• Home
• People
• Links
• Contact Us

Research Areas

• Bioenergy Resource & Engineering Systems
• Energy Analysis
• Global Change & Developing Countries Programs
• Landscape Ecology Regional Analysis
• Society-Technology Interactions

Energy and Oil Security Project

Oil Independence: Realistic Goal or Empty Slogan?
By David L. Greene and Paul N. Leiby
Oak Ridge National Laboratory

March, 2007

There is broad agreement that “U.S. Oil Dependence Remains a Problem” (Fri, Issues, Summer 2003), exposing our economy to disruption and putting our national security at increased risk.  In light of this, it is disturbing to find the notion gaining currency that it is impossible or even undesirable to actually achieve oil independence.  The problem arises from incorrectly defining oil independence as either zero oil imports or zero oil use.  Defined properly, oil independence is definitely achievable but will take time, persistence and strong yet practical policy measures.

A recent report by Resources for the Future declares that “Energy Independence is Unrealistic” (Parry and Anderson, 2005).  The Wall Street Journal reports that energy independence “…may be among the least realistic political slogans in American history.” (Fialka, 2006).  A task force of the Council on Foreign Relations declares the feasibility of achieving energy independence a myth (CFR, 2006).  This self-defeating conviction arises from a misunderstanding of the nature of the oil dependence problem and consequently leads to erroneous notions of what it will take to solve it.  Solving our oil dependence problem will not be easy.  It will take time and will require determination and persistence but it is most definitely achievable.  The good news is that once we seriously commit to solving the problem our energy security will gradually but steadily improve until energy independence is achieved, long before we stop using or importing oil.

Proponents of the pessimistic view define oil independence either as entirely eliminating oil imports or eliminating oil consumption altogether.  Neither definition is a useful basis for formulating an effective energy policy.  Indeed, for three decades U.S. energy policy has been hobbled by the lack of a pragmatic definition of oil independence derived from a sound understanding of the nature of the oil dependence problem and providing a measurable and attainable goal.

What do we mean by oil independence?  The common meaning of the words suffices. Merriam-Webster defines independence as, “not subject to control by others”, and to control is to “exercise restraining or directing influence over.”  To achieve oil independence as a nation we must reach a state in which our nation’s decisions are not subject to the restraining or directing influence of oil producers.  We therefore propose the following as an apt and meaningful (but not yet measurable) definition of oil independence.

“For all conceivable future world oil market conditions, the potential costs of oil dependence to the U.S. economy will be so small that they will have no effect on our economic, military or foreign policies.”

This definition captures the essential idea but it is not measurable.  A measurable definition needs to reflect our uncertainty about future oil market conditions and include a quantitative statement of how much the potential costs of oil dependence must be reduced.  This requires not only rigorous analysis, but judgment.  The following is put forward as a reasonable starting point for discussion and analysis and possibly an adequate, measurable definition.

“The annual economic costs of oil dependence will be less than 1% of U.S. GDP, with 95% probability, by 2030.”

Oil dependence is fundamentally an economic problem, albeit one with extremely important military and foreign policy implications.  However, the military and foreign policy dimensions arise almost entirely from the economic problem.  If there were no economic consequences of oil dependence the foreign policy and military implications would be minimal.  This seems self-evident with one exception: our military itself is dependent on oil.  Therefore, a cut-off in oil supply to our military would cripple our national defense.  Fortunately, this concern can be satisfied by a strategic reserve.  In fiscal year 2005, the U.S. Department of Defense consumed 133 million barrels of petroleum (Dimotakis, et al., 2006).  The U.S. Strategic Petroleum Reserve stands at 688 million barrels, enough to supply the entire requirements of the DoD for five years at the 2005 rate of use.  If the rest of the economy were unconcerned about the global supply of oil, this would seem to be an adequate military reserve, recognizing that there are important logistical issues that must be addressed. 

What are the economic costs of oil dependence?  To answer this question requires a clear understanding of the nature of our oil dependence problem. No one has summed up the oil dependence problem as tersely and precisely as Morris Adelman.

