全球智库信息集成服务系统

Regulation and government funding of R&D are necessary but not sufficient to slow climate change. To transform energy use and supply across the economy, greenhouse gas (GHG) emissions will have to be priced and the power of the market brought to bear.

Oil is the most demanding fossil fuel in this regard. It is largely used for a single purpose—transportation—for which it has few substitutes. It is the most diverse of all fuels—chemically, geologically, and geographically. And it is heavily capitalized and the most traded global commodity. Each characteristic makes the design of an effective and fair tax particularly difficult.

Moreover, the United States faces a hydrocarbon landscape transformed by new, unconventional oils. The long-standing expectation of a gradual, shortage-driven shift to clean fuels has been replaced by the need for a swift transformation in the face of abundant supply. National policy making has not begun to catch up. A new smart tax design offers a way to do so.

From Blunt Tax to Smart Tax

• Traditional tax designs are blunt instruments that treat all oils alike and tax only consumption of end products.
• Blunt taxes leave major unconventional oil emissions untaxed, treat dirtier oils more favorably than cleaner ones, provide no incentive for technological innovation, and offer no incentive for refiners to consider climate in determining product mix.
• For the first time it is possible to quantify the amount and profile of GHG emissions from all oils throughout the supply chain using the Oil-Climate Index developed by researchers at the Carnegie Endowment for International Peace, Stanford University, and the University of Calgary, allowing replacement of blunt taxes with a smart tax design.
• A smart tax differentiates among the chemical entities called “oil,” accounts for GHG emissions along the entire oil supply chain, and includes byproducts that do not fuel transport, thereby correcting the shortcomings of a blunt tax.

Moving Ahead in the United States

• A smart tax can be simply administered at the refinery level, allowing refiners to pass the charges up the supply chain to producers and crude transporters and down to product transporters and consumers in a mechanism comparable to netback pricing.
• A means of applying border tax adjustments will be necessary. A range of new data must be standardized, collected, and made available, and a lead government agency made responsible for the task.
• Political barriers to enactment can be lowered by making the smart tax revenue neutral with a formula for returning revenues to the economy that is simple, transparent to the public, and equally attractive to both major political parties.

Oil: The Toughest Climate Challenge

At the Paris climate summit in December 2015, 195 nations committed themselves to transform the world to a low-carbon economy through steep reductions in greenhouse gas emissions over the coming few decades.¹ Keeping the world’s average warming to the goal of 2 degrees or less above pre-industrial levels will require sweeping change across rich and poor economies in a very short space of time.² The task is enormous and will entail transforming virtually all areas of energy use and supply.

Yet, in Paris, countries’ pledges to reduce their emissions (the so-called intended nationally determined contributions, or INDCs³) largely ignored the world’s largest single source of fossil fuel energy and second-largest source of greenhouse gas (GHG) emissions—namely, oil. The U.S. INDC, for example, mentioned oil just once—in relation to reducing methane emissions from oil and gas operations—even though oil accounted for 36 percent of U.S. energy supply in 2013.⁴ The overall agreement in Paris explicitly exempted emissions from marine and air transport—both fueled almost entirely by oil.⁵

There are several reasons why oil is such a difficult challenge in the global effort to cut GHG emissions. Foremost among them is that oil is largely used for a single purpose—transportation—and in that role it today has far fewer alternatives than do the fuels used for heating, cooling, and generating electricity; in those cases, nuclear, wind, solar, biomass, and natural gas can all be used instead of coal. (The lack of alternative liquid fuels to oil would, of course, change if the transport sector came to rely heavily on electricity.)

Also, oil is massively capitalized and is the world’s most traded commodity.⁶ The enormous sunk costs in oil production and the global web of extraction, transport, refining, and use make it hard for any single country to push this huge, integrated industry in new directions.

The price signal, important to controlling emissions from all sources of energy, is vital in the case of oil.

Finally, oil is by far the most diverse of all fuels—ranging from light mixtures of liquids and gas to solid, heavy bitumen and kerogen, differing profoundly in their chemistries and in the refined products they produce. Oils come from a wide range of regions, are buried in very different kinds of geologic deposits, and require an array of complex technologies to extract and refine them before reaching consumers. This great diversity limits the applicability of regulatory and technological approaches to cutting emissions from oil.

