What's New?
 - Sitemap - Calendar
Trade Agreements - FTAA Process - Trade Issues 

espa�ol - fran�ais - portugu�s
Search

Chile � Taxes On Alcoholic Beverages

Report of the Panel

(Continued)


    5. "Direct Proportionality" Argument

  1. The European Communities also points out that the two panel reports on Japan � Taxes on Alcoholic Beverages I and II stand for the proposition that the application of specific taxes in direct proportion to the alcohol contained in each type of distilled spirit does not constitute "dissimilar" taxation of the spirits. The underlying reasoning is that, in such a system of taxation, the taxed product is not the spirit but the alcohol contained in the beverage. For example, the European Communities applies a uniform specific tax per hectolitre of pure alcohol to all the products containing ethyl alcohol, including, but not limited to, distilled spirits falling within HS 2208. 189
  2. The European Communities maintains that this reasoning does not apply in the case at hand. The measure applied by Chile is an ad valorem tax applied on the value of the beverage as a whole and not on the value of the alcohol content. Moreover, the value of the beverage is not directly related to the value of the alcohol content. Therefore, Chile's measure cannot be characterised as a tax on the alcohol content. For that reason, the European Communities considers that, in order to determine whether pisco and the other spirits are "similarly taxed", the Panel should compare the rates per bottle of each spirit, and not the rates per degree of alcohol.
  3. The European Communities claims that in any event, it has shown that pisco and the other spirits are also "not similarly" taxed when one compares the rates per degree of alcohol. Specifically, each degree of alcohol contained in a bottle of whisky, gin, rum, vodka or tequila is taxed at rate which is more than 50 % higher than the rate applied to each degree contained in a bottle of pisco especial.
  4. The European Communities notes that Chile has acknowledged that the rates per degree of alcohol vary from one spirit to another, but that it claims that the lack of proportionality between differences in taxation and differences in alcohol content does not amount to "dissimilar" taxation. In other words, according to Chile, if there is a difference in alcohol content, however small, between two spirits, then no conceivable tax differential which a Member may see fit to apply can be considered as "dissimilar" taxation of those spirits.
  5. In the view of the European Communities, Chile's position is logically untenable. If the taxed product is the alcoholic beverage, then one should compare the rates per unit of beverage volume, irrespective of their alcohol content. On the other hand, if the taxed product is the alcohol content, the comparison should be made between the rates per unit of alcohol volume, regardless of the beverage in which it is contained.
  6. The European Communities notes that in any event, as shown in Table 30, in the amended ILA the tax rate per degree of alcohol is not uniform but varies from product to product. Each degree of alcohol in whisky, gin, vodka, rum and tequila is taxed at 1.175 % while each degree of alcohol in pisco of 35° is taxed at only 0.771 %. Thus, pisco and the main types of imported spirits are not "similarly" taxed even if the comparison is made by reference to their alcohol content.
  7. Table 30 190

    Alcohol content

    Ad valorem rate

    Percentage points per degree of alcohol

    15°

    27 %

    1.8 %

    20°

    27 %

    1.35 %

    25°

    27 %

    1.08 %

    30°

    27 %

    0.9 %

    35°

    27 %

    0.771 %

    38°

    39 %

    1.026 %

    40°

    47 %

    1.175 %

    43°

    47 %

    1.093 %

  8. The European Communities also points out that if the rate per degree of alcohol was the same for all spirits as the rate currently applied to pisco of 35° , the resulting ad valorem rates would be as follows:
  9. Table 31 191

