Jean Monnet Center at NYU School of Law



Previous |Next |Title |Contents


III. The Aim-and-Effect Test


A. Rationales for the Aim-and-Effect Test

1. Fiscal Sovereignty Generally

In international trade agreements, fiscal powers remain within the realm of government sovereignty as long as the measures are not used to alter the competitive opportunities of products. Different governments make different social choices, many of which are manifested in the fiscal policies of the state. Whether it is to support a certain region, increase or decrease the use of a raw material or a means of production, or encourage small enterprises, governments frequently distinguish between products for non-trade policy reasons. However, an infinite number of non-trade fiscal policies can distort the market and affect trade. Should they all be struck down as discriminatory? The Contracting Parties cannot have intended such a restraint on their sovereignty. As the Panel in Malt Beverages stated, "the purpose of Article III is not to prevent contracting parties from differentiating between different product categories for policy purposes unrelated to the protection of domestic production."[13] To preserve their fiscal sovereignty, the Contracting Parties must have the opportunity to distinguish valid exercises of fiscal sovereignty from policies intended to undermine the goals of the GATT. Without the chance to show that measures fall within the permissible realm of non-trade policies, governments will be wary of enacting non-trade policies for fear that these policies may have unintended effects on trade and may result in dispute proceedings. It is precisely these concerns that the Panel and the Appellate Body did not adequate weigh in choosing the test to determine a violation of Article III:2.

2. Product Differentiation

Given that governments distinguish between products for both valid non-trade policy and discriminatory trade policy reasons, a panel's assessment of the product differentiations is crucial for the fiscal sovereignty of the Contracting Parties. The Panel in US-Auto Taxes concluded that notions of likeness and direct competition cannot be separated from the objectives of the regulatory scheme that draw a line between two otherwise similar products.[14] The aim-and-effect test therefore integrates the purpose of the fiscal measure into the determination of a violation of Article III:2. The Panel's test, by contrast, excludes the purpose of the measure from its assessment of the government's product differentiations. As a result, the purely descriptive test fails to create an objective means for distinguishing between products and limits the sovereignty of the Contracting Parties.

If a panel uses descriptive criteria to determine whether two products are "like" or "directly competitive" for the purposes of Article III:2, the panel can only make arbitrary distinctions between products based upon its "best judgment" and a case-by-case analysis. There is no guidance as to what constitutes "like" or "directly competitive" products. Under the Panel's test, the definition of "like" consists of physical characteristics, end-uses, and tariff classifications. But why does one physical characteristic determine likeness and another does not? Is the color of the alcohol most important or the distillation process? Does the type of sugar used in the process determine likeness or it is more pertinent to know if the alcohol is usually diluted before consumption? Under the Panel's test, products are considered "directly competitive" if they have similar consumer end-uses as shown through the marketplace. But how similar do the end-uses have to be? Which group of consumers determines the end-uses if different age groups, for instance, use a product differently? Depending upon how a product is characterized or how the consumer group is defined, an infinite number of products can be considered "like" or "directly competitive." Where should the line be drawn? Distinctions made by reference to the Panel's criteria are arbitrary because the criteria is only descriptive.[15] Panels are therefore singling out characteristics at random. They have no objective way to determine which characteristics or market analyses are most illustrative of likeness or direct competition.

Whether products are "like" or "directly competitive" under Article III:2 depends upon why the product differentiation is being made. The supposedly objective criteria of physical characteristics and end-uses are not objective because they have been singled out by policy-makers for a reason. From the government's point of view, if two products do not serve the same policy purpose, then they cannot be "like" or "directly competitive" or they would undermine the government's policy. It is only this policy that can guide a panel as to which criteria it should single out to determine if products are "like" or "directly competitive" for the purposes of Article III:2. For instance, are bicycles and motor scooters "like"? Yes, if one compares them to cars in price and fuel efficiency. No, if one compares them to each other in price, consumers, manufacturing processes, and environmental impact. Governments might classify bicycles and scooters together and tax them less than cars to discourage the use of cars or gasoline. Governments might place them in different categories and tax bicycles less than scooters to enable the poor to purchase a means of transportation or to limit noise pollution. Only by determining the policy reasons behind a measure can a panel objectively decide if a measure violates Article III:2.

The natural effect of this more objective legal reasoning process is the protection of the fiscal sovereignty of the Contracting Parties. In determining the government's rationale for differentiating between products, a panel is also assessing whether that rationale is a valid exercise of fiscal sovereignty. A panel must consider the aim of a measure to determine which differences between products are permitted to form the basis for tax distinctions. The permissible differences between products are those that stem from valid non-trade policies. It is these non-trade policies that lie within the fiscal sovereignty of the Contracting Parties. A test that does not assess the purpose of the measure cannot protect fiscal sovereignty because it fails to ask the crucial question. The aim-and-effect test, by contrast, inherently protects fiscal sovereignty because it has integrated the inquiry into the validity of the measure into its determination of a violation of Article III:2.

Although a panel must examine the aim of a measure before it can objectively determine a violation of Article III:2, a panel encroaches on the fiscal sovereignty of the Contracting Parties when it engages in an unrecognized or unspoken motive analysis. This type of analysis occurs when panels use their "best judgment" without any objective guidance to supplement the descriptive criteria.[16] Implicit motive analysis occurs because the categories of "like" or "directly competitive" and the distinctions between products reflect government policy decisions. When a panel begins its inquiry with the government's categories, it is immediately involved with and implicitly assessing the policy choices of the government. A panel may also knowingly weigh the government's claimed policy aim against the alternative reasons offered by the complainants but may refuse to admit this process for fear of calling the defendant government a liar. When a panel accepts the government's categories, it is, therefore, implicitly or explicitly, approving the government's reasons for distinguishing between products. The panel is stressing the same characteristics and end-uses as the government did. As the government singled out these elements to achieve its policy goals, the panel must find these policy goals compelling or else it would reject the government's categories and single out other characteristics.

