2.1 The interpretation of Article II:2(a) GATT and the delineation between Articles II and III GATT

The case thus revolved around the interpretation of Article II:1(b) GATT in relation to Article II:2 GATT. These paragraphs read:

The Panel and AB reached a fundamentally different reading of Article II:2(a) GATT and its relationship with Article II:1(b) GATT. Although not always clearly or consistently articulated, their approaches seem to fit the structure set out in, respectively, Figures 1 and 2.

Figure 1. Panel's interpretation of Article II:1(a) juncto Article II:2 GATT 1994

Figure 2. Appellate Body's interpretation of Article II:1(b) juncto Article II:2 GATT 1994

The AB agreed with the Panel that Article II:1(b) GATT draws a distinction between ‘ordinary customs duties’ (OCDs) and ‘all other duties or charges of any kind’ that are imposed ‘on or in connection with importation’ (ODCs) but the AB did not offer much guidance on how to interpret and delineate these charges left undefined in Article II:1(b).Footnote ¹¹ At first sight, the AB also seemed to confirm that both types of charges are subject to different sets of disciplines.Footnote ¹² While OCDs cannot be imposed on imported products in excess of those provided for in that Member's Schedule of Concessions, ODCs are not allowed upon imported products in excess of amounts imposed on the date of entry into force of the GATT,Footnote ¹³ as recorded and bound in that Member's Schedule of Concessions.Footnote ¹⁴

However, the Panel and AB fundamentally disagreed on whether the scope of Article II:1(b) could encompass charges on importation referred to under Article II:2.Footnote ¹⁵ In order to grasp their different point of view, we need to recall that the description of the items listed in Article II:2 could on its face be split into a part defining the scope of the charge in question (e.g. charge equivalent to internal tax) and into a part imposing substantive obligations upon that charge (e.g. imposed consistently with the provisions of Article III:2). The Panel focused on the scope and downplayed the substantive obligations imposed under Article II:2 itself (e.g. imposed consistently with III:2), whereas the AB did not distinguish both aspects.Footnote ¹⁶

In the Panel's view, charges falling within the scope of Article II:2 (i.e. charges equivalent to an internal tax, AD/CVDs, fees or other charges commensurate with the cost of the service) cannot fall within the scope of – and hence be disciplined under – Article II:1(b) but are disciplined under other GATT provisions (respectively Articles III, VI, and VIII GATT) (see Figure 1). To reach this conclusion, the Panel had to make the following interpretations.Footnote ¹⁷ First, the Panel found a clear distinction between charges under Article II:1(b) (i.e. ODCs and OCDs) and charges under Article II:2 as the former ‘inherently discriminate against imports’ whereas the latter do not. Second, according to the Panel, the requirement that the charge should be ‘equivalent’ to an internal charge under Article II:2(a) should be distinguished from the obligation of ‘consistency with Article III:2’Footnote ¹⁸ and is determined by whether the border charge and the internal charge have the same function, which does certainly not require that they are similar.Footnote ¹⁹ Such ‘equivalence’, which forms a necessary condition to fall outside the scope of Article II:1(b) by virtue of Article II:2(a), is thus a lower standard than the one imposed under Article III:2 (no taxation ‘in excess’).Footnote ²⁰ Third, the Panel found that the element of ‘consistency with Article III:2’ is not such a necessary condition for the application of Article II:2(b). The reference to Article III:2 GATT in this provision simply makes clear that, ‘in the view of the drafters of Article II:2(a), a border charge on the importation of a product which fulfils the same function as an internal tax on the like domestic product, should be, and is, subject to the provisions of Article III:2’.Footnote ²¹ A border charge equivalent to an internal charge falls outside the scope of Article II:1(b) and should instead be challenged under Article III:2 GATT. The question on whether such border charge on imported products is in ‘excess’ of equivalent internal charges on like domestic products will thus be addressed under Article III:2 GATT. More broadly, Article II:2 is included in Article II ‘to make clear that some charges, even though they may look like ordinary customs duties, or “other duties or charges”, are charges of a different kind and, as such, subject to different disciplines’.Footnote ²² Given that Article II:2 simply clarifies in the Panel's view that those items are subject to other disciplines, the charges listed in Article II:2 are no exceptions to the obligations set out under Article II:1(a).Footnote ²³ The Panel found support for its approach, inter alia, in a 1980 proposal of the Director-General, adopted by the GATT Council, concerning the introduction of a loose-leaf system for the Schedules of Concessions:

