1. What is a subsidy? And what data are available?

This paper depends on two conceptual steps. First, we need a definition of ‘subsidy’. The notion that some actions of governments are subsidies and others not may seem simple in economic theory, leading to claims that one definition would do, but in practice definitions serve particular policy and analytic purposes. What can be measured – and disciplined – depends on those definitions. Second, the availability of data has been shaped by the definitions, with the result that some debates about the extent and scope of energy subsidies are due to variations in how they are measured (Steenblik, Reference Steenblik and Pauwelyn2010). In thinking about a definition, we need not work though the full complexity of the meaning of the term (Jones and Steenblik, Reference Jones and Steenblik2010), but we need to make it operational for our purposes. That is, we should be able to identify fossil fuel subsidies, with common metrics to allow comparison.

The international organizations charged with action on fossil fuel subsidies differ in their definitions and in their calculation method. Some overlap; some parts are significantly different. The WTO definition of subsidies in general is based on legally binding reciprocal obligations aimed at improving the conditions of multilateral trade through disciplines on government policy. Other international organizations have more latitude to use specific definitions of fossil fuel subsidies designed for analytic purposes or, especially in the case of the International Monetary Fund (IMF), to support a process of domestic reform. In this section, we describe how each organization defines subsidies in general or fossil fuel subsidies in particular. We conclude the section with a discussion of the implications for the availability of data.

Summary: comparing definitions and data

Given these differing definitions used by international organizations, it is hardly surprising that what they report differs. The IEA, reflecting its long expertise in monitoring energy markets, uses market data and can, as a result of its methodology, miss government measures that clearly amount to economic support, or it might include elements that constitute a price gap that is not caused by government action.

Definitional ambiguities notwithstanding, our review of notifications under the ASCM found that Members have sporadically notified a variety of support measures (grants, loans, tax and excise exemptions, and others) that amount to fossil fuel subsidies, even if not labeled as such. The United States and Japan have been particularly forthcoming; the United States even notified several sub-federal subsidization programs, notably an oil- and gas-production tax credit offered by the state of Alaska to promote oil exploration and drilling. The notified Japanese measures were mostly related to overseas oil and gas exploration and to research and development (R&D). India has notified several sub-federal fossil fuel subsidy programs in the fishery sector, typically consisting of rebates on fuel taxes for ‘poor fishermen’. (Fuel subsidies, which may account for 20% of fisheries subsidies, were a major sticking point in the Doha Round negotiations.) The most recent Russian subsidy notifications include only one fossil fuel program, although independent researchers have been able to identify a range of fossil fuel subsidies based on the SCM definition (Gerasimchuk, Reference Gerasimchuk2012).

The OECD Inventory (OECD, 2012) relies on publicly available government data, which is information that could have been notified to the WTO. A comparison found that in most cases the OECD Inventory covered many more support measures than the WTO notifications, and in more detail, as shown in Appendix 3. Where the OECD inventory contains 640 measures for the 2008–2013 period, the WTO has only 64. Particularly on the sub-federal level, the OECD Inventory, compiled by the Secretariat, included more than governments themselves notified to the WTO, especially for Australia, Canada, and the United States, although the United States was able to notify some sub-federal level subsidies to the WTO. Our comparison in Appendix 3 includes reports to the G-20, since leaders have called for voluntary reporting and peer review. The G-20 data are also inadequate, finding only 35 measures thought to fall within the reporting obligation, because states are required to report only on inefficient fossil fuel subsidies that contribute to wasteful consumption, which is of course open to interpretation. Some inconsistencies between the G-20 and the WTO are puzzling, perhaps owing to the differing definitions, or the nature of subsequent surveillance. We see detailed reporting of fossil fuel subsidies by Japan in the SCM Committee, but in the G-20, Japan reports that it has no inefficient fossil fuel subsidies. By contrast, Argentina, Canada, Indonesia and Italy reported fossil fuel subsidies to the G-20, but not to the WTO.

One of the strengths of the WTO is the opportunity for Members to hold each other to account, for example by reviewing subsidies notifications in the SCM Committee.