    Alcohol content

    Current ad valorem rate

    Ad valorem rate if the rate per degree of alcohol was 0.771 %

    15°

    27 %

    11.57 %

    20°

    27 %

    15.42 %

    25°

    27 %

    19.28 %

    30°

    27 %

    23.13 %

    35°

    27 %

    27 %

    38°

    39 %

    29.30 %

    40°

    47 %

    30.84 %

  10. The European Communities then concludes that if the taxed product is the alcoholic beverage, then one should compare the rates per unit of beverage volume, irrespective of their alcohol content. On the other hand, if the taxed product is the alcohol content, the comparison should be made between the rates per unit of alcohol volume, regardless of the beverages in which it is contained.
  11. In rebuttal, Chile notes that the European Communities implies that it could accept the ad valorem and alcohol content standards of the Chilean law, if only the taxes were proportional. Then, Chile argues that there is no rule of proportionality in the WTO and, if there were, the European Communities would be in violation of its own rule because wine and beer are taxed less per degree of alcohol than whisky and other distilled spirits. Indeed, the EC's argument in this respect with regard to Chile is remarkably similar to the complaint of the Scotch Whisky Association about tax systems that favor wine and beer in Member Countries of the European Union. There are many other "non-proportional" taxes, including all specific taxes, for example. Furthermore, the European Court of Justice upheld a non-proportional tax on engine displacement under the EC's own national treatment provisions. The United States also taxes wine and beer at much lower rates per degree of alcohol, using a system that also produces large variations in tax for small differences in alcohol content of its wines. 192
  12. Chile points out that ultimately, the European Communities itself seems to recognize that Article III does not require taxes or regulations to have equal effects by all standards of measurement, as the European Communities ultimately appears to rely on a theory that the deficiency of the New Chilean System is not that the taxes vary by alcohol strength, but rather that the tax does not vary in direct proportion to alcohol strength, as shown for example in Tables 30 and 31 above. Chile does not deny the veracity of those charts, but Chile strongly disagrees that Article III requires a rule of direct proportionality with respect to alcoholic beverage taxes.
  13. Chile notes that the European Communities appears to suggest that decisions on Japan - Taxes on Alcoholic Beverages I and II support the EC proposition that taxes employing an alcohol content standard must establish tax rates directly proportional to the degree of alcohol in order not to constitute dissimilar taxation. In fact, the panels in those cases did not decide that issue. The panels only said that taxation must be based on objective criteria, and that Japan could not use a different scale of alcohol-based taxation for different types of directly competitive or substitutable alcoholic beverages, but the panels did not say that taxes must be directly proportional to alcohol strength.
  14. Chile maintains that GATT Article III:2 has never been interpreted to require such a direct proportionality rule. As Chile has already noted above, the Japanese systems were faulted for establishing significantly different tax scales for different types of distilled spirits that the panels held to be like, directly competitive or substitutable. The panels did not find fault, however, with the lack of direct proportionality to liquor content.
  15. The European Communities replies that contrary to Chile's protestations that this issue has not been decided by previous panels, Chile's position is refuted by the findings of the two Panel Reports on Japan � Taxes on Alcoholic Beverages I and II. The second panel report is particularly clear in this regard. In that report, the panel based its conclusion that whisky and shochu were not "similarly" taxed on the fact that the rate per degree of alcohol applied to whisky of 40° was higher than the rate per degree of alcohol applied to shochu of 25° . 193 This comparison would have been totally irrelevant if, as claimed by Chile, differences in alcohol content could justify non-proportional differences in taxation. If Chile's position was correct, the panel could not have reached the conclusion that shochu and whisky were not "similarly" taxed except by comparing the rate per degree of alcohol applied by Japan to whisky of 40° and to shochu of 40° , something which the panel did not consider necessary to do. 194 The European Communities maintains that similarly, the panel reached the conclusion that vodka was taxed "in excess of" shochu by comparing the rate per degree of alcohol applied to vodka of 38° to the rate per degree of alcohol applied to shochu of 25° and not the rate per degree of alcohol applied to shochu of 38° 195
  16. Chile further responds that it should also be born in mind that requiring a direct proportionality rule is not only unprecedented and unwarranted under the plain language of Article III:2, but it serves no valid policy purposes not already served simply by requiring objective standards. A large benefit for international trade in having objective standards such as the alcohol strength criterion of the New Chilean System is that domestic and foreign producers have the ability to adapt their production to reduce the tax or regulatory burden, if they so choose. As discussed below under the third element of the Article III:2 test, the New Chilean System allows foreign producers, if they so choose, to dilute their product to qualify for lower levels of Chilean taxes. While in some cases that may be inconvenient for exporters in much the way different national labeling or packaging requirements can be inconvenient, Article III does not require harmonization of standards to the convenience of large global exporters.
  17. In the view of Chile, thus, the European Communities recognizes the right to differentiate products based on objective criteria such as the alcohol content of a beverage and in doing so implement a neutral based system that does not infringe Article III. Chile has enacted such a law based on the very same objective criteria that the European Communities has recognized as valid: the alcohol content of the various spirits, which does not discriminate among the various domestic and foreign participants.
  18. Chile argues that similar views have been expressed before the European Court of Justice ("ECJ"). The ECJ in Case 170/78, Commission v United Kingdom, 1983 ECR 2265 noted that:
  19. [T]he Commission has recommended that spirits should be charged at a higher rate of duty according to alcoholic strength than liqueur wines. It appears, therefore, to have accepted that there are social reasons for imposing a relatively higher rate of taxation on beverages with a higher alcoholic content. 196