Without much explanation, the AB found this statement of ‘limited relevance’, even though it seems to indicate agreement among GATT Contracting Parties that Article II:2 charges are not the sort of charges that could be recorded as ‘other duties or charges’ under Article II:1(b).Footnote ²⁵

Instead, the AB came to a different interpretation of Article II:2 and its relation to Article II:1(b) (see Figure 2). Although not explicitly formulated as such, Article II:2 was approached by the AB as an exception to the obligations set under Article II:1(b). The AB agreed with both parties that if a charge satisfies the conditions of one of the items of Article II:2, it does not result in a violation of Article II:1(b).Footnote ²⁶ Conversely, in case these conditions would not be satisfied, such charges would fall within the scope and disciplines of Article II:1(b).Footnote ²⁷ To arrive at this conclusion, the AB had to override the Panel's specific interpretations spelled out above, all of which were appealed by the United States. First, the AB disagreed that charges under Article II:1(b) could be distinguished from those set out under Article II:2 by relying on the concept of ‘inherently discriminating against imports’. Charges under Article II:1(b) (i.e. OCDs and ODCs) could also be imposed without the rationale to inherently discriminate against imports (e.g. for raising revenue if no domestic production) and there is no textual basis in Article II:1(b) to limit its scope to inherently discriminatory charges.Footnote ²⁸ Conversely, charges under Article II:2(b) and II:2(c) are exclusively imposed on imports and offer therefore no contextual support that charges under Article II:2 are ‘universally non-discriminatory’ in respect of imports.Footnote ²⁹ Second, the AB considered that the obligation of ‘consistency with Article III:2’ referred to in Article II:2 must not be distinguished from but ‘read together with, and imparts meaning to’ the requirement that a charge and internal tax be ‘equivalent’, which calls for ‘a comparative assessment that is both qualitative and quantitative in nature’.Footnote ³⁰ In particular, ‘consistency with Article III:2’ forms ‘an integral part of the assessment’ of ‘equivalence’.Footnote ³¹ Third, as a logical inference, the AB also rejected the Panel's view that ‘consistency with Article III:2’ is not a necessary condition for the application of Article II:2(a).Footnote ³²

Accordingly, contrary to the Panel, the AB thus seems to draw a firm line between ‘border charges’, which are disciplined under Article II, and ‘internal charges’, which are disciplined under Article III.Footnote ³³ Applied to the facts of the case, the AB observed that ‘the Panel and the participants also agree that the Additional Duty and Extra-Additional Duty are border charges subject to the terms of Article II, and that they are not disciplined by the provisions of Article III as “internal taxes”’.Footnote ³⁴ Nonetheless, the AB's statement seems incorrect insofar that it contends that this view was also shared by the Panel. The AB seems to overlook that, according to the Panel, charges equivalent to an internal charge – and the United States failed to demonstrate according to the Panel that the (Extra-) Additional charges do not constitute such charges – are disciplined under Article III:2 instead of Article II. Under both interpretations, the consistency of the border charge with Article III:2 should thus ultimately be assessed, but according to the AB this should be done under Article II:2(a).Footnote ³⁵ Another difference between the two approaches is that the AB's reading does not mandate a Member challenging a border charge within the meaning of Article II:2(a) to formulate a separate claim under Article III:2.Footnote ³⁶ A border charge could (and should) be challenged under Article II GATT and not under Article III:2. This seems sensible given that a complaining party might, at least in theory, be unaware that a border charge to which its product is subjected is alleged to be counterbalanced by an internal charge.Footnote ³⁷

Nonetheless, the qualification of whether a charge is a ‘border charge’ subject to Article II or an ‘internal charge’ subject to III:2 is not always straightforward. As the AB acknowledged, Ad note to Article III might also come into play. This Ad note to Article III stipulates that Article III is applicable to any internal charge that applies to both domestic and imported products, but which is ‘collected or enforced’ in respect of the imported product at the time of importation. The delineation between such an ‘internal charge’ to which Article III:2 applies and a ‘border charge’ to which Article II:2(a) applies has to be assessed ‘in the light of the characteristics of the measure and the circumstances of the case’.Footnote ³⁸ Apparently, if a charge is imposed on domestic and imported goods (but for imported goods this charge is collected or enforced at the border), such charge is, pursuant to Ad note, deemed an ‘internal charge’ subject to Article III. On the other hand, if equivalent but different charges are imposed on imported and domestic products and the charge on imported products is imposed on importation, Article II:1(b) juncto II:2(a) is applicable for challenging such ‘border charge’.Footnote ³⁹

