There is no universally accepted definition of a subsidy. Organizations such as the Organisation for Economic Co-operation and Development (OECD) or the World Trade Organization (WTO) use definitions that align with their specific policy objectives. For instance, the WTO’s definition is narrow and contains three basic elements: “(i) a finanacial contribution (ii) by a government or any public body within the territory of a Member (iii) which confers a benefit.”
All three of these elements must be satisfied in order for a subsidy to exist in the legal parlance of the WTO. The narrow definition used by the WTO may reflect the need for verifiable and easily quantifiable evidence of support that can withstand legal scrutiny.
Economics, though, admits to a much wider range of definitions that could include nonfinancial policy support (for example, granting free access to a resource or favoring one firm or sector over another) as well as financial support (Sakai, Yagi, and Sumaila 2019). For instance, an uncorrected externality like air or water pollution would, in theory, be treated as a subsidy if the costs fall on some other party. When externalities are included in the definition, it is important to measure the extent of the subsidy correctly and to be precise about the definition being used.
Indeed, different types of expenditures and policies (and lack thereof ) may, at different times, be considered subsidies, public good provision, environmental policies, or social safety nets. Rather than focusing on any single definition of what a subsidy is, this report takes a practical view, acknowledging that different definitions are appropriate in different contexts and at different times.
1. A narrow definition of a subsidy would include only direct fiscal outlays from the government to producers or consumers that are intended to affect the production or consumption of goods and services. This definition corresponds to the WTO and other more traditional definitions. It could be broadened to include policies that do not include direct fiscal outlays from government, but result in transfers from producers to consumers or vice versa. These policies would include trade barriers, price ceilings, and price floors.
2. Expenditures on the provision of public goods can also be considered a subsidy if they are intended to benefit producers in a particular industry. For instance, research and development (R&D) of agricultural technologies, construction and maintenance of infrastructure like irrigation systems and ports, and even expenditures on weather observation and early warning systems may be considered subsidies if their uses are provided at below-market price and they benefit private producers.3. An even broader definition would include external costs that a person or firm generates and some other entity pays for—that is, an externality. This definition would include costs that are monetary (such as expenditures made to mitigate damages) as well as nonmonetary (such as health damages from air or water pollution).
In principle, there are good reasons for all three definitions, and the appropriate measure would depend on the intended use. Nevertheless, data and estimates of the three definitions vary widely. Data on direct, explicit subsidies like those in the first definition are generally more available, as they tend to be detailed in government budgets, and obtaining reliable estimates is usually simply a matter of data collection. Nevertheless, challenges remain in aggregating subsidies from different ministries, levels of government (federal versus state and local), and across sectors that use different definitions. Data on the provision of public goods can also be relatively easier to obtain, however, complicating decisions on what should be considered a subsidy and what should be considered a welfare or a safety net program. Obtaining estimates on external damages, or implicit subsidies, can be the most challenging. Accordingly, much of the new analyses presented in this report focuses on these damage estimates.
Subsidies are important tools that governments can use to encourage certain behavior, support economically or politically important industries, or achieve particular goals
Definitions of subsidies vary considerably. In general, there is more understanding of how explicit subsidies ought to be measured and defined—that is, as the financial value of support provided by the government to a sector. The support could be monetary (for example, a cash transfer or tax exemption) or in-kind (for example, free fertilizers or fuel). In principle at least, the explicit subsidy measures the financial value of such transfers. Measuring implicit subsidies raises a more complex set of issues. An implicit subsidy measures the externality resulting from an explicit subsidy or policy exemption. Since externalities are included in the definition of an implicit subsidy, it is important to distinguish the portion of uncorrected externalities caused by the subsidy from the total external cost of the activity. As an example, the use of a pesticide may have environmental impacts, even without subsidies. If a subsidy induces greater use of the pesticide, the additional impact may also count as an implicit subsidy.
1. The (narrow monetary) definition of an explicit subsidy would simply be SQs, which corresponds to the World Trade Organization and other more traditional definitions that are concerned mainly with fiscally tied definitions. In principle, such a definition is incomplete if a subsidy is concerned with all forms of policy support (tacit and explicit).
2. A broader definition of a subsidy would recognize that greater damage has occurred than would have occurred without the subsidy. In this case, the subsidy is measured as SQs + D(Qs − Qn), where SQs is the explicit subsidy and D(Qs − Qn) measures the implicit subsidy.
3. A more complicated definition would take deviations from the optimal level of pollution (EDQ*) into account and thus be SQs + D(Qs – Q*). This definition would be the most accurate definition of a subsidy, but it requires estimating the optimal level of damage (DQ*), which can be complicated. For this reason, the second definition is used more widely in practice.