Each of these characteristics makes the price signal, important to controlling emissions from all sources of energy, vital in the case of oil. Regulation and the development of new technology can be effective in the absence of a clear price signal, but both are far more effective if they work in step with market forces, rather than separate from or even against them. Government can pay for R&D, for example, pushing new technologies into the marketplace, but the flow of innovation will be stronger and faster if demand from the market makes commercial investment in new technology attractive, pulling those technologies into use.

The price signal most relevant to slowing climate change is a tax or fee that puts a price on the greenhouse gases that are emitted when fossil fuels are consumed. This is commonly known as a carbon tax even though carbon dioxide is not the only greenhouse gas involved (see box). Economists from across the ideological spectrum virtually unanimously hold the view that pricing carbon in this way is the most effective, most economically efficient, and least costly way to achieve the needed transformation. Since mid-2015, the urgent need for such pricing has been voiced by, among many others, the heads of the World Bank and the International Monetary Fund, the governor of the Bank of England, and the CEOs of six of the world’s major oil companies.⁷

A carbon tax is best designed expressly for oil because crude oil and petroleum products are so different from other energy sources—in their chemical diversity, their long international supply chains, their processing complexities, the value that causes them to trade at an economic premium, and the lack of ready substitutes. This is not to suggest that pricing carbon in other fuels is not also called for, only that oil presents a suite of unique challenges and opportunities.

The question to which this analysis is addressed is whether a workable carbon tax or fee can be designed that will account for oil’s chemical, geologic, economic, and geopolitical features. Ultimately, of course, the challenge is a global one, but as the Paris approach wisely recognized, the steps along the path to a safe climate will be designed and introduced country by country. Accordingly, the focus here is on the United States, one of the world’s largest users, producers, and refiners of oil; the second-largest emitter of greenhouse gases; and, at least potentially, a natural leader in the effort to keep the planet’s climate within safe bounds.

Oil Policy in the United States Today

For nearly a half century, Americans have believed that affordable oil would become increasingly scarce and have worried about the degree of U.S. reliance on foreign sources of it.⁸ The call for energy independence—a feature of campaign rhetoric for decades—has not been about energy in general, but about oil. Since the Organization of the Petroleum Exporting Countries (OPEC) oil embargo of 1973, policymakers have accordingly focused on safeguarding the security of the U.S. supply, employing an array of strategies including automobile mileage standards to reduce demand and an ethanol mandate to provide a source of domestically produced supply. As environmental and climate considerations rose in importance, regulatory policies such as low-carbon fuel standards and zero-emission vehicle requirements were adopted to guide a gradual transition to clean, low-carbon, alternative fuels.

Instead of running out of oil, however, the new millennium ushered in a paradigm-shattering change. Bolstered by surging demand, high prices, and technological breakthroughs, unconventional oils began to be produced in large volumes, including oil sands, hydraulically fractured tight oils, extra-heavy oils, ultra-deep oils, ultra-light condensates, and more. Faster than anyone believed possible, the United States left behind decades of worry over its dependence on oil imports and became a major oil producer and exporter almost overnight.

The United States became a major oil producer and exporter almost overnight. Neither public understanding nor national policy has begun to catch up to this profound change.

America is now confronting explosive production of an array of plentiful, little understood, still evolving, and, in some cases, environmentally damaging new oils. Old mind-sets, backed by insufficient information and outdated rules and governance structures, are struggling to adapt to a drastically altered hydrocarbon landscape. Security of supply has given way to the imperative of controlling climate change. The expectation of a gradual, shortage-driven transformation to clean fuels has been replaced by the wholly different challenge presented by the need for a swift transformation in the face of abundant supply. Neither public understanding nor national policy has begun to catch up to this profound change.

The Climate Significance of the New Oils

The oil supply chain is responsible for significant GHG emissions beyond the combustion of its end-use petroleum products. Extracting, processing, refining, and transporting oil is energy-intensive; large energy inputs are needed, and energy is wasted in operations. When the differences in energy inputs and waste in the supply chain and the variability of types of oil are accounted for, oil-to-fuel pathways differ dramatically in their full GHG emission profiles.

