(c) Directly Competitive or Substitutable

Physical Characteristics

1.175. According to the United States, Korea's submissions argue that some differences in production processes and physical characteristics prove that soju and imported distilled spirits are not in fact "like," "substitutable" or "competitive" products. The United States argues that from a basic economic perspective, however, differences in production or physical characteristics are not a priori determinative of whether two products are substitutable or directly competitive. For example, cane sugar and artificial sweeteners are totally different in terms of production process and chemical composition, yet they clearly compete directly in coffee houses and restaurants. In order to determine substitutability, the Panel should consider whether the products in question compete for consumer spending on a category of goods. In this case, all spirits should be considered as in competition because they compete for consumers' spending on various products within the category of alcoholic spirits. Similar production processes, physical characteristics, and end uses are indicative of "like" products, but some differences in these factors do not establish that two products are not substitutable. Another factor in determining substitutability is the extent to which consumers respond to relative price increases in one product by increasing purchases of another. These products are likely to be substitutes. Thus, Korea's reliance on minor differences in the production and physical characteristics as dispositive evidence of non-substitutability is misplaced. Furthermore, Korea greatly exaggerates the differences between soju and imported spirits.

1.176. The United States argues that it is unlikely that two identical products would not, to some degree, be substitutable, though they may not be perfectly substitutable. Even physically identical products might be packaged differently, marketed differently, or ultimately targeted at different consumer groups, but they would nonetheless remain "able" to be substituted. For example, two identical bottles of aspirin (with the same contents, price and packaging) are perfectly substitutable, whereas two bottles of aspirin with a different package size, branding, or price, but containing aspirin of the same chemical composition are not perfectly substitutable but remain substitutable, with demand for one being influenced by the price of the other.

1.177. According to the United States, with respect to other Western distilled spirits, the Korean submission stresses difference in color as between whisky and soju as a factor that rules out competition between these products. However, the United States argues that color may not only differ as between types of spirits, but also within spirit categories. For instance rum, tequila and shochu all have clear and amber versions, yet they do not become different non-competing products because of it.

1.178. According to the United States, it is also wrong to emphasize alcoholic strength as a dispositive factor in selecting a distilled spirit. Once they are prepared, many mixed drinks, such as vodka cocktails have a lower alcohol content than straight soju. As determined by the panel in Japan - Taxes on Alcoholic Beverages II, a difference in the alcoholic strength of two products "did not preclude a finding of likeness, especially since alcoholic beverages are often drunk in diluted form."

1.179. The United States agrees that flavour and aroma are important factors in selecting a distilled spirit, but again argues that a different flavor hardly precludes substitutability between classes of distilled spirits such as whisky, soju, vodka and rum. Indeed, the range of flavours and aromas within a class of spirit have as much of an impact on consumer choice as does the range of flavours and aromas between classes of spirits. In fact, soju itself is available in different flavorings, such as honey and wood. Under Korea's theory, different flavours of soft drinks such as Coke and Fanta do not compete, but it is doubtful that anyone familiar with the market would agree.

1.180. The United States further argued that Korea's invocation of so-called distinguishing physical characteristics for other Western distilled spirits is entirely arbitrary. The United States argues that Korea's citation of physical characteristics to distinguish soju from other Western distilled spirits are the same characteristics that happen to distinguish various kinds of Korean soju from each other. For example, Jinro's promotional material for its premium soju brands on its Internet home page cites factors such as a "rich smooth taste" and oak flavoring, the same factors Korea relied on in its first submission to distinguish soju from whiskey.²³⁵ According to the United States, Jinro's descriptions in fact confirm the competitive relationship between soju and Western spirits.

