1. Introduction
Under WTO law, can a WTO member enter a mutual recognition agreement with a second member without incurring a most-favoured nation obligation to other members? The answer is likely no, but the precise obligation owed to other members will vary depending on the situation. The most-favoured nation (MFN) obligations imposed by the General Agreement on Tariffs and Trade 1994 (GATT)Footnote 1 and the Agreement on Technical Barriers to Trade (TBT)Footnote 2 apply to mutual recognition agreements (MRAs) as they do other trade measures, but within certain limits. An MRA is an agreement in which two or more members recognize the equivalence of each other's technical regulationsFootnote 3 or conformity assessment procedures (CAPs)Footnote 4 for a certain sector of goods. An MRA on technical regulations allows products which comply with the technical regulations of one party to be offered for sale in the market of another party (and vice versa), just as if the products had complied with that second party's technical regulations. An MRA on CAPs authorizes the conformity assessment bodies of one party to certify products' conformity with the technical regulations of another party (and vice versa), so the products may be offered for sale in the market of the second party without undergoing its CAPs.
Both types of MRA have potential benefits and drawbacks. Technical regulation MRAs allow domestic producers to sell current products in a foreign market, but allow the sale in the domestic market of products which do not fully comply with domestic regulations. CAP MRAs only allow domestic producers a lesser benefit of using familiar domestic CAPs to assess their products' conformity with foreign technical regulations, but they maintain the application of domestic technical regulations to foreign products sold in the domestic market. CAP MRAs also require domestic conformity assessment bodies to develop the capacity to assess conformity with foreign technical regulations, while technical regulation MRAs do not.
The key question for WTO members entering an MRA is how another member might use the existence of the MRA to assert that the parties to the MRA owe an MFN obligation to that third member.Footnote 5 The analysis for answering this question centers on the MFN obligations of TBT Article 2.1 (TBT 2.1) and GATT Article I:1 (GATT I:1) and the textual differences between the two. Section 2.1 of this article begins the analysis with an overview of the WTO jurisprudence regarding TBT 2.1, starting with US–Clove Cigarettes Footnote 6 and continuing through US–Tuna II,Footnote 7 US–Cool,Footnote 8 and EC–Seals.Footnote 9 Section 2.2 examines the TBT 2.1 ‘legitimate regulatory distinction’ test (articulated by the Appellate Body in US–Clove Cigarettes) in detail, compares this test with the GATT Article XX necessity and chapeau test for exceptions to GATT I:1, and argues that in practice the application of the two tests should generally lead to the same result. Section 2.3 presents the remaining differences in the two tests, and suggests that despite its dedication to honoring the text, the Appellate Body is taking steps to avoid conflicting results between TBT 2.1 and the GATT provisions.
Section 3.1 of this article continues the analysis by applying these legal tests to technical regulation MRAs, predicting where WTO legal disputes might arise, and the likely outcomes. For technical regulation MRAs, there are four hypothetical situations which might create a dispute over MFN obligations: (1) the MRA includes rules of origin limiting the MRA to only cover goods originating in the parties to the MRA; (2) the MRA contains no rules of origin, but only covers goods which complete the CAPs of the second party to the MRA; (3) a party to an MRA rejects a third WTO member's request for recognition of its own technical regulations equal to what the first party has granted the second party to the MRA; or (4) the first party rejects the third member's request for recognition of its own technical regulations equal to what the first party has granted the second party to the MRA, and that third member has also entered a similar MRA with the second party.
Section 3.2 begins by arguing that the MFN obligation of TBT Article 5.1.1 (TBT 5.1.1) should be applied to CAPs in essentially the same way TBT 2.1 is applied to technical regulations, since TBT 2.1 itself likely does not apply to CAPs. The section goes on to apply the suggested TBT 5.1.1 test and the GATT I:1 / GATT XX test to CAPs MRAs. For CAPs MRAs, there are three situations which might create a dispute over MFN obligations: (1) the MRA includes rules of origin limiting the MRA to only cover goods originating in the parties to the MRA; (2) a WTO member party to an MRA rejects a third member's request for recognition of its CAPs equal to what the first member has granted the second party to the MRA; or (3) the first member rejects a third member's request for recognition of its CAPs equal to what the first member has granted the second party to the MRA, and that third member has also entered a similar MRA with the second party.
Applying the legal tests to each situation in turn yields varying results; overall it appears that members who enter an MRA will incur an MFN obligation to other members, but only those members whose products, technical regulations, or CAPs are truly equivalent to those of one of the MRA parties. Thus, in conclusion this article argues that although MRAs hold certain risks, the risks are manageable and should not prevent WTO members from utilizing this valuable tool for trade liberalization.