Finally, ignoring implicit subsidies (that is, externalities) for computational convenience brings problems of consistency: Would all in-kind (nonpecuniary) contributions be excluded? If not, what is the difference between one in-kind contribution (such as free pesticides) and another (the health externality from the pesticide)? One cannot rely on the fact that one kind of contribution affects profits and the other affects health (since a profit function is a subset of a societal welfare function). At the other extreme, it would also be inappropriate to measure the subsidy as SQs + DQs, since this approach wrongly assumes that without the subsidy there would be no external costs.
a. The pollution and damage functions are all linear for simplicity, and abatement is not considered. Introduction related to economic efficiency or equity. But they can also be distortive by reducing economic efficiency and exacerbating negative externalities. The litany of problems created by subsidies is widely recognized. Subsidies can reduce total factor productivity by shifting resources to less productive sectors. And, when imprudently applied to natural resource sectors, they can have harmful impacts on the environment. For instance, natural resource subsidies can lead to overcapitalization, which results in more land being devoted to agricultural use or more fishing boats attempting to harvest a shrinking supply of fish (Milazzo 1998). They can also send the wrong economic signals, indicating, for example, that scarce natural resources—like water in a desert—are abundant when, in fact, they are not. The consequence is overuse and inefficient use, which can result in a resource deficit that acts as a drag on economic progress and growth.
Perverse subsidies, especially for the use of natural resources, are also likely one of the most significant sources of inefficient spending (Arguedas and van Soest 2009; Yih et al. 2018). In 2020, total global debt surged to 263 percent of GDP, its highest level in half a century (World Bank 2022). In emerging markets and developing economies, government debt alone increased by 9 percent of GDP in 2020 as countries responded to the COVID-19 crisis by stimulating the economy while dealing with reduced revenues.
In normal circumstances, such levels of debt might be sustainable or even desirable if they were servicing prudent investments. But circumstances are not normal, as the world is dealing with multiple crises, including COVID-19, disrupted supply chains, rising inflation in many countries, and food and energy shocks stemming from the conflict in Ukraine. Credit markets are tightening, tax revenues are declining, and government spending is on the rise. With fiscal space shrinking quickly in many countries, economic stability will depend, at least in part, on better and more effective public spending. Assessing the magnitude and impact of subsidies on renewable natural resources is a key part of bringing greater efficiency and equity in public spending and addressing the unsustainable use of natural resources. Given the regressive nature of many subsidy schemes (Schuhbauer et al. 2020), subsidy reforms can also help to address concerns of rising inequality and poverty as well as enhance environmental sustainability. The magnitude of subsidies in natural resource sectors.
Although subsidies are such a large and important part of government budgets, reliable quantification has proven elusive due to the complexity, interconnections, and scale of support. This complexity arises partially because subsidies come in different forms, are provided by different levels of government (subnational as well as central governments), and can go by several different definitions—price ceilings or floors, direct support to producers or households, support through the subsidizing of inputs, tax expenditures, or unpriced or unaddressed negative externalities. This section presents new analyses combined with reviews of the literature to describe the magnitude of subsidies in the selected sectors.
Fossil fuel subsidies
Economists advocate price-based policy instruments as a central tool for addressing the adverse societal costs associated with fossil fuels, such as air pollution. In principle, this approach calls on governments to reflect the environmental and health costs of polluting activities in their prices—in particular, by taxing the fuels and activities that drive air pollution. However, rather than taxing polluting activities, many governments around the world provide explicit subsidies to lower the cost of using fossil fuels, thus entrenching polluting technologies and practices.
In an effort to promote industrialization and energy affordability—but also to cater to influential political interest groups—governments around the world are actively lowering the cost of polluting forms of energy through “explicit” subsidization schemes. These schemes have grown into expensive support programs for the consumers and producers of oil, gas, and coal products. Globally, explicit fossil fuel subsidies are estimated to have been around US$577 billion in 2021 (Parry, Black, and Vernon 2021). Thus they are almost three times more than global subsidies paid to the renewable energy sector (IRENA 2020); they are also almost six times more than the amount that countries have committed to raise in annual climate financing (US$100 billion) under the Paris Agreement on Climate Change. Online appendix A provides country-level figures for fossil fuel subsidies.2While US$577 billion is a vast amount to spend on propping up polluting fuels in one year, it is likely to be an underestimate. In particular, subsidies paid to polluting industries and the producers of fossil fuels are often far more difficult to define, observe, and quantify. Such producer subsidies can refer to various kinds of preferential treatment of fossil fuel exploration, extraction, or processing firms or other energy-intensive companies, industries, or products (chapter 3). Such producer subsidies could be explicit, such as grants, low-interest loans, or direct payments; they may be in-kind, such as credit subsidies, government guarantees to protect investment, or subsidies through public procurement.
The in-kind component of these subsidies is especially difficult to identify and measure, explaining the scarcity of studies. A study of G-20 countries estimates that producer subsidies amounted to US$444 billion in 2014. The largest share of these producer subsidies came in the form of fossil fuel investments by state-owned enterprises, amounting to US$286 billion (Bast et al. 2015).
