fossil fuel subsidy reform

Fossil Fuel Subsidy Reform: What Role for the World Trade Organization?

The removal of fossil fuel subsidies (FFS) would bring about many important and positive effects, among them helping to reduce air pollution and emissions of greenhouse gases that cause climate change and improving government’s finances. It could also reduce distortions affecting trade in not only the subsidized products, such as coal, fuel oil and natural gas, but also in goods that compete with fossil fuels, such as wind turbines and solar photovoltaic panels.

Natural gas fracking well in Louisiana

While the country, and indeed the world, struggles to contain the risks of catastrophic climate change it's easy to forget that the US has subsidized fossil fuel production for more than a century, and that we continue to do so. To assess how these subsidies affect domestic industry, Earth Track teamed up with the Stockholm Environment Institute to identify key subsidies to US oil and gas production and model how profits and production would change were these subsidies to be removed. My gratitude goes to both of my co-authors, Ploy Achakulwisut and Pete Erickson at SEI, for their tireless and stellar work modeling the subsidy impacts on thousands of fields expected to enter into production by 2030.

Our final analysis was recently published in the IOP Publishing journal Environmental Research Letters.  You can access the main article here and the detailed supplemental data here.  For subsidy wonks, the supplemental data provides background on the subsidies and how we modeled them that may prove of interest. We greatly hope that our analytical approach will be replicated in other large oil and gas producing countries using their country-specific data on producer subsidies. The carbon reduction benefits from subsidy reform tend to rise if declines in US production don't get displaced by subsidized fuel flowing in from other countries that have not implemented reforms.

This paper updates and expands on work we published in Nature Energy in 2017.  As was done in that analysis, we use field-specific economic cost data developed by Rystad Energy to evaluate how subsidy removal would affect cash flows and investment returns to the field's owners under different price and financial risk outlooks.  We also continued to evaluate discovered but not-yet-producing fields in order to incorporate the full capital cycle in our calculations.  This paper also includes three significant refinements.  First, we incorporated a wider set of subsidies, including industry-specific regulatory exemptions for the first time and a much more detailed look at potentially mammoth unfunded liabilities for well closure and reclamation.  Second, we integrated into our modeling the large changes to the tax code that took effect in early 2018; these changes, primarily through a large reduction in the baseline corporate tax rate, reduced the value of some tax breaks to the industry.  And third, we incorporated reductions in the cost structure achieved by many fracking operators in recent years, changes that reduced economic breakevens on production from these fields. 

Key Findings

At 2019 average market prices of oil and gas, the 16 subsidies could increase the average rates of return of yet-to-be-developed oil and gas fields by 55% and 68% over unsubsidized levels, respectively, with over 96% of subsidy value flowing to excess profits under a 10% hurdle rate (see Figure 1 below). At lower 2020 prices, the subsidies could increase the average rates of return of new oil and gas fields by 63% and 78% over unsubsidized levels, respectively, with more than 60% of oil and gas resources being dependent on subsidies to be profitable under a 20% hurdle rate.  This pattern is common:  without price triggers that automatically reduce or zero our subsidies during strong markets, most of the public support flows to higher profits during periods of high market prices.  In contrast, during periods of low prices the subsidies encourage investments into new production that would not otherwise have happened.  Because new investments in oil and gas will contribute to elevated carbon loadings for many years into the future, the large role government subsidies play in protecting investors from downside market risks is one that we need to eliminate if the country is serious about transitioning away from high-carbon fuels.  We looked at oil and gas extraction sites in this paper.  However, it is important to note that a similar dynamic is at play for very large infrastructure investments such as LNG facilities, pipelines, and petrochemical production -- all of which receive significant subsidies that reduce risks, and therefore financing costs, for these long-lived, fossil-intensive assets.

Under all price scenarios analyzed, the highest-value subsidies in our paper include federal tax incentives that have existed since 1916, as well as less recognized forms of support such as cost exemptions related to well cleanup and hazardous waste management.