The Oil-Climate Index (OCI)—developed by researchers from the Carnegie Endowment for International Peace, Stanford University, and the University of Calgary—is able to quantify, for the first time, the differences in both the amount and profile of GHG emissions produced throughout the supply chain.⁹ Some oils generate higher emissions in their extraction and others during refining. Still others have a more GHG-intensive slate of products. In the OCI’s first sampling of oils, overall emissions vary by nearly a factor of two.

The long-standing assumptions that oil is oil and that most or all of its GHG emissions occur when its end products are burned are simply wrong, as the examples shown in figure 1 make abundantly clear. Yet, both of these assumptions have shaped the design of virtually all oil taxes proposed over the past several decades. We call these measures blunt taxes. They are of two types. One type treats all barrels of oil alike through a flat fee based most commonly on volume (a per barrel tax) or energy content (a BTU tax). This made sense when barrels of crude oil were very similar. It no longer makes sense when a barrel of crude could be anything from ultra-light condensate to semisolid bitumen. The other type of blunt tax addresses only the end products of oil use—most often gasoline (a gas tax). This puts the entire burden of reducing GHG emissions from oil on consumers, thereby removing any incentive for producers, refiners, and transporters to reduce their own emissions. Targeting only consumers of end products would indeed dampen oil demand, but would do so at a needlessly high cost and with many unwanted effects.

Many oils have significant amounts of associated natural gas trapped with them in reservoirs. If this gas is not collected and sold upstream but is instead flared or vented, GHG emissions from upstream production make up a large share of each barrel’s much larger total emissions. The particular light, high-flare oil example shown in figure 1 produces 35 percent more emissions per barrel than light oil that is not flared, even though the high-flare oil generates emission credits when some of the associated gas is used for export and to produce electricity consumed in the oil’s production.

Depleted oil fields have generally been producing for several decades. The oil extracted from them is often heavier, more viscous, and more water-saturated than oil from younger fields. Pressure at the wellhead is often diminished in a depleted field, hampering its flow. Depleted oils therefore often require large energy inputs to produce steam, which reduces the oil’s viscosity to help force it out of the ground, and to run pumps to induce its flow and remove entrained water. The carbon-heavy, depleted oil requires the production and addition of hydrogen to convert the carbon contained in heavier oils into valuable petroleum products, which increases the emission intensity of midstream refining.

Climate policies must be designed to accommodate differences in oil resources and engineering processes with their attendant variance in GHG emissions.

The most emissions-intensive oil depicted in the figure, an extra-heavy oil, requires a physical and chemical transformation to make it flow more like a conventional crude and convert it into petroleum products. Different pathways for such conversion are currently in use and under development. For example, upgrading is an energy-intensive upstream process that removes the oil’s excess carbon to produce a synthetic crude oil along with the byproduct, petroleum coke (or petcoke, a solid fuel that is high in sulfur and other contaminants and used for power generation). When petcoke is combusted, its climate and air-quality impacts can be even greater than coal’s. Another such process involves diluting the extra-heavy oil with lighter hydrocarbons to make the oil flow to a refinery where hydrogen is added and excess carbon is removed.

Depending on future oil demand, these already large differences among oil types could widen substantially as advances in technology create ways to extract and transform new sources of hydrocarbons. Climate policies must be designed to accommodate these differences in oil resources and engineering processes with their attendant variance in GHG emissions.

A Smart Tax Versus Blunt Ones

Because of these large differences in the quantity and sources of emissions in the supply chain, a smart tax on oil must have the following three characteristics:

• It must differentiate among the very different chemical entities that today all go by the name “a barrel of oil.”
• It must account for GHG emissions along the entire oil supply chain: production, refining, transport, and end product use.
• It must include the emissions from oil byproducts that do not fuel the transport sector, the most important of which is the bottom-of-the-barrel material known as petcoke (which is sold, exported, and burned, but is classified and is often used as a solid fuel to blend with or displace coal).

Blunt taxes and a smart tax differ dramatically in economic efficiency, equity, and effectiveness in curbing GHG emissions. Specifically:

A Blunt Tax Leaves Major Emissions Untaxed

Figure 2 reveals the dimensions of this shortcoming. The OCI results of four U.S. oils (including both flaring and no-flaring scenarios in Texas’s Eagle Ford condensate formation) are used as proxies to illustrate the range of possibilities. A blunt tax applied to only combusted end products leaves 9–38 percent of emissions untaxed. In the case of Midway Sunset—a depleted, heavy oil from California—over one-half of the emissions are untaxed if petcoke is not counted.

