Introduction: ‘defensive’ policies and the notions of an economic and an environmental level playing field
The devastating effects of climate change are before our eyes. The bottom-up system of Nationally Determined Contributions (NDCs) laid out in the Paris Agreement is proving insufficient to meet the Agreement's goal of holding the increase in the global average temperature to below 2°C above pre-industrial levels.Footnote 1 As demonstrated by the recent proceedings of the COP26, the parties are still struggling to agree on an adequate roadmap in multilateral fora.
In this very difficult context, some parties are contemplating recourse to unilateral carbon border measures. These regulatory arrangements are different in nature from other environmental law measures giving rise to extra-territorial effects. In an increasingly interdependent and interconnected world, national or regional law is bound to have some degree of extra-territorial impact. The extra-territorial reach of EU law, in particular, has been the object of close scrutiny and detailed analyses over the past few years.Footnote 2 Carbon border measures, however, are part of a ‘new’ generation of ‘defensive’ policies, involving the creation of an environmental or an economic level playing field.
‘Defensive’ laws and policies are enacted with a view to safeguarding the environmental protection agenda of the regulating jurisdiction. Their aim is to ensure that the divergencies between external and internal (domestic) environmental standards will not undermine the regulating jurisdiction and jeopardise its attempts to pursue specific environmental goals. The rise of defensive policies and the development of the notion of a ‘level playing field’ are relatively novel phenomena, in an evolving scenario where globalisation and transnational trade flows produce far-reaching effects within national boundaries. In this perspective, defensive policies are an ‘imperfect answer’ to the externalities of globalisation and the trading-down dilemma.Footnote 3
Crucially, environmental and economic factors are intertwined in defensive policies. Compliance with environmental protection standards is associated with regulatory burdens and economic costs; the regulating jurisdiction's environmental protection agenda is threatened by external (transnational) divergencies in levels of protection because of the economic costs which are borne by domestic actors. First, transnational divergencies in the stringency of environmental regulations may undermine the environmental agenda of ‘virtuous’ jurisdictions in so far as firms may relocate to countries with more lenient standards. This has economic as well as environmental effects. Secondly, diverging regulatory compliance costs will distort competition between ‘greener’ (generally more expensive) and more polluting (generally cheaper) products, strengthening the geo-political leverage of jurisdictions with more lenient standards and hindering the development of environmentally sustainable strategies at the transnational level.
For this reason, the environmental component of defensive policies is not self-standing; the declared rationale for the enactment of these policies is the need to establish a level playing field. However, the scope of the relevant obligations may be narrower or broader. This will result from the decision to embrace a narrow focus on an economic (stricto sensu) level playing field, or a broader environmental level playing field perspective.Footnote 4 The notions of an economic and an environmental level playing field are not mere formal categories. As a matter of regulatory design, they pursue different goals, are characterised by different arrangements, and have a different scope of application.
An economic level playing field rationale is characterised by a focus on economic equivalence. This may be established by taking into consideration the economic costs borne by market actors to comply with specific environmental regulations, or the distortions of competition ensuing from the pursuit of divergent levels of environmental protection in different jurisdictions. As regards the available remedies, an economic level playing field will be maintained or restored as long as market actors in different jurisdictions bear the same exact economic costs, or as long as state parties may take ad hoc measures to redress distortions of competition in transnational trade and investment. Further, the scope of application of the economic level playing field is also demarcated by the distortions of competition which may arise due to divergencies in environmental standards or environmental levels of protection. Any products or any regulatory areas which bear no relevance from an economic perspective are thus excluded from the scope of the level playing field obligations.
Conversely, an environmental level playing field perspective is grounded on an assessment of environmental equivalence; this may be measured against specific environmental commitments, or the achievement of specific levels of environmental protection across jurisdictions. Turning to the question of the available remedies, these should not simply restore the equality of competitive opportunities; rather than ensuring that domestic and imported products ‘bear’ the same exact economic costs, or rather than ‘compensating’ a party for distortions of competition, they should create incentives for non-compliant parties to restore the environmental level playing field. Undeniably, the identification and calibration of these ‘punitive’ remedies may prove challenging. Finally, the scope of application of an environmental level playing field is more encompassing than the one of a stricto sensu economic level playing field; the establishment of an environmental level playing field puts the environmental dimension front and centre stage.
Both categories of measures will result in positive environmental spill-over effects. As already explained, economic and environmental factors are structurally intertwined under defensive policies; for this reason, measures informed by an economic level playing field rationale will also create incentives for regulators to enact more stringent standards or for firms to invest in ‘greener’ technologies. Nonetheless, an economic level playing field rationale suffers from specific limitations in terms of environmental integrity.
First, a determination of economic equivalence will only capture the economic implications of compliance with environmental standards. As demonstrated in the next section, environmental commitments and the effectiveness of environmental policies are beyond the radar of an economic level playing field rationale.Footnote 5 Secondly, establishing an economic level playing field and redressing distortions of competition is rather unambitious, from an environmental protection perspective. Comparing a focus on the economic costs ‘borne’ by domestic and imported products and recourse to economic remedies with an outright product ban imposed on environmental protection grounds is sufficient to shed light on the reductionist approach of an economic level playing field perspective.Footnote 6 Thirdly, measures informed by an economic level playing field rationale are more likely to come under challenge as a form of ‘green protectionism’.Footnote 7 Fourthly, in terms of regulatory rationale, measures informed by an economic level playing field perspective should not differentiate the treatment of developed, developing or least developed countries. The aim of these measures is to redress the distortions of competition associated with compliance with divergent environmental standards; whether specific countries should take the lead in tackling environmental challenges is thus irrelevant, from a (policy-neutral) economic perspective.Footnote 8 Finally, the scope of application of an economic level playing field is considerably narrower than the one of an environmental level playing field.
Overall, levelling the economic playing field is a ‘minimalist’ strategy. Defensive policies informed by an economic level playing field rationale cannot fully capture the environmental component and the effectiveness of different national measures. Further, these policies will only promote specific environmental standards or incentivise specific environmental practices indirectly. The opposite is true of measures which seek to establish an environmental level playing field. Nonetheless, as this paper illustrates, an environmental level playing field perspective is associated with several problematic aspects.
