1. Introduction
Proposing a three-step approach to the National Treatment Principle
Applying the WTO National Treatment Principle raises the difficult question of how to define and establish protectionism. The National Treatment Principle provides for equality of treatment between ‘like’ imported and domestic products and services. But what does equality of treatment mean? If a regulation has a negative competitive impact on imported products, is this enough to establish a breach? Or is it necessary to consider additionally whether the regulation has a non-protectionist purpose, which accounts for this negative impact?
This is one of the most debated issues in WTO case law. National treatment is a foundational principle of WTO law and the focus of much judicial activity. Because it ‘polices’ domestic regulation applied inside the border, it is of key importance to striking the balance between enabling governments to regulate and bringing into being trade liberalization commitments, an issue which is sensitive and controversial. Furthermore it covers not only de jure discrimination (from laws that explicitly distinguish between domestic and imported products), but also de facto discrimination (discrimination ‘in fact’ from origin-neutral laws). In both cases, and particularly the latter, the discrimination analysis normally goes far beyond the letter of the law. In other words, a regulation's text is not likely to reveal whether it has been written and applied in good faith vis-à-vis the National Treatment Principle. Adjudication requires the Appellate Body to analyse and interpret various contextual factors.
As early GATT Panels recognized, when establishing whether a disputed regulation is protectionist, there are two key elements: its intent, or policy rationale, and its effect on the market for competitive products. Both can be challenging to ascertain. Protectionist intent may be deliberately concealed, or at least not formally declared. Establishing protectionist effect necessitates complex causal analysis. For these and other reasons, the Appellate Body has formally rejected both intent and effect as a basis for establishing the existence of protectionism. Yet the analytical tests that it has adopted reflect the same inescapable factors: (policy-based) intent and (market-based) effect. This obfuscation at the core of the analysis creates an incentive to utilize discretionary ‘smell test’ (Hudec, Reference Hudec1998) approaches that conflate market-based and policy-based reasoning. This can take the form of findings about the existence and nature of competitive relationships that rest upon what appear to be quantitative market analysis, even after the Appellate Body has rejected (or misinterpreted) evidence submitted, leading to a loss of transparency and rigour. It can also lead to a compelling policy objective influencing the application of a supposedly ‘market-based’ analysis.
Adding to the complexity, the Appellate Body has emphasized minor differences in the wording between key subparagraphs of Article III and the TBT Agreement and leads to a discrepant approach to all three, and the EC–Seal Products ruling under Article III(4) increases this textual hair-splitting (Davey and Maskus, Reference Davey and Maskus2013: 181).Footnote 1 Legal and institutional challenges arise from the current divergent jurisprudence. To respond to these problems, this article proposes more convergence between these National Treatment Principle provisions: a rigorous market-based effects analysis, followed by a policy-based intent analysis. This takes the form of a three-step approach. The first two steps are the same as those the Appellate Body has affirmed in EC–Seal Products: discerning whether the products in dispute are in a competitive relationship, and whether there is an impact on the conditions of competition to the detriment of the imported product. There should then however be a third step of considering whether the detrimental impact can be explained based upon a non-protectionist regulatory rationale. While the proposed reforms are based in a lengthy tradition of scholarship, the article suggests a novel way of separating and integrating economic and policy-based components of the analysis.
The first and second steps: strengthening the ‘disparate impact’ analysis
To a surprising degree, the Appellate Body and the Panel have rejected a reliance on quantitatively derived thresholds or benchmarks to establish the existence and nature of competitive relationships between products. At the same time, they have often utilized the concepts and language of relevant econometric approaches to support qualitative conclusions.
A clearer picture of what the market suggests about the existence and nature of competitive relationships would improve the quality of dispute outcomes. The effective implementation of this suggestion, echoed in recent scholarship (Diebold, Reference Diebold2014; Pauwelyn, Reference Pauwelyn, Huerta-Goldman, Romaneti and Stirnimann2013; Bown, Reference Bown and Pauwelyn2010), would require reforms from the Parties' initial submissions to the Panel's findings of fact to the Appellate Body's final decision. A stronger affirmation from the Appellate Body that economic evidence informs dispute outcomes would help motivate these changes.
In practice, the rejection of market evidence may also contribute to an interpretive bias toward finding a disparate impact. This is implicit in its justifications: first, the regulation in dispute distorts competition, and econometric approaches may not capture this. Second, relying on assessments of trade flows will not capture discriminatory regulation whose impacts on imports are not clearly evidenced.Footnote 2 These problems are non-trivial, but the cure of conflating quantitative and qualitative reasoning does not fit the disease, but makes it worse by relying on a speculative approach that purports to be ‘objective’. As stated by Van Aaken in this context: ‘Being shortsighted is usually better than being blind’ (Van Aaken, Reference Van Aaken2011: 43).
The third step: considering the policy rationale
In spring 2014, the Appellate Body in EC–Seal Products affirmed that under the GATT National Treatment Principle, Article III(4), a negative competitive impact on imported products equates to a violation. This has been usefully termed a ‘disparate impact’ approach (Regan, Reference Regan2004b: 756). Disparate impact should not be an automatic grounds for a violation, as this gives Article III(4) too much influence over national regulation. As has been argued by many commentators over a number of years, it is essential to consider the reason (Hudec, Reference Hudec1998; Regan, Reference Regan2004a and Reference Reganb; Reid, Reference Reid2010). When competitive impacts are decisive, unrelated market circumstances can be tarred with the brush of protectionism.
The Appellate Body may still consider the reason for the regulation, and whether it can justify the disparate impact, under the General Exception to the GATT, Article XX. However, the list of designated exceptions under Article XX does not exhaust all of the possibilities that justify intervention. Furthermore, considering the purpose of a disputed regulation should comprise an integral component of the non-discrimination analysis. As Article XX is an exception, it cannot fulfill this function.
EC–Seal Products also threatens the proper functioning of WTO provisions. It imbalances the GATT Article III(4) and the analogous National Treatment Principle in the Agreement on Technical Barriers to Trade (‘TBT Agreement’). TBT Article 2.1 begins by establishing a disparate impact but additionally considers if it is based on a legitimate regulatory distinction (Flett, Reference Flett2013; Mavroidis, Reference Mavroidis2013). There is some overlap in the provisions, and the discrepancy is leading countries to avoid complaints under TBT Article 2.1, because it is perceived as an easier test to pass.Footnote 3 The outcome is to marginalize the TBT Agreement, a more recently negotiated provision whose legal test better reflects the purpose of the National Treatment Principle.