Except for the clarification under which provision a claim should be formulated, what is the relevance of this disagreement between the Panel and AB? Indeed, the disciplines of Article III:2 would apply anyway: either directly as ‘internal charge’ under Article III:2, or indirectly, as ‘border charges’ to which Article II:2(a) applies. Procedurally, the allocation of the burden of proof might be different, as will be discussed in the next section. But also on a substantive level, the AB seems to give relevance to its holding that a border charge to which article II:2(a) applies is not disciplined directly under Article III but under Article II. After all, the AB concluded that the (Extra-) Additional Duties could not be justified to the extent they result in the imposition of charges on imports in excess of internal charges on domestic products and that this would render these duties ‘inconsistent with Article II:1(b) to the extent that (they result) in the imposition of duties (on the product in question) in excess of those set forth in India's Schedule of Concessions’.Footnote ⁴⁰ Such a border charge inconsistent with Article II:2(a) (because applied inconsistently with Article III:2) thus falls within the scope of Article II:1(b), but the AB left open whether it is disciplined under the first sentence (referring to ‘ordinary customs duties’, OCDs) or the second sentence (referring to ‘other duties and charges’, ODCs) of Article II:1(b). Three possible but questionable explanations on the meaning of the AB's statement could be advanced (Figure 3).

Figure 3. Appellate Body's interpretation of Article II:1(b) juncto Article II:2 GATT 1994 (elaborated)

First, the AB might have assumed that the (Extra-) Additional Duties fall within the scope of ‘ordinary customs duties’ (OCDs) but this would, at least, have required some substantive discussion on the definition of OCDs, as it seems more likely to bring such charges under the residual category of ODCs (1).Footnote ⁴¹ Second, the AB might have meant that such charges are a type of ‘other duty or charge’ (ODCs) under the second sentence of Article II:1(b) and are allowed insofar as their nature and level are recorded in India's Schedule of Concessions (2).Footnote ⁴² It might be doubted that the AB had this in mind because of the broader wording employed by the AB in its statementFootnote ⁴³ and because the AB had previously observed that India had no recorded ODCs for the products in question.Footnote ⁴⁴ Third, Piérola (Footnote ²⁰⁰⁹) suggests that the AB might have implied that these charges are ‘other duties or charges’ (ODCs) but that such ODCs could be offered insofar as they do not lead to border charges in excess of the bound level for ordinary customs duties (OCDs) (3).Footnote ⁴⁵ As Piérola (Footnote ²⁰⁰⁹) rightly stresses, this interpretation would be problematic as it is generally understood that such ODCs are only valid if their nature and level was recorded. Yet, again, it seems doubtful that the AB had this problematic interpretation in mind as it considered that OCDs and ODCs are ‘two sets of charges … described and disciplined in separate sentences of article II:1(b) and, may, by their terms, not pertain to the same event of importation’.Footnote ⁴⁶ In sum, the AB's statement simply seems incoherent with other aspects of its report.

In a more recent case (China–Auto Parts), the AB unequivocally confirmed that ODCs ‘are permitted only when their nature and level are recorded in a Member's Schedule’.Footnote ⁴⁷ This would mean that charges referred to under Article II:2 but inconsistent with the stipulated obligations would be inconsistent with Article II:1(b) insofar as not recorded in a Member's Schedule (2)Footnote ⁴⁸, unless, of course, they would be considered OCDs, which seems not very likely. Nonetheless, the Panel in United States–Zeroing (Japan) (Article 21.5) seems to have endorsed the interpretation that such ‘Article II:2 inconsistent border charges’ are not inconsistent with Article II:1(b) if they do not result in duties exceeding the bound level for ordinary customs duties (3).Footnote ⁴⁹ After all, the Panel found a violation of Article II:1(b) because the United States had offered anti-dumping duties for the product in question in excess of its bound level and the excess amount could not be attributed to anti-dumping duties applied consistently with Article VI and could thus not benefit from the ‘safe harbour’ of Article II:2(b).Footnote ⁵⁰