2. The US–Clove Cigarettes test
To determine whether a WTO member can legally limit the extension of bilateral MRAs to other members and their products, it is necessary first to identify the legal test or tests which the WTO Appellate Body would apply in answering that question. For technical regulation MRAs, the Appellate Body will look to the MFN obligations of TBT Article 2.1Footnote 10 and GATT Article I:1.Footnote 11 However, the two obligations have obvious differences – GATT I:1 requires that ‘any advantage, favour, privilege or immunity’ granted to one member's products ‘shall be accorded immediately and unconditionally’ to other members, while TBT 2.1 instead echoes the ‘less favourable treatment’ standard of GATT Article III:4 (GATT III:4).Footnote 12 GATT I:1 is tempered by the general exceptions provision of GATT Article XX (GATT XX), but these exceptions do not apply to the TBT,Footnote 13 and the TBT has no general exceptions provision of its own. For these reasons, it is easy to imagine situations in which the application of TBT 2.1 and GATT I:1 will yield conflicting results. These potential conflicts have led to scholarly debates regarding the TBT as the possible lex specialis,Footnote 14 the scope of the General Interpretative Note to Annex 1A of the Marrakesh Agreement,Footnote 15 and the definition of what is or is not a conflict within WTO law.Footnote 16 The Appellate Body has now resolved many of these debates in a series of reports beginning with US–Clove Cigarettes Footnote 17 and continuing through US–Tuna II,Footnote 18 US–COOL,Footnote 19 and EC–Seals.Footnote 20 Footnote 21 The Appellate Body's reasoning throughout these decisions reveals its approach to the potential conflicts between the TBT and GATT – the Appellate Body seeks to interpret the two agreements in ‘a coherent and consistent manner’,Footnote 22 that is interpretations under which the TBT 2.1 analysis and GATT I:1 analysis yield the same result.
2.1 From US–Clove Cigarettes to EC–Seals
In US–Clove Cigarettes, the Appellate Body articulated the two-part test for whether a measure provides less favourable treatment under Article 2.1: (1) does the measure detrimentally modify the conditions of competition for the imported product, and (2) if so, does that detrimental impact stem exclusively from a legitimate regulatory distinction.Footnote 23 The second part functions as an affirmative defense for measures which fail the first.Footnote 24 In this case, the Appellate Body held that a US measure banning the sale of flavored cigarettes violated the national treatment obligation of Article 2.1, because the measure contained an exception for menthol cigarettes (which are made in the US) but not for clove cigarettes (which are made in Indonesia).Footnote 25 The measure detrimentally affected the conditions of competition for imported clove cigarettes versus domestic menthol cigarettes, and that effect did not stem from a legitimate regulatory distinction.Footnote 26
In US–Tuna II, the Appellate Body held that a US measure allowing a ‘dolphin-safe’ label for tuna caught under certain conditions violated Article 2.1,Footnote 27 because in effect the conditions only excluded the majority of Mexican tuna.Footnote 28 Interestingly, the measure violated both the national treatment and MFN obligations of Article 2.1, because it provided less favourable treatment to imported Mexican tuna versus both domestic US tuna and other imported tuna. Applying the test laid out in US–Clove Cigarettes, the Appellate Body found that US dolphin-safe labeling detrimentally modified conditions of competition for Mexican tuna.Footnote 29 It also found that the detrimental impact of the measure did not stem from a legitimate regulatory distinction.Footnote 30 The measure's goals of protecting consumers and protecting dolphins were legitimate regulatory objectives,Footnote 31 but the measure did not apply equally as between the dolphin-endangering fishing method used by Mexican fisherman and dolphin-endangering methods used elsewhere.Footnote 32 Therefore, the measure failed the test because it was not even-handed in the pursuit of its objectives.
In US–COOL, the Appellate Body held that a US country of origin labeling measure for beef and pork violated the national treatment obligation of Article 2.1.Footnote 33 The Appellate Body believed the record-keeping necessary to comply with the measure would drive US producers and retailers to deal only in livestock born, raised, and slaughtered in the US, as the most cost-effective method of complying with the labeling scheme.Footnote 34 Applying the test once more,Footnote 35 the Appellate Body found that the labeling scheme detrimentally modified conditions of competition for imported beef and pork,Footnote 36 and that the detrimental effect did not stem from a legitimate regulatory distinction.Footnote 37
In EC–Seals, the Panel held that an EU measure banning the sale of seal products violated the MFN and national treatment obligations of Article 2.1.Footnote 38 The violations arose from the measure's Inuit (IC) and Marine Resource Management (MRM) exceptions; in application, the IC exception allowed the sale of Greenlandic Inuit seal products but not those of Canada,Footnote 39 and the MRM exception allowed the sale of EU seal products but not Canadian products.Footnote 40 Applying the Appellate Body's US–Clove Cigarettes test, the Panel found that the exceptions detrimentally modified conditions of competition for Canadian seal products,Footnote 41 and that the detrimental effects did not stem from legitimate regulatory distinctions.Footnote 42 More importantly, based on the same reasoning, the Panel went on to find that these exceptions also violated the MFN and national treatment obligations of GATT Articles I:1 and III:4.Footnote 43
Prior to EC–Seals, panels had avoided applying the GATT in TBT cases, citing judicial economy.Footnote 44 The Panel's decision to instead complete the GATT analysis in EC–Seals became crucial to the outcome, because the Appellate Body voided the Panel's findings under TBT 2.1 after deciding that the EU Seal Regime was not in fact a technical regulation.Footnote 45 The Appellate Body upheld the Panel's finding that the regime violated GATT I:1Footnote 46 (the EU did not appeal the Panel's finding that the regime violated GATT III:4Footnote 47 ), and ultimately agreed that the EU failed to justify the regime under GATT XX.Footnote 48 Although EC–Seals in the end was not technically a TBT case, the Appellate Body's decision illustrated how it might apply the GATT in future TBT cases.