As Table 1 shows, our probe of the most recent five years' experience in the committee found that Members did raise questions about fossil fuel subsidies notification. Some questions were related to the identification of beneficiaries of the subsidies and eligibility criteria, information that according to the WTO questionnaire should have been provided with the notification, but the committee has not yet had a thorough discussion of fossil fuel subsidies. While not directly relevant to the focus of this paper, we note in passing that the record of notification and surveillance of renewable energy subsidies is somewhat stronger, as is shown in the three right columns of Table 1.

Table 1.

Questions asked about fossil fuel subsidies programs and renewable energy subsidy programs in the SCM Committee, 2008–2013

Note: * All fossil fuel subsidy programs in the Indian notification G/SCM/N/220/IND of 11 May 2012 were state-level programs providing fuel subsidies to fishermen.

Source: SCM Committee minutes G/SCM/M/62 through G/SCM/M/84.

Both the rate of notification of fossil fuel subsidies and the paucity of questions in the committee are no doubt a reflection first of the ambiguities with respect to the definition of subsidies discussed above, and of the usual ‘glass house’ syndrome. Members do notify some fossil fuel subsidies (perhaps those that are least likely to be challenged?), and they do ask for more detailed information on the notified support measures. If the information available from the WTO was comprehensive in its coverage of countries and measures, this paper would not be necessary. But the information on fossil fuel subsidies is inadequate, as it is for subsidies generally (on the inadequacy of subsidies notification in general, see Collins-Williams and Wolfe, Reference Collins-Williams and Wolfe2010).

In sum, the WTO already discusses fossil fuel subsidies, but inconsistently and based on data that are less complete than those available to other international organizations, but it has a treaty-based definition of ‘subsidy’, and robust surveillance mechanisms that could be better used. If we wish to improve discussion of fossil fuel subsidies, we need to think about how to improve the data available, but the prior question is where such data could be used at the WTO. In the next section, we discuss why the SCM Committee is not a candidate.

2. What happened to ‘non-actionable’ subsidies in the WTO?

The WTO struggles with disciplines on subsidies. On the one hand, subsidies have the potential of creating unfair competitive advantages, leading to trade distortions. On the other hand, subsidies can be a useful instrument for economic and social purposes. Unfortunately, Members do not always agree on which is which. Both the ASCM and the Agreement on Agriculture use a traffic light system. In Agriculture, Article 7 and Appendix 2, known as the Green Box, cover domestic support that is thought to be ‘minimally trade distorting’. The analog in the ASCM was Article 8 on ‘non-actionable’ subsidies, but that provision of the agreement has lapsed. A revived Article 8 that included energy subsidies is one of the possible mechanisms that would allow Members to notify their fossil fuel subsidies in a way that would increase the availability of information, and provide the basis for discussion on such subsidies in the SCM Committee.

While a revival of the category of non-actionable subsidies is often discussed within the context of government support for renewable energy (Howse, Reference Howse2010), in this section we explore whether a revival of Article 8 would be a way to bring more information on fossil fuel subsidies to the SCM Committee.Footnote ⁵ Notification of a measure as potentially non-actionable would not shield it from challenge in the Committee or the dispute settlement system.Footnote ⁶ Such notification would not increase the risk of a complaint (we discuss transparency and the risk of disputes below) but it would allow Members to discuss whether the harm to or benefit for the planet of a particular subsidy is a relevant policy consideration. In this section, we review the negotiating history of Article 8, concluding that a revival is improbable.

From the start, the Uruguay Round negotiations that created the ASCM reflected ideological differences among Members on subsidies. The United States, unlike other leading Members, such as the EU, Switzerland, and Canada, had never been in favour of subsidization in general because of possible market distorting effects (Stewart, Reference Stewart1993: 875–877). But negotiators recognized that because subsidies can be used to correct market failures and can be desirable when pursuing certain economic and social objectives, not all subsidies should be prohibited or subject to countervailing duties (GATT, 1987b). In the early days of the Uruguay Round, doubts were even raised as to whether Members would give up their sovereign right to use subsidies (GATT, 1987b: 3).

