Part III - Fossil Fuel Subsidies

4 Fossil Fuel Subsidies Key Issues

Subsidies for the production and consumption of fossil fuels exist in most, arguably all countries of the world, in spite of them undermining global efforts to curb climate change. Consumer subsidies are directed at the fossil fuel use of households or companies. They include free electricity or electricity at a reduced price, cooking fuels such as kerosene sold at below-market prices, petrol prices fixed at levels as low as USD 0.10 per litre and – depending on the definition – reductions in the value-added tax (VAT) and taxes on fossil fuels as well as prices that do not reflect the externalities associated with using the fuel. Producer subsidies are directed at the production of fossil fuels, and include inter alia tax rebates and loans, financial and technical support for exploring potential fossil fuel resources such as new oil or gas fields, direct financial transfers, and so forth.

Unlike most other policies in place to mitigate climate change, reforming such subsidies provides fiscal and macroeconomic benefits. Yet, fossil fuel subsidy reform received limited attention at the international level until the 2009 G20 commitment to phase out or reform inefficient fossil fuel subsidies (Van de Graaf and Blondeel, 2018). On the domestic level, policies constituting fossil subsidies have been reformed quite often (frequently only to be reintroduced or expanded at a later stage), but historically such reform has been driven by economic objectives rather than environmental ones (Skovgaard and van Asselt, 2018b). While the subsequent chapters will discuss this commitment and other efforts by the international economic institutions to address fossil fuel subsidies, this chapter will provide an introduction to the subject and the efforts to promote their reform by other institutions than the ones studied here. The chapter starts with a discussion of the different definitions of fossil fuel subsidies, definitions that have far-reaching political consequences, followed by an overview of the estimates of the size and scope of existing fossil fuel subsidies. Subsequently, I discuss the domestic politics of fossil fuel subsidies and their reform, followed by an overview of the efforts to address fossil fuel subsidies of other institutions than the ones studied in this book.

4.1 Definitions of Fossil Fuel Subsidies

There is no agreement on how to define energy subsidies (Gerasimchuk, 2014; Koplow, 2018; OECD Secretariat, 2010b). This disagreement has far-reaching consequences for the measurement of the global and national levels of fossil fuel energy subsidies and the countries that are considered as having fossil fuel subsidies. Few observers dispute that policies that lower the price paid by consumers below the market price, for example, fixing the price of petrol at USD 0.30 per litre as it has been the case in Iran (Kojima, 2016), constitute an energy subsidy. Yet, several other types of policies may be defined as fossil fuel subsidies depending on the definition. Few if any policies are defined as fossil fuel subsidies by the policymakers that adopt them, but they may subsequently be defined as fossil fuel subsidies by other actors.

An important distinction is the one between attempts to identify (and often also measure) fossil fuel subsidies that rely on an inventory approach and those that rely on a price-gap approach. These two approaches depend implicitly or explicitly on different definitions of fossil fuel subsidies, for example, the price-gap approach relies on definitions of fossil fuel subsidies that define such subsidies in terms of prices being below a given benchmark.

The inventory approach focuses on government policies and defines as fossil fuel subsidies those policies that confer benefits to particular fossil fuel producing or consuming activities. On the consumption side, the inventory approach identifies as subsidies policies including direct spending on the lowering of fossil fuel prices, reduced tax or VAT rates on fossil fuels, and so forth. On the production side, it identifies a broader range of policies as subsidies (although this varies somewhat between different kinds of inventory approaches), including the public provision of infrastructure for fossil fuels (e.g. pipelines, railroads); tax reductions; insurances, loans and guarantees provided with more favourable conditions than what the market offers; research and development; as well as government ownership of fossil fuel extraction enterprises (e.g. loss-making coalmines). The focus on policies means that the inventory approach often leads to debates regarding whether a given policy actually confers such benefits. The inventory approach requires extensive data gathering to identify the subsidies within a given country, and inventories sometimes do not include all subsidies within a country due to data limitations (Kojima and Koplow, 2015; Koplow, 2018). Importantly, inventory approaches rely on different definitions of fossil fuel subsidies, all of them characterised by defining fossil fuel subsidies in terms of policies conferring benefits on the consumption and/or production of fossil fuels. One prominent example of such a definition is the World Trade Organization (WTO)’s definition of subsidies (of all kinds not just those concerning fossil fuels) as a financial contribution by a government that confers a benefit to the recipient (World Trade Organization, 1994, Article 1). Yet, it is also possible to include non-financial contributions, for example, policies reducing risk, under the definition of subsidies (Koplow, 2018).

The price-gap approach focuses on the consumer price of fossil fuels rather than the policies influencing such prices. Specifically, it identifies whether the consumer prices are below a given benchmark price and estimates the combined value of the difference between the two prices. The benchmark price is generally based on the international market price of a given fossil fuel, often with the transport and distribution costs and/or VAT added, and in some cases also taxes corresponding to the externalities (e.g. air pollution, climate change, traffic accidents) of using the fuel (Clements et al., 2013; Coady et al., 2015; Gerasimchuk, 2014; Koplow, 2009; Steenblik and OECD, 2003). The level of the benchmark price is crucial for estimates of the size of total fossil fuel subsidies, as a high benchmark price will lead to high estimates of total fossil fuel subsidies at the global and national level. The price-gap approach only identifies the effects of fossil fuel subsidies that influence consumer prices, and hence producer subsidies are included in such studies only to the degree that they have an effect on consumer prices, which they rarely do, as most fossil fuels (e.g. gas or oil) are sold in global markets.

Definitional aspects are also important as regards determining which policies should be reformed or phased out. Both the G20 and the Asia-Pacific Economic Cooperation (APEC) made the commitment ‘to rationalize and phase out over the medium term inefficient fossil fuel subsidies that encourage wasteful consumption’ (APEC, 2009; G20 Heads of State and Government, 2009b). This wording raises questions regarding the exact interpretation of the terms ‘rationalize’, ‘medium term’ and most importantly for the issue of defining fossil fuel subsidies, ‘inefficient’¹ and ‘encourage wasteful consumption’. As is discussed in Chapters 5–8, much of the discussion has focused on whether a country’s fossil fuel subsidies are indeed inefficient and encourage wasteful consumption.

4.2 The Size and Scope of Fossil Fuel Subsidies

The size and scope of global fossil fuel subsidies depend on which of the aforementioned definitions is being used. Subsidies for the consumption and production of coal, natural gas, oil and products derived from these fuels (e.g. diesel, regular petrol, kerosene, liquid petroleum gas) are considered fossil fuel subsidies, as are subsidies for electricity and heat production based on fossil fuels (Kojima and Koplow, 2015). Subsidies for biofuels are generally not considered fossil fuel subsidies. In terms of geographical scope, most or virtually all countries (depending on the definition used) have some kind of fossil fuel subsidies in place. Policies subsidising the consumption of fossil fuels are more substantial in developing countries, whereas policies subsidising their production are common in both developed and developing countries. Price-gap estimates differ in their estimates of fossil fuel subsidies in developed and developing countries. Those which use benchmark prices without externalities, such as the International Energy Agency (IEA, 2018), find that fossil fuel subsidies are much smaller in developing countries than in developed ones, whereas those that include externalities, most notably the IMF (2019), find that developed countries account for more than a quarter of global subsidies. All estimates find that fossil fuel exporting countries have larger subsidies (compared to the size of their populations and GDP) than fossil fuel importing ones.

Studies of fossil fuel subsidies focus mainly on national policies, excluding development finance from multilateral development banks, multilateral and bilateral development institutions for fossil fuel production and consumption, which have been estimated at tens of billions of dollars (Kim and Urpelainen, 2013; Oil Change International et al., 2017). Political debates among policymakers, including the institutions studied in this book, have focused on national level subsidies, and consequently this book will mainly focus on how they have addressed this issue.

As mentioned previously, the different definitions translate into diverging estimates of the global economic costs or size of fossil fuel subsidies (not including support through development finance). The IEA uses a price-gap approach with a benchmark price including distribution, transportation and VAT but not externalities, and its estimate is not global but covers forty of the largest developing and emerging countries. This widely used estimate puts global consumption subsidies in 2017 at just over USD 300 billion, and generally fluctuating between USD 250 and 600 billion (IEA, 2015, 2016, 2017, 2018), depending mainly on the oil price. All told, the IEA estimate is at the low end of the range. The OECD has provided an estimate combining a price-gap and inventory approach (discussed in detail in Chapter 6) and covering the thirty-five OECD countries plus eight partner countries (Argentina, Brazil, China, Colombia, India, Indonesia, Russia and South Africa). The OECD estimates fossil fuel support in these countries studied in 2016 at USD 151 billion, and fluctuating between USD 150 and 250 billion in the years 2010–16 (OECD, 2018b). The OECD and the IEA have more recently started combining their estimates, and arrive at an estimate of USD 340 billion, fluctuating between USD 300 and 600 billion in the period 2010–17 for the countries covered by their combined estimate (OECD and IEA, 2019). The IMF’s estimate (discussed in detail in Chapter 7) covers 153 countries and includes both producer² and consumer subsidies, the latter calculated on the basis of a benchmark price including various externalities to arrive at its estimate of USD 5.2 trillion estimate for 2017 (Coady et al., 2019). The IMF’s estimate is about ten times higher than the IEA’s mainly due to the inclusion of non-priced externalities, but also due to its global scope. Importantly, these estimates do not tell us about who bears the costs of the subsidies. The estimates differ implicitly in this respect, as the IEA and OECD estimates concern the fiscal costs to public budgets of providing the subsidies, while the IMF estimates mainly concern the costs to society of using fossil fuels.

In terms of environmental consequences, fossil fuel subsidies are distinguished from other subsidies in targeting fossil fuels, which implies that they by definition have a negative impact on climate change. Different estimates exist of the direct effects of fossil fuel subsidies in terms of encouraging the use of fossil fuels and hence causing CO₂ emissions (Skovgaard and Van Asselt, 2019). These estimates differ in terms of their scope, in terms of the countries and subsidies covered (production subsidies and a range of consumption subsidies are often not included), as well as the methodology used and the time horizon. The estimates find that the emissions reductions alone from phasing out fossil fuel subsidies range from 1 to 23 per cent of the emissions in the countries covered (Burniaux and Château, 2011; Coady et al., 2015, 2019; Jewell et al., 2018). According to the conservative estimate of Jewell et al. (2018), reforming fossil fuel subsidies could deliver a quarter of the emissions reductions pledged under the Paris Agreement. These figures would be higher if the savings from reforms were redirected towards renewable energy (Jakob and Hilaire, 2015; Schmidt et al., 2017). Perhaps most importantly, these estimates cannot capture the political economic effects of breaking the lock-in of fossil fuel subsides in terms of fossil fuel infrastructure as well as the political power of fossil fuel corporate actors locking societies into fossil fuel-based modes of production and consumption (Erickson et al., 2020; Newell and Johnstone, 2018). Beyond climate change, fossil fuel subsidies lead to local air pollution, inter alia through the burning of coal and diesel, with effects on health that accounts for close to half of the global externalities of fossil fuel use according to the IMF (Coady et al., 2015, 2019; Parry et al., 2014).

In terms of redistributive consequences, proponents of fossil fuel consumption subsidies often justify them by framing them as a tool for poverty reduction, especially in developing countries (Rentschler and Bazilian, 2017a; Rentschler and Bazilian, 2017b). Yet, studies of the allocation of fossil fuel subsidies find that most of them are captured by the higher income segments of society. For instance, Arze del Granado, Coady and Gillingham (2012, p. 2241) in their study of twenty developing countries found that ‘the richest 20% of households capture on average six times more in fuel subsidies than the poorest 20%’. Fossil fuel subsidies are regressive because they tend to be universal while subsidising goods that people with a higher income have more opportunities to enjoy, for example, fuel for cars.

4.3 The Domestic Politics of Fossil Fuel Subsidies and Their Reform

Irrespective of the definition of fossil fuel subsidies that is used, such subsidies have proven difficult to reform (Skovgaard and van Asselt, 2018b, 2019). While the world has arguably witnessed an increase in the number of fossil fuel subsidy reforms since the Pittsburgh commitment in 2009 (Rentschler and Bazilian, 2017b; Van de Graaf and Blondeel, 2018), fossil fuel subsidies still persist globally, and the decline in the IEA’s estimates of global subsidies seem more driven by lower oil prices than by reform. Furthermore, it is far from certain to what degree fossil fuel subsidy reforms have been driven by the G20 commitment and the increasing international attention to fossil fuel subsidies. The reforms seem driven mainly by economic concerns, particularly fiscal deficits and the desire to provide more targeted social assistance to the poor (Rentschler and Bazilian, 2017b; Skovgaard and van Asselt, 2018a). The subsidies that have been reformed consist mainly of consumption subsidies in middle-income developing countries such as Egypt, India, Indonesia, Iran and the Philippines (Van de Graaf and Blondeel, 2018), as well as coal production subsidies in developed countries, especially EU member states such as Germany and Spain (Gençsü et al., 2017). The former group have increased and liberalised fuel prices and targeted subsidies at the poor, whereas the latter group have phased out coal subsidies gradually while providing support to communities dependent on coal mining (e.g. retraining of workers; see Zinecker et al., 2018).

An important aspect of the persistence of fossil fuel subsidies is that successful reform has often been followed by the reversal to old levels of subsidies After all, domestic actors have tried to reform fossil fuel subsidies as long as these subsidies have been in existence. Several attempts at fossil fuel subsidy reform have also failed, some before the adoption of the reform and some after implementation, inter alia due to public protests (e.g. in Ecuador and Sudan).

The literature on the politics of fossil fuel subsidies and their reform has identified several factors driving fossil fuel subsidies and the possibilities of reforming them. Here, I draw on the three kinds of factors identified in Skovgaard and van Asselt (2018c). First, the interests, strategies and organisation of actors – including both individuals and collective actors – that promote reform or try to keep subsidies in place. Their strategies include putting fossil fuel subsidies on the national political agenda or trying to block such efforts; framing fossil fuel subsidies in particular ways, building coalitions to promote or counter reform; and communicating the benefits of subsidies or their reform to policymakers and the public. Fossil fuel subsidies have been framed, on the one hand, in terms of their economic or environmental cost and on the other, as important tools for reducing poverty or improving national development and competitiveness. Beyond the strategies of actors, their degree of organisation also matters, particularly as regards actors benefitting from subsidies (Victor, 2009). Actors opposed to fossil fuel subsidies tend to be less organised in interest groups than those supporting subsidies, yet both form alliances cutting across different political parties, ministries, and non-governmental organisations (Skovgaard and van Asselt, 2019). One reason for the higher degree of organisation of the proponents of subsidies is that the benefits of fossil fuel subsidies are tangible and concentrated in specific groups (e.g. fossil fuel producers, beneficiaries of consumer subsidies), whereas the benefits of fossil fuel subsidy reform are less tangible and more diffuse across time and space (e.g. improved public budgets and environment; see Inchauste and Victor, 2017).

Second, ideational factors, including the aforementioned definitional issues as well as knowledge about fossil fuel subsidies and their environmental and socio-economic effects, also influence the politics of fossil fuel subsidies. Established discourses regarding issues such as development, competitiveness and environmental protection constitute important ideational contexts that may shape whether a particular framing is successful or not, e.g. may the framing of fossil fuel subsidies as environmentally harmful fail in countries in which environmental protection is not defined as important (Skovgaard and van Asselt, 2018c). Importantly, the existence of fossil fuel subsidies is a sensitive issue in several (especially developed) countries, and governments are often reluctant to acknowledge that a given policy constitutes a fossil fuel subsidy.

The third group of factors is more structural and includes macroeconomic developments and the socio-political characteristics of a country. In terms of macroeconomic factors, both fossil fuel reserves (Overland, 2010) and high fossil fuel prices (Benes et al., 2015; Rentschler and Bazilian, 2017b) are associated with higher subsidies, whereas rapid changes to fossil fuel prices have offered windows of opportunity for reform (Benes et al., 2015). Furthermore, states with weak institutional capacity and authoritarian rule are more likely to subsidise fossil fuels, inter alia because they lack the capacity to implement more complex welfare policy instruments such as cash transfers (Cheon et al., 2013; Lockwood, 2015; Victor, 2009). Finally, there is an element of path dependency to fossil fuel subsidies, which means that once in place they are difficult to remove. The path dependency may be due to fossil fuel subsidies empowering actors benefitting from them – particularly fossil fuel extraction companies – and thus contributing to carbon lock-in (Newell and Johnstone, 2018) or becoming part of the social contract between the state and its citizens (Moerenhout, 2018).

4.4 Other International Efforts to Promote Fossil Fuel Subsidy Reform

Beyond the institutions studied in this book, a range of other institutions have been important to the efforts to reform fossil fuel subsidies. In general, their involvement with fossil fuel subsidies has increased since 2009. First, the IEA stands out on the basis of its extensive work on defining and measuring fossil fuel subsidies. These efforts date back to before 2009, most notably the 1999 issue of the IEA’s World Energy Outlook, which included fossil fuel subsidies among its key foci (IEA, 2000). The most important part of the IEA’s work on fossil fuel subsidies has been its estimates of the total size of fossil fuel subsidies in major non-OECD economies, which was first published in 2006 and has since provided a crucial knowledge base for addressing fossil fuel subsidies. This estimate is probably the most widely used estimate of the size of global fossil fuel subsidies. As mentioned earlier, it covers only forty of the largest emerging and developing countries, and it is thus somewhat misleading to refer to it as an estimate of global subsidies, although it covers a very sizeable share of global subsidies. Importantly, the IEA employs a price-gap approach to measuring fossil fuel subsidies based on a benchmark price corresponding to ‘the full cost of supply or, where appropriate, the international market price, adjusted for the costs of transportation and distribution, and value-added tax’ (IEA, 2016, p. 97, fn. 8). The IEA was one of the four institutions (together with the World Bank, the OECD and Organization for Petroleum Exporting Countries [OPEC]) that was requested by the G20 to measure the magnitude and the consequences of such subsidies (G20 Heads of State and Government, 2009b).

Of the other institutions requested by the G20 to study fossil fuel subsidies, the World Bank also has a long-running track record. World Bank studies on energy subsidies in general date back to the 1980s (World Bank, 1983), and studies on fossil fuel subsidies specifically to the 1990s (Larsen and Shah, 1992). In terms of policy, the World Bank’s programmes induced developing countries to reform their energy subsidies (which almost always went to fossil fuels) as part of wider reform packages (Van de Graaf and Blondeel, 2018). However, the World Bank’s attention to fossil fuel subsidies waxed and waned until around 2009, when its focus on fossil fuel subsidies reached a consistent level. The World Bank published numerous publications on fossil fuel subsidies in developing countries, particularly on the best way to phase out such subsidies (Kojima, 2016; World Bank, 2013b; World Bank with contributions from International Monetary Fund (IMF), 2014).

