I. Introduction

1. This is an appeal by Korea from certain issues of law and legal interpretation developed in the Panel Report, Korea - Taxes on Alcoholic Beverages.¹ That Panel was established² by the Dispute Settlement Body (the "DSB") to examine the consistency of two Korean tax laws: the Korean Liquor Tax Law of 1949 and the Korean Education Tax Law of 1982, both as amended (the "measures"), with Article III:2 of the GATT 1994. The Liquor Tax Law imposes an ad valorem tax on all distilled spirits. The rate of that tax depends on which of the eleven fiscal categories a particular alcoholic beverage falls within. The Education Tax Law imposes a surtax on the sale of most distilled spirits, the rate of the surtax being a percentage of the liquor tax rate applied to the spirit in question. A detailed description of the operation of these two taxes is to be found at paragraphs 2.1 to 2.23 of the Panel Report.

2. The Panel considered claims made by the European Communities and the United States that the contested measures are inconsistent with Article III:2 of the GATT 1994 because they accord preferential tax treatment to soju, a traditional Korean alcoholic beverage, as compared with certain imported "western-style" alcoholic beverages. The Panel Report was circulated to the Members of the World Trade Organization (the "WTO") on 17 September 1998. The Panel "reached the conclusion that soju (diluted and distilled), whiskies, brandies, cognac, rum, gin, tequila, liqueurs and admixtures are directly competitive or substitutable products." ³ The Panel also concluded that "Korea has taxed the imported products in a dissimilar manner and the tax differential is more than de minimis" and that "the dissimilar taxation is applied in a manner so as to afford protection to domestic production."⁴ The Panel made the following recommendation:

3. On 20 October 1998, Korea notified the DSB of its intention to appeal certain issues of law covered in the Panel Report and legal interpretations developed by the Panel, pursuant to paragraph 4 of Article 16 of the Understanding on Rules and Procedures Governing the Settlement of Disputes (the "DSU"), and filed a Notice of Appeal with the Appellate Body, pursuant to Rule 20 of the Working Procedures for Appellate Review (the "Working Procedures"). On 30 October 1998, Korea filed its appellant's submission.⁶ On 16 November 1998, the European Communities and the United States filed their respective appellees' submissions⁷ and Mexico filed a third participant's submission.⁸ The oral hearing, provided for in Rule 27 of the Working Procedures, was held on 24 November 1998. At the oral hearing, the participants and the third participant presented their arguments and answered questions from the Division of the Appellate Body hearing the appeal.

II. Arguments of the Participants and the Third Participant

A. Korea - Appellant

1. "Directly Competitive or Substitutable Products"

4. Korea contends that the Panel misinterpreted and misapplied the term "directly competitive or substitutable product", especially the word "directly" which, in Korea's view, is at the heart of the term at issue. At some level all products are competitive, in that they compete for the consumer's limited budget, and it is therefore "directly" which gives meaning to the legal text and prevents Article III:2 from becoming an "unbridled instrument of tax harmonization and deregulation".

(a) Potential Competition

5. Korea claims that the alleged evidence of "potential" competition was essential to the Panel's finding of a directly competitive or substitutable relationship between the products at issue. ⁹ However, Article III:2 does not speak of "potential" competition. Accordingly, it is at least ambiguous whether "potential" competition is embraced by the second sentence of Article III:2 of the GATT 1994. Given that ambiguity, Article 19.2 of the DSU and the principles of predictability and in dubio mitius should have been respected by the Panel.

6. In Korea's view, the term "directly competitive or substitutable" is not meant to exclude products that do not compete directly or are not substitutable because of the contested measure itself. The absence of a competitive relationship on the market concerned should be taken as a powerful counter-indication that the products involved are not "directly competitive or substitutable". The Panel, however, has read Article III:2 as covering both products that "are either directly competitive now or can reasonably be expected to become directly competitive in the near future." ¹⁰ (emphasis added) In so doing, the Panel relieved the complainants of the need to prove that the lack of actual competition is caused by the contested measure, and opened the door to speculation about how the market could evolve in the future, irrespective of the measure in question. Korea warns against speculation about what consumers might (or might not) do, as opposed to looking at what they actually do. The Panel repeatedly excused the complainants' failure to produce evidence about actual competition by saying that preferences in the Korean market might have been frozen by government measures.¹¹ However, Korea points out that, at the time this case was argued, its market had been open for eight years.

7. Korea contends that the "potential" standard is impermissibly broad and speculative, and the wording and the purpose of Article III:2 do not permit this interpretation. There is nothing wrong with requiring complainants to wait until, if ever, their case becomes "ripe" and products actually compete directly. What if a Member has been forced to change its tax law because products might compete and then, in fact, they do not? Should that Member return to the panel to request permission to restore its tax system?