The next section analyses the European Commission's proposal for a carbon border adjustment mechanism (CBAM)Footnote 9 and a recent US proposal for the establishment of an import charge on carbon-intensive products (BCA or Coons-Peters Bill)Footnote 10 against the backdrop of the conceptual framework developed in this section. It illustrates that the two measures are broadly informed by an economic and an environmental level playing field rationale. The third section of the paper explores the practical obstacles connected to the implementation of the BCA: the establishment of environmental equivalence, the identification of adequate remedies and the demarcation of the scope of application of an environmental level playing field are all fraught with difficulties. The CBAM arrangements, on the other hand, are more feasible in practical terms. The fourth section turns to an in-depth examination of the structural constraints posed by WTO law. Yet again, the analysis illustrates that the proposed BCA is highly unlikely to be compatible with WTO law; the CBAM, by contrast, is overall more likely to be WTO law compliant. Taking stock of these findings, the paper concludes that practical obstacles and structural legal constraints push towards a narrower economic level playing field rationale, as a matter of regulatory design.
1. The regulatory design of the EU CBAM and the US BCA
The Commission's proposal for a Regulation establishing a CBAM was published on 14 July 2021, as part of the ‘Fit for 55’ package. Only a few days later, the Coons-Peters Bill was introduced in the Senate of the United States. Both measures purport to avoid leakage in carbon-intensive sectors; in other words, they aim to prevent national firms who operate in these sectors from relocating to countries with more lenient greenhouse gas (GHG) emission reduction policies.Footnote 11 Both measures are unequivocal examples of defensive policies. At first sight, and in so far as they focus on economic remedies, both the CBAM and the US BCA may be understood as measures informed by a stricto sensu economic level playing field rationale. Yet, a close analysis of the regulatory arrangements shows different nuances.
As extensively documented throughout the years, the EU has had recourse to price-based mechanisms in the regulation of carbon-intensive sectors; it has done so by setting in place a cap-and-trade system, in the form of an Emission Trading Scheme (ETS) involving the surrendering of allowances for GHG emissions.Footnote 12 The EU CBAM envisages a ‘quasi-extension’ of the ETS mechanism to importers. The proposal is informed by an economic level playing field rationale.
Starting from the question of remedies, the CBAM seeks to ensure that third country producers bear the same exact economic costs faced by European producers due to the operation of the ETS cap-and-trade scheme. This ‘re-allocation’ of economic burdens occurs via the calculation of the GHG emissions embedded in imported products, and the obligation for importers to purchase and surrender CBAM certificates; the price of the certificates mirrors the one of ETS allowances. This redresses the distortions of competition ensuing from the absence of carbon pricing in other jurisdictions.
More specifically, the Commission's proposal provides that ‘authorised declarants’ importing goods covered by the CBAMFootnote 13 shall submit a CBAM declaration on a yearly basis. This shall include the verified GHG emissions embedded in their products, and the (corresponding) number of CBAM certificates to be purchased.Footnote 14 The CBAM certificates are part of a separate ‘pool’; however, Article 21(1) stipulates that their price shall correspond to the average (closing) price of auctioned EU ETS allowances on a weekly basis. The number of CBAM certificates to be surrendered will also be adjusted to reflect the extent to which ETS allowances will keep on being allocated free of charge to EU firms in sectors exposed to carbon leakage.Footnote 15
As regards the calculation of GHG emissions, these shall be determined on the basis of the actual embedded emissions of the products, following the methods laid out in Annex III, points 2 and 3. Only direct emissions from the production processes of goods will be accounted for; indirect emissions from the production of electricity, heating and cooling are not covered in the Commission's proposal.Footnote 16 Where it proves impossible for the authorised declarant to communicate monitoring data on the actual embedded emissions, Article 7(2) provides for the application of default values; these shall refer to the average emission intensity of each exporting country, or the average emission intensity of the 10% worst performing EU installations. A more complex system applies when electricity is being imported.Footnote 17
An analysis of equivalence also demonstrates that the CBAM is informed by an economic level playing field rationale. First, products originating from countries whose emission trading schemes are fully linked to the EU ETS will be exempted.Footnote 18 Secondly, an authorised declarant may claim a reduction in the number of CBAM certificates to be purchased corresponding to any carbon price already paid in the country of origin of the product.Footnote 19 Equivalence is thus established in purely economic terms. Price-based policies enforced in third countries are taken into account, and any ‘explicit’ carbon price paid in the country of origin of the imported products will be waived; by contrast, the proposal makes no reference to the levels of environmental protection that third countries may have achieved via non-price-based GHG emission reduction policies, and the stringency of their climate change mitigation standards.Footnote 20 This reveals a direct focus on the existence or absence of carbon pricing in different jurisdictions, and on the ensuing distortions of competitive opportunities between products. The achievement of specific levels of GHG emission reductions in different jurisdictions, regardless of the policy instruments employed by regulators, is neither regarded as a goal nor as a benchmark for determinations of equivalence. Rather, economic evaluations surrounding the application and operation of carbon pricing are front and centre stage.Footnote 21
Finally, the scope of application of the CBAM also testifies to a narrow focus on an economic level playing field. The discussion of the CBAM's scope of application in the European Commission's Impact Assessment (IA) lays particular emphasis on matters of trade intensity and the ensuing distortions of competitive opportunities for EU products on the EU internal market.Footnote 22 These demarcate the scope of application of the arrangements; the proposed Regulation would apply to electricity and a very limited number of goods, as listed in Annex I. This includes products in the cement, fertilisers, iron and steel, and aluminium sectors.
A different set of considerations applies to the Coons-Peters Bill. The hybrid regulatory design of the BCA proposal partly reflects an environmental level playing field rationale. Starting from the question of equivalence, section 9904(b)(2)(B) of the Bill provides that any country which does not impose a BCA on US products and which is deemed to enforce ‘laws and regulations designed to limit or reduce GHG emissions that are at least as ambitious as Federal laws and regulations’ shall be considered as an exempt country. The US Bill's reference to national climate change mitigation regulations and the attempt to measure third country policies against the US ‘yardstick’ are tantamount to a rudimentary focus on environmental equivalence and a rudimentary form of environmental level playing field. The US Bill seeks to capture the environmental component of third countries’ policies, or at least to capture it more fully than the CBAM's mere reference to price-based measures and ‘explicit’ carbon prices. This may result from the absence of carbon pricing mechanisms in the US, rather than intentional design. Regardless, the US proposal reflects some elements of an environmental level playing field rationale.