In terms of concrete efforts to induce countries to reform their fossil fuel subsidies, the Bank has focused explicitly on fossil fuel subsidies (rather than just subsidies in general). It has done so in terms of country specific recommendations (see e.g. Diop, 2014; Peszko et al., 2019) and assistance to such reform, for example, in the shape of expertise; workshops for learning from other countries; and financial support for policy dialogue, communication and the targeting of subsidies, and so forth. (McCulloch, 2017; Skovgaard, 2018). Notable in this respect is the Bank’s Energy Sector Management Assistance Program (ESMAP), which has developed both an ‘Energy Subsidy Reform Technical Assistance Facility’ providing knowledge in the shape of an analysis of the environmental, fiscal, economic, political and social impacts of fossil fuel subsidy reform, as well as support for policy dialogue and the design of reform (World Bank, 2015). ESMAP has also developed the ‘Energy Subsidy Reform Assessment Framework’ (ESRAF), a guide to analysing energy subsidies, the impact of subsidy reform and the political context for subsidy reform (Flochel and Gooptu, 2016). However, the World Bank has also previously been criticised for providing billions of dollars in funding for fossil fuel production, inter alia coal- and gas-fired power plants, pipelines as well as oil and gas exploration, in spite of commitments to phase out such lending (The Big Shift Global, 2019).

The third of the four institutions requested to study fossil fuel subsidies by the G20, OPEC, had not previously addressed fossil fuel subsidies. OPEC was included among the four institutions due to the insistence of Saudi Arabia (interview with senior OECD officials, 29 April 2015) and has been less active than the other three institutions (Lang, 2011), and has been involved in fewer reports to the G20 than the other institutions. The lower level of involvement is evident in in that OPEC has contributed to fewer of the reports to the G20 than the other institutions (IEA, OECD, et al., 2010; OECD and IEA, 2019). The OPEC member states are among those countries with the highest fossil fuel subsidies total and per capita, and benefit from the fossil fuel subsidies in other countries in terms of increased exports (Jewell et al., 2018).

The institutions discussed in the preceding text cover energy and development. Other international energy and development institutions, such as the International Renewable Energy Agency (IRENA) and the UN Development Programme (UNDP), have been much less vocal concerning fossil fuel subsidies. As regards institutions addressing other issues than economic, development and energy issues, the picture is also rather mixed. Concerning environmental institutions, the United Nations Framework Convention on Climate Change (UNFCCC) is mainly notable due to its lack of attention to fossil fuel subsidies (van Asselt et al., 2018). The Kyoto Protocol contained a brief reference to the reduction or phasing out of subsidies in greenhouse gas emitting sectors (UNFCCC, 1997), but both the United Nations Framework Convention on Climate Change and the Paris Agreement remain silent on the issue. Opposition from oil-exporting countries as well as the general reluctance within the UNFCCC regarding addressing energy issues have meant that the attempts of some countries to place fossil fuel subsidies within the UNFCCC have been unsuccessful (van Asselt and Kulovesi, 2017; van Asselt et al., 2018). Yet, thirteen countries have chosen to mention fossil fuel subsidy reform in the Intended Nationally Determined Contributions (INDCs) they have submitted in the context of the Paris Agreement, several of them committing to such reform (Terton et al., 2015). More implicitly, Article 2.1.c of the Paris Agreement specifies that the objectives of the Agreement shall be met inter alia by making ‘finance flows consistent with a pathway towards low greenhouse gas emissions and climate-resilient development’ (UNFCCC, 2015). Although fossil fuel subsidies are not specifically mentioned, they are generally not consistent with a pathway to low greenhouse gas emissions. Yet, the provision does not place any obligations on states to reform fossil fuel subsidies.

Perhaps due to their broader scope covering sustainability and development, the Sustainable Development Goals (SDGs) include more specific commitments to the reform of fossil fuel subsidies in their Target 12.c, which commits all countries to undertaking efforts to rationalise inefficient fossil fuel subsidies that encourage wasteful consumption. The wording of Target 12.c is rather similar to the G20 commitment in its emphasis on inefficient fossil fuel subsidies encouraging wasteful consumption, but leaves even more freedom to states, especially as they commit only to rationalising and not to phasing out such subsidies. Furthermore, the Goal does not include a reference to when such subsidies should be rationalised the way the G20 commitment refers to as the medium term. The effects of the SDG commitment as well as the Nationally Determined Contributions commitments under the Paris Agreement remain to be studied but constitute a move towards more attention to fossil fuel subsidies among environmental institutions.

The UN Environment Programme (UNEP) has been active in promoting fossil fuel subsidy reform through the production of knowledge in terms of reports on fossil fuel subsidies and most importantly an internationally agreed approach to measuring fossil fuel subsidies in the context of the SDGs (developed together with the OECD and the International Institute for Sustainable Development [IISD]; UNEP, OECD and IISD, 2019). It has also promoted the norm of fossil fuel subsidy reform and linked it to the SDGs (UNEP, 2019).

Also trade institutions are more notable in terms of what they have not done than what they have done. In spite of persistent calls for the WTO to adopt measures disciplining fossil fuel subsidies the way that they discipline several other subsidies, it has not done so, and other trade institutions have generally not addressed the issue (Bièvre et al., 2017; Steenblik et al., 2018). In 2017, twelve WTO member states called for the WTO to adopt measures disciplining fossil fuel subsidies, thus utilising one of the most effective incentive-based instruments in international governance, namely retaliatory trade measures sanctioned by the WTO dispute settlement mechanism. However, this proposal has not found sufficient support from the rest of the WTO member states. One reason for the inaction is that most fossil fuel subsidies are not clearly trade distorting in the way that, for instance, several agricultural and renewable energy subsidies are (Steenblik et al., 2018). Countries could in principle (within the WTO or another trade institution) agree to sanction subsidies not because they are trade distorting but because of their environmental effects, as indeed was the case with draft versions of the currently abandoned Trans-Pacific Partnership Agreement (Steenblik et al., 2018). Yet, at the time of writing support for exploring this option within the WTO has come from only a few countries. The negotiations on an Agreement on Climate Change, Trade and Sustainability launched in 2019 by New Zealand, Costa Rica, Fiji and Iceland to use trade rules to tackle climate change and other environmental issues, specifically address fossil fuel subsidies (Costa Rica et al., 2019). Yet, it is too early to assess the eventual role of such an agreement in promoting fossil fuel subsidies.

Forums of smaller groups of states have been more successful in addressing fossil fuel subsidies, particularly promoting the norm of fossil fuel subsidy reform. Besides the G20, APEC adopted a commitment similar to the G20’s just a few weeks later, and has also adopted voluntary reporting and peer-review processes, in which member states can report their fossil fuel subsidies and some of them even undergo peer reviews (Verkuijl and van Asselt, 2020). In 2016, both the North American Leaders’ Forum – the heads of state of Canada, Mexico and the United States – and the G7 adopted commitments similar to the G20’s but with 2025 as the phase-out date (unlike the G20 commitment which does not include a phase-out date). Furthermore, the Friends of Fossil Fuel Subsidy Reform was established in 2010 on the initiative of New Zealand (Rive, 2018). The Friends is an informal group of – at the time of writing – nine non-G20 countries (Costa Rica, Denmark, Ethiopia, Finland, New Zealand, Norway, Sweden, Switzerland and Uruguay) working to promote the norm of fossil fuel subsidy reform. Its activities include the 2015 Communiqué on fossil fuel subsidies inviting states and non-state actors to support accelerated action to eliminate inefficient fossil fuel subsidies (Friends of Fossil Fuel Subsidy Reform, 2015), voluntary peer review and agenda-setting, including the aforementioned call for WTO to address fossil fuel subsidies.

Finally, among the civil society actors promoting fossil fuel subsidy reform, the International Institute for Sustainable Development (IISD) and its Global Subsidies Initiative GSI stand out (Lemphers et al., 2018). The IISD established the GSI in 2005 to provide knowledge about (initially mainly biofuel, since 2009 mainly fossil fuel) subsidies and promote their reform. It has been involved in international analyses of fossil fuel subsidies and concrete reforms of subsidies.

4.5 Summary

This chapter demonstrates the intricacies of the politics of fossil fuel subsidies. In spite of the widespread international commitments to reforming fossil fuel subsidies and their economic and environmental benefits, these subsidies persist globally. Domestic factors, inter alia the efforts of actors benefitting from the subsidies, lack of awareness of the subsidies, fossil fuel reserves and (weak) governance capacity, have been the main obstacles to fossil fuel subsidy reform. Surprisingly, international environmental institutions have been quiet as regards addressing such subsidies, which puts the activities of the economic institutions into perspective and underscores why it is relevant to study these activities.

5 The G20 and Fossil Fuel Subsidies The Catalyst

The September 2009 G20 commitment to reform fossil fuel subsidies took most spectators by surprise. Few of the limited number of people working on the topic were aware that such a commitment was being discussed (Van de Graaf and Blondeel, 2018), and the concept was largely unknown in broader circles. The surprise element only adds to the impression that there is a ‘before’ the September 2009 G20 commitment to reform fossil fuel subsidies and an ‘after’ (Skovgaard and van Asselt, 2018c). The commitment set in motion a range of efforts from other international institutions, which will be discussed in this chapter as well as the following ones. Its effects on the domestic level are less immediately evident, but nonetheless relevant. The G20 output from the Pittsburgh commitment and the subsequent, more technical output onwards is outlined in the next section. This is followed by a discussion of how US entrepreneurship was important in getting the G20 to address the issue, and how the output has been shaped by the membership circle and worldview of the G20 as well as interactions with the International Energy Agency (IEA), Organisation for Economic Co-operation and Development (OECD), Organization for Petroleum Exporting Countries (OPEC) and the World Bank. The subsequent section outlines the consequences of the G20 output, which was most pronounced in terms of promoting the norm of fossil fuel subsidy reform and of raising awareness of fossil fuel subsidies, in both cases both at the international and (to a lesser degree) the domestic levels.

5.1 Output: The Pittsburgh Commitment And The Subsequent Reviews

The G20 output has predominantly been formal and regulatory, most importantly in the shape of the 2009 commitment. The commitment reads as follows:

To phase out and rationalize over the medium term inefficient fossil fuel subsidies while providing targeted support for the poorest. Inefficient fossil fuel subsidies encourage wasteful consumption, reduce our energy security, impede investment in clean energy sources and undermine efforts to deal with the threat of climate change. We call on our Energy and Finance Ministers to report to us their implementation strategies and timeline for acting to meet this critical commitment at our next meeting.

(G20 Heads of State and Government, 2009b)

This commitment is most important in normative terms, as it defined and elevated the norm of fossil fuel subsidy reform to a new level (Van de Graaf and Blondeel, 2018). While the commitment referred to the OECD and IEA estimates that phasing out fossil fuel subsidies could reduce emissions by 10 per cent by 2050 (OECD, 2009), it did not provide a definition of fossil fuel subsidies, or specify what the terms ‘rationalize’, ‘medium term’ and ‘inefficient’ meant. In this way, the norm was left vague, especially as regards the policies that would fall under the category of inefficient fossil fuel subsidies and thence be targeted by the norm. Importantly, fossil fuel subsidies were primarily framed in terms of their impact on climate change, while the importance of maintaining support for poverty reduction was also stressed. Fossil fuel subsidies were also framed in terms of macroeconomic consequences (e.g. ‘inefficient fossil fuel subsidies encourage wasteful consumption, distort markets, impede investment in clean energy sources’) while the fiscal impact was not mentioned.

The 2009 statement also contained two important clauses regarding future efforts to promote fossil fuel subsidy reform. First, the IEA, the OECD, OPEC and the World Bank were tasked with measuring the magnitude and the consequences of such subsidies (discussed in detail in Section 5.3). Second, member states committed themselves to submitting strategies and timetables for phasing out their fossil fuel subsidies while taking into account the needs of the poorest citizens (G20 Heads of State and Government, 2009b), leading to various kinds of output that are most important in normative terms. The commitment to submitting strategies and timetables led to tasking member state experts under the authority of their finance and energy ministers with coordinating and overseeing the implementation of the commitment (Kim and Chung, 2012). The experts have been meeting in the context of a broader working group on energy, the G20 Energy Transitions Working Group (previously the Energy Sustainability Working Group), which focuses on the transition to sustainable energy systems. The expert output on fossil fuel subsidies has generally been reported to the finance ministers, and the largest group of experts also came from finance ministries (interview with senior OECD official, 3 February 2020).

Discussions of how to define fossil fuel subsidies (including whether to include production subsidies), as well as of how to define ‘inefficient’ and ‘wasteful consumption’ did not result in an agreement on common definitions. Rather it was agreed to leave these issues to the reporting countries (Lang, 2011). Starting in 2010, the G20 member states reported on an annual basis whether they had any inefficient fossil fuel subsidies and the progress on reforming or phasing out these subsidies, constituting regulatory output (Aldy, 2017). Seven countries (Australia, Brazil, France, Japan, Saudi Arabia, South Africa and the United Kingdom) have reported that they have no fossil fuel subsidies, whereas other countries have submitted plans of varying ambition for phasing out their subsidies (Kirton et al., 2013). The progress reports have focused mainly on measures taken to reform the subsidies identified in the 2010 country reports (Asmelash, 2017). The G20’s bottom-up approach leaving it to the member states to define which fossil fuel subsidies they have and how to phase them out has been criticised for only inducing countries to act to a limited degree (Van de Graaf and Westphal, 2011). Nonetheless, the reporting requirement constitutes important ideational output in terms of forcing G20 member states to acknowledge the salience of the norm of fossil fuel subsidy reform and argue whether it applies to them, as well as in terms of promoting the framing of policies as fossil fuel subsidies. The working group has also served as the forum for officials for discussions and the exchange of knowledge about fossil fuel subsidies on the basis of their own experience and the reports provided by the IEA, OECD, OPEC and World Bank. There are very few forums in which finance (and economics) ministry officials can discuss climate change, and the working group served as a useful forum for such discussions focusing on fossil fuel subsidies (Interview with former senior US Treasury official, 6 May 2014). In the first few years, fossil fuel subsidies were still a new issue where there existed only limited knowledge, and the working group expanded the knowledge about the issue among the participants, and – via the reports from the four International Organisations (IOs) – also among a wider public.

A subsequent development was the 2012 decision by G20 state leaders to request their finance ministers to explore the options for voluntary peer reviews of member states’ fossil fuel subsidies and their efforts to reform or phase them out (G20 Heads of State and Government, 2012). The peer review replaced self-reporting as the most important G20 (regulatory) output on fossil fuel subsidies (Rive, 2019). Currently, some member states (mainly those having undergone peer reviews) provide updates on their reform efforts at the meetings of experts, but no agreement has been reached regarding a proposal to reintroduce the mandatory self-reporting process with an IO review of the reports (interview with senior OECD official, 3 February 2020).

In 2016, the two largest economies and emitters, the United States and China, volunteered to be the first countries to undergo a pairwise peer review. In this review, they each first provided a self-report on their fossil fuel subsidies and the efforts to reform them; this was subsequently reviewed by the other country as well as the OECD, the IMF (in the case of China), Germany, Indonesia (in the case of China) and Mexico (in the case of the United States) (G20, 2016a, 2016b). In 2017, Germany and Mexico, and in 2019 Indonesia and Italy, underwent similar peer reviews, whereas at the time of writing Argentina and Canada have planned such reviews. The later reviews have been carried out by China, Germany, Italy, Indonesia, Mexico, New Zealand, the OECD (acting as chair for all the reviews), and in the case of the 2019 reviews also the IEA, the International Institute for Sustainable Development (IISD) and the World Bank. The peer reviews follow a logic in which a developed and an emerging economy undergo a review together to avoid criticism of double standards. So far, the countries undergoing a peer review are all countries that have acknowledged having inefficient fossil fuel subsidies in their reports to the G20. The peer reviews have been criticised for not including all fossil fuel subsidies in the reviewed countries (see e.g. the criticism of Germany’s peer review, Hansen, 2017). They are best understood as providing opportunities for learning (Verkuijl and van Asselt, 2020), getting states to accept the framing of particular policies as fossil fuel subsidies and acknowledging that the norm of fossil fuel subsidy reform is salient in regard to these policies. Importantly, although fossil fuel subsidies were addressed by officials from finance (and economics) ministries, and to a lesser degree also energy ministries, they were framed mainly as a climate change issue which also involved economic inefficiencies such as market distortions and the inefficient use of fiscal resources.

5.2 Causes

Regarding the factors influencing the adoption of the 2009 commitment (and hence the first aspect of economisation), entrepreneurship and relations with member states stand out. Before the Pittsburgh Summit, G20 member states including the United States had attempted to put fossil fuel subsidies on the G20 agenda for five years without success (interview with former senior US White House official, 17 February 2015). The difficulty of addressing fossil fuel subsidy reform in any international forum, particularly forums which include Saudi Arabia, meant that the commitment took spectators by surprise (Van de Graaf and Blondeel, 2018). Several G20 members, particularly Saudi Arabia, had blocked the previous attempts, underscoring the importance of which states are members of the institution and how the member states arrive at decisions (in this case consensus allowing one state to block proposals). The entrepreneurship of the US government (the G20 president) played a key role in placing the commitment on the agenda and also in terms of the US government drafting the commitment text (Van de Graaf and Blondeel, 2018). This draft text went fairly unchanged through the working groups of officials from the member states. One important change was the change in the timeframe for the phase-out/rationalisation from five years to ‘medium term’, a change which was at the insistence of the Chinese (interview with former senior US White House official, 17 February 2015). The notion of a deadline for reforming or phasing out inefficient fossil fuel subsidies has proven controversial in all forums debating these subsidies, and has only been possible to adopt in the G7 and the North American Leaders’ Forum, two smaller forums that do not include the largest emerging economies or oil producers, notably China, India and Saudi Arabia. This difference between the G20 and the two smaller forums regarding a deadline underscores the importance of which states are the members of the institution. Another important change to the draft commitment, was the BRICs (Brazil, Russia, India and China) successful insistence on adding ‘rationalize’ to the commitment to ‘phase out and rationalize over the medium term inefficient fossil fuel subsidies’ (Kirton and Kokotsis, 2015). Thus, the norm became less specific, since a more specific definition would have made it unpalatable to several G20 member states (Van de Graaf and Blondeel, 2018). The broad membership circle of the G20 (covering developed and emerging economies as well as fossil fuel exporters and importers) meant that the wording of the commitment was somewhat vague, but also increased its relevance to a broader set of countries and arguably also its legitimacy beyond the G20.

