Korean Market Developments

1.203. The United States argues that the development of the premium soju market is a most important indicator of the fact that Western spirits directly compete and are substitutable for soju. Soju sales increased significantly with the introduction of premium soju, whose sales volumes have surpassed the combined sales of all imported spirits. This development demonstrates that Korean consumers are receptive to trying and buying newly introduced types of spirits, including those with a higher prices. Furthermore, development of premium soju is clearly intended to move soju up market by borrowing the cachet from imported spirits, which contradicts Korea's contention that soju is only the cheap drink of "commoners" or "ordinary folks".

1.204. The United States argues that another development in the Korean market further dispels Korea's notion that soju is in a market of its own: the production of flavoured soju and juice cocktails starting in 1995. Although it may well be that much of soju is consumed as described in Korea's submission, these market developments show that new forms, venues and usage of various distilled spirits will continue to develop in Korea.

1.205. The United States considers that market developments and present conditions in Korea directly undermine Korea's reliance on pre-tax price differentials as precluding substitutability between soju and Western spirits. Despite competitive distortions created by Korea's liquor tax system, demand for Western spirits, particularly whisky, has increased as the relative prices of whisky and soju declined ( there were 100% and 250% reductions in the applicable liquor tax rate and import custom duties, respectively. During this period, sales of whisky in Korea increased over 136 percent as compared to a 13 percent increase in soju sales. These differing rates of growth in sales of whisky and soju amidst tax and duty reductions on imports demonstrates that the products are in competition.

1.206. The United States claims that without precise data regarding Korean consumer purchasing activity, it is virtually impossible to determine with complete certainty that product demand, in this case whisky, is driven by factors unrelated to the demand and price of the other product in question, i.e. soju. However, the United States considers that several market developments suggest that changes in market share between soju and whisky are not unrelated:

Thus, according to the United States, the two concurrent trends of Korean consumers trading up to Western spirits while trading up to premium domestic soju are a strong indication of the substitutability between Western spirits and soju.

4. The Dodwell Study

1.207. The United States argues that the results of the Dodwell study provide ample evidence that Korean consumers treat soju and imported Western spirits as substitute products. The United States notes first that a cross- price elasticity study might be a helpful addition to the Article III:2 analysis, but it is not a necessary one. The emphasis on case-by-case analysis in the drafting history of Article III, and the practice among GATT and WTO panels, make it clear that it would not be appropriate to establish a general rule concerning the percentage shift in consumer preferences. The complaining parties have not put forward any precise estimate, but some empirical results of studies conducted in various national markets indicate that there is a statistically significant trade-off between types of spirits.

1.208. First, the United States notes that the Dodwell study was patterned after the ASI study utilised in the Japan-Taxes on Alcoholic Beverages II,²⁴² in which the panel stated "contained persuasive evidence that there is significant elasticity of substitution among the products in dispute."²⁴³

1.209. According to the United States, the small anomalies in the Dodwell study should not lead the panel to a conclusion that the Dodwell study reflects "independent movements of independent variables." In the US view, drawing such a conclusion on the results of the Dodwell study is tantamount to concluding that relative prices of soju, premium soju and imported spirits have no impact on consumption decisions. In fact, the overall trends in the Dodwell study should lead the Panel to the opposite conclusion: consumer preferences in the Korean market for distilled spirits are significantly related to relative prices of soju and imported distilled spirits.

1.210. According to the United States, the anomalies pointed out by Korea are merely deviations from what is usual, normal or expected.²⁴⁴ The US argument is that a study that relies on sampling to draw inferences about an overall population may yield results that contain some random variation, and therefore will be susceptible to occasional anomalous results.²⁴⁵ In the US view, what is important in the Dodwell study is that the overall trend displayed in the sample responses suggests that relative prices are a factor in consumption of distilled spirits, and that consumers normally will substitute imported spirits for soju as the relative price of soju rises.

1.211. According to the United States, the analysis of the Dodwell survey results as described in the first US submission detailed the percentage changes in those choosing Western spirits (excluding premium soju) as the relative price of standard soju rises. Therefore, the aggregate effects of nearly all of the scenarios demonstrate that respondents increasingly choose Western spirits when faced with higher relative prices for standard soju.

