3. Administration of a US Carbon Tax

Ian Parry
Published Date:
March 2015
  • ShareShare
Show Summary Details
Jack Calder

Key Messages for Policymakers

  • A clear and preferably simple policy objective is important for both design and effective administration of a carbon tax.

  • Emissions from fossil fuel combustion are simpler to tax than other greenhouse gas emissions, and are by far the largest source of emissions. It is best, therefore, to concentrate on taxing these first.

  • If the policy objective is comprehensive taxation of those emissions, upstream taxation based on fossil fuel sales has significant administrative advantages over downstream taxation based on measured emissions.

  • It may be difficult to identify a single common point of upstream taxation for all fossil fuel sales, but the Internal Revenue Service can develop practical regulatory solutions, using its experience of administering excise tax on petroleum products and coal.

  • There is no need to re-invent the wheel for administration of an upstream carbon tax. It should use the established procedural framework for existing taxes, particularly excise.

  • If carbon capture and storage and measurement of carbon “sunk” in forestry become practicable, a simple system of government payments for carbon stored and sunk will be easier to administer than carbon tax repayments and credits.

  • If the policy objective is to exempt energy-intensive trade-exposed businesses, administration will be more challenging. Upstream taxation may be less advantageous.

  • Taxation of non-fossil fuel emissions will be more complex, and in some cases impractical, to administer. It will require development of new measurement systems, in some cases requiring specialist expertise, and special forms of tax or regulation. These could be introduced more gradually as capacity is developed.

  • If the policy objective is to tax emissions generated to produce goods and services for US consumption, border tax adjustments (to charge for embodied carbon in imports and rebate it for exports) will be required, but the administrative difficulties and complexities these would pose should not be underestimated.


This chapter considers the administration of a US carbon tax (CT). It does not consider administration of other methods of controlling greenhouse gas (GHG) emissions, such as emissions trading systems (ETS) or regulation.

The chapter starts with a brief illustration of US GHG emissions and of the revenues that would potentially arise from taxing them. It then briefly outlines alternative policy options for taxing them, not with a view to evaluating those options but merely with a view to considering their administrative implications. The options considered are first, comprehensive taxation of US GHGs so far as is practicable; second, taxation of US GHGs with exemptions for energy-intensive trade-exposed (EITE) or other businesses; and third, taxation of GHG emitted in the US or elsewhere to produce goods and services consumed in the US (described here in short as taxing GHG consumption).

For fossil fuel combustion emissions, the chapter discusses the practical advantages and disadvantages, administrative and otherwise, of “upstream” taxation (taxation at points where fossil fuels enter the economy) and “downstream” taxation (taxation at points of combustion or energy consumption). It considers the administrative implications of carbon capture and storage (CCS) and carbon “sinks.” It looks at the practicalities of extending CT to non-fossil fuel emissions. It considers administration of international transactions, and in particular the issues that would arise if it were intended to apply carbon pricing to carbon intensive goods and services imported into the US from countries that do not price carbon. Finally it outlines how a US CT might be administered in practice.

US GHG emissions and potential CT revenues

Table 3.1 shows 2010 US emissions of CO2 and other GHG measured in terms of their CO2 equivalence (CO2e). Total US GHG emissions vary from year to year (reflecting changes in energy demand, fuel prices, fuel efficiency, etc.). Various factors (such as expanding shale gas production) may increase relative contributions from some sources and reduce them from others. These figures are therefore just a snapshot, but can be used for illustration since the broad picture is unlikely to change significantly in the near future. Ignoring the role of carbon sinks, US GHGs were 6,821.8 million metric tons (MMT) in 2010. Fossil fuel combustion accounted for 70.0 percent of these emissions, with the major contributors including power generation (33.1 percent), transport (25.6 percent), industrial fuel use (11.4 percent), residential use (5.0 percent), and commercial uses (3.3 percent). Non-CO2 GHGs include methane (which accounted for 9.8 percent of GHG emissions), nitrous oxide (which accounted for 4.5 percent of emissions), and hydro-fluorocarbons (HFCs) (which accounted for 1.8 percent).