In conclusion, the AB in India–Additional Import Duties clarified that charges inconsistent with Article II:2 fall within the scope of Article II:1(b). However, the substantive obligation imposed on these charges under Article II:1(b) is not yet clearly sorted out in the case law. In our opinion, such charges should fall within the scope of ODCs and therefore only be deemed consistent with Article II:1(b) in the improbable case that they are explicitly recorded (2).Footnote ⁵¹

The AB's interpretation thus underlined the distinction between ‘border charges’, covered under Article II, and ‘internal charges’, covered under Article III. All ‘border charges’, even if they are imposed to counterbalance ‘internal charges’ (Article II:2(a)), fall within the scope of Article II. The delineation between ‘border charges’ and ‘internal charges’ was further clarified in the China–Auto Parts case, where the AB observed that ‘whether a specific charge falls under Article II:1(b) or Article III:2 of the GATT 1994 must be made in the light of the characteristics of the measure and the circumstances of the case’.Footnote ⁵² In this respect, the AB was of the view that:

Hence, the obligation to pay an ‘internal charge’, covered under III:2, is triggered by an internal factor. For example, the obligation to pay internal charges referred to in Ad note to Article III accrues because of an internal factor and is merely collected for imported products at the border.Footnote ⁵⁴ Conversely, the obligation to pay charges within the scope of Article II – so-called ‘border charges’ – accrues because of the act of importation. This fits with the Panel's observation in India–Additional Import Duties that also charges under Article II:2(a) are triggered by importation, even though ‘the internal taxes to which the relevant charges are equivalent link liability to events other than importation’.Footnote ⁵⁵

2.2 Burden of proof

The allocation of the burden of proof was a deciding factor for the outcome of this case. Neither party offered evidence on the operation of the internal charges that would allow a determination of whether the (Extra-) Additional Duties effectively resulted in charges on imported products in excess of charges imposed on like domestic products. Obviously, for a claim under Article III:2, the burden would be on the complaining party to make a prima facie case of excess taxation and thus to advance details on the operation of the internal charges. However, what if ‘consistency with Article III:2’ is not addressed under Article III:2 itself but in an indirect way under Article II:2(a)?

Importantly, the United States agreed that the ultimate burden on whether the conditions under Article II:2 are met (i.e. consistency with Article III:2) rests on the shoulders of the complainant given that Article II:2(a), though an exception, should not be considered an affirmative defence.Footnote ⁵⁶ But in the United States's view, it is up to the respondent to articulate a prima facie case that the charges fall within the scope and meet the conditions of Article II:2(a) in case the complainant has made a prima facie case under Article II:1(b). However, the Panel disagreed and held that the United States also had to make a prima facie case that the charges do not fall within the scope of Article II:2(a) to support its claim under Article II:1(b).Footnote ⁵⁷ This position seemed to be inspired by the Panel's conclusion that Article II:2 does not set out exceptions to the positive obligations under Article II:1(b). Although the AB adopted a different approach on Article II:2(a) juncto Article II:1(b), it generally agreed with the Panel's conclusion on the allocation of the burden of proof but seemed to emphasize, somewhat stronger, India's responsibility in this respect. The AB observed that Article II:1(b) and II:2(a) are ‘closely inter-related provisions’ and that, in this case, ‘the potential for application of Article II:2(a) is clear from the face of the challenged measures’.Footnote ⁵⁸ In light of these circumstances, in order to establish a prima facie case of a violation of Article II:1(b), the United States had to also present arguments and evidence that the charges are not justified under Article II:2(a).Footnote ⁵⁹ On the other hand, because India relied on Article II:2(a), it was also ‘required to adduce arguments and evidence in support of that assertion’.Footnote ⁶⁰ Yet, the AB did not give much guidance on when the complaining party would have met such a prima facie burden under Article II:2(a) as this ‘will to some extent vary, depending upon the particular substance of the challenged measure and the extent to which a relationship between the border charge and the corresponding internal taxes is identifiable’.Footnote ⁶¹ Regarding this particular case, the AB simply concluded that ‘both parties had a responsibility … to adduce relevant evidence at their disposal, both with respect to Article II:1(b) and Article II(a)’.Footnote ⁶² Referring to India's refusal to provide details on the charges upon the Panel's request, the AB reiterated its jurisprudence that ‘refusal will be one of the relevant facts of record, and indeed an important fact, to be taken into account in determining the appropriate inference to be drawn’.Footnote ⁶³ The AB even stressed that these ‘were particularly important pieces of evidence at India's disposal that should have been provided to the Panel’ and seemed to hint that the Panel should have attached more weight to India's refusal.Footnote ⁶⁴

The AB summarized its general approach to the burden of proof in WTO dispute-settlement procedures, but failed to explain why this approach would, in this case, support its conclusion to put the burden of making a prima facie case under Article II:2(a) on the complainant. Instead, the AB grounded this interpretation on a new criterion (‘closely inter-relatedness’ of two provisions),Footnote ⁶⁵ while it seems that the same conclusion could have been reached on the basis of more solid criteria developed in previous cases.

