API Comm (2)

API False Claim #5:  The subsidies documented do not affect oil consumption, but only the size and composition of the domestic industry relative to imports.


Some of the more obvious subsidies, such as many special tax breaks for domestic production, do primarily benefit domestic producers.  However, one of the interesting findings of the Greenpeace study is that subsidy elimination might actually benefit domestic producers relative to the major importers.  This is because there are very large subsidies that benefit primarily imported oil.  While API concludes there is no impact on consumption patterns, this is the inevitable result of them first "disappearing" as many of the government supports their industry gets as they can possibly stitch up a story for justifying.  As described in section 7 below, API excluded a nearly $2 billion worth of subsidies per year from their "true" subsidy total without even mentioning they had done so in their "Fueling Confusion" campaign.


With the exception of extremely narrow studies conducted by the Energy Information Administration, there are few other studies that come up with values anywhere close to the miniscule support API chooses to acknowledge.  A recent article in the Annual Review of Energy and the Environment examined all of the major studies of US fossil fuels over the past three decades illustrates this point clearly:



EIA's work, in both 1992 and 1999, suffered from a number of limitations.  View here for a detailed discussion of why EIA values are so much lower than other studies.  The subsidies are far too narrowly defined; important programs they view as subsidies were not added into their total values either due to how the study was scoped or to measurement problems; they do not look at off-budget subsidies in any detail; and they exclude any subsidies of large, but not sole, benefit to the energy sector.


API False Claim #6:  A significant part of the subsidies listed are incentives to produce alternative fuels such as natural gas, not oil.


Listed as a primary claim on the first page of API's evaluation of FGW, this item is not mentioned again throughout the entire more detailed discussion.  In fact, every single subsidy element in the FGW study has been pro-rated to reflect only the oil portion.  This fact is clearly stated in the introduction to the report, where it is inconceivable that API could have simply missed it.  Certainly, there are some subsidies to oil and gas that do primarily benefit natural gas.  However, for these subsidies, there has been no erroneous attribution of the value of them to oil.  For example, only a very small portion of the alternative (non-conventional) fuel production credit benefits oil; the vast majority supports natural gas.  This is why of the $970 to $1,370 million per year the Treasury loses from this provision, a scant $10-$27 million has been allocated to oil.[5]

Specific Subsidies:  API Attempts to Reclassify its Subsidy Problem Away


Faced with detailed, quantified estimates of how much their member firms receive in government support each year, API attempted to discredit each subsidy element one-by-one until there was nearly nothing left.  At that point, the Institute's researcher could piously claim that "oil consumption is not being artificially stimulated."  Below, we analyze API's specific claims against particular subsidy elements and demonstrate why they are incorrect.  We have applied basic concepts of finance and accounting in our evaluation of subsidies to oil, many of which are regularly applied by API's member companies within their own operations.  Where API itself is arguing that these basic approaches should not be used, we provide examples from outside of the oil sector to demonstrate how the API line of reasoning breaks down.  We also identify the areas that API eliminated from the oil subsidy total with no mention at all.


1.  Defending Persian Gulf Oil


API argues that defending Persian Gulf oil is not properly considered a subsidy to the oil imports because the defense operations there support more objectives than just oil security.  This claim in interesting for a number of reasons.  First and foremost, API is acknowledging that one of the functions the military in the region serves is to protect oil security.  This is an important admission.  Second, API points out that the military serves multiple functions in the region, which is exactly the same point made in the FGW report.  Third, when API presents the FGW estimate for this military presence, it does not acknowledge that the share we have attributed to oil is but a portion of the total cost of the military presence, reflecting the multi-mission element of the engagement.  Fourth, there is no recognition that we are not advocating for this security service cease, only that an equitable portion of the security cost be funded through a tax on oil, rather than from general tax revenues.
Since API implicitly acknowledges that the US military presence does provide oil security services to the industry for which oil producers or consumers do not pay for directly, its dismissal of the report's estimate is based solely on how to value those services.  API subscribes to the marginal cost approach, which, coincidentally also generates the lowest (and totally implausible) subsidy to oil.  The argument is that because the defense department would need to spend money on its other regional objectives even if there were no oil in the area, the incremental cost of protecting oil is near zero, and this is the only cost "properly" counted as a subsidy.  The problem with API's reasoning, however, is that each regional objective (ensuring access to oil supplies, preserving regional stability, and preventing the emergence of regional hegemonic powers)  can make exactly the same argument.


