4.118 Japan responded to the US claim that the Liquor Tax Law had a protective effect. For Japan, among the four reasons why the 1987 Panel Report found Japan's liquor tax protective, the first one (large specific tax rate differential) is non-existent now under the appropriate yardstick of comparison, the second one (ad valorem tax on imported liquors) is already abolished, the third one (exclusive production of shochu in Japan) is incorrect, and the last one (increased imports of western style liquors, and consumer use of shochu blended with whisky etc.) is a combination of an irrelevant fact and an incorrect assumption. As argued with the Community, for Japan, "directly competitive or substitutable relationship" between domestic shochu and imported distilled liquors of other categories is not proven. Alleged "sudden or dramatic difference in rates at the margin" is common to many tax systems including taxes on wine in the US and in the Community, and is not evidence of protectiveness. The product "Jun Legend" is a case of experiment at the margin which, Japan argued, accompanies any product categories with tax differentials. Shochu, for which Japan is only the second or third largest producer in the world, is not an "inherently domestic product", and whisky, for which Japan is the fifth largest producer in the world, is not an "inherently foreign product". For Japan, none of the US criteria are met.

4.119 Japan submitted that the Liquor Tax Law generally and the very nature of its tax classification system do not have the aim or the effect "so as to afford protection". For Japan, the absence of protective aims and effect of the regulatory distinction contained in the Liquor Tax Law, confirms that shochu and other imported liquors are not like products and that the legislation is not inconsistent with Article III:2, first and second sentences. Japan submitted three criteria to demonstrate that the regulatory distinction of the Liquor Tax Law is consistent with Article III:2: a) the nature of the categorization, b) the aims of the legislation, and more particularly the new policies of horizontal equity and neutrality, and c) the absence of protective effect evidenced by the fact that there is no competitive relationship between shochu and other imported liquors (no cross-price elasticity), the fact that shochu is produced abroad and the very neutrality of the Liquor Tax Law.

4.120 Japan submitted that shochu is readily distinguishable from the rest of the distilled liquors and these differences have resulted in different net-of-tax prices in relation to which tax rates are adjusted according to the tax categories of the Liquor Tax Law. Indeed, the categorization of distilled liquors under the Liquor Tax Law is not protectionist because it is based on three objective criteria: (i) the cost of the raw materials; (ii) the alcoholic strength; and (iii) the value added through the post distillation process. More specifically, the distinctions between the categories are the following:

Then Japan explained to the Panel that the related tax categories were established in light of the following three criteria: 1) the tax categories are based on socially established boundaries; 2) the level of consumption warrants an independent category; and 3) the tax categories are consistent with the policy goals. In particular Japan argued that:

4.121 Japan emphasized that the classifications of the Liquor Tax Law as amended correspond to the classifications of the Harmonized System (HS) nomenclature, as amended by the 1993 decision of the Customs Co-operation Council. As described above, shochu, as defined by the Liquor Tax Law, is a category of product which falls under HS 2208.90 (Other) and has a lower alcoholic strength. Japan argued that as a striking similarity, the HS nomenclature picks out brandy, whisky, rum, gin, vodka and liqueur as distinct categories. If the HS nomenclature is not protective, then, Japan argued, the Liquor Tax Law classification should not be regarded as protective either.

The HS nomenclature does not define whisky, vodka or liqueur. Japan submitted that under a principle common to civilized constitutions, however, any statute which imposes a fiscal burden on the people needs to be spelled out as clearly as possible. Such drafting work would inevitably involve an exercise of drawing a line between various concepts. It is certainly not an easy task, and involves a degree of judgment, because socially accepted concepts are not necessarily based on a scientific analysis but may result from a historical development. For instance, the criteria of purification by white birch charcoal in the vodka definition is, in Japan's view of the US argument, arbitrary. Is it not arbitrary, then, asked Japan, to define a sparkling wine by its mushroom stoppers held in place by ties or fasteners as in the Liquor Tax Directive of the Community. For Japan, the Community's argument against the Japanese criteria does not indicate arbitrariness inherent in the Liquor Tax Law, but rather the difficulty common to all product definitions. For Japan, despite allegations by other parties to the present dispute, there is nothing arbitrary in identifying these popular categories according to socially accepted concepts.