Within the US government, the initiative came from the White House (more specifically the Council of Economic Advisors). Owing to previous failed attempts to address fossil fuel subsidies in the G20, several officials doubted that the attempt would be successful, but still deemed it worthwhile (interview with former senior US Treasury official, 8 April 2014). The US government chose to act as a policy entrepreneur due to the perceived stalemate in the United Nations Framework Convention on Climate Change (UNFCCC) negotiations which led the government to look for issues which ‘were good economic politics’ as well as climate politics (interview with former senior US White House official and current IMF senior official, 17 February 2015). Previously, when the G20 member states had sought to address fossil fuel subsidies, the G20 meetings had been meetings of finance ministers (and central bank governors), but the G20 state leaders took over the issue in 2009 when they started to meet due to the economic and financial crisis. The transfer of fossil fuel subsidies from finance ministers to state leaders meant the issue was addressed by a set of actors with more power to adopt far-reaching decisions. Thus, it was a combination of external factors (the UNFCCC stalemate in the run-up to the fifteenth Conference of the Parties to the United Nations Framework Convention on Climate Change [COP15] and the crisis) and policy entrepreneurship which drove the adoption of the commitment. Arguably, had climate change not been as high on the international agenda in 2009, the Council of Economic Advisors would not have addressed fossil fuel subsidies in the first place. Furthermore, the resources and institutional set-up facing policy entrepreneurs mattered. Had the initiative not come from the United States and the Council of Economic Advisors (recognised as one of the most powerful entities within the US government) and had it not been addressed by state leaders, other member states would have had better prospects of blocking the initiative.

In terms of how fossil fuel subsidies were addressed (the second aspect of economisation), it was the entrepreneurship of the Obama administration that framed fossil fuel subsidies as a climate issue (as well as an economic one), something which was controversial among some countries including India (see Section 5.3. The perceived stalemate during the UNFCCC negotiations as well as in the recently published OECD–IEA report (2009) on the climate consequences of fossil fuel subsidy reform influenced the framing of fossil fuel subsidies as a climate issue. While the UNFCCC stalemate, as previously mentioned, led the United States (and other G20 states) to promote fossil fuel subsidy reform as a climate instrument, the OECD–IEA report provided important knowledge regarding the climate impact of fossil fuel subsidy reform, specifically that ‘eliminating fossil fuel subsidies by 2020 would reduce global greenhouse gas emissions in 2050 by ten percent’ (G20 Heads of State and Government, 2009b, item 29). In other words, institutional interaction with the UNFCCC, the OECD and the IEA influenced the G20.

The (macro)economic worldview inherent to the G20 is evident in the framing in terms of ‘inefficient fossil fuel subsidies encourage wasteful consumption, distort markets, impede investment in clean energy sources’ [author’s emphasis]. Although the main purpose of reforming fossil fuel subsidies according to the G20 is to fight climate change, the causal chain through which this impact takes place is economic, i.e. through impeding investment and encouraging wasteful consumption. Furthermore, distorting markets is framed as constituting a problematic consequence in itself. In other words, the worldview of the G20 shaped the framing of fossil fuel subsidies (the second dimension or aspect of the economisation of fossil fuel subsidies), yet was less influential regarding the G20’s decision to address fossil fuel subsidies (the first aspect of economisation), which was rather driven by climate concerns. This worldview was rooted in the G20’s origins as a forum for dealing with economic issues and the economic officials drafting the commitment.

The 2009 commitment set the tone for much of the subsequent G20 output on fossil fuel subsidies. The G20 state leaders reaffirmed the commitment at every summit until the 2017 Summit in Hamburg, when opposition from the United States meant that joint references to the commitment were removed (Asmelash, 2017, G20 Heads of State and Government, 2017). Only the G20 ‘Hamburg Climate and Energy Action Plan for Growth’, adopted by the remaining 19 G20 members referred to fossil fuel subsidy reform (G20, 2017a). The US decision to withdraw from the Paris Agreement caused major contention at the summit. Consequently, references to climate related issues including fossil fuel subsidies at subsequent summits were adopted by the other G20 members without the United States, although the 2019 Osaka Summit reintroduced the fossil fuel subsidy commitment in a joint G20 declaration (G20 Heads of State and Government, 2018, 2019). While factors such as the membership circle and worldview inherent to the G20 remained unchanged, other factors changed after 2009. The US presidency already under Obama did not engage in the same level of entrepreneurship as in 2009, although the United States and China were important in volunteering to be subject to the first pair of peer reviews. Post-2009 presidencies were a great deal less entrepreneurial than the US one, although some presidencies promoted the issue to a larger degree than others, e.g. the Mexican presidency that managed to convince the members to agree on the conditions for the voluntary peer reviews. Once fossil fuel subsidies were placed on the G20 agenda and a process set in motion, it remained there until the Trump administration took over. In this way, the Trump administration acted as an ‘antipreneur’ resisting and rolling back normative change (Bloomfield, 2016). Some of the countries that lowered the precision of the commitment (e.g. China) ended up being rather active in the process, whereas others (e.g. Saudi Arabia) argued that the commitment did not apply to them as they did not have any inefficient subsidies (Kirton et al., 2013). Interaction within other institutions mattered most in the cases of the four institutions tasked with providing an analysis of fossil fuel subsidies. They have continuously provided material to the G20 that has shaped the knowledge of participants in G20 meetings as well as the broader public. This knowledge concerned the nature, scope and consequences of fossil fuel subsidies (economic, environmental and distributive) as well as how to reform them (IEA and OECD, 2018; IEA, OECD, et al., 2010; IEA, OPEC, et al., 2010; IEA et al., 2011; OECD and IEA, 2019; OECD Secretariat, 2010a; World Bank with contributions from International Monetary Fund (IMF), 2014).

5.3 Consequences

5.3.1 International Consequences

Starting with the international level, the G20 set in motion a range of activities through interaction with other institutions. Most importantly, among the four institutions requested to provide an analysis, the request caused an increased attention to fossil fuel subsidies beyond the analysis, thus influencing their agendas. The OECD Secretariat was already working on fossil fuel subsidies before the Pittsburgh Summit, but the request lifted OECD involvement to a new level (interview with OECD officials, 29 April 2015). It was only following the G20 commitment that the member states gave the OECD Secretariat the mandate to scrutinise their national fossil fuel subsidies (interview with OECD officials, 29 April 2015), an activity that goes beyond the G20 request. At a later stage, the decision by the G20 members that have so far committed to peer reviews of their fossil fuel subsidies (China, Germany, Indonesia, Italy, Mexico, and the United States) to invite the OECD Secretariat to chair those peer reviews once again lifted the OECD Secretariat involvement to a new level (Skovgaard, 2017a). Today, the OECD involvement in fossil fuel subsidies extends well beyond servicing the G20 (see Chapter 6 for more detail). A similar picture emerges regarding the IEA, which also addressed fossil fuel subsidies prior to the 2009 commitment, but which has increased its activities regarding such subsidies, including the number of reports dedicated to the topic since the commitment.

The World Bank’s involvement with fossil fuel subsidies was arguably more significant prior to 2009 than those of the OECD and the IEA, as it had not only provided an early analysis but had also promoted reform as part of its programmes (see Chapter 4). After 2009, it continued these efforts while providing an increasing amount of analysis targeting fossil fuel subsidies as a distinct phenomenon (Kojima, 2016; Kojima and Koplow, 2015; Strand, 2013). Its Energy Sector Management Assistance Program (ESMAP) facility has also provided assistance and knowledge for countries considering fossil fuel subsidy reform (Flochel and Gooptu, 2016; World Bank et al., 2015). The drastic increase in World Bank attention to fossil fuel subsidies happened a few years after the Pittsburgh commitment, and can be attributed to the increasing attention to fossil fuel subsidies among member states, officials and management as much as the direct effect of the G20 request. The fourth institution requested to provide an analysis, OPEC, has unsurprisingly not paid the same kind of attention to fossil fuel subsidies as the other institutions beyond the reports to the G20, but has addressed the impact of fossil fuel subsidies and their reform on oil demand (OPEC, 2016).

Beyond the requested institutions, the G20 commitment has led to the adoption of similar commitments to reforming, rationalising or phasing out fossil fuel subsidies within forums including the Asia-Pacific Economic Cooperation (APEC), the G7, the North American Leaders’ Forum and the Friends of Fossil Fuel Subsidy Reform (Friends). Friends was established in 2010 on the initiative of New Zealand inspired by the G20 commitment and with the intention of promoting the reform of fossil fuel subsidies (Rive, 2018). The group deliberately consists of countries that are not members of the G20 to promote the reform of fossil fuel subsidies beyond this group and avoid duplication. Without the G20 commitment, this institution would not have been created in 2010. The APEC, G7 and North American Leaders’ Forum commitments would not have been adopted without the G20 commitment, and include similar language (see Chapter 4), except that the G7 and North American Leaders’ Forum commitments also include deadlines for the phase-out. These forums overlap considerably with the G20 in terms of membership. Finally, fossil fuel subsidies moving up the agenda of international institutions, particular among economic institutions, following the G20 commitment was also an important factor in the IMF addressing fossil fuel subsidies (see Chapter 7). Furthermore, although the adoption of Sustainable Development Goal (SDG) 12.c to ‘rationalize inefficient fossil fuel subsidies that encourage wasteful consumption’ (United Nations, 2015) was not directly inspired by the G20 the way the other institutions’ commitments were, the wording of the SDG is very similar to the G20 commitment.¹ The fact that there was an existing commitment covering twenty of the largest economies, as well as the member states of both APEC and Friends, paved the way for the adoption of SDG 12.c.

5.3.2 Domestic Consequences

Turning to the influence on national fossil fuel subsidies, the G20 influence on the five selected countries is less clear cut (Skovgaard, 2018).² In the case of the United States, federal fossil fuel subsidies (defined as policies rather than non-priced externalities) consist of tax expenditure in support of producers of oil, gas and coal, and as consumption subsidies, particularly those directed at the energy costs of low-income households, together valued at several USD billions but falling at least until 2017 (OECD, 2020a). As a comparison, in 2018, the United States had a GDP of more than USD 20,000 billion (World Bank, 2020c). The US federal government has long acknowledged the existence of US fossil fuel production subsidies. The Obama administration tried to end tax breaks for fossil fuel production, but failed in the US Congress due to opposition from Democrats from fossil fuel producing states and Republicans (Rucker and Montgomery, 2011). Regarding the G20 reporting, the Obama administration submitted various self-reports and most notably participated in the first peer review. The US self-report from 2015 of the federal policies it considered to be fossil fuel subsidies was reviewed by a team chaired by the OECD Secretariat and included China, Germany and Mexico. In this report and in the 2014 G20 progress report, the United States acknowledged that the tax reductions and support for low-income households’ energy costs constituted fossil fuel subsidies, although the latter was not inefficient and hence should not be reformed (US Government, 2014, 2015). The 2015 report included four tax exemptions and a liability cap (in the range of USD 0 to 342 million) not included in the 2014 report (US Government, 2014, 2015). These five subsidies were identified in an inter-agency process carried out in anticipation of the peer review with the intention of identifying additional subsidies that merited inclusion (interview with US Treasury official, 20 December 2016). The Trump administration’s unwillingness to address fossil fuel subsidies and other climate issues both within the G20 and domestically meant there was little scope for G20 influence on US fossil fuel subsidies.

On the public agenda, the attention to fossil fuel subsidies has waxed and waned over the years (Table 5.1), focusing in the beginning of the period on domestic proposals to end tax breaks and in 2019 on climate action) and only referring to the G20 in a few instances in 2009, 2010 and 2015. As Table 5.1 shows, the total number of articles referring to fossil fuel subsidies increased with a peak of twenty-two in 2012. However, only a few of them referred both to fossil fuel subsidies (in a way that related to US subsidies) and the G20, most notably in 2009 when referring to the Pittsburgh commitment and the Obama administration’s role in bringing it about (Eilperin, 2009b; Shin and Eilperin, 2009). None of the articles made a connection between the G20 commitment and domestic fossil fuel subsidy reform (e.g. by referring to the commitment when discussing fossil fuel producers’ tax breaks). Not even the peer review of US fossil fuel subsidies caught the attention of the newspapers.

Table 5.1 Fossil fuel subsidies and the G20 in the US media: New York Times and Washington Post

	2009	2010	2011	2012	2013	2014	2015	2016	2017	2018	2019	Total
Articles referring to US fossil fuel subsidy reform and the G20	3	1	0	0	0	0	2	0	0	0	0	6
All articles referring to fossil fuel subsidy reform (international and domestic)	3	6	20	22	9	8	16	0	0	1	15	100

In this way, the G20 changed the policymaking agenda by placing the identification of fossil fuel subsidies on the agenda of several agencies not usually taking much interest in the issue, and the ideational context of action by reframing specific policies as fossil fuel subsidies and making it difficult to argue that they did not constitute such subsidies. A liability cap and two royalty exemptions for oil and gas extraction – which amounted to tens of million dollars annually – were identified in the reports to the G20 as fossil fuel subsidies that could be reformed without congressional approval. They were reformed in 2014 and in 2016 respectively, the latter immediately following the election of Donald Trump as president (Bureau of Land Management, 2016; US Government, 2015).The three subsidies that were reformed are among the subsidies that were not acknowledged until the 2015 report (and the only ones not requiring Congressional approval), and in this way, the Obama administration lived up to the G20 commitment as far as possible. Yet, the decision to reform the subsidies was well under way before the peer review and was adopted by the Department of the Interior in isolation from the policy processes addressing the G20 commitment (interview with senior Department of the Interior official, 15 December 2016; interview with US Treasury official, 20 December 2016). Under the Trump administration, the 2016 decisions to reform the two royalty exemptions were weakened, while the reforms of the liability cap remained in place (Bureau of Land Management, 2018).

The peer review agreed with the US self-review regarding the subsidies identified (including the Low-Income Home Energy Assistance Program [LIHEAP] not being inefficient), but also argued that the support for inland waterway infrastructure mainly used to transport fossil fuels – not included in the self-report – constituted a fossil fuel subsidy (G20, 2016b). Altogether, the G20 commitment institutionalised the norm of fossil fuel subsidy reform, which the Obama administration sought to adhere to within the domestic constraints, and for which it was held accountable regarding policies it was reluctant to define as fossil fuel subsidies. Yet, this norm was challenged by the Trump administration, which explicitly made support for coal, gas and oil extraction a priority, and weakened two of the Obama administration’s three reforms (Hermwille and Sanderink, 2019).

Regarding the United Kingdom, according to the OECD, direct fossil fuel subsidies consist mainly of reduced rates of value-added tax (VAT) for fuel and power, the covering of liabilities related to coal mining and tax breaks for oil and gas production, together estimated at several billion pounds (OECD, 2020a). This can be compared to the UK’s 2018 GDP of USD 2,850 billion (World Bank, 2020c). In recent years, the UK government has introduced new measures subsidising oil and gas production by allowing for increased deductions of extraction costs from corporate taxes (OECD, 2019f). The UK government has promoted fossil fuel subsidy reform at the international level, including within the G20 (UK Treasury Official, interview, 24 November 2014). Yet, in its reports to the G20 (as well as domestically), the UK government has argued that the UK provides no inefficient fossil fuel subsidies (Kirton et al., 2013; UK Department for Business, 2019a, 2019b, 2019c, UK Department of Energy and Climate Change and HM Treasury, 2013). This argument is based on the definition of fossil fuel subsidies as ‘any Government measure or programme with the objective or direct consequence of reducing, below world-market prices, including all costs of transport, refining and distribution, the effective cost of fossil fuels paid by final consumers, or of reducing the costs or increasing the revenues of fossil-fuel producing companies’ (UK Department for Business, 2019b; UK Department of Energy and Climate Change and HM Treasury, 2013).

Importantly, this claim was challenged by members of the UK Parliament, first and most notably the House of Commons’ Environmental Audit Committee (with members from all major parties) in its report on energy subsidies (2013). The report opened new venues for actors – including environmental non-governmental organisations (NGOs) and renewable energy companies – opposed to fossil fuel subsidies, many of whom testified to the Committee and influenced its report. The Committee used inter alia a price-gap approach that (unlike the government) included VAT in the benchmark price, and consequently lower VAT on inter alia the electricity bills of households and small businesses, and were defined as a (GBP 3.6 billion) subsidy. The Committee also – unlike the UK Government – defined tax rebates for high-cost oil and gas fields and fracking as subsidies. In this way, the ideational influence from the G20 commitment brought fossil fuel subsidies onto the policymaking agenda. Specifically, the government’s international commitment to the norm of fossil fuel subsidy reform not only brought attention to the concept of fossil fuel subsidies (a cognitive and agenda-setting dynamic), it also meant that the government could be held accountable to the norm even if it thought it was not relevant to the UK (the ideational dynamic known as entrapment; see also Schimmelfennig, 2001). Actors including members of the House of Commons’ Environmental Audit Committee pointed to the perceived inconsistency between the UK government’s commitment to the norm and high international profile on fossil fuel subsidy reform and the existence of, even growth in, fossil fuel subsidies domestically (Carrington, 2015a). In subsequent years, petitions to Parliament as well as questions to the UK government raised by members of both Houses of Parliament calling for the reform of UK fossil fuel subsidies were met by the similar response that the United Kingdom does not subsidise fossil fuels gov (HM Treasury, 2017a, 2017b, UK Department for Business, 2019a, 2019b, 2019c, UK Department of Energy and Climate Change and HM Treasury, 2013). Although the government engaged in cognitive discussions of whether the norm was relevant to UK policies and price levels, it could not argue that the norm was not salient. With regard to the G20, the UK government ruled out participating in a G20 peer review of its fossil fuel subsidies on the basis that it did not have such subsidies (HM Treasury, 2017b).

These dynamics also played out on the public agenda (Skovgaard, 2018). The number of newspaper articles mentioning fossil fuel subsidies has increased substantially since 2011 (Table 5.2). Several articles linked the G20 commitment to fossil fuel subsidies in the United Kingdom, and referred to the debate concerning whether the norm of fossil fuel subsidy reform was relevant to UK policies and the alleged inconsistency between the UK government’s international profile on fossil fuel subsidy reform and its domestic policy (Carrington, 2015a, 2019). This link was most pronounced in the period 2011–15, whereas in the subsequent years attention to fossil fuel subsidies increased, but the attention to the link between the G20 and fossil fuel subsidies subsided.

