2.1 Climate change

Decarbonizing the global economic system to limit climate change is the foremost challenge of this century. If GHG emissions growth continues unabated, a global average surface temperature increase of 3–5°C and sea level rise of 0.5–1.0 m is likely by 2100.Footnote ⁴ The consequences of this altered climate will include increased human mortality from extreme weather events and vector-borne diseases, water and food shortages, mass coral bleaching, and risks to coastal infrastructure and low-lying ecosystems.Footnote ⁵ People in developing countries are particularly vulnerable to these impacts, and many will be displaced by future climate-related disasters, increasing the flow of refugees to the developed world (Yonetani et al., Reference Yonetani2015). It is ‘virtually certain’ (i.e. with 95–100% probability) that observed changes in the climate since the industrial revolution have been caused by humans.Footnote ⁶ CO₂ released from fossil fuel combustion and industrial processes amount to approximately 33 GtCO2-e (65%) of annual anthropogenic GHG emissions.Footnote ⁷ Thus, to avoid a 2°C average temperature rise, most of the world's fossil fuels will need to remain unburned.Footnote ⁸ Public financial support for the fossil fuel industry contradicts efforts to mitigate climate change by encouraging consumption of the very fuel sources that are largely responsible for it. It also diminishes the need to explore and invest in alternative, low-carbon, energy sources such as wind and solar (Coady et al., Reference Coady, Parry, Sears and Shang2015; IEA, 2014). Therefore, eliminating fossil fuel subsidies is a ‘no-brainer’ in mitigating climate change (Rockström et al., Reference Rockström, Gaffney, Rogelj, Meinshausen, Nakicenovic and Schellnhuber2017).

2.2 Fossil fuel subsidies

There is no consensus among international economic organizations as to how energy subsidies should be defined.Footnote ⁹ Moreover, methodology selection among these organizations is influenced by political considerations and their bureaucratic characteristics (Skovgaard, Reference Skovgaard2017). Nevertheless, whatever assessment method is used, there is general agreement that government support for the fossil fuel industry is substantial, comes in many forms (Steenblik, Reference Steenblik2008), and favours consumption-based, rather than production-based,Footnote ¹⁰ subsidies. The largest estimate of fossil fuel energy subsidies is provided by the International Monetary Fund (IMF), which estimates global fossil fuel subsidies at $4.9 trillion for 2013 (6.5% of global GDP), using a hybrid methodology incorporating macroeconomic analysis and taxonomic classification (Coady et al., Reference Coady, Parry, Sears and Shang2015). This IMF's estimate includes the cost of market externalities produced by climate change and local environmental pollution, arguably making it more comprehensive than alternative assessments. A more modest estimate is provided by the International Energy Agency (IEA, 2014), which estimates the value of global fossil fuel subsidies at US$548 billion in 2013 (IEA, 2014). The smallest estimate is given by the Organization for Economic Co-operation and Development (OECD), which estimate the government support among its members for fossil fuels at US$55–90 billion per year between 2005 and 2011 (OECD, 2013). The OECD's more limited approximation is explained by its narrow selection of measures and geographic scope, which extended only to the 34 OECD member countries (OECD, 2013).

Aside from the environmental necessity of curbing fossil fuel subsidies, there are sound economic reasons for their elimination. First, fossil fuel subsidies counteract other government attempts to lower the price of renewable energy by simultaneously deflating the price of fossil fuel energy. Second, fossil fuel subsidies redirect public spending away from productive purposes and play a part in producing government budget deficits in developing countries (Coady et al., Reference Coady, El-Said, Gillingham, Medas and Newhouse2006). Finally, fossil fuel subsidies are a poorly targeted mechanism for reducing the impact of fuel price rises on low-income households, which is one of the principal justifications provided by governments (Ellis, Reference Ellis2010). Although fossil fuel consumption subsidies do alleviate some of the effects of high-energy prices on low-income households, wealthier households accrue the majority of benefits due to their higher levels of consumption (Coady et al., Reference Coady, El-Said, Gillingham, Medas and Newhouse2006). Accordingly, government spending could be more efficiently targeted through programs that specifically insulate low-income households from fuel price rises, such as strengthening existing social safety nets (Coady et al., Reference Coady, El-Said, Gillingham, Medas and Newhouse2006). In these circumstances, the removal of fossil fuel subsidies would be a ‘win-win’ for climate change mitigation and economic policy (Brewer, Reference Brewer2003).