“The real problem we face over oil dates from after 1970: a strong but clumsy monopoly of mostly Middle Eastern exporters cooperating as OPEC.”  (Adelman, 2004, p. 16)

Over three decades ago two events permanently transformed the world oil market.  In 1970, crude oil production from the United States, until that year the world’s largest oil producer, peaked at a level that has never since been exceeded.  Shortly afterward, late in 1973, the Arab members of the Organization of Petroleum Exporting Countries embargoed oil exports to the United States and other countries supporting Israel in the 1973 October War.  World oil prices tripled, dramatically demonstrating the market power of the new cartel (figure 1).

What gives OPEC the power to affect world oil prices, and to cause or capitalize on price shocks?  Economic theory provides valuable guidance on this point.  The theory of partial monopoly (monopolists controlling a significant share but not all of a market), developed over half a century ago, says that market power depends chiefly on three factors: 1) the cartel’s market share, 2) the ability of world oil demand to respond to higher oil prices and, 3) the ability of non-OPEC oil suppliers to respond to higher oil prices.  In a dynamic market, the cartel’s market power is magnified when demand is growing and rest-of-world supply is shrinking. 

OPEC’s ability to secure a large share of the oil market is a consequence of the fact that OPEC members hold two thirds to three fourths of the world’s proven oil reserves.  The inability of world oil demand to respond strongly to higher oil prices is a result of the world transportation system’s strong preference for petroleum fuels combined with the continuing growth in demand for motorized mobility in both the developed and developing economies.  Non-OPEC oil producers cannot effectively replace oil supplies lost during a supply disruption (whether caused by OPEC or not) because of their more limited oil reserves and the fact that reserves outside of the Persian Gulf are generally more difficult and expensive to produce. 

A final, critically important fact is that the ability of oil consumers and non-OPEC producers to respond quickly (say within a year) to higher oil prices is roughly an order of magnitude smaller than their ability to respond over a period of many years.  This implies that the only way OPEC can sustain very high prices, such as those seen from 1979-85, is by continually cutting back on production.  That, of course, is precisely what OPEC and especially Saudi Arabia did over that period.   In the short run OPEC can influence the market through production quotas and production cutbacks.  In the long run OPEC members influence the market through their investments in production capacity.  The fact that OPEC members now have little spare capacity is the result of their investment strategy.  More money can be made by not fully accommodating the world’s growing demand for oil than by adding production capacity to accommodate it.

One should not expect OPEC to behave perfectly in accord with simple economic theory.  Adelman refers to OPEC as a “clumsy” cartel because it is made up of sovereign states who do not always agree on what is in their best interest and whose cooperation is far from perfect.  But the fact that OPEC may wield its market power clumsily does not imply that it has no power or that it is not really a monopolistic cartel.

If the problem of oil dependence is the result of a “clumsy monopoly” of oil producers, then the economic costs of oil dependence are a consequence of the higher price levels and price shocks caused by or capitalized on by those producers.  This is an important point. The market failure at the heart of the oil dependence problem is one of market power.  The costs of oil dependence are not external costs and neither a tax on oil nor a tax on imported oil is an adequate solution to the problem (although either would help). 

The U.S. economy suffers three kinds of economic costs as a result of its oil dependence and the actions of the OPEC cartel: 1) transfer of wealth, 2) loss of potential to produce and, 3) disruption losses.  When producers raise prices above competitive market levels through the use of market power, there is a transfer of wealth from oil consumers to oil producers.  This is often euphemistically referred to as a change in the terms of trade.  The fact that the change in terms of trade results from the use of market power is key, because the wealth transferred to OPEC producers is therefore pure surplus, i.e., profit.  It need not be reinvested in producing more oil.  It can be used to fund health care or terrorism, economic development or nuclear weapons programs, education or repression.  The transfer of wealth is not an economic loss from the perspective of the global economy.  The wealth lost by the U.S. economy is gained by oil producing economies.  From the parochial perspective of the U.S. economy, however, it is most definitely a cost.  Wealth transfer does not necessarily reduce U.S. GDP.  It is a change in ownership, not output.  For example, an oil producer might choose to buy Iowa farmland with his oil profits.  The same amount of corn would be produced but the ownership of the corn (as well as the land used to produce it) would change.  U.S. citizens become poorer, oil producers become richer.