Table 3.1Sources of US GHG emissions and potential revenue from a carbon tax, 2010
Gas/sourceCO2 or CO2e emissions, MMTPercent of total gross US GHGsRevenue from $20 per ton CO2 tax, $billion
Fossil fuel – combustion5387.879.0107.8
US Territories41.60.6
Fossil fuel – non-energy use125.11.82.5
Industrial processes193.52.83.9
Iron and steel54.30.8
Natural gas systems32.30.5
Natural gas systems215.43.2
Enteric fermentation141.32.1
Coal bed72.61.1
Petroleum systems31.00.4
Nitrous oxide306.24.56.1
Soil management207.83.0
Stationary combustion22.60.3
Mobile combustion20.60.3
Other non-CO2 GHGs19.70.30.4
Total Gross GHGs6821.8100136.4
Land use (sinks) CO2(1074.7)
Net total GHGs5387.8
Source: Author’s calculations based on EPA, 2012.
Source: Author’s calculations based on EPA, 2012.

The table also shows government revenues from a CT applied to 2010 emissions at an illustrative rate of $20 per ton. These would have amounted to around $136 billion, but it should be remembered that a CT could itself be expected to alter emissions levels and the relative contributions from different sources, perhaps significantly. As will be discussed, it would also be impracticable to tax all GHG emissions. If a CT reduced the illustrative 2010 emissions by, say, 15 percent, that would reduce revenues to around $116 billion; each percent of GHG taxed would produce $1.16 billion. And if, say, only 90 percent of those reduced emissions could be taxed in practice, that would reduce revenues further to around $104 billion. This “back-of-the-envelope” calculation1 is useful as an indication of potential order of magnitude of a US CT. For comparison total US federal taxes in 2010 were $2,163 billion, of which corporate income tax was $191 billion and excise taxes $67 billion. A CT would therefore be a substantial tax, comparable to other major business taxes, but, in the context of a $700 billion deficit, would not by itself transform US government finances or the overall US tax burden.

Policy and administration

Effective administration of a CT, as of any other tax, requires a well-defined tax base, with a clear underlying policy objective. Lack of clarity is often a major barrier to compliance and effective administration. The choice of tax base also needs to command broad public acceptance – without which compliance is inevitably problematic. Administration should not drive policy, but policy intentions must be capable of effective implementation in practice.

The general rationale for a CT is to internalize an externality (GHG emissions) that causes universal harm (global warming) to future humankind. If an all-seeing, all-powerful government representing future humankind were to impose a CT, it would probably impose it on those citizens whose consumption of fuels and products caused GHG to be emitted, and spend it to benefit humankind generally, particularly those who would suffer most from the consequences of global warming. This would transfer value from countries responsible for higher-than-average per capita GHG emissions, such as the US and Australia, to countries responsible for lower-than-average per capita emissions (i.e., broadly speaking, from rich countries to poor countries, though of course there is not an exact correlation between wealth per capita and GHG emissions per capita).

But future humankind does not have a government, and taxes are generally imposed and collected by national governments. Conceptually a national CT might seem an odd mechanism for tackling a global harm. But it is hard to imagine the US, or any other government for that matter, accepting an international CT imposed disproportionately on its own citizens and spent disproportionately for the benefit of citizens of other countries. A national tax is also a practical approach: it allows countries to use established tax collection and expenditure mechanisms to tax transactions or events within their own borders. It is hard to imagine the US ceding tax collection and expenditure to an international body like the United Nations Framework Convention on Climate Change (UNFCCC). Governments that have introduced carbon pricing elsewhere (e.g., in Australia, Sweden, British Columbia) have generally collected carbon taxes or auction fees on their own national emissions, and spent them primarily for the benefit of their own citizens.2