3.1 Relevant institutions

Subsidy reform sits at the nexus of a myriad of IEOs, such as the G20, WTO, IEA, Organization of the Petroleum Exporting Countries (OPEC), World Bank, IMF, United Nations Environment Programme (UNEP), Asia-Pacific Economic Cooperation (APEC), the International Renewable Energy Agency (IRENA), Global Subsidies Initiative, and the Friends of Fossil Fuel Subsidy Reform. These institutions have varying and overlapping roles to play in reforming fossil fuel subsidies, including multilateral decision-making, information sharing and analysis, lobbying, investment and program operations. For its part, the G20 first addressed the issue of fossil fuel subsidies in 2009, when world leaders agreed to ‘phase out and rationalise over the medium term inefficient fossil fuel subsidies’.Footnote ¹¹ In 2016, the G7 leaders placed a timeline on this pledge, committing to ‘the elimination of inefficient fossil fuel subsidies … by 2025’.Footnote ¹² Yet despite this rhetoric, no negotiations have commenced on a multilateral framework to eliminate fossil fuel subsidies. Further, although IEA estimates of fossil fuel subsidies have fallen from $500 billion in 2014 to $325 billion in 2015, much of this can be attributed to declining energy prices, rather than active efforts by governments to remove financial support (IEA, 2016b). Thus, there is an ongoing need to reach international agreement on eliminating fossil fuel subsidies as quickly as possible, perhaps drawing upon the lessons of fishery subsidy reform, where consensus is now emerging on the need to discipline subsidies that contribute to ecologically harmful fishing practices (Young, Reference Young2017). Admittedly, this may be difficult, as fossil fuel subsidies encourage consumptive and growth-oriented capitalism, which makes their removal unpalatable (Harris and Lee, Reference Harris and Lee2017).

3.2 Climate change law

In the absence of a binding multilateral agreement that specifically addresses fossil fuel subsidies, the relevance of other international legal instruments should be considered. Underpinning the international climate change law regime is the United Nations Framework Convention on Climate Change (UNFCCC),Footnote ¹³ beneath which sit the Kyoto ProtocolFootnote ¹⁴ and the Paris Agreement. The UNFCCC itself does not refer to subsidies. The absence of any explicit condemnation can be explained for a number of reasons, including the resistance of members to relinquish sovereignty over their natural resources, the unwillingness of developing countries to limit their options for economic development, and the disruptive influence of oil-producing members (Van Asselt and Kulovesi, Reference Van Asselt and Kulovesi2017).

In comparison, the Kyoto Protocol explicitly identifies fossil fuel subsidies in Article 2:1(a), which states that Annex I parties (i.e. developed countries) shall ‘implement and/or further elaborate policies and measures … such as: (v) Progressive reduction or phasing out of market imperfections, fiscal incentives, tax and duty exemptions and subsidies in all greenhouse gas emitting sectors that run counter to the objective of the Convention and application of market instruments’. Further, Article 2:3 states that Annex I parties ‘shall strive to implement policies and measures under this Article in such a way as to minimize adverse effects, including the adverse effects of climate change, effects on international trade, and social, environmental and economic impacts on other Parties’.

The strong language of the Kyoto Protocol was not, however, adopted in the Paris Agreement, which, like the UNFCCC, omitted any explicit reference to energy subsidies. Nevertheless, although the text of the Paris Agreement does not on its face deal with subsidy reform, its architecture allows for it through mechanisms such as ‘voluntary reporting on subsidies and their reform, putting forward subsidy reform as a nationally appropriate mitigation action, its inclusion in nationally determined contributions and discussing experiences in the context of technical expert meetings’ (Van Asselt and Kulovesi, 2017: 367).

Thus, although the fossil fuel subsidy reform under the climate change law regime is not impossible, it is unlikely to occur in the immediate future; the first revision of nationally determined contributions under the Paris Agreement is not due until 2022.Footnote ¹⁵ The development of a normative international climate change mitigation response that includes fossil fuel subsidies would also require a dramatic shift in the current approach of most countries. If a legal tool for compelling action that is more urgent is to be found, it must be located outside of the climate change law regime.

3.3 Trade law

International trade law has direct implications for climate change and fossil fuel subsidies. The linkage between trade and the environment is acknowledged in the preamble to the Marrakesh Agreement,Footnote ¹⁶ which states that the ambitions of the WTO parties should be achieved ‘in accordance with the objective of sustainable development’. More specifically, the WTO covered agreements with the most direct relevance for climate change governance and energy subsidy reform are the General Agreement on Tariffs and Trade (GATT)Footnote ¹⁷ and the SCM Agreement. The GATT principles of non-discrimination have been used to discipline a number of renewable energy subsidy programs. However, the GATT has never been invoked in order to challenge fossil fuel subsidies. The precise legal elements of the SCM Agreement are set out in greater detail in Section 6 of this paper. For now, it is sufficient to observe that, like the GATT, the SCM Agreement can only be invoked when a subsidy results in a trade-related market distortion. Subsidies with a purely domestic orientation will not be captured. It adopts a cascading ‘traffic light’ scale of prohibited (‘red-light’), actionable (‘amber-light’) and non-actionable (‘green-light’) subsidies. The green-light classification originally shielded all non-specific and certain other subsidies from discipline, even if they would have otherwise breached the SCM Agreement.Footnote ¹⁸ However, the provision was time-limited and expired on 1 January 2000.Footnote ¹⁹ Repeated attempts to revive it have been unsuccessful.Footnote ²⁰ The SCM Agreement has been used to discipline renewable energy subsidies, but not fossil fuel subsidies.

