4. Actual (or Potential) Competition

1.234. Looking carefully at the evidence and arguments that the complainant have presented, Korea submits that the arguments are misleading, flawed, or insufficient to meet the complainants' burden of proof. According to Korea, therefore, the complainants have not established that there is a direct competitive or substitutable relationship, actual or potential, between soju and the imported distilled spirits that are within the scope of this dispute.

(a) Physical Differences

1.235. Korea points out that there are important physical differences between the products at issue. These differences have an impact on consumer preferences and hence on the competitive relationship between the products. According to Korea, the complainants have attempted to trivialise those differences by looking at the products from a great distance, such that they can say that all the products have 'essentially the same characteristics'. In Korea's view, however, even a slight physical difference might be enough to render two products not competitive. For example, if a consumer does not like the taste of one particular additive, its addition to a product will eliminate that consumer as a potential buyer -- even if that additive does little to change the appearance or chemical composition of the product.

1.236. Korea also points out that the complainants display ignorance of the very real possibility that two products that are ostensibly similar might not compete. For example, Korea cites the US argument that 'under Korea's theory, different flavours of soft drinks such as Coca-Cola and Fanta do not compete, but it is doubtful that anyone familiar with the market would agree.' Korea disagrees with this. According to Korea, the directorate of the European Commission responsible for competition law matters has more than once drawn a distinction between cola-flavoured carbonated soft drinks and other types of soft drinks, deciding in the recent Coca-Cola/Amalgamated Beverages merger that the relevant market was the market for cola-flavoured carbonated soft drinks, explicitly excluding other flavours of carbonated soft drinks.²⁵² National competition authorities have come to the same conclusion.²⁵³

1.237. Korea argues that, the European Communities, in its discussion of physical differences, tries to take consideration back to the moment of distillation, at which point, it argues, 'all spirits are nearly identical' because 'all of them are concentrated forms of alcohol.' Korea argues that firstly, the principal test for whether products compete is the marketplace. Spirits are not sold at the point of distillation. Moreover, the difference in physical characteristics, starting with a difference in raw materials, is not negated by distillation. In addition, there are important post-distillation processes that also have an impact on the physical characteristics of products. Brown spirits, for instance, are generally matured in wooden casks and derive their flavour from this process, and from the original distilled ingredients.

1.238. According to Korea, the analysis at the point of distillation is meaningless as the subsequent addition of additives or ingredients can, and in the liquors discussed here, does, make a crucial difference for consumers when they choose a particular liquor in a particular market, such as the Korean market. Korea concludes, therefore, that although all spirits may be 'nearly identical' at the point of distillation, the most casual observation clearly shows that does not mean that drinkers are indifferent between them.

1.239. According to Korea, if physical characteristics at the point of distillation were meaningful, that would lead to the conclusion that products as disparate as fuel or pharmaceutical products are also directly competitive and substitutable with, for example, vodka.²⁵⁴

1.240. Korea also asserts that the complainants try to use evidence about production process to make up for the essential evidence about markets that is missing from their presentations. According to Korea, no evidence offered by the complainants shows that 'distilled spirits' is a group that is relevant to consumer choice in the Korean market. Korea asks: Why is 'distilled spirits' such a group rather than 'brown distilled spirits' or, 'white distilled spirits'? Why not 'alcoholic beverages' - spirits, wine and beer? Why not 'cold drinks', alcoholic or not?

1.241. In Korea's view, any one of these might be a grouping relevant to consumption - or might not be. Whether it is or not depends on the tastes and preferences of consumers. According to Korea, the effects of a failure to adequately research relations between products in consumption cannot be escaped by reference to modes of production. In particular, it cannot be demonstrated that distilled spirits is a relevant grouping for consumers on the basis that 'having essentially the same characteristics, soju and other distilled spirits and liqueurs are objectively apt to serve the same end uses'.

1.242. According to 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. Korea states that a high-end Ferrari and a low-end Renault Clio are both motor cars, and could be seen by a person of sufficiently limited imagination as 'having essentially the same characteristics' (four wheels, one engine, steering wheel) and to be 'objectively apt to serve the same end uses'. The fact that both are motor cars, though, is not enough to allow deduction of elasticity of substitution between them. Indeed, in the case of Ferraris and Clios, it seems very likely that the elasticity in substitution between the two products is nil: that if the price of Clios changes by 1 per cent (or 10 percent or 100 per cent), the effect on demand for Ferraris will be zero.