Subsidy analysis always requires the authors to define what should be viewed as a subsidy, and a counter-factual case to use to assess the subsidy value.  A global consensus on these definitions among practitioners has been growing over time.  Still, disagreements on both of these factors are common.  Sometimes the driver is differences in opinions on technical issues and baselines.  Often, however, claiming something is not really a subsidy is a political strategy deployed by the beneficiaries.  In any case, because of these definitional issues, our analysis also presents results for selected subsets of subsidies.  This allows readers with some definitional differences to nonetheless make productive use of our analysis and findings.

National estimates of dollars of subsidies flowing to the oil and gas sector are a useful metric of the scale of government intervention.  But the impacts of subsidies on investment and production decisions occur at the field level, and the same subsidy can be more or less significant in propping up producers in different types of geologic formations or parts of the country.  Our analysis helps to guide state and federal decisions on subsidy reform by illustrating which subsidies are most important on a regional level as well. 

Figure 1 below shows the average effect of each subsidy on the internal rate of return (IRR) of new, not-yet-producing oil and gas fields, at average 2019 prices of USD2019 64/barrel of oil and USD2019 2.6/mmbtu of gas. The charts on the left show the oil production-weighted average change across all oil-producing fields, and the charts on the right show the gas production-weighted average change across all gas-producing fields, in the US and in a given state. Fields that never reach a positive IRR (even with subsidies) are not included. 'NA' labels indicate that a given subsidy was not applied to that fuel and/or state.

The underlying data for this figure, as well as for other major oil- and gas-producing states not shown, can be found in the supplemental materials.  Many other detailed exhibits are also available in the published report and supplemental materials as well.

Fossil-Fuel Subsidies Must End: Despite claims to the contrary, eliminating them would have a significant effect in addressing the climate crisis

When it comes to tackling the climate crisis, ending $400 billion of annual subsidies to the fossil-fuel industry worldwide seems like a no-brainer. For the past decade, world leaders have been resolving and reaffirming the need to phase them out.

Why fossil fuel producer subsidies matter

This article in Nature explains how subsidies affect fossil fuel investment and why they deserve greater attention in global modelling analyses.

It responds to a 2018 study in Nature that used the results of integrated assessment models to infer that eliminating subsidies would yield “limited emission reductions…except in energy-exporting regions”, and described the emission reduction benefits as “small”.

1)  Solar parity and my cousin Jeff.  Congrats to Jeff Koplow, an energy researcher at Sandia National Labs, for being named the inaugural recipient of the Innovator in Residence Fellowship awarded by DOE's SunShot Initiative.  He'll lead a multi-disciplinary team in attacking key limits in current PV systems.  


Jeff's been working on energy innovation for a very long time, and some of his earlier inventions offer large benefits to the US energy sector.  These include the Sandia Cooler, a much more efficient way to cool heat exchangers and CPUs; and Twistact (winner of an Outstanding Technology Development Award), a new way to connect turbines to the gear box without a sliding contact, electrical arcing, or the need for rare earth metals.  

sandia cooler


2)  Fossil fuel subsidy reform update from COP21 (and nukes too). "The Beginning of the End of Fossil Fuel Subsidies," a blog update from Paris by staff of the Global Subsidies Initiate, summarizes some of the interesting and promising developments on fossil fuel subsidies at COP21.  Read their take here; there is real progress.

Here's one of the (albeit non-scientific) benchmarks I use to gauge progress on subsidy reform:

  • 25 years ago, almost nobody outside of NGOs would talk about energy subsidies.  (I once had a potential faculty advisor for my update on US energy subsidies ask why I didn't want to work on something "useful").
  • 15 years ago, heads of environmental and public health ministries would willingly talk about energy subsidies.
  • As of about 7 years ago, heads of financial ministries would willingly talk about energy subsidies.
  • Today, even many heads of state willingly talk about energy subsidies -- at least fossil fuel subsidies.  Conversations about subsidies to the nuclear fuel cycle will have to wait another few years, I suppose.

For a report on the push for massive new subsidies to nuclear at COP21, here's a summary from Michael Mariotte of NIRs.

