluted soju were directly substitutable and competitive products. The Panel held that if diluted soju were found to be directly competitive or substitutable with imported spirits, it would follow necessarily that distilled soju, which is more similar to imported spirits, would also be directly competitive or substitutable with those spirits.⁴³ Korea has not challenged the premise underlying the Panel's reasoning.

50. The European Communities contends that, in deciding to consider together all imported beverages, the Panel did not anticipate the outcome of the case nor did it find that the imported products were directly competitive or substitutable inter se. Korea has not shown that applying a different analytical approach would have led to a different result. There is no significant difference between Korea's strict product-by-product approach and the Panel's method. In practice, the Panel switched to a product-by-product approach whenever there were differences between the imported spirits in respect of a particular criterion.

51. According to the European Communities, the Panel's finding that Korea's measures are applied "so as to afford protection to domestic production", is based on three factors: the sheer magnitude of the tax differential, the lack of rationality of the product categorization, and the fact that there were virtually no imports.⁴⁴ There is no indication in the Panel Report that the Panel considered the second of these three factors to be particularly important.

52. Korea has not explained why it was necessary to add a series of exceptions to the definition of soju which resulted in the most important categories of imported spirits being placed in a much higher tax bracket than soju. The reasons why there are no imports of soju are irrelevant. What matters is that, in practice, imports of soju are and always have been negligible.

53. The European Communities considers the Korean argument that the measures do not appreciably change the competitive opportunities of the imported products to be factually wrong. In any event, comparing pre-tax price-differences is not sufficient to take account of all possible price distortions caused by the measures.⁴⁵ Furthermore, prices may be affected by extraneous factors, such as fluctuations in exchange rates.⁴⁶

54. The European Communities argues that the "so as to afford protection" requirement is concerned exclusively with whether the contested measures protect domestic production and not with how much protection is afforded. If two products are directly competitive or substitutable, then any tax differential which is more than de minimis may affect the competitive relationship between the products and, as a result, "protect" the less taxed product. The only remaining issue is whether protecting the less taxed product favours "domestic production".

55. The European Communities asserts that Korea's claims under this heading do not raise any "question of law", but only factual issues which, in principle, are not subject to appellate review. These claims can only be considered by the Appellate Body under Article 11 of the DSU. However, an appellant invoking this ground of appeal must show that the Panel abused its discretion in a manner which attains a "certain level of gravity".⁴⁷ The European Communities contends that the Panel did not make the errors Korea alleges. However, even if Korea could demonstrate that the Panel committed those errors, they would not come close to constituting "egregious errors that call into question the good faith of the Panel".⁴⁸

56. According to the European Communities, Korea's argument that flavour is one of the consumer's primary considerations when choosing a beverage is flawed. If two products are nearly identical, the consumer's choice between them will necessarily turn on very minor differences. For instance, the only reason for choosing a green necktie instead of a red necktie is the colour. Yet, colour remains a relatively minor feature of neckties and differences in colour do not prevent neckties from being "directly competitive and substitutable".

57. Korea also considers that the Panel erred in relying on the "commonality of raw materials" and the similarity of manufacturing processes as a decisive criterion. In the European Communities' view, Korea improperly characterizes the Panel's reasoning. The Panel stated in unequivocal terms that "commonality of raw materials" is a relevant factor, but not a dispositive one. Nor did the Panel consider that the similarity of manufacturing processes is, in and of itself, decisive.

58. The European Communities does not accept that Korea's tap and bottled water example is comparable with that of generic and branded aspirin. Generic aspirin and branded aspirin are identical or nearly identical products, even if they are marketed differently. Tap water and bottled water have the same appearance, but it may be highly questionable whether they have close physical characteristics.

59. The European Communities argues that the Nielsen Study refutes Korea's assertions on consumption patterns of soju and western-style spirits. It showed that some consumers drank whisky with their meals and that soju is not always drunk with meals. The Trendscope survey confirmed that western-style spirits are sometimes consumed with meals. The Panel did not base its conclusion that soju and western-style spirits have similar end-uses on the Nielsen Study's finding that 6 per cent of consumers drank whisky with their meals. Rather, the Panel rejected the relevance of the narrow distinction between consumption with meals and without meals, and also between consumption with "snacks" or with "meals".⁴⁹

60. The European Communities disagrees with Korea that there is a contradiction in the Panel's treatment of marketing strategies. Although the Panel stated that marketing strategies can be used to create primarily perceptual distinctions between products, it also stated that marketing strategies can be useful tools for analysis if they highlight fundamental product distinctions or similarities. ⁵⁰ The Panel thereafter relied on marketing strategies that highlight underlying product similarities.⁵¹

61. The European Communities recalls that the complainants adduced evidence before the Panel that certain pre-mixed drinks contained soju, thereby refuting Korea's claim that soju is always drunk straight. Whether pre-mixes are considered as soju or as liqueurs for tax purposes is altogether irrelevant. Pre-mixed "gin and tonic", "whisky and cola" or "piña colada" would not be classified as whisky, gin or rum. Yet, their very existence constitutes irrefutable evidence that some consumers like to drink those spirits mixed with non-alcoholic beverages.

To continue with Channels of Distribution

³⁴Panel Report, para. 10.92.

³⁵Panel Report, para. 10.94.

³⁶This factor is based on the prices of the Dodwell Study. However, the Panel mentions an even higher price difference: a factor of 24 (Panel Report, footnote 408).

³⁷Panel Report, para. 10.94.

³⁸Panel Report, para. 10.58.

³⁹Ibid.

⁴⁰Panel Report, paras. 5.72 and 6.182.

⁴¹See Panel Report, para. 10.38.

⁴²Panel Report, para. 10.98.

⁴³Panel Report, para. 10.54.

⁴⁴Panel Report, paras. 10.101 and 10.102.

⁴⁵See Panel Report, para. 10.94 and footnote 410.

⁴⁶Ibid.

⁴⁷European Communities' appellee's submission, para. 68, citing Appellate Body Reports in European Communities - Measures Affecting the Importation of Certain Poultry Products ("European Communities - Poultry"), WT/DS69/AB/R, adopted 23 July 1998 and EC Measures Concerning Meat and Meat Products (Hormones) ("European Communities - Hormones"), WT/DS26/AB/R, WT/DS48/AB/R, adopted 13 February 1998.

⁴⁸European Communities - Hormones, supra, footnote 47.

⁴⁹Panel Report, para. 10.76.

⁵⁰Panel Report, para. 10.65.

⁵¹Panel Report, para. 10.79.