Table 5.2 Fossil fuel subsidies and the G20 in the UK media: The Guardian and The Independent

	2009	2010	2011	2012	2013	2014	2015	2016	2017	2018	2019	Total
Articles referring to UK fossil fuel subsidy reform and the G20	0	0	2	5	5	4	3	0	1	1	2	23
All articles referring to fossil fuel subsidy reform (international and domestic)	0	0	8	11	10	9	27	11	18	16	46	156

Concerning ideational dynamics at the level of officials, the Treasury was the ministry responsible for developing the UK government’s definition of fossil fuel subsidies and for the G20. The two other ministries with important roles – the Department of Energy and Climate Change and the Department for International Development focused mainly on the international level (interviews with a Department for International Development official, 24 November 2014; Department of Energy and Climate Change, 7 October 2014, 28 April 2020). The interaction in the G20 working groups raised awareness of the issue but did not lead to fundamental cognitive and normative changes of ideas regarding British fossil fuel subsidies in the Treasury.

In the case of India, fossil fuel subsidies in India consist mainly of selling kerosene and liquid petroleum gas (LPG) at a loss, and are estimated at INR hundreds of billions or USD billions (OECD, 2020a), or 1–2 per cent of GDP (IISD, 2014). Indian GDP was USD 2,700 billion in 2018 (World Bank, 2020c). National production subsidies are estimated at USD 1.5 billion (Bast et al., 2015). The Indian government acknowledges the existence of Indian fossil fuel subsidies, and has carried out a series of major reforms of consumption subsidies since 2013, liberalising prices and focusing subsidies on the poor (Garg et al., 2020; Jain et al., 2018). India is often hailed as a showcase of successful reform.

The Indian government has been sceptical of the G20 commitment, especially the G20 framing of fossil fuel subsides as an environmental issue, since the Indian government preferred to frame it as a purely economic and fiscal issue (for an example of this perspective, see Dasgupta, 2013). The scepticism reflects the historically predominant (yet increasingly challenged) view within the Indian elite that climate change is the responsibility of developed countries and that developing countries should not commit to climate change actions (Sengupta, 2019; Thaker and Leiserowitz, 2014). Nonetheless, the Indian government has implicitly acknowledged the relevance of the norm to India by reporting its plans to reform fossil fuel subsidies to the G20.

The framing of fossil fuel subsidies as a domestic and economic issue is mirrored in the public agenda, where Indian subsidies increased in importance with a peak in 2012–13 (when there was substantial discussion of whether and how to reform). After 2015, most of the reforms had been successfully implemented, and subsequent (less path-breaking) reforms received less attention. Thus, G20 ideational influence on the institutions on the public agenda is extremely limited, as only one newspaper article linked the G20 with domestic reform, and focused on India’s status as a G20 member rather than the G20 commitment (Nandi, 2017; see also Table 5.3).

Table 5.3 Fossil fuel subsidies and the G20 in the Indian media: The Hindu and Times of India

	2009	2010	2011	2012	2013	2014	2015	2016	2017	2018	2019	Total
Articles referring to Indian fossil fuel subsidy reform and the G20	0	0	0	0	0	0	0	0	1	0	0	1
All articles referring to fossil fuel subsidy reform (international and domestic)	0	1	10	35	37	19	17	4	7	4	4	138

The fossil fuel subsidy reforms have been the responsibility of the Ministry of Finance and the Ministry of Petroleum and Natural Gas. According to all former and current officials of the two ministries interviewed, the main reasons for undertaking these reforms have been fiscal and macroeconomic: there are cheaper ways of alleviating poverty, and the fossil fuel subsidies were detrimental to the public budget and the balance of trade (as they increased oil and gas imports). Two contextual factors – none of them linked to the G20 – made the reform possible: low oil prices and the liberalisation of the Indian economy since the early 1990s. Low oil prices created the scope in which to liberalise fuel prices without causing public protests.

Like India, Indonesia has considerable direct subsidies, which, according to the OECD, were constituted mainly by setting the prices of oil products and electricity below the market price and were estimated at around IDR 100 trillion or USD 7 billion USD (OECD, 2020a), currently at around 5 per cent of public expenditure (G20, 2019b).³ As a comparison, in 2018, Indonesia had a GDP of USD 1,000 billion (World Bank, 2020c). The Indonesian government acknowledges that these policies constitute fossil fuel subsidies, and has since 2000 attempted, with varying success, to reform them (Beaton et al., 2017; Chelminski, 2018). Most of the sizeable production subsidies for oil, coal and gas have been difficult to quantify, yet it is safe to say that they amount to USD billions (Bast et al., 2015; G20, 2019). Since Joko Widodo became president in 2014, consumption subsidies for petrol have been phased out, and diesel and electricity subsidies reduced, although production subsidies have not been reformed (IISD, 2015b, 2018).

Unlike India, Indonesia has been supportive of the G20 commitment and underwent a peer review of its fossil fuel subsidies in 2019 simultaneously with Italy (G20, 2019b). The Indonesian government has also continuously reported its plans and efforts to reform fossil fuel subsidies to the G20. The peer review forced the Indonesian government to undertake more in-depth analysis of its fossil fuel subsidies, especially in terms of collecting more data about its production subsidies (G20, 2019b, Government of Indonesia, 2019). The peer review commended Indonesia for its reforms, including the way they were communicated and that the subsidies targeted the poor, but also noted more recent increases in fossil fuel subsidies and recommended the gathering of further information about production subsidies (G20, 2019b). The fossil fuel subsidy reform norm has generally had an influence on government policymakers, since failure to live up to the commitment is considered politically embarrassing (interview with Indonesian Ministry of Finance officials, 14 September 2016). The Indonesian government has also highlighted its fossil fuel subsidy reforms in its voluntary reporting to the G20 on measures supporting ‘energy transitions and global environment for sustainable growth’ (G20, 2019a).

Nonetheless, the reforms have been driven by domestic economic concerns rather than concerns about the G20 commitment, which instead was influential in cognitive terms of encouraging the Indonesian government to study their subsidies.

The G20 ideational influence on the public agenda has been virtually non-existent (Table 5.4). Most newspaper articles focus solely on domestic aspects of subsidy reform. Only one article referred briefly to the G20 efforts to phase out fossil fuel subsidies. Generally, the Indonesian public are unaware of the existence of fossil fuel subsidies or tend to underestimate them (Chelminski, 2018).

Table 5.4 Fossil fuel subsidies and the G20 in the Indonesian media: Kompass and Tempo

	2009	2010	2011	2012	2013	2014	2015	2016	2017	2018	2019	Total
Articles referring to Indonesian fossil fuel subsidy reform and the G20	0	0	0	1	0	0	0	0	0	0	0	1
All articles referring to fossil fuel subsidy reform (international and domestic)	0	1	4	61	28	45	18	19	48	60	26	310

Finally, in Denmark, fossil fuels subsidies consist of reduced energy taxes for fuels used for specific purposes and for oil extraction. The subsidies as identified by the OECD are estimated to amount to above DKK 1 billion or USD 200 million (OECD, 2020a). This can be compared to the Danish GDP of USD 350 billion (World Bank, 2020c). Denmark is not a G20 member, and hence not subject to the 2009 commitment. Yet, it is an active member of the Friends of Fossil Fuel Subsidy Reform, which emerged due to the G20 commitment (see Chapter 4). Danish membership of Friends has not led to public discussions of Danish fossil fuel subsidies. Furthermore, Denmark has not subjected itself to a voluntary peer review of its fossil fuel subsidies within Friends in the same way as Sweden and Finland have done. Despite the increasing focus on fossil fuel subsidies since 2010, only two articles linked the G20 and Danish fossil fuel subsidies (Nielsen and Andersen, 2015). Generally, fossil fuel subsidies have been framed as an international (mainly developing country) phenomenon rather than a Danish one. The G20 commitment has had an indirect influence on Danish discussions of fossil fuel subsidies by increasing attention to such subsidies among IOs, NGOs and civil servants, which again led to the aforementioned discussions of fossil fuel subsidies.

Table 5.5 Fossil fuel subsidies and the G20 in the Danish media: Politiken and Jyllands-Posten

	2009	2010	2011	2012	2013	2014	2015	2016	2017	2018	2019	Total
Articles referring to Danish fossil fuel subsidy reform and the G20	0	0	0	0	0	0	1	1	0	0	0	2
All articles referring to fossil fuel subsidy reform (international and domestic)	0	0	3	3	5	8	9	1	10	4	13	56

5.4 Summary

The G20’s 2009 commitment was a catalyst for action on fossil fuel subsidies. It constitutes an important normative output, since it elevated the norm of fossil fuel subsidy reform from relative obscurity to a level of salience in which several institutions and most countries – also beyond the G20 – had to address it. Beyond the normative impact, the commitment has also had an important cognitive (and agenda-setting) impact in terms of raising awareness of fossil fuel subsidies on the international and domestic levels, and in terms of the knowledge about such subsidies produced by the four IOs requested to do so by the G20. The G20 output has not altered incentive structures. Subsequent output from the G20 has been more low-key and to some degree hindered by the lack of precision concerning the norm of fossil fuel subsidy reform as formulated in the commitment. Yet, the self-reporting and especially the peer reviews have forced G20 member states to address the salience of the norm to their domestic policies (especially in the case of the United Kingdom) and in the case of the peer reviews, have led to new knowledge about the subsidies of the countries reviewed (in the United States and Indonesia). All things considered, the consequences of the G20 output have been significant at the international level (especially in leading to similar international commitments), whereas the domestic consequences have been more limited but still relevant. The commitment has not in itself brought about any major fossil fuel subsidy reform.

The G20 commitment was triggered by US entrepreneurship utilising a conducive moment and reacting inter alia to the inaction within the UNFCCC (a reaction which amounts to institutional interaction). The commitment was shaped by this entrepreneurship, the membership circle (including insistence from some emerging economies on avoiding the norm being too precise) and the G20’s economic worldview. Subsequent G20 output has been less shaped by entrepreneurship (except for the peer reviews) and more by the membership circle, the G20 economic worldview and interaction with the IEA, OECD, OPEC and the World Bank. The pro-fossil fuel stance of the Trump administration has played a small but still significant role in limiting G20 efforts on fossil fuel subsidies, although the other nineteen G20 members have moved forward without the United States.

6 The OECD and Fossil Fuel Subsidies The Knowledge Provider

Both before and after the G20 commitment, the OECD has played a central role in providing knowledge about fossil fuel subsidies and promoting their reform. As an institution focusing on knowledge production, the OECD’s involvement has not attracted the kind of public attention that the G20 commitment has done. The OECD’s efforts have been more low key and, in many ways, linked to those of the G20, but undisputedly important to the international efforts to reform fossil fuel subsidies. Furthermore, the OECD involvement with these subsidies dates further back than that of the G20 and most other institutions.

This chapter proceeds with an outline of the OECD’s (formal and informal) knowledge output on fossil fuel subsidies, which have an important cognitive dimension in terms of defining what fossil fuels are and how we can understand their implications. The subsequent section explains that the OECD initially addressed fossil fuel subsidies on the OECD bureaucracy’s own initiative, but their involvement was lifted by interaction with the G20 and shaped by the institutional worldview inherent to the OECD Secretariat. This worldview emphasised the economic aspects of fossil fuel subsidies, but its influence was restricted by the limited autonomy of the OECD Secretariat. Finally, the consequences of the OECD output are discussed, with the finding that the output has had a cognitive influence on the domestic level and especially the international level.

6.1 Output: Knowledge Reigns Supreme

The OECD output on fossil fuel subsidies is knowledge oriented, either in the shape of formal publications or providing informal venues for learning about such subsidies and being socialised into the norm of fossil fuel subsidy reform. The OECD addressed fossil fuel subsidies before the G20 commitment as part of the regular environmental performance reviews of individual member states, studies of pricing policies and more general studies. Already in 1999, the OECD discussed fossil fuel subsidy reform as an instrument to achieve the Kyoto Protocol targets (OECD, 1999). The OECD Secretariat organised three workshops on environmentally damaging subsidies in 2002, 2003 and 2005 (OECD, 2002, 2003, 2005b), focusing mainly on finding a common definition for a subsidy and on methods for measuring subsidies and their environmental impact (Potier, 2002), that is, providing a forum for producing cognitive output in terms of learning and new knowledge. Most of the workshop presentations focused on subsidies in general or agricultural and fisheries subsidies, but a few addressed subsidies for coal, energy or transport (e.g. Franz, 2005). This was arguably a reflection of the fact that the OECD Directorate of Food, Agriculture and Fisheries was the main Directorate leading both workshops and publications, albeit in close cooperation with the Directorates of Environment, Trade as well as the IEA. The subsidies were framed in terms of their impact on sustainable development (OECD Secretariat, 2005; Pearce, 2002a). The Economics Department (the most influential OECD directorate) and the Environment Directorate have also – both before and after the 2009 G20 Pittsburgh commitment – focused on fossil fuel support in their ‘Economic Surveys’ and ‘Environmental Performance Reviews’ respectively, which analyse individual OECD member states. In the summer of 2009 – before the Pittsburgh G20 Summit – the OECD Secretariat published modelling of the climate change (and economic) impact of eliminating the subsidies measured by the IEA in The Economics of Climate Change Mitigation (OECD, 2009). This book was drafted by the Economics and Environment Directorates, and unlike subsequent OECD output framed fossil fuel subsidy reform in the context of adopting global carbon pricing. Their estimate that eliminating these subsidies would reduce emissions by 10 per cent was taken up by the G20 in the Pittsburgh commitment (see also Chapter 5 and Section 6.3).

Following the Pittsburgh commitment, the OECD increased its output on fossil fuel subsidies. This output was mainly drafted by the Trade and Agriculture Department with important input from the Environment Department and the Centre for Tax Policy and Administration. The OECD Secretariat co-authored reports to the G20 (IEA and OECD, 2018; IEA et al., 2011; OECD and IEA, 2019; OECD Secretariat, 2010a), and organised workshops on the topic. Perhaps most importantly, in 2011, the OECD published its first inventory of fossil fuel subsidies or ‘estimated budgetary support and tax expenditures for fossil fuels’ in twenty-four member states (OECD Secretariat, 2011). Given the sensitive nature of defining policies as fossil fuel subsidies in most OECD countries, this was an important cognitive output that reframed several national policies as subsidies, highlighting their negative environmental and economic impact. It also constituted important cognitive output in terms of collecting and producing new knowledge about these subsidies, including about their fiscal costs. The inventory has been updated regularly in the following years (OECD, 2012b, 2015b, 2018b), expanding the countries covered to all thirty-six member states as well as Argentina, Brazil, Colombia¹, China, India, Indonesia, Russia and South Africa). The OECD has also published country briefs about fossil fuel subsidies in the individual countries covered in the inventory since 2015, a potentially controversial output due to the sensitive nature of fossil fuel subsidies (see Section 6.4).

More recently, the OECD Secretariat has chaired the G20 voluntary peer reviews, thus playing an important role in the production of knowledge about the subsidies in the countries reviewed and to some degree also in the institutionalisation of the norm of fossil fuel subsidy reform. The peer reviews are central to the G20’s work on fossil fuel subsidies (see Chapter 5), and with the OECD Secretariat chairing the reviews, the sharing of best practices among countries are facilitated (Interview with senior OECD official, 3 February 2020). Importantly, the OECD Secretariat has significantly more expertise on the issue than the states participating in the review (International Organisations [IOs] such as the World Bank and the IEA also participate in some of them), and is a continuous presence in the peer reviews, unlike the other IOs. Altogether, the OECD Secretariat plays an influential role within the peer reviews. The OECD Secretariat’s expertise is also the background for the OECD Secretariat and the IEA reviewing review Dutch fossil fuel subsidies in a similar way to how they have reviewed fossil fuel subsidies as part of the G20 peer reviews.

In terms of defining fossil fuel subsidies, the OECD has opted for what is referred to as a Total Support Estimate, combining an inventory and a price-gap approach to identify both consumption and production subsidies (Jones and Steenblik, 2010; Koplow, 2018). Importantly, the Total Support Estimate does not include externalities in its benchmark price. The OECD tends to speak about support rather than subsidies, since the term subsidy is seen as referring to a smaller set of measures than support², and since it may be legally problematic in relation to WTO disputes to define a measure as a subsidy (interview with OECD officials, 29 April 2015). The OECD uses the term fossil fuel support in the way that others use fossil fuel subsidies. Hence, when analysing how the OECD has addressed fossil fuel subsidies, I will focus on its efforts regarding what it itself refers to as fossil fuel support. More recently, the OECD and the IEA have combined their respective estimates to produce joint estimates of global fossil fuel subsidies to the G20 (IEA and OECD, 2018; IEA, OPEC, et al., 2010; OECD and IEA, 2019) and have combined their knowledge outputs within a joint portal (OECD and IEA, 2020).

Importantly, the OECD Secretariat placed a strong emphasis on environmental (particularly climate) and fiscal consequences of fossil fuel subsidies, and rather less on the macroeconomic dimensions (e.g. the costs of subsidies to society and their distortionary nature) and significantly less on distributive consequences. For instance, the OECD Secretary-General Angel Gurría wrote in the foreword to the 2011 Inventory that ‘reforming or eliminating support for the consumption or production of fossil fuels can contribute to achieving economic and fiscal objectives, while also helping tackle environmental problems like climate change’ (OECD, 2011a). The fiscal emphasis is evident in that subsidies are measured in terms of budgetary expenditure and tax expenditure (reductions in tax rates also known as tax rebates), that is their impact on public budgets, an emphasis also present in the OECD’s work on agricultural subsidies. The macroeconomic costs of their climate impact are not measured.

6.2 Causes

The OECD’s output addressing fossil fuel subsidies was particularly shaped by two factors: institutional interaction and the OECD worldview. Institutional interaction constitutes the most important factor increasing OECD interest in fossil fuel subsidies (the first aspect of economisation) with less influence on how the OECD addressed the issue (the second aspect of economisation). The G20 is the most important institution in this respect. Its request to the OECD Secretariat to analyse fossil fuel subsidies and the implementation of the Pittsburgh commitment lifted OECD involvement to a new level (interview with OECD officials, 29 April 2015). In the following years, the OECD Secretariat arranged workshops for representatives of member states and reported (individually and with the IEA, OPEC and the World Bank) to the G20 on fossil fuel subsidies. Furthermore, it was only following the G20 commitment that the member states gave the OECD Secretariat the mandate to scrutinise their national fossil fuel subsidies (interview with OECD officials, 29 April 2015). More recently, the decision by the G20 members that have committed to voluntary peer reviews to invite the OECD Secretariat to chair those peer reviews once again lifted OECD Secretariat involvement to a new level (see also Chapter 5). Interaction with other institutions was more important in shaping how the OECD addressed fossil fuel subsidies. As regards the drafting of the reports, especially country-specific ones, the OECD staff collaborated with the World Bank and the IEA, which in this way influenced the OECD approach to the subject (interview with OECD officials, 29 April 2015).