This article thus takes stock of developments under Article III(4) post-EC–Seal Products. It proposes an approach which addresses some of the latter's negative implications, and which fits within the constraints of the provisions as drafted. Under the proposed approach, both market-based and policy-based components of the Article III analysis are strengthened and made distinct.
2. The National Treatment Principle in GATT Article III and the TBT Agreement
The key provisions of GATT Article III are Article III(2) and III(4), which cover trade-restrictive taxation and regulation, respectively. This article focuses primarily on the former, while taking into account the Article in its totality and the interpretive differences between these two subparagraphs, which support a reformed interpretation of Article III(4). Both subparagraphs focus on internal measures and whether they favour domestic products, and both are interpreted with reference to the same chapeau. Article III(1), the chapeau, states that measures should not be applied to imported or domestic products ‘so as to afford protection to domestic production’.Footnote 4
The first sentence of Article III(2), on taxation, stipulates that a government must not tax ‘like’ imported products ‘in excess of’ domestic products. The second sentence stipulates that if the products are not ‘like’, but directly competitive, differential taxation is also in violation if it results in the protection of domestic products.Footnote 5
Under Article III(4), on regulation, the key questions are whether the products in dispute are ‘like’, and if so whether the imported product is treated no less favourably than the domestic product.Footnote 6
The TBT Agreement deals with technical regulations, standards, and conformity assessment. Due to the overlap in their jurisdictions, there may be initial uncertainty with respect to which WTO provision, Article III(4) or the TBT Agreement, is most relevant to a measure in dispute. Because the definition of ‘technical regulation’ is selective, it had not been the focus of many disputes until the quick succession of three TBT Agreement disputes in 2011‒2012. TBT Article 2.1 contains a national treatment provision, whose wording is similar to that of Article III(4).Footnote 7 Yet its interpretation has differed. The key questions are the same as under Article III(4): first, establishing whether domestic and imported products are ‘like’, and, second, if the imported product is receiving treatment ‘no less favourable’ than that accorded to the domestic product. However, when establishing the presence of ‘less favourable treatment’, the Appellate Body has taken into account the purpose of the regulation in dispute by considering whether it is based upon a ‘legitimate regulatory distinction’. These differences and their implications are further discussed below.
3. Assessing competitive relationships under the National Treatment Principle
Davey and Maskus linked the narrower approach that the Appellate Body has taken to Article III (in contrast with TBT Article 2.1) with the need for sounder economic analysis of the impact of the regulation in dispute on the market. After the 2011–2012 TBT disputes, they wrote: ‘If the Appellate Body is unwilling to import the “so as to afford protection” test in to the less favourable treatment analysis under Article III:4, perhaps it will … require a more rigorous showing of a modification of conditions of competition in de facto discrimination cases' (Davey and Maskus, Reference Davey and Maskus2013: 181). In other words, if it is going to base findings of discrimination on conditions of competition, the Appellate Body should be sure that at least it has understood these conditions correctly.
Greater evidentiary rigour can complement, rather than replace, an approach to Article III that incorporates both ‘effect’ and ‘intent’ (the latter discussed in the second part of this article). There are a number of examples in case law under Article III and TBT Article 2.1 where more use of market analysis would have led to sounder dispute outcomes. A body of literature from those better qualified than this author, cited throughout, considers how to make better use of economic methodologies. The following critique is far from comprehensive, but highlights the disconnection between the market evidence submitted and the legal conclusions it supports. Like the ambiguous treatment of regulatory ‘intent’, it is symptomatic of ambivalence toward fundamental elements that inform the analysis of whether a measure is protectionist. It leads to lack of transparency and the undesirable conflation of quantitative and qualitative analysis. As the Panel is responsible for findings of fact, including evaluating such evidence, much of the following focuses on Panel decisions.
Like and directly competitive or substitutable products
Under the National Treatment Principle, ‘like’ products are competitive products. More specifically, in EC–Asbestos, the Appellate Body affirmed that ‘a determination of “likeness” under Article III:4 is, fundamentally, a determination about the nature and extent of a competitive relationship between and among products', andFootnote 8 the TBT disputes have adopted this interpretation;Footnote 9 EC–Seal Products has maintained the status quo in this respect.Footnote 10 Article III(2) contains the additional category of ‘directly competitive or substitutable’ products; the importance of competition to this category is self-evident.Footnote 11
This might suggest that the dispute settlement bodies currently adopt a quantitative approach, evaluating the market evidence to determine whether products are like (i.e. competitive). There are standard econometric approaches to quantifying competitive relationships between goods. These include evaluating the cross-price elasticity, which aids in establishing whether a change in the price of one product affects another product; in other words, do consumers consider the products as substitutes and purchase product B instead of product A if the price of product A rises.
However, the Appellate Body has repeatedly affirmed in both GATT and TBT disputes that decisions should not rest upon such evidence. Instead, when establishing whether products are ‘like’, it relies primarily on qualitative criteria drawn from the 1970 GATT Border Tax Adjustment Working Party report (‘BTA criteria’). These include: ‘the product's end-uses in a given market; consumers' tastes and habits, which change from country to country; the product's properties, nature and quality'.Footnote 12 The BTA criteria are meant to define what ‘like’ means, and product competitiveness is an essential component of ‘likeness’. The Appellate Body has stated that BTA criteria constitute a non-exhaustive list of criteria for assessing evidence relating to whether the products in dispute are in a competitive relationship.Footnote 13 In this enquiry, the degree to which cross-price elasticity studies feature varies greatly.
Following from Japan–Alcohol, ‘like’ products are also often described as an accordion which stretches or narrows depending on the circumstances of the dispute.Footnote 14 The statements above both enshrine product competitiveness as decisive and then emphasize the need for flexibility and qualitative analysis.
