(f) Place of Consumption

1.89. The European Communities refers to Korea's claim that soju is typically drunk in places where western-style spirits are not available yet, such as "Korean restaurants, mobile street vendors and inexpensive Chinese restaurants".

1.90. The EC view is that it is true that western-style spirits still have little presence in certain types of outlets, but that cannot be explained, as Korea pretends, simply as the result of "consumers' tastes and habits." In the EC view, it is surely not coincidental that those outlets where the penetration of western spirits remains the lowest are also the least expensive.

1.91. According to the European Communities, Korea's submission totally disregards the existence of a clear trend towards wider availability of western-style spirits. The EC position is that only a few years ago, western-style spirits could be found only in upmarket restaurants and entertainment establishments. Since then, as shown by the Hankook survey included in the EC First Submission, western style-spirits have gained considerable distribution penetration and are now available at a wide range of outlets. According to the European Communities, the continued application by Korea of protective internal taxes stands as an obstacle to that trend.

1.92. The European Communities states that in this regard, it is instructive to compare the Korean and Japanese markets. According to the European Communities, only a decade ago, Japanese izakayas (the equivalent of traditional Korean restaurants) used to serve only shochu, sake and beer, whereas western-style snack bars would serve western drinks but not shochu. Today, shochu and western style spirits are allegedly usually available at both izakayas and snack bars.

1.93. The European Communities further argues that it is important to note that a considerable and growing proportion of both soju and western-style spirits is purchased for consumption at home. In the EC view, this is totally ignored in Korea's submission. Specifically, the EC industry estimates that between 30 % and 35 % of western spirits and more than 20 % of soju are consumed at home.

1.94. According to the European Communities, as evidenced by the Hankook survey, western-style spirits are sold for home consumption through the full range of retail channels, where they compete head-on with soju.¹⁹⁸ The European Communities further claims that as shown by the photographs attached to its First Submission, retail establishments of different types often display soju and other distilled spirits side-by-side on the same shelves.

(g) Gift Giving

1.95. The European Communitie notes that in the case of distilled soju, Korea advances the additional argument that its main use is for gift giving. The EC view is that although this may be so, western-style spirits, and in particular whisky and brandy, are also often offered as a gift. The advertisements for Robbie Dhu and Johnny Walker Gift Set includes two advertisements promoting specifically the purchase of whisky for gift-giving.¹⁹⁹ According to the European Communities, Korea incurs in an even more embarrassing contradiction when, in comparing diluted soju and cognac, it states that the latter is "very suitable for gifts."

(h) Pricing

1.96. The European Communities also notes that Korea contends that the pre-tax prices of soju are much lower than the pre-tax prices of western style spirits. Korea alleges that, because of that price difference, there cannot be competition between soju and western-style spirits.

1.97. The European Communities takes issue with Korea's contention that it is sufficient to compare pre-tax prices in order to exclude "any possible distortive effect from the disputed taxes."

1.98. The European Communities states for example, that tax differentials may affect the relative importance of the different price segments within each tax category. The European Communities refers to the Panel Report in Japan - Taxes on Alcoholic Beverages II, in which it was stated that one of the consequences of a protective system of internal taxes may be to make it more difficult for the cheaper brands of the more heavily taxed products to enter the market.²⁰⁰ The EC view is that in Korea, this effect is attested by the fact that premium brands account for a disproportionate share of imports,²⁰¹ whereas the cheapest brands of western spirits are virtually absent. On the other hand, lower taxes have given to soju producers an advantage to target in particular the low end of the market.

1.99. The European Communities also argues that protective taxes limit the sales growth of the more heavily taxed imported categories and, as a result, keep their unit costs at an artificially high level, as compared with less taxed domestic products which are sold in much greater volumes.

1.100. In any event, according to the European Communities, absolute price differences are not of themselves determinative of whether two products are "directly competitive or substitutable." In the EC view, what really matters is the consumers' response to changes in relative prices. The Dodwell study allegedly provides evidence of that type of response. The European Communities notes that Korea has criticised the supposed methodological flaws of the Dodwell study, but it has not put forward any contrary evidence showing that there is no significant degree of cross-price elasticity between soju and western spirits.