2.2 Legitimate Regulatory Distinction and GATT Article XX
In articulating the US–Clove Cigarettes test, the Appellate Body attempted to rationalize the TBT by bringing it into harmony with the GATT. It interpreted TBT 2.1 within the context provided by GATT III:4 and the absence in the TBT of a general exception provision equivalent to GATT XX.Footnote 49 Against that background, the Appellate Body redefined ‘less favourable treatment’ for TBT 2.1. Logically, the Appellate Body first borrowed the ‘detrimental effect on conditions of competition’ definition of less favourable treatment from GATT III:4,Footnote 50 citing its Korea–Beef Footnote 51 and EC–Asbestos Footnote 52 decisions. However, while the GATT III:4 analysis of less favourable treatment ends there and moves to the GATT XX exceptions, the TBT 2.1 analysis adds a second prong of ‘legitimate regulatory distinction’.
In creating this second prong, the Appellate Body noted that all technical regulations establish distinctions between products; therefore, TBT 2.1 cannot prohibit all regulatory distinctions as less favourable treatment, or by extension it might prohibit all technical regulations.Footnote 53 Implicit in this argument is the apparent need for a general exceptions provision in the TBT. So, the Appellate Body looked to the object and purpose of the TBT, giving particular consideration to three items in the preamble sixth recital and Article 2.2 of the TBT which mirror GATT XX: a list of legitimate regulatory objectives which members should not be prevented from pursuing, a prohibition of arbitrary discrimination and disguised restrictions on trade, and a necessity requirement.Footnote 54 Based on these items, the Appellate Body decided the object and purpose of the TBT required a second step in the less favourable treatment test of TBT 2.1,Footnote 55 which in application is rather similar to GATT XX.
2.2.1 The Legitimate Regulatory Distinction analysis
The analysis for this second prong asks:
whether the detrimental impact on imports stems exclusively from a legitimate regulatory distinction rather than reflecting discrimination against the group of imported products. In making this determination, a panel must carefully scrutinize the particular circumstances of the case ‒ that is, the design, architecture, revealing structure, operation, and application of the technical regulation at issue, and, in particular, whether that technical regulation is even-handed ‒ in order to determine whether it discriminates against the group of imported products.Footnote 56
In practice, based on the Appellate Body's findings, this analysis has two elements, that is a legitimate regulatory distinction is one that: (1) pursues a legitimate regulatory objective, and (2) is designed and applied such that no reasonableFootnote 57 alternative distinction would more effectively pursue the objective. (Interestingly, the analysis appears to seek alternatives which better achieve the legitimate objective, rather than those which are less trade restrictive.Footnote 58 ) Any distinction which does not meet this definition is deemed discriminatory.
The first element implicitly includes the three steps of identifying potential objectives, whether an objective is legitimate, and whether the distinction pursues (contributes to) that objective. For example, in US–Clove Cigarettes the US cited the regulatory objectives of avoiding the health care cost of treating menthol cigarette addicts and the development of a black market, but these objectives were not legitimate.Footnote 59 Additionally, the Appellate Body considered the legitimate objective of reducing youth smoking, but the distinction between menthol cigarettes and clove cigarettes did not contribute to that objective.Footnote 60 Therefore, the distinction failed the first element, and was found to be discriminatory.Footnote 61 The two subsequent cases illustrate the second element. In US–Tuna II, the distinction between tuna-fishing methods satisfied the first element; the identified regulatory objectives of consumer protection and protection of dolphins were legitimate, and the distinction pursued those objectives.Footnote 62 However, the distinction could have been better calibrated to also address the risks to dolphins from methods other than the method used in Mexico.Footnote 63 Therefore, the distinction failed the second element, and was found to be discriminatory.Footnote 64 In US–COOL, the regulatory objective of providing information to consumers was legitimate, and the distinctions created by the labeling scheme pursued that objective.Footnote 65 However, the labeling scheme could have been much better designed to convey to consumers all the product-origin information which producers were required to track.Footnote 66
2.2.2 The GATT Article XX analysis
The analysis for whether a measure meets the requirements of the GATT XX exceptions and chapeau is somewhat more complicated, but yields similar results. As defined by the Appellate Body, this analysis begins with identifying the objective of the measure,Footnote 67 whether the objective falls within the exceptions,Footnote 68 and whether the measure contributes to the achievement of the objective.Footnote 69 Next, as to whether the measure is necessary:
[A] panel must assess all the relevant factors, particularly the extent of the contribution to the achievement of a measure's objective and its trade restrictiveness, in the light of the importance of the interests or values at stake. If this analysis yields a preliminary conclusion that the measure is necessary, this result must be confirmed by comparing the measure with its possible alternatives, which may be less trade restrictive while providing an equivalent contribution to the achievement of the objective pursued.Footnote 70
Such alternatives must be reasonably available to the responding member, without imposing an undue burden such as prohibitive costs or substantial technical difficulties.Footnote 71 And, finally, as to the chapeau restriction on arbitrary or unjustifiable discrimination and disguised restrictions on trade, the analysis is generally whether the cause or rationale of the measure's discriminatory effect is legitimate under the objectives listed in Article XX.Footnote 72 As demonstrated by the Panel and Appellate Body decisions in EC–Seals, in application this more complicated analysis should yield the same results as the TBT 2.1 ‘legitimate regulatory distinction’ analysis.
