(d) Legal Analysis

1.185. Korea notes that the complainants argue that because Korea imposes a different rate of tax on 'soju' than the rate which is imposed upon a number of alcoholic beverages, its tax system unfairly burdens imported spirits, to the benefit of the domestic spirit 'soju'.

i) The Purpose of Article III of the GATT is to Prevent Protectionism, not to Harmonise Taxes

1.186. According to Korea, Article III, perhaps more than any other provision of the GATT, steps into a realm which is traditionally the province of national sovereignty -- the domestic tax system.

1.187. Korea notes that as each Member joins the WTO, it brings with it its individual tax system, an intricate web of rules built up over many years reflecting a mixture of government objectives (ranging from social policies to the need to raise revenues).

1.188. Korea further states that no Member of the WTO has the same rate of internal taxation for all products, and the tax differences that have built up throughout the years may not always seem completely coherent. However, the argument continues, it is not the role of the GATT to attempt to render each tax system internally coherent, and nor is it the province of GATT to harmonise the tax systems of its many Members.

1.189. Korea takes note of the 1992 Panel report in United States - Measures Affecting Alcoholic and Malt Beverages in which it was stated:

1.190. Korea further argues that it should be noted that Article III is not intended to encroach upon the powers of WTO Members to pursue their legitimate objectives through their tax systems. In this regard Korea draws attention to a statement by the Appellate Body put in Japan -Taxes on Alcoholic Beverages, in which it said:

1.191. Korea, therefore asserts that Article III of the GATT is only meant to step into this sensitive realm of taxation in certain narrowly circumscribed circumstances, in order to pursue a specific objective. According to Korea, the fundamental purpose of Article III is to avoid protectionism and to oblige the WTO members to provide for equality of competitive conditions for imported products in relation to domestic products. As the Appellate Body put it in Japan - Taxes on Alcoholic Beverages II:

ii) The Analysis Required under Article III:2

1.192. Korea states that reconciling the anti-protectionist purpose of Article III:2 with the need to recognise each WTO Member's discretion to maintain and develop its own tax policy has not been an easy exercise.

1.193. Korea further states that the Malt Beverages panel construed the so-called 'aims and effects' test. Under that test, as long as the tax system introduced product distinctions for policy purposes that were unrelated to the protection of domestic production, the resulting tax differences could not violate Article III:2.¹¹² According to Korea, this approach did not find favour with the Appellate Body in Japan - Taxes on Alcoholic Beverages II. In Korea's view, the Appellate Body (as well as the Panel in the first instance) ruled that the aims and effects test strayed too far from the text of Article III:2, ignoring the distinction made between the two prohibitions contained, respectively, in the first and the second sentence of this provision. The first sentence tolerates no differentiation whatsoever between tax rates that apply to 'like' products. The second sentence imposes constraints on tax differentials that apply to 'directly competitive and substitutable products'.

1.194. Korea further argues that according to the Appellate Body, the wording of the first sentence leaves no room to inquire whether a tax rate differential, which applies to 'like' products, was really introduced with protectionist 'aims and effects'. By this argument, differential tax rates applying to 'like' products violate Article III:2.

1.195. Korea argues, however, that this does not mean that the Appellate Body was insensitive to the concerns of the WTO membership about preserving discretion to develop individual tax policies. Korea notes that in its report, the Appellate Body took great pains to underline 'how narrow the range of "like" products is meant to be'.¹¹³ Korea further notes that the Appellate Body, affirming the Panel,¹¹⁴ also clarified that 'like' products are a subset of directly competitive and substitutable. In Korea's view, this indicates that Article III:2 only comes into play when the products at issue are in a particularly close relationship (they should be at least "directly competitive or substitutable"). Where there is no relationship, or an insufficiently close relationship between two products, argues Korea, the WTO Members retain the sovereignty to impose taxes according to their own discretion.

1.196. Korea argues that the determination of whether two products are 'like' or directly competitive and substitutable products is based upon an overall appreciation, on a case-by-case basis, of their qualities and their relationship. Korea refers to one general remark regarding both the determination of 'like' products and of directly competitive and substitutable products which was made in Japan - Taxes on Alcoholic Beverages II, which allegedly emphasised that a particularly appropriate test to define whether two products are 'like' or 'directly competitive or substitutable products' is the 'marketplace'.¹¹⁵ Korea notes that while noting that there might be other means to identify the broader category of directly competitive and substitutable products, the Appellate Body upheld this reasoning, stating that the: 'GATT 1994 is a commercial agreement, and the WTO is concerned, after all, with markets.' ¹¹⁶

1.197. Korea also states that it is useful to make some preliminary comments about the threshold condition of Article III:2: i.e. how does one analyse whether there is a sufficiently close competitive relationship between the Korean sojus and the remaining imported liquors? Korea poses the question, when are products sufficiently competitive for the purposes of comparing their tax rates under Article III.2?