In 1988, Switzerland proposed a traffic light approach, establishing three categories of subsidies, and attaching different legal effects to each of them (GATT, 1988b; Croome, Reference Croome1998). A first category consists of export subsidies and local content subsidies that are aimed at giving domestic industries competitive advantages. These subsidies would be prohibited. A second category consisted of actionable subsidies and could be subject to countervailing duties when it has been established that material injury is caused. A third category was the ‘green light’ category and consisted of subsidies that do not have a trade-distortive effect or that should be allowed because they serve a purpose that would be accepted as legitimate.

While there was naturally a relatively broad consensus on the non-actionability of generally available subsidies, perhaps because they were considered to have a minimally distortive effect on trade, the list of other non-actionable subsidies was more problematic (GATT, 1989d; GATT, 1989e; GATT, 1990a). Some negotiators questioned why the ‘green light’ category had to exist in the first place (GATT, 1989c); some wanted a strict definition of non-actionable subsidies, and others advocated for a broad category in which the intent of a subsidy has to be taken into account when assessing its non-actionable status. Other Members underlined the need for taking the effect of a subsidy into account, not its intent (GATT, 1988a; GATT, 1988c: 11). Developing countries wanted to include structural adjustment subsidies under the non-actionable category (GATT, 1989f); other countries did the same for regional subsidies, and the EC, Canada, and Switzerland were strongly in favour of carving out environmental subsidies and R&D subsidies (GATT, 1987a; GATT, 1989a; GATT, 1989b; GATT, 1989g; GATT, 1989h).

During preparations for the Brussels Ministerial Conference in 1990, the United States remained strongly opposed to the category of non-actionable subsidies (Croome, Reference Croome1998: 173–174), while the EC pursued a more nuanced perspective in the sense of being prepared to allow subsidies with no significant effect on international trade (Stewart, Reference Stewart1993: 875–877). The US approach to non-actionable subsidies changed after Bill Clinton became President in 1993. The general view that subsidies were bad was more nuanced than before because the new administration saw R&D as particularly important for the US economy (Doane, Reference Doane1995). This shift in the position of the United States helped to facilitate a conclusion to the negotiations on the ASCM.

Uncertainty about the non-actionable category was reflected in Article 31, which required a review after five years with a determination of whether to extend or modify Article 8. In the last meeting of the SCM Committee before the 1999 Ministerial Conference in Seattle, some Members (e.g., Switzerland and Canada) worried that not extending Article 8 would give the wrong signal to the public and would ‘reduce the ability of the ASCM to address environmental concerns’ (Bigdeli, Reference Bigdeli2011: 9). Several developing country Members countered that extension of the provision in its present form would be unacceptable since the category provided advantages only for developed countries (WTO, 1999). The United States had mixed views on the extension, noting that in the five years of its provisional application, no non-actionable subsidy had been notified (Bigdeli, Reference Bigdeli2011). In the absence of consensus, the category lapsed almost unnoticed shortly after the 1999 Seattle Conference.

Members showed little interest in negotiations on any subsidies issues during the Doha Round, with the exception of fisheries subsidies, but Venezuela, Cuba, and the EU each submitted proposals on non-actionable subsidies. The proposal of Venezuela emphasized that Article 8 could not be reactivated without modification to ensure that developing country structural adjustment subsidies would not be subject to countervailing measures or challenges in a dispute (WTO, 2002a). Developed countries were skeptical of the proposal, and other developing countries argued that the provision would have to be reformulated and be more balanced with regards to the interest of developing countries (WTO, 2003a; WTO, 2003b; WTO, 2003c). The EU proposal to reactivate Article 8 started from the perspective of why the category was never used (WTO, 2002b). The discussions quickly showed that the diverging opinions that were already present in the Uruguay Round negotiations had not disappeared. Similarly, the discussions on fisheries subsidies, where some proposals were made to classify ‘good’ and ‘bad’ subsidies, resulted in the same outcome: countries could not agree on the classification criteria and expressed considerable scepticism about whether the real purpose of the proposal was conservation of fish stocks, not the effects of subsidies on trade.