3)  Addressing tax exemptions to fuel used in international shipping.  Subsidies that almost nobody sees or thinks about can be particularly distortionary.  In darkness, the biggest mushrooms grow -- or something like that.  Wholesale exemptions of fuels used in international ship, air, and rail transport from taxation is one example of issue (the subsidy even made my top 10 most distortionary energy subsidies list).

Particularly for activities that cut across multiple countries, the options available to correct the problem can be heavily constrained by pre-existing international agreements.  Those agreements were often developed to with trade or sovereignty issues in mind; transparent pricing of natural resources was generally not a factor.

One possible solution to this conundrum is put forth in a recent working paper ("Drying up tax havens-A mechanism to unilaterally tax maritime emissions while satisfying extraterritoriality, tax competition and political constraints") by Dirk Heine, Susanne Gäde, Goran Dominioni, Beatriz Martínez Romera, and Arne Pieters.  As the title implies, there are a fair number of hurdles any solution needs to meet.  The authors focus on using cargo as the tax base rather than the fuel itself; and structuring the tax levy formula such that higher efficiency vessels end up with a lower tax.  To avoid distorting port-use decisions, transhipped cargo would not incur the tax.  There is contact information for the authors in case any of you have additional suggestions or ideas.  There is also a summary description of their plan here, which provides a quick overview of the approach.

4)  Limiting export credit support for coal plants.  Export credit subsidies have been an area much talked about, but too often ignored when subsidy reform plans are put forth.  The subsidies are often tough to value, but can be quite large and tip high risk projects in environmentally-sensitive regions of the world from "no-go" to actionable.  The subsidies generally work by offering direct loans or guarantees at below the market rates that firm or industry would normally be able to receive.  The Kusile power stationsupport may also generate an incremental subsidy by enabling high risk enterprises to utilize higher debt-to-equity ratios than would otherwise be possible, reducing the project's weighted average cost of capital.

Against this backdrop, an agreement among OECD nations to significantly reduce their ability to finance the export of low-efficiency coal-fired power plants and equipment is a very positive step.  Read the agreement here.

5)  Charles Koch, one of the Koch brothers, argues for subsidy elimination.  It's hard to imagine the whipsawing that must have been ripping through Charles Koch as, year-after-year, his push for small government, free market libertarianism ran smack into his affinity for the array of special government subsidies that have benefited his oil, gas, and now paper operations for generations.  Maybe his latest book, Good Profit: How Creating Value for Others Built One of the World's Most Successful Companies, is his effort to finally reach an internal cease fire.

To his credit, Koch does acknowledge he's been a subsidy beneficiary.  And, I do agree with a core premise of his criticism:  that our convoluted and complex system of subsidies tends to favor the well-connected and the wealthy, replacing economic viability with political connections in the selection of markplace winners.  Lower income quintiles clearly need the support more than the rich.  Similarly, on the corporate site, the new companies that will create the jobs and industries of our future are generally too small to have lobbying arms.

How big are subsidies to Koch-related firms?  This article pegs the figure at about $200 million since 1990, based on data from DC NGO Good Jobs First.  I consider this a low estimate, probably really low.  The Good Jobs First data sources don't always pick up subsidies through credit or insurance markets.  Non-standard ones, such as the "black liquor" tax break that many of the paper companies tapped into by stretching a pre-existing rule targeted as new fuels, are also often missing.  Data on privately-owned Koch subsidiary Georgia Pacific are hard to come by, but given its scale, this tax break alone could have exceeded $200m.  In fact, this analysis estimated black liquor subsidies to Georgia Pacific just in 2009 at $1 billion.  Finally, the private nature of Koch means commonly-available tax breaks to oil and gas may not show up in the $200m figure either.

The point here isn't to attack Koch for saying corporate welfare should be eliminated.  The subsidy trough is a crowded one, so the limiting the push for reform only to those who are "pure" enough not to have been subsidized would be foolhardy.  If Charles is serious about finally reforming subsidies, this is a good thing.  Properly done, the changes would improve the economic vitality of the country going forward. 