As stated by Neven and Trachtman:
the set of factors examined in traditional Article III product comparisons seems less sophisticated, less precise, and less plausible than a modern economic analysis of the extent of competition between two products. (Neven and Trachtman, Reference Neven and Trachtman2013: 302)
The alcoholic beverage disputes: the rejection of market evidence
The Appellate Body set out its ambivalent stance toward economic evidence, also evident under Article III(4) and TBT Article 2.1, in several Article III(2) disputes on taxation of alcoholic beverages. In the early WTO dispute Japan–Alcoholic Beverages II, cross-price elasticity analyses were central to the Party's submissions about whether shochu is directly competitive and substitutable with imported liquors. Japan argued that consumers would not necessarily switch to shochu in the absence of spirits.Footnote 15 The Panel dismissed Japan's studies on two grounds. First, it was concerned that these models might not take into account the impact of the measure in dispute, namely higher tax levels on whiskey, on the competitiveness of the products. Second, it criticized Japan's study because it suffered from the problems of autocorrelation and multicollinearity.Footnote 16 The Panel affirmed that, despite these limitations, Japan's studies did indeed show some competitive relationship. Because 10% of consumers would switch to spirits if shochu was not available, this was sufficient to prove that the spirits were ‘directly competitive or substitutable’.Footnote 17
While the Panel critiqued Japan's methodology, its determination seems random. The uncertainty that the Panel rejected is smaller than the uncertainty that it embraced in coming to its own conclusion. Also, by utilizing the language and concepts of cross-price elasticity, the Panel gave the impression that its conclusion was based upon the rejected studies. In accepting the Panel's analysis, the Appellate Body furthered this misperception:
the decisive criterion in order to determine whether two products are directly competitive or substitutable is whether they have common end-uses, inter alia, as shown by elasticity of substitution [emphasis added].Footnote 18
Likely in response to this problem, in the later Korea–Alcoholic Beverages, the Appellate Body de-emphasized cross-price elasticity:
‘[Q]uantitative analyses, while helpful, should not be considered necessary’ [footnote eliminated]. Similarly, ‘quantitative studies of cross-price elasticity are relevant, but not exclusive or even decisive in nature’.Footnote 19
It further clarified that: ‘We believe that the Panel uses the term “nature of competition” as a synonym for quality of competition, as opposed to quantity of competition.’Footnote 20 It rejected quantity for quality. In Philippines–Distilled Spirits, the Appellate Body reiterated this conclusion,Footnote 21 and the Panel exhibited a similar response to submitted economic studies, heavily critiquing the methodology employed on grounds that were somewhat spurious (Neven and Trachtman, Reference Neven and Trachtman2013: 319–321).
In both disputes, the conceptual basis for rejecting a quantitative approach was that the regulation in dispute distorted competition. In other words, the tax made imported products more expensive, and as a result consumers were less likely to see them as substitutes for domestic products. Thus, competition should have existed, though it could not be evidenced.Footnote 22
This is a serious problem, though it can be accommodated to some extent in cross-price elasticity studies by making sure that the price differentials that the tax imposes are accounted for in surveys or by looking to comparable markets (Neven and Trachtman, Reference Neven and Trachtman2013: 314). However, the Appellate Body's response has been to undertake qualitative analysis of ‘potential’ competition. The Appellate Body in Korea–Alcoholic Beverages stated:
the scope of the term ‘directly competitive or substitutable’ cannot be limited to situations where consumers already regard products as alternatives. If reliance could be placed only on current instances of substitution, the object and purpose of Article III:2 could be defeated by the protective taxation that the provision aims to prohibit … In this case, the Panel committed no error of law in buttressing its finding of ‘present direct competition’ by referring to a ‘strong potentially direct competitive relationship’.Footnote 23
The Panel considered the possibility that lack of experience would artificially dampen consumer demand in Korea. Thus, the Panel suggested that there was potential for future competition.Footnote 24 As Korean consumers' familiarity increases, their perception may change so that they eventually consider products to be ‘like’. This rests upon a problematic assumption that if consumers do not know about a product because of a measure in dispute, this in itself could suggest that the country is not complying with its GATT obligations. Further, the treaty text focuses on products that are ‘like’ or ‘directly competitive or substitutable’.
The Panel also stated that evidence about consumer preference from a market with similar characteristics may have some relevance.Footnote 25 The Appellate Body agreed, but the dispute settlement bodies offered no criteria with which to establish if markets were sufficiently similar. It may be appropriate to consider such evidence, but what amounts to complex modeling should not be done on a conceptual basis. This reasoning is necessarily reductionist in the variables it considers.
Importantly, in these disputes, the concept of potential competition introduced a bias toward finding products competitive. Further to this, the Appellate Body also asserted that only a small degree of substitutability in one segment of the market can establish competitiveness. Philippines–Distilled Spirits focused on higher taxation rates for largely imported distilled alcoholic beverages as compared to, in particular, domestically produced alcoholic beverages derived from sugarcane. The Philippines argued that the majority of Filipino consumers could not purchase distilled alcohols because of their high prices. Therefore, higher taxes on non-sugarcane derived alcohol were not protectionist.
The Appellate Body concluded that there was a small amount of competition between imported and domestic distilled spirits among the (luxury) market that had access to both. According to the Appellate Body:
Article III of GATT 1994 does not protect just some instances or most instances, but rather, it protects all instances of direct competition. It follows that the competitive relationship does not need to occur throughout the whole market for a panel to find that a measure is inconsistent with the second sentence of Article III. We thus conclude that, even if the Philippine distilled spirits market were segmented, actual direct competition exists within at least a segment of that market.Footnote 26
Thus the Appellate Body asserted as a matter of principle that Article III should be interpreted to mean that there only needed to be competition in a small market segment for products to be in competition. It then referenced the submitted cross-price elasticity studies to support the finding that there was a significant degree of competitiveness or substitutability between the products, a dubious conclusion with respect to this evidence (Neven and Trachtman, Reference Neven and Trachtman2013: 318).Footnote 27
Less Favourable Treatment (Article III(4) and TBT Article 2.1) and ‘so as to afford protection’ (Article III(2))
The Appellate Body has stated repeatedly that it will not rely upon the effect of a measure, in terms of trade volumes of domestic versus imported products, to establish discrimination. This is because there may be discriminatory treatment that is not reflected by trade volumes, but exists in the application of the regulation. According to the Appellate Body in Korea–Alcoholic Beverages: ‘the Panel stated that if a particular degree of competition had to be shown in quantitative terms, that would be similar to requiring proof that a tax measure has a particular impact on trade … We do not consider the Panel's reasoning on this point to be flawed.’Footnote 28 This is because Article III outlines a code of conduct for trade partners. Regardless of current import levels, a measure should not limit ‘competitive opportunities’.Footnote 29 The goal of protecting competitive opportunities has become a de facto precedent, as affirmed by the Appellate Body in EC–Seal Products.Footnote 30
There is clearly a tension between, on the one hand, emphasizing detrimental impact as decisive, and, on the other, focusing on competitive opportunities. Like ‘potential competition’, ‘competitive opportunities’ are difficult or impossible to measure. Following from this, in a number of disputes, the reasoning seems less than robust. For example, in US–COOL, a recent TBT dispute, the Appellate Body agreed with the Panel's analysis that the measure provided a negative impact on conditions of competition for imported cattle and stated that the low market share of imported beef was one component that suggested this detrimental impact.Footnote 31 This dispute was unusual for the degree to which it took account of trade flows in establishing less favourable treatment (Pauwelyn, Reference Pauwelyn, Huerta-Goldman, Romaneti and Stirnimann2013: 10). The Panel noted that the measure led to segregation between domestic and imported animals; while this in itself did not constitute discrimination, segregation led to higher costs.Footnote 32 It also concluded that the measure created incentives for participants to process domestic rather than imported livestock because it would be cheaper to do so.Footnote 33 Thus, the measure created a lack of competitive opportunities for imported livestock, and constituted less favourable treatment.Footnote 34
The Panel's finding centred on the fact that the technical requirements of the measure, for example, segregation and increased compliance costs for imported cattle, created detrimental conditions of competition. However, this finding left unexamined whether, in practice, the measure created a situation where domestic beef was being favoured over imported or mixed-origin beef.Footnote 35 It is possible, for example, that even with the regulatory burden of the labeling requirement, imported beef was still cheaper.