As a thought experiment, if the nearly 5 billion barrels of crude oil and condensates produced in the United States in 2015 were allocated in equal shares to the four cases in figure 2, an estimated 730 million metric tons (MMT) of carbon dioxide equivalent emissions would go untaxed annually under a blunt tax applied only to end products. This is not a minor omission: it represents an estimated one-third of currently reported emissions from the U.S. oil sector.¹⁰

Blunt Taxes Send a Perverse Market Signal

If all oils were treated the same way, or if the tax were assessed only on end products, then the most polluting—in the climate sense—oils would receive the most favorable tax treatment because more of their emissions would go untaxed. Figure 3 illustrates how a flat tax and a tax on only end products consumers use send a distorted market signal for lower- versus higher-emitting oils. At a carbon price of $20 per metric ton of carbon dioxide equivalent emissions, a smart tax would assess $8 on a barrel of Eagle Ford ultra-light condensate, not flared—less than the other sample oils. A blunt, per barrel tax would charge Eagle Ford $2 a barrel more—higher relative to other sample oils. Oils with even larger total climate footprints, including extra-heavy oils, oil sands, and other future oils, would introduce even greater market distortion.¹¹

Rather than encourage the production of dirtier oils, a tax designed to slow climate change should indicate to the market that the GHG emissions of these oils must be reduced, bringing them into line with other oils. Or it should send a clear sign that these oil types should be developed last, if at all—hence discouraged rather than favored

A Blunt Tax Creates No Incentive for Technological Innovation in the Supply Chain—Another Perverse Effect

Oil production and refining are highly competitive activities, often operating with small profit margins. A small change in price can make a big difference in operations. Innovation and greater efficiencies are always being sought in this constantly evolving industry.

A tax should encourage and reward technological advances that reduce GHG emissions. An example would be the incentive to develop cost-effective processes for collecting and using any associated gas that is now flared. Figure 2 shows that, if only end product combustion is taxed, flaring its associated natural gas improves the Eagle Ford condensate’s competitive position compared to safely managing its associated gas by leaving 37 percent of total emissions untaxed rather than 9 percent.

Other examples of innovations for dealing with depleted and extra-heavy oils are illustrated in the Oil-Climate Index. For example, the same California Midway Sunset oil illustrated in figure 2 requires process heat to generate the steam needed for its production. The necessary energy could come from a fossil fuel, or it could be generated from concentrated solar energy. A blunt oil tax would penalize that step forward. Discouraging innovation in this way is, of course, the very last thing a climate control policy should do.

A Blunt Tax Creates No Incentive for Changes in the Product Slate

Important petroleum products—including petcoke, residual fuel oils, petrochemical feedstocks, and asphalt—fall outside the traditional boundaries for liquid transport fuels. But the emissions from their use are equally relevant to what matters: the amount of GHGs in the atmosphere.

Petcoke is the most important of these ancillary petroleum products. If this very high carbon, solid, oil byproduct were not included, approximately 25 percent of total emissions from extra-heavy oils would go untaxed, as highlighted in figure 4.

A smart carbon tax on oil must give refiners reasons to consider climate in determining their product mix. A smart tax creates the necessary incentives; a blunt tax does not.

A Blunt Tax Imposed Only on End Products Requires a Higher Tax Rate Than a Smart Tax for Comparable Emission Reductions

Because oil demand is relatively inelastic, with few ready substitutes, a blunt tax that charges only end product consumers would likely necessitate a much higher tax rate in order to achieve a given level of emission reduction. Moreover, assessing the tax on all stakeholders in the supply chain—producers, transporters, refiners, and consumers—greatly expands the opportunities for cost-effective emission reductions.

Administering a Smart Tax

How, in practical terms, could a smart tax, levied throughout the whole supply chain, be collected? Would it require millions of collection points? In fact, there is an elegantly simple way to manage such a tax. It could be levied on the very small number of refiners, allowing them to pass the charges up the supply chain to producers and crude transporters, and down the chain to product transporters and consumers. There were just 137 refineries, run by 59 corporate entities, in the United States processing 18 million barrels of oil a day as of January 2015. Just seventeen of these companies accounted for 85 percent of all refining capacity.¹²

Refiners know oil. In order to process crude into products, they need to accurately characterize it, both chemically and physically. And because they can procure crude oils from a wide array of producers, refiners have leverage over those from whom they choose to buy. The price they charge is based on the crude’s quality and its transport costs (known as netback pricing).