4.122 Japan then responded to another alleged arbitrariness in the Liquor Tax Law which is the ceiling on the alcoholic strength of shochu A (35 per cent) and B (45 per cent). The perceived arbitrariness, however, reflects the difficulty inherent in defining a historically developed concept. As an illustration of such common difficulty, Japan noted the following minimum alcoholic strengths for various products under the European definitions:

For Japan, these floor rates are by no means arbitrary; they merely capture basic features of historically developed and socially accepted concepts. As stated earlier, the relatively low alcoholic ceiling for shochu reflects the fundamental features of this type of low-cost alcoholic beverage. The Liquor Tax Law does leave room for experimentation at the margin, as demonstrated by examples of Juhyo vodka or Jun Legend. Japan insisted that this is not inherent in the Japanese taxation system. As long as there is a tax differential between product categories, the opportunity for experimentation arises. This is an issue of optimal tax policy and not of GATT rules. Japan argued, for example, that Juhyo vodka contained only 20 per cent alcohol and would not meet the criteria of alcoholic strength for vodka in either Europe (37.5 per cent) or the United States (40 per cent). Although the filtering method of vodka was used, it was shochu in terms of alcoholic strength. This borderline product was consumed in much the same way as shochu and was subsequently reformulated into genuine shochu with alteration of raw materials to give a typical shochu taste.

4.123 For the Community, neither the Liquor Tax Law nor economic reality bear out Japan's exercise of ex-post facto rationalization since

4.124 In response to the Community's argument, Japan pointed out that similar differentials in net of tax prices between categories do exist in countries which apply a flat-rate tax on distilled liquors as well: For Japan, the Community claimed similarity of prices of brandy and gin based on the promotional leaflet of a Belgian supermarket, picking the least expensive item out of five brandy brands advertised in the leaflet (armagnac, calvados and cognac) and the most expensive gin brand from the six on the leaflet (gin and geneva, most expensive when adjusted for differences in bottle size). Japan called it a classical case of a selective reference, saying that calculation of the average net-of-tax prices of all items appearing on the leaflet results in a significant price differential. Against the Community's argument that grapes are very inexpensive and are used to produce industrial alcohol in the Mediterranean area, Japan questioned why is it then that brandy brands are more expensive than other distilled liquors in the Belgian supermarket. Moreover, according to a survey done by Business International, Inc., an affiliate of The Economist magazine, brandy is more expensive than whisky and whisky is more expensive than gin in 32 cities of the world where a flat tax rate is applied. Against the Community's claim that no category of distilled spirits is inherently more expensive than others, Japan questioned why is it then that such price differentials prevail across the world. For Japan, the Community tends to ignore the fact that differences in the market value of distilled liquors correspond to subtle product differences.

4.125 The Community reiterated that the prices and taxation systems of whisky in countries other than Japan are outside the terms of reference of the present dispute.

4.126 The Community also referred the Panel to a recent recommendation of the Administrative Reform Council, a public law advisory body attached to Japan's Prime Minister's office, which passed the following judgement on the alleged rationality of the Liquor Tax Law:

4.128 The Community and Canada responded that the prices and taxation systems of liquors in countries other than Japan were outside the terms of reference of the present dispute.

4.129 Japan continued its explanation of the neutrality of the categorization in stating that the tax rates are set corresponding to the average net-of-tax price of each category. While net-of-tax prices vary from one category to another, the ratio of the tax over the retail price stays roughly constant between categories. Looking at the figures contained in Annex IV, one can see that on the basis of the weighted average of the suggested retail prices of some of the 20 best selling brands, per quantity containing the same amount of alcohol as a 750 ml, 40 per cent bottle, the amount of tax burden on each category varies substantially. The ratio of the tax burden over the retail price, on the other hand, is roughly the same; as measured in December 1995 vis-à-vis average suggested retail prices, at around 20 per cent. According to the figures, a consumer of any category, on average, paid roughly 20 per cent of the price for tax. Japan referred the Panel to 20 individual prices, the tax/price ratio of which is distributed within a similar range. If these categories were subject to the same tax rate per alcoholic content as currently applied to whisky, any shochu would bear a heavier tax burden than any imported whisky.