In addition, while the marginal costs of meeting any one objective are relatively small, the total costs of the baseline military capability are enormous.  Were every objective to pay only the incremental cost of their particular mission, there would be no funds left to pay for the massive fixed costs of having personnel and equipment in the region to meet the combined needs.  The simple and incontrovertible fact here is that a marginal cost approach to costing Persian Gulf oil defense is simply not a viable metric.  It is no different than an airplane flight.  The incremental cost of any additional passenger on the airplane is miniscule:  a little ticket agent time, an ever-smaller bag of peanuts, and perhaps a soda.  Yet if every passenger paid only $5 for a flight, there would be no revenues to cover the larger baseline costs, like the airplane, fuel, and pilots.


As described in detail in chapter 4 of FGW (http://archive.greenpeace.org/~climate/oil/fdsuboil.pdf) subsidies to Persian Gulf oil must be evaluated through the lens of "common costs." Common costs refer to situations where two or more outputs are produced simultaneously from the same "production" process.  In this case, the presence of the U.S. military in the Persian Gulf region, along with all of the general overhead that makes this presence possible, is the "production process" in question.  Rather than a factory making two physical outputs off the same production line, the security presence represents a joint provision of services.  The fixed costs of this presence are properly allocated among all of three regional security objectives, of which protecting oil supplies is one.  For reasons described in more detail in the report, we have allocated the costs equally to each objective; thus, the estimates contained in FGW that support oil markets is equal to one-third of the entire cost of the US presence in the region.

2. Strategic Petroleum Reserve (SPR)


The Strategic Petroleum Reserve holds hundreds of millions of barrels of oil, owned by the federal government, to buffer supply shocks.  The cost of this reserve has been continuously understated since its inception, and what costs have been recognized have been funded by the general taxpayer.


API makes two arguments about the SPR subsidy estimate.  First, they claim that we quantified the cost of SPR using “an assortment of questionable methods,”  and second, that benefits from the SPR flow to more than oil consumers and therefore should not be paid only by oil consumers.  In fact, the valuation methods we used simply recognize that multi-year investments must be financed, and that these finance costs need to be recognized in the cost of the good or service produced.  We agree that there are broad macroeconomic benefits from having an SPR to buffer supply shocks, and for that reason do not advocate discontinuing it.  Rather, we advocate improved accounting for the Reserve's real costs and a change in how these costs are funded.


Although API acknowledges that SPR stabilizes petroleum prices in the United States, it fails to recognize that this government intervention in oil markets makes oil a less risky source of energy, thereby helping it compete with alternative energy sources.  Increased price stability encourages oil use and discourages the development of alternatives.  Because the need for price stabilization is driven by an over-reliance on oil, it is oil consumers, and not the general taxpayer, who should pay for the Reserve.  Only by forcing oil consumers to pay for this insurance policy against supply disruptions can the proper price signals enter energy markets.  This encourages the development of a more diversified and flexible pattern of energy demand over the longer-term.


Valuing Subsidies to SPR


We encourage interested parties to review the detailed description of how the subsidies to SPR were calculated (see Chapter 4 of the report), and we trust that most reasonable people will agree that the method was both accurate and fair.  As is common throughout the “Fueling Confusion” initiative, API does not specify which methods it considers questionable, allowing the organization to try to discredit the subsidy numbers without having to take a clear position on items where it would likely be proved wrong in short order.


The bulk of the subsidies to SPR come through the cost of financing construction and oil inventories, which is not currently recognized on entity's books at all.  We guess that API views these imputed interest charges as “questionable” because of the fact that they are not currently levied by the Treasury (the source of capital for these items).  This, in large part, is our point.  Private companies, including all of API’s members, must finance both their capital investments and inventory.  If they cannot earn a sufficient return to pay for all direct costs, plus the financing costs, plus an adequate return on capital, they will go out of business.  We are sure that anyone reading this paper would love to have the use of a billions of dollars for decades without having to pay any interest on the money, which is essentially what SPR does.