The worldview proved more important regarding the second aspect of economisation, how the OECD addressed fossil fuel subsidies, particularly in cognitive terms. The OECD Secretariat is a bureaucracy characterised by emphasising the economic aspects and consequences of policy issues and instruments, and prioritising economic growth and development (see Chapter 3). Addressing fossil fuel subsidies fit such a worldview that in cognitive terms focuses on the economic impacts of policies and in normative terms prefers free-market policies. Fossil fuel subsidy reform has in many countries meant deregulation and leaving price-setting and investment decisions to the market without government interference. Such reform also agrees with the norm complex of liberal environmentalism, which the OECD Environment Directorate was instrumental in developing and which is prevalent in global environmental policy and predicates international environmental protection on a liberal economic order (Bernstein, 2001). Furthermore, the OECD Secretariat – consisting predominantly of officials with degrees in Economics – has the production of knowledge and data as its main task and prefers producing data that can be and are analysed econometrically and which highlights the economic consequences (Lehtonen, 2009; Ruffing, 2010).

In more specific terms, the OECD’s Total Support Estimate is rooted in its work on other subsidies, particularly agricultural subsidies. The Trade and Agriculture Directorate was the most important Directorate because of its agricultural subsidy expertise. Because of this institutional legacy, the OECD definition of fossil fuel subsidies is derived from the OECD definition of agricultural subsidies, again derived from the WTO’s definition of subsidies (of all kinds) as ‘direct transfers, fiscal incentives and provision of goods and services’ (OECD Secretariat, 2005, p. 17). The past experience of working with agricultural subsidies was entrenched in the organisation and hence influenced the definition of fossil fuel subsidies and the general framing of fossil fuel subsidies.

Other factors have also influenced the OECD efforts regarding fossil fuel subsidies to some degree. Policy entrepreneurs in the OECD Secretariat played an important role in seizing the opportunity provided by the increasing attention to the issue. Yet, they played a more important role regarding how the OECD addressed fossil fuel subsidies. Particularly Ronald Steenblik, until his retirement in 2018 the OECD’s chief fossil fuel subsidy expert and perhaps the foremost expert on the topic globally, has shaped the OECD’s approach both through internal discussions and policy and academic publications on fossil fuel subsidies published long before Pittsburgh (Jones and Steenblik, 2010; Steenblik, 1999, 2003, 2016; Steenblik and OECD, 2003; Steenblik et al., 2018). The influence of Steenblik and other policy entrepreneurs within the OECD Secretariat is most pronounced regarding the Total Support Estimate, which they have been committed to, and the fact that they have often been critical of alternative approaches such as the IMF’s (Steenblik, 2014).

The influence of policy entrepreneurs and the institutional worldview of the Secretariat have been circumscribed by the role of another set of factors, namely the relations with member states. Member states have not been directly involved in drafting the most important OECD publications, which were published on behalf of the OECD Secretariat. Yet, member states played an indirect role by limiting how far the OECD staff could go (interview with Swedish Ministry of Foreign Affairs official, 30 April 2015). The consensual nature of OECD policy processes, even as regards OECD Secretariat publications (which are discussed but not approved by member states) means publications that go directly against the preferences of large groups of members states are highly unlikely (Carroll and Kellow, 2011). Thus, the relatively low degree of OECD Secretariat autonomy (at least compared to the IMF or the World Bank) acted as a scope condition for the influence of the institutional worldview and policy entrepreneurs, especially before Pittsburgh. If the Secretariat had had more autonomy than is actually the case, the institutional worldview and policy entrepreneurs could have been more influential.

The membership circle of the OECD covers more countries than the G20, but the thirty-seven OECD countries only include developed countries plus a few countries (Colombia Mexico, Chile and South Korea) classified by the World Bank as upper-middle income countries. The OECD addressed the fossil fuel subsidies of its member states as well as a number of other selected states, and addressed their subsidies differently from how the IMF or the IEA did, although the membership circle is very similar to the IEA’s (only Chile, Colombia, Iceland, Israel and Slovenia are members of the OECD but not the IEA). Altogether, there is far from a perfect correlation between the preferences of the member states and the output of the OECD. Rather, the output was shaped predominantly by the institutional worldview and to some degree entrepreneurs within the OECD within the limits of the OECD Secretariat’s autonomy.

6.3 Consequences

6.3.1 International Consequences

Given that the OECD’s output has been ideational, mainly in terms of providing cognitive knowledge about fossil fuel subsidies and how to reform them, but also promoting the norm of fossil fuel subsidy reform, it is not surprising that the consequences have been ideational. At the international level, the most important consequence of the OECD’s output on fossil fuel subsidies has been its effect on the G20. The OECD together with the three other IOs has provided several reports to the G20, which have shaped the debate within the G20 working groups and consequently also G20 output. In the working groups, the OECD could disseminate knowledge about fossil fuel subsidies to officials, who especially in the early years had little knowledge of what was still an emerging subject. More recently, the OECD Secretariat chairing the G20 peer reviews has meant significant influence over the definition of what constitutes a fossil fuel subsidy in the countries reviewed. The G20 has de facto adopted an approach similar to the OECD’s, identifying fossil fuel subsidies in terms of policies, rather than price levels, and focusing on a wide range of policies including reduced tax rates, the provision of services at below-market rates, etc., including for the production of fossil fuel subsidies. When the OECD approach was used in the 2016 G20 peer reviews of fossil fuel subsidies in the United States and China, it became the de facto standard for identifying fossil fuel subsidies, as is also evident in the efforts to measure fossil fuel subsidies in the context of the Sustainable Development Goals (SDGs) (see later in this section).

Counterfactually, had the peer reviews used a different definition of fossil fuel subsidies, e.g. one based on the IEA’s price-gap approach or the IMF’s inclusion of non-priced externalities, both the identification of fossil fuel subsidies and the policy recommendations would have been different (see also Chapter 7). The OECD’s framing of fossil fuel subsidies in economic, especially fiscal, terms thus influenced how the G20 addressed the issue. Hence, the OECD’s economisation of fossil fuel subsidies contributed to the ongoing economisation of the issue in the G20. The G20 treatment of fossil fuel subsidies is also shaped by other factors than the OECD Secretariat, which was only one of the participants in the peer review, but still constituted an important and constant factor.

Beyond the G20, the OECD has also contributed technical expertise to APEC peer reviews. The OECD Secretariat officials are closely involved in the development of a definition of fossil fuel subsidies in the monitoring of the SDGs. This definition is important, as it will be used to monitor all signatories to the SDGs’ (virtually all countries in the world) efforts to live up to SDG 12.c and to ‘rationalize inefficient fossil fuel subsidies that encourage wasteful consumption … ’ (United Nations, 2015). More precisely, it is used to measure progress in terms of indicator 12.c.1 the ‘Amount of fossil fuel subsidies per unit of GDP (production and consumption)’, an economic framing of fossil fuel subsidies (UNEP et al., 2019). The definition was adopted by the SDG Expert Group on Fossil Fuel Subsidies chaired by Ron Steenblik from the OECD Secretariat and Pete Wooders from the International Institute for Sustainable Development (IISD). It draws on the fossil fuel definitions of the WTO, which the OECD also draws on and which identifies fossil fuel subsidies in terms of policies, as well as the IEA definition which identifies them in terms of prices (UNEP et al., 2019). More specifically, it includes direct budget transfers (based on the WTO definition), so-called induced transfers (based on the IEA definition and measured in terms of price-gaps rather than policies and also used by the OECD), as well as ‘tax expenditures, other government revenue foregone and under-pricing of goods and services, including risk’ (UNEP et al., 2019, p. 41) which is optional for governments to report. Risk transfers, part of the OECD definition, are not included.

6.3.2 Domestic Consequences

The domestic consequences of the OECD output are harder to discern.³ The OECD’s output is cognitive and normative, and it is difficult to distinguish the influence of the OECD from other international institutions on the domestic level. In the case of all the five countries studied, the OECD published country briefs focusing specifically on fossil fuel subsidies and also addressed the issue in environmental performance reviews. It also arranged several workshops for officials on fossil fuel subsidies, which was important especially immediately after the Pittsburgh commitment. At this time, fossil fuel subsidies constituted a new concept the understanding of which was still to be shaped, also among officials from finance and energy ministries (author’s observation as a Danish official participating in one of these workshops in 2010).

Regarding the United States, the OECD identified federal fossil fuel subsidies to producers of oil, gas and coal, as well as to the energy costs of low-income households, together valued at several USD billions (OECD, 2020a). OECD influence on the United States was arguably most salient in the case of the G20 peer review, which as discussed in Chapter 5 constituted a small but significant effect on the Obama administration’s efforts to reform US fossil fuel subsidies. Besides the peer review, the OECD had already recommended reforming environmentally harmful subsidies in its environmental performance reviews (OECD, 1996, 2005a) and in its country brief on US fossil fuel subsidies (OECD, 2019 g). Yet, the OECD defining specific policies as subsidies in these reports had little impact, since these policies were already acknowledged as subsidies, and actors in the United States, including environmental NGOs and parts of the Democratic Party, sought to reform them but faltered due to domestic opposition (see also Chapter 5). On the public agenda, very few references were made to the link between the OECD and fossil fuel subsidies (see Table 6.1), but in one case referring to the OECD’s definition of fossil fuel subsidies as including tax breaks for fossil fuel companies and to such tax breaks in the United States (Schwartz, 2015).

Table 6.1 Fossil fuel subsidies and the OECD in the US media: New York Times and Washington Post

	2009	2010	2011	2012	2013	2014	2015	2016	2017	2018	2019	Total
Articles referring to US fossil fuel subsidy reform and the OECD	0	0	0	1	0	0	2	0	0	0	0	3
All articles referring to fossil fuel subsidy reform (international and domestic)	3	6	20	22	9	8	16	0	0	1	15	100

In the case of the United Kingdom, the OECD identified fossil fuel subsidies worth billions of pounds targeting heating and power consumption by households as well as the production of oil, natural gas and coal (OECD, 2019f, 2020a). In terms of promoting the norm of fossil fuel subsidy reform, the OECD Secretariat commended the United Kingdom for phasing out sizable subsidies for coal production, but also noted the introduction of tax reductions for oil and gas production (OECD, 2019f). The OECD’s influence was most pronounced with regard to the House of Commons’ Environmental Audit Committee’s (which includes members of all major parties) report on energy subsidies (2013) and the related debate about UK subsidies.

While the report and the debate were brought on by the G20 commitment and the UK government’s claim that it provides no inefficient fossil fuel subsidies (Kirton et al., 2013), the OECD’s output on fossil fuel subsidies shaped the report and the debate. The OECD’s definition of fossil fuel subsidies as including all policies that confer benefits on fossil fuels, including reduced tax and value-added tax (VAT) rates, differed from the UK government’s definition of fossil fuel subsidies as ‘any Government measure or programme with the objective or direct consequence of reducing, below world-market prices [author’s emphasis], including all costs of transport, refining and distribution, the effective cost of fossil fuels paid by final consumers, or of reducing the costs or increasing the revenues of fossil-fuel producing companies’ (UK Department of Energy and Climate Change and HM Treasury, 2013, item 112). In other words, policies that lowered fuel prices below what they would be if they were fully taxed but above the world market price including transport, refining and distribution costs, would be considered a subsidy by the OECD but not by the government. Hence, the OECD’s finding that the United Kingdom’s reduced VAT rate of just 5 per cent on the heating fuel and power consumption of private households constituted a subsidy worth billions of pounds annually ran counter to the UK government’s claims but was used by the Environmental Audit Committee (2013). Likewise, the Environmental Audit Committee drew on OECD fossil fuel subsidy reports to identify reduced tax rates for fossil fuel extraction as subsidies (House of Commons Environmental Audit Committee, 2013). In 2019, the OECD’s definition became again the topic of debate, when the UK government responded to questions from the House of Lords’ European Union Committee and a petition to redirect fossil fuel subsidies to renewable energy with reference to the United Kingdom not having such subsidies according to the IEA definition, and that the OECD’s inclusion of individual tax reliefs is ‘devoid of their context within our overall tax regime’ (HM Treasury, 2019, item 2; UK Department for Business, 2019c). This reply led the House of Lords’ European Union Committee to suggest that the UK government should use the OECD’s definition, as this was more compatible with UK climate leadership (House of Lords European Union Committee, 2019).

As is evident in Table 6.2, the OECD had only a limited impact on the UK public agenda, and only in the period 2012–15. Moreover, few of these articles directly linked OECD estimates and UK fossil fuel subsidies, although the Overseas Development Institute and Oil Change International’s (two NGOs) use of OECD data to criticise G20 countries’ fossil fuel subsidies was referred to (Vidal, 2014b).

Table 6.2 Fossil fuel subsidies and the OECD in the UK media: The Guardian and The Independent

	2009	2010	2011	2012	2013	2014	2015	2016	2017	2018	2019	Total
Articles referring to UK fossil fuel subsidy reform and the OECD	0	0	0	1	1	1	2	0	0	0	0	5
All articles referring to fossil fuel subsidy reform (international and domestic)	0	0	8	11	10	9	27	11	18	16	46	156

Thus, the OECD had an important cognitive influence on the policy debate regarding fossil fuel subsidies, especially regarding the question of what constitutes a subsidy. It played a more indirect normative role, as the House of Lords’ European Union Committee associated the OECD’s approach to identifying fossil fuel subsidies with climate leadership. Yet, this did not lead to reform of fossil fuel subsidies, as the United Kingdom has rather increased fossil fuel subsidies for oil and gas production through improving the possibilities for deducting extraction costs from corporate taxes (OECD, 2019f).

India is not an OECD member state but has nonetheless been part of the OECD’s inventories of fossil fuel subsidies. According to the OECD, fossil fuel subsidies in India consist mainly of selling diesel, kerosene and liquid petroleum gas (LPG) at a loss, and are estimated at INR hundreds of billions or several USD billions (OECD, 2019d). The OECD acknowledged the substantive Indian fossil fuel subsidy reforms. These conclusions have not caused debate in India, since the policies are widely acknowledged as constituting subsidies. Indian fossil fuel subsidies have been significantly reformed since 2013 (Garg et al., 2020; Jain et al., 2018), but these reforms were driven by domestic fiscal and macroeconomic concerns (see also Chapter 5). The OECD played a more background role compared to the GSI and the World Bank as regards providing opportunities for learning about how to best undertake such reform (Lemphers et al., 2018; Skovgaard, 2018). Regarding the public agenda, it is perhaps unsurprising given that India is not a member of the OECD, that the Indian newspapers studied have not even once linked the OECD and Indian fossil fuel subsidies (see Table 6.3).

Table 6.3 Fossil fuel subsidies and the OECD in the Indian media: The Hindu and Times of India

	2009	2010	2011	2012	2013	2014	2015	2016	2017	2018	2019	Total
Articles referring to Indian fossil fuel subsidy reform and the OECD	0	0	0	0	0	0	0	0	0	0	0	0
All articles referring to fossil fuel subsidy reform (international and domestic)	0	1	10	35	37	19	17	4	7	4	4	138

Indonesia, like India, is not an OECD member state but has been part of the OECD’s inventories of fossil fuel subsidies. The OECD identifies fossil fuel subsidies in Indonesia as consisting of setting the price on oil products and electricity below the market price as well as of production subsidies, and estimated them at around IDR 100 trillion or USD 7 billion (OECD, 2020a). The OECD also commended Indonesia for its subsidy reforms, but noted that fuel prices were fixed in 2018 in response to rising international oil prices (OECD, 2019e). More importantly, the OECD Secretariat chaired the G20 peer review. As discussed in Chapter 5, the peer review forced the Indonesian government to undertake a more in-depth analysis of its fossil fuel subsidies, especially production subsidies, and commended Indonesia for its reforms (G20, 2019b). The G20 followed an approach to defining fossil fuel subsidies in line with the OECD’s (see Section 6.4.1). The OECD’s chairing of the G20 peer review meant it could play a cognitive role in shaping the G20 analysis of Indonesian subsidies, including the focus on production subsidies, which the OECD has prioritised to a greater extent than the IEA or the IMF. Nonetheless, it may be difficult to precisely distinguish when the OECD’s influence began and the G20’s influence ended in this respect. Yet, Indonesian reforms have been driven by domestic economic concerns rather than the G20 or the OECD. In terms of cognitive influences such as learning about successful reform, the OECD has participated in workshops about Indonesian reform, but have not played the same role as the World Bank and the IMF (Diop, 2014; interview with senior Bappenas official, 20 December 2016). Just as in India, none of the newspapers studied have linked the OECD and Indonesian fossil fuel subsidies.

Table 6.4 Fossil fuel subsidies and the OECD in the Indonesian media: Kompass and Tempo

	2009	2010	2011	2012	2013	2014	2015	2016	2017	2018	2019	Total
Articles referring to Indonesian fossil fuel subsidy reform and the OECD	0	0	0	0	0	0	0	0	0	0	0	0
All articles referring to fossil fuel subsidy reform (international and domestic)	0	1	4	61	28	45	18	19	48	60	26	310

Finally, regarding Denmark, according to the OECD, the Danish government subsidises fossil fuels by reducing energy taxes for fuels used for specific purposes (combined heat and power generation) and diesel. The subsidies that the OECD was able to estimate amounted to above DKK 1 billion or USD 200 million, although falling significantly after 2015 due to reductions in tax rebates for diesel and in the support for combined heat and power (Danish Ministry of Climate Change, 2019, OECD, 2019c, 2019h, 2020a). These reforms seem to have been driven at least as much by economic concerns as by environmental ones. The 2019 OECD Environmental Performance Review mentioned Danish fossil fuel consumption subsidies, but did not include the reform of these subsidies in its forty-four recommendations for improving Danish environmental policy, and unsurprisingly these subsidies were not subsequently addressed in Danish policy or public agendas, which focused on the recommendations. Otherwise, opposition Members of Parliament have raised the issue of Danish fossil fuel subsidies a couple of times in the Danish Parliament (Danish Ministry of Climate Change, 2015; Danish Ministry of Taxation, 2019). In this context, the Danish government has acknowledged that fossil fuel production is subsidised, but argues that tax expenditure on consumption (defined by the OECD as subsidies) do not constitute subsidies since total fossil fuel taxes exceed the total externalities (Danish Ministry of Climate Change, 2015). Later, in its 2019 National Energy and Climate Plan submitted to the EU, the Danish government acknowledged the existence of indirect fossil fuel subsidies, highlighted that some of these subsidies were being phased out, and that the government would look into further fossil fuel subsidy reform (Danish Ministry of Climate Change, 2019). In general, the various Danish governments have been hesitant to acknowledge that particular policies constitute fossil fuel subsidies as this was seen as contrary to the Danish promotion of the norm of fossil fuel subsidy reform.