The EC–Seals Panel gave full effect to the Appellate Body's attempts to harmonize the TBT and GATT. As discussed above, the Panel first examined Canada's claims under TBT 2.1, finding that the exceptions in the EU's regulatory regime violated the MFN and national treatment obligations of TBT 2.1. When the Panel then reached Norway's similar claims under GATT I:1 and III:4, it equated the steps of the GATT analysis with previous steps under TBT 2.1, and essentially said ‘see above’. First, in analyzing whether the regulatory regime violated GATT I:1 and III:4, the Panel simply referred back to its previous finding that the regime detrimentally modified the conditions of competition for Canadian products under TBT 2.1.Footnote 73 Next, after reiterating that the protection of public morals on seal welfare is a legitimate objective, the Panel did conduct a separate analysis of whether the regime was ‘necessary’ under GATT XX(a), finding that it was in fact necessary.Footnote 74 However, in its final analysis of whether the regime met the requirements of the GATT XX chapeau, the Panel referred straight back to its TBT 2.1 analysis of legitimate regulatory distinction, stating again that the regime's exceptions constituted arbitrary and unjustifiable discrimination.Footnote 75
The Appellate Body placed far more emphasis on the textual differences in the GATT and TBT tests, but eventually reached the same conclusions. For example, the Appellate Body noted that the TBT list of legitimate objectives is open ended, while the GATT XX list is closed; however, in principle the balance struck between avoiding unnecessary trade obstacles and recognizing members' rights to regulate in the TBT is the same as the balance between the GATT I:1 and III:4 obligations and the exceptions in Article XX.Footnote 76 In another example, the Appellate Body actually held that the Panel erred in equating the GATT XX chapeau analysis with the TBT 2.1 legitimate regulatory distinction analysis; although there are ‘important parallels between the analyses’, the Appellate Body could not overlook the ‘significant’ textual differences.Footnote 77 However, after conducting a separate GATT XX chapeau analysis, the Appellate Body still found that the EU Seal Regime failed the test.Footnote 78
2.2.3 Remaining differences
The most problematic differences between the TBT 2.1 legitimate regulatory distinction analysis and the GATT XX analysis are the identification and legitimacy of objectives, particularly for exceptions within measures. First, under the GATT XX analysis, the relevant objective is always that of the overall measure, even when the discrimination is caused by an exception within the measure;Footnote 79 while under the TBT 2.1 analysis, the relevant objective is that of the discriminatory distinction and need not relate to the objective of the overall measure.Footnote 80 So, a measure might pass the TBT 2.1 test based on the legitimate objective of an exception, but fail the GATT XX chapeau test because the exception does not pursue the objective of the overall measure. Second, as noted by the Appellate Body, the lists of legitimate objectives in GATT XX and the TBT sixth recital are not identical; so, the objective of a regulatory measure could theoretically be legitimate under one list and not the other. However, the Appellate Body appears to have sidestepped both of these issues in EC–Seals, by stating that the relationship of the discrimination and objective of the measure in question is only one important factor, not the sole test, for deciding whether the measure meets the GATT XX chapeau requirements.Footnote 81 The Appellate Body thus created a two-step chapeau test: (1) whether the discrimination-causing exception pursues the objective of the whole measure (which is unlikely to ever be true of an exception), and (2) whether the discrimination is non-arbitrary and justifiable. Since these steps are ‘factors’, it appears that a measure could pass the chapeau test by passing either of the steps. In EC–Seals, the measure failed both steps – the discrimination-causing IC exception did not pursue the whole measure's objective of addressing EU moral concerns for seal welfare, and the discrimination was arbitrary and unjustifiable due to the ambiguous criteria and over-broad discretion for regulatory bodies in the IC exception.Footnote 82
This test gives the Appellate Body the flexibility necessary to avoid conflicts between the GATT and TBT analyses, because it avoids the question of whether the exception pursues one of the legitimate objectives listed in GATT XX. Similar to the TBT 2.1 legitimate regulatory distinction test, here in the second step the Appellate Body examines whether an alternative measure might better pursue a legitimate objective; also as in the TBT 2.1 test, the Appellate Body accomplishes this by pointing out errors within the existing measure which ought to be corrected (rather than calling the corrected measure an alternative). However, here the Appellate Body neither truly identifies the objective (presumably protecting the economic and social interests of the Inuit) nor determines its legitimacy. The measure was arbitrary because the Appellate Body believed it could be better designed to pursue an unnamed and unlisted objective; but, had the Appellate Body believed differently, the measure could equally have passed the chapeau test because it properly pursued an unnamed and unlisted objective.Footnote 83
3. Applying the test to MRAs
These recent cases have greatly clarified the analysis for claims of discriminatory technical regulations, and the potential inconsistencies between TBT 2.1 and GATT I:1. In doing so, they also provide the framework within which a panel or the Appellate Body would analyze a claim of discriminatory mutual recognition of technical regulations or CAPs between two WTO members, as raised by a third member excluded from the recognition. The following subsections demonstrate the application of the TBT 2.1 and GATT I:1 tests in the context of mutual recognition, analyzing the situations in which a third member may raise such a claim, first regarding technical regulation MRAs and second regarding CAP MRAs.