This is important because a preference for domestic beef was one of the assumptions upon which the outcome rested.Footnote 36 As Mavroidis stated: ‘How can the AB know which beef is being favoured when it has not conducted market analysis? It seems that the AB came up with one theoretically probable (even plausible) but unproven scenario (segregation will push traders to US beef) and then built its finding on this score around it’ (Mavroidis, Reference Mavroidis2013: 521).
US–Tuna II: a challenging scenario
US–Tuna II is a 2011 TBT dispute which raises complex questions regarding the use of market analysis, but ultimately highlights the same problems documented above; namely the need for a more rigorous use of evidence. It concerned a US law that tuna could not be labeled dolphin safe if it were fished by encircling dolphins with purse seine nets to catch the tuna that congregated underneath, a requirement that applied specifically within the Eastern Tropical Pacific (‘ETP’).Footnote 37 The Panel focused its analysis of whether tuna that did and did not qualify for the label were ‘like’ on the BTA criteria. Based on physical characteristics, end-uses, and customs classification, the products were ‘like’ – indeed, identical. However, the Panel would not pass judgment on Mexican fishing practices by implying they were unsafe to dolphins. Therefore, the Panel dismissed consumer preference for one type of tuna over another as irrelevant.
Adopting a more market-based approach, the first question is which tuna products to compare. This may seem straightforward: the dispute concerned Mexican versus US tuna, so the products could be compared on the basis of nationality to confirm what would be expected to be a very high degree of substitutability. However, most Mexican tuna were caught by setting on dolphins, so would not have received the US ‘dolphin safe’ label, and this label (the measure in dispute) may have influenced the competitiveness of the products.
In this context, the challenge is to try and ascertain consumer views of the product independent of the regulation. In other words, did consumers specifically prefer tuna that were not caught by setting on dolphins, to the extent that the products would not be in competition even if the label were removed or its conditions modified so that Mexican tuna could achieve market access? Indeed, there was evidence that consumer's preferences on this issue were so strong as to lead industrial consumers of tuna products to deem tuna caught by setting on dolphins as non-competitive. The Panel stated:
We further note that it is undisputed that US consumers are sensitive to the dolphin-safe issue. This is acknowledged by both Mexico and the United States, and is also confirmed by the evidence presented with the amicus curiae brief to which the United States has referred to in its answers to questions … The evidence presented to the Panel also shows that major tuna processors reacted to these dolphin-safe concerns, and that this led to changes in their purchasing policies as of April 1990. These policies are still in place: such companies will not purchase tuna from vessels that fish in association with dolphins.Footnote 38
The Panel here extrapolated that the dolphin-safe label had a commercial value, and that producers actually refused to buy uncertified tuna for fear of consumer boycotts. It is difficult to reconcile these statements, made in the context of the ‘less favourable treatment’ analysis, with its conclusion that consumer preference was irrelevant to the competitiveness of these products.
The Panel then went on to examine whether Mexican tuna was subject to less favourable treatment. It established that only 1% of the imports in the responding country, the US, came from Mexico, the complaining country. On first glance, this low figure suggests that the measure is having a negative impact on Mexican imports. However, the Appellate Body stated:
Moreover, it is well established that WTO rules protect competitive opportunities, not trade flows. [footnote omitted] It follows that, even if Mexican tuna products might not achieve a wide penetration of the US market in the absence of the measure at issue due to consumer objections to the method of setting on dolphins, this does not change the fact that it is the measure at issue … that denies most Mexican tuna products access to a ‘dolphin-safe’ label in the US market.Footnote 39
The Appellate Body suggested that consumer preferences might be the reason there were not many imports from Mexico, and there would not be many even if the measure were removed. In other words, it was very possible that the products were not competitive (or ‘like’), and that the measure had no negative competitive impact on Mexican tuna. However, the measure was in violation because it did not preserve ‘competitive opportunities’. It is hard to say exactly what this means; the possibility that future consumers might care less about setting on dolphins is one troubling interpretation.
This regulation poses deeper challenges than the taxation of alcoholic beverages, due to the difficulty of identifying the influence of the regulation in dispute and the fact that the submitted evidence did not fit a traditional cross-price elasticity model (indeed some came from an amicus curiae brief). Nonetheless, it underlines the degree of discretion afforded by the BTA criteria and ‘competitive opportunities’ test. It seems feasible to conduct cross-price elasticity analysis of the extent to which end-of-the-line consumers would consent to ‘setting on dolphins’ given particular price differentials; such evidence would certainly aid in coming to more rigorous conclusions regarding the competitive relationships at stake. A legal culture of relying more on market analysis is an essential component of improving the quality and quantity of evidence that dispute Parties submit.
Practical considerations for a more quantitative approach
These examples demonstrate how the continued insistence that the National Treatment Principle is founded in an assessment of competitive relationships gives a false impression. The core concepts of ‘potential competition’ and ‘competitive opportunities’ are not reflected by the treaty text. In practice, assessing these concepts involves utilizing legal reasoning to undertake complex market analysis, and in some cases this seems to serve the aim of establishing a violation. Decoupling quantitative and qualitative reasoning, and tying dispute outcomes to a more rigorous use of the former, would lead to simpler, more accurate, and more transparent dispute outcomes.
Given the strength of the above condemnation, it is important to acknowledge that quantitative analysis of competitive relationships has its limitations, both capacity-related and methodological.