4.130 In support of its argument that the categorization of the Liquor Tax Law is reasonable, Japan submitted to the Panel that EU members, Canada and the United States apply significantly different tax rates between distilled liquors, wine, beer and intermediate products. Moreover, in eight Community Member States and the United States, the tax applicable to still wine is different from that applied to sparkling wine. In Japan's view, this treatment of different categories is apparently not based on the degree of competitive relationship between products. For example, there seems to be little competition, if any, between cream liqueur and vodka, while the study by Bossard Consultants found a strong competitive relationship between wine and distilled liquors, and between beer and distilled liquors. In Japan's view, the other parties to the dispute reject tax distinctions between whisky and shochu, a pair of products which have not been demonstrated to be competing against each other. On the other hand, Japan submitted that the complainants take for granted tax differentiation among beer, wine and distilled liquors, categories which the Bossard study found to be competitive products. Japan suggested to the Panel that if Japan should not apply different rates to distilled liquors by virtue of GATT rules, wine-producing France should not tax distilled liquors more heavily, nor should Germany impose a higher tax on distilled liquors than on beer. Japan added that seven Community Member States apply different rates to distilled liquors. Japan argued that not all of these practices are inconsistent with GATT rules and that Article III is not an instrument for harmonization of internal taxes.

4.131 The complainant parties responded that the taxation systems in countries other than Japan were not included in the terms of reference of the present dispute, and, in any event, were not relevant to the issue of whether Japan's Liquor Tax Law is consistent with GATT rules.

4.132 For Japan, the classification of the Liquor Tax Law is for purposes other than protection of domestic production. Categorization of distilled liquors under the Law is indeed for the purpose of a legitimate policy to ensure neutrality and equity, and not for the purpose of protection. There has been a shift in the policy objective of legislation in the 1989 amendment: from vertical equity to neutrality and horizontal equity, or from the pursuit of differentiation in the tax/price ratio to equalization. Until 1987, the taxation placed the greatest emphasis on achieving "vertical equity". "Vertical equity" requires a greater tax burden be born by those who have a greater tax-bearing ability, compared to those who are less able to pay. This principle emphasizes the income redistribution function of taxation, and forms the basis of "progressive taxation", which applies a higher rate of tax to a higher income level. In contrast, the "horizontal equity" requires the same burden for those similarly situated. The value added tax achieves horizontal equity better by imposing an equal amount of tax on the same amount of consumption, although it does not redistribute income or achieve progressivity. Although a tax authority would never ignore any of the three goals of neutrality, horizontal equity and vertical equity, relative emphasis among the three may vary in response to prevailing economic and social circumstances and kinds of taxes at issue. Japan submitted, as an example, the US Congressional Budget Office's report on excise taxes, which starts the analysis by examining the three criteria. The taxation on distilled liquors at the time of the 1987 Panel Report was designed with the utmost priority put on vertical equity. With the 1989 reform, however, a higher priority was given to neutrality than to equity in general, and to horizontal equity than to vertical equity. For Japan, the Community and the United States erred in ignoring the policy shift.

4.133 Japan reiterated that the primary objective of the policy adopted since the 1989 amendment is to achieve neutrality and horizontal equity to consumers� choice or minimization of distortions in competitive conditions among products. Japan argued that the Liquor Tax Law succeeded in doing so in ensuring that the ratio of the tax over the retail price stays roughly constant between categories of distilled liquors. Japan therefore argued that the complainants failed to prove the existence of protective intent. Concerning the lack of a statement on the purpose of the Liquor Tax Law, Japan responded that omission is only a standard practice of tax legislation and that the government representatives cited neutrality and equity repeatedly before the National Diet and on other occasions. The legislative record shows that neutrality and equity were cited as primary objectives of the current Law. For example, the Minister of Finance stated at the National Diet that the fundamental principles of the 1989 amendments were to ensure equity in distribution of the tax burden and to maintain neutrality toward economic activities, under the recognition that the tax should spread the common cost in a broad, fair manner among the people, and to simplify the tax system. With respect to the 1994 amendment, the Director General in charge explained its objective and purpose in the following manner: "The present amendment reflects the last year's recommendation by the Tax Commission to appropriately rearrange the level of the taxation burden which has been declining for some categories. The amendment intends to redress the burden, and to readjust the tax rates in light of changing consumption patterns, in order to equalize the tax burden among various categories". Japan argued that the 1994 amendment was not for the purpose of implementing the recommendations of the 1987 Panel report, but to redress the decline of the tax/price ratio of shochu and spirits which resulted from the rising prices to maintain the consistency with the recommendation already attained by the 1989 amendment. The amendment raised the tax rates for all categories, except whisky/brandy. For example, the tax rate for shochu A has been raised 30 per cent and that for shochu B by 44 per cent. For Japan, these statements confirm that the policy intent of the Liquor Tax Law was and is to ensure neutrality and equity taking the tax/price ratio as a principal indicator.

