C. Korea (Cont.)

15. In response to a question about the physical differences between exports of soju and shochu to Japan, Korea states that:

(a) The three leading brands of standard soju exported to Japan are Jinro, Doosan�s Green and Bohae. Jinro and Doosan only use sugar and citric acid in their products exported to Japan. Bohae�s standard soju exported to Japan uses no additives. On the other hand, Korean standard soju can use seven additives.

(b) The difference in additives can be illustrated by the difference in additives by the two products Jinro exports to Japan.

	Jinro Gold	Jinro Export
Alcohol content	25%	25%
Citric acid	x	x
Sugar (natural)		x
Fructose	x
Oligosaccharide	x
Stevioside	x
Refined salt	x
Amino acid	x

(c) The difference in the number of ingredients leads to different consumption patterns. In Japan, shochu A is almost always consumed with water, either warm or cold, other beverages and on the rocks. In Korea, standard soju is almost always consumed straight.

(d) The source of this information is enterprises such as Jinro that are engaged in marketing these products on the Japanese market, and who have found it necessary to export soju to satisfy Korean expatriates living in Japan, and to produce a different product to meet the needs of Japanese consumers.

(e) This can be easily verified by the Panel by looking at the bottles of Jinro Gold (the Korean soju, which targets the Korean residents in Japan) and Jinro export (targeting Japanese consumers) provided by Korea. The labels of Jinro Gold contain Korean characters, whereas those of Jinro Export contain no Korean characters.

(f) The labels of these bottles also give an indication regarding their tax treatment in Japan. The label on the back of Jinro Gold refers to "spirits", whereas the label on the back of Jinro Export refers to "shochu A".

16. In response to a question concerning a hypothetical comparison of an expensive bottle of wine and a cheaper table wine, Korea states that:

(a) For the purposes of Article III:2, in order to find that two products are "like", one must show to begin with, that the two products are "directly competitive or substitutable". According to the panel in the recent Japan - Taxes on Alcoholic Beverages II:

"Like" products should be viewed as a subset of directly competitive or substitutable products. ³⁰⁸

Accordingly, a finding of "likeness" in Article III:2 presupposes an even stronger competitive relationship between two products than a finding of directly competitive or substitutable products.

(b) Price has an impact upon the competitive relationship between products. Where prices vary greatly, this can mean that two products do not compete. In the example given by the Panel of cheap wine and a rare Bordeaux, it might be that in a particular market the large difference in price means that the two products do not compete, and are therefore neither "like" nor directly competitive or substitutable products. This is so because most consumers will not consider a $1 000 bottle of wine to be a substitute for a $5 bottle of wine, notwithstanding apparent similarities concerning physical characteristics between the two products (colour, packaging, alcoholic content).

(c) That conclusion is likely to be supported by other divergences. Indeed, the fact that a person chooses to pay more for a rare bordeaux when he or she has the option of buying a cheap table wine is a measure of the differences between the two products. Between a cheap wine and a vintage bordeaux, for example, important differences might be the age of the wines, the type, year, and provenance of the grapes used, the blending, the way the wine was cared for and stored, all of which have an impact on the organoleptic qualities of the wine. The rare bordeaux might come from a famous vineyard or chateau, and the purchaser might not be interested in drinking the wine at all, or at least not with a regular meal. He might see it as an investment or as a wine for very special occasions These types of differences would also support the conclusion that these products are not "like" or directly competitive or substitutable products.

(d) In the case at hand, Korea has shown that price is an important factor in the directly competitive or substitutable products, and therefore, "like" product analysis. The complainants� own evidence (notably, the Dodwell Study) indicated that there are no overlaps in the prices of standard and premium soju on the one hand, and western-style spirits on the other hand. Thus, the most expensive form of standard soju (i.e., premium) is still much less expensive than the cheapest western-style spirits. In contrast, prices of wine might cover a whole range of prices.