Not only is the international trade law regime more legally relevant to energy subsidies, it also presents greater opportunities for tangible outcomes. The Paris Agreement's enforcement approach is ‘facilitative, non-adversarial, and non-punitive’, in keeping with the ‘bottom-up’, voluntary character of the rest of the Agreement (Doelle, Reference Doelle2016: 16). In contrast, the WTO DSB's ‘system of compulsory third-party adjudication’ is a ‘rarity in international relations’ (Mavroidis, Reference Mavroidis2012: 840), and has been held out to be the ‘jewel in the crown of the WTO’.Footnote ²¹ It also enjoys high rates of participation and compliance (Leitner and Lester, Reference Leitner and Lester2015). In this context, the WTO makes sense as the forum for international climate litigation because it is the most promising avenue for legal remedies.

4.1 Renewable energy

As of May 2017, nine disputes had been initiated through the DSB against renewable energy programs (Meyer, Reference Meyer2017). These programs were challenged under the GATT and SCM Agreement rules regarding national treatment, antidumping, and subsidies (Meyer, Reference Meyer2017). Five of these disputes initially based their claims on violations of the SCM Agreement (Meyer, Reference Meyer2017). However, the only case where a Panel was actually established to adjudicate an SCM Agreement claim against a renewable energy program was in Canada–Renewable Energy. Footnote ²² In that case, the Appellate Body, in a much criticized ruling (Genest, Reference Genest2015; Rubini, Reference Rubini2015), declined to rule on whether the scheme was in breach of the SCM Agreement.Footnote ²³ Instead, the scheme was found to violate Canada's national treatment obligations under the GATT.Footnote ²⁴ It could be concluded that the complete absence of SCM Agreement specific cases indicates that the Agreement is less problematic for renewable energy programs than first thought. However, while the GATT has been the instrument of choice for claimants, this is likely the result of tactical decisions during litigation, namely the relative simplicity of legal argument under the GATT compared with the SCM Agreement (Birhanu Asmelash, Reference Birhanu Asmelash2015). Under the GATT, renewable energy programs that attach domestic content requirements (e.g. Canada's feed-in-tariff scheme) are a ‘straightforward’ violation of the national treatment obligation, which is designed to avoid protectionist economic policies (Meyer, Reference Meyer2015). In India–Solar Cells, the United States’ decision to withdraw its claim under the SCM Agreement claim but persist with the GATT provides further evidence that claimants are satisfied that the GATT alone is sufficient to successfully challenge renewable energy programs (Birhanu Asmelash, Reference Birhanu Asmelash2015).

Nevertheless, while the GATT may be the subject of more litigation than the SCM Agreement, that does not diminish the legal relevance of the SCM Agreement for renewable energy subsidies. In fact, as well as breaching GATT national treatment obligations, subsidies that are contingent upon the use of domestic over imported goods also violate the SCM Agreement.Footnote ²⁵ Further, unlike fossil fuel subsidies, many renewable energy subsidies are provided to energy producers rather than consumers as these subsidies encourage the international competitiveness of domestic firms (Birhanu Asmelash, Reference Birhanu Asmelash2015). These subsidies are vulnerable to the SCM Agreement as specificity under Article 2 is usually more obvious in the case of production subsidies (Marhold, Reference Marhold, Hancher, De Hauteclocque and Salerno2017). Moreover, de facto specificity will arise if the subsidy is specific to a certain industry that captures many programs that explicitly provide incentives to renewable energy manufacturers (Birhanu Asmelash, Reference Birhanu Asmelash2015). Finally, several other energy programs that are likely in breach of the SCM Agreement have been identified, including Italy's feed-in-tariff program (D'Orsi, Reference D'Orsi2014) and Germany's allocation of free emissions allowances under the European Union's carbon trading scheme (Shah, Reference Shah2007). This suggests that many renewable energy subsidies will fall under the red or amber-light classifications in the SCM Agreement, and that the Agreement remains relevant for renewable energy subsidies, despite its lack of litigation.