The different opinions on which subsidies could be considered non-actionable and which ones not indicate that a revival of the non-actionable category only for the purpose of environmental subsidies would be difficult, and would inevitably lead to long negotiations with all kinds of other subsidies proposed for the non-actionable category (Horlick, Reference Horlick2013). And yet this history shows that negotiators have long understood that subsidies that have an environmental effect are an issue. For years after the lapse of Article 8 it seemed that governments acted as if the provision was still in place; that is, Members understand the logic of why certain subsidies ought to be non-actionable, and have acted accordingly. Until 2010, cases on energy-related subsidies were rare, perhaps reflecting a tacit agreement not to challenge programs designed to promote renewable energy (Rubini, Reference Rubini2012). That tacit agreement now seems fragile.

3. Role of the Trade Policy Review Mechanism (TPRM)

With no prospect of restoring the non-actionable category to the ASCM, a helpful alternative forum to generate more information about fossil fuel subsidies with an opportunity for discussion is the Trade Policy Review Body (TPRB).Footnote ⁷ The central objective of the TPRB is ‘to contribute to … the smoother functioning of the multilateral trading system, by achieving greater transparency in, and understanding of, the trade policies and practices of Members'. The TPRB generates three sorts of reports – the periodic Trade Policy Review (TPR) of each Member (WTO, 2011a: paras. 178 ff.), the annual review of the state of the trading system, and the monitoring reports on measures taken in response to the financial crisis. In these reports to the TPRB, issued on the authority of the Director-General, not Members, the Secretariat sometimes warns or expresses worries on the basis of its analysis, but never criticizes Members explicitly, and never comments on their rights and obligations under the WTO agreements. Discussion in the TPRB therefore does not imply either that a measure is or is not actionable.

Monitoring subsidies falls within the mandate of the TPRB. Subsidies, called ‘General Economic Support Measures’, have been a particular concern in the crisis monitoring exercise because they clearly increased after September 2008, with undoubted effects on the trading system for good as well as ill, but neither the SCM Committee nor the TPRB has enough data to form a complete picture. Annex 4 of the most recent annual monitoring report (WTO, 2012c) is a valiant attempt, but the Secretariat observed that assessment is inevitably biased because of the paucity of information provided by Members, sometimes because they claim that the relevant supports are not ‘new measures’ and hence not covered by the process. The TPR reports on individual Members face the same difficulty. Data on support measures are found in various sections of the reports rather than in one place. While the reports are among the few sources of systematically collected subsidies data, they are not strictly comparable, the Secretariat observes, because the absence or presence of data in the report on any one country may be an artifact of the information problems, rather than an indication that the member does or does not maintain such subsidies (WTO, 2012c: para. 118).

The reports already take notice of energy subsidies. Indonesia's energy consumption subsidies are large, and well known, at close to 29% of central government spending in one recent year (WTO, 2013c: para. 1.11). The TPR report by the Secretariat includes information on energy subsidies and the problems of their reform that were not included in the report that was prepared by Indonesia. In a recent periodic monitoring report, Annex 4 on ‘General Economic Support Measures’ in the Director-General's annual monitoring report (e.g. WTO, 2012c) is a valiant attempt, but the Secretariat observed that assessment is inevitably biased because of the paucity of information provided by Members, sometimes because they claim that the relevant supports are not ‘new measures’ and hence not covered by the process. There is a confirmed Brazilian subsidy program for ethanol mentioned, and an unconfirmed US tax credit for the wind power industry (WTO, 2013b: 68 and 72).

Scholars recognize the potential of the TPRM as a forum in which both the trade policies of Members and issues that go beyond trade to include the broader economic policy of a country can be discussed without prejudice to legal compliance with WTO obligations (Chaisse and Matsushita, Reference Chaisse and Matsushita2013: 13–14; see also Charnovitz, Reference Charnovitz2010: 755–756), but opinions diverge about the scope of its mandate and on how to improve the mechanism. Developing countries are hesitant to give the monitoring and surveillance mechanisms in the WTO more weight out of a fear that the bigger, more powerful Members would prescribe their trade policies (Ghosh, Reference Ghosh2010). Some academics consider the current mandate too restrictive (Ghosh, Reference Ghosh2010: 437–438), others hold that the TPRM within its current mandate can be improved with respect both to information gathering and the participation of civil society (Zahrnt, Reference Zahrnt2009: 11–12). The Director-General stretched the mandate in 2009 to justify monitoring measures related to the financial crisis as requested by the G-20 (Wolfe, Reference Wolfe2012). In 2011, the Ministerial Conference formally recognized the crisis-monitoring exercise and asked that it be continued and strengthened (WTO, 2011b).