Turning to the available remedies, the BCA incorporates part of the CBAM's economic rationale; it artificially creates a domestic carbon price and provides for imported goods to bear the same (presumed) economic costs as US products. The amount of BCA to be imposed on importers results from the ‘emission value’ associated with the specific production processes of the covered fuel or good, multiplied by the ‘domestic environmental cost’ incurred in the US for the purposes of producing the relevant fuel or good. As briefly mentioned above, no price-based mechanisms for the reduction of GHG emissions are in force at the Federal level; for this reason, there is no ‘explicit’ carbon price in the US. As a consequence, the Bill provides for the identification of an overall US ‘domestic environmental cost’. This value shall be determined annually on the basis of the average cost incurred by companies within a sector ‘to comply with any Federal, State, regional or local law, regulation, policy or program which is [in force and] designed to limit or reduce GHG emissions, including (A) the Clean Air Act […]; (B) GHG emission standards for passenger cars and light trucks; (C) any State, regional or local law, regulation, policy or program that imposes a cap-and-trade system […] or a tax or fee on carbon dioxide’.Footnote 23 As specifically regards the applicable ‘emission values’, it is worth clarifying that a petition procedure is available. Importers may petition the Secretary of the Treasury to revise the determination of the production GHG emissions of a specific covered good; this enables the former to demonstrate that the emissions embedded in the imported goods are lower than presumed.Footnote 24
Although the proposal provides for economic remedies, some elements suggest that the BCA is located halfway through a stricto sensu economic level playing field rationale and an environmental level playing field perspective. In this sense, the proposed scheme operates as a punitive import charge, quasi-calibrated to the climate change mitigation commitments of third countries.
First, unlike the CBAM, the US Bill does not target the emissions of each and every product and the specific economic costs borne by all producers. The emissions embedded in a specific product and the latter's competitiveness, in the light of the economic costs associated with its production, are irrelevant if the country of origin of the product has sufficiently ambitious (price-based or non-price-based) climate change mitigation policies in place. In this sense, the US Bill shifts the focus from specific products and the economic costs associated with their production, to specific countries and their climate change mitigation commitments.
Secondly, all countries included on the OECD Development Assistance Committee's most recent list of least developed countries (LDCs) will be exempt from the application of the BCA. The exemption of all LDCs testifies to the absence of a pure economic level playing field rationale. From a policy-neutral, economic perspective, the exclusion of LDCs is unwarranted; indeed, the CBAM follows this rationale. By contrast, the US exemption reflects a broader environmental perspective and an acknowledgment of the principle of Common but Differentiated Responsibilities and Respective Capabilities (CBDR-RC).Footnote 25
To conclude, the scope of application of the proposed BCA is considerably broader than the one of the CBAM; this also reflects an environmental level playing field rationale. The proposed BCA would encompass both ‘covered fuels’Footnote 26 and ‘covered goods’;Footnote 27 the latter category includes goods produced within the steel, aluminium, cement and iron sectors,Footnote 28 any goods composed of 50% or more of materials produced within the above sectors,Footnote 29 and any further products prospectively identified by the Secretary of the Treasury.Footnote 30 The GHG emissions which are covered in the context of the US scheme are also much broader in scope than the ones covered by the EU CBAM. In the case of covered fuels, all ‘upstream GHG emissions’ resulting from the extraction, processing, transportation, financing or other preparation of for use of the fuel are taken into account.Footnote 31 In the case of covered goods, all relevant ‘production GHG emissions’ from production, manufacture and assembly are considered.Footnote 32 Pursuant to section 9905(b), the Secretary of the Treasury shall identify the methodologies to determine all relevant emissions. If reliable data is unavailable, the ‘benchmark emissions’ of the sector shall apply;Footnote 33 these are defined as the GHG emissions of the 1% highest emitting production sites of the specific sector in the US.Footnote 34
This confirms that the measures are not informed by a purely economic logic. The Coons-Peters Bill covers more sectors than the EU CBAM. The wording of the Bill suggests that the scheme is meant to encompass as many goods as possible within the relevant sectors; the GHG emissions which are covered are also very broad in scope. It is dubious whether substantial evidence of potential carbon leakage could be provided for all covered sectors and goods; in this sense, unlike the European Commission, the US legislator has hardly focused on trade intensity and the ensuing distortions of competition. Yet, for this very reason, the US Bill exemplifies an attempt to establish a broader environmental level playing field.
As illustrated by this concise overview, the EU CBAM and the US proposed BCA are respectively informed by an economic and an environmental level playing field perspective. The CBAM suffers from all the limitations which are typical of an economic level playing field rationale; these, as explained in the first section, are associated with a narrow focus on economic equivalence, recourse to economic remedies, and the determination of an economic scope of application. Conversely, the regulatory design of the BCA scores relatively better, from an environmental law perspective. The next section turns to the question of the obstacles associated with an environmental level playing field perspective, focusing on practical implementation matters.
2. Practical obstacles in the establishment of an environmental level playing field: equivalence, remedies and scope of application
Is the attempt to establish an environmental level playing field feasible in practical terms? An analysis of the BCA proposal helps provide an answer to this question. The first relevant consideration relates to the identification of the applicable remedies. The Coons-Peters Bill provides for the determination of the ‘domestic environmental cost’ incurred by US firms in the relevant sectors. This is meant to capture the economic costs associated with compliance with several non-price-based environmental policies, well beyond the ‘explicit’ price of carbon under the EU ETS. Nonetheless, the identification of consistent and reliable models to determine an ‘implicit’ domestic environmental cost raises highly complex methodological questions.
The possibility to convert the ‘implicit’ carbon price associated with recourse to non-price-based environmental policies in other jurisdictions into an ‘explicit’ economic value was taken into consideration at the EU level, in the context of the CBAM proposal. While this option was explored, its practical feasibility was excluded at an early stage.Footnote 35 The determination of ‘implicit’ carbon prices was considered too complex, and potentially arbitrary. More recently, an alternative suggestion has been put forward; this involves the determination of the marginal abatement cost associated with the most recent reduction of a ton of carbon dioxide emissions in a jurisdiction.Footnote 36 Although reference to marginal abatement costs provides a policy-neutral value, allowing comparison between a different mix of price-based, partial-price-based and non-price-based policies, it raises several methodological hurdles and postulates the availability of very detailed and up-to-date data from every jurisdiction. This is the first, considerable problem associated with the Coons-Peters Bill. By contrast, the CBAM's reference to the EU (ETS-derived) ‘explicit’ carbon price, typical of an economic level playing field perspective, makes the relevant regulatory arrangements feasible in practical terms.