1.181. The United States further argues that, with respect to the taste and sensation of the products, it is not the raw materials that are responsible for the so-called "stinging sensation" supposedly imparted by soju. If this were true, then all spirits could and can claim a "stinging sensation in the mouth and throat", since the raw materials are similar in many other spirits. For example, Archer Daniels Midland Company's grain neutral spirits are used for both soju and Smirnoff vodka (produced in the United States). The claimed unique "cold" mouth feel may come from the fact that soju, like vodka, is usually refrigerated before consumption. Moreover, although Korea cites to "harshness" as a unique desirable characteristic of standard soju compared to the smoothness or mildness of Western-style spirits, as the representative from the European Community pointed out, the advertisement in its submission for Jinro Bisun, produced by Korea's largest producer of soju, boasts how it is a "mild" one, the same way Western-style spirits are marketed.

1.182. The United States also challenges the Korean claim that soju is uniquely suited to spicy food and unlike Western spirits, is consumed exclusively during meals in Korean restaurants or at home. According to the United States, that does not mean that Western-style spirits are not equally suitable for such use. Other countries also have hot and spicy food, and consume distilled spirits other than soju. For instance, in the United States, when consuming hot and spicy Mexican food, it is common to consume tequila in the form of a Margarita. In Poland, vodka is drunk with herring. Consumers' habits are not fixed, and can change with the introduction of alternative products. Indeed, in Korea a considerable proportion of imported Western-style distilled spirits is not consumed in bars and posh hotels, but at home, similar to soju. The United States says it understands that after dinner at home, both soju and Western- style spirits are an option.

1.183. The United States argues that, as the GATT drafting history demonstrates, in order to be directly competitive or substitutable, products need not share a majority of physical characteristics, and a basic commonality of physical properties is sufficient. The basic physical properties of soju and Western distilled spirit categories are essentially the same: all are concentrated forms of alcohol that are produced through distillation and used for human consumption. Their variations -- distillation method, appearance, taste, alcoholic content, and raw material inputs -- do not create any particular product not substitutable for the other. The 1987 Japan liquor panel did not consider the minor variations in distilled spirits important, instead stressing the flexibility of use: "Alcoholic drinks might be drunk straight, with water, or as mixes. . . . [T]he flexibility in the use of alcoholic drinks and their common characteristics often offered an alternative choice for consumers leading to a competitive relationship."²³⁶ According to the United States, this approach, consistent with the consideration by the drafters that oranges and apples are competitive, confirms that it is most appropriate to consider a broad commonality of physical characteristics.

End Uses

1.184. According to the United States, in examining the current uses of distilled spirits in the Korean market, it should be recalled that already both soju and Western-style distilled spirits are sold and advertised side by side. In the Korean Air duty free magazine²³⁷ (US Exhibit D), presumably catering to both Korean and foreign travellers, Johnnie Walker Blue Label, Johnnie Walker Gold Label and Moon Bae-Sool soju are advertised on the same page. Moreover, as shown in the pictures in U.S. Exhibit G, soju and Western-style spirits are sold together in a range of retail establishments. In the first picture, Cherry 15 is next to Seagram Extra Dry Gin, and Alexander vodka is next to Korean premium soju (aged in oak) in a convenience store. In the fourth picture, Something Special Scotch whisky is next to Kim Sat Gat premium soju in a Seoul supermarket. The fact that these products are sold through the same retail channels is important evidence of direct competition in the market place.

1.185. The United States argues that Korea has implied that soju, because it is less expensive, is marketed to an entirely different group of consumers than Western spirits and therefore is not directly competitive with Western spirits. The United States disagrees with that implication. In fact, according to the United States, given the greater degree of availability of Western spirits, it is more likely that many Koreans will consume both soju and Western spirits, differentiating the timing, degree and occasion of their consumption mix based on personal preferences and the relative prices of the two products. Moreover, The United States agrees with the representative of the European Communities that, as shown in Attachment 6 of the Korean submission, soju is also aimed at businessmen in shirts and ties, and not farmers or factory workers. This is exactly the same group of people Western spirits companies are targeting -- Korea's middle and upper class professionals.