Given the inevitable motive analysis in the Panel's test, a panel could find itself in an awkward situation. It might agree with the government's policy motives but the application of the descriptive criteria would place products in categories that would undermine the government's policy. To divide the products as the government did, the panel would be forced to distort either the categories or the product characteristics. For instance, if a government wants to support small breweries by taxing them less than large breweries, and there are equal numbers of both sized breweries in the country, a panel would have to decide that beer from small and large breweries are neither "like" nor "directly competitive." It could be quite difficult to argue that the beers have different characteristics, different end uses, and different consumers. By relying on the descriptive criteria, the panel undermines its credibility by engaging in a convoluted reasoning process.

This unacknowledged or unspoken motive analysis interferes with the sovereignty of the Contracting Parties both during the dispute and before the dispute, in legislative and administrative processes. During the dispute, the defendant government has no opportunity to rebut the claims against its policies. Rather, the panel decides for itself, based upon limited information and its own intuition, whether to accept or reject the government's categories and policies. Similarly, it gives policymakers no guidance as to acceptable non-trade policies and distinctions between products. Governments will hesitate to enact any non-trade measures that may have an unintended impact on trade because they will have no opportunity to show their valid motives. Both the lack of notice about a panel's reasons for a judgment and the inability of a defendant government to respond to a motive analysis encroach on the sovereignty of the Contracting Parties.

The weaknesses of the Panel's test, namely arbitrary line-drawing and an elusive and convoluted legal reasoning process, illustrate the need for panels to openly determine the purpose of the measure. By inquiring into the aim of the government's policy, the aim-and-effect test creates an objective and transparent method to assess the differences between products and determine a violation of Article III:2. Even more importantly, in making this crucial assessment about product differentiations, the aim-and-effect test protects the fiscal sovereignty of the Contracting Parties.

3. Article III:2 and Article III:4

When a panel considers the aim of a measure before determining if a government has validly differentiated between two products under Article III:2, it also expands the tools a government can use to implement its policy. By equalizing the incentives for governments to pursue taxation or regulation, the aim-and-effect test protects the sovereignty of the Contracting Parties. Discriminatory regulation is prohibited in paragraph 4 of Article III and the accepted test for Article III:4 includes an evaluation of the aim of the regulation.[17] As both Article III:4 and Article III:2 seek to prevent governments from undermining the GATT through internal measures, the two paragraphs should produce the same outcomes. This equality would enable governments to choose freely between taxation and regulation. Such a result occurs when the aim-and-effect test is used, but not when a panel relies exclusively on descriptive criteria.

If a panel only uses descriptive criteria for Article III:2, it undermines its own credibility and skews the choice of policy tools of the Contracting Parties. By relying on descriptive criteria for Article III:2, a panel is using a narrower test than the accepted test for Article III:4.[18] The discrepancy between the tests for Article III:2 and Article III:4 may force a panel to manipulate the categories and product differences for Article III:2. Consider the case where imported wine is taxed more heavily than domestic beers and advertisements for wines are more restricted than those for beer and the aim of both measures is clearly discriminatory. A panel would find the difference in advertising regulations violates Article III:4 because wine and beer would be "like" products under the broader Article III:4 test. However, it would be difficult for a panel to determine that wine and beer are "like" for the purposes of Article III:2 because these products have different physical characteristics, potentially different end-uses, and possibly different tariff classifications.[19] In order to find both measures discriminatory, a panel would either have to consider the aim of both measures or would be forced to manipulate categories and product characteristics to find a violation of Article III:2.

From a government's perspective, a discrepancy in the tests for Article III:2 and Article III:4 limits its fiscal sovereignty on two levels: during the dispute and during legislative and administrative decisions. During a dispute, a government can comply more easily with Article III:4 than with Article III:2. A government can argue that a regulation is valid because it was enacted for non-trade reasons but it does not have the same defenses for its tax measures. Moreover, the convoluted legal reasoning process of panels gives a government no notice ex ante as to how the two definitions of "like" will be interpreted or how far the coverage of the two paragraphs will extend. If a panel refuses to consider the aim of a measure in its test for a violation of Article III:2, it skews a government's choice of method for implementing a policy towards regulation and away from taxation.

When the aim-and-effect test is used for Article III:2, there are no discrepancies in the definitions of "like" in Article III:2 and Article III:4 or in the coverage of the paragraphs. Both definitions of "like" are interpreted broadly and include an evaluation of the purpose of the measure. The coverage of the paragraphs is the same because both paragraphs are focused on the protective application of the measure. The equality means that a government can defend the aim of both its tax and regulatory measures and is on notice about the important criteria for a panel in both situations. There is no longer an incentive to implement a policy through regulation. As a result, the aim-and-effect test enables governments to choose freely between the two methods and protects the sovereignty of the Contracting Parties on this level, as well as during the assessment of the government's product differentiation.