(e) Moreover, the products in this case also differ in a number of other ways, in particular in their physical characteristics and end uses. Korea maintains that an overall assessment of the products at issue in this case must lead to the conclusion that these products are neither "like" nor directly competitive or substitutable products.

17. In response to a question about the likeness or competitiveness of physically similar products, Korea states that:

(a) When two products are physically identical in the sense that no known test or procedure can distinguish between them, it seem to Korea that the products must be like. Physical identity in this strict sense, however, seems to imply either that the market prices of the products are the same, or that only one of them is sold, or that consumers are ignorant of their likeness. If two products both appear in a market, selling at different prices, but chemists, physicists, or other experts maintain that they cannot distinguish between them, there is therefore a problem. In effect, there is a conflict of evidence experts say they cannot detect any differences but consumers of the products act as if they can.

(b) In that event, Korea does not believe that the expert evidence alone is enough to declare the products either "like" or "directly competitive or substitutable products". To take that step, it is necessary to refute the hypotheses that the expert tests are incapable of detecting differences that are significant to consumers; or that the experts have not properly designed or targeted their tests to detect them; or that they are misinterpreting their results.

(c) With respect to "physically identical or nearly so", it seems to Korea that the addition of "or nearly so" to "physically identical" raises difficulties. One problem is circularity; defining "nearly so" so that any change in the product that has only a small effect on consumer demand is judged to leave a "nearly identical" product, whereas any change that has a substantial effect on demand is judged to lead to a different product. Such a definition of "nearly identical" would render this question meaningless - if such a definition is used, products that fall under it are likely to be directly competitive or substitutable products.

(d) Korea believes, however, that many changes that are "small" in a technical sense will lead to products that are very different by the test of market performance. To take one example, the addition to a food or drink product of small amounts of an unpleasant substance, or a substance perceived to be unpleasant, may, if known to consumers, cause demand for the product to fall to zero, even though the substance does not alter taste or threaten health and its presence cannot be detected by consumers. The "small" addition may convert a sought-after product into one that is no longer saleable at any price. Moreover, the elasticity of substitution between the original product and the "nearly identical" product may be zero or insignificant. Korea is far from convinced that "nearly identical products" with different market demands and prices and a low or zero cross-price elasticity of demand should be considered directly competitive or substitutable products.

(e) Korea believes that products may be physically identical or nearly so but not directly competitive or substitutable products. Korea believes that differences that are small in a technical sense may be important to buyers, and that performance in the market place is the ultimate test of directly competitive or substitutable. Alternatively stated, Korea does not believe that there is any technical short-cut that can determine direct competitiveness or substitutability without reference to market performance.

(f) The relevant physical distinctions are those that are important to customers. The WTO is concerned with markets, and markets are ultimately dependent on consumer tastes and habits. In the present case, regarding spirits, Korea has pointed to the following distinguishing physical characteristics:

(i) raw materials, additives

(ii) production process

(iii) alcohol percentage

(iv) flavour, smell and colour

(g) Together with the difference in end-use and price, the distinctions indicate the lack of a directly competitive and substitutable relationship between Korean sojus on the one hand, and the western-style liquors at issue on the other hand.

(h) When looking at physical characteristics, the appropriate measure of relevance is the importance to consumers. In the present case, Korea has argued that the differences in physical characteristics, together with other market-related factors (such as price and end use), indicate that none of the western-style liquors competes directly with any of the two Korean sojus on the Korean market.

18. In response to a question about comparing premium diluted soju to standard diluted soju, Korea states that:

(a) Korea included premium diluted soju in its comparison of diluted soju with other liquors because the differences between premium and standard diluted soju, as compared to the differences between diluted soju (including premium soju) and imported liquors, are of minor importance. Premium diluted soju is only an "upgraded" version of standard soju. Thus, premium diluted soju could be compared to a Renault�s compact car "Clio" with leather seats which, although it is a luxury version, is still a compact car. To proceed with this analogy: imported liquors are Mercedeses, Jaguars and Rolls Royces compared to diluted soju. The price of a Renault Clio, even equipped with leather seats, is substantially lower than the cheapest Mercedes, let alone a Jaguar or Rolls Royce.