Concerning cognitive influences, participation in workshops on fossil fuel subsidies arranged by the OECD enhanced knowledge on and awareness of the topic within the Finance Ministry and other ministries (author’s observation, including as a participant on one of those workshops). Yet, within the Danish ministries, fossil fuel subsidies have been perceived as mainly a developing country phenomenon, which did not necessitate changes to Danish policy (interview with senior Danish government official, 13 January 2014). While there were discussions within the Danish government of undertaking a self-review of Danish fossil fuel subsidies using the OECD’s definition, it was decided not to carry out the review as the timing would collide with the ongoing analysis of Danish energy subsidies, and (at the time) could be politically sensitive (interview with former Danish senior civil servant, 5 May 2020). Consequently, learning has mainly been relevant in terms of changing Danish beliefs regarding how to best undertake fossil fuel subsidy reform in developing countries, not in developed ones.

The OECD reports have not led to any changes to the public agenda (see Table 6.5), where they have not appeared (although one article mentioned the OECD’s role in the G20 peer review of Chinese fossil fuel subsidies; Andersen, 2016).

Table 6.5 Fossil fuel subsidies and the OECD in the Danish Media: Politiken and Jyllands-Posten

	2009	2010	2011	2012	2013	2014	2015	2016	2017	2018	2019	Total
Articles referring to Danish fossil fuel subsidy reform and the OECD	0	0	0	0	0	0	0	1	0	0	0	1
All articles referring to fossil fuel subsidy reform (international and domestic)	0	0	3	3	5	8	9	1	10	4	13	56

6.4 Summary

The OECD, a knowledge producing institution, has unsurprisingly had the greatest impact in terms of cognitive output defining what fossil fuel subsidies are and how we can understand their implications. Thus, the OECD has produced an extensive amount of cognitive knowledge about what is important about fossil fuel subsidies and which policies in individual countries can be understood as fossil fuel subsidies. A crucial aspect of this output has been the framing in terms of economic, particularly fiscal impact. The extensive data gathering endeavour as well as the efforts to promote the OECD’s definition has had a significant impact on how fossil fuel subsidies – until 2009 an extremely marginal issue – have been addressed, especially at the international level. At the domestic level, the influence of the OECD is more difficult to discern, not least because the OECD Secretariat has often been most influential when it has chaired G20 peer reviews, making it rather difficult to precisely distinguish the OECD’s influence from the G20’s. It is worth contemplating how different the international efforts to promote fossil fuel subsidy reform might have been had the OECD adopted a different definition, for example, if it had included the non-pricing of externalities in the way the IMF (Coady et al., 2015, 2019) has done, or development finance supporting fossil fuel production, the way the NGOs such as Oil Change International have done (Oil Change International et al., 2017). The OECD output has primarily been driven by institutional interaction with the G20, which lifted OECD involvement to a new level, and the institutional worldview inherent to the OECD Secretariat, which emphasises the economic aspects of fossil fuel subsidies. The relatively limited autonomy of the OECD Secretariat has acted as a scope condition for the influence of this worldview.

7 The IMF and Fossil Fuel Subsidies The Unexpected Environmentalist

The IMF is one of the most powerful institutions in the world, often criticised for forcing governments to adopt fundamental policy changes that reflect the worldview of IMF officials rather than the preferences of the government or the electorate (Barnett and Finnemore, 2004). The criticism has particularly focused on the IMF’s insistence on promoting policies rooted in the so-called Washington Consensus including trade, financial and labour market liberalisation and cutting fiscal deficits (Babb, 2013; Chwieroth, 2008). Environmental protection has rarely been included among the IMF’s top priorities, and the IMF has been criticised for the environmental consequences of its policies (Harvey, 2005; Shandra et al., 2011). Against this backdrop, the IMF’s involvement with fossil fuel subsidies is striking: not only did it seem to emerge out of the blue in 2013, but it also placed the IMF among the most radical of the institutions addressing climate change. This chapter proceeds with an outline of this seemingly radical change as well as the IMF’s use of conditionalities to promote fossil fuel subsidy reform, followed by a discussion of how this output was driven by IMF officials and shaped by the economic worldview of the IMF. Finally, the chapter finds that the output had some cognitive impact at the international level but had more important ramifications at the domestic level.

7.1 Output: Conditionalities and Carbon Pricing

IMF output has addressed fossil fuel subsidies along two strands, both increasing in importance. First, a strand consisting of bilateral interactions with countries experiencing fiscal problems exacerbated by fossil fuel subsidies and including policy recommendations and IMF programmes inter alia promoting subsidy reform to improve fiscal balances (interview with IMF senior official, 17 February 2015). This strand is composed mainly of distributive output in the shape of conditionalities, particularly that the country in question would receive IMF loans only if it embarked on reforms including subsidy reform. Yet, to a lesser degree, the strand also consists of knowledge output in the shape of analyses of the consequences of such reforms and the best ways to undertake them. For decades, the IMF has been opposed to subsidies, especially consumer subsidies fixing the prices of goods, because of their fiscal and macroeconomic impact. The former has been problematic in the eyes of the IMF because subsidies constitute sizeable budget items including at times of fiscal deficits, the latter because subsidies distort the allocation in markets by fixing or supporting prices. At times, subsidies (for fossil fuels as well as other products such as food) have been phased out or cut significantly as a result of IMF insistence, for example, in Egypt (Sherry et al., 2014) and Sudan (Michael, 2013). Over the past ten years, fossil fuel subsidies have been singled out in an increasing number of IMF recommendations to individual countries and have increasingly been treated as distinct from other kinds of subsidies. The IMF has suggested phasing out such subsidies because they are an inefficient, fiscally costly and often economically distorting way of providing welfare benefits (interview with IMF officials, 9 April 2014). The ongoing IMF Extended Credit Facility Arrangement (a kind of IMF programme) for Burkina Faso constitutes an example of this, with detailed recommendations on how Burkina Faso should liberalise the government-fixed fuel prices (IMF, 2015a; interview with senior IMF official, 22 June 2016). This strand mainly emphasises the fiscal consequences of the fossil fuel subsidies, while also stressing the distributional (both positive and negative) and macroeconomic consequences. Environmental consequences (including local externalities) have been accentuated to a lesser extent. In the case of Burkina Faso, the IMF policy was developed by local IMF officials and officials from the Fiscal Affairs Department, and focused on what the IMF refers to as pre-tax subsidies, namely the subsidies lowering consumer prices below the international market price plus distribution costs, not the absence of a full pricing of the externalities (interview with senior IMF official, 22 June 2016). Not including an incomplete pricing of externalities distinguishes the IMF’s approach to fossil fuel subsidies in Burkina Faso from its approach to fossil fuel subsidies globally, which did include such non-pricing (see subsequent discussion of the IMF’s fossil fuel subsidy definition).

The second strand consists of knowledge output and focuses on the lack of a carbon price (and environmental taxes generally) and on solving this problem from the perspective of an economist, that is, getting the price right (interview with IMF senior official, 17 February 2015). Prior to 2008, the IMF only occasionally addressed subsidies including energy (rather than fossil fuel) subsidies in policy reports¹ (Baig et al., 2007; Gupta et al., 2000) and in country-specific policy recommendations (see e.g. IMF 2004). Energy subsidies were framed in terms of fiscal and macroeconomic impact, without referring to the environmental impact. Thus, subsidising fossil fuels was framed as being similar to subsidising any other product. The IMF used price-gap approaches to measure all kinds of subsidies, and did not include externalities in its fuel benchmark prices (e.g. Said and Leigh, 2006), which meant that getting the price right at this stage did not signify including the costs of externalities in the price.

In 2008 the first official publications² to address fuel subsidies as a distinct concept and to include environmental externalities (priced at USD 0.50 per litre petrol and diesel) were published (IMF, 2008a, 2008b). They would provide the blueprint for future IMF output in fossil fuel subsidies. After 2008 the IMF increasingly addressed fossil fuel subsidies and their environmental impact, while maintaining the emphasis on fiscal and macroeconomic consequences. In 2013, they published the report ‘Energy subsidy reform: lessons and implications’ (Clements et al., 2013) which raised the IMF’s engagement with fossil fuel subsidies to a new level. Crucially, the report used a price-gap approach based on a benchmark price including both value-added tax (VAT) and the social cost of externalities, particularly climate change and other environmental externalities such as local air pollution. This approach was adopted on the basis of the IMF’s work on fiscal instruments (in the Fiscal Affairs Department) and the emphasis on – as the title of a key publication on fossil fuel subsidies says – Getting Energy Prices Right (Parry et al., 2014). The IMF has also offered an online course on Energy Subsidy Reform based on these publications. The notion of ‘getting prices right’ constitutes a framing of fossil fuel subsidies as being problematic in normative terms because they cause prices to be ‘wrong’ in the sense of preventing the optimal allocation in markets. A price is right if it reflects all production-related costs to society as well as all benefits to the buyer, as determined in a market where the only government intervention consists of taxes addressing negative externalities, or subsidies addressing positive externalities. Importantly, the notion of including non-priced externalities in the definition of a subsidy was hardly discussed outside the IMF prior to 2013 (one exception being Hodas, 2006).

The cognitive and normative dimensions of this notion are closely intertwined: fossil fuel subsidies are defined (cognitively) in terms of getting the prices wrong, and they are normatively problematic because they get the prices wrong. One can compare this framing of the right price to the proponents of fossil fuel consumption subsidies, who argue that the price is right if it is low enough to allow everybody to buy it, and – in the case of fossil fuel producing countries – allows for people to get ‘their share’ of the fossil fuel resources of that country (Cheon et al., 2013). Regarding distributive consequences, the IMF framed fossil fuel subsidies as a highly ineffective way of supporting low-income households, since most of the subsidies are captured by citizens with above-average incomes (Arze del Granado and Gillingham, 2012).

The framing also closely linked fossil fuel subsidies to another policy issue, namely carbon pricing. The IMF reports explicitly referred to Pigouvian or corrective taxes as the optimal way of correcting externalities, drawing on the works of the economist Arthur Pigou. Pigou (1932) invented the concept of an externality defined as the cost or benefit of an activity undertaken by one actor that affects another actor not involved in the activity, thus creating a suboptimal situation since the costs of the activity do not reflect the true costs or benefits to society (see also Chapter 1). Furthermore, Pigou developed the notion that externalities should be corrected by placing taxes (or financial support) on the activity creating the externality that corresponds to the social costs (or benefits). For instance, burning coal incurs costs to other parts of society that are not borne by the polluter, and consequently should be corrected by a Pigouvian tax on coal burning corresponding to the environmental damage. Or a positive externality such as non-patented research should receive support corresponding to the social benefit. The linking of fossil fuel subsidies and carbon pricing is closely related to the IMF increasingly advocating carbon pricing, particularly carbon taxes, as the most important climate mitigation instrument (IMF, 2019c, 2019e; Lagarde, 2015; Mooij et al., 2012; Parry, 2019; Parry et al., 2018).

Importantly, the IMF explicitly endorsed environmental taxes over other environmental policy instruments, including not only regulatory standards such as emission standards, which were never popular among neoclassical economists (Lauber and Schenner, 2011; Meckling and Allan, 2020), but also over emission trading schemes, an alternative carbon pricing instrument. Neoclassical environmental economists have been divided over whether environmental taxes or emissions trading schemes constitute the most optimal instrument. Whereas advocates of taxes draw on Pigou’s arguments for letting the polluter bear the social costs of pollution, advocates of trading schemes draw on economist Ronald Coase (1960), who argued that the right to pollute should be divided between actors, who could then trade these rights and thus create a price on emissions through trading. According to Coase, the social costs of pollution should not necessarily be borne by the polluter. Other actors working with fossil fuel subsidies had not linked it to carbon pricing, and in carbon pricing circles, fossil fuel subsidies had been mentioned only as a detriment to effective carbon pricing without framing the absence of carbon pricing as a subsidy (see OECD, 2009). It is worth comparing the IMF’s endorsement of Pigouvian taxes to that of the World Bank, which has been very active in promoting carbon pricing, but initially promoted emissions trading and only at a later stage promoted carbon taxes on a par with emissions trading (Skovgaard and Canavan, 2020).

The findings of the report were extended and updated in a 2015 IMF working paper (Coady et al., 2015), whose estimate of global subsidies at USD 5 300 billion³ in 2015 – compared to estimates of USD 1900 billion in the 2013 report USD (Clements et al., 2013) and USD 550 billion in the IEA’s 2014 World Energy Outlook (IEA, 2014) – received significant attention. This increase was due to a revision upwards of the assessments of externalities – especially air pollution. In 2019, the IMF updated its 2015 analysis with new data and found that very little had changed: global fossil fuel subsidies were estimated at USD 5,200 billion for 2017 (Coady et al., 2019).

According to both the 2015 and 2019 analyses, local air pollution accounted for little less than half of the subsidies, climate change for little less than one quarter, with reductions to consumption taxes, direct subsidies and non-air local externalities (congestion, traffic accidents, etc.) for the remaining part (Coady et al., 2015, 2019). Regarding climate change, the social cost of emitting a ton of CO₂ was in 2015 estimated at USD 35 (based on Parry et al., 2014) and USD 40 in 2019 (Coady et al., 2019), figures originating from the US government’s Interagency Working Group on Social Cost of Carbon (respectively Interagency Working Group on Social Cost of Carbon, 2013, 2016). The USD 35 and 40 per ton social costs are low estimates, based on William D. Nordhaus’ DICE model, which has been criticised for seriously under-estimating the costs of climate change (Howard and Sterner, 2017). In comparison, Richard Tol’s (2011) meta study of different estimates of the social cost of carbon arrives at a mean of USD 177 per ton (although most of the estimates are below this mean), and the High-Level Commission on Carbon Pricing (2017) estimated that carbon prices should be between USD 40 and 80 per ton by 2020 to meet the Paris Agreement’s target of temperature increases ‘well below 2 degrees’ (UNFCCC, 2015). Interestingly, one of the IMF’s reports from 2019 advocating carbon taxes operated with a carbon price of USD 75 per ton being required to meet the 2-degree target (IMF, 2019e).⁴ Yet, the USD 35 and USD 40 estimates are high compared to actual levels of carbon taxation. Only about 20 per cent of global emissions are currently covered by such taxes, with about half of these emissions priced at below USD 10 per ton, and only 5 per cent of them above USD 40 per ton (World Bank, 2020b).

According to the IMF’s definition, practically all states subsidise fossil fuels, even countries with carbon taxes that do not fully price externalities. Developed countries account for a quarter of energy subsidies, and emerging and developing Asia for half (Coady et al., 2015). Consequently, the 2013 and 2015 reports break with previous analyses which – using different definitions – identified fossil fuel subsidies as a phenomenon primarily seen in developing countries. The reports also contradicted the claims of influential IMF member states (including the United Kingdom and Japan) that they do not subsidise fossil fuels these (Kirton et al., 2013), something which made the report unpopular among the governments of those countries but popular among NGOs. In 2020, the IMF planned to integrate climate mitigation issues, including carbon pricing, in its bilateral Article IV consultations (which are carried out with all member states), although these consultations were temporarily postponed because of the coronavirus pandemic (interview with senior IMF official, 19 May 2020).

7.2 Causes

The most important factor influencing the decision to address fossil fuel subsidies (the first aspect of economisation) was policy entrepreneurs within the IMF, who together with the institutional worldview shaped how the IMF addressed the issue. From 2005 onwards, fossil fuel subsidies gained increasing attention, particularly driven by the fiscal impact of high oil prices (2005–8 and 2011–14) and the ‘Poverty and Social Impact Analysis Group’ within the Fiscal Affairs Department. More specifically, IMF officials, including members of this analysis group, provided technical assistance to countries suffering from the costs of subsidising expensive fossil fuels and wanting to reform these subsidies, and this experience increased the awareness of the subsidies and how to address them (interview with IMF official, 25 February 2015). Consequently, the IMF stressed the fiscal impact of fossil fuel subsidies while stressing the importance of mitigating measures to protect the poor (interview with IMF official, 25 February 2015). Thus, the contextual factor of high oil prices provided policy entrepreneurs within the IMF with a window of opportunity. As mentioned in Section 7.1, from 2008 onwards, the IMF has increasingly singled out fossil fuel subsidies from other subsidies and stressed their environmental impact; a development in framing that was due to policy entrepreneurship within the organisation.

The notion of including environmental externalities in measures of efficient fuel prices had been floating around for some time among a circle of economists working for the IMF, the World Bank, the US government and various environmental think tanks in Washington DC (interview with senior IMF economist, 24 April 2014). The IMF economists within this group also promoted the notion of including undercharging for environmental costs in a broad definition of fossil fuel subsidies, and such undercharging has received considerable attention from IMF management since 2011 under Christine Lagarde (interview with senior IMF economist, 24 April 2014). Lagarde took a more active interest in climate politics than her predecessors, which opened a window of opportunity for the aforementioned economists. With few exceptions, all publications addressing fossil fuel subsidies since the 1990s have been authored by at least one of the key IMF economists working on fossil fuel subsidies: David Coady, Ian Parry and Benedict Clements. Of these, Parry has a background in environmental energy taxation (with a PhD from the University of Chicago), and has argued the case for carbon taxes in academic and IMF publications (Parry, 2015, 2019; Parry et al., 2014, 2018) as well as being the Fund’s leading expert on such taxes and environmental fiscal policy in general. He arrived at the IMF in 2010, when the Fund’s interest in the under-pricing of externalities from fossil fuel use was increasing. David Coady is the Chief of the Expenditure Policy Division of the Fiscal Affairs Department and has mainly worked and published on the distributional effects of public policies (and earned his PhD from the London School of Economics). Similarly, Benedict Clements (at the time of writing Chief of the Fiscal Policy and Surveillance Division of the Fiscal Affairs Department, holding a PhD from the University of Notre Dame) has worked and published on public and fiscal policy. Both Coady and Clements have been working on subsidies at the IMF since the Fund first started addressing the issue (Baig et al., 2007; Clements et al., 1998).