3.1 Mutual recognition of technical regulations
As stated in the Introduction, for technical regulation MRAs an MFN claim could arise in four situations: (1) the MRA includes rules of origin limiting the MRA to only cover goods originating in the parties to the MRA, (2) the MRA contains no rules of origin, but only covers goods which complete the CAPs of the second party to the MRA, (3) a WTO member party to an MRA rejects a third member's request for recognition of its own technical regulations equal to what the first member has granted the second party to the MRA, or (4) the first member rejects the third member's request for recognition of its own technical regulations equal to what the first member has granted the second party to the MRA, and that third member has also entered a similar MRA with the second party. The first two situations involve a complaining member (Member C) whose products comply with the technical regulations of a second member (Member B) with whom the respondent (Member A) has entered an MRA. The third and fourth situations involve Member C asserting that its own technical regulations are equivalent to Member B's regulations (and therefore entitled to the same recognition from Member A), even if Member C's products do not fully comply with Member B's regulations.Footnote 84 These hypothetical situations assume that the relevant products of Member C and Member B are ‘like products’. The underlying technical regulations are assumed to be legitimate under TBT 2.1 and GATT I:1, and might regulate product characteristics and/or related process and production methods.
The identification of the measure and its objective is an issue which arises in each of the situations. In the context of MRAs, the ‘measure’ consists of multiple pieces: the underlying technical regulations of Member A and Member B, the MRA and (in some situations) Member A's denial of Member C's request for recognition. The analysis should include all these as parts of the measure in question because they are substantially related; that is, the MRA and denial of recognition cannot properly ‘be understood strictly by themselves, totally divorced from’ the underlying technical regulations.Footnote 85 Neither the MRA nor the denial of recognition pursues a legitimate objective of its own, but both might pursue the legitimate objective of the underlying technical regulations. An MRA creates in effect an amended set of technical regulations for Member A, which state that foreign products may now qualify for sale in Member A's market by complying with either of two lists of criteria (Member A's existing regulations or Member B's regulations). This new set of technical regulations (the MRA) continues to pursue the same legitimate objectives (i.e. health, safety, consumer protection, etc.) as the previous regulations. Member A's denial of recognition also pursues the legitimate objectives of the underlying technical regulations, if Member C's technical regulations do not equally fulfill those objectives.
3.1.1 First situation
In the first situation, if the MRA contains rules of origin excluding goods from all but Member B, an excluded Member C whose products comply with Member B's technical regulations might claim that Member A is violating the MFN obligations of TBT 2.1 and GATT I:1. Member C will likely win both claims. Applying the TBT 2.1 test, the MRA's rules of origin modify the conditions of competition in Member A's market to the detriment of Member C's products versus Member B's products, because the rules ban the sale of Member C's products and allow those of Member B. This is simple de jure discrimination based on origin. Also, the distinction in the rules of origin between the products of Member B and Member C is not a legitimate regulatory distinction. The measure as a whole (the MRA and the underlying regulations) continues to pursue the legitimate objectives of the underlying regulations, but the distinction contained in the rules of origin does not pursue those objectives or any other legitimate objective. Under the GATT I:1 analysis, the rules of origin grant an advantage to Member B without immediately and unconditionally according the same advantage to Member C. The measure as a whole should be found necessary for one of the exceptions of GATT XX, based on the objectives of the underlying technical regulations; but the rules of origin fail the GATT XX chapeau requirements, just as they fail the legitimate regulatory distinction test under TBT 2.1.