(b) By taking the same criteria by which one distinguishes products in a like-or directly competitive or substitutable products-analysis, the differences resulting from a premium and standard comparison are small. For example, the price difference between standard diluted soju and premium diluted soju amounts to a factor of 1.76, as opposed to a factor of 5 for the cheapest imported liquor (gin), and more than 19 for the most expensive imported liquor (cognac/brandy).

(c) For one of the leading premium diluted soju brands, Kimsatgat, the primary difference has been that one of the seven possible additives (stevioside) is replaced by honey. In contrast, for the imported liquors at issue (whisky, vodka, gin, brandy, rum) there are differences in taste, physical characteristics, and end-use. The taste of gin, for instance, is quite different from the taste of diluted (including premium) soju, due to the different physical composition. Gin is not drunk with Korean meals, whereas diluted (including premium) soju is, and so on.

(d) This analysis led Korea to the conclusion that standard diluted soju and premium diluted soju are sufficiently similar and sufficiently competitive that they should be grouped together in the analysis required by this case. To be sure, Korea has treated distilled soju separately from diluted soju, as between these two products the differences are significant (price, use, marketing, raw material, tax classification).

19. In response to a question about competition between types of whiskies, Korea states that:

(a) There is no reason to assume that imported whiskies are not directly competing with or unlike domestic whiskies because of the price differences cited. The prices cited concern individual brands. Both domestic and imported whiskies cover the whole gamut of prices (for instance, cheap whiskies are bottled in Korea, but also expensive ones). Taking into account all sales, domestic whisky is on average somewhat more expensive than imported whisky. ³⁰⁹

(b) There are no other differences (such as physical characteristics or end use) to suggest that imported and domestic whisky are positioned differently on the Korean market. Next to the small difference in price, this is relevant as well. As Korea has emphasized before, the decision of whether products are "like" or directly competitive or substitutable products is a matter of an overall appreciation of their relationship, weighing all the relevant factors. Of course, no distinction is made in Korea�s liquor and education tax rates between imported and domestic whisky.

20. In response to a question about the negotiating history of Article III and Ad Article III, Korea states that:

(a) These examples show that physically different products may be in a sufficiently close competitive relationship for Article III:2 to apply. Whether that is, in fact, the case, depends on a case-by-case analysis, according to the Appellate Body in Japan - Taxes in Alcoholic Beverages II.

(b) It may well be that apples and oranges are directly competitive or substitutable products in certain markets. Then again, one can conceive of a number of reasons why such fruits are not directly competitive or substitutable (e.g., with breakfast it is more common to have orange rather than apple juice; most consumers can make orange juice themselves, but not apple juice; apple pie is more common in many countries than orange pie; in countries where oranges are not grown they are more expensive than apples; etc.). In fact, certain fresh fruits, such as bananas, have been found to be in a market of their own, at least in the EC market. ³¹⁰ Such determinations, also in respect of the other products cited, cannot be made in the abstract.

(c) These examples then are relevant to the present case, in that they illustrate that products with different physical characteristics may be directly competitive or substitutable products. This was also shown in the Japan - Taxes on Alcoholic Beverages II, where whisky and certain other spirits were found to be in a directly competitive or substitutable product-relationship with Japanese shochu. That case also shows, however, that such findings depend on a factual analysis of individual markets.