1.212. The United States notes that Korea's critique of the Dodwell study alleges flawed results and methodological weaknesses. Yet, despite these assertions, the study establishes a connection between changes in relative prices of soju and western spirits and purchasing behaviour. The Dodwell survey followed the methodology of the ASI study, attempting to determine whether the typical Korean spirits customer varies consumption preferences between soju and Western spirits as relative prices of spirits change. In the US view, the study did not attempt to determine the actual shares or shifts in market shares between spirits products, an objective that would not be relevant to this proceeding. Rather it sought to establish whether Korean spirits consumers view the products as substitutable. The fact that the observed percentages of survey responses do not perfectly correlate with actual soju market shares does not undermine the study's conclusions, and focusing on it overlooks the study's objective.

1.213. The United States notes that Korea further asserts that the Dodwell study reflects a sample bias and posed ambiguous questions. According to the United States, with respect to the survey questions, it is unreasonable to impute confusion to the entire survey population. With respect to the sample, inclusion of survey respondents having recently purchased soju and whisky suggests that, at least for brown spirits, respondents' familiarity with these products runs contrary to the assertion that the survey questions implied a one-time or experimental purchase. Finally, the separation of brown and white distilled spirits does not affect the fundamental movement of respondents from standard soju to Western spirits as standard soju's relative price increased.

1.214. The United States also notes that Korea questioned, as did the Panel, the grouping of premium soju with other Western distilled spirits in the Dodwell study. According to the United States, as with whisky (standard and premium), it was considered appropriate to establish more than one price point for diluted soju given that diluted soju products differed in price. Setting a price benchmark for the soju category based on the actual prices for the higher price premium soju to the exclusion of standard soju would have ignored the majority of the overall diluted soju market. Standard soju, due to its market share significance, was chosen as the benchmark for comparison. The fact that premium soju was selected as an alternative product, alongside Western spirits, to standard soju, and that some respondents switched to premium soju, does not undermine the validity of the study's conclusion.

5. The Measures are Applied "So as to Afford Protection to Domestic Production"

1.215. The United States does not address this issue in its rebuttal submission, but in its oral statements argues that, with respect to this element of the case, the Appellate Body stated in Japan - Taxes on Alcoholic Beverages II, the protective application of a tax can most often be discerned from "the design, the architecture and the revealing structure of the measure," which includes the very magnitude of the dissimilar taxation.

1.216. The United States further argues that Korea has largely ignored this element of the analysis. However, the Korean measures at issue do present a structure applied so as to afford protection to domestic production. According to the United States, the very large differentiation in tax rates between imported and domestic products, on the basis of a law that consists mainly of arbitrary exclusions from the definition of Korea's domestic product, soju, can only be considered protective. Moreover, in contrast to the facts presented in the Japan case, the protection of soju in Korea can be equated even more directly with the protection of a domestic industry. In Japan, there were significant imports of soju from Korea (which the panel in Japan - Taxes on Alcoholic Beverages II found were the same product as shochu), but in Korea, there have been only negligible imports of soju -- or shochu -- from any other country.

1.217. The United States further notes that Korea makes the claim that soju cannot be considered "domestic" production at all, because its raw material is often imported. According to the United States, this assertion ignores Korea's own legal structures, however. It is true that the main ingredient for soju is ethyl alcohol, 70% of which is imported. But, unlike whisky, ethyl alcohol is a raw material that can be used for a variety of end products. Ethyl alcohol is classified separately under HS heading 2207, and the process of manufacturing soju results in a substantial transformation of that raw material, while any imported soju is classified under HS heading 2208. Whisky, whether bottled or in bulk, is classified under HS heading 2208. Thus, it is plain that soju is a domestic product, while whisky is exclusively imported.

1.218. Finally, according to the United States, Korea's invocation of progressive social policies as a pretext for this discrimination is unrelated to the structure of its law, which is drawn on the basis of arbitrary physical characteristics and not on price of the product. While Korea may have a social policy objective of imposing lower internal taxes on inexpensive products purchased by lower income consumers, Article III does not permit Members to draw artificial product categories for tax purposes so as to discriminate against imports and protect domestic production. According to the United States, the same or similar arguments by Japan have been rejected in two panel proceedings.