In terms of the second aspect of economisation, how fossil fuel subsidies were addressed, the policy entrepreneurs particularly from the Fiscal Affairs Department were important in framing the environmental impact in neoclassical terms (as discussed in Section 7.1). There is nothing surprising about the Fiscal Affairs Department adopting a neoclassical approach, since it – even more than other IMF Departments – has been consistently informed by neoclassical economics even following changes to the Washington Consensus (Ban, 2015; Park and Vetterlein, 2010b). While the IMF has always framed fossil fuel subsidies as well as other subsidies in terms of their negative fiscal consequences, the framing in terms of macroeconomic and especially environmental consequences increased in salience after 2008 because of the aforementioned entrepreneurship. The framing of subsidies as producing suboptimal societal outcomes has long been present within the IMF (see e.g. Gupta et al., 2000), but the focus on the macroeconomic costs of pollution (framing environmental degradation in terms of economic costs) only gained traction after 2008 (Arze del Granado and Gillingham, 2012; Clements et al., 2013; IMF, 2008a, 2008b). Hence, the entrepreneurs successfully defined fossil fuel subsidies in a way that resonated with the organisation’s institutional worldview including the professional background of IMF officials, and leading to a very ideal-typical instance of economisation. This resonance is evident not just in the way in which the Fund has addressed fossil fuel subsidies, but also in how it has addressed climate change in general. Most notably, the 2019 report was part of a broader trend of IMF output addressing the economic aspects of climate change such as fiscal policies to meet the Paris Agreement (IMF, 2019c), the greening of the financial system (Carney, 2019; Grippa et al., 2019), using whales as carbon sinks (Chami et al., 2019) and a bilateral assessment of the climate policies of Belize and Grenada (IMF, 2019b, 2019d). Thus, the economisation of fossil fuel subsidies was part of a larger wave of economisation of climate related issues. While the economisation of fossil fuel subsidies was a frontrunner for the IMF’s economisation of climate issues, arguably the more overarching economisation and inherent attention to climate issues was an environment conducive to continued attention to fossil fuel subsidies.

Other factors were less important. Relations with member states were not particularly important, as is evident in the IMF reports stating that virtually all countries have fossil fuel subsidies, conclusions that ran directly against the claims of influential member states such as the United Kingdom and Japan. The IMF’s de facto call for carbon taxes also goes against the current policy situation in the majority of member states (including the most influential member state the United States), which have not adopted such taxes. These states broadly endorsed the IMF’s 2019 efforts to address mitigation when they were presented as proposals to the Board (interview with senior IMF official, 19 May 2020). Institutional interaction was also limited, beyond the aforementioned group of Washington-based economists at inter alia the IMF and the World Bank, who discussed the notion of including environmental externalities in estimates of efficient fuel prices, and the use of OECD production subsidy data. The IMF was not one of the institutions requested by the G20 to provide reports fossil fuel subsidies in the way that it provided reports on climate finance (see Chapter 12). Although IMF officials have participated in various meetings of fossil fuel experts inter alia in the context of the G20, this interaction has had no significant influence on the decision to address fossil fuel subsidies and how it addressed them.

7.3 Consequences

7.3.1 International Consequences

The IMF’s notion of including the non-pricing of externalities in the definition of fossil fuel subsidies constituted a radical break with previous definitions of fossil fuel subsidies, especially those by the IEA and the OECD. Yet, it has not led to similarly radical changes in the politics of fossil fuel subsidies at the international level. It did not have an incentive-based influence over other institutions, and although ideational and agenda-setting influences were present, they were also limited. Perhaps because of its radical break with more established definitions or because it ran against the preferences of most states, there were limits to the willingness to use the IMF’s definition in other intergovernmental institutions. Regarding the G20, IMF officials have participated in the peer review of China, which also referred lengthily to the IMF’s reports and their analysis of the non-pricing of the externalities from fossil fuels in China (G20, 2016a). Other peer reviews also brought up the issue of addressing externalities, inter alia through ‘corrective taxation’ (G20, 2016b), but did not rely on an IMF analysis or the notion of ‘getting prices right’ to the same degree as the China peer review (G20, 2016a). Furthermore, the attention to addressing externalities has gradually declined since the US and China peer reviews in 2016. The 2017 peer reviews of Mexico and particularly Germany discussed non-pricing of (especially environmental) externalities as possibly constituting a fossil fuel subsidy and in the German review referring to IMF data, the 2019 reviews of Indonesia and Italy did not address the issue (G20, 2017b, 2017c, 2019b, 2019c).⁵

The differences between the peer reviews underscore that the influence of the IMF output was to a large degree contingent on IMF participation in the peer review (the China peer review) and has declined over time. None of the peer reviews included non-priced externalities in their definition of fossil fuel subsidies. Thus, there was some agenda-setting influence in terms of urging the G20 to address such non-pricing, cognitive influence in terms of providing new knowledge about such non-pricing in the countries reviewed and normative influence in terms of recommending addressing this non-pricing. Without the IMF reports, it is unlikely that the non-pricing would have been addressed to the same extent. Yet, crucially, the non-pricing of externalities was not defined by the G20 as a fossil fuel subsidy, and hence, it was not framed as an issue where the norm of fossil fuel subsidy reform would be salient.

Beyond the G20, the actors tasked with measuring progress towards Sustainable Development Goals (SDGs, led by the United Nations Environment Programme [UNEP] and the International Institute for Sustainable Development [IISD] with significant OECD involvement) discussed the IMF’s definition at great length and compared it to the OECD and IEA definitions, but chose not to include the non-pricing of externalities in its definition, instead opting for an approach closer to the OECD’s (see Chapter 6). IMF staff were not part of the development of the SDG 12.c indicator. The IMF’s approach has proven more popular among environmental NGOs, a community not otherwise known for embracing the IMF. The Health and Environment Alliance (Nandi, 2017) and the Climate Action Network Europe (2017) are among the NGOs that have been inspired by the IMF’s large estimates and the inclusion of non-priced externalities under the heading of subsidies, and used these cognitive ideas to promote climate policies.

7.3.2 Domestic Consequences

The IMF’s output has had more pronounced consequences at the domestic level than at the international level. ⁶ Particularly the Fund’s programmes have had an important incentive-based impact on domestic fossil fuel subsidy reform in a range of countries, including Egypt, Sudan and Ukraine (IMF, 2015c; Michael, 2013; Sherry et al., 2014). The IMF made the provision of loans that the country was not able to obtain from other sources conditional upon far-reaching fiscal reform, including fossil fuel subsidy reform. Identifying ideational and agenda-setting consequences requires a more detailed look at individual countries, which is why the chapter turns to the five country cases.

In the case of the United States, the IMF estimated that fossil fuel subsidies in the United States totalled USD 649 billion in 2015, of which non-priced externalities constitute US 647 billion (Coady et al., 2019). The IMF had no incentive-based influence, as the United States has not been subject to a lending arrangement since 1965 and limited impact on the public and policymaking agendas. Regarding the public agenda (see Table 7.1), the IMF’s 2015 and 2019 reports on the size of global fossil fuel subsidies were referenced a few times in newspaper articles that also explicitly linked them to US subsidies, but without initiating a major policy debate (Mooney, 2015; Schwartz, 2015). The IMFs reports on fossil fuel subsidies and the notion of non-priced externalities as a subsidy did not have a significant influence on the policy agenda at the national level, where carbon pricing has been very controversial since 2009, when the Waxman–Markey proposal for a US emissions trading system floundered in the Senate. Thus, there have not been any attempts (IMF-inspired or otherwise) to link fossil fuel subsidy reform and carbon pricing at the federal level in the United States⁷, although there have been separate attempts to introduce both carbon pricing and fossil fuel subsidy reform at the federal level. As mentioned in Chapter 5, the debate over fossil fuel subsidy reform in the United States has been framed in domestic terms.

Table 7.1 Fossil fuel subsidies and the IMF in the US media: New York Times and Washington Post

	2009	2010	2011	2012	2013	2014	2015	2016	2017	2018	2019	Total
Articles referring to US fossil fuel subsidy reform and the IMF	0)	0	0	0	0	0	3	0	0	0	1	4
All articles referring to fossil fuel subsidy reform (international and domestic)	3	6	20	22	9	8	16	0	0	1	15	100

In terms of ideational consequences, Treasury officials interacted with the IMF officials who drafted the IMF definition on fossil fuel subsidies, something which developed the cognitive understanding of the issue in both organisations (interview with senior IMF economist, 24 April 2014). Yet this collaboration did not induce the Treasury to adopt a price-gap approach including environmental externalities in the way the IMF’s definition of fossil fuel subsidies does (Clements et al., 2013). Perhaps the most important ideational influence of the IMF was its participation in the G20 peer review of the United States (see Section 5.3.2). In this case, the IMF managed to ensure that the non-pricing of externalities was included on the agenda in the context of the United States, but without any longer-term changes to the policy agenda or to policy itself.

Regarding the United Kingdom, the IMF estimated British fossil fuel subsidies at GBP 28 billion, which virtually only consisted of non-priced externalities (Coady et al., 2019). Incentive-based influences were absent since the last lending arrangement ended in 1979. Agenda-setting and ideational influences were relevant, especially in the context of the inquiry of the House of Commons Environmental Audit Committee into British fossil fuel subsidies (see also Chapter 5). The Committee referred to the IMF’s 2013 report and the notion of defining non-priced externalities as a subsidy, but in the end opted for excluding such externalities in the definition they used (House of Commons Environmental Audit Committee, 2013). The IMF’s 2013, 2015 and 2019 reports were also picked up by British media (see Table 7.2), which highlighted its high estimates of global and (in 2013 and 2015) British fossil fuel subsidies (Carrington, 2015b; Watkins, 2013). The contradiction between the promises to fight climate change made in 2015 in the run-up to twenty-first Conference of the Parties to the United Nations Framework Convention on Climate Change (COP21) in Paris and the persistence of British subsidies were highlighted in particular (Carrington, 2015b).

Table 7.2 Fossil fuel subsidies and the IMF in the UK media: The Guardian and The Independent

	2009	2010	2011	2012	2013	2014	2015	2016	2017	2018	2019	Total
Articles referring to UK fossil fuel subsidy reform and the IMF	0	0	0	0	2	1	5	0	0	0	1	9
All articles referring to fossil fuel subsidy reform (international and domestic)	0	0	8	11	10	9	27	11	18	16	46	156

In terms of cognitive influences, the UK Treasury was responsible for both the UK government’s definition of fossil fuel subsidies and for the IMF, and hence Treasury officials interacted regularly with IMF officials. This interaction increased awareness of the issue but did not amount to fundamental ideational influences on the Treasury officials. This was mainly because even before the IMF became closely involved in fossil fuel subsidies, the Treasury perceived fossil fuel subsidies in terms similar to that of the IMF’s overarching view, namely as undesirable first because of their macroeconomic effects and secondly on the grounds of the environmental effects (interview with UK Treasury official, 24 November 2014; for an example of how the Treasury perceived fossil fuel subsidies through an environmental economics perspective, see Stern, 2006). Yet, the Treasury did not concur with the Fund that non-priced externalities constitute fossil fuel subsidies.

Concerning India, the consequences of IMF output have been more limited. The IMF estimated that Indian fossil fuel subsidies amounted to USD 209 billion in 2015, of which non-taxed externalities constituted more than USD 200 billion (Coady et al., 2019). India’s last lending arrangement with the Fund (the largest ever) ended in 1993, ruling out incentive-based influences. The influence on the public agenda is also non-existent: only one article in the two leading Indian dailies referred to the IMF output, and this was only a reference to an analysis by an NGO made using the IMF’s fossil fuel subsidy data (see Table 7.3).

Table 7.3 Fossil fuel subsidies and the IMF in the Indian media: The Hindu and Times of India

	2009	2010	2011	2012	2013	2014	2015	2016	2017	2018	2019	Total
Articles referring to Indian fossil fuel subsidy reform and the IMF	0	0	0	0	0	0	0	0	1	0	0	1
All articles referring to fossil fuel subsidy reform (international and domestic)	0	1	10	35	37	19	17	4	7	4	4	138

India’s successful fossil fuel subsidy reform has been driven by fiscal and macroeconomic concerns: there are cheaper ways of alleviating poverty, and fossil fuel subsidies are detrimental to the public budget and the balance of trade (see also Chapter 5). Two contextual factors made the reform possible: low oil prices and the liberalisation of the Indian economy ongoing since the early 1990s. Low oil prices created the scope in which to liberalise fuel prices without attracting public protests. Although the liberalisation of the Indian economy is arguably the result of ideational influences promoting the belief in free-market economic governance (Mukherji, 2013), more specific ideational influences concerning fossil fuel subsidies have not been significant. Rather they have been limited by scepticism of the framing of fossil fuel subsidies as an environmental issue, since the Indian government frames them as an economic and fiscal issue (Dasgupta, 2013). The scepticism reflects the historically predominant (yet increasingly challenged) view within the Indian elite that climate change is the responsibility of developed countries and that developing countries should not commit to climate change actions (Sengupta, 2019; Thaker and Leiserowitz, 2014).

Indonesia constitutes a different case as regards IMF influence. The IMF estimated Indonesian fossil fuel subsidies at USD 97 billion in 2015, of which non-taxed externalities constituted USD 86 billion (Coady et al., 2019). Incentive-based effects were important in 2002 (and thus before the period in question here), when the IMF programme following the 1997-8 Asian financial crisis led to increases in fixed fuel prices (Government of Indonesia, 2002). After this programme ended in 2003, the absence of direct leverage meant that the IMF played the part of a trusted policy advisor rather than an active stakeholder (former senior IMF official, interview, 14 December 2016). However, the drivers of more recent Indonesian fossil fuel subsidy reforms are primarily domestic (interviews with officials from the Indonesian Ministry of Finance and Bappenas, 14 September 2016 and 20 December 2016). The Indonesian Ministry of Finance has been an important driver of such reforms (and interacted closely with the World Bank) because of concerns about their impact on the budget (interview with senior Indonesian Ministry of Finance official, 26 February 2016).

The IMF’s knowledge output in the shape of reports has also to some degree been picked up by the Indonesian media (see Table 7.4), including IMF data on the magnitude of Indonesian non-priced externalities and suggestions for how to reform fossil fuel subsidies (Nasution, 2013). Furthermore, the IMF has been cognitively influential in providing analyses of Indonesian fossil fuel subsidies in collaboration with the World Bank. The IMF collaborated with the World Bank, following a standard division of labour in which the IMF focused more on the monetary exchange rate and broad fiscal setting, while the World Bank focused on sectoral and microeconomic issues (former senior IMF official, interview, 14 December 2016).

Table 7.4 Fossil fuel subsidies and the IMF in the Indonesian media: Kompass and Tempo

	2009	2010	2011	2012	2013	2014	2015	2016	2017	2018	2019	Total
Articles referring to Indonesian fossil fuel subsidy reform and the IMF	0	0	0	4	2	0	0	0	0	2	0	8
All articles referring to fossil fuel subsidy reform (international and domestic)	0	1	4	61	28	45	18	19	48	60	26	310

Finally, in the case of Denmark, according to the IMF, fossil fuel subsidies in Denmark amount to USD 6.3 billion, only consisting of non-taxed externalities (Coady et al., 2019). Denmark has never been subject to an IMF programme, and incentive-based influences did not play a role. As regards influences on the public agenda, only one newspaper article has referred to the IMF’s estimates (see Table 7.5). Yet, concerning the policymaking agenda and cognitive influences, green politicians have referred to the IMF’s estimate that Danish fossil fuel subsidies amount to USD 1,000 per capita and forced the government to admit to having fossil fuel production subsidies (Danish Ministry of Climate Change, 2015; Poll, 2016). Various Danish governments have consistently argued that the high levels of the Danish carbon tax means that it does not make sense to speak about Danish consumption subsidies, thus disputing the IMF’s definition. When there were discussions within the Danish government of a self-review of Danish fossil fuel subsidies (see also Chapter 6), the OECD definition was preferred to the IMF one, which was considered – by the Danish government as well as by several other countries – too far-reaching (interview with former Danish senior civil servant, 5 May 2020).

Table 7.5 Fossil fuel subsidies and the IMF in the Danish media: Politiken and Jyllands-Posten

	2009	2010	2011	2012	2013	2014	2015	2016	2017	2018	2019	Total
Articles referring to Danish fossil fuel subsidy reform and the IMF	0	0	0	0	0	0	0	1	0	0	0	1
All articles referring to fossil fuel subsidy reform (international and domestic)	0	0	3	3	5	8	9	1	10	4	13	56

7.4 Summary

As regards fossil fuel subsidies, the IMF turned out in many ways to be the unexpected environmentalist. The Fund’s output can be divided into two strands. First, the IMF has for decades induced countries to reform their fossil fuel subsidies through the conditionalities of its programmes, but until recently without treating fossil fuel subsidies differently than any other subsidy. Second, since 2008, the IMF has defined the non-pricing of externalities including climate change as constituting a subsidy. This definition constitutes a game changer in IMF policy on the subject, which was driven by policy entrepreneurs among the IMF officials and was shaped by the economic worldview of the IMF in the context of the IMF bureaucracy’s autonomy from its member states. The IMF’s output has had some ideational impact on the international level, where other institutions such as the G20 took the IMF’s definition of fossil fuel subsidies into account but ended up relying on other definitions. The output has had more important consequences for the domestic level. The first strand had decisive incentive-based consequences for countries such as Indonesia under IMF lending arrangements, which required them to reform their fossil fuel subsidies. The IMF’s extensive definition of fossil fuel subsidies, according to which practically all countries in the world subsidise fossil fuels, also had an impact. It was utilised by actors who wanted fossil fuel subsidy reform in countries such as Denmark and the United Kingdom.