1.243. Likewise, according to Korea, given the notable differences between standard soju and the western-style liquors there is no reason to assume that there would be elasticity of substitution between them. The complainants need to show this, but have failed to do so.

1.244. Korea also claims refers to the EC claim that 'according to Korea, the main difference between soju and gin is that gin is flavoured with juniper berries.' According to Korea it is true that Korea argued that juniper imparts a very particular flavour to gin that some consumers do not like. However, this was one of many differences, not least differences in price and end use, that Korea pointed out with respect of gin.

1.245. Korea notes, furthermore, that in interpreting the term 'substitutable', the United States, rather than proving that for Korean consumers a western-style liquor like whisky is directly²⁵⁵ substitutable for standard soju, despite the very considerable price difference, cites the example of bottled water and tap water, two products that ostensibly have similar physical characteristics and similar end-uses.

1.246. According to Korea, products that conceivably can be substituted for each other yield a very broad field. Korea states for example, that one could say that woollen sweaters and coal can be substituted for each other. If you are cold, you could either put on a sweater or throw another lump of coal on the fire. Washing machines and socks could be substitutable. The easier and cheaper the availability of laundry facilities, the fewer socks a person needs in order to have a constant clean supply. With this view of substitutability, baby carriages and wheelbarrows could be substitutable!

(b) Price

1.247. According to Korea, on the basis of the facts provided in the complainants' Dodwell Study, it is apparent that there is a huge discrepancy in the pre-tax prices of the products at issue. Korea states that taking whisky and standard soju as an example, the Dodwell Study data shows that the ratio of their prices varies from 6.3 to 12 times, with no overlap in price at all. Furthermore, according to Korea, the weighted average figures that Korea provided in its first submission showed that whisky is on average 11 times more expensive than standard soju.²⁵⁶ Large pre-tax price differences are shown for all the imported liquors at issue. This casts serious doubts on the existence of a directly competitive and substitutable relationship between these products, and highlights an important difference between this case and Japan - Taxes on Alcoholic Beverages II .

1.248. Korea points out that there are enormous pre-tax price differences. Based on the figures provided by the Dodwell study, Korea found the following pre-tax price differences:

• Standard soju v. whisky(premium) ratio of 1 to 12
• whisky(North American) ratio of 1 to 10.8
• whisky(standard scotch) ratio of 1 to 7.2
• whisky(bottled in Korea) ratio of 1 to 6.3
• brandy/cognac ratio of 1 to 19.2
• vodka ratio of 1 to 5.7
• gin ratio of 1 to 5
• rum ratio of 1 to 6.2

1.249. Korea indicates that the Dodwell study data shows that the ratio of price difference varies from 6.3 to 12 times with no overlap in price at all. Furthermore, according to Korea, the weighted average prices showed that whisky is 11 times more expensive than standard soju. In response to the European Communities' claim that weighted average prices are "meaningless," Korea states that a weighted average price is an accepted means of getting a typical price for a product. Large pre-tax price differences are shown for all the imported liquors at issue. This casts serious doubts on the existence of a directly competitive and substitutable relationship between these products, and highlights an important difference between this case and Japan-Taxes on Alcoholic Beverages II.

1.250. Korea considers that pre-tax price differences of this magnitude must be taken into account in discussing whether it can reasonably be argued that any of these products are "like" or directly competitive or substitutable. Korea argues that these price differences refute any argument that it was taxes that have "frozen" consumer preferences. Furthermore, argues Korea, as to directly competitive or substitutable products, it must be addressed whether, in the face of these price differences prior to the application of tax, the tax differentials at issue in this case can be said to "afford protection" to domestic production.

1.251. Korea argues that the complainants have been unable to provide a response to the proposition that large pre-tax price differential leads to the absence of a directly competitive or substitutable relationship. Rather than addressing the price discrepancy between standard soju and the imported liquors head on, the complainants attempt to divert the Panel's attention to premium diluted soju, which is the somewhat more expensive variety, representing some 5% of total diluted soju volume.