As summarized by Howse and Levy:
Consumer-preference estimates are not always precise. These characterizations are not universal constants, waiting to be discovered … They emerge from particular samples in a data set and are based on particular modeling of preferences. Even then, they emerge with error bands. (Howse and Levy, Reference Howse and Levy2013: 341)
The assessment of whether consumers find particular products to be in a competitive relationship is based upon a set of choices made by modelers. There is inevitably room for error (and bias) not only in the sampling process itself but also in the choice of which population to sample. There are also limitations related to data capture. Self-selecting participants in surveys on purchasing preferences may not be representative of the population as a whole. They may report their preferences inaccurately; for example, they may underestimate the quantity of alcohol they consume due to societal prohibitions (Sousa, Reference Sousa2014: 12). In sum, as the WTO itself stated in the World Trade Report of 2005, ‘Elasticity values are not normally known with precision’ (World Trade Organization, 2005b: 177).
Further, it seems doubtful that any country would provide evidence that did not serve its own aims, and the respondent will likely generate the majority of the economic analysis. Complainants will not have access to the same information about the domestic situation.
However, as stated by Howse and Levy, ‘the perfect ought not to be the enemy of the good’ (Howse and Levy, Reference Howse and Levy2013: 341). The recommendations made here are based in a slightly different universe, in which Panels are always supplied with sufficient and high-quality evidence from the Parties and WTO Panels have sufficient economic expertise to evaluate this evidence in a rigorous manner. The current jurisprudence effectively advises against relying too much on sophisticated econometric models in this context. If WTO Panels took quantitative evidence more seriously, this would introduce a motive for Member States to do the same. With respect to the limitations in the capability of WTO Panels to interpret the evidence, and the lack of transparency in their approach, persuasive arguments have been made about the need to incorporate more economic expertise into the WTO dispute settlement process through appointing Panelists with economic expertise and providing more specialists in economics to adjudicators (Bown, Reference Bown and Pauwelyn2010; Pauwelyn, Reference Pauwelyn, Huerta-Goldman, Romaneti and Stirnimann2013).
4. The role of regulatory intent in the National Treatment Principle: evolution of the controversy
After more rigorously establishing that disputed products are in competition, and that there is a negative competitive impact on imported products, the Appellate Body should then turn to the question of whether this impact can be explained with respect to a non-protectionist regulatory purpose.
The issue has been actively contested since the pre-1995 GATT era. The case law is evolutionary and at times discrepant. The incorporation of an analysis of regulatory rationale under TBT Article 2.1 suggested that the National Treatment Principle in WTO law as a whole might evolve in this direction. With its clear affirmation in EC–Seal Products, the Appellate Body has apparently brought to a close nearly twenty-five years of debate on this issue, at least with respect to GATT Article III(4).
The early debate: aim and effect and the ‘like’ products test
Pre-WTO GATT Panels recognized ‘intent’ and ‘effect’ as key components of establishing protectionism. In a 1992 GATT dispute under III(2), US–Malt Beverages,Footnote 40 the Panel decided that the determination of whether the products in dispute were ‘like’ should also regard the broader purpose of the Article articulated in the chapeau: a measure should not be applied so as to afford protection to domestic products. This argument was applied and extended in the unadopted GATT Panel report in the dispute US–Taxes on Automobiles.Footnote 41 The Panel concluded that, though a US luxury excise tax on automobiles had a de facto effect of discriminating against imports, the tax did not have a discriminatory intent.Footnote 42 In this context, the Panel should examine not only the effect of the measure, but its aim, a synonym for ‘intent’.
In the early WTO dispute Japan–Alcoholic Beverages II, the Appellate Body rejected this ‘aim-and-effect’ approach. First, the approach had a clear methodological failing of integrating the analysis of whether a measure was protectionist into the ‘like’ products test. Article III(2) makes reference to the chapeau and its concept of ‘protectionism’ in its second sentence. Second, government intent was too subjective.Footnote 43 In Chile–Alcoholic Beverages, the Appellate Body reaffirmed that ‘The subjective intentions inhabiting the minds of individual legislators or regulators do not bear upon the inquiry, if only because they are not accessible to treaty interpreters.’Footnote 44
Since Japan–Alcoholic Beverages II, the Appellate Body has consistently affirmed that ‘likeness’ is a market-based, not a policy-based, concept. In EC–Asbestos, it stated that ‘a determination of “likeness” under Article III:4 is, fundamentally, a determination about the nature and extent of a competitive relationship between and among products’.Footnote 45 However, in this very dispute, some analyses argued that the compelling policy justification for differentiating a carcinogenic and non-carcinogenic product weighed more in the decision than whether the products in dispute were in a competitive relationship (Horn and Weiler, Reference Horn, Weiler and Horn2003: 31; Lydgate, Reference Lydgate2011: 178–180).
‘Less favourable treatment’
The rejection of aim and effect formed the primary focus of Hudec's noted article lamenting, the rejection of ‘regulatory purpose’ under the National Treatment Principle analysis (Hudec, Reference Hudec1998). Yet perhaps now, in the wake of EC–Seal Product, is the time for a more decisive requiem in the Article III(4) context. The ‘aim and effect’ approach left open the question of whether an examination of regulatory purpose could be integrated into the rest of the discrimination analysis, but it has been de-emphasized.
After rejecting ‘aim and effect’ in Japan–Alcoholic Beverages II, the Appellate Body affirmed that it is necessary under the second sentence of Article III(2) to examine whether a measure is applied ‘so as to afford protection’ to directly competitive or substitutable domestic products. Rather than being based upon subjective ‘intent’, however, this analysis ‘objectively’ examines the structure and application of a measure.Footnote 46
The rejection of ‘aim’ in favour of ‘objective’ structure likely shifted the emphasis away from certain types of ‘subjective’ evidence, such as reviewing preparatory materials of the legislation, and suggests that any statements regarding the purpose of the regulation, even within the regulation, are not a valid source of evidence. While the position remains formally consistent, in some Article III(2) disputes,Footnote 47 the Appellate Body has made reference to just such evidence. Even in Chile–Alcoholic Beverages, Footnote 48 after rejecting ‘subjective’ intent, the Appellate Body affirmed that a measure's purposes are relevant to whether it is applied so as to afford protection to domestic products.Footnote 49 This is also symptomatic of how the disavowal of ‘subjective’ intent has led to a kind of double consciousness, perpetuating the Appellate Body's discomfort with this component of the analysis and yet maintaining its unavoidable centrality.