21. In response to a question concerning legal requirements for sweetness in soju, Korea states that there is no legal requirement regarding the minimum sugar content for soju in Korea. The sweeter taste of Korean soju can be explained by the use of such additives as stevioside and/or aspartam, which are 150-300 times sweeter than sugar.
22. In response to a question about consumption of soju with food, Korea states that the bulk of standard soju is consumed with meals. Other than this, a small proportion is consumed in some other on-premise locations. Some standard soju will also be consumed at home without a meal (finishing a bottle after the meal is over, for instance).
23. In response to a question about uses of distilled soju as a gift, Korea states that, as has been emphasized in Korea�s first submission, distilled soju is an artisanal product that occupies a "niche" in the Korean market, and is mainly given as a gift. When distilled soju is received as a gift, it is usually consumed with meals on traditional occasions such as New Year�s Day and Korean "Thanksgiving" (August 15). Other instances in which distilled soju is consumed are rare. In some very expensive and traditional Korean restaurants and Japanese restaurants, distilled soju is offered, at very high prices.
24. In response to a question whether the questions and methodology in the Dodwell study are similar to those used in the ASI study in Japan- Taxes on Alcoholic Beverages II, and whether Korea disagrees with that Panel's use of the ASI study in reaching its conclusions, Korea states that it is confident that the Panel in Japan � Taxes on Alcoholic Beverages II examined the ASI study with appropriate care. Korea, however, has not studied the detail of the ASI report, which was submitted to a different panel in a different case, and is concerned with a market in a different country. It is therefore unable to comment on the merits of the ASI study. Moreover, Korea doubts the value of a post-mortem on either the ASI study itself, or the use of it by the Panel in Japan � Taxes on Alcoholic Beverages II. If the ASI study is free of the flaws of the Dodwell study, its use by the Panel was proper, but that fact cannot provide a sound argument for reliance on the flawed Dodwell study in this proceeding. If the ASI study is as defective as the Dodwell Study, the fact that the Panel relied on it cannot make a sound case for repeating the same mistake.
25. In response to a question concerning cross-price elasticity, Korea stated:

(a) Korea agrees with the implied observation that the position of the US on this issue is inconsistent with that of the EC.

(b) To discuss the comments of either party, however, a context is needed. One important element of such a context is the kind of world to which te comments are intended to apply � is it a theoretical world in which information is fully and freely available, or the real world in which accurate information is difficult to get?

(c) The US suggestion that +any shift � should be interpreted as a sign of positive cross-elasticity and therefore substitution" seems to Korea to cast the Article III:2 net too widely if it is intended to apply to a full-information world, and therefore to be a statement of principle. That criterion would make spirits, beer and wine DSSP, spirits and soft drinks quite possibly DCSP, and beer and soft drinks almost certainly DCSP. It seems unlikely that any WTO member thought that membership entailed an obligation to apply the same rate of tax to such a wide range of products � and few, if any, do.

(d) If the US criterion is intended for application to the real world, however, imperfections of information provide additional reasons to reject it. In the real world, there are problems of error in the construction and organisation of samples and problems of statistical error � to say nothing of the blundering of those "fallible human beings" that have become a feature of recent EC argument in this case. Even if the US criterion were accepted in principle, the existence of such sources of mistakes in estimation would require a substantial margin to allow for errors in estimation � a margin that the US position denies.

26. In response to a question about a de minimis standard for the question of "so as to afford protection", Korea stated:

(a) If products that are only very weakly substitutable can be DCSP, a tax on a domestic product that is lower than the tax on a DCSP foreign product may have only a very small effect on the quantity demanded of the foreign product in the country applying the tax. Korea considers that such a small effect should not be sufficient to meet the "so as to afford protection" requirement of the second sentence Article III:2.

(b) Indeed, Korea submits that de minimis protective effects do not trigger the application of the second sentence. This is based on the text and structure of this provision, which is unlike the hard-and-fast prohibitions incorporated in the first sentence of Art. III:2, or such other GATT provisions as Art. I or XI.

(c) To begin with, where DCSP are concerned, not just any tax differential is problematic. Only tax differentials that are more than de minimis can become a problem. WTO members are thus left with some flexibility in designing their tax systems.

(d) Similarly, the insertion of the "so as to afford protection "requirement in the second sentence of Article III:2 must also have been meant to allow governments a measure o flexibility. Article III:2 only interferes in a Member's tax system where an (appreciable) tax discrepancy raises real concerns about protectionism.

(e) Therefore, Korea considers that where the tax differential can only be said to have a minimal protective effect, it should not be considered to have met the "so as to afford protection" threshold.