The fact that SPR can just pretend none of these costs exist is a huge subsidy, and one with very real costs.  For every dollar tied up in SPR’s oil inventory for 20 years, the government needs to issue a dollar of debt, on which it pays interest to other parties.


SPR as Benefiting More Parties than Just Oil Consumers


API argues that because smoothing price hikes in oil protects many others in the economy as a whole, the entire world economy should pay for the service.  We agree that SPR does have broad benefits, and should be retained.  However, it does not follow that because SPR has spillover benefits that the cost of the Reserve should be borne by the general taxpayer.  API’s position is flawed in a number of respects, and certainly does not reflect how most other markets function.


It is the scarcity of oil, driven by a lack of short-term substitutes, that triggers the macroeconomic effects of concern.  Market pricing is one of the most powerful mechanisms to clearly demonstrate where scarcities exist, and to signal that more of a commodity should be produced and/or the existing stock should be used more judiciously.  Hiding these price signals makes little sense; nor does it match what happens in other sectors of the economy.  By protecting consumers and refiners from oil market disruptions, SPR reduces both the need for private sector entities to establish their own inventories and the incentives for oil consumers to increase their ability to shift fuels in times of oil shortages.  API’s arguments that the entire world benefits from the US Petroleum Reserve is also overly-simplistic, since the International Energy Agency requires member countries to have their own reserves independent of SPR.


It makes perfect sense to attribute the full costs of SPR to domestic oil markets.  Consider a number of examples to illustrate this point:

  • Economic Multiplier Effects.  When a factory hires workers and pays them wages, these wages benefit others in that local economy as these workers buy groceries, rent homes, and purchase other goods and services.  Yet, the grocer doesn’t pay 2 percent of the factory workers' wages to say “thank you” for this uptick in sales, though using API’s logic he should.
  • Constrained Resources in Production Systems.  Modern production systems can be extremely complex, and a scarcity of a single input can trigger large costs on the system as a whole, as production slows or stops entirely.  A common technique used to optimize production systems is “shadow pricing,” which uses mathematics to set internal prices on particular aspects of the production system.  As the system becomes constrained, the shadow price on the constrained (or bottle-necked) resource rises dramatically, illustrating the economic value to the system overall of investing in ways to alleviate that bottle-neck.  The price on the constrained resource is what rises steeply, even though all parts of the production system would benefit if the constraint were removed.  Oil is no different, except here the macro-economy is the “production system” and oil is the constrained resource.
  • Pollution.  Consider the hypothetical example of an older oil refinery in the Southeastern United States.  The facility continues to pollute the surrounding area quite heavily, to the detriment of surrounding residents and wildlife.  Using API’s logic, since these surrounding populations also benefit if the plant reduces its emissions, they should pay part of the costs to do so.  Such as strategy would violate the “polluter-pays principle,” at the core of much domestic and international environmental policy.

3.  Accelerated Depreciation


Accelerated depreciation provisions enable companies to write off plant and equipment expenditures for tax purposes faster (sometimes much faster) than the assets actually wear out.  This treatment of costs departs from the general accounting principal of matching income with the cost of producing that income and of depreciating assets over their useful lives.  According to the Congressional Research Service, while these subsidies were reduced substantially in the Tax Reform Act of 1986, the economic decline rate for both equipment and buildings is still "much slower than that reflected in tax depreciation methods."[7]  Only subsidies from oil-related capital are counted in FGW as a subsidy to oil.


API hopes to dismiss the substantial benefits that the oil industry receives through accelerated depreciation by arguing that because the provision is available to everybody, it is therefore not a subsidy. As so adeptly stated by George Orwell in Animal Farm, "All animals are equal, but some animals are more equal than others."  Capital-intensive industries, especially those with long-lived capital, are "more equal" under accelerated depreciation rules, and important economic distortions can result.