If this situation could be described as ambivalent, the Appellate Body's treatment of regulatory purpose under Article III(4) is inexplicably dismissive. In the vast majority of disputes, it has affirmed that a negative impact on conditions of competition to the detriment of imported products is indicative of a violation. There has been some inconsistency. In EC–Asbestos, the Appellate Body made reference to the chapeau's emphasis on preventing protectionism and developed a two-step analysis: first, establishing whether the ‘like’ products were treated differently; second, determining whether that differential treatment constituted less favourable treatment.Footnote 50
In Dominican Republic–Cigarettes, the Appellate Body further clarified:
However, the existence of a detrimental effect on a given imported product resulting from a measure does not necessarily imply that this measure accords less favourable treatment to imports if the detrimental effect is explained by factors or circumstances unrelated to the foreign origin of the product [emphasis added].Footnote 51
Here, a measure with negative impact on conditions of competition may not violate Article III(4), if that impact is explained by something besides protectionism.
These rulings, coupled with the recent evolution of a line of jurisprudence under TBT Article 2.1 examining the existence of a ‘legitimate regulatory distinction’, led the EU, the responding country in EC–Seal Products, to argue that a similar test should be incorporated under GATT Article III(4).Footnote 52 The Appellate Body closed the door on this possibility. It stated:
a determination of whether imported products are treated less favourably than like domestic products involves an assessment of the implications of the contested measure for the equality of competitive conditions between imported and like domestic products. If the outcome of this assessment is that the measure has a detrimental impact on the conditions of competition for like imported products, then such detrimental impact will amount to treatment that is ‘less favourable’ within the meaning of Article III:4 [emphasis added].Footnote 53
The Appellate Body clarified that, in Dominican Republic–Cigarettes, it had not intended that there be an additional step of considering explicitly whether a detrimental impact could be explained by factors unrelated to the foreign origin of the product.Footnote 54 However, it stated that in that dispute, ‘the detrimental impact on competitive opportunities for like imported products was not attributable to the measure at issue’.Footnote 55 It affirmed that there had to be a ‘genuine relationship’ between the regulation in dispute ‘and its adverse impact on competitive opportunities for imported versus like domestic products’.Footnote 56 When determining this, the relevant question is whether the regulation in dispute is what affects the conditions under which the ‘like’ products compete.Footnote 57
The ‘genuine relationship’: a portal for regulatory purpose?
Previous hopes that the rulings of EC–Asbestos and Dominican Republic–Cigarettes provided an inroad for regulatory purpose under Article III(4) (Porges and Trachtman, Reference Porges, Trachtman and Trachtman2006: 84–86) may migrate to this ‘genuine relationship’ test. After all, the requirement for a genuine relationship between a measure and its effects could be interpreted as requiring a negative impact to have a causal link to a protectionist motive. Yet a close reading reveals this is not the case. The Appellate Body affirmed that the relevant question is whether the regulation in dispute is what affects the conditions under which the ‘like’ products compete. The causal analysis thus serves to establish that it was the measure in dispute that caused the negative competition impact, and not for example other market factors.Footnote 58 A more rigorous standard would be a welcome development, but cannot stand in for examining whether a detrimental impact could be explained vis-à-vis a non-protectionist purpose.
Arguments against the disparate impact approach
It makes GATT Article III a deregulatory instrument
The only possible justification for differentiating Article III(2) from III(4) is that the former refers explicitly to the chapeau's ‘so as to afford protection’. This excessively textual approach misses the forest for the trees. It implies that taxation measures should be treated less strictly than regulatory measures, when Article III(1), the chapeau, equally informs the interpretation of all of Article III (Flett, Reference Flett2013: 57).Footnote 59
This leads to the more fundamental question of whether protectionism equates with ‘disparate impact’. There is some evidence that this is a conventional interpretation under WTO law. A 1999 Report from the WTO Secretariat summarized that: ‘The essence of the principle of national treatment is to require that a WTO Member does not put the goods or services or persons of other WTO Members at a competitive disadvantage vis-à-vis its own goods or services or nationals [emphasis added]’.Footnote 60 This interpretation is oft-reinforced in the disputes, such as the Appellate Body's statement in Korea–Alcoholic Beverages that ‘the object and purpose of Article III is the maintenance of equality of conditions of competition for imported and domestic products [emphasis added]’.Footnote 61
However, the treaty text does not support this, and the discrepant approach to TBT Article 2.1 disputes cast more doubt on the interpretation. Further, in Japan–Alcoholic Beverages II, the Appellate Body declared that that ‘the broad and fundamental purpose of Article III is to avoid protectionism in the application of internal tax and regulatory measures … Toward this end, Article III obliges Members of the WTO to provide equality of competitive conditions for imported products in relation to domestic products.'Footnote 62 Equality of competitive conditions is a tool toward the end of determining protectionism, not the final goal.Footnote 63 These discrepant interpretations again capture the double consciousness regarding the interpretation of protectionism.
Also, even if WTO Members support disparate impact when they initiate national treatment disputes, it seems unlikely to appeal to them when they are on the receiving end. As a third party to the EC–Seal Products dispute, the United States summarized the problem:
It is … difficult to understand how a ‘detrimental impact’ on imports from one Member compared to another Member can by itself be sufficient to find that those imports are being treated less favorably. One would expect that any measure will affect some products differently from others. Yet that different treatment would not amount to discrimination unless one also looks at the reason why there was such a difference in treatment.Footnote 64
The EU made similar arguments in its submission.Footnote 65
If taken at face value, this ruling will lead to spurious findings of violation. Under WTO law, Member States have an ‘undisputed’ right to ‘set the level of protection they wish to achieve’.Footnote 66 They are free to pursue regulatory goals, even if this is detrimental to imported products. As Verhoosel has argued, if every measure with a disparate impact automatically violates the National Treatment Principle, it becomes an instrument of deregulation rather than non-discrimination (Verhoosel, Reference Verhoosel2002: 48).Footnote 67
Article XX is not sufficient to capture the ‘policy-based’ component of the non-discrimination analysis
A measure found to violate Article III may be saved by the General Exception, Article XX, if it falls under its subparagraphs, which delimit particular public policy goals such as public morals, human, animal, and plant life and health, prevention of deceptive practices, and protection of exhaustible natural resources, inter alia.Footnote 68 In the language of the Article XX chapeau, the measure cannot be ‘applied in a manner which would constitute a means of arbitrary or unjustifiable discrimination between countries where the same conditions prevail, or a disguised restriction on international trade’.Footnote 69
Article XX forms an important component of enabling governments to regulate in the public interest. Yet under a disparate impact approach, Article XX becomes the only provision to consider whether a trade-restrictive regulation has a non-protectionism motive. This makes an exception to the GATT necessary to the complete application of one of its articles. Article XX covers a finite list of exceptions; regulations not covered may also have a non-discriminatory motive. Also, not all regulatory goals are clearly identified under Article XX. One example is trade-restrictive regulation that protects products with particular cultural value. Cultural value is recognized only very narrowly under the list of exceptions (Burri-Nenova, Reference Burri-Nenova2009). In practice, of course, if the Appellate Body is ‘stricter’ with Article III, it can apply Article XX ‘generously’. Even so, this approach seems artificial.