4. The Article XX Issue

Admittedly, after finding a violation of Article III:2, the Panel's test permits the defendant government to defend its tax measure based on the list of exceptions in Article XX.[20] However, because the list provides only seven possible defenses, it restrains the Contracting Parties rather than liberates them from the Panel's narrow Article III:2 test. For instance, if a government wants to impose a luxury tax, it has no recourse under Article XX because there is no exception for redistributive policies. The government would have to rely on a panel to determine that a Mercedes Benz is not "like" or "directly competitive" with a Honda and the government can tax the vehicles differently. The Panel's test forces a panel to either reject valid exercises of fiscal sovereignty or, in the Article III:2 inquiry, to develop arbitrary distinctions and similarities between products and engage in a motive analysis.

Even though Article XX does not mitigate the violations of the sovereignty of the Contracting Parties in the Panel's test, the Panel and Appellate Body used Article XX to reject the aim-and-effect test. They claimed the aim-and-effect test would render Article XX redundant. A panel should not examine the aim of a measure during the Article III:2 inquiry because Article XX provides the only permissible policy rationales. Article XX plays a different role than the Panel and Appellate Body envisioned, however, and does not require the rejection of the aim-and-effect test.

Article XX only applies when there is a facial violation of the GATT obligations, as when a measure is origin-specific. Although Article XX does not refer to origin-specific measures, limiting Article XX to such violations can be inferred from the fact that Article III:2 can only apply to origin-neutral measures. The Article III:2 inquiry into whether products are "like" or "directly competitive" is only relevant if a panel is unsure if the measure is "applied...so as to afford protection to domestic production." Uncertainty about the protective aim and effect of a measure does not arise with origin-specific measures. These measures target imports and are, by definition, applied to protect domestic production. Uncertainty about the aim and effect of a measure only arises when the measure is origin-neutral. It is in these situations where the aim-and-effect test is necessary to determine if products are "like" or "directly competitive." If Article III:2 applied to origin-specific measures, the ensuing Article III:2 inquiry would be redundant because the protection is evident. As Article III:2 cannot apply to origin-specific measures, a panel must look elsewhere for the violation and the exceptions. That facially discriminatory measures are per se violations of the GATT is found throughout the Agreement, such as in Article I:1 or Article III:1. Article XX, in turn, enables a government to argue that its origin-specific measure is necessary to achieve a non-trade policy. As such violations are egregious, the list of exceptions must be narrow and well-defined.

Another approach to the role of Article XX is found in the language of the Appellate Body in United States-Standards for Reformulated and Conventional Gasoline. The purpose of Article XX is to "enumerate the various categories of governmental acts, laws or regulations which WTO Members may carry out or promulgate in pursuit of differing legitimate state policies or interests outside the realm of trade liberalization" (emphasis added).[21] Article XX secures areas of government policy that are not subject to the GATT. The language of Article XX supports this analysis. It states, "Nothing in this agreement shall be construed to prevent the adoption or enforcement by any contracting party of measures necessary to...." Article XX outlines what the Agreement cannot do, rather than what governments can do. It defines the parameters of the Agreement, not what is permissive within the bounds of the Agreement. Questions of permissive behavior within the bounds of the Agreement are analyzed under the main provisions of the Agreement such as Article III:2. Article III:2 recognizes the interests within the realm of trade liberalization, i.e., that taxes cannot be applied with the aim and effect of protecting domestic products. The Contracting Parties have already agreed that, while most taxes are exercises of the fiscal sovereignty of the parties, some taxes are not permissible. They are giving up a measure of fiscal sovereignty in return for trade benefits. Article XX, by contrast, protects the areas of government policy that will not be subject to an accord or restricted for the purpose of trade liberalization.

Finally, Article XX cannot be an exclusive list of the exceptions to the GATT because the Contracting Parties have included additional exceptions based on the larger principles of the GATT in related contexts. For instance, the Agreement on Technical Barriers to Trade (TBT), incorporated into the GATT and considered an elaboration on Article III, permits governments to enact technical regulations that fulfill "legitimate objectives." The TBT Agreement sets out legitimate policies that permit governments to distinguish between products that are otherwise "like" or "directly competitive." The list of policies in the TBT Agreement includes measures not contained in Article XX. As the TBT is consistent with the principles of the GATT, then there are policies in addition to those listed in Article XX which are still compatible with these larger principles. Article XX is only a subset of this larger group of acceptable policies. Article III:2, like the TBT Agreement, is consistent with the larger principles of the GATT because it is concerned with protective measures. As Article III:2 embodies these principles, then acceptable non-trade policies under Article III:2 are similarly a sub-set of the group of policies compatible with the GATT, just like the policy exceptions in the TBT Agreement and in Article XX.

If the Panel and the Appellate Body had understood the significance of their decisions, they would have been more sensitive to the restrictions on the fiscal sovereignty of the Contracting Parties that result from their reading of Article XX. Under accepted rules of interpretation in international law, the preferred interpretation of an agreement is the one that best preserves the sovereignty of the contracting parties. All other things being equal, it is assumed that parties have given up as little power as necessary to achieve the goals of the agreement. Article XX should be interpreted not to apply to cases of origin-neutral measures falling within the realm of trade liberalization. Otherwise, the fiscal sovereignty of the Contracting Parties would be limited to seven permissible policies. Once the argument about the redundancy of Article XX is eliminated, there remains little basis for rejecting the aim-and-effect test in favor of the Panel's test.