1.186. The United States also claims that Korea's narrow approach to end uses looks at whether a distilled spirit is consumed before, during or after a meal; as a mixed drink, with ice, or straight; and what type of food is served at a restaurant. In fact, Korea attempts to draw strict categories where none exists in the Korean market or any place else US manufacturers are familiar with. Before addressing the issue of mealtime consumption, the United States recalls recall that Korea says little regarding other sub-categories of usage in which even Korea admits Western distilled spirits compete, such as consumption in bars, after meals, and in the home, or providing as a gift. Korea, does, however, concede in its answer to U.S. question 3, that some standard soju will also be consumed at home without a meal, as are western spirits.

1.187. According to the United States, Korea's narrow approach to end use is also undermined by the GATT drafting history citing to apples and oranges as directly competitive products. As stated by Korea, "A member of a group of products that are related in consumption need not be a substitute for all other members of the group -- just for some other members of the group." In its second submission Korea refers to the U.S. citation of differing end uses for apples and oranges as evidence that such products may not be competitive or substitutable. Although it is unlikely that oranges would be substituted as filling for apple pies, this does not preclude their substitutability for several other end uses, such as fruit juices, fruit jellies and jams, and overall fresh fruit consumption. Differences in product end uses may vary, but they do not erect a wall preventing substitution between differing types products within the same category.

1.188. The United States also argues that Korea's focus on differences in consumption within the on-premise and off-premise market segments is not appropriate for purposes of Article III:2. On-premise consumption covers a broad range of establishments that must be considered together for purposes of substitutability, particularly since the availability of Western spirits has continually expanded. In addition, in the on-premise consumption segment, serving particular distilled spirits is up to the discretion of the owner of the particular establishment. In general, establishments such as inexpensive Korean restaurants that appeal to low-income patrons are less likely to serve expensive high-end distilled spirits, and vice versa. However, with all of the variability in types of restaurants, bars and night clubs in Korea, it is not realistic to assume distinct rigid categories with respect to marketing and distribution of distilled spirits, and Korea has shown no support for such allegations.

1.189. The United States further argues that in the off-premise sector, making generalizations about particular kinds of stores would be equally incorrect. In Korea's retail sector, virtually all types of spirits are available in all types of establishments, yet in different proportions. For instance, a small family-owned shop will have cases of soju available and a few bottles of premium Scotch whisky on the shelves. Conversely, upscale department stores, such as Lotte and Shinsegae in downtown Seoul, will carry a preponderance of imported spirits, and may only have a small display for domestically-produced soju. The wide range of availability of imported spirits is clearly established in the Hankook study submitted by the EC as well as photographs of a range of retail establishments in which soju and Western spirits share the same store shelves submitted by the United States. Korea's argument that such placement is unimportant because in one photograph submitted by the United States, Gillette shaving foam is also apparent, overlooks the fact that the shaving foam is on a neighbouring set of shelves, not in the same group with the distilled spirits.

1.190. The United States notes that Korea's market for imported distilled spirits has been open for less than a decade, and is not a mature market with respect to consumer's awareness of, and receptivity toward, different types of spirits, their uses and places of consumption. Even though there are some differences in the methods of consumption compared to other countries, such as consumption of soju with Korean meals, it is wrong to assume, that end use must be identical for all uses under Article III:2. Given the fact that the spirits industry has been barred from the Korean market until recently, the industry has not had the opportunity to address and capitalize on every possible usage of its products in Korea. However, since all distilled spirits are fundamentally interchangeable, as Western spirit products become more familiar to Korean consumers, it is expected that methods of consumption will continue to expand.

1.191. The United States claims that with respect to on-premise consumption, Korea has also implausibly maintained that there is no overlap between the types of restaurants or bars that serve soju, and those that serve imported spirits. In fact, in its answers to U.S. question 4, Korea has already conceded at least some potential overlap between distilled soju and Western spirits when it recognizes that soju is offered in so-called "very expensive and traditional Korean and Japanese restaurants." Given its earlier statement that Western spirits are sold in Japanese restaurants, Korea minimally acknowledges substitutability in such restaurants.