The core of each TPR report is based on the notifications from Members, but each report depends on a far wider range of information. The Secretariat collects data from official sources (questionnaires to Member under review) and non-official sources, including from other international organizations, media reports, and NGOs (Ghosh, Reference Ghosh and Deere-Birkbeck2011: 431; Chaisse and Matsushita, Reference Chaisse and Matsushita2013; Hoekman, Reference Hoekman2011: 18–19). To ensure accuracy, the Secretariat seeks verification of the data from non-official sources when discussing the draft of its report with the Member (WTO, 2011a: para. 180).

The crisis monitoring reports and annual reports on the trading system use a similar method. The 2012 monitoring report notes that responses were received from around 60% of the 43 delegations (counting the EU as one) who were asked to verify information received from other sources (WTO, 2012c: para. 19). Many delegations either verify the information, or provide more accurate information, but some responses are less helpful: certain delegations request the Secretariat to delete all information because the measures are not based on legal instruments, or they simply refuse to verify any information whose source is the public media (WTO, 2013b: 21). In such cases, the Secretariat cannot use the data, but if they have evidence that a disputed measure is based on legislation, or if a delegation fails to respond to a request for verification, the measure will be still included in reports as ‘unverified’.

We think that all three sorts of TPR reports ought to have a specific section on fossil fuel subsidies. Given the poor record of notification on subsidies in general, and the ambiguities about whether subsidies that might not be specific even ought to be notified, the Secretariat needs help in generating the data from non-WTO sources.

4. The role of NGOs and third-party notification

Many international organizations make use of third-party data (Mitchell, Reference Mitchell1998). Of particular relevance, many NGOs have a recognized role in contributing to the work of multilateral environmental agreements (MEAs) where they can enhance the knowledge base available to the MEA (Oberthür et al., Reference Oberthür, Buck, Müller, Pfahl, Tarasofsky, Werksman and Palmer2002). Of course, many of the relevant NGOs are based in or funded from the G7 countries, which makes the emerging economies wary (Kahler, Reference Kahler2013: 719), but their role can be essential. One of the strengths of the Convention on International Trade in Endangered Species of Wild Fauna and Flora (CITES) is its ability to work closely with Trade Records Analysis of Flora and Fauna in Commerce (TRAFFIC), an NGO whose purpose is essentially to support CITES by providing information on illegal trade in endangered species (Oberthür et al., Reference Oberthür, Buck, Müller, Pfahl, Tarasofsky, Werksman and Palmer2002; Wolfe and Baddely, Reference Wolfe and Baddeley2012: 7–12). Other examples include Greenpeace International providing information on illegal trade to the Secretariat of the Basel Convention on Transboundary Movements of Hazardous Wastes and their Disposal (Okaru, Reference Okaru2011), or Greenpeace and the WWF providing information to the scientific body of the UNFCCC. The Alternative Fluorocarbon Environmental Acceptability Study provides information on the production of ozone-depleting substances used by Parties to the Montreal Protocol. Finally, the Ramsar Convention has concluded several memoranda of cooperation with NGOs and has given certain duties to the International Union for Conservation of Nature (IUCN) (Oberthür et al., Reference Oberthür, Buck, Müller, Pfahl, Tarasofsky, Werksman and Palmer2002).