A US CT would almost certainly likewise be collected by the US government and spent primarily for the benefit of US citizens. It would thus be broadly neutral for the US as a nation, transferring value from US citizens affected by the CT to US citizens who benefitted from the disposition of its revenues (obviously those classes would overlap). This would probably be necessary (though perhaps not sufficient) to secure US government and public acceptance of a CT. The policy objective might be as much to increase government revenue (to reduce the deficit, lower other taxes, or increase spending) as to curb GHG emissions. But a US CT on US emissions, even if collected and spent within the US, would still benefit humankind globally to the extent it reduced global emissions and spurred emissions pricing initiatives in other countries. Indeed if other countries similarly taxed their own emissions, the combination of national CTs would have the same impact on global emissions as an international CT.

Unfortunately, however, it cannot be assumed that all countries will tax their GHG emissions. In that case a US CT on US emissions would reduce global emissions so far as it applied to non-mobile sources of emissions. But if it applied to mobile sources (i.e., emissions from activities that could be carried out abroad), there is a risk that these would gravitate to countries that did not price emissions. This would negate the benefit of taxing those sources, and at the extreme could even result in an increase in global emissions, if the activities generated higher GHGs abroad then they would have done in the US. A number of policy responses to this are possible. For example:

  • Tax all US GHG emissions anyway. Treat the risks of business moving as not significant (for example, because the advantages of cheap energy resulting from shale gas might compensate for the disadvantage of a CT); or manage those risks in some other way, for example, by output-based rebates.

  • Tax US GHG emissions but with exceptions or exemptions for particular industries, especially energy-intensive trade-exposed (EITE) industries.

  • Tax GHG emitted in the US or elsewhere to produce goods and services consumed in the US by means of border tax adjustments.

These different options (and no doubt there are other possibilities) would have implications for CT design, which would in turn have implications for administration.

The first policy option is very clear and straightforward. Implementing it would be largely a question of administrative practicality. To internalize the cost of US GHG emissions accurately, a CT should ideally apply to them comprehensively and at a uniform rate based on their global warming potentials (GWP). In practice administrative difficulty or cost may make this difficult to achieve. Gaps in coverage and non-uniformity of rates would not necessarily make a CT ineffective, just less effective than it might be. The aim should be to find the best possible balance between coverage/uniformity and administrative practicality.

The second option is more complex. Exemptions may be difficult to define and monitor. As will be discussed, administrative options that are suitable for comprehensive GHG taxation – for example, collection based on “upstream” sales of fossil fuels – may be less suitable for selective GHG taxation.

Taxation of US GHG consumption would be more complex still. This is discussed in more detail later under International Considerations.

It is important to define policy objectives clearly. If a CT is planned and designed on the basis that its objective is to tax US GHG emissions, and it emerges later that its true objective is to tax non-mobile GHG emissions or US GHG consumption, it may be difficult to re-structure it to meet that revised objective in practice.

CO2 from fossil fuel combustion

CO2 from fossil fuel combustion accounted for 79 percent of 2010 US GHG (see Table 3.2), and so is the main source of GHG to which a CT would need to apply. (Non-energy use of fossil fuels, discussed later, accounted for a further 1.8 percent.)

Table 3.22010 CO2 fossil fuel combustion emissions, and revenue potential, by fuel type
MMT CO2 or CO2e emissionsPercent of gross US GHG emissionsRevenue potential from $20 per ton CO2 tax
CO2 from fossil fuel combustion5387.879.0$107.85bn
Natural gas1261.618.5$25.2bn
Source: Author’s calculations based on EPA, 2012.
Source: Author’s calculations based on EPA, 2012.

Distinct stages can be identified in the use of fossil fuels:

  • Production (discovery and extraction)

  • Processing

  • Combustion

  • Consumption (of heating, transport, goods, etc., that use energy from combustion).