In theory, the size of the wealth transfer loss is simply the quantity of oil imported times the difference between the actual market price of oil and what the price would have been in a competitive market.  Since there is no competitive market, the latter price cannot be observed.  Inferring a competitive price inevitably involves some uncertainty but it can be done well enough to get a reasonable estimate of the wealth transfer cost.  For example, in 2005 wealth transfer probably cost the U.S. economy between $100 billion and $150 billion dollars.

The two other kinds of economic costs do reduce GDP.  When the price of any fundamental input to production increases the potential to produce given the same endowments of capital and labor shrinks.  Higher prices signal scarcity to the marketplace.  It matters not whether the scarcity is physical (less oil in the ground) or the result of market power.  Either way, higher oil prices mean it’s a tougher world in which to make a living.  The size of this loss of GDP depends on how much oil our economy consumes, how much it produces, and on the sensitivities of domestic oil demand and domestic supply to the price of oil.  Estimating these costs requires more assumptions and information about the nature of U.S. oil supply and demand but estimates can be made.  A reasonable range for the potential GDP costs incurred in 2005 is from $10 billion to $50 billion.

Unlike wealth transfer losses and potential GDP losses, disruption costs occur only when there is a sudden, unexpected jump in the price of oil.  Disruption costs are caused by temporary dislocations in the economy as a result of an oil price shock.  Complex reallocations of labor and capital have been observed throughout the U.S. economy in response to oil price shocks.  The size of disruption losses depends on the flexibility of production, the ability of wages and prices to adjust quickly, the expectations of consumers and producers, monetary policy and, of course, the importance of oil to the economy.  All else equal, however, disruption losses should be roughly proportional to total expenditures on petroleum relative to the size of the GDP.  While there is greater uncertainty about the magnitude of these costs, a plausible range for 2005 is from $50 billion to $170 billion. 

A reasonable range for total U.S. oil dependence costs in 2005 is $175 to $330 billion. The midpoint of this range is about 2.25% of U.S. GDP.  The average cost of a barrel of oil was $50 in 2005.  With oil prices averaging $60 through October 2006, costs for 2006 are certain to be larger.

What are future oil dependence costs likely to be?  There is one thing we can be certain of.  Any single prediction of how the world oil market will evolve through 2030 will be wrong.  Therefore, the most appropriate way to describe future oil market costs is probabilistically.  For this purpose, we have developed a simulation model that allows us to estimate oil dependence costs in many possible futures (Greene, et al., 2006).  We calibrate the model to Energy Information Administration projections for High, Low and Reference Oil Price projections and allow it to pick randomly among them (EIA, 2006).  The model also simulates oil supply disruptions of variable sizes and durations, based on the characteristics of past supply disruptions, and it estimates the resulting oil price shocks.  It incorporates the effect of reduced U.S. oil demand on world oil prices.  It also allows for uncertainty in other critical parameters, such as the price elasticities of U.S. and world oil supply and demand.  Based on a simulation of 10,000 alternative futures, the distribution of oil dependence costs shown in figure 2 was generated. 

The simulations suggests that, with 95% confidence (conditional on the myriad assumptions embedded in the analysis) oil dependence costs in 2030 will be less than 3.5% of U.S. GDP.  The solid black line shows the path of expected cost, the dark gray shaded area indicates +/- one standard deviation, while the light gray shaded areas indicate the 5th and 95th percentiles.  With a projected $23 trillion GDP in 2030, 3.5% is approximately $800 billion per year.  The simulations indicate a 50/50 chance that costs will exceed $400 billion per year.  Figure 2 also shows estimated historical costs.  The uncertainty bands on these estimates are smaller because historical oil market conditions are known.

The remaining uncertainty has to do with key parameters of the model.  We do not claim that these estimates are definitive but rather that they are reasonable and useful for illustrating the concept of a measurable definition of oil independence.  Clearly, “business as usual” as reflected in the EIA’s forecasts will not get us there.