Secondly, despite the precondition that the US ‘implicit’ environmental cost shall be determined, the US Bill does not provide for an assessment of economic equivalence on the basis of the other jurisdictions’ ‘implicit’ carbon price; as already explained, it focuses on the environmental commitments of third countries. While this reflects an environmental level playing field perspective, it triggers the question whether environmental equivalence may be assessed accurately and consistently in specific sectors. Environmental and climate change mitigation policies may include maximum emission standards, technical or performance-based standards, tradable performance standards, cap-and-trade systems, carbon taxes, financial incentives and subsidies for renewable energy.Footnote 37 The picture becomes increasingly complex if planning laws and transport regulations are taken into account. Taking different regulatory scenarios into account and measuring the relative ambition and effectiveness of a different mix of policies in different jurisdictions is an impossible endeavour. On these grounds, the very assessment of environmental equivalence becomes an unachievable task. Any such determination is liable to come under challenge for being inconsistent and arbitrary. The opposite is true of the assessment of economic equivalence, under the CBAM. The ‘explicit’ carbon prices existing in different jurisdictions and derived from different price-based policies can be easily and straightforwardly determined.
The final point relates to the broad scope of application of the US Bill. Unlike the proposal for a CBAM, the US Bill does not lay emphasis on trade intensity and the ensuing distortions of competition; as already explained, it aims to encompass within its scope of application several sectors, a plurality of products within these sectors, and all GHG emissions embedded in the relevant products. In practical terms, this broader perspective is associated with several problems. These are explored in great detail in the Commission's IA on the CBAM.
Starting from the question of the covered sectors, the IA acknowledges that a comprehensive scope of application ‘could make the largest contribution towards enhancing the effectiveness […]’ of carbon border measures.Footnote 38 Nonetheless, it also cautions against the inclusion of high-emission industrial sectors for which reliable data are unavailable. Examples are organic chemicalsFootnote 39 and refineries; in the latter sector, for instance, several products are produced simultaneously and any decision on how to attribute emissions to different outputs is considered impracticable at the current stage.Footnote 40 Only a handful of sectors, out of the 63 identified at the EU level as potentially at risk of carbon leakage, have thus been included within the CBAM.Footnote 41 The selection of the sectors was based on three criteria: trade intensity (as directly connected to distortions of competition and potential leakage); emission levels; and technical complexity.Footnote 42 This triggers the question whether the much more encompassing scope of the proposed US measures is feasible in practical terms.
Similar considerations apply to the determination of the specific products covered by the US import charge. The US Bill provides for a very broad scope of application. The EU IA, however, warns of the highly complex technical evaluations connected to a broad product coverage. The IA emphasises that a narrow scope of application may create incentives for the production of more finished or semi-finished products outside the EU; these could then be imported without being covered by the CBAM.Footnote 43 As a result, the question of how far down the value chain the CBAM should be applied has been the object of very careful scrutiny.Footnote 44 Despite the intention to cover as many products as possible, the CBAM proposal identifies a very limited number of goods. The practical arguments behind this choice are illustrated in great detail in the IA: these include the need to ensure that the covered products and materials are clearly and unambiguously identifiable in practice,Footnote 45 and the need to determine, monitor and verify the relevant embedded emissions ‘with reasonable robustness and credibility’.Footnote 46 This becomes all the more difficult at every step down the value chain, as the number of production steps increases and the carbon value relative to value generation varies.Footnote 47
Finally, several practical obstacles are associated with a broad determination of the ‘covered’ emissions. A ‘full carbon footprint’ approach came under analysis in the EU IA; this includes all emissions from the mining of raw materials, the production of materials and components and all relevant manufacturing processes, indirect emissions from energy production, emissions from the transport of the materials, components and final product, emissions from use, and emissions from the disposal of the product.Footnote 48 The scope of application of the US proposed BCA is largely informed by this approach. The European Commission, by contrast, backtracked on its original plan to include indirect emissions from energy production and use; yet again, reasons surrounding practical feasibility are behind the non-inclusion of indirect emissions.Footnote 49
The considerations developed in this section call into question the practical feasibility and the technical robustness of the arrangements of the Coons-Peters Bill. A broader environmental level playing field rationale is associated with a number of practical obstacles surrounding the determination of environmental equivalence, the identification of ‘implicit’ carbon prices, and the demarcation of the scope of application of the relevant regulatory arrangements. The opposite is true of a narrower focus on an economic level playing field.
3. Structural legal constraints: compatibility with the GATT 1994
This section focuses on the structural legal constraints associated with the obligations ensuing from the GATT 1994. The analysis demonstrates that the regulatory design of the CBAM, informed by an economic level playing field rationale, makes the scheme more likely to be WTO law compatible. An environmental level playing field rationale, by contrast, is connected to a number of problems. On these grounds, (compliance with) the WTO law system creates incentives for regulators to adhere to a narrower economic level playing field rationale. The first sub-section below provides a brief overview of the substantive obligations laid out in the GATT 1994, drawing the conclusion that both the CBAM and the BCA would have to be justified under Article XX GATT. The second sub-section turns to a detailed examination of the potential justification of the two schemes under Article XX GATT and the Chapeau thereof. This analysis paves the way for the final considerations of the paper.