Work to protect renewable energy subsidies has focused on ways to strengthen the GATT and SCM Agreement to ensure that climate change mitigation efforts are not hindered by international economic law. As a starting point, the SCM Agreement contains no GATT Article XX equivalent, which would allow for public policy purposes to be accounted for in assessing the legality of a measure (Shadikhodjaev, Reference Shadikhodjaev2015). However, Article XX is also unlikely to be read into the SCM Agreement, without the adoption of a new instrument explicitly endorsing its application in other WTO covered agreements (Waltman, Reference Waltman2016; Shadikhodjaev, Reference Shadikhodjaev2015). On the other hand, invoking Article XX exceptions may prove unhelpful, as they currently do not provide protection for domestic content requirements (Shadikhodjaev, Reference Shadikhodjaev2015). In order to ensure that the Article XX exceptions are available for domestic content requirements, Meyer proposes the introduction of a three-limb ‘political necessity’ test, which incorporates an assessment of the measure's contribution to net global welfare (Meyer, Reference Meyer2015). Other suggested improvements to the trade law regime include removing the domestic content restriction specifically for renewable energy or other climate change mitigation policies (Waltman, Reference Waltman2016), amending the actionability criteria in the SCM Agreement Articles 6.3 and 15.7 to allow mitigation policies to usurp trade liberalization and providing specific exceptions for mitigation or renewable energy policies (Waltman, Reference Waltman2016).

4.2 Fossil fuels

Fossil fuel subsidies have never been litigated under the WTO. That is not because these subsidies cannot be litigated, merely that a number of their features generally preclude them from the application of the GATT and SCM Agreement. First, most fossil fuel subsidies are targeted at energy consumption (IEA et al., 2011). As a result, it is rare for such subsidies to be caught by the exhaustive definitional list contained in SCM Agreement,Footnote ²⁶ which applies more directly to production-based subsidies. Second, the definition of a subsidy contained in the SCM Agreement does not extend to ‘regulatory subsidies’, which stem from government decisions to omit the cost of negative externalities, such as local and global environmental pollution, in the price of goods (Birhanu Asmelash, Reference Birhanu Asmelash2015). This precludes WTO Members from being held to account for regulatory failures to price the externalities of fossil fuels. Second, consumption-based subsidies are usually provided economy-wide and therefore many fail to meet the test of specificity under the SCM Agreement (Birhanu Asmelash, Reference Birhanu Asmelash2015). Third, fossil fuel exports are less commonly subsidized; export taxes and restrictions are far more common (Birhanu Asmelash, Reference Birhanu Asmelash2015). The only specific example of a trade-oriented fossil fuel subsidy to receive significant academic attention is dual pricing, which is not covered by WTO rules (Marhold, Reference Marhold, Hancher, De Hauteclocque and Salerno2017).

On the other hand, a specific analysis of fossil fuel subsidies can reveal surprising vulnerabilities under the SCM Agreement. For example, contrary to the perception that renewable energy programs are the sole beneficiaries of local content requirements, Tordo et al. (Reference Tordo, Warner, Manzano and Anouti2013) demonstrate that these policies are also popular amongst countries that have recently started developing their oil and gas resources. This links to the central purpose of this paper, which is to show that fossil fuel subsidies are not immune from the features disciplined by international trade law. However, as Verkuijl et al. (Reference Verkuijl, van Asselt, Moerenhout, Casier and Wooders2017) note, the lack of official data on fossil fuel subsidies necessitates that any analysis occurs on a case-by-case basis, thus overcoming the information gaps presented by generalized analysis. This paper therefore advances the literature on fossil fuel subsidies by analysing in detail a specific example of government support for the coal industry.

5.1 Northern Australia Infrastructure Facility

The Northern Australia Infrastructure Facility (NAIF) is an investment agency, established by the Australian Government through the Northern Australia Infrastructure Facility Act 2016 (NAIF Act).Footnote ²⁷ The NAIF Act is a ‘major initiative’ of the Australian Government's White Paper on Developing Northern Australia (Our North, Our Future),Footnote ²⁸ which describes Northern Australia as the ‘trade gateway for all of Australia’.Footnote ²⁹ The functions of the NAIF are, inter alia, ‘to grant financial assistance to States and Territories for the construction of Northern Australia economic infrastructure’.Footnote ³⁰ The NAIF is governed by an independent board, whose investment decision-making power is guided by Ministerial directions contained in the Northern Australia Infrastructure Facility Investment Mandate Direction 2016 (NAIF Investment Mandate).Footnote ³¹

The NAIF Investment MandateFootnote ³² specifies seven mandatory criteria that projects must meet in order to be eligible for financial assistance, including, relevantly, that the proposed project: ‘involves construction or enhancement of economic infrastructure’; ‘will be of public benefit’; ‘is unlikely to proceed, or will only proceed at a much later date, or with a limited scope, without financial assistance’, and; ‘is located in, or will have a significant benefit for, Northern Australia’.Footnote ³³ In performing its duties, the NAIF is supported by Infrastructure AustraliaFootnote ³⁴ and the Export Financing Insurance Corporation (Efic).Footnote ³⁵