B. The Mechanics of the Aim-and-Effect Test

1. Textual Basis

The aim-and-effect test would not be a viable option, even with its sensitivity to the fiscal sovereignty of the Contracting Parties, if it did not stem from a solid textual and contextual basis. In considering Article III:2, a panel should look to both the text and the context of the paragraph. The context is important for two reasons. First, the texts of Article III:2 and its interpretive note, setting out the requirements for "like" products and "directly competitive" products, respectively, are ambiguous. Article III:2 states:

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 interpretive note, Ad Article III, paragraph 2, adds:

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.

The interpretive questions are numerous. For instance, does "moreover" indicate another category of products or another set of criteria for "like" products? "Moreover" means "besides," is a synonym for "also" and "likewise," and stresses the importance of what is to come. Does Ad Article III:2 mean that the first and second sentences of Article III:2 set out different standards for different categories of products? Or, does it close a loophole by stating that other product differentiations, not just those between "like" products, also violate Article III:2?

The second reason it is necessary to look beyond the text is that the Vienna Convention on the Law of Treaties requires examining the context of the terms in question. Specifically, Article 31(1) provides:

A treaty shall be interpreted in good faith in accordance with the ordinary meaning to be given to the terms of the treaty in their context and in the light of its object and purpose.[22]

As shown, the meanings of the words of Article III:2 are not "ordinary" and, therefore, the context and the object and purpose of the Article become even more important.

GATT 1994 Article III:1 forms the crucial context of Article III:2. Article III:1 sets out the general obligations of the Contracting Parties on which the subsequent paragraphs elaborate:

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.

The key language, "internal taxes...should not be applied...so as to afford protection to domestic production," indicates that both the aim and the effect of the measure are important for a panel to evaluate. "So as to" means "in order to" and indicates a conscious decision to apply the measure in a certain way to achieve the chosen goal. Although this phrase could be ambiguous as to whether it includes both aim and effect, the active decisional quality of this phrase is unmistakable. If only the effects of the measures are examined, an infinite number of tax measures could be "applied...so as to protect domestic production," thereby limiting the sovereignty of the Contracting Parties. However, it is neither the object nor the purpose of the Agreement to restrict the sovereignty of the Contracting Parties by encroaching on valid non-trade policies. Rather, the Agreement seeks to prevent overt discriminatory trade policies and disguised restrictions to trade. Any interpretation that limits a government's internal non-trade polices would be more than the Contracting Parties intended and should be rejected .

As Article III:1 outlines the goals and meaning of the whole Article, its emphasis on protective measures should permeate all the tests designed for the Article.[23] The most important element of any test is to determine if the measures and product designations are "applied...so as to afford protection to domestic production." To further underscore this concern, Article III:2 actually refers to the principles of Article III:1 as the standard. Both measures designating products as "like" or "directly competitive" must overcome the hurdle in the second sentence of Article III:2. The conjunction, "moreover," links the two sentences of Article III:2 so that potentially "like" products are included in the same test as potentially "directly competitive" products. As a result, whether products are "like" or "directly competitive" can only be determined by referring to the language of Article III:1, in other words, to the aim and the effect of the measure. This textual basis supports the legal reasoning and sovereignty arguments in favor of the aim-and-effect test.

2. The Operation of the Test

The aim-and-effect test is not only viable on theoretical and textual levels, but also in implementation. The inquiry into the aim of the measure does not require investigating the subjective understandings or intents of individual legislators or the legislature. Rather, it examine the purpose of the measure as embodied in the measure itself. As stated in the US Appellate Submission, "the focus [is] on the direction in which the gun [i]s pointing rather than whether the person holding the gun even want[s] to pull the trigger."[24] Purpose is shown by examining the nexus between the means and ends of the legislation. For instance, what is the structure of the measure, does it draw a line between imports and domestic products, what incentives does it create, what is the wording of the legislation as a whole, and how "exceptional" are the categories it articulates? Is the measure proportional to its stated goals and does it create the least degree of inconsistency with the GATT provisions generally? It is also important to examine if the defendant government has presented a plausible alternative to the complainants' claims of protective intent.

To clarify, an "exceptional" category means the government has drawn an distinction between products that is not normally part of a general scheme of taxation. For instance, it would require drastic changes in the nature of the product for another product to fit into the protected category. The category may include detailed requirements or may be defined by the exclusion of other potentially similar products.[25] It could be argued that if a product is forced to change its nature to qualify for the "exceptional" category, then it is a different product. However, because a government can single out any characteristic to include in a category, it is arbitrary rather than illustrative to say that a product cannot fit into a government's category without changes. A government cannot be absolved from drawing an "exceptional" category by stating that the products are "directly competitive." A narrowly drawn category indicates that the government may have been pursuing a protective purpose mind when designing the measure. Nor is it acceptable for a government to claim the categories follow "socially established boundaries."[26] Such a claim illustrates why "exceptional" categories may be problematic: they can reinforce the historical biases of consumers which may have been determined, in part, by the protection of the products in these categories.