Independent researchers can generate useful reports on fossil fuel subsidies from publicly available data – see for example Oosterhuis (Reference Oosterhuis, Alexandru, Blumberga, Bobinaite, Cauchi, Nikolova and Papyrakis2013) – but the effort can be resource-intensive. Nevertheless, as long as data exist in the public domain, anybody who understands subsidies and taxes (and the local language, if important) can put together a set of estimates. One obvious source of data is the OECD, as suggested by Tim Josling and Klaus Mittenzwei (Reference Josling and Mittenzwei2013) in an approach to the problem of incomplete and late notification of agriculture subsidies. They argue that OECD Producer Support Estimates (PSE) data on farm subsidies are complementary to WTO notification data, despite the slightly differing conceptual bases. A virtue of exploiting that complementarity would be the ability to calculate provisional or ‘pre-notifications’ (synthetic notifications for years for which official notifications are not yet available) from recent OECD data, since the OECD dataset is updated autonomously, and nearly continuously, meaning its data are much more timely than WTO agriculture notifications. The issue has apparently provoked lively debate in the TPRB, participants say, since developing countries led by Brazil and India argue that the WTO should not use OECD data (WTO, 2009). However, that should not stop NGOs from doing so.

Where information is incomplete or unclear, NGOs can resort to ‘provocative notification’, a process of notifying what information is available or has been dug up. Where this information is incorrect – and in particular where it suggests subsidies larger than they are in reality – this kind of notification provides an incentive for the country in question to provide the correct information, if only to counter the impact of the flawed NGO notification. The GSI's reporting of EU biofuel subsidies provoked the Commission to respond by giving more accurate and up-to-date figures than those available to the GSI (Charles et al., Reference Charles, Gerasimchuk, Bridle, Moerenhout, Asmelash and Laan2013). The transparency objective had been achieved.

In the WTO, NGOs can contribute as knowledge brokers helping to facilitate the discussion between Members and contribute to better policies (Perez-Esteve, Reference Perez-Esteve2012). In the dispute settlement system, this function is already formalised by allowing NGOs to submit amicus curiae briefs. NGOs also have a role to play, however, before issues reach the dispute settlement phase. Guidelines adopted by the General Council in 1996 (WTO, 1996) include the possibility of ‘informal arrangements to receive the information NGOs may wish to make available for consultation by the interested delegations’. While no formal process exists (yet) for NGOs to submit notifications, and the general economic support measures identified in the context of the trade monitoring exercise are not included in the public database, the Secretariat does use NGO data that are relevant. For example, in the crisis monitoring undertaken by the TPRB, the Secretariat makes use of data published by Global Trade Alert (Wolfe, Reference Wolfe2012; Wolfe, Reference Wolfe2013). WTO rules for engagement with NGOs are unlikely to become more formal (Pandey, Reference Pandey2013), but current rules do allow sufficient scope for NGOs to provide information on energy subsidies in a form that would be useful for the WTO. We know that if an NGO provides good data, the Secretariat can and will use it. The 2014 TPR report on Malaysia has an extensive discussion of energy subsidies, over half the total (WTO, 2014). The report draws heavily on third-party information, notably a report by our GSI colleagues (IISD, 2013).

5. Proposal for a new template for notification of fossil fuel subsidies

If information provided by an NGO is to be useful, it has to be in the language of the WTO; that is, the data must fit the ASCM definition of subsidy, must be comparable across measures and countries, and must use a standard template comparable to a WTO notification. The information in a WTO notification is the answer to a question posed implicitly or explicitly by a WTO agreement. The generic definition of a notification in the WTO Glossary is ‘a transparency obligation requiring member governments to report trade measures to the relevant WTO body if the measures might have an effect on other Members’. In most cases, the notification is structured by a questionnaire or template developed by the relevant body, and increasingly the resulting data are being placed in the new Integrated Trade Intelligence Portal (I-TIP), a publicly available and searchable database.

The virtue of templates is that their use ensures that the same types of data are collected for every Member, and every subsidy, and that the results can be compiled in a database that allows simple comparison and aggregation. If NGOs are to contribute data on energy subsidies, they must provide it in such a standard format. Having a template helps provide a clear path for researchers. As more people use it and test it, there is a double dividend: feedback helps improve estimation and reporting methods, and the data become more comparable across different studies – which can only happen if the WTO or some other body ensures that all such third-party notifications are publicly available.