Financial records indicate that NAIF will spend approximately AU$3 million per annum outsourcing operations to Efic (Swann, Reference Swann2017), which will provide ‘all of the assessment and the grunt work to support [NAIF's] CEO and the board’.Footnote ³⁶ NAIF is supported by AU$5 billion in Australian Government funding,Footnote ³⁷ which it has the power to spend until 30 June 2021.Footnote ³⁸ The default mechanism for financial support under the NAIF is a concessional loan,Footnote ³⁹ but the Board is not prohibited from utilizing alternative financing mechanisms, other than providing equity.Footnote ⁴⁰

5.2 North Galilee Rail Line

The North Galilee Rail Line (the NGRL) is a 311.6 km rail line in north-central Queensland, Australia, with an estimated capital investment cost of AU$2.2 billion and a peak operating capacity of 100 million tonnes per annum (Mtpa) of coal.Footnote ⁴¹ The NGRL connects the Abbot Point international coal port with the proposed Carmichael coalmine and Galilee Basin, satisfying the Queensland Government's ‘pit-to-port’ development policy.Footnote ⁴² At peak operating capacity, the Carmichael coalmine will supply 60 Mtpa of thermal coal, while the remaining 40 Mtpa will be provided by other coalmines in the Galilee Basin, six of which are currently awaiting approval.Footnote ⁴³ In December 2016, media sources indicated that an Adani Group company had tendered for a AU$900 million loan from the NAIF to fund the NGRL (Viellaris, Reference Viellaris2016; Burke and Clark, Reference Burke and Clark2016; Robertson, Reference Robertson2016; Koziol and Wroe, Reference Koziol and Wroe2016). In March 2017, it was reported that the rail freight company Aurizon Holdings Ltd (Aurizon) (formerly Queensland Rail) launched a competing AU$1.25 billion bid (Stevens, Reference Stevens2017; Norris, Reference Norris2017). The NAIF board is currently considering these proposals.

6.1 Subsidy

Under the SCM Agreement, a subsidy will exist where the elements of Articles 1.1(a) and (b) are satisfied: i.e. a ‘financial contribution’ and a ‘benefit’ are conferred. The Appellate Body has confirmed that these are the two separate legal elements ‘which together determine whether a subsidy exists’.Footnote ⁴⁴

The concessional loan provided by the NAIF to fund the Abbot Point rail-link likely constitutes a subsidy under the SCM Agreement.

6.2 Subsidy – financial contribution

Article 1.1(a)(1) lists four circumstances where a financial contribution will exist. This is an exhaustive list,Footnote ⁴⁵ including, inter alia, ‘a direct transfer of funds’, foregoing government revenue, the provision of goods or services, and making payments to a ‘funding mechanism’ to carry out the aforementioned functions.Footnote ⁴⁶ The existence of any one of these factors is sufficient to meet the required legal standard.

The NAIF loan constitutes a financial contribution in the form of a direct transfer of funds.Footnote ⁴⁷ The NAIF is authorized to ‘grant financial assistance to States and Territories’;Footnote ⁴⁸ practically, this assistance flows to the proponent.Footnote ⁴⁹ Thus, while the Queensland State Government may act as an intermediary for the loan, it is uncontroversial that the Commonwealth Government is making a direct financial contribution to the proponent. The default financing mechanism for investments by the NAIF is a loan.Footnote ⁵⁰ The SCM Agreement explicitly includes ‘loans’ as an example of a ‘direct transfer of funds’.Footnote ⁵¹

The financial contribution must also be provided ‘by a government or any public body’.Footnote ⁵² A public body ‘must be an entity that possesses, exercises or is vested with governmental authority’.Footnote ⁵³ When a ‘statute or other legal instrument expressly vests authority in the entity’, this analysis will be straightforward.Footnote ⁵⁴

The NAIF is clearly a public body. The NAIF Act vests executive government decision-making power in the agency, with oversight from the relevant Minister. ^{Footnote 55}

6.3 Subsidy – benefit

On its face, the text of SCM Agreement Article 1.1(b) does not define when a benefit will be provided. The Appellate Body has held that the legal test for establishing the existence of a benefit is ‘whether the recipient has received a “financial contribution” on terms more favourable than those available to the recipient in the market’.Footnote ⁵⁶ The Appellate Body has highlighted that in order to perform this comparison, it is necessary to define the relevant market considering the demand and supply sides.Footnote ⁵⁷ Shadikhodjaev (Reference Shadikhodjaev2015) notes that the Appellate Body's reasoning on this point ‘suggests that government action towards the creation of a new market does not, in itself, amount to a subsidy’. This appears to be an important distinction; fossil fuel subsidies are designed almost exclusively to support existing energy markets rather than facilitate the creation of new ones.

