|
|
The question whether or not a finding that a formally origin-neutral measure violates the non-discrimination rules requires some kind of discriminatory impact related to origin arises in the context of both most-favored-nation and national treatment.
In the GATT a violation of Article III:4 7 presupposes, among other things, that the imported products are "like" domestic products and accorded "less favorable" treatment. It is obvious that any internal regulation openly distinguishing between the two categories of national and foreign goods to the detriment of the latter violates GATT Article III:4. When the regulatory differentiation does not follow the line between domestic and imported, but distinguishes according to an origin-neutral criterion (physical characteristics, price, sale modalities etc.), its scrutiny becomes more complicated.
Such a measure distinguishing on the basis of product related characteristics can only violate Article III:4, if goods on both sides of the regulatory dividing line are "like". Hence, the scope of Article III:4 with regard to origin-neutral national rules firstly depends on how broadly or narrowly one defines the "likeness" of goods. This is also the reason why the debate on the limits of national regulatory autonomy and "aims and effects" mostly relates to the question whether elements beyond end-uses such as health or environmental risks should result in goods not being "like". Accordingly, the WTO Member whose rule is under scrutiny usually defends itself by claiming "unlikeness".
However, the reach of Article III:4 also depends on when to conclude that imported products are treated "less favorably" than like domestic goods. Whether and in what manner this requires discrimination or protective effect to the benefit of domestic goods depends on how one applies this condition. Once the differently treated products have been found to be "like", there are, both in theory and in practice, two fundamentally different and inconsistent ways of identifying less favorable treatment.
The first of the two practiced approaches, called in this article the "diagonal test", 8 asks whether there are any imports receiving less favorable treatment than any like domestic products. The concepts of GATT Article III:4 and the origin-neutral measure divide the entire universe of like products into four possible categories. The rule favors some products and disfavors others, potentially both among imports and domestic products. An example would be a ban on asbestos in construction materials, a limitation of high-alcohol beer sales to liquor stores or a mandatory percentage of recycled glass in glass containers. Such measures favor asbestos substitutes, recycled glass and low-alcoholic beer; they disfavors asbestos, all-new glass and strong beer.
Under the diagonal approach, one compares the treatment accorded to the disfavored foreign goods with that enjoyed by the favored domestic goods, e.g. domestic low-alcohol beer receives more favorable treatment than imported strong beer. This means that it is always possible to find a violation of Article III:4, as long as the disfavored type is (potentially) imported and the favored type exists domestically. Even when imported products overwhelmingly receive the better treatment and most domestic goods fall in the less advantageous category, there is still a violation.
The alternative approach, called in this article the "asymmetric impact test", 9 compares in aggregate both domestic sub-groups with both foreign sub-groups. The treatment received by imports is only less favorable than that accorded to like domestic goods if the burden arising from the measure is heavier for imports than for domestic goods. This can be the case because, despite the equal treatment flowing from the origin-neutral measure, it is comparatively more burdensome for foreign suppliers to comply with the rule. Due to this disparate intensity of the burden, this first scenario may be described as asymmetry between the respective intensity of the measure's effect on imports and domestic products (asymmetric intensity).10 Otherwise, if the burden appears to be equivalent in intensity, it is necessary that most imports fall into the disfavored regulatory category and/or most domestic products fall into the favored group. This latter consideration, an asymmetric distribution of imports and domestic goods, is precisely what the diagonal test omits and what this paper is about. There must be an asymmetry in the composition of imports and domestic goods which is responsible for the rule's disparate impact on imports to the advantage of national products. If that is the case, the origin-neutral national rule treats imports and like domestic goods identically de jure. But de facto or indirectly, like imports bear a heavier overall burden than the like domestic products. The origin-neutral distinction negatively affects the competitive relationship of the group of imports in relation to the group of like domestic products and thus protects the latter against the former.
Conversely, there is no automatic violation of Article III:4, if the quantitative ratio between favored and disfavored products is equivalent on both sides. For this equivalence to exist, it is not sufficient that some domestic products bear the heavier regulatory burden. Instead, it is necessary that, as a whole and relatively, the entire groups of imports and domestic products bear the burden in an equivalent proportion. In that case, the aggregate group of like imported products not only de jure, but also de facto receives no less favorable treatment than the entire group of domestic products. Simultaneously, the disadvantage arising from the measure for parts of both domestic and imported like products is not even indirectly related to origin. It is only related to the origin-neutral (e.g. physical) characteristics on which it depends.