But one should not take the input of Koch or other current beneficiaries blindly.  It is vital for anybody working on subsidy reform to read the small print on any proposal to be sure there are no surprise gaps.  Varying definitions (or more perjoratively, a definitional sleight-of-hand) can coincidently eliminate the subsidies to competitors while keeping their own.

Fossil Fuel Subsidies: Building a Framework to Support Global Reform

Keynote presentation at the Expert Workshop on Subsidies to Fossil Fuels and Climate Mitigation Policies in Latin America and the Caribbean (LAC), held at the Inter-American Development Bank in Washington, DC on January 14, 2014.  Slides review recent global estimates of fossil fuel subsidies, highlighting both the tallies and the reasons the estimates differ widely from one another. 

In advance of the G20 meeting in Los Gatos, Mexico this past June, Earth Track and Oil Change International did a detailed review of how the member countries were doing in meeting their 2009 commitments to phase out inefficient fossil fuel subsidies.  The resulting report, Phasing Out Fossil-Fuel Subsidies in the G20: A Progress Update, summarizes the limited progress that has been made to date.  Our initial report, released in 2010, can be found here.

Something to Talk About, Even if It Brings Little Real Change

g20 mexicoInternational agreements are always political.  The wording is often vague, allowing countries to "agree" on a document that sometimes means very different things to each of the signatories.  In some cases, this vagueness allows at least some steps to be taken on important international issues, with scrutiny and compliance rising over time as institutions, measurement, and systems of accountability are established. 

However, if the agreement lacks any enforcement mechanism, compliance may decline over time rather than increase.  Circumstances may change for signatories, or countries may see peers doing little to meet the intent of the agreement and be unwilling to continue to pay the economic or political cost of compliance while others free-ride.  Unfortunately, the G20 fossil fuel subsidy reform commitment falls into this second category.  These factors likely explain the fairly abysmal results of the most recent update provided by member countries after the Los Gatos meeting:

As in 2011, members were this year invited to submit reports on their progress in phasing out inefficient fossil fuel subsidies.  All members responded; eleven respondents indicated having no fossil fuels to report.

That's 11 out of 20 -- more than half -- claiming they have no reportable fossil fuel subsidies.  This number has been rising with each reporting sequence.  And the countries that did report?  They provided only a smattering of information on what is generally wide-spread support for fossil fuels around the world (2012 country submittals are here).  There is far more missing than what has been included.

Defining Your Way Out of a Politically Tricky Situation

As part of our most recent review, we looked at how countries were defining their responsibilities under the G20 phase-out effort.  Specifically, we examined the detailed definitions of which policies were reportable under their interpretation of the G20 phase-out communique.  These are summarized in the table below.  To highlight potential reporting gaps that the particular wording creates, we have italicized selected text in the left column. 

There are striking differences in how countries view the same commitment.  The table clearly illustrates how selective definitions enable countries to limit reporting and reform to a very narrow set of policies while still symbolically supporting the international reform effort. The main themes were:

  • Ignoring opportunity cost. A group of countries, including large exporter Saudi Arabia, but also South Korea and Turkey, have excluded the sale of domestically-produced fuels at below-market prices from their definition of subsidies so long as direct production costs are covered.
  • Excluding targeted subsidies so long as average fuel taxes stay high. Italy has most clearly defined this approach, arguing that because national prices remained above world levels, variation in incremental fuel taxes above that base were not subsidies. Relative to a domestic baseline, however, the differential subsidies by industry create distortions in investment patterns and returns, and replace economic signals with political ones.
  • Selective coverage of policy types included within country definitions. Many countries say they have adopted a fairly broad definition of subsidies promulgated by the IEA, but then report only a small portion of the policies (primarily tax breaks) that the definition includes.  
  • No formal definition. Avoiding a formal commitment to a specific definition of subsidies gives countries more flexibility to omit without mention particular policy interventions. Definitional gaps are evident in reporting by key countries such as Russia and China. The US submittal does not include an accepted definition either, yet focuses almost exclusively on a narrow set of tax breaks to the fossil-fuels industry.