1.198. Korea argues that to answer this threshold question, it is useful to recall first the factors that are relevant to establish a competitive relationship between two products and then analyse the degree of competition that will trigger the application of Article III.2, first and second sentences.

iii) Relevant Factors

1.199. According to Korea, first, one must determine whether the products at issue are at least roughly comparable. Where potable liquids are concerned, it makes little sense to compare soft drinks with alcoholic beverages for example. Consumers who want a soft drink will not consider alcohol a suitable alternative.

1.200. Korea further states that within the broad category of alcoholic beverages, a further distinction can be made according to product characteristics, such as:

• Raw materials - Are products rice-based, grain-based, potato-based, tapioca-based, etc? Which additives are used? In Korea's view, these elements may influence the taste of the beverage, and the consumer's perception in general.
• Production process - Are the alcoholic drinks based on fermentation only, or on distillation? Are they aged, and if so how, before being sold? According to Korea, these are factors that may influence the final appearance and taste of the drink, but also its image in the marketplace. Ageing in wooden casks, for instance, gives a drink an 'artisanal' image, that may make quite a difference to consumers, even though objectively it is debatable whether more industrial production processes (e.g., mere storage in metal containers) result in inferior products.
• Physical characteristics - of the finished products. Are the products similar in terms of their appearance, the degree of alcoholic strength?

Other notable factors determining whether two products compete revolve around price and end use.

• Price - Are the products at issue in the same price range? According to Korea, if prices differ greatly, two seemingly similar products cannot be considered directly competitive. Korea notes that in Japan - Taxes on Alcoholic Beverages II, the complainants argued that Japanese shochu and the imported liquors were in the same price range before tax.¹¹⁷ Korea further notes that Japan does not appear to have contested this. Korea asserts that the price differences in Korea between standard soju and the disputed imported liquors are very considerable.
• End uses - Do consumers use these products in the same way? Do they drink a particular drink straight or mixed; with meals, or on other occasions? Korea states that these elements were highlighted in Japan - Taxes on Alcoholic Beverages I.¹¹⁸ Korea argues for instance, that depending on the food they usually eat, consumers will show a preference for different alcoholic drinks during meals. In countries with a spicy food tradition, consumers may favour strong alcohol that cleans the palate, when eating meals. In countries where spicy food is not a tradition, consumers will by definition have a different preference for an alcoholic drink that goes best with their national cuisine.

1.201. Korea argues that the existence, or absence, of a competitive relationship can also be gleaned from other factors, such as the places of sale and consumption. According to Korea, one relevant question in this connection is whether the disputed liquors are sold through the same outlets, so that consumers are actually offered a choice between them. If not, this is an indication that they are not, in fact, competing.

iv) The Degree of Competition or Substitutability

1.202. Korea poses the question, how strong must competition between two products be, before one can speak of a directly competitive and substitutable, or even a 'like' product relationship within the meaning of Article III:2?

1.203. Korea argues that it is good policy to give a strict reading to the terms of Article III:2. In Korea's view, given that the WTO is an organisation which brings together a large number of diverse countries, and that it has no mandate to harmonise tax policies, the threshold to review taxes under Article III:2 GATT (i.e., the competitive relationship between differently taxed products) is both important and substantial.

1.204. According to Korea, for 'like' products, the competitive relationship must be very strong indeed. Korea cites the Appellate Body report on Canada - Certain Measures Concerning Periodicals, in which it was stated that 'like' products are 'perfect substitutes'.¹¹⁹

1.205. Korea argues that in relation to the examination of the legal obligation under Article III:2 second sentence, the existence of competitive relationship is also an essential requirement for the determination of whether products can be determined to be directly competitive and substitutable. In Korea's view, unlike 'like' products, substitution does not have to be perfect between the products. Korea asserts that the texts of Article III:2 second sentence and its Interpretative Note require that the competitive relationship must still be strong before the products can be considered directly competitive and substitutable. This is why the text refers to "directly competing or substitutable products".

1.206. In Korea's view, the complainants have not shown that there is any close relationship between the imported products and soju. Korea therefore, is of the view that they have been unable to show actual competition on the market between the products and are now trying to introduce the argument that the absence of actual competition was due, at least in part, to the tax measures.

1.207. According to Korea, such a loose interpretation of Article III.2 runs counter to the text of the Interpretative Note to Article III.2. This Note adds that 'only where competition was involved' between two 'directly competitive or substitutable' products could a tax differential conceivably amount to a violation of the second sentence of Article III:2. In Korea's view, this language strongly suggests that there must be actual competition between the directly competitive or substitutable products before their tax rates are to be compared.

1.208. Korea further argues that it is clear that in the present case the complainants have neither shown that imported liquors actually compete with the Korean sojus, nor have they shown that these liquors would directly compete with each other in the absence of the tax differential. Korea refers to the argument by the complainants that there is a substantial measure of price elasticity between the Korean sojus and imported liquors. According to Korea, the complainants seek to argue that various product combinations are in competition with each other, on the grounds that relatively small changes in the price of soju will persuade Korean consumers to switch to imported liquors such as whisky. Korea notes that in support of this argument the European Communities submitted a market survey,¹²⁰ which has been endorsed by the United States.