  20. Chile reiterates that the New Chilean System does not make distinctions based on type of distilled spirits, but rather taxes every distilled spirit (except wine and beer) according to the same scale based on alcohol content and value. The European Communities uses a different objective scale for taxation of alcoholic beverages, levying a specific tax per degree of alcohol content for distilled spirits. Both the Chilean and EC systems by some measures discriminate against products based on price and alcohol content. Ad valorem systems, measured by specific rate per volume unit, fall most heavily on high priced goods, but economists agree that ad valorem systems are fairest in preserving competitive relationships. Specific rate taxes by comparison distort competition in favor of high price goods, if measured on an ad valorem basis.
  21. According to Chile, the European Communities continues to attack the New Chilean system based on an unfounded theory that taxation based on alcohol content must be proportional. In some respects, the EC's argument seems to be motivated by an effort to make the EC's own tax system into a fixed GATT rule, as the European Communities insists on the use of not just directly proportional but specific taxes. There is no basis for such a claim. Chile has already pointed out that the European Communities does not even follow its own rule of proportionality with respect to beer and wine. The U.S. system for wine is even less proportionate.
  22. Chile contends that if the EC's rule is applied to the effect that tax rates must be constant per degree of alcohol in order to meet the non dissimilar taxation test, it follows that an ad valorem flat rate would breach that test, as it shown in Table 32.
  23. Table 32 197

    Tax Per Degree of Alcohol Assuming a 30% Flat Ad Valorem Rate

    Alcohol Content

    Tax per degree of alcohol

    25�

    1.20 %

    30�

    1.00 %

    35�

    0.86 %

    40�

    0.75 %

    45�

    0.67 %

  24. Chile maintains that such a proportionality test would invalidate luxury taxes that fall more heavily on luxury goods, or more lightly on low-priced staples, since such taxes are unlikely to be directly proportionate, which would be contrary to their very purpose.
  25. Chile states that while the EC practices are not at issue, it hopes that the broader perspective of the Panel will lead it to reject the EC's insistence on a proportionality test and preference for specific duties as the preferred or only analytical perspective for assessing compliance with Article III of the GATT 1994.
  26. According to Chile, both the EC and Chilean systems treat wine and beer separately. Chile taxes beer and wine relatively higher, measured in terms of degrees of alcohol, and the European Communities taxes wine and beer lower per degree of alcohol.
  27. Chile notes that in another effort to bolster its complaint, the European Communities argues at several points that the Japan - Taxes on Alcoholic Beverages II case settles the issue whether there can be non-proportionate increases in tax based on alcohol content. Japan's system taxed shochu and whisky according to two very different scales of specific taxes, which significantly favoured shochu over whisky. Contrary to the EC's rather surprising claim, the panel did not reject the Japanese system because of a non-proportionate distinction in tax per degree of alcohol content, but rather because the scale for two different types of spirit was entirely different and advantageous only and at all points to shochu.
  28. Chile further maintains that it is likewise inconceivable that members of the WTO, particularly developing country members, thought or think that, in joining the WTO and accepting thereby the obligations of Article III:2, they were foregoing the right to use fiscal policy tools such as luxury taxes or exemptions or reduced taxes for goods purchased primarily by poor consumers, even if such policies result in higher taxes on many imports than on many like or directly competitive products.
  29. In the view of Chile, the kind of flawed interpretation proposed by the European Communities in this case would set a bad precedent for not just alcoholic beverage cases, but also for other cases in which countries differentiate in the application of internal taxes on the basis of objective criteria. For example, virtually any luxury tax will have the effect of taxing high priced products more, even proportionately more than low-priced goods. Borrowing one of the examples from the Panel's questions, a Boss suit is surely as substitutable with a suit sold in a discount store as Scotch whisky is with Pisco Corriente. The EC's analysis would therefore prohibit a luxury tax on high priced suits, if most taxed suits were imported and there was substantial production of untaxed suits domestically. An even more closely analogous case would be an exemption from tax for low-price cooking oil, which would inevitably result in olive oil bearing a higher tax than, for example, palm oil or soybean oil. Such a tax would be illegal under the EC's analysis of Article III, and no exemption would save it.
  30. Chile further argues that similarly, horsepower or engine displacement taxes, both of which have been common in Europe, would also breach Article III, if, as a result, some types of imported vehicles were treated less favorably than some types of vehicles predominantly produced domestically. Under the EC's theory, the United States and Canada would have justifiable complaints against the EC Member States because such taxes result in imported types of vehicles bearing higher taxes than domestic types.
  31. Chile emphasizes that the New Chilean System does explicitly impose taxes according to alcohol content. However, that kind of objectively based discrimination, like discrimination by horsepower, engine displacement, or (in luxury tax systems) value, has not been considered to violate Article III:2. Yet, by the EC's logic, if it could be shown, for example, that an engine displacement tax resulted in higher taxes on, for example, sport utility vehicles that were largely imported, and lower taxes on small commuter cars that were largely produced domestically, then the engine displacement tax would become inconsistent with Article III:2 for distinguishing in its effects between types of directly competitive or substitutable vehicles.
  32. Chile further puts forth that the European Communities itself does not observe a rule of direct proportionality for its own alcoholic beverages. Many EC Member States impose proportionately much lower taxes per degree of alcohol on beer and wine than on distilled spirits. The Scotch Whisky Association shows the following discrepancy in tax per degree of alcohol in various Member States of the European Communities:
  33. Table 33 198