The aim prong of the aim-and-effect test may also use the descriptive criteria as evidence of the protective intent of the measure. As a result, the inquiries for determining the likeness of or direct competition between products are slightly different. For arguably competitive products, a panel might evaluate the incentive structure of a measure by comparing it to current or potential end uses of the products and determining how the incentives might alter product usage. For arguably "like" products, whether a product can fit into an "exceptional" category will depend upon its raw materials and manufacturing processes. The descriptive criteria are not free-standing factors or arbitrarily applied but, rather, they are grounded in and given significance by the inquiry into protective intent.[27]

As the inquiry into the aim of the measure is only concerned with the purpose of the measure as revealed by the measure itself, the burdens on the complainant and a panel to show a discriminatory aim are not onerous.[28] Moreover, the WTO dispute resolution process is designed so that both parties have the roles of claim and rebuttal. The claim of a legitimate non-trade policy does not automatically override obligations under the Agreement. There are also other options for designating the burden of proof. For instance, after the complainant presents a credible case that the measure protects domestic production, the government would have to rebut the claim by demonstrating a legitimate purpose. The burden of persuasion thus falls on the defendant government and the burden of going forward with the evidence lies with the complainant. Alternatively, after the complainant shows a protective effect, the defendant would have to rebut a presumption of protective intent. The burden of persuasion is thus split between the complainant and the defendant.

Concurrent with the inquiry into the purpose of the measure, a panel examines the effects of the measure on the equality of the competitive opportunities of the products. Thus, like the aim prong of the test, the effects inquiry is integrated into the determination of whether products are "like" or "directly competitive" and whether there is a violation of Article III:2. The effects inquiry is not concerned with trade volumes but if the tax rates afford greater competitive opportunities to the domestic product than to the imported product. As such, the complainant only needs to show that circumstances can lead to a protective effect rather than proving a protective effect. The effects of the measure are evident in changes in consumer behavior and in the market share of the products, as shown by data on sales and trade flows. Signs of unequal competitive opportunities include definitions that draw a clear line between domestic and imported products and an increase in the development of domestic production facilities. In some instances, the inquiries into the aim and the effect of a measure will overlap. For instance, examining the incentive structure of the measure for a discriminatory purpose may involve examining the effects of that incentive structure on consumer behavior. Evidence of the protective effects of a measure are often presented in the complaint, so a panel may be less involved in ascertaining the effects of a measure than the aim of the measure. The effects prong of the inquiry may not even be necessary if a panel decides the measure does not have a protective aim.

One piece of evidence used in the inquiry of the effects of the measure is the tax differential.[29] Depending whether the panel's inquiry is into the likeness or competitiveness of products, the use of the levels of taxation might be slightly different. If products are claimed to be "like," the relevant text of Article III:2 states, "the products imported...shall not be subject...to...charges of any kind in excess of those applied...to like domestic products." Any difference in the level of taxation seems protective, although it is not conclusive evidence of protective effect. The main concern is still whether the measure has altered the competitive opportunities of the products.[30] If the products are claimed to be "directly competitive," the language of Ad Article III:2 indicates the taxes must be similar. "Directly competitive" products can be taxed dissimilarly to reflect the dissimilarities of the products, as long as there is no protection. Although the aim-and-effect test cannot determine the tax differential at which protection begins, finding the exact point where the competitive relationship changes is not crucial because the the tax differential is not dispositive.[31]

As a piece of evidence for both potentially "like" and "directly competitive" products, the tax differential can be used to show a protective effect in a variety of ways. For instance, if more domestic products fall into the lower tax category and more imported products fall into the higher tax category, then the structure of the measure raises suspicions of protection. If this division occurs as a by-product of a non-trade policy, the purpose inquiry will protect the government's measure. If a tax is progressive, the tax differential can show whether the rise is consistent or climbs steeply as imports become more prevalent. The magnitude of the difference in taxes levied on the categories could also indicate a protective effect. The tax differential is used in conjunction with other pieces of evidence such as the existence of exceptional categories or changes in consumer behavior in response to the incentive structure and, therefore, is not dispositive.

When a panel uses the descriptive criteria only as pieces of evidence of protection in the aim or the effect prong of the test, it is freed from relying on potentially erroneous factors to determine a violation of Article III:2. For instance, tariff classifications, one of the descriptive criteria, should not be a conclusive indication of the likeness of two products. First, a panel is accepting the government classifications as a starting point for the analysis. It is such government classifications that are often at issue, however. Second, tariff classifications result from compromises among signatories to an agreement and reflect the parties' concessions rather than the likeness or competitiveness of products. Third, governments would be bound to these classifications for more purposes than they originally intended. The reliance on tariff classifications limits the excise taxes a government can charge for non-trade policy reasons. Tax policy would be subordinate to tariff policy and, once again, the fiscal sovereignty of the Contracting Parties would diminish. Given that the aim-and-effect test refers to many pieces of evidence, a panel would not even have to consider tariff classifications or, if it did, it could limit the evidentiary value of the classifications.

Another piece of evidence that the aim-and-effect test incorporates but prevents from becoming dispositive is the cross-price elasticity of demand. As a methodology, cross-price elasticity is too problematic to be indicative of the direct competition between products. For example, it is hard to use and easy to manipulate. The absence of cross-price elasticity does not mean there is no competition between the products. The longer a consumer has to respond to a change in price, the more likely that cross-price elasticity will exist. Cross-price elasticity also varies from one price point to another, with the least elasticity of demand occurring at higher prices. There are also the inevitable effects of income on purchasing behavior. As an interpretive tool, cross-price elasticity of demand is arbitrary. There is no objective way to specify which level of cross-price elasticity determines direct competition for the purposes of Article III:2. On a policy level, cross-price elasticity is difficult for governments to anticipate or prevent. It can occur for reasons outside the control of the government, like technological advances. It also forces policy-makers to change taxes whenever the existence of cross-price elasticity can be shown, such as when the market changes or when a new study is issued. The aim-and-effect test does not need to rely on this evidence to find a violation under Article III:2. In the aim-and-effect test, cross price elasticity of demand serves as one factor to determine if a measure has a protective intent and effect.