Subsidy Definitions Vary by Country,
Lead to Gaps in Reporting and Reform Commitments

View table in PDF or full report.

*Extracted from Doug Koplow, Phasing Out Fossil-Fuel Subsidies in the G20:  A Progress Update, June 2012.

Country Definition of Subsidies Subject to Phase Out

(Emphasis Added)


Potential Gaps



European Union

“For the purposes of the exercise launched by the G20 Pittsburgh summit, the EU and its Member States have chosen to take as a working definition of fossil fuel subsidies the following, based on the approach of the International Energy Agency:


A fossil-fuel subsidy is any government measure or program 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’.


Definition is actually fairly broad, in theory capturing any type of government program, regardless of intent, that either modifies energy prices or changes revenues or costs for producers. In practice, though, most of the countries have picked up only taxes and a few direct expenditures.


As noted below with Italy, the definition seems to focus on national average levels, missing sector-specific tax breaks (e.g. , reduced taxation of fuels used by fishing fleets).


Risk transfers may also not be well captured by this definition, particularly if the connection to costs or revenues is not immediate, but through changes in the expected returns (lower volatility returns) in a particular sector.


Australia does not have measures related to the production of fossil fuels that fall within the scope of the G20 commitments.


Australian Government budgetary support for fossil fuel production is limited to measures that are intended to support production of clean energy.”


Australia does not have any sector-specific tax expenditures for fossil fuel production (although fossil fuel producers are able to access general measures that apply across the economy or across the mining and quarrying sectors as a whole).


 • Subsidies to less polluting forms of fossil-fuels (e.g. , clean coal or pollution controls) even though they may still be “dirtier” than renewable alternatives.


Policies that have the effect, though not the intent, of subsidizing fossil-fuels seem to be excluded.


Special tax breaks for extractive industries (e.g. , percentage depletion) that are generally viewed as subsidies in most other countries in the world.


“There are two broad possible approaches that Canada could take to this commitment: 1) Use the commitment as an opportunity to undertake selective rationalization of Canadian measures (which we recommend), or 2) If Canada is not prepared to undertake any substantive reforms, minimize the obligation so that Canada can still position itself as meeting the commitment (Horgan 2010).


This leaked memo illustrates the political aspects of subsidy definitions in how a countrys response to the G20 commitment is framed.


The limited items reported illustrate the country chose the second option.


It was decided that all the countries would provide their own definition of inefficient subsidies. Accordingly, following [sic] definition of subsidies has been adopted by India:


A fossil fuel subsidy is any Government measure or budgetary support that has a consequence of reducing the effective cost for fossil fuel paid by consumer, (after accounting for taxes on these fuels) or of reducing the costs or increasing the revenue of fossil fuel producing companies.’”


Adjustments for taxes may mask important user subsidies to fuel sector (e.g. , roads or other transport infrastructure, tank cleanups).


Indias own 2010 progress report submission illustrates what they believe is excluded from reform: “It may be mentioned that this list does not include the indirect subsidy provided for energy services like tax benefits on profits derived from commercial production and refining of mineral oils and natural gas; investment linked incentives for expenses on new pipelines; sales tax concessions by State/local government etc.


According to the Indonesian Budget Law, fuel subsidy defined [sic] as a budgetary allocation given to a company or institution that produces and/ or sells the oil fuel and Liquefied Petroleum Gas (LPG), with the purpose of providing access to energy at an affordable price for consumers.


Non-budgetary transfer approaches (e.g. , tax, credit, insurance subsidies).


Programs that subsidize costs for reasons other than providing energy access at an affordable price.


“Italy considers favorably the International Energy Agencys (IEA) definition of fossil fuel subsidies as: any government measure or program with the objective or direct consequence of reducing below world-market prices, including all costs of transport, refining and distribution, the effective cost for fossil fuels paid by final consumers, or of reducing the costs or increasing the revenues of fossil-fuel producing companies.