First of all, the depreciation rules vary by industrial sector and equipment class.  In fact, there are specific rules that deal only with oil-related capital.  Depending on the degree to which the tax write-off period deviates from the actual service life across the various asset classes, the rules can introduce significant cross-sectoral distortions.  In addition, the provision overall tilts energy markets away from less capital-intensive methods for meeting energy demand.  The policies directly disadvantage energy conservation and efficiency, since capital equipment is eligible while consumer durables or home improvements such as insulation are not.

4.  Foreign Source Income


Fueling Global Warming evaluates two subsidies related to earnings abroad by oil companies:  foreign tax credits and deferral of foreign income.  API does not address the first (and larger) item in its rebuttal at all, though it does zero out the subsidy from its list of "true" subsidies to oil.  The Institute argues on the second that there is no subsidy to oil because, again, it is a provision applicable to all industries, not just oil.  Both of these provisions confer substantial economic benefits to the oil sector.


Foreign Tax Credits


It is standard treatment for all industries that taxes paid to foreign countries on foreign corporations be credited against US taxes owed to avoid double taxation.  FGW agrees that this approach is a sensible one, and does not state that all foreign tax credits taken by oil producers constitute subsidies.  While we are in agreement with API on the theoretical concept here, it is in the actual implementation of the foreign tax credits where things get interesting, and where API is totally silent.


There is a fairly extensive history of oil companies receiving tax credits rather than tax deductions for payments to foreign governments that are actually royalties rather than taxes paid.[8]  This shifting results in the loss of hundreds of millions of dollars per year to the U.S. Treasury.  The tax games are especially apparent when oil companies report paying taxes in countries that have no corporate income taxes on other sectors.  In other cases, tax rates are higher for oil companies than for other sectors, suggesting similar, though less extreme, shifting.  Our low estimate of $486 million per year in subsidies counts only situations where the oil companies are claiming credit for income taxes in countries where no other sectors pay these taxes.  The high estimate ($1.06 billion per year) assumes that some of the payments in countries that do have corporate income taxes on other sectors are disguised royalties.


Interestingly, API's argument that the oil companies actually pay worldwide taxes above the statutory rate supports our position that there is significant reclassification of royalty payments into the foreign tax category to obtain benefits related to US taxes owed.  Should API truly wish to discredit this item as a subsidy to the oil industry, we would be happy to work with them on an acceptable data reporting format for their members that would demonstrate what is really happening one way or another.


Deferral of Foreign Source Income


When a U.S. firm earns income through a foreign subsidiary, that income (net of foreign tax credits) is taxed only when it is repatriated as dividends or other income.  Because the parent firms are able to time when this happens, they can defer their U.S. tax liabilities for many years.  FGW includes the oil portion of this subsidy in its totals for the sector.


Once again, API claims that the deferral of foreign source income is not a subsidy because all firms can do it.  Were this really the case, the US government's tax expenditure estimates would not track the provision year after year.  Nor would the European Union have brought a case before the World Trade Organization -- which the US subsequently lost.  At best, the fact that multiple industries can use this provision could, theoretically, mean that the subsidy does not distort economic decisions.  Yet, given the large US role in oil multi-nationals, and the much larger importance that foreign trade has in the oil industry than in most other sectors of the economy, it is likely that the subsidies associated with this provision flow disproportionately to oil.  We estimate the subsidy at between $60 and $300 million per year.


Overall Tax Levels


Using data collected from the major oil producers by EIA, we estimated that the majors paid an average effective tax rate of only 12 percent in 1995.  API attacks this figure as being "particularly egregious in misrepresenting the facts" because it includes some international income in the calculation.  While there is some international income included, we have included only the total income subject to US taxes.  In fact, including this foreign income is properly done since the corporations are US firms that have repatriated earnings to this country.  By arguing that income subject to US taxation from all countries is not a legitimate basis for calculating taxes paid, API is essentially claiming that a US-based multinational should pay absolutely no US taxes on earnings abroad.  This is a that position many in the government are unlikely to agree with.


Our calculations in deriving the 12 percent figure are crystal clear:  we divided actual taxes paid by the total income subject to U.S. taxation.  International income not subject to US taxation was properly excluded.