3.1.2 Second situation
In the second situation, if the MRA has no rules of origin but applies only to goods which complete Member B's CAPs, Member C still has an arguable claim that Member A is violating the MFN obligations of TBT 2.1 and GATT I:1, based on de facto discrimination. Member C's claims are unlikely to succeed, but the possibility exists. Applying the TBT 2.1 test, Member C can argue that the MRA detrimentally modifies the conditions of competition, due to additional shipping expenses. In this situation, Member C's producers must ship their products first to Member B for assessment, while Member B's producers may ship their products directly to Member A, giving Member B's producers a cost advantage. The lack of familiarity and likely language barrier for Member C's producers in using Member B's CAPs may create a further cost advantage for Member B's producers. Given the Appellate Body's willingness to consider hypothetical effects on competition,Footnote 86 these potential added costs likely qualify as a detriment to conditions of competition for Member C's products in Member A's market.Footnote 87
However, this distinction between the products of Member C and Member B is likely a legitimate regulatory distinction. The distinction in this situation is caused by the requirement that Member C's products complete Member B's CAPs before sale in Member A's market. This requirement pursues the legitimate objectives of the underlying technical regulations by ensuring that Member C's products actually comply with those regulations.Footnote 88 As to whether the distinction could be better designed to pursue the objective, a possible alternative is that Member C's products could undergo CAPs in Member A, with Member A's conformity assessment bodies assessing the products' conformity with Member B's regulations. But, this alternative may be unreasonable given the cost to Member A of adapting its conformity assessment bodies to assess conformity with both its own technical regulations and Member B's technical regulations, a cost which the technical regulation MRA is specifically designed to avoid (as opposed to a CAPs MRA). Also, this alternative may not be a valid alternative according to the analysis – it is less trade restrictive than the measure in question, but no more effective in pursuing the legitimate objectives of the underlying technical regulations. So, in this situation the distinction may in fact be legitimate, meaning the MRA would not violate the MFN obligation of TBT 2.1. However, in practice no regulatory distinction has yet passed the legitimate regulatory distinction test (the distinctions at issue in US–Clove Cigarettes, US–Tuna II, US–Cool, and EC–Seals all failed the test); so there remains a substantial risk that the Appellate Body would find this distinction illegitimate as well, contrary to the above reasoning. .
Again, under GATT I:1 the result is the same. Although in this situation the measure does not de jure discriminate as before, in effect the MRA grants advantages to Member B and not Member C. Such de facto discrimination is also covered by the GATT I:1 MFN obligation.Footnote 89 Moving on to GATT XX, the measure as a whole should pass the necessity test for an exception based on the underlying technical regulations, and the MRA likely meets the GATT XX chapeau requirements given the legitimate objective and the unreasonable expense of the alternative. However, given its past distaste for regulatory measures with discriminatory effects, there remains a risk that the Appellate Body would instead decide the MRA is arbitrary and unjustifiable.Footnote 90
3.1.3 Third situation
In the third situation, if Member A enters an MRA with Member B and denies Member C's request for equal recognition of its technical regulations, Member C may claim that Member A's denial of its request violates the MFN obligations of TBT 2.1 and GATT I:1. The outcome of these claims depends solely on the factual finding as to whether, when compared to Member B's technical regulations, Member C's technical regulations equally or better fulfill the objectives of Member A's relevant technical regulations. Applying the TBT 2.1 test, Member A's decision detrimentally modifies the conditions of competition in Member A's market for Member C's products versus Member B's, because Member B's products can be sold in Member A's market while Member C's products are banned. (Here the products in question are those which comply with the exporting member's technical regulations, but not Member A's technical regulations.) If Member C and Member B's technical regulations equally fulfill the objectives of Member A's technical regulations, then there is no legitimate regulatory objective for Member A's distinction between the products of Member C and Member B. Conversely, if Member C's technical regulations do not equally fulfill the objectives of Member A's technical regulations, then the distinction also pursues the legitimate objectives of Member A's regulations, and no alternative distinction could better pursue those objectives.
Applying the GATT I:1 analysis, the result is the same. Member A has granted the advantage of market access to Member B and not to Member C. Under GATT XX, the measure as a whole (the underlying technical regulations, the MRA, and the denial of recognition) should be found necessary for the legitimate objectives of the underlying technical regulations. But, if Member B and Member C's technical regulations equally fulfill Member A's regulatory objectives, then Member A's denial of Member C's request does not meet the GATT XX chapeau requirements – it is arbitrary and unjustifiable discrimination. Conversely, if Member C's technical regulations do not equally fulfill Member A's regulatory objectives, the denial of Member C's request meets the chapeau requirements.
3.1.4 Fourth situation
The fourth situation is identical to the third, with the addition that Member C has also concluded an MRA with Member B covering the same technical regulations as the MRA between Member A and Member B. The analysis is also identical, except that Member C now has highly persuasive evidence that its technical regulations equally fulfill Member A's regulatory objectives, namely Member B's official recognition that Member C and Member B's regulations are equivalent.Footnote 91 (On the positive side, unlike in the previous situations, here Member A can equally claim that Member C has an MFN obligation to recognize Member A's technical regulations, due to Member C's MRA with Member B.)