Price

1.192. With respect to Korea's arguments on the different prices of the various products, the United States claims that although Korea glosses over the concept of substitutability, it does not go so far as to claim that the existence of large pre-tax price differentials precludes two products from being substitutable. In fact, Korea acknowledges such substitutability when it states that standard soju and premium soju, two products with a typical price differential of more than 100%, are considered "close substitutes" in Korea.

1.193. The United States notes that the substitutability of alcoholic beverages in a wide price range is not uncommon. Moreover, a range of prices exists within several product types. For example, whisky prices can range from ten to a hundred dollars for a bottle. The Korean submission frequently cites whisky as 11 times higher in price than soju, ignoring the price ranges within the whisky product category. For example, pre-tax imported whisky prices (for 375 ml bottles) range from over 5,000 won for premium scotch whisky to over 3,000 won for standard scotch whisky. Thus, at the extremes, imported scotch whisky pre-tax prices are 7.2 to 12 times higher than standard soju, not including domestic bottled whisky which is 6.3 times higher. Notably, imported standard scotch whisky pre-tax prices are only 3.6 times higher than premium soju.

1.194. The United States further argues that price differences are not of themselves evidence of a lack of actual or potential competition. Changes in consumer purchase behaviour are dependent on relative price changes, not absolute prices between competing products. Constructing a price demarcation where products with relative prices exceeding a specific threshold are considered not directly competitive or substitutable ignores the reality that these products are arrayed along a price continuum available to consumers in the marketplace. The availability of products across a spectrum of prices in the Korean market attests to a commensurate range of differing consumer tastes and preferences for distilled spirits, as well as the desire of individual consumers to vary their individual consumption choice on the basis of occasion, place of consumption and other factors. Moreover, purchases of distilled spirits, unlike purchases of items such as automobiles or homes, occur on numerous occasions over a relatively short time frame. Consequently, at the margin, changes in relative prices between spirits, such as soju and Western spirits, may alter some individual consumption decisions, but not all.

1.195. Furthermore, the United States notes that Korea's price analysis devotes substantial attention to comparisons of standard soju and the more upscale whisky products, giving limited attention to the variability of prices within spirit categories such as whisky and soju. For example, the pre-tax price of premium soju is two to three times higher than standard soju, and in some instances distilled soju pre-tax prices exceed those of Western spirits. In comparing soju and whisky weighted average prices, the United States claims that Korea's analysis failed to account for the variation of prices among differing whisky products. Furthermore, when adjustments are made for alcoholic content, pre-tax price differences between standard soju and Western spirits range from 3 to 6 times higher and for premium soju, 2 to 3 times higher.

Broader Substitutability of Distilled Spirits

1.196. The United States argues that there is evidence that Korean Consumers are not substantially different from other consumers around the world in the ways that they form tastes for alcoholic beverages. According to the United States, the US, EC and Korean submissions show that the advertising for the products is similar and aimed at similar audiences. In fact, the advertisements would not be out of place in Western magazines except for the Korean print. In addition, in Korea, as in the rest of the world, distilled spirits are sold in stores, bars and restaurants, among other locations. In Korea, soju and Western spirits are purchased from the same shelves in different retail outlets. Thus, there is m market evidence supporting the US statements that all spirits should be considered as in competition because they compete for consumers' spending on various products within the category of alcoholic spirits.

1.197. The United States argues that Korea's implication that all distilled spirits are not a recognized category of competing goods is contradicted by both international and Korean practices. There is a reason that the international convention on the Harmonized System, to which Korea is a party, has grouped all distilled spirits, including soju, under the same customs heading. This classification reflects the fact that on the international scene, distilled spirits are considered a distinct product group that includes soju. One need only look at a recent issue of an industry trade journal, Drinks International²³⁸ which puts out a list of the top spirits brands in the world each year, to see that soju is considered a competitor to all other distilled spirits in HS heading 2208. On page 35 of the March 1998 edition, one sees that Jinro Soju is ranked the number one spirits brand in the world.