We also carefully described our findings:


"integrated producers have paid roughly 25 percentage points less in taxes than their statutory rates suggest they owe.  This differential is evidence of the substantial tax breaks they have received over the past 20 years.  Special provisions reduced integrated producers' tax liabilities by roughly $7.0 billion in 1995.  This approach yields subsidy estimates nearly $3 billion higher than what we calculated on a provision-by-provision approach.  About $1 billion of this differential can be accounted for by the fraction of foreign tax credits claimed and state/local tax deductions that are properly excluded from U.S. taxable income to avoid double taxation.  This leaves a $2 billion discrepancy between the two estimation methods that we are unable to reconcile given available data.  Due to this limitation, we use the lower estimates for tax subsidies, calculated on a provision-by-provision basis, in our totals.  Although this approach is more conservative, it may understate the value of tax breaks to oil."

Had we been able to fully reconcile the differences, the subsidy values to oil presented in the report could well have been billions of dollars higher.


5.  "General Government" Expenses


Not every type of energy requires the same level of government oversight.  For example, nuclear power plants require far more government oversight than a plant using gas turbine technology to deliver a kilowatt hour of electricity.  If the consumers paid nothing for this oversight, they would believe that the two forms of power were equal, and there would be no price signals to shift to energy sources that require less public resources to be sure they operate safely.  This same line of reasoning applies to the provision of many types of government services, including market regulation, health and safety oversight, and information gathering.  While beneficiaries of these activities often refer to the services they receive as "basic functions" of government, the activities often benefit one sector (in this case oil) far more than others.


With respect to the FGW report, API was both selective and inaccurate in its list of subsidies it mentions as inappropriate.  For example, API claims that FGW counted the costs of "administering programs associated with the development of oil resources on public lands" as subsidies.  In fact, while they have been costed out, we explicitly excluded them from our subsidy totals.  As stated in the report, we do not "treat the government cost to manage oil sales as a subsidy to oil.  This decision reflects the necessity of incurring some costs in order to earn the oil royalties in the first place, and the relatively low level of overhead costs in comparison to other natural resource areas, such as timber."[10]


6.  Royalty Payments


There are two issues driving royalty-related subsidies:  improper payments by the oil companies ($31-$130 million/year) and lapses in proper auditing of leases by the Bureau of Land Management ($50-$75 million/year).  API argues that the first item is incorrect because the oil companies have paid royalties properly in the past.  We understand that they must take this position because their members have litigated this issue (though they have generally lost).  However, there is substantial evidence that past payments, which relied on artificial transfer prices set between different branches of the same company, were improperly done.  This would not be the first time that transfer pricing had been used by the oil industry to minimize required payments to the U.S. government.  A common practice in the 1960s used transfer pricing to shift profits from the entire oil production chain to a tanker subsidiary.  These subsidiaries were then registered in countries such as Liberia or Panama that had extremely low tax rates.  As a result, most of the global profits of these integrated producers were able to escape taxation.[11]


7.  The Mysterious Final Slice:  API Eliminates Subsidies Without Even a Mention


Among the more egregious departures from accuracy in the API "Confusion" initiative is the final slice of pie, presented as depicting the only "true" subsidies to the oil sector.  With only this single slice of pie resting on the API sky and clouds background, it is reasonable for the reader to assume that the $0.4 - $1.3 billion in subsidies described in the text comprise what is left from the FGW report.[12]   The reader would be wrong.  In fact, even if one -- for the sake of argument -- subtracts every subsidy that API questions, the residual is between $2 and $3.5 billion per year.  This is between three and five times higher the what API says is left.


So what has mysteriously disappeared?  Government subsidies to water transport infrastructure, of which the oil industry is one of the largest users, has disappeared with nary a mention.  Subsidy totals:  between $690 and $775 million in 1995.  Subsidized lending to foreign sales of oil field equipment or development of foreign oil fields is also gone, though worth between $205 and $270 million per year.  Another disappearance is federal spending on oil research and development.  Though much lower than in the past, it is still provided nearly $120 million in public funds in 1995.  The entire issue of inadequate bonding at operating wells was also not mentioned, but provided between $170 and $550 million in reduced bonding premiums to well operators in the year of the study.