3.2 Mutual recognition of CAPs
The analysis for CAPs under the TBT is less settled than the analysis for technical regulations, because the CAPs analysis depends on a separate set of TBT articles which have not been interpreted by a panel or the Appellate Body. TBT 2.1 does not apply to CAPs; CAPs by definition are not technical regulations, and TBT 2.1 by its terms applies only to technical regulations.Footnote 92 Within the TBT, technical regulations are covered by the section Technical Regulations and Standards containing articles 2 through 4, while CAPs are covered by the section titled Conformity With Technical Regulations and Standards containing articles 5 through 9. Within this section on CAPs, TBT Article 5.1.1 (TBT 5.1.1) is the counterpart to TBT 2.1, providing MFN and national treatment obligations for members in the preparation, adoption, and application of their CAPs.Footnote 93 The first question then is whether TBT 5.1.1 applies to MRAs. Both Bartels and Rigod have argued that, by its wording, the obligations of TBT 5.1.1 apply only to members' grants of access to their domestic CAPs, and not to recognition of another member's CAPs.Footnote 94 Unlike TBT 2.1, whose obligations apply broadly ‘in respect of technical regulations’, TBT 5.1.1 covers only the preparation, adoption, and application of CAPs by a member.Footnote 95 However, it can also be argued that by recognizing another member's CAPs, a member is ‘adopting’ and ‘applying’ those CAPs in a way which grants favourable access to the second member's products. Given the applicability of GATT I:1 and the Appellate Body's efforts to harmonize the TBT and GATT, the Appellate Body is more likely to accept this interpretation of TBT 5.1.1, rather than the argument that the TBT contains no MFN obligation for recognition of CAPs.
The second question is what legal test to apply to the MFN obligation of 5.1.1. Given the similarities between TBT 2.1 and TBT 5.1.1, the Appellate Body will most likely interpret ‘conditions no less favourable under TBT 5.1.1 as it has ‘treatment no less favourable under TBT 2.1. Presumably then, the first prong of the TBT 5.1.1 test is the same as that for ‘treatment no less favourable under TBT 2.1 and GATT III:4 – whether conditions of competition have been detrimentally modified. The final question is whether the TBT 5.1.1 test includes a second prong, matching the Appellate Body's test for TBT 2.1. Unlike technical regulations, CAPs do not by their nature create distinctions between products; CAPs simply enforce the distinctions created by technical regulations. However, given the context provided by the sixth recital of the TBT and TBT Article 5.1.2, the MFN and national treatment obligations of TBT 5.1.1 should be qualified to allow a member to ensure that all products sold in its market adequately conform to its technical regulations. So, the TBT 5.1.1 test needs a second prong, which might be called the ‘necessary procedural distinction’ prong. In this context, the distinction is that aspect of a CAP which is causing a discriminatory effect between the like products of two members. Similar to a legitimate regulatory distinction, a necessary procedural distinction should be one which has a legitimate objective, and which could not be better designed or applied to pursue that objective. However, for a necessary procedural distinction the only legitimate objective must be ensuring that a product conforms to a relevant technical regulation which itself pursues a legitimate regulatory objective.
Applying this test to CAP MRAs, claims of discriminatory mutual recognition could arise in three situations: (1) the MRA includes rules of origin limiting the MRA to only cover goods originating in the parties to the MRA,Footnote 96 (2) a WTO member party to an MRA rejects a third member's request for recognition of its CAPs equal to what the first member has granted the second party to the MRA, or (3) the first member rejects a third member's request for recognition of its CAPs equal to what the first member has granted the second party to the MRA, and that third member has also entered a similar MRA with the second party.Footnote 97 The first situation involves a complaining member (Member C) whose products are intended to comply with the respondent's (Member A) technical regulations, but will benefit from using the CAPs in the second party to the MRA (Member B) rather than Member A's CAPs. The second and third situations involve Member C asserting that its CAPs offer an assurance of conformity with Member A's technical regulations equal to that of Member B's CAPs.Footnote 98
3.2.1 First situation
In the first situation, if the MRA contains rules of origin excluding Member C's products, Member C may claim that Member A is violating the MFN obligations of TBT 5.1.1 and GATT I:1.Footnote 99 Both claims will likely succeed. Applying the TBT 5.1.1 test discussed above, the rules of origin detrimentally modify the conditions of competition for Member C's products versus Member B's in Member A's market, because Member B's products may complete the CAPs of either Member B or Member A (whichever is cheaper), while Member C's products must complete Member A's CAPs. This is de jure discrimination by Member A against Member C. For the second prong, Member A can offer no legitimate objective for the distinction created by the rules of origin. The underlying technical regulations may have legitimate objectives, but the rules of origin do not serve to ensure that products conform to those regulations. If Member B's CAPs offer adequate assurance of conformity with Member A's technical regulations, then the CAPs offer that assurance equally for the products of Member B and Member C, with no need to distinguish between the two. Therefore, the MRA would fail the ‘conditions no less favourable test of TBT 5.1.1.