27. In response to a question concerning the product mix of imports, Korea stated:

(a) Under a system of specific taxes, all bottles of scotch, for example, have the same tax whatever their price, so the specific tax raises the price of high-price brands by a smaller percentage than the price of low-price brands. Alternatively stated, the specific tax lowers the price of high-priced scotch relative to low-priced scotch, and therefore provides an incentive to the purchase of high-price brands.

(b) Korean taxes on sprits, however, are ad valorem � they have no effect on the price of one spirit in a particular class (for example, whisky, brandy) relative to another member of that class. Korea therefore sees no basis for the proposition of the EC that the Korean tax system favours the purchase of high-price rather than low-price brands of scotch.

(c) Korea does not know whether its residents import a higher proportion of high-priced scotch than those of similar countries. If it is a fact, however, Korea believes that an explanation for it must be sought elsewhere than in its tax system.

VIII. Third-Party Arguments

A. Canada

1. Canada's submission is limited to the issue of the criteria to be taken into account in assessing whether a difference in tax rates is applied so as to afford protection to domestic production.
2. Canada asserts that it welcomed the outcome of the Japan - Taxes on Alcoholic Beverages II case and was pleased with the principles for the interpretation and application of Article III of GATT 1994, set out by the Appellate Body in its report. Canada notes that the issues which arise in the context of the Korean liquor tax regime bear strong resemblance to matters which were under dispute in Japan - Taxes on Alcoholic Beverages II and accordingly, the Panel�s disposition of the present dispute should be guided by the principles established in the reports of the Panel and Appellate Body in that case.
3. Canada notes that all of the participants appear to have embraced the principles of the Appellate Body report in approaching this case and in particular with respect to the interpretation of the second sentence of Article III:2 of GATT 1994. In Canada's view, all participants appear to agree that the phrase "so as to afford protection" in Article III:1, as it applied to the second sentence of Article III:2, should be interpreted only with respect to objective effects and that no subjective element of intent should be taken into consideration. Canada urges the Panel to base its decision in the present case on these principles.
4. In that context, having examined the submissions and evidence presented thus far, Canada agrees with the European Communities and the United States that the assessment of whether the measures are applied so as to afford protection to domestic production involves an objective analysis. Canada further agrees that the facts and circumstances regarding the "structure of the measures" as well as their "overall application on domestic as compared to imported products" demonstrate that the Korean measures at issue are applied so as to afford protection to the domestic production of soju in Korea.
5. Canada stresses that in examining the "design, architecture and revealing structure" of a measure, only objective factors should be taken into account. For example, as noted by the Appellate Body in Japan - Taxes on Alcoholic Beverages II, the magnitude of dissimilar taxation (which is an objective factor that can be discerned from the structure of a measure) in and of itself can be evidence of protective application. ³¹¹ In fact, in Canada�s view, given the magnitude of the tax differentials in the dispute at hand, it is unnecessary to examine any other factors. ³¹²
6. Thus, in Canada�s view, assessing whether a measure is applied so as to afford protection to domestic production is an analysis to be based on objective criteria, and in this case, the amount of the tax differential is sufficient to make a finding that the measures at issue are applied in a manner that protects domestic production.

To continue with Mexico

³⁰⁸ See Panel Report on Japan - Taxes on Alcoholic Beverages II, supra., at para. 6.22.

³⁰⁹ See Korea's, Attachment 5.

³¹⁰ Judgment of the Court of Justice of the European Communities of 14 February 1978, Case 27/76, United Brands Company and United Brands Continental BV v Commission of the European Communities, 1978 ECR 207.

³¹¹ Appellate Body Report, supra, p. 29.

³¹² Canada notes that according to the EC and the US, the magnitude of the tax differentials at issue in this dispute appear to be larger than those found to be sufficient evidence of protective application in Japan - Taxes on Alcoholic Beverages II. In Canada's view however, even if the Panel were to find it necessary to take into account additional evidence, the other factors presented by the EC and the US are more than sufficient to establish protective application.