Disparate impact ‘imbalances’ TBT Article 2.1 vis-à-vis GATT Article III(4)
The outcome of EC–Seal Products has formalized a disparity between Article III and TBT Article 2.1. Under both TBT Article 2.1 and GATT Article III(4), the Appellate Body first establishes a disparate impact. The TBT Agreement contains an additional step, under which the Appellate Body considers whether a disparate impact is explained by a ‘legitimate regulatory distinction’. To determine whether a regulatory distinction is legitimate, it takes into account whether a measure has been applied in an even-handed manner to foreign and domestic ‘like’ products by examining the design, architecture, and revealing structure of the measure. These elements parallel the ‘protective application’ test under Article III(2). This test differs, however, by additionally considering whether there is a rational relationship between the measure in dispute and the regulatory goal.Footnote 70
The TBT rulings offered a promise of an evolutionary interpretation of the National Treatment Principle more broadly. Responding to the rulings, for example, Flett wrote: ‘In the most recent cases the tail (the TBT Agreement) has wagged the dog (the GATT 1994), re-affirming the regulatory space in Article III:4 and, by extension, Article III:2, and heralding a further improvement in the previously unsettled balance between the trade interest and national regulatory autonomy’ (Flett, Reference Flett2013: 39).
Instead of consolidating this improvement, the Appellate Body deepened the divide between Article III(4) and TBT Article 2.1, in the process affirming its difference from Article III(2)). Differences in the ‘context, object and purpose’ of the TBT Agreement justified this differing approach.Footnote 71 Primarily, this difference was the availability of Article XX under the GATT Agreement but not the TBT Agreement. The Appellate Body stated:
In our view, the fact that, under the GATT 1994, a Member's right to regulate is accommodated under Article XX, weighs heavily against an interpretation of Articles I:1 and III:4 that requires an examination of whether the detrimental impact of a measure on competitive opportunities for like imported products stems exclusively from a legitimate regulatory distinction. In the light of the immediate contextual differences between the TBT Agreement and the GATT 1994, we do not consider that the legal standard for the non-discrimination obligation under Article 2.1 of the TBT Agreement applies equally to claims under Articles I:1 and III:4 of the GATT 1994.Footnote 72
In fact, neither the existence of Article XX nor the contextual differences between the agreements justify the divergence. The TBT Agreement contains a component similar to GATT Article XX, Article 2.2, though it does not contain a closed list of specific negotiated exceptions, but covers any ‘legitimate objective’. If a measure complies with Article XX, it complies with the GATT. Conversely, if measure complies with TBT 2.2, it can still violate TBT 2.1 and therefore the TBT Agreement as a whole. In this sense, the TBT test is more intrusive into domestic regulation, justifying a less strict approach to TBT Article 2.1 (Howse and Levy, Reference Howse and Levy2013: 350; Ming Du, Reference Ming Du2007).
In practice, however, the ruling has made Article III(4) a more difficult test to pass. The EU argued that creating a divergent test would ‘“render Article 2.1 of the TBT Agreement irrelevant” as complainants would have a strong incentive not to invoke Article 2.1 of the TBT Agreement, and, instead, to bring claims under the GATT 1994, even if the measure at issue qualified as a technical regulation’.Footnote 73 In EC–Seal Products, Norway made its complaint against the EU only under GATT Article III, and not under TBT Article 2.1. Clearly, Norway perceived GATT Article III as being tougher. The difference in interpretation has the likely result that complaining countries will follow Norway in focusing their complaints on the GATT Agreement.
This was also noted by the US following the dispute:
we are not fully persuaded by the Appellate Body's finding that the national treatment provisions of the TBT Agreement are to be interpreted differently from the national treatment provisions of the GATT 1994 in light of the fact that these two provisions contain identical wording….. Indeed, these findings raise the very real possibility, as demonstrated in this dispute, that Article 2.1 of the TBT Agreement will become superfluous, and the legal approach developed in the recent TBT disputes will become just an historical footnote.Footnote 74
The primary contextual difference between the provisions is that TBT Article 2.1 deals exclusively with technical regulations. The definition of technical contains several components,Footnote 75 but with respect to the importance of considering the regulatory rationale, there does not seem to be any clear reason to differentiate the legal approach.
As summarized by Davey and Maskus,
Of course, the justification is that the context of TBT Article 2.1 requires this result, but it is hard to see why the context of Article III(4) – and specifically the context supplied by Article 3.1 – should not also require such a result … [The Appellate Body's divergent] approach in [the TBT dispute] US–Clove Cigarettes…highlights the problems with the way it has largely read the less favourable treatment requirement out of Article III:4. (Davey and Maskus, Reference Davey and Maskus2013: 180–181)
The TBT Agreement was negotiated approximately fifty years after the GATT Agreement and its Article III to build upon past experience and create a more specialized legal instrument. While there have been only a handful of disputes under this provision, becoming an ‘historical footnote’ would be an unfortunate outcome for this valuable jurisprudence.
How to incorporate regulatory rationale into Article III(4)
This article has argued for greater convergence between Article III(2) and III(4) and TBT Article 2.1, and that examining ‘aim’, or regulatory intent, should be a part of the test of protectionism under all three. One important question is the depth of the reforms necessary to create this convergence. Verhoosel proposed the deep reform of replacing both Article III/XX and TBT Article 2.1/2.2 with integrated tests that incorporate the same fundamental stages (Verhoosel, Reference Verhoosel2002). This elegant solution would reduce the complexity that arises from separating these components. In deference to the practical difficulties of treaty amendment, the reforms proposed here are more moderate: maintaining the existing provisions and trying to work around problems of replication. These reforms depart from existing case law, but there is no bar to this departure. Existing case law clearly plays an important role in shaping Appellate Body decisions, and yet, unlike common law systems, the Appellate Body is not bound by its own precedents. In fact, this very article documents how the case law on regulatory purpose has evolved over time.