As the AB reiterated, the general principle for allocating the burden of proof is that it ‘rests upon the party, whether complaining or defending, who asserts the affirmative of a particular claim or defence’.Footnote ⁶⁶ Consequently, the burden of proving a violation rests on the complainant's shoulders, whereas the burden of proving an exception is put on the defendant. Nonetheless, the AB has attempted in previous case law to distinguish ‘exceptions’, which establish an exception to a rule, from provisions that exclude the application of other provisions (so-called ‘excluding provisions’) (Grando, Footnote ²⁰⁰⁶). With respect to such ‘excluding provisions’, the complainant has the burden of proving that ‘the defendant does not fall under the situation or has not complied with the requirements of a provision that excludes the application of the general rule’ (Grando, Footnote ²⁰⁰⁶). The AB's understanding of Article II:2(a) seems to fit surprisingly well in this category of ‘excluding provisions’.Footnote ⁶⁷ Paraphrasing AB statements from previous cases, items listed in Article II:2 seem to be ‘positive rules establishing obligations in themselves’ and not affirmative defenses.Footnote ⁶⁸ If a Member complies with Article II:2(a), Article II:1(b) ‘simply does not apply’ and, conversely, if a Member does not comply with those obligations set out in Article II:2(a), Article II:1(b) applies.Footnote ⁶⁹ Indeed, this conforms to the AB's reading of Article II:2(a) juncto Article II:1(b) as sketched out above (see Figure 2): consistency with article III:2 should be assessed under the analysis of Article II:2(a) itself and, in case these obligations are not fulfilled, the charges fall within the scope of – and are disciplined by – Article II:1(b).Footnote ⁷⁰

Hence, the AB could have based its interpretation on previous case law instead of opting for a new criterion.Footnote ⁷¹ The choice for a new approach without linking it to previous case law adds a new layer of legal uncertainty to a field that urgently needs some coherence. To illustrate this point: Piérola (Footnote ²⁰⁰⁹) wonders whether this new approach could be extrapolated to other provisions such as Article XX GATT in case the challenged measures call for the application of this provision. It is, however, well-settled case law that Article XX spells out affirmative defences, which put the burden on the defendant for formulating a prima facie case.Footnote ⁷²

3.1 WTO rules and efficient trade and domestic policies

The main goal of the GATT/WTO is to facilitate the exchange of reciprocal reductions in trade barriers. This goal is often perceived to clash with other policy interests of the Members. The fear is that governments may feel constrained from unilaterally raising tariffs because of GATT obligations, and may instead choose to lower domestic standards to improve the competitive position of their domestic firms. In particular, many labor and environmental groups claim that competitive pressures will lead either to a ‘regulatory chill’, with governments resisting the use of tougher regulations, or to a ‘race to the bottom’, with governments setting even less restrictive policies.

How can we reconcile the objectives of the WTO with the national objective of national governments? In an influential paper, Bagwell and Staiger (2001) argue that the answer to this question can be found in existing GATT rules, which are aimed at securing ‘property rights over negotiated market access commitments’.Footnote ⁷³ They consider a simple general equilibrium framework in which two countries trade two goods and governments make decisions over their trade policies (e.g. tariffs) and their domestic standards (e.g. labor and environmental standards) in pursuit of their own national objectives. The objectives of each government can be represented as a general function of its local prices and terms of trade, which are affected by both trade and domestic policies. Domestic policy instruments are imperfect substitutes for tariffs, implying that the same level of market access can be achieved by different ‘policy mixes’ (e.g. a low tariff and a weak labor standard, or a stricter labor standard and a higher tariff).