Between stages there will often, though not always, be a transmission or distribution stage. For example, natural gas is distributed from gas processing plants to users via a pipeline network; electricity generated by fuel combustion is distributed to consumers via an electricity “grid”; in other cases (such as petroleum-powered transport) combustion and consumption happen at the same point.

Most fossil fuel undergoes processing after production. Crude oil is distilled and subjected to other processes at refineries to produce a range of usable petroleum products, most but not all of which are combusted as fuel. Natural gas is normally processed at gas processing plants to separate gas liquids and to remove water and impurities. Coal is normally “washed” and graded at a coal preparation plant to remove rocks and dirt and other unwanted content. Processing may be carried out as a separate business – most common in petroleum refining – or as part of a production business.

CO2 emissions from fossil fuels almost all occur at the combustion stage. But a CT can be imposed indirectly at earlier “upstream” stages. “Upstream” is traditionally used in the extractive industries to mean exploration and production (and “downstream” to mean other stages), but in the carbon pricing literature (and this chapter) it just means imposed at a stage prior to combustion/consumption, that is, it may include processing.

CT could be imposed at different stages for different fuels or uses, with an obvious requirement that it should apply only once. For example, producers could be required to pay CT only on sales but not sales to approved processing or refining companies, who would then pay CT on their sales instead; refiners could pay CT on their sales of petroleum products, but not on gasoline sales to wholesalers, who could be taxed on their sales instead.

The choice of which stage should form the CT base will depend on various factors. Administrative considerations – finding the best balance between coverage and accuracy of measurement of carbon content on the one hand, and administrative cost, convenience, and practicality on the other – may be important, but non-administrative political and economic considerations may be at least as important.

Advantages and disadvantages of upstream CT3

Administrative advantages

  • Limited number of measurement points/taxpayers. Tax is paid by taxpayers, not points, but the number of points at which measurement is required affects taxpayers’ compliance costs and burdens, and the ease of government monitoring.

    • CT on petroleum could be applied to output from refineries, of which there are around 150, and to imports, which occur at a small number of points (diminishing as tanker sizes increase).4 The number of refining firms who would measure output from their own (multiple) refineries and pay the tax would be considerably smaller. CT based on production measured at the well-head would involve a larger number of taxpayers (oil producing firms), and a much larger number of measuring points, and would have no obvious advantages.

    • CT on coal could be applied to output from mines measured at the mine mouth (and again to imports measured at import points). There are around 1,300 US coal mines, but a much smaller number of producer firms who would measure the output from their mines and pay the tax. In 2010, 26 producers accounted for more than 85 percent of total US coal production, and the total number of producers was fewer than 500. CT could alternatively be applied to output from coal preparation plants – likely to be fewer in number than mines – where coal is routed through these.

    • For gas there are over 450,000 gas wells, but a much smaller number of gas producing firms (around 8,000, of whom no more than 500 are of significant size). CT could therefore be paid by producer firms, based on total well output. A better option might again be to tax output from gas processing plants – there are around 500 significant ones (though again also some small ones) and again the number of processor firms that measured and paid the tax would be smaller than the number of plants. Again imported natural gas would have to be taxed at import points (of which there are around 50).5

So possibly not more than, say, 1,250 to 1,500 taxpayers could account for CT on the vast bulk of fossil fuels if taxed upstream. Even if that number were increased 20-fold (for example, because producing firms rather than processing firms were taxed) it would still be a small number for a tax yielding $100 billion. Contrast company income tax, for example, where in 2010 there were more than 1.8m US C corporations paying $191 billion. Furthermore, if it was decided to impose CT on methane emissions from gas, coal, and petroleum production (see later discussion) – which produce up to 4.5 percent of GHG emissions – some of that CT might also be collected from those same taxpayers.