Regulatory distinctions thus remain possible as long as they do not indirectly lead to the result which Article III outlaws: that domestic products receive more favorable treatment than imports and protection against imports. This approach allows more flexibility in applying regulatory distinctions 11 and makes considerations of "aims and effects" unnecessary in cases involving a symmetric distribution of imports and like domestic goods across the regulatory sub-categories. 12
A graphic illustration helps to conceptualize the problem addressed in this paper by describing the operation of both diagonal and asymmetry tests. It also serves to highlight the fundamental contrast between these two approaches.
* Banned (High Tax)13
The asymmetric factual matrix of the first table reflects a situation of de facto discrimination against imports under both asymmetry and diagonal approaches. Most of the goods, which are or could be imported and fall within the group of like products, face the less advantageous regulatory treatment. Simultaneously, most like domestic products receive the more favorable of the two available treatments. A comparison of the two entire groups of like products, domestic and imported, along the lines of the thick arrow demonstrates the imbalance or asymmetry. One would obviously also find that some of the imports receive less favorable treatment than some like domestic products by comparing the diagonally opposite sub-categories along the lines of the wavy arrow (diagonal approach).
The second table reflects a symmetric distribution of goods across the regulatory sub-categories. Since the ratio between the numbers of goods in both sub-categories is equivalent for imported and domestic like products, the asymmetry test (thick arrow) would not result in a finding of discrimination. In contrast, using the diagonal comparison (wavy arrow) as a tool, one would nevertheless find a violation.
The third table illustrates a situation that one could call "reverse asymmetry" in the distribution of goods across the regulatory sub-categories. The majority of domestic goods fall within the less favorably treated sub-category, whereas most like imports receive the more beneficial treatment available under the formally origin-neutral law. According to the asymmetry approach (thick arrow), the regulatory differentiation thus results in a de facto reverse discrimination against domestic goods that is not outlawed by WTO law. Nonetheless, the diagonal approach (wavy arrow) again permits a finding of less favorable treatment to the detriment of some like imports, irrespective of the fact that imports on the whole benefit from the distinction in far greater proportion than like domestic goods.
Only in theory, there is also a third way to apply the test, which would never result in an origin-neutral measure violating GATT Article III:4, unless it imposed a qualitatively higher burden on foreign goods. If the comparison of the treatment accorded to like products follows the dividing line of the regulation and takes place within the horizontal sub-categories created by the regulatory differentiation, 14 the disadvantaged sub-group among imports receives no less favorable treatment than the corresponding domestic disadvantaged sub-group. However, such a subdivision of the whole group of like products on both sides finds no basis in the wording of Article III. 15 It would also contravene its purpose, preventing discrimination, be it de jure or de facto discrimination. 16 The limitation of the comparison to the sub-categories would enable WTO Members to circumvent their obligations by designing measures with facially origin-neutral criteria, which in fact result in better treatment of the majority of domestic products. Such a scenario is, therefore, correctly understood as de facto discrimination through (formally) equal treatment, 17 as is the scenario in which a measure imposes a requirement that is more burdensome for foreign suppliers than for the domestic industry (asymmetric intensity). 18
Whilst the approach of comparing the treatment of foreign and national goods only within each regulatory sub-category is not viable, a very narrow interpretation of "likeness" can sometimes achieve a similar result. If the interpretation of "likeness" is sufficiently narrow, it may prevent the differently treated goods from entering the same product category (like products). If the differently treated goods are not "like" there can, as a consequence, be no violation of GATT Article III:4. However, it is clear that the regulatory distinction itself cannot, as a general rule, be sufficient for the conclusion that goods are no longer "like" one another. Such an approach would again render the concept of de facto discrimination meaningless and permit an easy circumvention of many non-discrimination disciplines.
Admittedly, the two approaches outlined above (asymmetry and diagonal tests) are mutually exclusive only from a strictly conceptual standpoint. Practically speaking, it is conceivable to use them both in combination. The asymmetric impact of a measure on imports could serve as a contributing factor when assessing the compatibility of national measures with the WTO Agreement. Asymmetric impact would strengthen the case of the complainant that the defendant denies national or most-favored-nation treatment. However, it would not be a necessary condition for a finding of less favorable treatment. 19 If the measure has no asymmetrical impact on imports, other factual aspects of the measure (its application, or the objective legislative design) could still result in a finding of violation. Conversely, an asymmetrical impact could merely have probative, but no dispositive value for a finding of less favorable treatment. Similarly, the "aims and effects" test sometimes includes the considerations presently discussed as asymmetric impact. 20
The same question, whether or not asymmetric impact is necessary, arises in connection with GATT Article III:2. 21 It is of particular relevance in the field of taxation, given that domestic laws often provide for a multitude of different consumption tax rates, be they general sales or value added taxes with different rates or special excise levies on particular product categories. The delimitations between the different rates at times cross within groups of like products and even more often within directly competitive or substitutable products.