1.252. Korea refers to the EC argument that using pre-tax prices for the purposes of comparison does not remove the "distortive effect from the disputed taxes", because the tax has kept low-priced imports from entering the market. Korea reminds the parties that, unlike Japan, it has ad valorem taxes. In Korea's view, while it might be argued that a high specific tax would keep low-cost products out of the market, an ad valorem tax is just that - it is linked to the price of the product, and therefore, the lower the price, the lower the tax amount.

1.253. Korea states that in determining whether two products are directly competitive or substitutable products, price cannot be excluded. In Korea's view, it is obvious that when one product is many times more expensive than another, it is difficult to argue that those two products are in competition. Korea notes that the US and EC competition authorities consider price as one of the most relevant factors.²⁵⁷ Korea also recalls that during the Panel meetings, the EC representative stated that competition law analysis was relevant to an analysis of Article III:2.

1.254. Korea draws attention to the importance of the reasoning in Japan - Taxes on Alcoholic Beverages II, wherein it was stated that:

1.255. Korea argues that common sense tells us that the larger the price differences between two products, the less influence a change in the price of one will have on the demand for the other. In Korea's view, the complainants have not been able to overcome this common sense presumption.

1.256. According to Korea, the price difference between premium diluted soju and standard diluted soju is far less important than the price difference between standard diluted soju and western-style spirits. Korea claims that premium diluted soju is, on average, less than 80% more expensive than standard diluted soju,²⁵⁹ and the difference in price reflects the other very small differences between premium and standard diluted soju.

1.257. Korea also refers to an assertion by the European Communities, which in Korea's view is unfounded: 'western-style spirits are more expensive, to a large extent as a result of discriminatory taxation. If western-style spirits were taxed as soju, they would be less expensive and Korean consumers could afford to drink them with meals more often.' Korea's argument is that western-style spirits are indeed more expensive, but the rest of this statement does not follow.

1.258. Korea points out that using the Dodwell data provided by the EC, the pre-tax price differences between these beverages is great. These price differences would thus be maintained if tax rates were harmonised, and they are too great for it to follow that consumers would suddenly find imported alcoholic beverages an affordable alternative to diluted soju for meal-use, even disregarding the matter of taste. According to Korea, even the European Communities cannot think that its evidence shows that the addition of a few hundred won to the price of a bottle of diluted soju will have any substantial effect on demand for, for example, whisky.

1.259. Korea also states that another unfounded allegation appeared in the EC oral statement, where it stated that:

1.260. Korea states that it is probably true that if the price of whisky falls, the volume of whisky purchased will increase. According to Korea, that the quantity of whisky purchased is responsive to the price of whisky, does not, however, further the EC and US case. According to Korea, they need to show that this increase was at the expense of soju - that the demand for soju is responsive to the price of whisky.

1.261. Korea argues that if whisky sales increase at a greater percentage rate than soju sales, the share of whisky in the hypothetical 'soju-whisky market' will rise, and that of soju will fall. In Korea's view, that is entirely consistent with sales of soju rising at the same rate both before and after the increase in the share of whisky. The increase in whisky market share and the fall in soju market share says nothing about whether whisky grew at the expense of soju.

1.262. According to Korea, the point can be underlined by taking two goods that most people will accept to be unrelated - say whisky and floor polish - and aggregating them into an artificial market. In that 'market', the share of whisky might rise: whisky sales might grow at a higher percentage rate than sales of floor polish. But this does not prove or even suggest that whisky sales have grown at the expense of floor polish sales. Indeed, since the two products are, ex hypothesi, unrelated, it is clear that they have not.

1.263. In Korea's view, this 'proof' that whisky has grown at the expense of floor polish is in no way different from the 'proof' offered by the European Communities and the United States that whisky has grown at the expense of soju in the 'whisky-soju market'. Whisky's gain of market share in the 'whisky-soju market' shows that sales of whisky have grown at a faster percentage rate than soju: it shows no more and no less.

1.264. According to Korea, the arguments of the European Communities and the United States that market evolution in Korea supports their claim that whisky and soju are substitutes deserve no weight. In Korea's view, sales in whisky have certainly increased in Korea as its price fell, which is to be expected. Korea argues that in itself that fact has no bearing on the case. What the European Communities and the United States need to show is that the increase in sales of whisky was at the expense of sales of soju, which in Korea's view, they have not done.