  • Clearly identifiable taxpayers. Upstream CT would apply to taxpayers who sell coal, gas, or petroleum. Those are recognized business activities, making CT taxpayers easier to identify. Contrast combustion, which, unlike refining, is not a business, but a business by-product. Almost all businesses combust fossil fuel to some extent, so defining a manageable number of taxpayers requires identification of those emitting CO2 above a chosen threshold. This may be difficult at the margin since emissions (or fuel inputs) are not measured and reported for normal business purposes, and may present avoidance opportunities.

  • Coverage. An upstream tax could achieve comprehensive coverage whereas by definition a tax on emissions above a chosen threshold implies less than perfect coverage. Of course if de minimis limits were applied to exclude small upstream businesses from CT, then an upstream CT would likewise not be comprehensive, but these often complicate rather than simplify administration, for example, by creating avoidance opportunities, and would probably not be necessary or worth the trouble, given the small total number of upstream taxpayers.

  • Measurability. The CT base must be visible and measurable, and must provide an accurate measure of GHG emissions, even if indirect. This reduces taxpayer costs, and avoids disputes. The technical difficulty of measurement is also clearly relevant to administration. It is an advantage if the base is measured for normal business purposes, since again this reduces costs and makes measurement easier to monitor. The volumes of output from fuel producers and processors are relatively easy to classify and measure, and the CO2 output is in turn broadly measurable on the basis of those measurements, limiting technical difficulty and scope for dispute. It is worth noting, however, that likely CO2 output can generally be more accurately measured from processed fuel outputs than from unprocessed outputs, which may contain varied amounts of impurities (which lower the carbon content per unit of the fuel). For petroleum, measurement of refinery outputs also has the advantage of allowing different distillates to be taxed at rates reflecting their different CO2 content or different likelihood of combustion. Coal has four different grades (anthracite, bituminous, sub-bituminous, and lignite) with even more significant variations in CO2 content, and again it is likely to be easier to measure CO2 content after coal has been washed and graded.

  • Limited scope for avoidance. Note that avoidance by reducing GHG emissions is acceptable – indeed is one of the prime objectives of a CT. By-passing the point of measurement to avoid tax on US GHG emissions would, on the other hand, be unacceptable, as it would frustrate the policy intention. There would be limited opportunities for fuels to by-pass upstream measuring points, since they are generally measured at those points as part of normal business processes. The fossil fuel industries are closely regulated for environmental and other reasons, further reducing opportunities for avoiding measurement points. Avoidance of CT by relocation of GHG emissions abroad is a more difficult issue, as discussed later under International Considerations, but the location of upstream activities would not be relevant to this – imports of fuels destined to create emissions from combustion in the US would be taxed, and exports destined to be combusted abroad would be exempted.

  • Use of existing tax mechanisms. This (rather than the limited number of taxpayers) is probably the key advantage of upstream taxation. An upstream CT on fossil fuels would be broadly similar in operation to an excise tax. The US Internal Revenue Service (IRS) already collects excise6 on petroleum products and coal (though not on natural gas unless compressed for use as a transport fuel). CT would involve some refinements that do not apply to current excise – for example, different rates for different fuel types, possibly credits or refunds for non-combustion uses. But broadly speaking, key procedures developed for excise purposes would be similar to those required for CT: for example, procedures to ensure that fuels do not by-pass excise measuring points and are taxed only once, to provide refunds where fuels are used for excise-exempt purposes, and to impose excise on imports and exempt exports. Extending this kind of regime to natural gas should raise no major new issues of principle. The existing excise legal framework for taxpayer registration, returns, payments, audit, and dispute resolution could probably be adapted for CT without too much alteration, and would have the advantage of familiarity to most CT taxpayers. The training and skills of IRS staff working in this field would be well suited to applying a similar regime to CT and making any necessary modifications. Excise is encrusted with the idiosyncrasies of any long-established tax, and full integration with CT might therefore be difficult, but prospects for rationalization of their relationship would likely be stronger if they were administered together rather than separately.