Our colleagues in the Global Subsidies Initiative (GSI) have long argued that the subsidies questionnaire used by the SCM Committee (discussed above) is inadequate, and have proposed a template for governments to use in preparing subsidies notifications to the WTO in a more structured and consistent manner (Steenblik and Simón, Reference Steenblik and Simón2011). The GSI encouraged governments to use the template for their notifications to the WTO, and carried out a project to assess and organize German subsidies using the template (Thöne and Dobroschke, Reference Thöne and Dobroschke2008). For reasons discussed elsewhere (Wolfe, Reference Wolfe2013: 18–19), WTO notification remains weak, and Members do not use the template. When voluntary reporting by governments does not work because they seemingly lack the incentive to notify, third parties have a role to play in identifying and collecting information about subsidies that would be accessible in the WTO format (Thöne and Dobroschke, Reference Thöne and Dobroschke2008: 23–24). We think that NGOs and individual researchers could use a modified version of the GSI template, particularly in countries with good budgetary transparency or countries in which information is available from other international organizations (whether NGOs have sufficient means to notify this information to the WTO is a question we have yet to explore.)

The GSI is concerned with fossil fuel subsidies and energy subsidies more broadly because of their effect on environmental quality, economic development, governance, and hence on sustainable development in general. The GSI therefore applies an expansive meaning to fossil fuel subsidies, but one based on the definition of a subsidy formulated in the ASCM – see the checklist in Appendix 1 below. The main difference with the ASCM is that the GSI does not include a benefit analysis (ASCM Article 1.1(b)) when identifying energy subsidies, nor does it limit its research to specific subsidies (Article 2). Consequently, a broad range of economic support measures are included in the GSI's work, including both production and consumption subsidies.

The template that we propose in Appendix 2 is thus based on the WTO definition, taking into account the subcategories that the GSI developed to facilitate the understanding of the definition. While initially developed by the GSI for use by governments, this template as modified can serve as the basis for notifications of fossil fuel subsidies by NGOs.

The proposed template contains two parts. The notification summary in the first part provides a quick overview of what different subsidization programs look like in a given country, organized in a table according to a chosen classification system. The Harmonized System (HS), which would organize subsidies by product, has the advantage of being familiar to most WTO Members. The International Standard Industrial Classification of All Economic Activities (ISIC) would allow organization of energy subsidies by economic activity (Steenblik and Simón, Reference Steenblik and Simón2011). National account statistics of countries are already often structured according to economic activity (agriculture, fishing, mining, manufacturing, retail, etc.). Using ISIC also offers the advantage of having a discussion on subsidies where Members can immediately compare the amount of subsidies related to a particular economic activity, rather than to a product. Finally, when subsidies are organized according to the ISIC system, Members could discuss what kinds of activities related to fossil fuels they do not consider problematic and those that they do, which should be subject to reductions.

The second part is a detailed description by subsidy program of the conditions and objective of the subsidy. The checklist of energy subsidies (Appendix 1) allows researchers to identify consumption and production subsidies for fossil fuels, as well as subsidies for renewable energy (Beaton et al., Reference Beaton, Gerasimchuk, Laan, Lang, Vis-Dunbar and Wooders2013). The type of subsidy will subsequently determine the method of measuring the amount. For example, direct government expenditure can easily be defined by looking at the government's budget, while tax exemptions have to be measured against a reference level. The concept of ‘price support’ in WTO jurisprudence includes direct government intervention in the market intended to fix the price of a good at a particular level, for example, through the purchase of surplus production when price is set above the market-clearing level (see WTO, 2012d: para. 7.85).