How can we reduce future oil dependence costs to an acceptable level?  To become oil independent we must reduce our exposure to high oil prices and oil price shocks and undermine the cartel’s market power.  From the theory of market power and the description of oil costs it is clear that this means reducing the oil intensity of our economy and increasing our economy’s ability to substitute energy efficiency and alternative energy sources for oil, as well increasing our ability to produce oil from conventional and unconventional resources.  Reducing our oil consumption will reduce world oil demand, putting pressure on the cartel’s market share.  To the extent we can do this by advancing technology, we may be able to increase the price responsiveness of supply and demand and undermine the cartel’s market power.

Some might say this sounds easy but isn’t really possible.  But there is an historical precedent.  From 1973 to 1985, higher oil prices working in increasingly deregulated markets, strong fuel economy standards that required a near doubling of automotive fuel economy, fuel switching in non-transport sectors, and increased domestic production from the North Slope of Alaska (as well as similar actions on the part of many other oil consuming nations) produced a steep and steady decline in OPEC’s market share and the collapse of world oil prices in 1986.  This collapse of OPEC’s market power was followed by a sustained period of relatively low oil prices and sufficient supply, and low oil dependence costs that lasted for more than 10 years, i.e., a decade of energy independence.  Unfortunately, low prices and abundant supplies persuaded us that the oil dependence problem had been solved once and for all, or that there really never had been a problem.

Addressing oil dependence will require a comprehensive, robust and sustained policy strategy.  Rather than discuss all the possible policies that might be used to achieve these ends, we choose as an illustration the long-term strategy proposed by the non-partisan National Commission on Energy Policy.  The Commission attempted to insure that the policies it proposed were both economically and politically practical.  They explained their policy criteria as follows.

“In choosing among a large number of potential policy options, the Commission applied several general criteria, including economic efficiency; cost-effectiveness and consumer impacts; ability to provide appropriate incentives for future action; flexibility for adjustment in response to further experience, new information, and changed conditions; equity; political viability; and ease of implementation, monitoring, and measurement.”  (NCEP, 2005, p. viii)

The NCEP strategy is wide-ranging but still incomplete; e.g., it recommends raising fuel economy standards but offers several possible levels without recommending one, and it does not fully consider non-transportation oil uses in the buildings and industrial sectors.  Where necessary, we have augmented the NCEP proposals in a way that we hope retains the spirit of their strategy.  For example, we assume that passenger car and light truck fuel economy will increase by 75% by 2030.  We assume that electricity and natural gas will be substituted for distillate oil in home heating.  Of course, the NCEP is not responsible for the modifications we have made to their strategy.  We invite the reader to draw his own conclusions about the feasibility of the changes in U.S. oil supply and demand in 2030 shown in Table 1.  This illustrative case is intended neither as endorsement nor criticism of the NCEP strategy. 

The NCEP strategy does not come close to eliminating U.S. oil use or oil imports.  Petroleum use is reduced by only 26% relative to the U.S. Energy Information Administration’s Reference Case projection.  Greenhouse gas emissions are reduced by roughly the same percentage.  The most significant factors are increased passenger car and light truck fuel economy and increased use of biofuels.  Supply is increased by 29% over the Reference Case by petroleum production from the Arctic National Wildlife Refuge and greater use of Coal-to-Liquids technology to produce synthetic petroleum fuels.  To be compatible with the goal of protecting the global climate, the NCEP advocates producing liquids from coal via integrated gasification combined cycle technology and sequestering the carbon emissions.  This should result in roughly the same fuel cycle carbon emissions as combustion of conventional petroleum fuels.

TABLE 1: Estimated Changes in U.S. Oil Supply and Demand in 2030 for the Modified NCEP Oil Independence Strategy (Millions of Barrels per Day)

In the simulation, OPEC is allowed two strategies: 1) no response, maintain the production schedule of the EIA projection and, 2) reduce production to maintain the oil price path of the EIA projection.  An extremely important result of the simulation is that OPEC’s choice of strategy has a negligible impact on the distribution of future oil dependence costs for the United States.  The benefits of reducing oil dependence do not depend on how OPEC responds.