So in short, an upstream CT on fossil fuels appears likely to provide a good balance of coverage and administrative cost and convenience. It would cover 80.8 percent of GHG emissions (including non-combusted fossil fuels), the vast bulk of which would be collected from no more than, say, 1,500 taxpayers, with the possibility of further methane emissions being taxed on broadly the same population.

Non-administrative advantages

  • Opposition and pressure for compensation or exemption might be less widespread than if CT were imposed directly on downstream industries. (At any rate they are unlikely to be worse.) It would also make it slightly more difficult to grant special exemptions in practice, which proponents of a universal CT would consider an advantage.

  • The low visibility of an upstream CT to consumers and voters might make it more acceptable. An upstream CT might have the same price impact on consumers as a downstream CT, but it might have the same low visibility as a corporate income tax, which consumers tend to think of as a tax on someone else, even if passed on to them in higher prices. Taxing big mining and petroleum companies might even be popular with voters, increasing general public acceptance. There are competing considerations here. Politicians sometimes advocate transparency in taxation (particularly if opposed to the government imposing the taxes); and there is a special case for visibility of CT since it is meant not just to influence consumer behavior but raise consumer awareness. But in practice governments hoping for re-election might find low visibility attractive.

Administrative disadvantages/problems

  • As discussed, it may be difficult to measure CO2 content accurately if production rather than processing is used as the CT base. But it may be possible to use processed fuels in most cases. In any case minor inaccuracies in measurement of CO2 content, while undesirable, would not make a CT ineffective.

  • In some cases fuels may by-pass processing facilities. For example, dry gas may be fed directly into transmission pipelines after limited well-head processing. It has been estimated that CT based on outputs from gas processing plants plus imports would catch only 70 percent of emissions from natural gas combustion. Similarly coal may be delivered direct from mine mouth to end user. But it should be possible to tax fuels at an alternative upstream point where they are not routed through processing plants. Similar problems are successfully managed in applying excise taxes.

  • A small proportion of gas and petroleum outputs may not be combusted (e.g., asphalt, chemical feedstocks), or CO2 from combustion may be captured and stored. CT should not apply in those circumstances. But practical ways of dealing with them could be developed, as discussed later.

  • Taxation of non-fossil fuel emissions (e.g., nitrous oxide, CO2 from industrial processes) would have to be administered separately, whereas it could be combined with downstream taxation of fossil fuel emissions.

Non-administrative disadvantages

  • Upstream CT may be seen as inconsistent with the “polluter pays” principle. This principle may partly explain why most countries tax combustion emissions. (Other anti-pollution programs, for example on SO2, have tended to focus on emissions.) Upstream CT may therefore be seen as less “fair.” But a case can be made for allocating some GHG responsibility to producers and processors – they let the genie out the bottle.

  • There is likely to be industrial opposition. This will happen however CT is imposed, and, as discussed earlier, the lower visibility of upstream taxation might actually reduce opposition. But if policymakers wished to respond to it, for example by giving reliefs or exemptions to industry, or more generally by grandfathering existing CO2 production levels, it might be more difficult if fuels were comprehensively taxed upstream. This would depend on the type of exemption proposed. For example, if the aim was to exempt EITE industry from CT on power consumption but not transport costs, upstream taxation might make this more difficult.

  • As discussed above, it might be argued that the low visibility of an upstream CT to consumers would lessen its impact on consumer behavior and choice (although economists tend to be skeptical of this argument).

  • In countries with well-established downstream fuel and energy taxes, governments may be reluctant to replace them with a new upstream tax that is likely to be controversial, however much preferred by economists. (An old tax is a good tax.) This, however, is of limited relevance to the US, which does not currently impose any significant downstream federal fuel and energy taxes.

  • There is a treaty bar on taxing air fuel used for international flights. But a CT on refinery output might not fall foul of this.

Are there advantages to applying CT at the point of combustion?