In this setting, governments' domestic-policy autonomy can interfere with the maintenance of ‘reciprocity’ – the balance of negotiated market-access commitments. In particular, a country could commit to reduce its tariff on a particular product and subsequently impose internal taxes on the sale of the product in a manner that favors domestic over foreign producers. Bagwell and Staiger (2001) argue that GATT's rules on ‘nonviolation complaints’ can be used to avoid the erosion of tariff commitments and a ‘race to the bottom’ in domestic regulations. As stated in Art. XXIII.1(b) of GATT:

Nonviolation complaints are based on the ‘right to redress’ and on the concept of ‘nullification or impairment’ of reasonable expectations of a benefit accruing from a negotiated concession and agreement. Under a successful nonviolation complaint, the complaining country is entitled to a ‘rebalancing’ of market-access commitments, whereby either its trading partner finds a way to offer compensation for the trade effects of its domestic-policy change (e.g. by lowering its trade barriers) or the complaining country is permitted to withdraw an equivalent market-access concession of its own.

Consider, for example, a government that is facing pressure from domestic producers to offer import relief in an industry where it has agreed, as a result of WTO negotiations, to hold tariffs low. The prospects of nonviolation constraints can deter this government from offering unilateral import relief to its producers by lowering domestic standards.

In our view, Bagwell and Staiger's (2001) view of the role of nonviolation complaints may be overoptimistic. As the authors themselves admit, nonviolation complaints have proven difficult to carry out in practice. Since the creation of the GATT, very few cases have centered on such complaints, and none of these explicitly involved labor or environmental standards. Moreover, in the very few disputes in which complainants have resorted to the idea of ‘nullification or impairment’, they have not succeeded in fulfilling the burden of proof. This is due to the difficulty for adjudicators in determining what the negotiating parties could have reasonably expected when they signed the agreement, as well as to the difficulty of assessing the trade effects of given changes in domestic standards (see also Horn, Footnote ²⁰⁰⁶).

We also somewhat disagree with Bagwell and Staiger (2001) on a more important point, which is directly related to the dispute between India and the United States considered here. In their paper, they argue:

Thus, in their view, efficient combinations of trade and domestic policies cannot be implemented if a government enters tariff negotiations with domestic policies that discourage access to its markets relative to efficient domestic policy.

We believe that this negative view may apply to domestic regulations that take the forms of standards, but not to domestic taxes. This is because in the latter case Articles II:2 and III:2, and their interpretation by the AB in its ruling on India–Additional Import Duties, can provide the necessary flexibility to allow Members to strengthen their domestic policies without affecting the balance of market-access concessions. For example, manufacturers in an importing country faced with the imposition of higher energy taxes, may argue that the resulting cost increase reduces their competitiveness vis-à-vis imported goods. In such circumstances, governments could offset this competitive disadvantage by using a corresponding border tax.

Thus, global efficiency can be achieved if countries negotiate tariff reductions – to internalize terms-of-trade externalities – and then use domestic taxes and equivalent border charges – to internalize negative domestic externalities.

Three important considerations are in order. First, in a standard two-country trade model like the one described by Bagwell and Staiger (2001), one of the two countries may wish to increase taxation on the domestic producers, as a result of an increase in the extent of (or in the awareness of) the negative environmental externalities associated with domestic-production activities. This more stringent domestic policy could be combined with the use of border taxation, so as to keep terms of trade unchanged. Notice, however, that the trading partner would be negatively affected by these policy changes, since they lead to a reduction in trade volumes. To keep the welfare of the trading partner unchanged, the border tax adjustment would need to be such that the terms of trade actually improve for the trading partner.

Second, Bagwell and Staiger (2001) consider a situation in which governments use standards rather than taxes to deal with the local externalities. From a legal point of view, it is not clear whether governments would be able to use border taxation to countervail the market-access effects of domestic standards. According to the AB's ruling on India–Additional Import Duties, Article II:2(a) allows border charges which are equivalent (and thus consistent with Article III:2) to domestic charges. The AB said that Members could, by imposing border charges, counterbalance domestic charges. But the AB did not say that Members are allowed to impose border charges so as to counterbalance non-fiscal charges on domestic products, which are disciplined under Article III:4. The text of Article II:2(a) might, on its face, not allow border charges on imported products to counterbalance domestic standards, since it does not read: ‘charges on importation equivalent to domestic charges or other domestic regulations consistent with Article III’.Footnote ⁷⁴ Instead, Article II:2(a) refers only to ‘internal tax’ and to ‘consistency with Article III:2’ (not III:4). Thus, the exception under Article II:2(a) seems to require that there exist a domestic charge (e.g. not product standard) that is counterbalanced by a border charge.Footnote ⁷⁵

Our analysis above may thus only apply to when governments used taxes – rather than standards – to regulate negative domestic externalities. In this case, they would be able to optimally adjust their domestic policies to their national objectives, without distorting the balance of negotiated market access.