1.198. The United States notes that Korean government itself groups imports of soju with other distilled spirits in its implementation of the Harmonized System, and groups all alcoholic beverages together -- soju, beer, vodka, whiskey, sake etc. -- in a single liquor tax law. In its responses to our questions, Korea admits that sake and beer compete with soju. It is self-serving to insist that the other products in its liquor law (such as imported distilled spirits) do not compete, despite their being grouped together in Korea's own laws.

1.199. According to the United States, several other factors support what should be the obvious conclusion that Korean consumers can be presumed to recognize the similarity of soju and imported spirits. Western spirits and Korean soju are sold alongside each other in retail outlets, and their advertisements are aimed at a similar clientele with similar sales pitches. The development, by Korea's own soju producers, of bottling operations for imported whiskey is also significant. These producers have obviously recognized that the products compete and are taking advantage of the distribution channels already developed for traditional soju brands.

1.200. The United States argues further that Korean manufacturers have advertised soju in the same international media as Western distilled spirits and seek to take advantage of characteristics of soju that are similar to those of Western distilled spirits. The United States refers again to a recent edition of Drinks International. Page 14 of the March 1998 issue contains an advertisement for Jinro Soju, which asks potential purchasers to "experience its unique smooth taste" and boasts of its "incomparable versatility for cocktails." It even shows soju in a glass "on the rocks" with a lemon slice and a maraschino cherry. Obviously, Korea's top soju producer, Jinro, considers soju substitutable for Western distilled spirits and eminently suitable for a variety of their uses -- like Western spirits, not only drunk straight in a shot glass, but also "on the rocks" or mixed with other products to make cocktails.

1.201. The United States argues that Jinro's recognition of soju's substitutability with other spirits in its international export markets is not reconcilable with the position of the Korean government in these proceedings that soju is not "like" or "directly competitive or substitutable" with any Western distilled spirit. Overseas, Korean soju competes on a level playing field with all other distilled spirits, but at home Korea rejects any notion that the products are in competition, and insists on a reading of Article III that would protect soju through burdensome taxes on imports. The United States considers that this position finds no support in the letter or spirit of Article III:2.

1.202. The United States considers that there is also some empirical evidence supporting the general notion that consumers budget for alcoholic beverages separately than other goods.²³⁹ This study of consumers (Alley et al.) reveals that the consumption of alcoholic beverages is separate from their consumption of other goods. Hence, "after compensating for income changes, consumption of other goods is unaffected by price changes" of alcoholic beverages.²⁴⁰ Furthermore, in Alley and in two additional studies²⁴¹ there is evidence of substitutability between domestic and imported spirits in Canada and Finland, respectively. Accordingly, there is further support for the conclusion that all imported spirits compete with soju in the distilled spirits market in Korea.

To continue with Korean Market Developments

²³⁵ US Exhibit Q.

²³⁶ See panel Report on Japan - Taxes on Alcoholic Beverages I, supra., at para. 5.7.

²³⁷ US Exhibit D.

²³⁸ US Exhibit P.

²³⁹ The United States cites A.G. Alley, D.G. Ferguson, K.G. Stuart, An Almost Ideal Demand System for Alcoholic Beverages in British Columbia, 17 Empirical Economics 416 (1992).

²⁴⁰ Id. at 414.

²⁴¹ The United States cites Andrikopoulos, et al., The Demand for Domestic and Imported Alcoholic Beverages in Ontario, Canada: A Dynamic Simultaneous Equations Approach, 29 Applied Economics 945-953 (1997); Holm, Alcohol Content and Demand for Alcoholic Beverages: A System Approach, 20 Empirical Economics 75-92 (1995).