(a) The substantive obligations of the GATT 1994: border tax adjustments and adjustable regulations
The question of the compatibility with WTO law of different kinds of measures aimed at redressing regulatory divergencies in the field of climate change mitigation has been extensively debated.Footnote 50 One point on which all commentators agree, unsurprisingly, is that any such measure should be framed as the import equivalent of domestic climate change mitigation measures. This triggers the application of Article III:2 or III:4, excluding the more stringent discipline of Article II or Article XI.Footnote 51
From a WTO law perspective, claiming that carbon border measures qualify as a border tax adjustment (BTA) would be the best strategy. In this context, the first relevant question is whether the EU CBAM and the US BCA could qualify as BTAs. Article II:2(a) GATT demarcates the scope of the notion of a BTA. It provides that nothing in Article II shall prevent a party from imposing at any time on the importation of a product ‘a charge equivalent to an internal tax imposed consistently with the provisions of paragraph 2 of Article III in respect of the like domestic product or in respect of an article from which the imported product has been manufactured or produced in whole or in part’. As further clarified in the note ad Article III, any internal tax or other internal charge which applies to an imported product and to the like domestic product and which is collected in the case of the imported product at the time or point of importation shall be regarded as an adjustable tax subject to the provisions of Article III.Footnote 52
This means that in order for a tax to be subject to border adjustment, three requirements must be met. The first precondition is the existence of an equivalent internal tax; a charge imposed on the importation of a product may only qualify as a BTA where an equivalent internal tax is levied on the like domestic product. Secondly, the internal tax must be imposed consistently with the provisions of Article III:2.Footnote 53 Finally, the internal tax must be levied on a ‘product’ or in respect of an ‘article from which the product is manufactured or produced’.
The question whether carbon taxes would qualify as adjustable product taxes or non-adjustable producer taxes has been widely debated in the literature.Footnote 54 Any detailed analysis of this point would go beyond the scope of the present enquiry. If we assume that carbon taxes may qualify for a BTA, their adoption could provide the most promising way forward to enact WTO law compliant carbon border measures.Footnote 55 In this sense, WTO law compatibility may influence the legal design of these measures and push towards an economic level playing field rationale. The adoption of carbon taxes and their adjustment at the border provides for a specific economic remedy and reflects a focus on economic equivalence, in that imported products end up ‘paying’ the same carbon tax already ‘paid’ by domestic products.
This triggers the question whether the CBAM and the BCA could qualify as BTAs. The answer is negative. The CBAM extends the carbon price ‘created’ by the EU cap-and-trade system to imports, employing a mandatory requirement for importers to purchase and surrender CBAM certificates. However, there is no internal tax to be adjusted at the border; nor does the proposed CBAM impose a charge on importers. Unlike the CBAM, the US proposal involves the payment of an import charge; the Bill provides for the US authorities to impose a ‘charge’ ‘on the importation of a product’, as per Article II:2(a). Nonetheless, such charge has no equivalent in the internal taxation system. Nor is it levied in the form of an ‘explicit’ carbon price on like domestic products.
The following question is whether the CBAM and the BCA may be considered as ‘adjustable regulations’ under Article III:4. This stipulates that imported products shall be accorded treatment no ‘less favourable’ than that accorded to ‘like’ domestic products in respect of all laws, regulations and requirements affecting their internal sale, offering for sale, purchase, transportation, distribution or use. A violation of Article III presupposes three elements.Footnote 56 The first element is the existence of a law, regulation or requirement applied to domestic and imported products. The relevant measures need not be identical.Footnote 57 The CBAM mandatory requirement to annually purchase and surrender certificates qualifies as a law affecting the internal sale of imported products.Footnote 58 The non-identical ‘domestic counterpart’ is the EU ETS system. On these grounds, the CBAM falls within the scope of application of Article III:4. The opposite applies to the US BCA. The US Bill provides for a ‘fee’ to be levied on the importation of specific products. This can hardly qualify as a ‘law, regulation or requirement’. This leads to the conclusion that the US BCA would neither qualify as a BTA, nor as an adjustable regulation.
An analysis of the compatibility of the CBAM proposal with Article III:4 then postulates a focus on the second and third elements of a violation of that article. These are that the imported and domestic products at issue qualify as ‘like’ products, and that imported products are accorded ‘less favourable’ treatment than like domestic products. A determination of ‘likeness’ under Article III:4 is considered to be ‘[…] fundamentally, a determination about the nature and extent of a competitive relationship between and among products’.Footnote 59 Although ‘likeness’ must be established on a case-by-case basis,Footnote 60 the WTO dispute settlement organs have consistently taken four broad indicators into account. These are: (i) the properties, nature and quality of the products; (ii) their end uses; (iii) consumers’ tastes and habits, encompassing consumer perception and behaviour; and (iv) the tariff classification of the products.Footnote 61 Regulatory distinctions relating to the characteristics of a class of productsFootnote 62 as well as consumer perception or behaviourFootnote 63 may be of relevance to a determination of ‘likeness’ in so far as they also have an impact on the competitive relationship between the products under analysis.Footnote 64 This triggers the question whether highly polluting goods (eg steel produced in blast furnaces) and ‘greener’ goods (eg steel produced with hydrogen-based technologies) may be deemed to be ‘like’ products. The answer is that these products are to be regarded as ‘like’ products, unless consumer behaviour towards highly polluting products shifts to such an extent as to alter the competitive relationship between carbon-intensive and green products. At the current stage, it is difficult to envisage such scenario.
The following step involves an analysis of whether the CBAM would afford ‘less favourable treatment’ to imported products than the one afforded to ‘like’ domestic products. A determination of ‘less favourable treatment’ rests on a finding that the measures under analysis affect the equality of competitive opportunities of imported and ‘like’ domestic products.Footnote 65 Would the proposed CBAM afford ‘less favourable treatment’ to imported products vis-à-vis the ‘like’ domestic products? It is legitimate to suggest that it would. The prevailing test in the context of Article III:4, known as the ‘disparate impact’ test, involves an assessment of whether the ratio of imported products which are the object of ‘less favourable’ treatment outweighs the corresponding ratio of domestic products.Footnote 66 An application of this test in the context of the CBAM implementation would lead to the conclusion that the economic costs imposed on imported products are on average greater than the ones which are borne by domestic ‘like’ products under the ETS. EU products are likely to be on average less carbon-intensive (and thus more expensive, pre-CBAM) than imported products. Imported products originating from countries with more lenient climate change policies, conversely, will be more carbon-intensive (and thus cheaper, pre-CBAM) than EU products. As a result, the GHG emissions embedded in imported products will on average be higher than the ones embedded in EU products; imported products will then ‘pay’ more under the CBAM than the average EU product ‘pays’ under the ETS. Indeed, if national environmental commitments had been comparable and economic competition between imported and domestic products had been fair, there would have been no need for the EU to attempt to level the playing field via the CBAM.Footnote 67
Regardless, and even if the CBAM were not to be held in breach of Article III:4, the WTO dispute settlement organs would find a violation of the Most Favoured Nation (MFN) principle. Pursuant to Article I:1, any advantage granted to any product originating in (or destined for) any of the contracting parties shall be accorded immediately and unconditionally to a ‘like’ product originating in (or destined for) all other contracting parties. It is fair to suggest that the proposed CBAM would violate the MFN principle de jure as well as de facto.