1.101. According to the European Communities, in the present case, absolute price differences are even less determinative in view of the nature of the products concerned. The European Communities adds that spirits are not like cars, which are purchased by most consumers in developed countries only once every four or five years. Spirits are consumable products, which can be purchased many times over a relatively short period of time. According to the European Communities, Korea's argument assumes that each Korean consumer drinks always the same type of spirit. In practice, however, most consumers, even if they prefer a certain type of spirit, may drink also other spirits depending not only on the occasion but also on the prevailing prices for each of them.

1.102. In the EC view, the above reasons explain why the two Panels on Japan - Taxes on Alcoholic Beverages did not take into account absolute prices differences, even though such differences were also substantial in the Japanese market.²⁰²

1.103. The European Communities further argues that at any rate, the comparisons made by Korea grossly overstate the actual price differences. The pre-tax prices for diluted soju shown in the Dodwell study, on which Korea has based its price comparisons, are prices for standard diluted soju. The pre-tax price for premium soju is between two and three times higher.²⁰³ In the EC view, if an adjustment is made to take into account the differences in alcohol content, the pre-tax price for a bottle of standard whisky is about two to three times higher than the pre-tax price for a bottle of premium soju.²⁰⁴

1.104. The European Communities also argues that the differences between the pre-tax prices for distilled soju and standard diluted soju are even larger than the differences between the pre-tax prices for standard whisky and standard diluted soju.²⁰⁵

1.105. The European Communities notes that as additional evidence, Korea submits a comparison of the "weighted average prices" for whisky and soju. According to the European Communities, that comparison, however, is meaningless. The prices for whisky vary enormously. Even if one considers only mainstream brands, the prices for standard blended whisky may be as much as ten times less than the prices for de luxe and single malt whiskies. The EC view is that although to a lesser extent, the prices for soju also vary. The pre-tax prices of premium diluted soju may be two to three times higher than the prices for standard diluted soju. According to Korea, premium soju accounts for 5 % of the total sales of diluted soju. According to the European Communities, therefore, as a result, the weighted average price for all diluted soju calculated by Korea is virtually the same as the average price for standard diluted soju.

i) Other Alleged Differences

1.106. The European Communities refers to Korea's claim that the advertising for soju is targeted at the "common man" whereas the advertising for western-style spirits targets the up-market consumer. In the EC view, the soju advertisements included in Attachment 6 to Korea's submission fail to support those allegations. They do not show farmers or labourers but rather business men in shirts and ties. According to the European Communities, Korea has carefully omitted to include in Attachment 6 to its submission any example of recent soju advertising for premium brands, which is even more clearly targeted to the up-market Korean consumer.²⁰⁶

1.107. The European Communities contends that Korea's allegations regarding the differences in advertising imply that while soju is the poor man's drink, whisky and other imported spirits are a luxury drink for the most affluent classes. In the EC view, this is totally misleading, as both soju and whisky are now widely consumed across social boundaries.²⁰⁷

1.108. The European Communities also notes that Korea makes the argument that the fact that some Korean companies which sell both whisky and soju do so under different trademarks is an indication that the two products do not compete in the same market, yet some of those companies use also different brands for premium diluted soju and standard diluted soju.

1.109. According to the European Communities, even less convincing is Korea's argument that diluted soju does not compete with vodka, gin and rum because the latter are sold in small volumes whereas soju is a mass volume product. In the EC view this is because only a few pages later Korea reverses this argument without any apparent embarrassment in order to claim that distilled soju is a "tiny niche product."

6. The Dodwell Study

1.110. According to the European Communities, until the early 1990s, western spirits were virtually excluded from the Korean market by a combination of trade barriers. In the EC view, too few sales figures are available to allow an econometric analysis of the substitution relationship between soju and Western style spirits. This leaves a survey such as the Dodwell study as the only method. In the EC view, Korea does not appear to question the rationale for the survey. However, as Korea has presented a number of criticisms to the Dodwell report, the European Communities responds to those criticism as follows-

(a) Alleged "Inconsistency" of Results

1.111. The European Communities notes that Korea points to the fact that there are a number of anomalies in the study. Notably, on a few charts there appears anomalous behaviour in terms of slightly higher soju purchases when prices for the soju increase. According to the European Communities, these anomalies are far less troublesome than Korea is suggesting:

In the EC view, this is far from being a bad result for a survey because, as Korea pointed out, it has to be kept in mind that the survey deals with fallible human beings. A perfectly consistent result from interviewing five hundred people cannot be expected, and would actually be highly suspicious. The anomalies also run counter to the implicit insinuation that the survey results are biased by the patron of the study.