4.134 Japan explained that two factors prompted this shift in policy priorities. First, the policy considerations behind Japanese taxation in general were shifting since 1986 towards putting more emphasis on horizontal equity, as opposed to vertical equity, and on neutrality. The Tax Commission issued the following recommendations:

4.135 Japan went on to argue that the 1987 Panel Report was the second motivation. It was as a result of Japan's commitment to faithfully fulfil the recommendation of the panel report to "eliminate protective effects prohibited by Article III:2 of GATT", that the tax/price ratio was equalized across categories of distilled liquor. Neutrality and horizontal equity, the two purposes prioritized in the whole reform package, were pursued with special vigour for distilled liquor taxation because of the 1987 Panel Report. In fact, the Tax Commission's recommendation of December 1987 referred to the "requirement to faithfully implement GATT recommendation", and that of April 1988 specifically recommended that "the tax disparity be minimized between various alcoholic beverages". Based on these two factors, the amendment to the Liquor Tax Law minimized the tax disparity between alcoholic beverages in general, and, with respect to distilled liquors, equalized the tax/price ratio across categories of distilled liquors in stricter pursuit of neutrality. It also intends to attain equity, or fair distribution of the tax burden among consumers, to the extent compatible with the neutrality requirement.

4.136 In response to Japan's statement that between 1986 and 1988 the tax policy underlying the Liquor Tax Law shifted from "vertical equity" to "horizontal equity" and that this shift led to an amendment to the Liquor Tax Law in 1989 that, according to Japan, "emphasizes neutrality", Canada referred the Panel to the Deregulation Subcommittee of the Administrative Reform Council which stated the opposite. The function of the Council, which is an independent body and whose members are appointed by Japan's Prime Minister and approved by the Diet, is to promote "rational" reform of Japan's "administrative system". In December, 1995, the Subcommittee recommended fundamental changes to the Liquor Tax Law. According to the Subcommittee, the different tax rates levied on the various categories pursuant to the Liquor Tax Law have a wide impact on consumer choice at the time of purchase, constitute "virtual restrictions on buyer's activities"; are not based on logical taxation standards; are not based on "fairness of the tax burden"; are not neutral in relation to consumer choice. Thus, Canada argued, contrary to Japan's assertion that "the liquor tax is similar to VAT in terms of "tax discrimination indices" and thus is not "trade-distortive", the Administrative Reform Council concludes otherwise, and considers that as currently structured, the Liquor Tax Law distorts neutrality in relation to consumer choice.

4.137 Japan argued that what Canada read from the Subcommittee�s recommendation differed from the language of the Subcommittee report. Japan then submitted a chart which compares the relative tax burden between shochu A and imported vodka and whisky under the three yardsticks. For example, the left column indicates that the liquor tax per litre of vodka is 2.7 times higher than that of shochu A, and that of whisky, 7 times higher than that of shochu.

Note: Calculated on the basis of weighted average of 20 most selling brands.

4.138 Japan argued that this figure demonstrates that i) the liquor tax is similar to VAT in terms of "tax discrimination indices", and that ii) VAT would be regarded more "trade-distortive" than the liquor tax as long as a comparison is made on the basis of the taxes by the tax amount per litre of beverage or of pure alcohol. Japan submitted that VAT is regarded as one of the most trade-neutral indirect taxes and its introduction is one of the conditions to join the European Union. On the other hand, a comparison made with the amount of tax per litre of beverage or of pure alcohol would find such VAT as trade-distortive. In fact it is the use of those two yardsticks as tools of comparing taxes which is problematic, rather than the tax itself. Japan emphasized that a consumer usually does not buy a product exclusively on the basis of the size of the bottle or on the basis of the alcoholic strength. Consumers choose products by comparing the price and the overall value of a product, which depends upon the taste, flavour and other features and is not confined to the volume and strength. Japan argued that this is why the tax/price ratio is a better criterion to evaluate the effects of taxes on competitive conditions, and neutrality is achieved when the tax/price ratio is equalized, as is the case with the Japanese tax.