8 The Alignment of the Economic Institutions on Fossil Fuel Subsidies Synergies, but Definitions Can Be Divisive

The three preceding chapters have explored the economisation of fossil fuel subsidies by the G20, the OECD and the IMF, all three of which addressed fossil fuel subsidies as an economic issue. Even when the environmental consequences of fossil fuel subsidies were highlighted, it was done in a manner framing these consequences in economic terms, for example, as wasteful consumption or as non-priced externalities. While the three institutions agreed on the importance of a reform of fossil fuel subsidies due to their environmental and economic consequences, they also differed in how they addressed these subsidies. Most notably, the IMF adopted a radical definition of fossil fuel subsidies based on the Pigouvian notion of corrective taxes, which stood out against the more established definition of the OECD. Thus, the economisation was more ideal-typical in the case of the IMF. The G20 skirted the issue of fossil fuel subsidy definition but worked closely with the OECD. The chapter proceeds with a comparison of the three institutions and demonstrates that economisation may lead to diverging framings of the issue in economic terms. Altogether, they constitute core institutions within the fossil fuel subsidy complex, a complex characterised by synergistic relations (Verkuijl and van Asselt, 2020). Subsequently, the chapter outlines how this divergence was driven by the differences in worldview, policy entrepreneurship and the degree of autonomy of the International Organisation (IO) bureaucracy from principals. Yet, the similarities between their worldviews (they agreed on a range of fundamental issues), institutional interaction and overlapping memberships pulled in the direction of convergence between the institutions. Finally, there is a discussion on the consequences of this divergence at the international and domestic levels (conflicting estimates of fossil fuel subsidies and discussions of which definition to use), while the convergence between the institutions was important for the attention to the issue and the norm of fossil fuel subsidy reform.

8.1 How They Align

8.1.1 Types of Output

The institutions differed considerably in the form of their output. The G20’s output was mainly regulatory with a crucial normative component, most notably in 2009 when it propelled both the concept of fossil fuel subsidies and the norm of fossil fuel subsidy reform high up international and domestic agendas. Subsequent G20 output focused on gathering knowledge about fossil fuel subsidies (especially by requesting such knowledge from four IOs) and directly holding G20 member states accountable to the norm through self-reporting and voluntary peer reviews. The output of the OECD consisted of knowledge but focused more on the cognitive aspects of fossil fuel subsidies by producing extensive knowledge about fossil fuel subsidies, including within individual countries. This knowledge production had a normative aspect in that its purpose was to identify subsidies that should be reformed or phased out. The OECD also disseminated knowledge about fossil fuel subsidies and how to reform them through formal and informal channels such as workshops and direct involvement with countries, including G20 peer reviews, which they chaired. The IMF’s output can be divided into two strands. The first strand provides incentives for fossil fuel subsidy reform through the conditionalities of the programmes. The second strand is knowledge based and centres on the IMF’s definition of fossil fuel subsidies as including non-priced externalities. This definition has a normative component in ‘getting the price right’ through pricing externalities and expanding the applicability of the norm of fossil fuel subsidy reform, as well as a cognitive component in framing fossil fuel subsidies in terms of macroeconomic consequences and producing knowledge about the level of fossil fuel subsidies in countries around the world.

The relationship between the G20 and the OECD can be described as synergistic, with the two institutions focusing on different tasks, and the OECD supporting the G20’s efforts to promote the norm of fossil fuel subsidy reform through cognitive output. This division of labour is due to the G20 requesting that the OECD provides such support, and, in more general terms, the OECD fulfilling secretariat functions for the G20 in several areas. The relationship between on the one hand the two institutions and on the other the IMF has rather been one of co-existence, with a sometimes conflictive relationship between the IMF and the OECD as regards the issue of subsidy definitions (for the argument that non-priced externalities should not be treated as a subsidy, see Steenblik, 2014).

8.1.2 Scope and Actors Addressed

For all three institutions, it is possible to distinguish between output that is global in reach and output that is more targeted. The G20 commitment and the resulting self-reports and peer reviews focused on the G20 members, yet much of the cognitive output produced by the IEA, OECD, OPEC and the World Bank at the request of the G20 was global in reach. Likewise, most of the OECD’s output concerned its members plus seven of the largest emerging economies, yet other output, for instance its publications on how to define fossil fuel subsidies, had a global reach. The IMF’s first, incentive-based strand targeted countries with financial problems, whereas its second, ideational output on the non-pricing of externalities targeted virtually all countries. Hence, there is a significant overlap in the countries that they address. In the case of synergistic relationships, such as the one between the OECD and the G20, this should not imply any risk of conflicting output. Conflicting output are more likely when institutions address a set of countries with regard to a particular issue but use differing approaches, such as the divergence between the OECD and the IMF regarding the definition of fossil fuel subsidies. Yet, this did not result in conflict at the domestic or international levels, as discussed in detail in Section 8.3. In terms of the government institutions they interacted with, there is also a considerable overlap, as the circle of fossil fuel subsidy experts in each country is limited and these experts are often based in finance ministries, which interact with all three institutions.

8.1.3 Cognitive Dimensions

All three institutions frame fossil fuel subsidies in economic and environmental terms. The environmental framing is perhaps most interesting, as fossil fuel subsidies were previously grouped together with other subsidies and not singled out because of their environmental impact. Distinguishing fossil fuel subsidies from other subsidies implies an emphasis on their environmental consequences.

The G20 left open both the definitional question and the closely related questions of which fossil fuel subsidies exist in individual countries and their size. Importantly, the G20 has primarily framed fossil fuel subsidies in terms of climate change as well as macroeconomic consequences, while the importance of reducing poverty was also stressed. The fiscal impact was not mentioned. The OECD has mainly framed the subsidies in environmental and fiscal terms. While the macroeconomic consequences of fossil fuel subsidies have also received some attention (but not been included in the estimates of their consequences), the distributional impact has hardly been addressed at all. The IMF has adopted a Pigouvian framing, which includes the non-pricing of externalities as a subsidy and highlights the environmental consequences while framing them in terms of their macroeconomic impact. The fiscal and distributional consequences have mainly been emphasised by the Fund outside its knowledge output regarding getting prices right through the pricing of externalities, the fiscal consequences being important to the IMF’s first strand. These differences in framing are not conflictive in themselves, since the institutions merely highlight different aspects of fossil fuel subsidies while agreeing on the importance of their environmental and economic consequences. Yet, the difference between the IMF’ economisation of fossil fuel subsidies based on textbook economics and the OECD’s economisation based on fiscal concerns underscores how economisation was more ideal-typical in the case of the IMF (see also Chapter 1).

In political terms, the differences in definitions between the OECD and the IMF are more important, since they lead to different estimates of the size of fossil fuel subsidies. The IMF estimated that fossil fuel subsidies were ten, in the case of some countries such as the United States even a hundred, times higher than the OECD did (IMF, 2018a; OECD, 2020a). In this way, rather conflicting bodies of knowledge were produced by the two institutions. While the two subsidy definitions de facto measure different phenomena, they are both generally referred to as measuring fossil fuel subsidies. Yet, it is important to bear in mind that both the IMF and the OECD mainly measured the size of subsidies in monetary (although the former focused on macroeconomic costs and the latter on fiscal ones) and quantitative terms, rather than, for example, the number of policies or the climate impact measured in tons of CO₂e. This monetary and quantitative approach to measuring and addressing a problem constitutes one fundamental cognitive aspect of economisation. Hence, I argue that the cognitive differences between the institutions is best understood as different kinds of economisation, rather than as economic and non-economic ways of addressing fossil fuel subsidies.

8.1.4 Normative Dimensions

All three institutions have promoted the norm of fossil fuel subsidy reform. It is notable that three economic institutions all defined the climate consequences of fossil fuel subsidies as an important reason for reforming them. They also shared a concern over inefficiency and opposition to subsidies as distortionary and leading to economically suboptimal outcomes (e.g. ‘wasteful consumption’ as the G20 2009 put it). Defining economic efficiency as a key objective is a crucial dimension of the economisation of fossil fuel subsidies of the three institutions, although most pronounced within the IMF. Divergence appeared rather in terms of the IMF’s linking of the norm with carbon pricing, a link that was rooted in Pigouvian economics.

This divergence had implications for the application of the norm of fossil fuel subsidy reform, which the IMF found was salient to several situations (whenever a fuel was sold at a price that did not include its social costs) in which the OECD did not find that it applied. There are also situations in which the OECD identified fossil fuel subsidies that the IMF did not, for example, consumption subsidies that do affect prices. This divergence is important, because the two institutions identify different issues as problematic in terms of fossil fuel subsidies. For the IMF, it is the non-pricing of externalities, for the OECD policies subsidising the consumption and production of fossil fuels. Both agree that consumption subsidies lowering the price of fossils below the benchmark price including value-added tax (VAT) constitutes a subsidy, as do production subsidies. The IMF also included production subsidies in its estimate and used OECD estimates of these subsidies, which nonetheless amounted to a minuscule part of the total global fossil fuel subsidies in the IMF estimate (about USD 17 billion out of 5,300 billion). Likewise, the OECD is generally also in favour of environmental taxes, including carbon taxes (Bernstein, 2001; Ruffing, 2010).

Consequently, the two institutions recommend different actions to implement the norm of fossil fuel subsidy reform. While the IMF recommends adopting taxes corresponding to the externalities from fuel use, the OECD recommends phasing out specific policies. The IMF’s approach is clearly rooted in the notion of Pigouvian taxation, whereas the OECD’s approach is less specific in terms of ideational underpinnings but fits with the paradigm of liberal environmentalism. Liberal environmentalism differs from neoclassical environmental economics in that it is a more encompassing norm complex rather than an academic paradigm with a more specific problem definition (pollution as an externality) and policy solutions (pricing). Nonetheless, there are considerable overlaps, and it is worth noting that one of the fathers of liberal environmentalism and of the OECD’s approach to sustainable development, David Pearce, gave a presentation at one of the earliest OECD workshops on environmentally harmful subsidies. Although Pearce advocated Pigouvian taxation to address externalities (Pearce, 2002b), he did not treat the absence of such taxation as a subsidy (a notion that was developed only later; see Clements et al., 2013; Hodas, 2006), but as a way of addressing environmental degradation, a problem defined as distinct from the existence of subsidies (Pearce, 2002a).

8.1.5 Incentives

In terms of incentives, the relationship between the three institutions is more synergistic. The IMF is the institution that has provided the strongest incentives for fossil fuel subsidy reform, but these incentives suit the output of the other two institutions. Most importantly, the IMF making fossil fuel subsidy reform a condition for loans under the adjustment programmes has led to fossil fuel subsidy reforms in the G20 members Indonesia and (after the Pittsburgh commitment) Argentina¹ as well as non-G20 countries including Egypt and Ukraine. Thus, IMF incentives have supported the norm of fossil fuel subsidy reform both within and beyond the G20 by inducing countries to reform these subsidies.

8.2 Causes of Alignment

Factors pulling in the direction of convergence and divergence shape the overall pattern of alignment in terms of synergy on fundamental issues and some degree of diverging (although rarely conflicting) ideas concerning definitions, especially between the IMF and the OECD. Regarding convergence, the fundamental elements of the institutional worldviews shaped the basic shared approach to fossil fuel subsidies as an issue that was problematic for environmental and economic reasons. The environmental reasons were mainly conceived of in economic terms as problematic due to their economic costs. Institutional interaction is another important factor for convergence, particularly between the G20 and the OECD, which interacted with other as well as several of the same institutions, particularly the IEA and the World Bank. On a fundamental level, interaction also mattered in terms of the Pittsburgh commitment lifting the topic to a new level internationally, including within the OECD and more indirectly the IMF (see Chapters 6 and 7). Overlap in terms of membership also pulled in the direction of convergence, since several states are members of all three institutions, most notably the largest developed economies. Yet, also institutions that do not overlap to the same extent, most notably the Friends of Fossil Fuel Subsidy Reform, have produced similar output, indicating that convergence can also happen without overlapping membership.

On the other hand, the divergence between the institutions, particularly the IMF and the OECD, has been driven by factors, which influenced the institutions in different ways. The institutional interaction that influenced the OECD did not influence the IMF to the same extent, since there was no request to address fossil fuel subsidies and thus the IMF did not interact in the same way with the G20, the IEA, the World Bank and OPEC. Rather the IMF addressed the issue on the initiative of IMF staff, which more than OECD staff induced their organisation to address fossil fuel subsidies. IMF staff acting as policy entrepreneurs are also the main reason why the IMF to a larger degree than the OECD framed fossil fuel subsidies in terms derived from environmental economics: IMF staff linked the macroeconomic and environmental framing of such subsidies on the basis of the notion of getting the price right.

The differences in how the organisations addressed fossil fuel subsidies was also influenced by more specific differences between their worldviews: the IMF defined the environmental impact in the aforementioned neoclassical way, and the OECD defined fossil fuel subsidies in a fashion reflecting how it had addressed other subsidies. The G20 was much less specific in how it defined fossil fuel subsidies and the applied the norm of fossil fuel subsidy reform, but to a large extent relied on its member states and other institutions, notably the OECD.

The degree of autonomy of the IO bureaucracies proved to be an important scope condition for the influence of the worldview and policy entrepreneurs, and thus also shaping the divergence between the institutions. This is evident in that the IMF, to a greater degree than the OECD, adopted positions running against the preferences of its member states, most notably the claim that developed countries have significant fossil fuel subsidies. The OECD had less autonomy and only received the mandate to scrutinise its members’ subsidies after the G20 commitment. The differences between the two organisations demonstrate that differences in membership and voting rules did not have an impact, as the IMF contradicted member states that are more influential within the IMF than within the OECD, especially the United States. Arguably, if the IMF had the same degree of autonomy as the OECD Secretariat, its position would have been more similar to that of the OECD.

8.3 The Consequences of Alignment

The alignment of the three institutions had important consequences both at the international and the domestic levels.

8.3.1 The International Level

The international level is the level where the combined output of the three institutions had the greatest impact. Especially the G20 managed to ensure the inclusion of the issue of fossil fuel subsidies and the norm of fossil fuel subsidy reform on the agenda of other international institutions, which led to new commitments to fossil fuel subsidy reform within Asia-Pacific Economic Cooperation (APEC), the North American Leaders’ Forum and indirectly the Sustainable Development Goals (SDGs), and even to the creation of an institution dedicated to fossil fuel subsidy reform, Friends. In addition, the G20 commitment led to other international institutions producing new knowledge about these subsidies, including the OECD. Together with the IMF’s cognitive output on fossil fuel subsidies, they significantly increased the knowledge of such subsidies based on the framing of them as economically and environmentally costly. While it is not possible to provide a full counterfactual analysis, I would argue that without the G20 commitment, the international efforts to address fossil fuel subsidies would have looked different, with fewer or no commitments to reform. Likewise, without the OECD output, there would have been less knowledge about fossil fuel subsidies especially in developed countries. Moreover, without the IMF’s output, carbon pricing and the non-pricing of externalities would not have been linked to fossil fuel subsidies.

The synergy between the G20 and the OECD reinforced their respective outputs. The OECD in particular benefitted from the G20 commitment, which lifted its own involvement to a new level (see Chapter 6) and opened the doors for OECD subsidy experts to be part of G20 peer reviews, SDG working groups on the reporting of efforts to fossil fuel subsidies, and so forth. The differences between the OECD and IMF definitions of fossil fuel subsidies also played out at the international level, particularly in the context of the G20 peer reviews and the SDG fossil fuel subsidy working group. In both cases, there was a discussion of which definition to use, and in both cases, approaches closely aligned with the OECD definition won the day. The IMF’s approach was discussed and to some degree also applied in the SDG report on how measure fossil fuel subsidies (UNEP et al., 2019) and the first G20 peer reviews (of China and the United States), but not in subsequent peer reviews. Thus, the potential for conflict between the two different definitions did not undermine the efforts to address fossil fuel subsidies, but merely led to expert discussions of their relative merits and in the case of the Chinese and US peer reviews, to discussions of the pricing of externalities in the two countries.

8.3.2 The Domestic Level

The three institutions had less of a discernible effect at the domestic level than at the international level (see Chapters 5, 6 and 7). In the five countries studied, those that reformed their fossil fuel subsidies did so mainly because of domestic pressure. Nonetheless, the fact that the institutions all promoted the norm of fossil fuel subsidy reform meant that the prevailing combined effect was one of synergy with the institutions reinforcing each other. The OECD and the IMF were important in picking up the baton from the G20 and ensuring that the issue remained on the (political to a larger degree than the public) agenda after the initial attention caused by the G20 commitment had died down. The IMF and the OECD also had the ability to produce new knowledge about the issue in a way that the G20 did not. The synergistic relationship between the G20 and the OECD played an important role primarily regarding G20 countries – especially those undergoing peer reviews – in terms of the OECD using its expertise on fossil fuel subsidies in general and about the countries subject to peer review in particular.

Regarding the definitional divergence between the IMF and the OECD, actors in developed countries studied picked up the IMF estimates of fossil fuel subsidies within their country in particular. Most notably, the UK House of Commons Environmental Audit Committee discussed the different approaches and the different estimates of UK subsidies provided by the two institutions (House of Commons, 2013). This Committee based its conclusions on the OECD definition. In the context of this debate and in the media in all the three developed countries, the IMF’s estimate sometimes received attention because of the higher figures. Yet, all things considered, the OECD definition has been more important (also in the context of the G20 peer reviews), and the relationship between the two definitions has been one of co-existence rather than conflict.

8.4 Summary

The three institutions have worked to promote the norm of fossil fuel subsidy reform. More specifically, they framed fossil fuel subsidies as undesirable because of their environmental and economic consequences, though the environmental consequences were framed in economic terms. Although the OECD and the IMF diverged regarding the definition of fossil fuel subsidies, the relationship was generally one of co-existence rather than conflict. Particularly the G20 and the OECD had a synergetic relationship, in which the G20 requested the OECD to provide an analysis of fossil fuel subsidies, which lifted the OECD involvement with the topic to a new level. The OECD also influenced the content of the G20 output. The divergence between the OECD and the IMF centred on the IMF’s inclusion of the non-pricing of externalities in its definition of fossil fuel subsidies, a framing rooted in the neoclassical notion of getting prices right. and constituting an even more ideal-typical case of economisation than the OECD’s framing. This divergence had ramifications for both the international and domestic levels in terms of conflicting estimates of fossil fuel subsidies and discussions of which definition to use, for example, in the context of the SDGs. Yet, the synergy between the institutions regarding the overarching framing of fossil fuel subsidies and the more specific synergy between the G20 and the OECD helped draw attention to the issue and promote the norm especially at the international but also the domestic levels. The convergence between the institutions was based on the shared elements of their worldviews as economic institutions, institutional interaction and to some degree also their overlapping membership. The divergence between the IMF and the OECD was driven by the differences in worldview, policy entrepreneurship among IMF officials and the IMF’s greater autonomy from principals, which allowed for the intra-institutional factors to play a greater role in the case of the IMF. In this way, IO autonomy acted as a scope condition for the other factors.