Under the proposed Regulation, importers may claim a reduction in the number of their CBAM certificates corresponding to any carbon price already paid for in the country of origin of a product; this grants a direct advantage to products imported from countries which have established a carbon pricing system, as opposed to countries which employ non-price based mechanisms. Further, the WTO dispute settlement organs have consistently focused on de facto violations of the MFN principle; any advantage which is granted in practice to the products originating from one country must be accorded to all other countries.Footnote 68 Under the CBAM proposal, the number of certificates to be purchased is calibrated to the GHG emissions embedded in the imported products; this grants an indirect advantage to imported products originating from countries with more stringent climate change mitigation policies in place, vis-à-vis ‘like’ imported products originating from other countries.
As this concise analysis has endeavoured to show, the CBAM is likely to violate Article III:4 and very likely to violate Article I:1. The US Bill will probably fall for assessment under the stringent discipline of Article II. This triggers the question whether the two measures may be justified under Article XX. The next sub-section thus turns to an analysis of this point.
(b) Article XX and the Chapeau
Recourse to Article XX enables Members to justify measures which would otherwise be incompatible with the substantive obligations of the GATT 1994. In accordance with the well-entrenched ‘two-tiered’ test, a measure will have to be provisionally justified under one of the exceptions provided for in Article XX, sub-paragraphs (a) to (j). An evaluation of compatibility with the Article's Chapeau (introductory clause) will then ensue.Footnote 69
The EU CBAM and the US BCA could be provisionally justified under sub-paragraphs (a), (b) and (g). The conditions enshrined in sub-paragraph (g) are easier to meet. Pursuant to this sub-paragraph, measures relating to the conservation of exhaustible natural resources may be provisionally justified where they are made effective in conjunction with restrictions on domestic production and consumption. The WTO dispute settlement organs have embraced an evolutionary, extensive interpretation of the notion of ‘exhaustible natural resources’; clean air as well as endangered animal species have been deemed to fall within this category.Footnote 70 As regards the requirement that measures shall ‘relate to’ conservation goals, the dispute settlement organs have found that the relevant measures must be ‘primarily aimed at’ the conservation of exhaustible natural resources,Footnote 71 and that their structure and design must show a ‘close and genuine relationship of ends and means’ between the measures and the relevant objectives.Footnote 72 The final requirement is that the measures shall be made effective in conjunction with domestic restrictions. This has been interpreted as a requirement of ‘even-handedness’ in the application and implementation of the measures,Footnote 73 involving complementary regulations directed to both domestic and imported products.
None of these requirements seem to pose any obstacles to the provisional justification of the CBAM and the BCA. Both measures aim to avoid carbon leakage in carbon-intensive sectors; in this sense, their goal is the reduction of atmospheric GHG emissions and the protection of the climate system (‘conservation of an exhaustible natural resource’). It is legitimate to presume that both measures would be considered to be ‘primarily aimed’ at and rationally connected to this goal. Equally, they comply with the ‘even-handedness’ criterion.Footnote 74
Sub-paragraphs (a) and (b) respectively enable Members to provisionally justify measures which are necessary to protect public morals, or necessary to protect human, animal or plant life or health. In so far as they seek to protect the climate system, both the CBAM and the BCA aim to protect human, animal and plant life and health; this point is not likely to be controversial, as long as the parties can produce evidence of potential carbon leakage in carbon-intensive sectors. As regards ‘public morals’, the dispute settlement organs have found that this denotes standards of right and wrong conduct which are bound to vary in different contexts and in correspondence to different value systems.Footnote 75 According to the Appellate Body (AB), a measure which is ‘designed’ to protect public morals is a measure that is not incapable of achieving this goal; the dispute settlement organs shall assess the design of the measure at issue, including its content, structure and expected operation.Footnote 76 The EU and the US could make a case that carbon border measures help tackle climate change and promote ‘greener’ technologies in third countries; in this perspective, and in so far as these measures promote environmental goals to which national constituencies attach the highest importance, the design and the expected operation of the CBAM and BCA are perhaps not incapable of protecting public morals.
An assessment of the ‘necessity’ of the relevant measures, on the other hand, involves ‘weighing and balancing’ different factors.Footnote 77 In this context, a Panel must undertake a holistic assessment of the extent to which a measure contributes to its goals vis-à-vis its degree of trade restrictiveness, in the light of the importance of the interests or values at stake. If a Panel reaches the preliminary conclusion that a measure is necessary, it must then compare it with any less trade restrictive alternatives which would achieve the same level of protection.Footnote 78
The dispute settlement organs have always emphasised that the balance between the trade restrictiveness of a measure and the degree to which it contributes to its objective is bound to be differently appreciated in cases involving different policy goals.Footnote 79 Further, they have consistently found that public health and environmental protection are goals of the highest importance.Footnote 80 This suggests that invoking sub-paragraph (b) – rather than sub-paragraph (a) – could facilitate the EU and the US. Under the ‘weighing and balancing’ process, the dispute settlement organs must establish the degree of trade restrictiveness of the measures and the extent to which they contribute to the relevant policy goal in qualitative or quantitative terms.Footnote 81 However, as the AB has clarified, the ability of a measure to produce a material contribution in the long-term should also be taken into due consideration.Footnote 82 In the case of these carbon border measures, which do not involve bans or restrictions, it is reasonable to suggest that the contribution of the measures to the achievement of their (crucial) policy goal would be deemed to outweigh the measures’ trade restrictiveness. Equally, it is difficult to envisage the existence of reasonably available and less trade restrictive alternatives which would achieve the same level of climate protection.Footnote 83
The following step would then involve an evaluation of compliance with the Chapeau. This provides that (provisionally justified) measures shall not 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. The Chapeau's requirements are an expression of the principle of good faith; they underscore the attempt to draw a ‘line of equilibrium’ between the right of a Member to have recourse to the Article XX exceptions, and its duty to respect the rights of the other Members.Footnote 84 Compliance with the Chapeau must be evaluated on a case-by-case basis; the ‘line of equilibrium’ moves ‘as the kind and the shape of the measures at stake vary […]’.Footnote 85
The dispute settlement organs have remarked that the Chapeau targets and addresses the question of the application of a measure, ie the manner in which it is applied in practice.Footnote 86 Such assessment, however, will also involve an examination of the ‘design, architecture and revealing structure’ of the measure.Footnote 87 In this sense, the boundaries between the two analytical dimensions (‘regulatory structure’ and ‘practical application’) are blurred.Footnote 88 The same exact factual circumstances which have resulted in a finding of a violation of the GATT's substantive provisions may come under analysis under the Chapeau;Footnote 89 nonetheless, the relevant factual elements will be analysed under a different light and from a different perspective.