A third important consideration is that international trade negotiations alone – combined with domestic and border taxation – can only yield globally efficient outcomes in the absence of nonpecuniary externalities across countries. This is because in this case countries are not affected by each other's domestic policies directly, but only through the trade effects of such choices. If instead there are nonpecuniary externalities across countries, as in the case of transboundary pollution problems, global efficiency would clearly require coordinated policy efforts on both trade and environment (see also the discussion at the end of Section 3.3).

In conclusion, GATT rules of border tax adjustments can help to achieve efficiency of trade and domestic policies. Our analysis suggests that Articles II and III can help to counter fears that trade pressures associated with a country's WTO market-access commitments can cause a ‘regulatory chill’ or a ‘race to the bottom’ in domestic regulations.

3.2 Problems with border tax adjustments

The rationale for tax border adjustments is simple: they are a way to maintain the competitiveness of domestic industries, when responding to stricter domestic regulations. However, as discussed below, their application is often complex and can lead to abuses. First, border tax adjustments need to comply with the National Treatment obligation of Article III, which requires imported products to be treated no less favorably than ‘like’ domestic products. While there is no legal definition for ‘likeliness’, the Interpretative Note to Article III reads

The interpretation of the concepts of ‘like’ and ‘directly competitive or substitutable’ products is far from obvious, as it refers to the extent of demand substitutability, which needs to be assessed based on econometric or other evidence. Another limitation of GATT rules on National Treatment is that they do not put any discipline on domestic instruments in cases where there is no ‘like’ or ‘directly competitive or substitutable’ domestic product.Footnote ⁷⁶

A second problem with the application of border taxes arises when the domestic excises tax is applied to an intermediate good, but it is the final good that is imported. For example, Poterba and Rotemberg (Footnote ¹⁹⁹⁵) stress administrative problems in the tax treatment of imports of final goods produced using intermediate goods that are subject to environmental taxes. In this case, border taxes are difficult to implement, since they require arbitrary assignments of intermediate-good inputs to final goods, for example on the basis of relative output weight or value. However, not taxing such imports would place domestic producers of final goods at a cost disadvantage and may encourage offshore production of these final goods.

In the case of India–Additional Import Duties, the United States argued that some of its products that are charged Extra-Additional Duty are subsequently used in India as inputs in the manufacturing of other products and are subject to state VAT, state sales tax, Central Sales Tax, and/or ‘other local taxes or charges’ in the same way as like domestic products.Footnote ⁷⁷ As noted by the Panel, in the absence of a credit for the Extra-Additional Duty, these types of imports would be subject to duties ‘in excess’ of the internal taxes on like domestic products.

A third complication in the use of border taxes arises for countries characterized by a decentralized fiscal system, where charges tend to vary across constituencies and it is thus difficult to establish the correct rate for a common border tax. For example, under the Indian Constitution, excise duties on alcoholic beverages are established and collected by the individual states, not the central government, and the different Indian states are permitted to levy such excise duties at varying rates. Individual states are empowered to levy excise duties on alcoholic liquor ‘manufactured or produced’ in the relevant state. When different states levy varying rates of excise duty, it is by definition impossible for the government to fix a single rate for the border tax that is ‘equal to the excise duty’.

As discussed in the previous section, one of the controversial issues with this dispute was the fact that India did not provide information on the various rates applied by its states and the methodology by which the central government averaged them in order to establish the corresponding border charge. Had this information been available, it is still not clear what kind of methodology to compute border taxes would have been considered consistent with the National Treatment obligation of Article III.

3.3 Implications for the debate on carbon border taxes

Parties to the UNFCCC and the Kyoto Protocol are currently in talks designed to help shape a climate-change regime to follow the Protocol's first commitment period, which ends in 2012. At this point, the nature of that regime and the commitments it will entail are uncertain, but the emissions reductions needed will be significant.