1.265. Korea argues that the complainants display an interesting approach to the alcoholic strength of products. According to Korea, the complainants argue that a difference in alcoholic strength should not be taken into account in certain instances (such as, it should not affect the determination of whether products meet the strict 'like' product criteria). In Korea's view, the complainants only analyze price in this way in a last-ditch attempt to deal with evidence that is very damaging to their case. Korea further argues that the EC cannot bring itself to compare alcohol-adjusted prices of an average standard soju price -- instead, it compares an atypical bottle of standard whisky to a bottle of premium soju, and still comes out with a 2 to 3 times price difference.

1.266. Korea argues that the EC is focusing on exceptions in describing the Korean market. When it comes to the price of premium soju in this case, they take the most expensive premium soju brand (Kimsatgat) as representative. The sales volume of such high-priced premium soju is minimal. Korea further argues that he European Communities takes inexpensive whisky as an example, which is atypical as well.

1.267. Korea also argues that the calculations of the European Communities are misleading. When taking the price per degree of alcohol based on weighted averages, so as to obtain a representative price, whisky is still 7.96 times the price of standard soju per degree of alcohol. Looking at premium soju alone, whisky is 4.71 times the price of premium soju per degree of alcohol.²⁶¹ Again, these are pre-tax prices.

1.268. Korea asserts that contrary to its contention in Japan - Taxes on Alcoholic Beverages II, that pre-tax prices were within a relatively short range, the European Communities now claims that there were actually significant price differences in Japan, and submits figures that are supposed to show this. According to Korea, the European Communities even goes so far as to suggest that the Japan - Taxes on Alcoholic Beverages consciously disregarded substantial price differences.²⁶²

1.269. Korea observes that prices of western-style liquors and the Japanese shochus, once the tax is removed, appear to be rather close, with a number of overlaps.²⁶³ Furthermore, Korea notes that Japan chose not to dispute the EC's contentions about prices and price competition. However, rather than speculating about the evidence, and the arguments presented to the panel in Japan - Taxes on Alcoholic Beverages, Korea submits it is safer to recall the explicit finding by that panel:

To continue with Price

²⁵² Korea cites Commission Decision of 22 January 1997 declaring a concentration to be compatible with the common market and the functioning of the EEA Agreement, Case No IV/M.794 Coca-Cola/Amalgamated Beverages, 1997 OJ L 218, p. 15.

²⁵³ Korea cites for example Decision no 96-D-67 of the French Conseil de la Concurrence (Competition Council) of 29 October 1996, Coca Cola Beverages.

²⁵⁴ Korea alleges that the EC stated during its oral statement (p.12) that: neutral spirits are a raw material which can be used to produce a variety of alcoholic beverages, including for example vodka and gin, as well as other products, such as fuel or pharmaceutical products.

²⁵⁵ Korea maintains that the United States, the European Communities and Korea are in agreement that the word 'directly' applies to both 'competitive' as well as to 'substitutable.

²⁵⁶ Korea refers to Attachment 5 to Korea's first submission.

²⁵⁷ Korea notes that in its written rebuttal the EC states that only consumers' responses to changes in relative prices are relevant and not absolute price differences). In competition law, price is one of the assessed criteria to determine whether there is competition between products. In such an analysis, prices are assessed both in nominal and relative terms (see Commission Decision of 22 July 1992, Case No. IV/M 190 Nestle Perrier, OJ L 356, p.1). In Korea's view, the difference in nominal terms is of such a degree that it renders improbable any change in the consumers' responses resulting from relative price changes.

²⁵⁸ Panel report, para. 6.31.

²⁵⁹ As appears from Korea's answer to question 5 of the second set of questions from the Panel, the weighted average price of standard diluted soju (excluding premium soju) is 306.58 won, while the weighted average price of premium diluted soju alone is 539.70 won.

²⁶⁰ See EC oral statement, p. 6.

²⁶¹ Calculations based on data from the National Tax Administration.

²⁶² EC written rebuttal, at para. 130.

²⁶³ See Annex 2 to EC written rebuttal.

²⁶⁴ Panel Report on Japan - Taxes on Alcoholic Beverages, para. 4.82.