Finally, by considering diverse pieces of evidence in determining a violation of Article III:2, the aim-and-effect test weighs many of the same factors that consumers consider in their purchasing decisions. Given that Article III:2 is concerned with the equality of the competitive opportunities of the products, it is counterintuitive for a test under Article III:2 to rely on factors unimportant to the consumer. For consumers, the most important factor in a purchase is generally price, not, for instance, raw materials or manufacturing processes. Even consumers who are concerned with the environmental and health impacts of the raw materials or the manufacturing processes will only purchase environmentally-friendly or healthy products within his or her price range. The second concern of most consumers is likely to be quality, which is often reflected in the price of the product. In the aim-and-effect test, the aim inquiry looks to the incentive structure of the measure which would impact the price and thus consumer behavior. Similarly, the effects inquiry examines the tax differential, which would also alter prices, as evidence. By including factors important to the consumer in determining a violation of Article III:2, the aim-and-effect test can assess the competitive opportunities of products as perceived by the consumers and as evidenced in their behavior.[32]

The discussion of the mechanics of the aim-and-effect test is not complete without a word about the options of the defendant government after a panel recommendation. Under the aim-and-effect test, if a measure does not have a protective purpose or protective effect, then the products are not "like" or "directly competitive." A government can tax the products as it wishes in keeping with its fiscal sovereignty. Once a protective purpose and a protective effect are shown, then the defendant government must bring its measure into compliance. If the products are "directly competitive," there may be a point at which a tax no longer evidences protective effect. Governments could conceivably adjust the tax rates so that the taxes are similar. However, the tax differential is not the only evidence of a protective aim and effect. If a government lowered a tax on a "directly competitive" product but did not eradicate it, there would still be other pieces of evidence pointing to the protective aim and effect of the measure. The government would have to change the incentive structure and the definitions of the categories, for instance, in order to argue that the measure no longer violated Article III:2.

C. Measuring the Tax Differential

In order to fully operationalize the aim-and-effect test, it is necessary that the methodology for measuring the tax differential best reflects the competitive relationship between products. The choice of methodology is the last major piece in establishing the viability of the aim-and-effect test and in protecting the fiscal sovereignty of the Contracting Parties. The tax differential is initially important because without one, it is unlikely that the competitive opportunities have been distorted and the complainants would not have a cause of action. As discussed, a tax differential also serves as evidence of the protective aim and effect of the measure. This evidence is then used to determine if the measure violates Article III:2. Although the tax/price ratio is not the prevailing methodology, it is the most appropriate method for the aim-and-effect test because it measure the competitive relationship between products on a number of levels.[33]

The crucial feature of the tax/price ratio for the aim-and-effect test is that it incorporates the most important factor to consumers. If consumers base their decisions more on price than on other factors, then price should be the basis by which the competitive opportunities of products are measured. To a large extent, consumers determine whether products are competitive. Factors such as prices or taxes that change the perceptions of consumers and their purchasing behavior influence the level of competition between products. Therefore, all other things being equal, the competitive relationship between products and distortion of that relationship are reflected in a comparison of the prices of the products. The tax/price ratio is the only methodology that measures the characteristic of a product which most affects consumer behavior and alters the competitive relationship of products.

If the aim-and-effect test does not use the tax/price ratio to measure the tax differential, then the inquiries into the aim and effect of a measure will be distorted and could threaten the fiscal sovereignty of the Contracting Parties. Because the tax differential serves as evidence that the measure has altered the competitive relationship between the products, it is crucial that the differential measure the change in the competitive opportunities as the consumer perceives the change. Otherwise, the tax differential may reflect alcohol content or volume, arguably of less importance to the consumer. This same tax differential is then used to show that a tax measure may change consumer behavior and may have a protective aim and effect. The result of this incongruity is that the different steps of the analysis into an Article III:2 violation measure different relationships between products.

To illustrate the danger of using the aim-and-effect test with a methodology other than the tax/price ratio, consider the following scenario. The tax/price ratio shows no tax differential between potentially "like" products and only a minimal differential between potentially "directly competitive" products. By contrast, a ratio based, for instance, on alcohol content shows a differential. If a panel uses the latter methodology to decide the measure has the purpose of protecting the domestic product, this conclusion is not necessarily correct. There is no distortion of the competitive relationship if the taxes are the same percentage of the price. The consumer only sees that the higher priced product has a larger final price, i.e., tax plus price, as it should. Consumer behavior is thus unlikely to change, raising questions about whether the measure is protective. The panel, however, by relying on a different yardstick than the tax/price ratio believes that it has found evidence showing protection for the aim-and-effect test. In the end, a panel may not find a violation of Article III:2 depending on the other pieces of evidence. However, it is likely that a panel would decide a case erroneously if it does not use the tax/price ratio in conjunction with the aim-and-effect test. The panel would be encroaching on the fiscal sovereignty of the Contacting Parties by invalidating a legitimate non-trade measure which does not even have a protective effect.

The tax/price ratio is also appropriate for the aim-and-effect test because it reflects multiple elements of the competitive relationship between products. On one level, as discussed, price impacts consumer behavior more than other product characteristics. On another level, price incorporates the alcohol content of the product, volume discounts, economies of scale, the differing costs of production and transportation, the quality of the product, and the costs of complying with government regulations. Price thus reflects both the competitive relationship between products and ensures that all other things, such as the cost of inputs, are equal in that relationship. The tax/price ratio is an objective piece of evidence when a panel is determining if a measure has a protective effect.