In the context of Article III:2, first sentence, the interpretative choice has features identical to Article III:4. Once the likeness of differently taxed products has been established, the question is whether the imports are taxed "in excess" of the rate applied to like domestic products. A tax regime, which formally applies identically to both the national production and imports, automatically creates four possible categories if the distinction does not respect the borders of likeness. 22 Is it sufficient for a finding of "excess" taxation that there are high-taxed (potential) imports and low-taxed like domestic products? In other words, need there only be some imports taxed in excess of some like national goods? Or is that an excessively low threshold, given that it only requires that the mentioned sub-groups exist and that it is met even when both imports and like domestic products suffer and benefit in an equivalent manner?
In the alternative, the comparison would have to take place between imported and like domestic products, taken as entire groups. It would be necessary for excess taxation that (potential) imports fall predominantly under the higher tax and/or that the like domestic goods predominantly benefit from the lower tax. 23 The question of diagonal versus asymmetry test must not be confused with a de minimis threshold for "excess taxation", as the latter only relates to the amount of the difference, not the dimension in which it exists.
The second sentence of Article III:2 read in conjunction with the Ad Note raises the same issue notwithstanding the different wording. Once established that the disfavored imports directly compete with the favored imports, the test is whether the taxation is dissimilar and applied "so as to afford protection to domestic production." Both the questions of dissimilar taxation and afforded protection again depend on how to compare domestic goods and substitutable imports: as aggregates or diagonally. 24 Since the criteria of dissimilar taxation and afforded protection are cumulative conditions, it would be sufficient for one of these two requirements to include an asymmetry test, but of course they could also both include it. 25 The other conditions of a violation of Article III:2, second sentence, inter alia that the rate differential be de minimis in order for the taxation to be dissimilar, remain unaffected.
Mutatis mutandis, the most-favored-nation principle under Article I:1 26 raises the same question. The main difference is that the comparison operates in another dimension, not between national and imported goods, but between products from different source countries. 27 A violation presupposes that the advantage is not accorded immediately and unconditionally to like products originating in the territories of all other WTO Members. In the event that some imports from one country enjoy an advantage under objective, origin-neutral conditions, can another country always claim this advantage for its like exports which do not meet the objective conditions? In fact, could both countries have that claim under Article I:1? Or is it necessary that the conditions disfavor the entire group of like exports in comparison with that group of the other exporting country? 28
Article I:1 gains additional importance through its cross-reference to Article III. As a matter of substantive coverage, the most-favored-nation obligation reaches beyond border measures to the fields of taxation and regulation. While this aspect of Article I:1 has been qualified as being superfluous, if all imports receive national treatment,29 that only holds for formal discrimination on the basis of origin. In relation to de facto discrimination, the reference maintains importance since it adds another dimension in which the comparison occurs: not only between imported and like domestic products (Article III), but also between imports from different countries (Article I:1 in conjunction with Article III). In other words, in taxation and regulation, the legislating Member owes, in relation to imports from WTO Members, not only treatment as favorable as that accorded to like national goods, but also treatment as favorable as that of like imports from other countries. 30 This additional dimension is valid under both the diagonal and the asymmetric impact approach and may matter depending on the kind of products which are imported from particular countries as opposed to those produced domestically.
Most-favored-nation and national-treatment provisions in other provisions of the GATT (e.g. Article XIII) and other covered agreements (e.g. GATS Articles II and XVII) equally raise the question whether to apply the diagonal or the asymmetry test irrespective of minor differences in wording.
7 Under the Article's title "National Treatment on Internal Taxation and Regulation" the first sentence of GATT Article III:4 reads:
The products of the territory of any contracting party imported into the territory of any other contracting party shall be accorded treatment no less favourable than that accorded to like products of national origin in respect of all laws, regulations and requirements affecting their internal sale, offering for sale, purchase, transportation, distribution or use.
8 "Traditional interpretation" according to William J. Davey & Joost Pauwelyn, MFN Unconditionality: A Legal Analysis of the Concept in View of its Evolution in the GATT/WTO Jurisprudence with Particular Reference to the Issue of "Like Product", in Regulatory Barriers and the Principle of Non-Discrimination in World Trade Law 13, 41 (Thomas Cottier and Petros C. Mavroidis eds., 2000).
9 "Geographical distribution of discriminatory effects" according to Ole K. Fauchald, Environmental Taxes and Trade Discrimination 225, 232, 245, 400 (1998). "Discriminatory effect requirement" according to William J. Davey & Joost Pauwelyn, supra note 8, at 38-41. In order to make the notion of "discriminatory effect" unambiguous, it is necessary to add that the requirement is an overall or aggregate discriminatory effect to the benefit of domestic products or, as the authors do, to require a discriminatory impact related to origin, id., at 40. Otherwise, the diagonal test also requires some discrimination or protection, but a protection of a part of the domestic industry at the expense of a part of like imports.