(b) Expectations, the "Trade Effects" Test and the "Nature" of Competition

8. Korea notes the considerable emphasis the Panel placed on "expectations" of an "equal competitive relationship" between imported and domestic products.¹² However, Korea argues that these "expectations" exist only for those products which are "like" or "directly competitive or substitutable". If products are not currently directly competitive or substitutable, there can be no relevant expectations with respect to them.

9. The Panel erroneously considered that to "focus on the quantitative extent of competition instead of the nature of it, could result in a type of trade effects test being written into Article III cases." ¹³ This is a misunderstanding of the "trade effects" test. While past cases held that a lack of "trade effects" is not a defence to an Article III:2 violation, in those cases the products involved had already been shown to be "like" or "directly competitive or substitutable".

10. Korea observes that the Panel referred to the "nature" of competition many times in its findings, making statements such as: "the question is not of the degree of competitive overlap, but its nature." ¹⁴ By examining the nature of competition, the Panel added a vague and subjective criterion which is not present in Article III:2, and dispensed with the complainants' obligation to show direct competitiveness or substitutability and also with the need to look at actual markets.

(c) Evidence from other Markets

11. The Panel stated that it could "look at other markets and make a judgement as to whether the same patterns could prevail in the case at hand." ¹⁵ In Korea's view, this amounts to little more than guesswork and constitutes an impermissible broadening of the scope of Article III:2. The Panel also disregarded the fact that consumer responsiveness to different products "may vary from country to country".¹⁶ Moreover, there was no basis for the Panel to assume that the Korean and Japanese markets were, or were becoming, the same. Korea further contends that, even if evidence from other markets were relevant, the Panel should not have limited itself to looking at only one other country's market. To ensure a balanced view, evidence from more than one other market ought to have been reviewed.

12. Korea submits that all of the above misinterpretations of Article III:2 constitute a violation of provisions of the DSU and general principles of law, namely, the principle that neither panels nor the Appellate Body can add to or diminish the rights and obligations provided in the covered agreements (Article 19.2 of the DSU), the principle of predictability, and the principle of in dubio mitius.

(d) Grouping of the Products

13. Korea stresses the importance of the methodology used to compare domestic and imported products under Article III:2. It considers that the Panel committed a major legal error in wrongly defining the comparison it had to undertake. The Panel grouped together products that are not physically identical; are produced in different ways by different manufacturers using different raw materials; taste differently; are used differently; are marketed and sold differently at considerably different prices and are subject to different tax rates in Korea. The Panel also failed to carry out a separate analysis for diluted and distilled soju.¹⁷ Korea urges that the Panel erred in conducting its analysis on the basis of an agglomeration of the characteristics of two such different products. To extend conclusions that are primarily based on diluted soju to distilled soju is unacceptable logic. Further, by treating diluted soju and distilled soju together, the Panel overlooked the relevance of the considerable price differential between diluted soju and whisky.

14. Korea points out that the Panel decided to treat all the imported distilled spirits as one group.¹⁸ In Korea's view, the Panel's decision to group all these beverages together was, in effect, a decision (or at least a presumption) that they are directly competitive or substitutable everywhere, without considering whether that was true in the Korean market. By grouping all the imported beverages together, the Panel made it impossible to appreciate the differences between the imported products. The Panel could not, for example, conclude that in Korea diluted soju was directly competitive or substitutable for vodka, but not whisky.

15. Korea acknowledges that when the Panel considered the product characteristics, it examined the products in the group one by one. But the Panel dismissed as insignificant, differences between the products regarding such characteristics as colour, taste and price. In so doing, the Panel "trivialized" actual consumer perceptions which are at the heart of the "directly competitive or substitutable" standard. The erroneous approach adopted by the Panel makes it impossible to determine what the outcome of the case would have been if the Panel had not erred at the outset.

2. "So As to Afford Protection"

16. According to Korea, the Panel erred in finding that the Korean taxes had a protective effect mainly on the basis of an analysis of the structure of the law itself. The Panel ignored Korea's explanation for the structure of the law.¹⁹ The Panel also made too much of the fact that there is virtually no imported soju, overlooking the fact that there has simply been a lack of interest abroad in the manufacture of these typically Korean products. More importantly, the Panel did not follow the Appellate Body's ruling in Japan - Alcoholic Beverages, to the effect that, even though the tax differential may, in some cases, show that the tax is applied "so as to afford protection", "in other cases, there may be other factors that will be just as relevant or more relevant to demonstrating that the dissimilar taxation was applied 'so as to afford protection'." ²⁰ (emphasis added)

17. Korea reiterates the argument it made before the Panel that, in view of the large, intrinsic pre-tax price-differences between diluted soju and the imported products at issue, the tax differential cannot be said to have the effect of "afford[ing] protection" to diluted soju. Where the price-difference between two products is so significant, the additional difference created by the variation in tax can have no protective effect. Korea also maintains that demand for distilled soju is specific and static, and that it would not be affected a great deal by altering the price, especially not to the degree at issue in this case. Korea, therefore, claims that the tax differential does not "afford protection" to distilled soju, contrary to the conclusion reached by the Panel.