Applying GATT I:1, the results are the same. By modifying the conditions of competition, the rules of origin grant an advantage to Member B without according the same advantage to Member C. The measure as a whole (the underlying technical regulations, the CAPs and the MRA) should be found necessary under a GATT XX exception, but it fails the GATT XX chapeau requirements due to the rules of origin.
3.2.2 Second situation
In the second situation, if Member A enters a CAP MRA with Member B and denies Member C's request for similar recognition of its CAPs, Member C may claim that Member A's denial of its request violates the MFN obligations of TBT 5.1.1 and GATT I:1. As in the similar situation for technical regulations, the outcome of these claims depends on the factual determination of whether the CAPs of Member C and Member B offer an equal assurance of conformity with Member A's technical regulations. Applying the TBT 5.1.1 test, Member A's decision detrimentally modifies the conditions of competition for Member C's products versus Member B's, because Member B's products may be sold in Member A's market after completing their familiar domestic CAPs, while Member C's products must complete the foreign CAPs of either Member B or Member A. For the second prong, if Member C and Member B's CAPs offer equal assurance of conformity with Member A's technical regulations, then there is no legitimate objective for the distinction created by Member A's denial of Member C's request. Conversely, if Member C's CAPs do not offer an equal assurance of conformity, then the distinction pursues the legitimate objective of ensuring conformity with Member A's technical regulations, assuming those regulations themselves pursue legitimate objectives. Finally, no alternative distinction could better ensure conformity with Member A's technical regulations.
Applying the GATT I:1 analysis, the result is the same. Member A has granted Member B's products the advantage of using their domestic CAPs to qualify for sale in Member A's market, and denied the same advantage to Member C's products. The measure as a whole (the underlying technical regulations, the CAPs, the MRA, and the denial of Member C's request) should be found necessary under a GATT XX exception. If Member C's CAPs offer an assurance of conformity with Member A's technical regulations equal to Member B's CAPs, then the denial of Member C's request fails the GATT XX chapeau requirements – it is arbitrary and unjustifiable discrimination. Conversely, if Member C's CAPs do not offer an equal assurance of conformity, then the denial meets the chapeau requirements.
3.2.3 Third situation
The third situation is identical to the second, with the addition that Member C has also concluded an MRA with Member B covering the same CAPs as the Member A/Member B MRA. The analysis is also identical, except that Member C may now have persuasive evidence that its technical regulations offer equal assurance of conformity with Member A's technical regulations, namely Member B's official recognition that Member C and Member B's CAPs are equivalent. However, this evidence is not as persuasive in the CAPs context as it is for mutual recognition of technical regulations. In the technical regulations context, Member B has recognized the equivalence of Member C's technical regulations and its own technical regulations, which are also equivalent to Member A's technical regulations. In the CAPs context, Member B has only recognized the equivalence of Member C's CAPs and its own for assessing conformity with Member B's technical regulations, while the CAPs of Member B and Member A are equivalent for assessing conformity with Member A's technical regulations. So, if the technical regulations of Member B and Member A are quite different, in the CAPs context the MRA between Member C and Member B is not particularly persuasive evidence.
4. Conclusion
In conclusion, a WTO member does incur an MFN obligation to third members by recognizing the technical regulations or CAPs of a second member. However, the only third members who could successfully invoke that obligation and demand recognition of their own technical regulations or CAPs are: (1) those whose products fully comply with the relevant technical regulations of the first or second member, and (2) those whose technical regulations or CAPs are equivalent to the second member's technical regulations or CAPs.Footnote 100 Given these limitations, WTO members have no great need to fear the effects of mutual recognition due to the MFN obligation. On the other hand, members could take steps to minimize the risk of losing a dispute in the WTO and potentially having to recognize another member's technical regulations or CAPs under less than optimal conditions. First, members should not apply rules of origin to mutual recognition, as this is very likely a violation of the TBT and GATT. Second, members who are party to a current MRA should seek to negotiate MRAs with any third members whose technical regulations or CAPs in the relevant sector are arguably equivalent to those of the second party to the current MRA. Otherwise, members with current MRAs risk being compelled to recognize a third member's technical regulations or CAPs, without securing mutual recognition of their own technical regulations or CAPs by that third member.Footnote 101 Third, when two members mutually recognize each other's technical regulations, they should each also secure the other's recognition of domestic CAPs assessing conformity with the other's technical regulations, and equip their domestic conformity assessment bodies to conduct those assessments. This step reduces the risk of a third member's claim of de facto discrimination, by allowing the third member's products to be shipped directly to the either the first or second member for an assessment of conformity with the other's technical regulations. A third member's de facto discrimination claim is unlikely to succeed even without this step, but the risk exists and should be addressed. As the focus of global trade negotiations shifts from the reduction of tariffs to the reduction of technical barriers, WTO members can use these simple steps to maximize the potential of mutual recognition while reducing unwanted risks.