1.209. Korea points out that given the considerable price differences (even pre-tax) and different end uses between these liquors, the complainants' conclusion flies in the face of common sense. Korea adds that another study funded by the European Communities, and conducted by the same organization but not for legal proceedings, found that "soju in particular remains unaffecetd by imported drinks".¹²¹

1.210. Korea submits that this particular study (the 'Dodwell' study) constitutes poor evidence. Korea puts forward the following reasons for its assertion that the Dodwell study cannot be relied upon as constituting evidence of product substitutability:

v) Problems in the Dodwell Report

Internally Inconsistent Results

1.211. Korea expounds upon the peculiarity and problems, namely flaws in design or oddities in the reported results. Korea mentions as an example that in chart 2, the percentage choosing soju rises when price of soju increases from 1,100 to 1,200 won, even though all other prices are held constant. Korea argues that this seems to imply a positively-sloped demand curve, which is, as the United States noted before the Japan - Taxes on Alcoholic Beverage II case, "contrary to one of the fundamental tenets of micro-economic theory.¹²² Moreover, Korea observes, the percentage choosing premium and standard scotch rises as the price of soju rises from 1,000 to 1,100 won but falls as the price of soju rises from 1,100 to 1,200 won. Korea notes that such odd findings are frequent: indeed, they appear in every chart.¹²³

1.212. Korea argues that there is no theoretical reason why the quantity demanded of one good should not fall as the price of a related good rises, although the implication that soju and whisky, for example, are complements in consumption, so that a fall in the price of whisky would increase consumption of soju, might seem to many economists implausible. Korea notes that the difficulty with the Dodwell report is that it sometimes shows the quantity of scotch demanded rising as the price of soju rises, and sometimes falling. Korea finds the facts reported, though curious, are less troubling than their lack of internal consistency.

1.213. Korea suggests that consumers in actual markets are unlikely to be so random in their responses to price changes. Korea concludes that the members of the Dodwell sample may have treated the Dodwell question with less gravity than the report's authors might like to believe.

To continue with Standard and Premium Soju

¹⁰⁹ Panel Report on United States - Measures Affecting Alcoholic and Malt Beverages, adopted on 19 June 1992, DS23/R, BISD 39S/206, 276, at para. 5.71 - 5.72.

¹¹⁰ See Appellate Body Report on Japan - Taxes on Alcoholic Beverages II, supra., p. 16.

¹¹¹ Appellate Body Report, supra., p.16.

¹¹² Panel Report on United States - Measures Affecting Alcoholic and Malt Beverages, supra., para. 5.25.

¹¹³ Appellate Body Report on Japan - Taxes on Alcoholic Beverages II, supra., p. 20.

¹¹⁴ Panel Report on Japan - Taxes on Alcoholic Beverages II, supra., supra, para. 6.22.

¹¹⁵ Ibid.

¹¹⁶ See Japan - Taxes on Alcoholic Beverages II, Appellate Body Report, supra., p. 25. The Appellate Body reiterated this reasoning in its more recent report in Canada - Certain Measures Concerning Periodicals, supra, p. 28.

¹¹⁷ See Panel Report in Japan - Taxes on Alcoholic Beverages II, supra., at para. 4.82. See also the more recent Canada - Certain Measures Concerning Periodicals, Appellate Body report, at p. 30, where the Appellate Body, in finding a directly competitive and substitutable relationship between certain imported and domestic periodicals, approvingly cited evidence of 'considerable price competition' between the two.

¹¹⁸ See Panel Report on Japan - Taxes on Alcoholic Beverages I, supra., para. 5.7.

¹¹⁹ Appellate Body Report on Canada - Certain Measures Concerning Periodicals, supra., p. 31.

¹²⁰ See EC Annex 13.

¹²¹ EC Commission, Your Guide to Exporting Food Products to Korea, Alcoholic Beverages 1997 (Sofres Report), p.22.

¹²² Panel Report in Japan - Taxes on Alcoholic Beverages II, supra., at para. 4.86.

¹²³ In chart 1, the percentage choosing premium scotch falls as the price of soju rises from 1,100 to 1,200 won even through all other prices-including that of premium scotch-are held constant; in chart 3, the percentage choosing premium soju falls as the price of standard soju rises from 1,100 won to 1,200 won; in chart 4, the percentages choosing gin and tequila both fall as the price of soju rises from 1,000 won to 1,100 won; in chart 5, the percentage choosing tequila falls as the price of soju rises from 1,000 to 1,100 won, and the percentages choosing vodka, gin and premium soju all fall as the price of soju rises from 1,100 to1,200 won; and in chart 6, the percentages choosing liqueur and premium soju both fall as the price of soju rises from 1,000 to 1,100 won, and the percentages choosing vodka and gin both fall as the price of soju rises from 1,100 to 1,200 won.