3. Application of Article III:2 of the GATT 1994

18. Korea submits that the Panel erred in several ways when assessing the evidence. While Korea recognizes that appellate review is limited to questions of law, it considers that, in reviewing a panel's interpretation and application of Article III:2, second sentence, the Appellate Body cannot avoid considering the factual underpinnings of the panel's assessment. In this case, the Panel drew conclusions which the evidence before it did not support. Errors of this type were decisive in the adjudication of the dispute in favour of the complainants, and thus constitute reversible legal errors.

19. The Panel also erred in applying different standards of proof to the evidence. The Panel was far more exacting when looking at evidence submitted by Korea than when considering evidence brought by the complainants. The Panel, in effect, applied a "double standard of proof ".²¹ The Panel also misapplied the requirements on the burden of proof which follow from the Appellate Body Report in United States - Measure Affecting Imports of Woven Wool Shirts and Blouses from India ("United States - Shirts & Blouses").²²

20. Despite evidence to the contrary provided by Korea, the Panel relied upon the notion that consumer preferences in the Korean market might have been frozen by the measures at issue. By so doing, the Panel unfairly put Korea in the position of having to prove a negative -- that the lack of competition was not due to the contested measures -- rather than requiring the complainants to prove positively that consumer preferences in Korea had been frozen.

(a) Product Characteristics

21. Korea notes that the Panel found that "[a]ll the products ... have the essential feature of being distilled alcoholic beverages." ²³ In essence, the Panel considered this sufficient to raise a presumption that all distilled alcoholic beverages are "directly competitive or substitutable". Korea disagrees that such a generic statement could give rise to a presumption of this type. The Panel erred in dismissing the importance of flavour in a case concerning beverages. The flavour of products is one of the consumer's primary considerations when choosing a beverage and distinctions between flavours are, therefore, not "minor" ²⁴ from the consumer's perspective.

22. The Panel's focus on the fact that all the alcoholic beverages at issue are produced by distillation means that certain industrial products (e.g., paint thinner) or medicinal products (e.g., rubbing alcohol) would also be in a directly competitive or substitutable relationship with the beverages in question. Similarity in raw materials and the methods of production are, therefore, meaningless in defining a directly competitive or substitutable relationship between products.

23. That the Panel applied a "double standard of proof" is shown by its rejection, on the one hand, of Korea's example of bottled and tap water, which Korea believes demonstrated that close physical similarity is not always probative evidence of a directly competitive or substitutable relationship. On the other hand, the Panel relied on the United States' example of branded and generic aspirin to show that physical similarity was highly significant. Moreover, the Panel dismissed Korea's bottled and tap water example, in part, because it refered to "different products in different countries".²⁵ Yet, in another part of its Report, the Panel said that evidence from "other countries" was relevant.²⁶

(b) End-Uses

24. Before the Panel, Korea showed that, in Korea, the overwhelming end-use of diluted soju is consumption during meals whereas western-style drinks are hardly ever consumed with meals. The Panel, however, found that this distinction did not suffice to prevent the products from being considered as directly competitive or substitutable.²⁷ Korea believes the Panel erred, both as to the application of Article III:2 and as to requirements of the burden of proof, in accepting that all the beverages at issue were drunk for the same purposes, inter alia, socializaton and relaxation. In reaching this finding, the Panel drew upon three sources: (a) trends and anecdotal evidence; (b) marketing strategies; and, (c) the presence of admixtures.

25. The Panel placed emphasis on "trends and changes in consumption patterns" ²⁸ which it said were demonstrated in the Nielsen Study and the Dodwell Study. However, neither of these studies nor the Trendscope Study contains evidence of trends. They show, instead, a "snapshot" of the market at a particular moment in time; they do not show changes over time. The Panel erred in considering that these studies contained evidence of trends, and the Panel's statements amount to mischaracterization of the evidence presented. Moreover, the Panel erred in speculating that those trends were "likely to continue",²⁹ without pointing to any supporting evidence.