    Proportionality of taxes per degree of alcohol on several The European Communities members

    Erreur! Signet non défini.

    Austria

    Belgium

    Denmark

    Finland

    France

    Wine

    Nil

    426

    856

    2,162

    30

    Beer

    347

    425

    780

    2,888

    257

    Spirits

    723

    1,651

    3,674

    5,097

    1,440

    Germany

    Greece

    Ireland

    Italy

    Luxembourg

    Wine

    Nil

    Nil

    2,559

    nil

    Nil

    Beer

    196

    309

    2,049

    351

    197

    Spirits

    1,297

    999

    2,847

    648

    1,035

    NL

    Portugal

    Spain

    Sweden

    UK

    Wine

    441

    Nil

    Nil

    2,937

    1,911

    Beer

    424

    275

    168

    1,746

    1,619

    Spirits

    1,497

    814

    686

    5,955

    2,843

    (ECU per hectoliter of pure alcohol) 199

  34. Chile notes that it cites this chart not to criticize the EC Member State tax systems nor to argue that beer and wine, which are not at issue in this dispute, should be regarded as competitive or substitutable products (though it is very clear that the Scotch Whisky Association believes that to be the case and is promoting arguments in that regard that are remarkably similar to the EC's arguments in this dispute). Rather, Chile refers to the chart to demonstrate that even the European Communities (despite the Scotch Whisky Association) does not really believe in a rule of direct proportionality for alcohol taxes. It might be added that tax systems of other countries exempt low value items from some forms of indirect taxes, such as low cost clothing or food. Luxury taxes and other kinds of progressive taxes similarly are particularly likely to have substantial effects on imports -- but such taxes are not considered to violate Article III merely because of a lack of direct proportionality.
  35. The European Communities notes that Chile incorrectly argues that the fact that the EC Member States apply different rates per degree of alcohol to wine, beer and spirits involves a "recognition" by the European Communities that tax differentials do not have to be proportional to alcohol content. Contrary to what appear to be Chile's expectations, the European Communities does not consider it necessary to deny that wine, beer and spirits are not "similarly" taxed in the European Communities. It is self-evident from the table produced by Chile that they are not. This, however, is far from constituting per se a violation of Article III:2, as Chile implies. To begin with, it would have to be established that wine, beer and spirits are "directly competitive or substitutable" products, something which could be rather difficult, especially by Chile's own strict standard. Even then, it would still be necessary to show that the tax differentials afford protection to the EC domestic production. One may suspect that the very successful Chilean wine exporters would not subscribe the argument that, by taxing more heavily whisky than wine, the United Kingdom affords protection to its non-existent production of wine.
  36. The European Communities also notes that Chile states the following in support of its allegation that the European Communities has "recognised" that the application of taxes that do not vary in proportion to alcohol content does not constitutes dissimilar taxation:
  37. The ECJ in case 170/78, Commission v. United Kingdom, 1983 ECR 2265 noted that:

    ... the Commission has recommended that spirits should be charged a higher rate of duty according to alcoholic strength than liqueur wines. It appears, therefore, to have accepted, that there are social reasons for imposing a relatively higher rate of taxation on beverages with a higher alcohol content.