With reference to the NAIF loan, it is clear that the subsidy would confer a benefit. The NAIF is explicitly authorized to provide loans on terms that are ‘concessions’ on those offered by the relevant market (‘commercial financiers’), including, inter alia, ‘longer loan tenor’, ‘lower interest rates’, ‘extended periods of capitalization of interest’, and ‘deferral of loan repayments’.Footnote ⁵⁸ To use the language of the Appellate Body, the NAIF loan constitutes ‘government interventions in support of certain players in markets that already exist’.Footnote ⁵⁹ The existence of any one of these concessional terms would thus be sufficient to constitute the conferral of a benefit.

6.4 Subsidy-specific

If a measure is found to be a subsidy, as defined by Article 1, it will not fall within the ambit of the SCM Agreement unless it is ‘specific’. Prohibited subsidies are specific by default as specified in Article 2.3, whereas actionable subsidies must satisfy the criteria of specificity set out in Article 2. Under the now defunct green-light protections, specific subsidies that provided assistance to disadvantaged regions were protected. However, specific subsidies are still lawful under the SCM Agreement, provided they are not prohibited or actionable. Given the emphasis of these classifications on preventing trade distortions, most regional subsidies that are not export oriented are likely to be permitted.

A subsidy may be specific if it is ‘limited to certain enterprises located within a designated geographical region within the jurisdiction of the granting authority’.Footnote ⁶⁰ In EC and Certain Member States – Large Civil Aircraft, the Panel concluded that ‘a subsidy available in a designated region within the territory of the granting authority is specific, even if it is available to all enterprises in that designated region’.Footnote ⁶¹ The same conclusion was reached by the Panel in US–Anti-Dumping and Countervailing Duties (China).Footnote ⁶² There is also no requirement for the geographic region to be drawn with reference to any existing or proposed political or economic boundaries; it ‘can encompass any identified tract of land within the jurisdiction of a granting authority’.Footnote ⁶³ The Panel's analysis of Article 2.2 outlined above was not challenged by any party on appeal to the Appellate Body.

Any financial contribution provided by the NAIF is likely specific within the meaning of SCM Agreement Article 2.2. The mandatory criteria for eligibility include that ‘the project is located in, or will have a significant benefit for, Northern Australia’.Footnote ⁶⁴ Northern Australia is defined as a geographic area that includes large swathes of Australia that are north of the Tropic of Capricorn, as well as some smaller sub-regions that are below this line.Footnote ⁶⁵ Australia may argue that the constraint set out in the eligibility criteria is not strictly based on geography, as projects that ‘produce significant benefits to Northern Australia’ may also be funded.Footnote ⁶⁶ However, the surrounding text of the enabling legislation,Footnote ⁶⁷ the Minister's introductory speech to Parliament,Footnote ⁶⁸ and the relevant White Paper on Developing Northern AustraliaFootnote ⁶⁹ reveal a clear bias in the way funding will be channeled towards projects in Northern Australia. This is underscored by the observation that of the 96 subsidies NAIF is actively considering, all are based in Northern Australia.Footnote ⁷⁰

6.5 Prohibited subsidies

There are two types of subsidies that are prohibited outright by the SCM Agreement: ‘subsidies contingent, in law or in fact … upon export performance’ and ‘subsidies contingent, whether solely or as one of several other conditions upon the use of domestic over imported goods’.Footnote ⁷¹ The blanket ban on these subsidies is justified on the principle that their raison d’être is to affect trade and that they are therefore likely to adversely affect the interests of other members (Van den Bossche and Zdouc, Reference Van den Bossche and Zdouc2013).

The text of Article 3.1(a) provides that subsidies may be de jure or de facto contingent upon export performance. As the Appellate Body noted in Canada–Autos,Footnote ⁷² a subsidy that is de jure contingent upon export performance can be revealed by the explicit or implicit meaning ‘of the very words of the relevant legislation, regulation or other legal instrument constituting the measure’.Footnote ⁷³ Where the legal instrument lays out a requirement for export contingency on its face, it will be relatively straightforward to demonstrate that the measure is in violation of Article 3.Footnote ⁷⁴