4.139 The Community submitted that the system established by the Liquor Tax Law was radically different from an ad valorem tax system. In a true ad valorem system, taxes are proportional to the actual sales value of each shipment and, therefore, neutral, provided that the rates are the same for all the categories. In contrast, under the Liquor Tax Law system, the tax amount is unrelated to the actual sales price. Instead, the assumption is made that certain categories of liquors are a priori more expensive than others. The tax rates are then set so as to reflect this "assumed value" and uniformly applied to all shipments, regardless of their actual price. The outcome of this system is not "neutrality" but arbitrariness because no category of distilled spirits is inherently more expensive than others. The evidence provided to the Panel shows that, even under the current tax system, the pre-tax price of some brands of whisky is lower than the equivalent price for some brands of shochu.

4.140 Japan argued that the lack of plausible alternatives further testifies to the lack of protective intent. Conceivable alternatives to ensure neutrality and equity are: (i) to raise the ad valorem value-added tax to a level comparable to that of the European Union, which applies not only to liquor consumption but to almost all consumption or (ii) to alter the liquor tax into an ad valorem tax. However, for Japan, neither of these is practical. First, the decision to raise the ad valorem consumption tax from the present three per cent to five per cent beginning April 1997 was made in 1994 only after a prolonged, heated debate. It is not very likely that the rate would be raised to the Community level in the near future. Second, an ad valorem excise tax could easily invite tax evasion by way of transfer-pricing, particularly if applied at the shipping stage. Canada's Federal Manufacturers Sales Tax suffered from the same difficulty and was abolished in 1991. On the other hand, enforcement cost of an ad valorem tax would be very substantial if applied at the retail level.

4.141 The Community responded that the reasons given by Japan for not pursuing the alleged objective of neutrality through the application of an ad valorem tax system are groundless. In the Community's view, Japan's definition of neutrality assumes that specific excise taxes are passed on in full to consumers. Thus, if the current tax system was truly neutral, there should be no reason for the Diet to object to its replacement by an ad valorem consumption tax. The Diet's opposition merely reflects the fact that under the current system shochu is much less taxed than it would under a truly neutral ad valorem consumption tax. In any event, internal political difficulties may not provide a valid justification for infringing GATT rules. For the Community, the tax evasion problems invoked by Japan are common to the application of all ad valorem systems. Similar issues arise, for instance, in connection with the application of ad valorem customs duties to import transactions between related parties. It is therefore suggested that any tax evasion problems related to transfer pricing could be appropriately tackled by using any of the alternative methods for the calculation of transaction values provided for in the Agreement on Implementation of Article VII of GATT 1994 (the Customs Valuation Code). In this respect, it is worth noting that Japan currently applies ad valorem customs duties on imports of a fair number of liquors, apparently without this giving rise to any major duty evasion problems. Moreover, until 1989, Japan applied ad valorem excise duties to certain liquors. These duties were abolished because they were found inconsistent with Article III:2 by the 1987 Panel Report and not because they were an invitation to tax evasion. Ad valorem excise duties on alcoholic beverages are currently applied by other WTO Members (e.g., in Denmark). Contrary to Japan's assertions, the application of a flat rate tax to all distilled spirits would not necessarily render the Japanese system less "neutral" (according to Japan's own definition of neutrality). For the Community, the scenario depicted by Japan is unrealistic because it does not take into account the likely dynamic effects of a tax readjustment. If shochu was taxed at the same rate as whisky, the cheapest brands of shochu would be expelled from the market, just like former second grade whisky was wiped out of the market by the 1989 tax reform. This would have the consequence of lowering the average tax/price ratio for the remaining brands of shochu. On the other hand, an increase of the rates on shochu would allow the entry into the market of cheaper brands of the other categories of liquors, thus driving up their respective average tax/price ratios. Moreover, there is no reason why a flat tax rate should be applied at the level of the current rate for whisky. It could as well be set at the level of the current rate for shochu or at any intermediate level between the current rates for shochu and "whisky/brandy". The tax systems of other WTO Members which apply a flat tax rate to all distilled spirits are not less "neutral" in terms of tax/price ratios than the current Japanese system but rather the opposite.