Most other countries apply carbon pricing to large emitters at the point of combustion. The reasons for this may be mainly non-administrative. As discussed, it may be seen as more consistent with the “polluter pays” principle. It may also be seen as more visible to consumers. (It may, however, be difficult for business reasons for CT taxpayers to make CT visible to consumers even if applied at the point of combustion – for example, a power company with power stations using different fuels might find it awkward to discriminate between customers and prefer just to pass on CT as an average price increase.) Another (political) advantage of taxing at the downstream level is that it facilitates targeted “fiscal cushioning” (such as special CT exemptions) to placate powerful industrial lobbies (as has happened in various countries).

One administrative advantage of a combustion-based CT is that it avoids the need under an upstream CT for a mechanism to exempt non-combusted fossil fuels (asphalt, etc.). But this is a relatively minor issue, because only a small proportion of fossil fuels are not combusted and, as discussed later, exempting them under an upstream CT should not be a major problem.

One claimed administrative disadvantage of taxing at the point of combustion is lower coverage (most countries that tax on this basis cover less than 50 percent of their GHG emissions). It is not practical, for example, to tax transportation fuel at the point of combustion. But again low coverage may to some extent reflect political rather than administrative considerations, for example, the desire to protect certain industries; or the fact that many countries already have high taxes on transport fuels, and may be reluctant to impose carbon charges on top of them. There is no reason in principle why CT at the point of combustion for some fuel uses (such as electricity generation) should not be combined with CT at the point of wholesale distribution for other uses (such as transportation).

But there may be other reasons why it is inherently difficult to make a combustion-based CT comprehensive. There are a larger number of taxpayers (and measuring points) where CT applies at this point rather than upstream. It has been estimated that the number within the European ETS is 12,000 and that a similar number would be subject to a US combustion-based CT, whereas under an upstream system broad coverage could probably be achieved with around 1,500 taxpayers. But 12,000 is not a large number of taxpayers for a major tax (again contrast the numbers for US corporate income tax).

A more important issue may be the technical difficulty of measuring CO2 emissions at this point. For example, the EU rules on measurement for the purposes of its ETS are highly technical and complex.7 It imposes major bureaucratic burdens on businesses, such as measuring fuel inputs and applying emissions coefficients, which it would not do for normal business purposes. It is often considered impractical to impose such burdens on any but the largest businesses (for example, the Australian CT applies only to only around 300 companies emitting more than 25,000 tons p.a., and is estimated to cover only around 60 percent of CO2 emissions). Restriction to large companies thus reduces coverage, and may distort markets and create opportunities for avoidance and perceptions of unfairness. Combustion emissions need special technical expertise to monitor – they are generally not monitored by normal tax administrations. So the increase in administrative cost and complexity from taxing at the point of combustion may be much greater than the mere increase in taxpayer numbers would suggest.

CT at point of distribution to final consumer?

It would be impractical to impose CT at the point of consumption if by this is meant taxing millions of end consumers. But where consumers buy fuel directly for heating or transport purposes, it would be relatively straightforward to impose CT on retail distributors in the same way as a retail sales tax. At the other extreme it would be impracticable to impose CT on retail distributors of manufactured goods, since it would be extremely difficult and administratively onerous to calculate the GHG emissions involved in the manufacture of complex goods such as cars, which have hundreds of components with different origins, manufacturing histories, power inputs, and so on. In between, there are cases where consumers do not purchase fuel directly, but where it may be possible to quantify the GHG input reasonably accurately (for example, electricity – but even there it could be difficult if the supplier provided electricity from a different kind of power plant).

For similar reasons, it is in general unlikely to be possible for CT to be passed on cumulatively through a supply chain in the same way as a VAT, where tax accumulating on each addition to value is eventually borne by the final consumer (though not collected by the government direct from the final consumer). The accumulation of GHG inputs and outputs would be far more complex and onerous to measure than sales inputs and outputs used to calculate VAT. (Again car manufacture is an obvious example where this would be impractical.) In many cases CT would for practical purposes have to be passed on as a price increase (as is normal for excise taxes) rather than as a tax (as is normal for sales taxes).