“However, and according to this definition, Italy as much as most other EU member states does not have subsidies that lower the price of fossil fuels below international market price levels. Furthermore, State aid within the EU is clearly limited by the Treaty on the Functioning of the European Union (TFEU), which forbids any public support not compatible with the TFEU.



Although Italy generally adopted the same subsidy definition as other EU members, the country added the clarification on taxes that other members seemed also to have applied, though did not state.


As a result, any situation where specific sectors are receiving higher subsidies (or tax reductions) than others may be missed. Even if overall tax rate results in prices above EU minimums, there can be inter-sectoral distortions and these can cause important environmental problems.

South Korea

Korea defines fossil fuel subsidy as a government measure with the objective or direct consequence of reducing below production costs for net importers, world price instead or production cost the effective cost for fossil fuels paid by end consumers, or reducing the costs or increasing the revenues of fossil fuel producing companies.


South Korea has adopted the OPEC viewpoint that selling above production costs but below world prices is not a subsidy. The opportunity cost of these programs can be large, and the allocation of windfalls via political means often results in corruption, black markets, and shortages.


While current policies in Mexico are consistent with the goals of the G20 commitment, we believe that in order to make a stronger commitment regarding the phase out of our fossil-fuel subsidies, it would be necessary for all countries to agree on a uniform methodology for calculating subsidies. Using such a common methodology, peer monitoring would be an effective tool to gauge progress across countries in removing fossil fuel subsidies in an objective and clear manner.



No specific definition of what counts as a subsidy to Mexico; only a recognition that absent a formal process for establishing a common standard there are likely to be problems.


No clear definition of subsidies has been put forth by Russia, though there is recognition that reform of consumer prices for energy would be included.


While the provision of a working definition does not ensure all important subsidies will be captured, the absence of a working definition means exclusions are even more likely.

Saudi Arabia

“Saudi Arabia has considered a definition of inefficient subsidies on the basis that there is no cost to the Government that outweighs the social and economic benefits of the pricing mechanism, leading to wasteful rather than natural growth in consumption, and that these benefits, including in the form of economic diversification, cannot be provided by equally effective ways or by the use of available alternative sources of energy.


“Based on these criteria, the Government would like to articulate that while domestic fossil fuel prices in Saudi Arabia could be below international prices, these prices reflect the countrys comparative advantage in oil production and are above the production costs. Indeed, the Government is not paying any fossil fuels-related subsidy from the treasury. Therefore, Saudi Arabia is not implementing any measures that fit the criteria for inefficient fossil fuel subsidies. The G20 proposal for phasing out inefficient fossil fuel subsidies does not therefore apply to Saudi Arabia.




Large opportunity cost of selling fuel domestically at an artificially low price is not being recognized. NGO assessments of Saudi Arabia have indicated that the underpricing has resulted in a wide array of problems regarding over-consumption, inefficiency, and poor investment decisions.


There seems to be little data on producer subsidies within the Kingdom, such as via credit support, subsidized insurance, or post- operational cleanup and closure of drilling sites.



“The appropriate definition for ‘Inefficient Fossil Fuel Subsidy’ is stated below:


A fossil-fuel subsidy is any government measure or program with the objective of reducing, below production cost, the effective cost for fossil fuels paid by consumers or of reducing the costs or increasing the revenues of fossil-fuel producing companies through measures other than efficiency improvement measures and/or measures for the penetration of new technologies (e.g., clean coal technologies).’”


Any subsidy to a new technology” would not meet the definition of an inefficient subsidy according to Turkey.


Consumer subsidies exempted as well so long as prices remain above production cost.


Unlike the standard IEA definition, Turkey has excluded government measures that have the “direct consequence of distorting markets, even if that end was not an intent of the policy.

United States

“There are a number of tax preferences, described below, available in the United States to producers of fossil fuels. The preferences below are all permanent provisions in the tax code.


Subsidy mechanisms other than tax breaks.

Subsidy quantification based on single source (Treasury), though estimates from other parts of government often disagree.