4.142 In response to the Community's argument that the evasion problems are manageable since there are other cases of the ad valorem tax, Japan made the following counterarguments. First, the Danish ad valorem tax is expected to be abolished shortly. According to Japan, the abolition was prompted by the fact that, as prices are determined on a case-by-case basis, the system became enormously confusing, and that, in particular, it was easy to manipulate prices while such manoeuvres were hard to detect. The Community now prohibits members from adopting an ad valorem tax. Second, although it is true that an ad valorem tax used to be part of the Japanese liquor tax before the 1989 amendment, detection of transfer pricing was possible then because retail prices were stable at the suggested retail prices and the margin rates for retailers and wholesalers were broadly constant. However, liquors are now traded, in some cases, at prices substantially different from suggested retail prices and it has become virtually impossible to distinguish manipulation from normal trading. In fact there was another ad valorem excise tax before the 1989 amendment which applied to consumer goods such as automobiles, cameras, watches and audio equipments. However, since these items were often sold at discount, there were a series of complaints over taxable prices, and the integrity of the tax was put into question. Third, for Japan, an ad valorem excise tax is fundamentally different from an ad valorem customs duty; and the success of an ad valorem customs duty does not mean that an ad valorem excise tax is feasible. The total tax burden under Japan's liquor tax is 20 billion US dollars, and is more than 50 times greater than that of customs duties on liquor. The aggregate magnitude of incentives for tax evasion is therefore far greater for the liquor tax than for the customs duty. Nevertheless, it is far more difficult to secure compliance with the excise tax. Customs duties are charged before the goods are withdrawn, while the liquor tax is levied after they are shipped. Importers must declare the prices to the customs authorities and pay applicable duties before they are authorized to withdraw the goods. In contrast, the liquor tax is levied after the goods have been shipped, on the basis of declaration for the preceding one month. Japan argued that careful examination of the examples of ad valorem duties put forward by the Community demonstrates the impracticality of an ad valorem liquor tax. In response to the Community's claim concerning "the likely dynamic effect of a tax adjustment", Japan raised the following points. First, since 1987, the tax rate on whisky has been halved. However, the share of premium whisky in the imported Scotch whisky market expanded from 33 per cent in 1987 to 51 per cent in 1994. Japan argued that, contrary to the allegation of the Community, the decrease of tax promoted premium whisky sales. Also, during the same period, in which the shochu tax rate increased twofold, containers of shochu A shifted from a medium size to a large size. Again, contrary to the Community's allegation, higher tax has led to an increase in the share of low-cost brands. The Japanese market configuration shifted in a direction opposite to the Community's "likely dynamic effect of tax adjustment". Second, according to Japan, differentials in net of tax prices between categories similar to those in the Japanese market do exist in countries which apply a flat-rate tax on distilled liquors as well. According to a survey done by Business International, Inc., an affiliate of The Economist magazine, brandy is more expensive than whisky and whisky is more expensive than gin in 32 cities of the world where a flat tax rate is applied. The Community argued that the application of a flat rate tax to all distilled spirits would not necessarily render the Japanese system less neutral. However, Japan calculated the tax burden in Belgium on the basis of average prices quoted in the promotion leaflet of the supermarket in Brussels submitted by the Community and found that any bottle of vodka or any bottle of gin is burdened with a higher tax/price ratio than any of the bottles of brandy in the shop (left hand side of the exhibit in Annex V). On the other hand, under the Japanese system, tax/price ratios are broadly equal across categories (right hand side of the exhibit in Annex V). Japan concluded from the exhibit that a flat-rate tax was far inferior to the current Japanese tax in terms of neutrality, horizontal equity and vertical equity. Japan submitted two tables on "Tax/Price Ratio in Belgium and in Japan" (see Annex V).

TO CONTINUE WITH JAPAN - TAXES ON ALCOHOLIC BEVERAGES