There are, however, potential objections to using this methodology. First, if most tax systems are structured around volume or alcohol content, it could be inappropriate for a panel to use a different method. It might encourage governments to engage in ex-post facto reasoning. Governments might claim their measure does not create a tax differential when the tax/price ratio is examined, even though the measure was designed to protect domestic production. If a panel suspects that a government is using the tax/price ratio to rationalize its tax scheme, then the panel could use both the tax/price ratio and the method designated in the measure to verify the government's claims.[34] Governments are also prevented from rationalizing their tax schemes by the tax/price ratio because the fiscal measure would need a mechanism to change the taxes as prices change.[35] Even if a government has implemented such a mechanism, the reevaluation of the taxes has to be frequent enough to keep the ratio constant in a given period.

A second problem with the methodology could arise if the sales markets of the products have divergent structures. For instance, wines vary greatly in price while beers vary less in price. In order to compare the tax/price ratio of these two products, it is necessary and feasible to choose wines that are representative in availability, market share, and quality. Only representative products can reflect a distortion of the competitive relationship between products with different market structures. Another objection is that this methodology says nothing about the appropriate ratios or differentials. However, it is the government's decision to set the ratio in pursuit of its chosen policy. The role of a panel is to ensure that the government maintains the stated ratio and is not protecting the domestic market.

Finally, because this methodology is a ratio, any distortion in the absolute price resulting from consumer biases or from the historical or current cost of taxes does not appear in the tax differential. The canceling out of consumer preferences occurs when the tax changes to reflect consumer bias by the same percentage that the price changes to reflect consumer bias. If the tax does not change, the government is evidently favoring preferred products. Similarly, the portion of the price that reflects the cost of the tax cancels out of the equation. A panel can then objectively determine if the government has levied a higher tax on a product and if the competitive opportunities of the products differ.

In sum, the aim-and-effect test is not fully operable unless the tax/price ratio is used to determine the tax differential. This methodology complements the inquiries into the aim and the effect of a measure because it also assesses the competitive opportunities between products. The sensitivity of the tax/price ratio to the relationship between products is important because a panel uses the tax differential as evidence of protection for Article III:2. If the tax differential does not reflect the competitive relationship of the products, the panel might invalidate a measure that falls within the fiscal sovereignty of the defendant government.


[13]Malt Beverages, para. 5.25.

[14]Panel Report, United States-Auto Taxes, DS31/R, dated 11 Oct. 1994 (unadopted)(92-0658) noted in Second Submission of the United States, WTO Panel, Japan-Taxes on Alcoholic Beverages, 26 Feb. 1996, para. 41 [hereinafter referred to as Second Submission]..

[15] The Panel tried to compensate for the arbitrariness of its test by advocating a flexible line between "like" and "directly competitive" products. Panel Report, para. 6.22. Flexibility emerges during the case-by-case analysis of product distinctions. Such flexibility, however, does not answer the original objection that a panel's inquiry is arbitrary unless it examines the aim of the measure. It only acknowledges that the Panel's delineation between the definitions is problematic for those products that could fall in either category. Moreover, flexibility itself is an arbitrary concept because there is no indication of how much flexibility is permissible. Finally, flexibility could engulf the definitions so that there is no difference between the categories.

[16] Although the Panel claimed to rely only the descriptive criteria, it stated, "Nowhere in the contested legislation was it mentioned that its purpose was to maintain a roughly constant tax/price ratio" (emphasis added). Panel Report, para. 6.25(iii). The Panel's legal reasoning illustrates that the descriptive criteria only gain significance when examined in conjunction with the aim of the measure. It is clear that the Panel could not avoid an analysis of the aim of the measure.

[17] The relevant portion of Article III:4 states:

The products of the territory of any contracting party imported into the territory of any other contracting party shall be accorded treatment no less favorable 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.

[18] See First Submission of the United States, WTO Panel, Japan-Taxes on Alcoholic Beverages, 1 Dec. 1995, paras. 25-26, 30-31 [hereinafter referred to as First Submission]; Appellate Submission, paras. 107-108.

[19] Admittedly, wine and beer could be considered "directly competitive" for the purposes of Article III:2 and the tax measure would be an Article III violation. However, the ECJ, with similar facts and a similarly mechanical test, has held that only cheap wine and beer are "directly competitive." There would still be problems with finding all wines and beer to be "directly competitive."

[20] To summarize, the Article XX General Exceptions to GATT 1994 include policies designed to protect public morals and human, animal, and plant life or health; policies relating to the importation and exportation of gold or silver; policies necessary to comply with laws or regulations not inconsistent with the GATT, such as customs, anti-trust, and intellectual property laws; policies relating to the products of prison labor; policies to protect national treasures; policies relating to conservation of exhaustible natural resources, as long as they also apply to domestic goods; policies taken in pursuance to intergovernmental commodities agreements that meet the criteria of the contracting parties; policies restricting the exports of materials, implemented as part of government stabilization program; and policies essential to the acquisition or distribution of products in short supply. The pursuance of these policies, however, cannot be a disguised restriction to trade or an arbitrary or unjustifiable discrimination between countries where the same conditions prevail.

[21] Panel Report, United States-Standards for Reformulated and Conventional Gasoline, WT/DS2/AB/R, at 17, cited in Appellate Submission, para. 65.