This amounts to the same variable pricing mechanism that would be applied through a smart tax. Refiners can include in netback pricing the upstream carbon tax—the amount determined by the GHG emissions of upstream operations including, for example, the flaring and venting of associated gas. Imported oil can be managed through the same mechanism. Foreign producers would either have to provide the information needed for U.S. refiners to determine the amount of the upstream tax or refiners would take their business elsewhere. Refiners would be directly responsible for the portion of GHG emissions they generate in their operations. Finally, they would pass down to consumers that portion of the tax attributed to combusting each end product.

The administrative burdens of a smart carbon tax would pale in comparison to many other common taxes.

By contrast, a tax assessed on producers at the wellhead (point of extraction) could involve more than 1.5 million unique assessment points in the United States alone, and would have no plausible advantages over a tax assessed on the small number of refiners.¹³ Moreover, it would have to cover in some way the nearly 8 million barrels a day of imported crude, which would involve multiple foreign taxpayers, many of whom are not prepared to adopt a carbon tax.¹⁴

Assessing the full tax on producers is far less feasible than placing it on U.S. refiners. To do so, assumptions will have to be made with regard to the refinery in which the oil is destined to be processed, the particular refining units through which it will be passed, and the final slate of products that will be produced. In cases where certain oils have consistently been sent to a single, easily characterized refinery, reasonable life-cycle emission extrapolations can be made at the point of extraction. But in most cases, these assumptions are likely to be rough and inaccurate. A tax is then vulnerable to exploitation and arbitrage by those who understand how to use such assumptions to their advantage, to the ultimate detriment of the policy’s intended purpose.

In any case, the administrative burdens of a smart carbon tax would pale in comparison to many other common taxes. The U.S. corporate income tax, for example, in addition to taxing an economic “good” (corporate income) rather than a “bad” (pollution), also suffers from administrative complexity. In 2010, there were more than 1.8 million corporations collectively accountable for $191 billion in corporate income taxes. A carbon tax, regardless of its incidence point, could bring in a sizeable portion of this sum via dramatically fewer taxpayers.¹⁵

International Considerations

Crude oil and petroleum products are highly mobile commodities. Crude produced in one country can be transported to other countries to be refined, with those products transported again, over long distances, to be consumed in many other countries. Thus, until or unless a uniform price on carbon is adopted worldwide, countries that choose to use the market will also have to consider the use of border tax adjustments in order to avoid a loss of competitiveness if oil production and/or manufacturing enterprises move to countries that do not charge a carbon tax. These adjustments at the border are also needed to prevent some emissions from escaping taxation entirely.

When it comes to meeting the emission reduction pledges made at the Paris summit, a country’s responsibility to cut GHGs will depend on where along the supply chain it takes control of crude oil or petroleum products. In the parsing of oil sector GHGs, multiple INDCs could be involved depending on the country in which an oil and its products are produced, refined, and consumed. As countries report on their current INDCs and revise future INDCs, they will want to think strategically about what types of domestic and foreign oil resources they choose to utilize and how they account for their particular GHG emission responsibilities.

The case of Canadian oil sands exported to a U.S. refinery offers an example of the implications of this approach to allocation. There are many pathways to convert these bituminous resources into usable products, and more are under development. Depending on the method chosen, Canada’s obligation could change dramatically, as illustrated in figure 5. In the case where the United States has a carbon tax that properly accounts for all emissions and Canada does not, different oil sands pathways are materially different in financial terms. If Canada opted to upgrade its oil sands, stripping off the excess carbon in Alberta, it would then ship synthetic crude oil to the United States and would retain possession of the petcoke byproduct. In this case, Canadian producers would pay a carbon tax of $6 per barrel, and U.S. refiners would pay $9 per barrel, a $3 difference. However, if Canadian producers shipped diluted bitumen (dilbit) to U.S. refineries—which then converted the bitumen a