The AB has held that the notions of ‘arbitrary or unjustifiable discrimination’ and ‘disguised restriction on international trade’ impart meaning to each other.Footnote 90 A ‘disguised restriction’ includes forms of ‘disguised discrimination’ of an arbitrary and unjustifiable nature.Footnote 91 As regards the reference to discrimination between ‘countries where the same conditions prevail’, the relevant ‘conditions’ should be identified by having regard to the sub-paragraph of Article XX under which the measure was provisionally justified, as well as the substantive obligations of the GATT violated by the measure.Footnote 92
Would the CBAM and the BCA breach the requirements of the Chapeau? The conclusive part of this sub-section aims to address this question against the background of the dispute settlement organs’ findings in previous disputes. The examination demonstrates that the CBAM's economic level playing field rationale overall scores better under an analysis of compliance with the Chapeau's requirements. By contrast the BCA arrangements, informed by an environmental level playing field perspective, are highly unlikely to be WTO law compatible.
To begin with, the criteria for the evaluation of the environmental equivalence of different national commitments vis-à-vis US policies and the very complex determination of the ‘domestic environmental cost’ incurred by US firms (the ‘manner in which the measure is applied in practice’) are highly likely to be challenged as forms of ‘arbitrary discrimination’. In US – Shrimp, the AB laid particular emphasis on the need to adopt transparent and predictable procedures, with a view to safeguarding due process and basic fairness; it also highlighted that the processes followed by the US were ‘singularly informal and casual’.Footnote 93 Similar considerations may apply to the over-complex and potentially arbitrary system laid out in the Coons-Peters Bill.
Another major flaw in the design and structure of the US import charge lies in the disregard for the ‘implicit’ carbon price existing (and ‘levied’) in other jurisdictions. If the policies of a third country are not recognised as equivalent in environmental terms to the US ones, a product will be subject to the proposed import charge (US ‘domestic environmental cost’ multiplied by embedded emissions). The imposition of the full (‘implicit’) US ‘domestic environmental cost’, however, fails to take account of the existence of an ‘implicit’ (albeit lower) carbon price in jurisdictions with non-equivalent environmental commitments. The CBAM may lend itself to the same criticism, in so far as it also fails to take into account the ‘implicit’ carbon prices existing in other jurisdictions. However, unlike the BCA, the CBAM takes the ‘explicit’ (ETS-derived) price of carbon as a benchmark; symmetrically, it enables importers to claim a reduction in their number of CBAM certificates corresponding to any ‘explicit’ carbon price paid in other jurisdictions. This is an important difference.
Further problematic aspects are associated with the environmental level playing field rationale of the BCA. Products originating from countries whose environmental policies are deemed equivalent to the US ones would not be subject to the BCA, regardless of whether they are ‘green’ or highly polluting products. ‘Green’ products originating from countries whose environmental policies are not regarded as equivalent to the US ones, by contrast, would be subject to the BCA import charge. This aspect is difficult to reconcile with the Chapeau requirements.Footnote 94
A closely related point is the one of the broad scope of application of the US measure in terms of sectors, covered products and relevant emissions. This is also very likely to be WTO law incompatible, if solid evidence of the potential for carbon leakage in specific sectors is missing. An analysis of the broad scope of application of the US import charge may then lead to a finding of a ‘disguised restriction on international trade’. Further, and importantly, the exemption of LDCs from the application of the US BCA is also likely to come under challenge in so far as it undermines the environmental rationale of the measure. Exempting LDCs while failing to exempt other Members is likely to discriminate between countries where the same (environmental) conditions prevail, and very likely to ‘promote’ carbon leakage in these specific countries.Footnote 95
As the above analysis has illustrated, the application of the US proposed BCA is associated with several problematic aspects; the elements under examination are all typical of an environmental level playing field rationale.Footnote 96 The last step of the enquiry involves an assessment of the CBAM and of the economic level playing field rationale. Two aspects deserve particular attention. The first point regards the question of arbitrary and unjustifiable discrimination. As already seen, the CBAM proposal only accounts for (and waives) the ‘explicit’ carbon prices of different jurisdictions. This may amount to arbitrary and unjustifiable discrimination between countries where the same environmental conditions prevail;Footnote 97 as already explained, an economic level playing field rationale straightforwardly focuses on the existence or absence of carbon pricing and thus fails to capture the effectiveness of environmental policies. In this perspective, ‘extending’ the EU carbon price to products originating from countries which may have achieved the same levels of GHG emission reductions as the EU by having recourse to non-price-based policies has a twofold effect: it fails to account for the ‘implicit’ carbon prices existing in these jurisdictions, and it fails to achieve environmental goals.Footnote 98 It is impossible to predict whether a Panel would make such a finding; the key question would probably be whether countries with and without price-based policies may be regarded as countries where the same environmental conditions prevail.
The final dimensions to address relate to the requirement to engage in international negotiations, and ‘coercion’. The dispute settlement organs have consistently expressed a strong preference for multilateral and consensual solutions, as opposed the imposition of unilateral measures.Footnote 99 The corollary under the Chapeau is that the regulating Member shall demonstrate that prior to enacting the relevant unilateral measure, and after that moment, it has negotiated in good faithFootnote 100 with all (affected) MembersFootnote 101 and it has taken the (relevant) conditions prevailing in different countries into account.Footnote 102 This requirement will weigh heavily on both the EU and the US. More specifically, the EU's efforts to negotiate and consider the position of developing countries and LDCs may come under the spotlight. Despite being justifiable from a carbon leakage policy perspective, the CBAM's ‘one-size-fits-all’ approach may be considered irreconcilable with the Chapeau requirements in the absence of repeated attempts by the EU to negotiate in good faith for a multilateral solution.Footnote 103
This aspect is also closely related to the assessment of the ‘coercion’ element. In US – Shrimp, the AB famously laid emphasis on the coercive effects of the application of the challenged measure and its rigid and unbending requirements.Footnote 104 ‘Coercion’ encompasses two different dimensions: the requirement for all exporting Members ‘to adopt essentially the same policy’ as that enacted by the importing Member,Footnote 105 ie a focus on specific ‘means’ rather than on the relevant policy goals and ‘aims’, as well as a failure to take into consideration the ‘different conditions which may occur in the territories of […] other Members’.Footnote 106 Could an examination of ‘coercion’ lead to a finding that the application of the CBAM is in breach of the Chapeau requirements? This cannot be categorically excluded.