In line with the GSI approach, our template does not ask if a subsidy is specific – a judgment that in the end can only be made by Members through questions in the SCM Committee or the dispute settlement system on a case-by-case basis. Our objective, and a further reason for using the TPRM, is to shine a light on all subsidies that cause market distortions and therefore have an impact on the trading system. Any support measure from a central or sub-central government that is applicable to natural gas, oil, coal, or peat will fall under that category and could be notified by an NGO. We have found in examining subsidy notifications that Members hardly ever provide data to calculate the trade effect of a subsidy, but some discussion on this point will be essential in each case in order to help Members see the WTO relevance of the measure. Both consumer and producer fossil fuel subsidies have an effect on the trading system. Producer subsidies provide a competitive advantage to the recipient. Consumer subsidies favour local industries that are energy intensive. If energy as an input becomes cheaper through fossil fuel subsidization, the producers of energy intensive goods will gain an advantage.

Providing for the amount of a subsidy in the notifications would allow for better discussions. Small subsidies might not be a concern to other Members, while very large subsidization programs might provoke questions and debate. Direct government expenditure will be easier to measure than tax exemptions. NGOs can rely on different methods to calculate subsidies or rely on the data of the OECD, IMF, and IEA to provide the information for this column.

Finally, verification by the Secretariat will be easier if the NGO notification contains the source of the information, including legislative authority for the measure, and any other documentary evidence. The notification should also contain the source for any quantitative data or the calculation of the amount of a subsidy.

6. Conclusion: shining a light on fossil fuel subsidies

Trade and the environment are entwined, and one nexus is subsidies. The preamble of the WTO Agreement records Members' commitment to enhanced trade ‘allowing for the optimal use of the world's resources in accordance with the objective of sustainable development’. Our aim in this paper was to consider the role of transparency in advancing toward this goal of a trading system that meets the needs of the present without compromising the ability of future generations to meet their own needs. This preambular goal remains hard to define in legal obligations, even if new negotiations were in prospect, but we think Members can nevertheless hold themselves accountable for progress in this direction. We think that civil society can help, as it did in the long process of developing the information and analysis to underpin the framing of fisheries subsidies as an appropriate subject for negotiations in the WTO.

Fossil fuel subsidies offer an example of a significant climate change problem where new trade rules are unlikely in the short term, and where an NGO contribution to increased transparency could make a difference. Fossil fuel subsidies divert precious resources from other pressing needs and add colossal sums to national debt. These perverse subsidies provide a strong incentive to prefer carbon-based fuels to the alternatives, and they affect resource allocation in ways that are unfavourable to other, cleaner forms of energy. Their use clearly affects patterns of international trade, even if specific harm to trading partners is not always easily identified. The ASCM ought to be a useful tool to discipline such subsidies, but this agreement is not effective: Members either fail to notify their subsidies, or notify only sporadically. As a result, harmful subsidies continue to undermine sustainable development and the WTO's overall goal remains distant, or even recedes.

We think that imaginative use of existing transparency mechanisms can help. We focus not on the so-called ‘enforcement’ of precise legal obligations, because the real life of the trading system is elsewhere, and because it is hard to enforce a preamble. This paper therefore offers a proposal on how more information about fossil fuel subsidies can be brought into the WTO as the basis for analysis and further discussion. With more information, Members can learn from each other while publication in TPR reports may help citizens learn that certain kinds of policies in favour of both consumers and producers are in fact perverse subsidies.

Our analysis of the negotiating history of ASCM Article 8, and the broader difficulties in the Doha Round, lead us to conclude that no new provisions for energy subsidies will be added to the WTO any time soon. We must work with what exists, but what exists can make a significant contribution to the analysis of fossil fuel subsidies within the context of both an individual country's trade and economic policy and the effects of subsidies on the trading system. We have found no example of the SCM Committee using third-party notification, so the target for our proposal is the TPRB, but we remain consistent with the definition of subsidies in the ASCM, ‘as if’ Article 8 were still in force; that is, a Member who wished could use an NGO TPR notification as the basis for a legitimate question in the SCM Committee. We think that the Secretariat should make energy subsidies a separate subsection of every TPR and every monitoring report. Having more comprehensive data can only help discussions about the trade implications of energy subsidies and climate change. Such enhanced surveillance, far from providing raw material for more disputes, will help to clarify the principles and rules applicable to energy, thereby avoiding disputes. And that clarification is essential if the WTO is to be in a position to support rather than obstruct global action to reduce greenhouse gas emissions, one of the defining issues of our time.