C. Korea

1. The Nielsen Study

1.219. Korea states at the outset that in order to rebut assertions of the complainants in the Dodwell Study and elsewhere in their submissions, it has commissioned another study, carried out by the company A.C. Nielsen (Nielsen study).²⁴⁶

1.220. Korea states that the Nielsen study concluded that while all Korean restaurants, Chinese restaurants and mobile street vendors sell standard soju, most cafés/western style restaurants and bars sell whisky. The study also found that 29.3% of the respondents consumed alcoholic beverages at home with their meals, while 81% were found to have consumed such beverages with meals at restaurants. The study claims that diluted soju was the alcoholic beverage predominantly consumed with meals. Drinking diluted soju with meals was most popular at Korean restaurants (73%), followed by Japanese restaurants (18%). According to the study, of the seven beverages offered to the respondents, none were consumed with meals at cafés/western style restaurants, bars and hotel bars. Finally, the survey found that soju is predominantly consumed straight (98.6%), while whisky is usually consumed "on the rocks" (with ice) (63.8%).

1.221. In Korea's view, therefore, the Nielsen study substantiates its argument that there is no demand substitutability between soju and the western-type drinks at hand in the Korean market.

2. General Comments

1.222. Korea argued in its first submission and continues to maintain that a violation of Article III:2 cannot be made in the abstract. According to Korea, in order to begin to prove the existence of a violation, the complainants must show that there is a 'like' or directly competitive or substitutable relationship between specific products in a specific market. The decision as to whether two products have such a relationship is based upon an overall appreciation of many factors, such as their physical similarities, their end uses, price, and consumers' tastes and habits. Two products may be similar to some degree in some ways, and different to some degree in other ways.

1.223. Korea claims that the complainants have preferred to discuss each relevant criterion in the abstract and for all the products at once. Taking each criteria separately allowed them to make generalisations about all the products without addressing the specific arguments that Korea raised for each individual product. In Korea's view, as a result of that approach, the complainants do not provide a total view of the relationship between any particular product pair.

1.224. According to Korea, the effect of this approach is three-fold. First, it allows the complainants to gloss over or ignore differences between products. Second, it allows them to highlight and exaggerate the importance of exceptional cases. Third, through this approach they make it more difficult for the Panel to have a clear overall view of any one product pair.

1.225. Korea also asserts that what the complainants lack in evidence and argument, they try to make up for by allusions to Korea's past, its alleged protective measures, its lack of imports etc. According to Korea, its past as a developing country is not at issue in this case. What is at issue is Korea's market today. Korea refers to a neutral study funded by the EC Commission which observed that:

3. Generalisations About the Korean Products

1.226. Korea notes that consistent with their contention that all distilled spirits are the same, the complainants also refuse to recognise differences in the Korean spirits at issue in this case, claiming that 'soju', whether it be standard or distilled, is one product.

1.227. Korea has insisted from the beginning that standard soju and distilled soju are too different for these two products to be lumped together. Korea seeks to show that in trying to rebut this fact, the complainants first draw support from the fact that standard soju and distilled soju share, in part, the same name. According to Korea, the similarity in the names of standard soju and distilled soju is meaningless (for example, one would not consider 'beer' and 'root beer' to be 'like' or directly competitive or substitutable on the basis of their names).

1.228. Korea further notes that the complainants then go on to argue that the physical differences between standard soju and distilled soju are insignificant. However, those physical differences are obviously enough to affect consumer behaviour, as standard soju and distilled soju do not compete, indeed do not have the same end uses, and are sold at vastly different prices. The complainants have not been able to counter those arguments. In addition, it should be noted that the distinction between standard soju and distilled soju existed prior to this case, that these two products fall into different tax brackets,²⁴⁸ and that despite the EC's statements to the contrary, distilled soju is not exempt from tax.

1.229. Korea states that another point made by the complainants is that the difference between distilled and standard soju is really not important because, "they can be, and are, often blended with each other."²⁴⁹ The complainants then point out that whisky is often blended too. Korea points out that the complainants again offer no proof whatsoever for their allegation that distilled and standard soju are often blended. Accordingly, this example of the reference to whisky is irrelevant to a comparison between standard and distilled soju.