In response to that challenge, a number of countries are pursuing or considering strong domestic action to address climate change. They are doing this either in anticipation of future regime obligations, as part of their obligations under the current treaties, or out of a desire to address the challenge of climate change irrespective of what might develop at the international level. In those countries, one of the key obstacles to such action is the fear that it may put their domestic industries at a disadvantage relative to producers in countries that do not take similarly strong action.

One policy option that has been repeatedly proposed to deal with such challenges is border carbon taxes, which are seen as a trade measure that would level the playing field between domestic producers facing costly climate-change measures and foreign producers facing very few.

The recent debate on border carbon taxes has been particularly heated in various countries. In the United States, two bills were proposed before the Senate,Footnote ⁷⁸ both of which involve a cap-and-trade scheme and both of which foresee border taxes as part of the regime. Although they eventually failed to pass the Senate, these bills will likely inform whatever future climate-change legislation is passed. In Europe, similar proposals have been circulated. The EC-mandated High Level Group on Competitiveness, Energy and Environmental Policies has proposed the use of border carbon adjustments in its second report in 2006. Various politicians support the idea of imposing a carbon tax on goods imported from countries with no emission curbs under the Kyoto regime as part of the so-called EU's ‘carbon equalization system’.Footnote ⁷⁹

While such measures increase the political feasibility of national climate-change legislation, they pose a serious threat to the international trading system and potentially violate international trade law under the WTO.

The India–Additional Import Duties case cannot really help us to assess whether or not the proposed carbon border taxes would breach WTO obligations. As discussed above, the AB ruling on this dispute establishes that, in the case of goods subject to indirect taxes (e.g. sales taxes and value-added taxes), border taxes are allowed under Article II:2(a) as a way to level the playing field between taxed domestic industries and untaxed foreign competitors. The extent to which they could also be applied to energy inputs of products, however, is unclear.Footnote ⁸⁰

To be compatible with GATT/WTO rules, carbon border taxes should not discriminate between domestic producers and foreign producers of like products – both should be treated similarly according to the National Treatment principle. Also, they should not discriminate between ‘like’ products based on the country of production, in line with the Most Favored Nation principle.

In the case of environmental taxation, goods that are ‘like’ from the point of view of their use may actually be considered very different because of the technology used to produce them: is a tonne of cement produced with solar energy ‘like’ a tonne of cement produced using coal?

While, as mentioned in Section 3.2 above, there is no legal definition for ‘likeliness’, the AB has ruled that likeness ‘is, fundamentally, a determination about the nature and extent of a competitiveness relationship between and among products’.Footnote ⁸¹ Import restrictions on the basis of non-product-related process and production methods are generally not permitted. This would seem to mean that steel is steel, no matter how it is produced. Going further, likeness has been defined as being determined by four criteria: (i) the (physical) properties, nature, and quality of the products; (ii) the end-uses of the products; (iii) consumers' perceptions and behavior in respect of the products; and (iv) the tariff classification of the products.Footnote ⁸² It might be argued that consumers perceive ‘dirty’ steel as different from ‘green’ steel, but this would probably be something of a legal long shot, but a WTO dispute panel would probably consider the two products to be ‘like’.Footnote ⁸³

Even if carbon border taxes were considered to be compatible with GATT/WTO rules, there are doubts about the effectiveness of such schemes. This would depend on whether they cover only basic materials (such as raw aluminum) or also cover manufactured products made from those materials (such as aluminum-frame bicycles). As described in the previous section, a broader scheme will be particularly difficult to manage, but a scheme that is more narrowly cast may have unintended adverse impacts. Specifically, it will raise the price of aluminum as an input good to domestic manufacturers of, say, bicycles, but it will not levy any charges on imported bicycles. Such a scheme would protect the aluminum sector from competitiveness impacts, but not the sectors that add value to aluminum.

A border carbon tax scheme could also be evaluated on the extent to which it might exert pressure on some countries to adopt stricter policies, or to take on tough treaty obligations. This potential will of course vary from country to country and sector to sector. In those cases where the percentage of a given good exported to the implementing country is particularly small, imposing a carbon tax will have little or no policy impact on the exporter.

Finally, as already mentioned in Section 3.1, GATT/WTO rules on tariff bindings and border tax adjustments can only help to achieve global efficiency when domestic policies affect foreign countries only indirectly, through their effects on market access. In the case of carbon emissions and other transboundary negative externalities, countries are affected by each other's domestic policies directly. In this case, global efficiency could only be achieved through trade and environmental negotiations.Footnote ⁸⁴