A relatively simple means for reforming Article III(4) is to bring it in line with Article III(2), so that it incorporates the same test of protective application. This would go a significant distance toward rectifying the current imbalance, but there would still be a gap between Article III(4) and TBT Article 2.1, as under the latter the Appellate Body additionally considers whether the policy objective at stake rationally accounts for the difference in treatment between ‘like’ domestic and imported products. One complication to utilizing the TBT Article 2.1 approach under Article III(4) is that the discrimination analysis under the Article XX chapeau contains this same test.Footnote 76 Though there are some differences,Footnote 77 this stage of the analysis would effectively be replicated in Article III(4) disputes in which the regulating country invoked Article XX.
One solution is to incorporate what might be termed a ‘negative’ approach, rather than a ‘positive’ one. A negative approach would examine whether any factors unrelated to the origin of the product explained the difference in treatment. This would be similar to the ‘genuine relationship’ test, but focus on establishing a policy-based rather than market-based causal link. Considering any potential explanation differs from the positive approach taken under TBT Article 2.1, which focuses on the specific policy justification put forth by the regulating country. Pragmatically speaking, a positive approach is likely less cumbersome, and avoids an open-ended process of elimination of what other objectives the regulation in dispute might pursue. Nonetheless, embracing this negative approach would have the benefit of incorporating an analysis of regulatory rationale without replicating the chapeau test.
Incidentally, there is the possibility that regulation discriminates against imported products as the result of lack of foresight or poor drafting. In this scenario, neither a protectionist nor a ‘legitimate’ aim would account for the disparate impact. The reason to examine the intent of a measure is to enable an appropriate amount of regulatory space for legitimate policy goals, so if there is what might be termed accidental protectionism, this should indeed fall foul of WTO rules.
An explicit and consistent consideration of regulatory purpose, following any of the approaches proposed above, enables an expanded use of evidence. There should be no bar to examining statements of intent in the wording of the legislation itself, or its preparatory documents. Such evidence can enhance a rigorous examination of whether a measure is protectionist.
Any feasible reform proposal must of course also consider why the Appellate Body has taken such a cautious and circumspect approach. There are political difficulties involved with empowering an international judiciary to determine the purpose of domestic regulation. These proposed reforms may appear to give carte blanche to the Appellate Body, empowering it with judicial authority that goes beyond its limited competencies. This argument rests in the assumption that eliminating a consistent and explicit consideration of regulatory purpose will somehow prevent the Appellate Body from undermining national regulatory autonomy.
This is false. Under TBT Article 2.1, the Appellate Body undertakes precisely this type of analysis when evaluating the legitimacy of the regulatory objective against an open-ended list of objectives, without such ill effect. Indeed, these disputes provide a model for how such an analysis can take place. Also, as argued above, the Appellate Body is already taking regulatory intent into account when adjudicating Article III disputes. By denying that this is a factor in its decisions, it has made its reasoning on this point less rigorous and more implicit. Thus, acknowledging this as a formal step of the analysis will actually hold the Appellate Body more to account in its evaluation of domestic regulatory priorities.
Discerning the existence of protectionism by its very nature necessitates a sophisticated act of judgment, as there is often an elusive relationship between the stated and ‘real’ purpose of regulation in dispute. Beyond including it as a formal stage of the analysis, it is not possible to prescribe a precise formula for a finding of protective intent. The same type of evidence may lead to opposite conclusions in different disputes. To take the example of preparatory statements for the legislation, while these should be taken into account they cannot be determinative. False negatives could occur if there were no statement of protectionist intent and the Appellate Body concluded that regulation was therefore not protectionist despite clear evidence to the contrary. False positives might also occur if a politician portrayed non-protectionist regulation as protectionist in order to gain support from certain constituencies. Instead, a combination of intent- and effect-based evidence must inform the analysis on a case-by-case basis. Despite the political difficulties involved in the Appellate Body determining the purpose of Member States' regulation, there is no avoiding the fact that the judges must judge.
7. Conclusion
Under the current approach to the National Treatment Principle, various components of the protectionism analysis are conflated. The Appellate Body has affirmed that assessing the competitiveness of products in dispute, and existence of a detrimental competitive impact, rests on market analysis. In practice, however, it has been circumspect in utilizing submitted quantitative evidence, and relied on qualitative concepts of potential competition and competitive opportunities. These interpretive strategies, not reflected in the treaty text, give rise at times to a false impression that qualitative conclusions are based on quantitative evidence, leading to a lack of transparency. Furthermore, they are speculative and sometimes questionable with respect to the evidence submitted.
There is a similar conflation, both explicit and implicit, between market-based and policy-based elements of the analysis. Following from EC–Seal Products, under Article III(4), the latter has simply been denied. As considering the policy intent of a regulation is an unavoidable part of the reasoning, this leads to a situation where it can only be incorporated implicitly. For example, the Appellate Body might defer to a compelling policy-based justification by deciding that a wide market segment of ‘like’ imported products has to be subject to a detrimental competitive impact (making violation more difficult to establish), or conclude that there has to be a large degree of detrimental impact.
This article advocates a return to the clarity of separate analysis of market-based ‘effect’ and policy-based ‘intent’ under the National Treatment Principle in TBT Article 2.1 and GATT Article III. Both should contain the same fundamental tests: considering whether products in dispute are in a competitive relationship and whether there is a detrimental impact on competition for imports, then whether this impact can be explained vis-à-vis a non-protectionist regulatory purpose. More convergence will reduce unnecessary complexity and rectify the current marginalization of the TBT Agreement, and the analyses can still take account of the differences between the provisions (for example, the relationship between GATT Article III and Article XX) in ways that are proposed above.
Establishing the existence and nature of competitive relationships based on market evidence will lead to more rigour and transparency. A politically unpalatable outcome will be offset by the subsequent step which enables a compelling regulatory rationale to play a role in the analysis. This will eliminate the need to acknowledge policy-based factors through market-based reasoning. It is even-handed in its approach to all policy objectives that a government may wish to pursue, rather than relying on implicit hierarchies of regulatory importance, or the policy objective falling under a delimited list of ‘exceptions’ under Article XX.
The National Treatment Principle, and the balance it strikes between national sovereignty and WTO obligations, is one of the most sensitive issues in WTO law. As the above-documented disputes demonstrate, in some cases moving away from an evidence-based approach and relying upon more speculative methods to establish disparate impact has been used to extend the jurisdiction of the National Treatment Principle. Denying the role of regulatory intent in the protectionism analysis under Article III(4) has the same impact, as detrimental competitive impact on imports is not enough of a basis for findings of non-compliance. The result in both cases is an imbalance toward trade liberalization at the expense of domestic regulatory autonomy.