The standard of de facto contingency is more complex, but will be met ‘if the facts demonstrate that the subsidy is in fact tied to actual or anticipated exportation or export earnings’.Footnote ⁷⁵ In Canada–Aircraft, the Appellate Body observed that there are three distinct substantive elements to the test of de facto contingency contained in footnote four: (1) the ‘granting of a subsidy’, (2) that ‘is … tied to’, and (3) ‘actual or anticipated exportation or export earnings’.Footnote ⁷⁶ No single fact will be decisive,Footnote ⁷⁷ and footnote 4, second sentence of the SCM Agreement, expressly precludes using the ‘mere fact that a subsidy is granted to enterprises which export’ to justify a conclusion of contingency. Accordingly, de facto export contingency must be ‘inferred from the total configuration of facts constituting and surrounding a subsidy’.Footnote ⁷⁸ The test is objective,Footnote ⁷⁹ and factors that may be considered include: a Member's awareness that its own domestic industry is too small to absorb domestic production of a subsidized product,Footnote ⁸⁰ the export orientation of the recipient,Footnote ⁸¹ the level of a particular company's exports,Footnote ⁸² the terms of any loan contract,Footnote ⁸³ if a subsidy brings a product closer (e.g. faster or more likely to be manufactured) to sale on the export market,Footnote ⁸⁴ statements by government ministers, members of parliament, and government officials,Footnote ⁸⁵ and any other fact that demonstrates the export contingency of the subsidy.Footnote ⁸⁶ With respect to the volume of exports produced by a subsidy, the DSB has given no indication as to what level the increase must take before it will fall foul of the SCM Agreement. There is no hard limit; this factor will be weighed alongside others in determining whether a subsidy is de facto export contingent.

A NAIF loan for the NGRL cannot be challenged on the basis of de jure contingency on export performance. However, it is vulnerable to a finding of de facto contingency under Article 3.1(a) for the following reasons:

1. (a) Regarding the scale of Australia's domestic industry, it is evident that Australia does not require new coalmines for domestic consumption. Domestic demand for coal has ceased growing and is met by current on-shore production.Footnote ⁸⁷ In Australia–Automotive Leather II, the Panel was persuaded that the Australian Government's support for domestic hide production facilities was conditioned on sales targets that could only be met from exports.Footnote ⁸⁸ In the case of the NGRL, it is obvious that the loan will solely facilitate export trade and not domestic consumption.

2. (b) The scale of exports from the Galilee Basin is likely to be significant. The most recent estimate of Economic Demonstrated Resources in the Galilee is 5340 Mt, while inferred estimates place the volume of reserves at 23843 Mt.Footnote ⁸⁹ The NGRL will provide access to the Abbot Point international coal export port for the Galilee Basin, with a peak operating capacity of 100 Mtpa.Footnote ⁹⁰ In 2014/15, Australia's coal exports were 392 Mt.Footnote ⁹¹ Thus, the NGRL has the potential to increase Australia's coal export capacity by more than 25%. Assuming current market volumes, the operating capacity of the NGRL also represents about 10% of the current world trade in steam coal (IEA, 2016a).

3. (c) The subsidy for the NGRL brings the sale of products from the Galilee Basin substantially closer to export. Under the NAIF Investment Mandate, a mandatory eligibility criterion is that ‘the proposed project is unlikely to proceed, or will only proceed at a much later date, or with a limited scope, without financial assistance’.Footnote ⁹² As a result, the construction of the NGRL, and its attendant impact on Australia's coal export capacity, can be causally attributed to the provision of finance by the NAIF.

4. (d) Statements by government ministers have repeatedly emphasized the export orientation of the loan (Bennett, Reference Bennett2017; Killoran and McCarthy, Reference Killoran and McCarthy2016).

5. (e) The White Paper underpinning the NAIF Act makes it clear that the development of Northern Australia means ‘unlocking new investment’ through exports and international trade.Footnote ⁹³

The existence of any one of these factors alone would likely be insufficient to allow a finding of export contingency. However, the aggregation of factors in this case study is unique in WTO jurisprudence. The significant scale of exports produced by this subsidy and the weight of other evidence in this matter points strongly towards a finding of de facto export contingency.

6.6 Actionable subsidies

Actionable subsidies are not a prima facie violation of the SCM Agreement. However, they are open to challenge should they cause ‘adverse effects’ to the interests of other Members.Footnote ⁹⁴ As a first step, an actionable subsidy must satisfy the definition of a subsidyFootnote ⁹⁵ and meet the requirement of specificity.Footnote ⁹⁶ Adverse effects may then be established through, inter alia, ‘serious prejudice to the interests of another Member’.Footnote ⁹⁷

6.7 Serious prejudice

Serious prejudice is concerned with ‘negative effects on a Member's trade interests’ including ‘lost … export volume or market share’.Footnote ⁹⁸ Accordingly, serious prejudice to the interests of another Member includes the circumstance where ‘the effect of the subsidy is to displace or impede the exports of a like product of another Member from a third country market’.Footnote ⁹⁹ In US–Upland Cotton,Footnote ¹⁰⁰ the Panel ruled that ‘prejudice’ occurs when ‘a detrimental impact on a complaining Member's production of, and/or trade in, the product concerned’ can be shown. The threshold level of ‘serious’ is met where the effect is ‘important, not slight or negligible, or meaningful’.Footnote ¹⁰¹ The test of seriousness cannot be analysed prospectively, as observations must be made by a ‘reference period’.Footnote ¹⁰²