[22] Vienna Convention on the Law of Treaties, May 23,1969, art. 31(1), 155 U.N.T.S. 331, 8 I.L.M. 679.

[23] The Appellate Body argued that Article III:1 applies to the first and second sentences of Article III:2 differently because Article III:2(1) does not refer to Article III:1. Appellate Report, sec. H.1. However, the context, object and purpose of the Article do not change from one sentence to another. The fact that only the second sentence of Article III:2 refers to Article III:1 does not mean Article III:1 informs the sentences differently. Most of the other paragraphs in Article III do not mention Article III:1 and, yet, Article III:1 still informs their interpretations.

[24] Appellate Submission, para. 43.

[25] Shochu had its own category that was very detailed, including requirements about the types of still and sugar. Japanese Liquor Tax Law, Article 3.

[26] Japan argued that socially established boundaries were one of three criteria for its tax categories. Panel Report, para. 4.120.

[27] Even this Panel found this to be the case. See supra note 16.

[28] Indeed, the Panel examined the purpose of the Japanese Liquor Tax Law. See supra note 16. In addition, the Appellate Body outlined the very type of analysis required for the aim-and-effect test when it stated, "[a] protective application can most often be discerned from the design, the architecture, and the revealing structure of the measures." Appellate Report, sec. H.2.c.

[29] The Panel's test treats the tax differential as a second step in its inquiry after it has determined whether products are "like" or "directly competitive."

[30]The Panel's test, by contrast, presumes a measure has a protective effect once a tax differential is shown between "like" products. This presumption is inappropriate because is assumes products can be found "like" without considering the effects of the taxation. Moreover, the existence of a tax differential between products becomes dispositive rather than an actual or potential change in the competitive relationship.

[31] The Panel's test, by contrast, permits a de minimus difference in the tax rates of "directly competitive" products. This standard obstructs the more important question whether any change in the competitive opportunities of the products has occurred. A de minimus standard assumes that no protection of domestic products occurs before a certain tax differential but protection does occur after that differential. As the level of competition is determined by more factors than the tax differential, its is arbitrary to single out the rate of taxation as the main indicator of protective effect. In addition, once a panel accepts that a government can tax the dissimilarities between products dissimilarly, the dissimilarities may be significant enough to warrant a larger than de minimus difference in taxes. A de minimus standard is also arbitrary because it offers no guidance as to the amount of change permitted in the competitive relationship before a violation of Article III:2 occurs.

[32] Even if the descriptive definitions of "like" or "directly competitive" were expanded to include price or focus more on end-uses, price and end-uses reflect consumer preferences which are often the result of historical biases. In addition, this change would neither address the arbitrariness of drawing distinctions between products nor the hidden or implicit motive analyses that result when panels do not consider the purpose of the measure.

[33] The prevailing methodologies in GATT and ECJ alcohol tax cases compare the taxes levied on products by the tax/alcohol content or tax/volume ratios. E.g., Panel Report, paras. 6.24, 6.33; Case 170/78, Commission v. United Kingdom of Great Britain and Northern Ireland (Wine and Beer Case), 1983 E.C.R. 2265.

The test based on alcohol content is generally considered the most objective possibility although it is not as objective as it appears. Within each category of alcohol, there are varying strengths of alcohol and an average must be chosen. The average strength the defendant government or the complainants propose is manipulable and arbitrary. For instance, a government could compare mid-strength imported beers with mid-strength domestic wines. However, these beers may be taxed less than other beers or they may not be competitive with wines of mid-strength. The average strength beer may also not be the most consumed variety of beer. In addition, the differences in strengths of beer may vary more than differences in other factors like price. As alcohol content has less impact on consumer behavior than other factors, it is arbitrary to use the tax/alcohol content ratio as evidence of a change in the competitive relationship in the aim-and-effect test. Alcohol content as a yardstick also loses objectivity when a progressive system of taxation is introduced. Suppose a government chooses to tax stronger alcoholic beverages more than weaker alcoholic beverages. A ratio of 1/3/5/10 could become 1/9/25/100 but the tax/alcohol content methodology would not be concerned with the size of the differentials between the categories which raises questions about its objectivity.

A tax differential determined by volume is thought to be the most consistent methodology because governments frequently establish their tax classifications by volume. However, the logistical ease of placing a tax on volume does not indicate that the tax/volume ratio is a good measure of the rate of taxation. Similarly, the government's choice to assess the tax by volume does not mean that a panel cannot evaluate the effects of a tax by a different ratio. The volume of the container is even less likely to affect consumer choice than alcohol content. Again, a tax differential measuring a characteristic of little importance to the consumer is not the most objective or accurate method to show a change in consumer behavior for the aim-and-effect test. Governments and panels would also need to incorporate a corrective factor when comparing the ratios. Otherwise, it might be difficult, for instance, to compare alcohol that is usually diluted to thirst-quenching alcohol that is usually drunk in large quantities. A methodology based on volume turns out to be less consistent than originally thought.

[34] The Panel here examined the tax differential by both tax/volume and tax/alcohol content. It considered the comparison by alcohol content to be most significant because Japan had chosen that method to apply the tax. Panel Report, para. 6.33.

[35] The Japanese government had no mechanism for maintaining a consistent tax/price ratio. The Panel interpreted this omission as one sign that the government was engaging in ex-post facto reasoning. Panel Report, para. 6.34(iii).


Previous |Next |Title |Contents

 

 


Questions or comments about this site?
Email Enfellows@exchange.law.nyu.edu

Top of the page