Implementing the modified NCEP strategy steadily reduces both the expected level and the variability of oil dependence costs.  By 2030 the expected annual oil dependence costs are less than 1% of GDP but, unfortunately, the 95th percentile is 1.5% of GDP and so the oil independence goal is not met (1% is the 69th percentile).  More is needed.  The gap might be filled by a tax on carbon or on petroleum.  It might also be filled by advanced technologies.  Perhaps hydrogen fuel cell or internal combustion engine vehicles will be market-ready well before 2030.  Hydrogen can be made from a wide variety of energy resources but petroleum is not likely to be one of them.  Plug-in hybrid vehicles that can take some energy from the grid could also play a role given advances in battery technology.  Once again, electricity is generated from a wide array of energy resources but only 3% of U.S. electricity is generated using petroleum.  Advances in heavy truck or aircraft technology are needed to permit greater efficiency improvements in these rapidly growing components of the transportation sector.  Advances in conversion technology might expand the potential for biofuels.  Alternative feedstocks for petrochemicals and new materials to replace asphalt in road building are other significant areas of opportunity.  Regardless of what other polices are adopted, continued technological progress will be essential to maintaining U.S. oil independence beyond 2030.

Oil independence is about reducing our vulnerability to oil dependence costs to an acceptable level.  It is not about eliminating oil imports or eliminating oil use.  From the admittedly preliminary analysis presented here, a one-third reduction in oil use and a one-third increase in oil supply over projected levels would probably get the job done by 2030.  Achieving oil independence will take time.    It might be possible to achieve oil independence before 2030 through other policy actions and with technological advances not included in the modified NCEP strategy.  However, it is not possible to achieve oil independence next year or even in five years.  The good news is that once we get started we will see steady improvement all along the way.  Achieving oil independence will require strong but practical policies.  Improving light-duty vehicle fuel economy by 75% even by 2030 will require powerful fuel economy policy, given the stagnation of fuel economy over the past two decades.  Strong policies will be needed in other sectors as well.  Maintaining oil independence will require sustained effort.  History teaches that if we relent, believing that the problem has been solved, our oil dependence problem is likely to return.

If we accept the judgment that achieving oil independence is unrealistic, we will be defeated before we begin.  By adopting a pragmatic definition of energy independence, grounded in a scientific understanding of our oil dependence problem, we take the first step towards achieving oil independence.

***

References

1. Adelman, M.A., 2004. “The Real Oil Problem”, Regulation, Spring 2004.
2. Auerswald, P.E., 2006. “The Myth of Energy Insecurity”, Issues in Science and Technology, https://www.issues.org/22.4/auerswald.html , Summer 2006.
3. Council on Foreign Relations (CFR), 2006.  “National Security Consequences of U.S. Oil Dependency”, Report of an Independent Task Force, New York, New York.
4. Dimotakis, P., R. Grober and N. Lewis, 2006. “Reducing DoD Fossil-Fuel Dependence”, The MITRE Corporation, McLean, Virginia.
5. Energy Information Administration (EIA), 2006. Annual Energy Outlook 2006, DOE/EIA-0383(2006), Washington, D.C.
6. Fialka, J.J., 2006. “Energy Independence: A Dry Hole”, Wall Street Journal, July 5, 2006.
7. Fri, R., 2003. “U.S. Oil Dependence Remains a Problem”, Issues in Science and Technology, vol. XIX, no. 4, Summer, The National Academies.
8. Greene, D.L., D.W. Jones and P.N. Leiby, 1998. “The outlook for US oil dependence”, Energy Policy, vol. 26, no. 1, pp. 55-69.
9. Greene, D.L., P.N. Leiby, P.D. Patterson, S.E. Plotkin, M.K. Singh, 2006. “Oil Independence: Achievable National Goal or Empty Slogan?”, draft report, Oak Ridge National Laboratory, Oak Ridge, Tennessee, July 2006.
10. National Commission on Energy Policy (NCEP), 2004, Ending the Energy Stalemate, National Commission on Energy Policy, Washington, D.C.
11. Parry, I.W.H. and J.W. Anderson, 2005. “Energy Independence is Unrealistic”, Resources, no. 156, pp. 11-15. Winter, Resources for the Future, Washington, D.C.