Even when fossil fuels are not subsidized explicitly, their prices do not fully reflect the vast societal and environmental damages they cause. The polluting activities that drive these externalities are reinforced and incentivized by the underpricing of fossil fuels. The International Monetary Fund (IMF) calls this failure to price externalities “implicit subsidies” to fossil fuels (Parry, Black, and Vernon 2021). The IMF estimates the cost of these implicit fossil fuel subsidies at US$5.4 trillion in 2020, with local air pollution and global climate change impacts constituting more than 75 percent of the total. At US$2.5 trillion a year, local air pollution was the single largest unpriced environmental externality from fossil fuels in 2020—far more than the size of explicit subsidies. An important implication is that removing explicit subsidies alone is unlikely to bring fuel prices to their socially optimal level.
Agricultural subsidies
As with fossil fuel subsidies, agricultural support can be considered in terms of explicit and implicit subsidies. However, unlike air pollution, much of the effort to quantify global subsidies is restricted to identifying explicit support. Even when it comes to explicit support, however, quantifying global magnitudes can be extremely difficult, largely because countries and organizations measure support in different ways. In addition, several definitions exist for agricultural support, which are discussed in more depth in chapter 6. The most comprehensive measure of agricultural support is the total support estimate (TSE). This estimate includes support for both outputs (final crops produced) and inputs (seeds and fertilizer, for example), the sum of which is Introduction 7 called the producer support estimate (PSE), as well as any taxpayer-to-consumer transfers (TCTs) and general services support estimates (GSSEs). GSSEs mainly track the provision of public goods like R&D, infrastructure financing and maintenance, and educational programs.
Globally, between 2016 and 2018, annual TSE for 84 countries with available data amounted to US$635 billion per year (see online appendix A for country-level values). This amount equals approximately 0.9 percent of GDP and nearly one-fifth of agricultural value added for these countries. The share of this TSE that was transferred to individual producers—that is, PSE—was about 71 percent, with the remaining share split between GSSE (18 percent) and TCT (11 percent). Thus the bulk of support goes to producers. Around 61 percent of this support is in the form of coupled support, such as market price support or payments for input use, which distort producers’ decisions. As discussed in chapters 8 and 9, this type of support is responsible for much of the harmful environmental impacts. Recent trends, however, suggest that some countries have increased their funding for decoupled support, like direct payments and agricultural investments. This information comes from a database assembled by Gautam et al. (2022), which is discussed more in chapter 6. The 84 countries included accounted for 67 percent of the global value of agricultural production in 2016 (FAO 2022). Thus total explicit subsidies are likely to be significantly higher, perhaps approaching US$1 trillion, based on a simple extrapolation.
Estimating the magnitude of implicit agricultural subsidies is much more difficult. Total greenhouse gases from agriculture are estimated to be approximately 13.7 gigatons of CO2-eq or approximately 26 percent of total annual greenhouse gas emissions (Poore and Nemecek 2018). Given that nearly all of these emissions are untaxed and unregulated, these emissions can be considered an implicit agricultural subsidy. At a shadow price of between US$40 and US$80 per ton of CO2-eq, this subsidy is the equivalent of US$548 billion to US$1.1 trillion worth of external damages that are not internalized by producers or consumers of agricultural products. Other studies have estimated this value to be much higher, at US$1.5 trillion per year (Pharo et al. 2019).
In addition to greenhouse gas emissions, agriculture is responsible for damages to other types of natural capital, like forests and other natural habitats as well as freshwater stocks, which can be considered an implicit subsidy. In many regions, nitrogen fertilizer is applied in such large quantities that much of it is not absorbed by crops and ends up in water runoff, collecting in water supplies. In countries like India, where nitrogen fertilizer is heavily subsidized, a mere 32 percent of nitrogen is absorbed by plants. Even in regions like Europe and North America, where subsidy rates are much lower, only about 52 percent and 68 percent of nitrogen, respectively, is absorbed by plants (Zhang et al. 2015). This excess nitrogen in waterways has enormous environmental and health impacts, which are discussed in chapter 8. While it is difficult to put a single dollar figure on their impact, environmental, health, and productivity damages can be very significant, as chapter 8 discusses.
Fishery subsidies
Rich countries are subsidizing the destruction of fisheries in all corners of the world’s oceans. Explicit subsidies in the fishery sector total an estimated US$35.4 billion per year, of which US$22.2 billion are considered to enhance capacity and contribute to overfishing. This spending includes subsidies for fuel, fishing access agreements, boat construction and renewal, fishery development projects, fishing port development, tax exemptions, and marketing and storage infrastructure. Fishery subsidies are not distributed evenly around the world. Five entities—China, the European Union, Japan, the Republic of Korea, and the United States—contribute 58 percent of the total estimated subsidy. Indeed, the vast majority of subsidies in the fishery sector are spent by high-income or upper-middle-income countries, often to support fishing fleets that traverse the global oceans.
The largest implicit subsidies for fisheries are surely the lack of regulations, which enable open access and lead to overfishing. Overfishing results in a loss of economic benefits estimated at US$83 billion per year, representing an implicit subsidy that is nearly 20 percent of the size of the total sector.