26. Korea argues that the Nielsen Study was used "selectively" by the Panel. For example, even if it showed some overlap in beverages available in Japanese and western-style restaurants, it also showed that in the large majority of outlets there was no overlap. Thus, the overlap shown in the Nielsen Study is very limited and, in light of contrary evidence, cannot be considered as conclusive proof of the similarity of end-uses between the drinks at issue. In other words, the overall consumption pattern sufficiently rebuts any presumption of common end-uses raised by the minor overlap indicated by the Nielsen Study. The same is true of the figures given in the Nielsen Study for home-consumption of alcoholic beverages with meals. Even if 5.8 per cent of respondents stated that they drink whisky with their meals, that still leaves 94.2 per cent who do not. This evidence supports Korea's argument regarding the "meal" end-use of particular alcoholic beverages. Instead, it was turned around to become evidence of "overlap" in end-use. The speculation engaged in by the Panel was made worse by the Panel's consideration of trends on the Japanese market.

27. The Panel's treatment of Korean companies' marketing strategies discloses again a "double standard of proof ". Where the Panel considered that marketing strategies supported a finding that the products were "like" or "directly competitive or substitutable", they became important evidence ³⁰, whereas the Panel dismissed evidence from marketing strategies that it considered did not support such a finding.³¹

28. The Panel also erred when assessing the evidence submitted concerning admixtures. Korea argued that diluted soju and stilled soju are consumed "straight" in Korea (unlike some of the imported beverages at issue in this case), a fact borne out in its market study.³² That soju cocktails are different from diluted soju and distilled soju is reflected in Korea's tax law. Korea maintains that the presence of diluted soju in admixtures cannot support a finding of similarity with other drinks which are drunk in a mixed form, just as the existence of Bailey's³³ is not proof that whisky is often drunk mixed. Like Bailey's and whisky, diluted soju, distilled soju and admixtures are different drinks and are treated as such under the Liquor Tax Law. The Panel wrongfully rejected Korea's point which rebutted the evidence on admixtures.

To continue with Channels of Distribution

¹WT/DS75/R, WT/DS84/R, 17 September 1998.

²The Panel was established 16 October 1997 with standard terms of reference (see WT/DS75/7, WT/DS84/5, 10 December 1997) that were based on requests for the establishment of a panel made by the European Communities (WT/DS75/6, 15 September 1997) and the United States (WT/DS84/4, 15 September 1997).

³Panel Report, para. 11.1.

⁴Ibid.

⁵Panel Report, para. 11.2.

⁶Pursuant to Rule 21(1) of the Working Procedures.

⁷Pursuant to Rule 22 of the Working Procedures.

⁸Pursuant to Rule 24 of the Working Procedures.

⁹Panel Report, para. 10.97.

¹⁰Panel Report, para. 10.48. Korea also refers to paras. 10.40 and 10.73 of the Panel Report.

¹¹See, for example, Panel Report, para. 10.94.

¹²Panel Report, para. 10.48.

¹³Panel Report, para. 10.42.

¹⁴Panel Report, para. 10.44. Korea also refers to Panel Report, paras. 10.42, 10.44 and 10.66 in this respect.

¹⁵Panel Report, para. 10.46.

¹⁶Panel Report, Japan - Taxes on Alcoholic Beverages ("Japan - Alcoholic Beverages"), WT/DS8/R, WT/DS10/R, WT/DS11/R, adopted 1 November 1996, para. 6.28.

¹⁷Panel Report, para. 10.54.

¹⁸Panel Report, para. 10.60.

¹⁹Korea's explanation of the structure of its tax regime is set out at paras. 5.172 to 5.181 of the Panel Report. In its arguments before the Appellate Body, Korea placed particular emphasis on the arguments summarized at para. 5.176 of the Panel Report.

²⁰Adopted 1 November 1996, WT/DS8/AB/R, WT/DS10/AB/R, WT/DS11/AB/R, p. 30.

²¹Korea's appellant's submission, para. 85.

²²Adopted 23 May 1997, WT/DS33/AB/R, WT/DS33/R, p. 14.

²³Panel Report, para. 10.67.

²⁴Ibid.

²⁵Panel Report, para. 9.23.

²⁶Panel Report, para. 10.45.

²⁷Panel Report, para. 10.76.

²⁸Panel Report, para. 10.48.

²⁹Panel Report, para. 10.76.

³⁰Panel Report, para. 10.79.

³¹Panel Report, paras. 10.65 and 10.66.

³²See in particular Panel Report, paras. 5.268 and 5.273.

³³Panel Report, para. 7.11. Bailey's Irish Cream is an alcoholic beverage which is a mixture of whisky and cream.