The NAIF loan is highly likely to result in serious prejudice to the interests of another Member, through the displacement of the Member's trade in coal. While any of India's coal trading partners could be considered (e.g. Russia), Indonesia is the party likely to be most severely affected by an influx of Australian coal onto international energy markets. The NGRL will have an operating capacity of 100 Mtpa, the majority of which is expected to be exported to India. This is about 74% of Indonesia's current steam coal exports to India (IEA, 2016a), and would represent a substantial displacement of existing exports.Footnote ¹⁰³ Therefore, the scale of exports from the Galilee Basin is likely to meet the test of seriousness.

6.8 Like product

In order to prove serious prejudice to Indonesia, it must be demonstrated that Indonesian and Australian coal are ‘like products’.Footnote ¹⁰⁴ The phrase ‘like products’ is defined in footnote 46, which provides that, ‘[t]hroughout this Agreement the term “like product” (“produit similaire”) shall be interpreted to mean a product which is identical, i.e. alike in all respects to the product under consideration, or in the absence of such a product, another product which, although not alike in all respects, has characteristics closely resembling those of the product under consideration.’

The Panel in Indonesia–Autos referred with approval to the Appellate Body Report in Korea–Alcoholic Beverages (1999), which determined that under the GATT, ‘the issue of “like product” must be considered on a case-by-case basis’.Footnote ¹⁰⁵ In applying the likeness test under the SCM Agreement, the Panel then looked closely at the physical characteristics of the two products, but noted that it was not precluded from considering other criteria.Footnote ¹⁰⁶ As such, it appears that while the physical characteristics are an important feature of the test for likeness under the SCM Agreement, the other elements highlighted by the Working Group on Border Tax Adjustments ought to also be considered, specifically: ‘the product's end-uses in a given market; consumers’ tastes and habits, which change from country to country; the product's properties, nature and quality’.Footnote ¹⁰⁷ Nevertheless, it is important to acknowledge that the precise extent of the like product analysis under the SCM Agreement has not been identified by a Panel or the Appellate Body.

In terms of end-uses, coal from Australia and Indonesia is currently used in India for identical purposes: electricity generation.Footnote ¹⁰⁸ Coal from the Galilee Basin would also be used for the same purpose.Footnote ¹⁰⁹ The consumer taste and habits criteria is less relevant for analysing coal as a product, and may be collapsed into the consideration of a product's quality, which is usually the differentiating factor for electricity producers.Footnote ¹¹⁰

Two important factors determining the quality of coal products in India are energy and ash content.Footnote ¹¹¹ Indonesian export coal is typically 2–12% ash content (Ewart, Vaughn and Marston & Marston Inc., Reference Ewart and Vaughn2009), whereas coal from the proposed Carmichael Coal Mine, in the Galilee Basin, is expected to be about 26%.Footnote ¹¹² In terms of calorific content, Indonesia export coal is typically 5 300-6 700 kCal/kg (Ewart, Vaughn and Marston & Marston Inc., Reference Ewart and Vaughn2009), whereas the Carmichael Coal Mine expects to export coal with 4 950 kCal/kg.Footnote ¹¹³ Although the quality of the two coals is slightly different, both can be used at the same power stations in India.

Finally, although tariff classifications will not be determinative,Footnote ¹¹⁴ it is relevant to note that the Indonesian and Australia's schedule of concessions delineates between ‘anthracite’, ‘bituminous coal’, and ‘coal (excluding anthracite and bituminous)’.Footnote ¹¹⁵ Simply put, these classifications distinguish between different types of coal, used for distinct purposes. Relevantly, almost all thermal coal is classed as bituminous,Footnote ¹¹⁶ indicating that Indonesia and Australia regard coal used for electricity production as ‘like products’, despite its varying quality. Accordingly, while there are subtle differences between Australian and Indonesian coal, it is highly likely that the products would be found to be like.

6.8 Like product: renewable energy

Although electricity produced by fossil fuels and renewable energy are often in direct competition, this does not allow renewable energy exporters to claim serious prejudice as a result of subsidized Australian coal products. First, Australia is exporting raw coal, not electricity. Thus, even if electricity produced by fossil fuel and renewable sources are like products, this finding is irrelevant to the matter at hand. Rather, for renewable energy exporters to claim serious prejudice, they must demonstrate that renewable energy products (e.g. solar panels and wind turbines) are like coal. This is highly unlikely under the current test for likeness (Wold et al., Reference Wold, Wilson and Foroshani2012).