  38. The European Communities contests that to begin with, this is not something which the ECJ said. Instead, this was an argument made by the United Kingdom, the defendant in that case. In any event, the quoted passage does not support Chile's position. The quoted statement argues in favour of imposing a higher rate on spirits than on liqueurs "for social reasons". It does not argue that doing so would not constitute "dissimilar" taxation. To the contrary, to the extent that it refers to a "higher rate of taxation" the quoted statement admits explicitly the obvious fact that under the proposed system liqueurs and spirits would not be "similarly" taxed.
  39. The European Communities also states that for the same reasons, the arguments drawn by Chile from the fact that some EC Member States apply car taxes that vary according to horsepower are also irrelevant. 200 If a country applies higher taxes to large cars than to small ones, then it is indisputable that large cars and small cars are "not similarly" taxed. A different matter is whether small cars are "directly competitive or substitutable" with large cars. And still a different matter, whether applying higher taxes to large cars "affords protection to domestic production".
  40. The European Communities goes on to argue that in this connection, it is worth noting that, contrary to Chile's assertions, the European Court of Justice ("ECJ") has never given an unconditional endorsement to the application of car taxes linked to engine power. The ECJ has ruled that those taxes may be compatible with Article 95 of the EC Treaty provided only that they are "free from any discriminatory or protective effect". 201 In a number of decisions, the ECJ has found that the application of car taxes linked to engine power was contrary in casu to Article 95 because it afforded protection to the domestic car production of the Member State concerned. The tax systems condemned by the ECJ had features which made them very similar to Chile's liquor tax system, such as arbitrary thresholds or disproportionately steep increases beyond the point where there is no significant domestic production. 202

To continue with "so as to afford protection to domestic production"


189 See Council Directive 92/83/EEC, of 19 October 1992 (OJ No L 316 of 31.10.92) and Council Directive 92/84/EEC, of 19 October 1992 (OJ No L 316 of 31.10.92).

190 EC First Submission, Table 21.

191 Ibid., Table 22.

192 26 U.S.C.A. � 5001.

193 The European Communities notes that under Japan's Liquor Tax Law, the tax rate varied within each of the four tax categories ("shochu", "whisky/brandy", "spirits" and "liqueurs") according to alcohol content. For example, the tax rate on shochu B of 25° was Y102,100 per kiloliter and the rate on shochu B of 40° Y284,100 per kiloliter. Shochu is produced within the range of 20° to 45° , but 25° is the most usual strength.

194 Panel Report on Japan - Taxes on Alcoholic Beverages II, supra., paras. 6.33 and 2.3.

195 Ibid., paras 6.24 and 2.3.

196 The ECJ in Case 170/78, Commission v United Kingdom, 1983 ECR 2265.

197 Chile Oral Statement at the Second Substantive Meeting, Table I.

198 Chile First Submission, Table 3.

199 The Scotch Whisky Association, Bulletin Board - Case for Taxation Reform (visited 24 Sept. 1998) <http://www.scotch-whisky.org.uk/bb-txrfm.htm>.

200 The European Communities notes that the suggestion to the effect that the US "is more likely" to produce cars with higher horsepower than the European Communities is simply wrong. Most Mercedes, BMWs and Jaguars, as well as Porsches or Ferraris, are still assembled in the European Communities.

201 See e.g. the Judgement of 9 May 1985, Case 112/84, Michel Humblot v. Director des services fiscaux (ECR 1985, pages 1367-1380); and the Judgement of 17 September 1987, Case 433/85, Jacques Feldain v. Directeur des services fiscaux du departement du Haut-Rhin (ECR 1987, p. 3521).

202 The European Communities points out that for example, in Humblot (para. 16) the ECJ ruled that:

"Article 95 of the EEC treaty prohibits the charging on cars exceeding a given power rating for tax purposes of a special fixed tax, the amount of which is several times the highest amount of the progressive tax payable on cars of less than the said power rating for tax purposes, where the only cars subject to the special tax are imported, in particular from other member States".