Despite API's efforts to redefine, reclassify, and sometimes just erase federal programs that support oil producers and consumers, it is clear that oil receives significant subsidization within the United States.  This makes little sense from an environmental, economic, or energy security perspective.  It may also make little sense for domestic producers:  if subsidies to both domestic production and imported oil were removed simultaneously, there is a reasonable possibility that the domestic producers would actually become more competitive.


Efforts to reduce carbon emissions should pursue subsidy reform as a first-tier strategy.  This includes both the elimination of certain provisions, and the restructuring of funding for others so that oil consumers, rather than the general taxpayer, pay the cost of the programs.


[1] Doug Koplow and Aaron Martin, Fueling Global Warming:  Federal Subsidies to Oil in the United States, prepared by Industrial Economics, Inc., for Greenpeace, June 1998.  Available on line at http://www.greenpeace.org/~climate/oil/fdsub.html.


[2] American Petroleum Institute, "Fueling Confusion:  Deceptive Greenpeace Study Premised on Flawed Estimates of Subsidy," written by Rayola Dougher, November 1999.


[3] Figures are from:  DeLuchi, Mark et al., A Comparative Analysis of Future Transportation Fuels (Berkeley, CA:  Institute of Transportation Studies, UCAL-Berkeley, Oct. 1987); Sperling, Daniel and M. DeLuchi, "Transportation Energy Futures," Annual Review of Energy, 1989, 14: 375-424; Yim, Man-Sung, J. Evans, and R. Wilson, "Health and Environmental Risks of Energy Systems," (Boston, MA: Harvard School of Public Health, 1991); U.S. General Accounting Office, "Motor Vehicle Regulations:  Regulatory Cost Estimate Could be Improved," July 1992 (RCED-92-100); and Green, David and Jin-Tan Liu, "Automotive Fuel Economy Improvements and Consumers' Surplus," Transportation Res., Vol. 22A, #3, pp. 203-218, 1988.  A more detailed discussion of over- and under-regulation in energy markets can be found in Koplow, Doug, Federal Energy Subsidies:  Energy, Environmental, and Fiscal Impacts, (Washington, DC:  Alliance to Save Energy, 1993), pp. 57-59.


[4] Organisation for Economic Cooperation and Development (OECD), Environmental Effects of Liberalising Trade in Fossil Fuels, Internal Draft, November 25, 1998.  Prepared for OECD by Doug Koplow, Industrial Economics, Inc.


[5] Fueling Global Warming, Appendix Exhibit A-2.


[6] Graham Fuller and Ian Lesser, "Persian Gulf Myths," Foreign Affairs, May/June 1997, pp. 42-52.


[7] Congressional Research Service, Tax Expenditures: Compendium of Background Material on Individual Provisions, Senate Committee on the Budget, December 1996, pp. 228, 233.


[8] See Edwin Rothschild, Oil Imports, Taxpayer Subsidies and the Petroleum Industry, Washington, DC:  Citizen Action, May 1995, pp. 13-15, for a detailed history of the foreign tax credit and oil companies.


[9] Fueling Global Warming, pp. 2-9, 2-10.


[10] Fueling Global Warming, p. 6-5 and Exhibit A-1.


[11] See Glenn Jenkins, "United States Taxation and the Incentive to Develop Foreign Primary Energy Sources," in Gerard Brannon, editor, Studies in Energy Tax Policy, Cambridge, MA:  Ballinger Publishing Company, 1975, pp. 203-250.


[12] Among the "true" subsidies API recognizes are the payments to low income oil consumers through the Low Income Home Energy Assistance Program.  API author Rayola Dougher than states that by "including LIHEAP in their oil subsidy pie, Greenpeace in effect is advocating that the government end such payments designed to help the poor."  What the report actually (and quite clearly) says is that an "increased emphasis on weatherization in the short term could help reduce the need for subsidized oil purchases over the long term."  It is hard to interpret an intention to end the support for heating assistance for the poor from this wording; we find it disturbing that API chose to do so.