First, the CBAM may have indirect coercive effects in so far as it fails to account for ‘implicit’ carbon prices; this may push third countries to adopt price-based mechanisms. However, it is worth noting that effective non-price-based policies are bound to result in significant GHG emission reductions; conversely, non-effective price-based policies will result in a low ‘explicit’ carbon price. In these circumstances, products originating from a country with ineffective price-based policies and a low ‘explicit’ carbon price are still likely to ‘pay’ more under the CBAM than products originating from a country with effective non-price-based policies and a high ‘implicit’ carbon price. Although the CBAM does not account for ‘implicit’ carbon prices, it partially captures them through the calculation of the embedded emissions. Secondly, the CBAM may have coercive effects in so far as the revenues of the scheme are collected and used in the EU. This element could indirectly push third countries which employ non-price-based policies to adopt price-based mechanisms; this would enable the third country to levy the relevant ‘explicit’ carbon price internally and make use of the relevant revenues, rather than have it levied in the EU.Footnote 107
It is impossible to predict whether these elements in the application of the CBAM could lead to a finding that the requirements of the Chapeau have been breached. Such a finding would make it extremely difficult to enact unilateral measure to tackle carbon leakage. Indeed, as this sub-section has illustrated, an environmental level playing field perspective is associated with a plurality of problems; nor are measures informed by this rationale likely to be WTO law compliant. Although both the CBAM and the US proposal have weaknesses, it is fair to suggest that the former has more chance of surviving scrutiny. Overall, under WTO law, the CBAM's structure and design and its narrow focus on an economic level playing field score better than an environmental level playing field rationale.
4. The future of defensive policies?
This paper has analysed the EU proposal for a CBAM and the US Coons-Peters Bill against the backdrop of the notion of ‘defensive’ policies. The second section has demonstrated that the two measures are broadly informed by an economic and an environmental level playing field rationale. The third section has explored the practical obstacles associated with an environmental level playing field perspective, illustrating the technical difficulties in the determination of environmental equivalence, the identification of adequate remedies, and the adoption of a broad scope of application. The fourth section has then turned to an examination of structural legal constraints, focusing on the obligations ensuing from the GATT 1994. The analysis has shown that the CBAM's economic level playing field perspective and its reference to an ‘explicit’ carbon price overall score better under WTO law. This demonstrates that the CBAM is more feasible in practical terms and more likely to be WTO law compliant than the US proposed BCA. As a matter of regulatory design, practical obstacles and structural constraints push towards a narrower focus on an economic level playing field.
As illustrated throughout the analysis, an environmental level playing field rationale is associated with technical and legal difficulties and high levels of complexity. However, as explained in the first section, an economic level playing field perspective has considerable limitations in terms of environmental integrity. These emerge from an analysis of the dynamics of economic equivalence, recourse to (environmentally unambitious) economic remedies, and the role played by distortions of competition in the determination of the scope of application of the relevant measures.
At the time of writing, the proposal for a CBAM is at an early stage in the EU legislative process. The EU institutions still have plenty of time to refine the CBAM's regulatory arrangements and remedy some of its weaknesses. Economic equivalence and reference to ‘explicit’ carbon prices are both inherent to the regulatory structure of the CBAM; the same is true of the application of economic remedies, providing for all imported products to ‘bear’ the same exact economic costs as EU products. For this reason, these elements are highly unlikely to be the object of revision; further, as this paper has explained, assessing environmental equivalence and determining ‘implicit’ carbon prices pose a plurality of challenges. Nonetheless, there are margins for improvement on different regulatory aspects.
First, the revenues of the CBAM should be destined for LDCs and used to enable technology transfer and capacity-building.Footnote 108 The application of the CBAM already remedies distortions of competition between products, ‘altering’ the price of (generally more polluting and) originally cheaper imported products. On these grounds, there is no valid reason for the EU to keep the CBAM revenues and use them internally. On the contrary, redistributing the revenues to LDCs would be in line with the tenets of the CBDR-RC principle and would provide a more appropriate alternative than exempting LDCs from the application of the CBAM. As already explained, exempting specific countries undermines the environmental effectiveness of the scheme, promotes carbon leakage, and is difficult to reconcile with the Chapeau requirements.
A second and related point is the one regarding free allowances in the EU. These should be phased out as soon as possible, regardless of the specific provisions on the corresponding adjustment in the number of CBAM certificates to be purchased by importers. From an environmental protection perspective, the CBAM and the allocation of free allowances to firms operating in sectors at risk of carbon leakage pursue the same exact goal. From an economic perspective, however, the continuation of the free allowances system helps redress the distortions of competition that will materialise on foreign markets between exported EU products, on the one hand, and products originating from third countries, on the other. Clearly, the coexistence of the free allowances system and of the CBAM arrangements undermines the latter's environmental rationale and testifies to the narrow economic focus of the scheme.Footnote 109
Finally, the scope of application of the CBAM should be broadened in so far as technically and practically possible; indirect GHG emissions should also be included.Footnote 110 The EU institutions should obviously collect sufficient evidence surrounding potential carbon leakage; however, an overly narrow focus on trade intensity and on the products which are more likely to face competition on the EU internal market reinforces the arguments on ‘green protectionism’.
In the years to come, we will certainly witness the adoption of more ‘defensive’ policies.Footnote 111 In a rapidly evolving technical and regulatory landscape, it is impossible to predict what these policies will look like. As this paper has illustrated, no easy solution is available. This final section above has pointed to potential improvements in the specific context of the CBAM; these would have to take place in the short term, throughout the EU legislative procedures. In the long term, however, regulators should certainly focus on new ways to overcome the problems associated with an environmental level playing field rationale.