(b) Standard Vs. Premium Soju

1.112. The European Communities notes that the inclusion of premium diluted soju as an object of choice biases the results upwards. According to the European Communities, this is not the case because:

(c) Choice of Respondents

1.113. According to the European Communities, the criticisms made by Korea with respect to the choice of respondents are also unfounded because:

(d) Pair-Wise Choices

1.114. The European Communities reiterates that the fact that the survey sample might have a higher preference for drinking western beverages than the Korean sales volume figures suggest is of no consequence to the validity of the results. In the EC view, what matters is that a cross-price relationship is established.

1.115. The European Communities also states that it should be mentioned in this context that the separate paring of brown and white liquors actually biases the reaction to price increases downwards rather than upwards. The European Communities refers to Korea's explanation that a preference for white liquor does not imply a preference for brown liquor over soju. A person with a preference for brown liquor might therefore not be impressed by the rising soju prices to drink white liquor, and vice versa. In the EC view, this means the price reaction of the survey will be underestimated.

(e) Single Drink Choice

1.116. The European Communities notes that Korea claims that the phrasing of the survey question may be ambiguous. In particular, Korea presumes that the question could be interpreted as a unique and non-repeated sales offer.

1.117. In the EC view, it is difficult to see why this should be the case, in particular since the phrasing of the question belongs to the standard repertoire of market surveys. However, even if it were interpreted thus, it is quite clear that people react to price changes and choose more imported liquors, at the expense of soju. This means clearly that the consumer interprets the liquors for choice as substitutes.

1.118. According to the European Communities, Korea would have preferred a phrasing "would these prices cause you to change your drinking habits?" In the EC view, this question would produce lower figures (which is actually why it is proposed). The reason for this is simply that people are less able to make a statement about permanent behavioural change. The European Communities add that the proposal of Korea also sits oddly with a remark earlier on the internal consistency of the results.

1.119. In the EC view, the conclusions of the survey are quite clear: consumers are sensitive to the relative prices of soju and other drinks and change their behaviour accordingly. This indicates very strongly that the consumers view Western liquors and soju as substitutes. According to the European Communities, Korea's critique does not affect these conclusions.

(f) Overall

1.120. The European Communities also argued that the spirits markets have two defining characteristics. First, spirits consumption is habitual behaviour in that people tend to order the same drink they ordered on a previous given occasion. Thus, behaviour changes only gradually. Second, spirits are experience goods in that they must be purchased and consumed to be evaluated by consumers. Descriptions do not suffice. Market penetration increases slowly as it is necessary to get consumers actually to try the products first. Market surveys such as the Dodwell study must be evaluated in light of these factors.

To continue with The Sofres Report

¹⁹⁸ Ibid.

¹⁹⁹ Attachment 6 to Korea's First Submission.

²⁰⁰ Panel Report on Japan - Taxes on Alcoholic Beverages II, supra., para. 6.33 (d).

²⁰¹ See the Sofres Report, at p. 26.

²⁰² According to the European Communities, in Japan - Taxes on Alcoholic Beverages I, supra., the Panel ruled that price differences were irrelevant for a like product determination. Prices are not mentioned in the Panel's analysis under the second sentence of Article III:2. In Japan - Taxes on Alcoholic Beverages II, supra., differences in prices were examined by the Panel only in connection with Japan's argument that, because tax/price ratios were "roughly the same", the products were not taxed "similarly" (paras. 6.33-6.34) but not for the purposes of determining whether the products were "directly competitive or substitutable" (paras. 6.28-6.32)

²⁰³ See EC Annex 6.

²⁰⁴ See EC Annex 7.

²⁰⁵ See EC Annex 6.

²⁰⁶ EC Annex 12.

²⁰⁷ The Sofres Report describes as follows the profile of the whisky drinkers (at p. 25):