10 For reasons of space and clarity, this scenario - in which the disparate impact of an identical legal provision de facto results in less favorable treatment of imported products (asymmetric intensity) and a WTO Member would thus have to apply different legal provisions to domestic and imported products in order to ensure equally favorable treatment (see Panel Report, United States - Section 337 of the Tariff Act of 1930, adopted 7 November 1989, BISD 36S/345, para. 5.11) - is not the focus of this paper. This does not mean that the conceptual framework of this paper does not apply to such a scenario: the impact of a measure may well be disparate not for all, but only for a part of imports. It may simultaneously be disparate for a part of domestic goods. Then, too, the question is whether it is sufficient that a measure, which applies to all goods irrespective of origin, burdens only some isolated imports more heavily than some domestic like products.
11 See also William J. Davey & Joost Pauwelyn, supra note 8, at 40.
12 See also Ole K. Fauchald, supra note 9, at 236.
13 The content of the brackets relates to the context of taxation, i.e. GATT Article III:2. This question will be discussed infra in section III.B.
14 See the two dotted arrows in the first illustration supra in section III.A.3.a), where the two rows designate the two regulatory sub-categories resulting from the legislative differentiation.
15 See, in relation to Article III:2, Ole K. Fauchald, supra note 9, at 216.
16 John H. Jackson, World Trade Rules and Environmental Policies: Congruence or Conflict?, 49 Wash. & Lee L. Rev. 1227, 1236-1237 (1992). For the opposing view, see David M. Driesen, What is Free Trade? The Real Issue Lurking Behind the Trade and Environment Debate, 41 Va. J. Int'l L. 279, 350-351 (2001).
17 Aristotle has already expressed that injustice can arise both from unequal and equal treatment. See Aristotle, Nicomachean Ethics, 1131a (Martin Oswald trans., 1962); Aristotle, The Politics, 1280a (Carnes Lord trans., 1984).
18 See supra section III.A.2, note 10.
19 See also Ole K. Fauchald, supra note 9, at 224, for an equivalent hypothesis operating in the opposite sense: the "diagonal test" is understood to be a presumption of violation, which might be refuted with the proof of absent disparate impact of the measure on imports. However, according to the author, it is uncertain whether this presumption can actually be rebutted given the statements of panels and the Appellate Body, id., 223.
20 Ole K. Fauchald, supra note 9, at 224-225 and note 32.
21 Under the title "National Treatment on Internal Taxation and Regulation" Articles III:1 and III:2 provide:
1. The contracting parties recognize that internal taxes and other internal charges, and laws, regulations and requirements affecting the internal sale, offering for sale, purchase, transportation, distribution or use of products, and internal quantitative regulations requiring the mixture, processing or use of products in specified amounts or proportions, should not be applied to imported or domestic products so as to afford protection to domestic production.
2. The products of the territory of any contracting party imported into the territory of any other contracting party shall not be subject, directly or indirectly, to internal taxes or other internal charges of any kind in excess of those applied, directly or indirectly, to like domestic products. Moreover, no contracting party shall otherwise apply internal taxes or other internal charges to imported or domestic products in a manner contrary to the principles set forth in paragraph 1.
The interpretative Note Ad Article III:2 states:
A tax conforming to the requirements of the first sentence of paragraph 2 would be considered to be inconsistent with the provisions of the second sentence only in cases where competition was involved between, on the one hand, the taxed product and, on the other hand, a directly competitive or substitutable product which was not similarly taxed.
22 See the illustrations supra in section III.A.3 with the references to taxes.
23 See also William J. Davey & Joost Pauwelyn, supra note 8, at 39.
24 Id., at 39-40.
25 Regarding the question of whether the asymmetry test applies already for "not similarly taxed", see infra note 71 and accompanying text.
26 Under the title "General Most-Favoured-Nation Treatment" the relevant part of GATT Article I:1 states:
With respect to customs duties and charges ... and with respect to all matters referred to in paragraphs 2 and 4 of Article III, any advantage ... granted by any contracting party to any product originating in or destined for any other country shall be accorded ... to the like product originating in or destined for the territories of all other contracting parties.
27 For the purpose of the most-favored-nation obligation, the illustrations supra in section III.A.3 must be read in the following manner: the column with the title "domestic" represents the imports from one WTO Member and the column with the title "imported" represents the imports originating in other countries.
28 William J. Davey & Joost Pauwelyn, supra note 8, at 40.
29 Alan O. Sykes, Product Standards For Internationally Integrated Goods Markets, 67 note 21 (1995).
30 See also Ole K. Fauchald, supra note 9, at 216.