1.230. Korea refers to what it calls another astonishing, assertion of the complainants that, 'swift transition' occurred in the mid 1970s from distilled to standard soju, which and 'was possible only because, in the eyes of Korean consumers, the two varieties of soju are the same product'²⁵⁰ Korea recalls that standard soju was introduced in 1962 because, due to food shortages, distilled soju made of rice could no longer be produced. Production of distilled soju only started again in 1991. If a swift transition from one product to another following food shortages is an indication of a close relationship between them, Korea recalls that Parisians also made a 'swift transition' from eating beef meat to eating rat meat during the Fall of Paris in the 1870s, due to a food shortage.²⁵¹ According to Korea, peoples' behaviour in times of food shortage says very little about which products consumers would consider substitutes in normal circumstances.

1.231. Korea states that the complainants argue that the only reason Korea wants to distinguish between standard soju and distilled soju is because it wants to 'sacrifice' distilled soju, which makes up 0.2% of the 'soju market' in order to 'spare' standard soju. According to Korea, it has no intention of sacrificing distilled soju. While distilled soju differs from the imported beverages in different ways than standard soju, it still differs from them too greatly to be 'like' or directly competitive or substitutable for those products.

1.232. Korea further argues that this allegation that Korea is willing to 'sacrifice' distilled soju to 'spare' standard soju is characteristic of another facet of the complainants' approach to this case. In Korea's words, where it suits them, the complainants trade their inverted telescope for a microscope, 'zooming in' on exceptional cases that seem to support their argument. In the same way that looking at the situation from a great distance obscures the reality, zooming in too closely also means that your view is distorted.

1.233. Korea notes that the complainants allege that Korea is willing to 'sacrifice' distilled soju because it only represents 0.2% of the 'soju market'. What this highlights is the fact that the complainants are asking the Panel to do the opposite: they would like to use examples drawn from this 0.2% in order to prove a point about the other 99.8% of 'soju' sold in Korea, which is standard soju. For example, the complainants want the Panel to ignore the substantial price differences between standard soju and the western-style liquors at issue in this case, on the basis of the high prices of that 0.2%.

To continue with Actual (or Potential) Competition

²⁴² Panel Report on Japan - Taxes on Alcoholic Beverages II ,supra., para. 6.29.

²⁴³ Ibid., para. 6.32.

²⁴⁴ In the US view, Korea's critique of the Dodwell study focuses on random anomalies to the exclusion of the underlying findings that Korean consumers view standard soju and imported spirits as substitute products. For example, Korea's critique highlights that in chart 1 of the Dodwell study, the percentage of respondents choosing premium scotch falls as the price of standard soju rises from 1,100 to 1,200 won. The critique ignores that in the same scenario a rising percentage of respondents chose standard whisky and cognac as soju prices rose. In aggregate, the percentage choosing Western spirits in this scenario actually rose, indicating respondents' responsiveness to changes in the relative prices of Western brown spirits and soju. Moreover, the aggregate trends in nearly all the scenarios indicate increasing percentages of respondents choosing Western spirits amidst the relative rise of standard soju prices. In focusing narrowly on selective inconsistencies Korea has obfuscated the underlying study results evidencing that respondents frequently choose to substitute imported spirits for standard soju as the relative price of these products narrowed. See First Korea Sub., Attachment 2, at pp. 4-5.

²⁴⁵ In the US view, in only one of the price scenarios (medium brown spirit price levels) in the Dodwell study did the aggregate results run counter to theory. However, it should be noted that in terms of respondents, a net of 3 respondents choose non-Western spirits as the relative price of standard soju reached its highest level.

²⁴⁶ Korea claims that unlike the Dodwell Study, the Nielsen Study has focused on factual evidence rather than speculation. Rather than asking respondents 'what would you do?', it has asked them 'what do you do?'

²⁴⁷ See Sofres Report, Introduction.

²⁴⁸ According to Korea, while the US notes that standard soju and distilled soju have the same rate of Education Tax (10%), they fail to mention that the liquor tax on distilled soju is 50%, while the liquor tax on standard soju is 35%. Another mistaken attempt at trivialising the distinctions between distilled and standard soju is the EC's assertion that the distinction in the tax law was introduced only in 1991 (see EC written rebuttal, at para. 34), supposedly in response to pressure from the EC. In fact, the distinction was already made in 1962.

²⁴⁹ See EC written rebuttal, at para. 38.

²⁵⁰ See EC written rebuttal, at para. 35.

²⁵¹ See Horne, Alistar, The Fall of Paris: The Siege and the Commune 1870-71, London 1965.