Point of taxation of fossil fuel emissions – general conclusions

On balance an upstream CT on fossil fuel sales seems likely to be an administratively better option than a downstream CT imposed directly on power and industrial plant emissions, providing a better balance of coverage/accuracy and administrative practicality; and it has no conclusive non-administrative disadvantages. But countries have in practice managed to administer downstream CT on power and industrial plant emissions, and if combined with CT on retail sales of fuel for transport and domestic consumption, this could achieve quite high coverage, even though not as high as an upstream CT. If CT is imposed upstream on fossil fuels there does not have to be a single point of taxation. The experience of collecting excise taxes is likely to be useful for working out the most practical points of measurement to ensure that they are comprehensively taxed in practice. If petroleum excise tax is applied to refinery output, and coal excise to mine output, these might be best points of measurement for CT purposes too. This chapter cannot claim to have identified and addressed every practical problem that might arise with an upstream CT. But in general it seems likely that such problems will be similar in kind and degree to those that arise with excise taxes, and should therefore be capable of being addressed in practice. British Columbia probably comes closest to an upstream CT, and seems to combine reasonable coverage (estimated at 77 percent of Canada’s total GHG emissions8) with relatively simple administration.

Sequestered CO2 from fossil fuels

Non-energy use of fossil fuels

Non-energy use produced 1.8 percent of GHG emissions in 2010, but if CT applied upstream, it would be necessary to ensure that it applied only to CO2 emitted and not to CO2 stored, for example, in asphalt (where 100 percent is stored), chemical feedstock (62), waxes (58), lubricants (9), others (25). If CT applied to refinery output it would be possible to identify and exempt asphalt at that stage, and similarly to identify and exempt (or partially exempt) other products destined for non-combustion use. If not possible, then purchasers would have to apply for refunds (as currently happens with excise where petroleum fuels are used for exempt purposes such as agriculture).

Carbon capture and storage (CCS)

CCS is not an immediate concern since there has been limited development of CCS technology in practice so far.9 If it did become viable in future, steps would have to be taken to exempt captured CO2 from CT. If CT applied upstream, it would not be practical to identify and exempt in advance fuels whose CO2 would be captured by CSS. Retrospective linking of CCS to the relevant upstream supply, refund of CT on that supply, and revision of the price charged to the customer for it would be administratively complex, and if CCS ever became common, with CO2 piped from numerous sources to CCS plants, it would probably negate some of the administrative advantage of upstream CT. The problem with CT credits or refunds in an upstream system is that upstream CT payers might not be involved in CCS projects, and downstream businesses who were involved would be unable to use them. One option would be to make CT credits tradable, but this would add complexity and might be open to fraud. A simpler option would be for the government just to pay an amount equal to CT per ton to a company that implemented CSS.10 The government would monitor CCS facilities, and charge CT on any CO2 that escaped. Monitoring CCS would be a specialist technical function, and monitoring and payment could be allocated to a specialist technical agency either in the IRS or outside it. International issues might have to be considered in due course – for example, import or export of liquefied CO2 for CCS. Credit for export would need international cooperation (to ensure CO2 was genuinely stored). The US would give no credit for imported CO2, but would impose CT on any that escaped.

Land use (forestry, etc.) carbon sinks

Forestry provides the main potential carbon “sink” (though of course deforestation can increase GHG emissions). In 2010, forestry and other land use were estimated by the EPA to have reduced US CO2 emissions by a substantial 1,075 MMT, primarily through increases in forest density or acreage. In principle there is a case for the government to provide some form of CT set-off for additional CO2 sunk in forestry. As with CCS, it would probably be better for the government not to try and embed forestry within the CT regime through CT refunds, tradable CT credits, and the like, but instead simply to make payments equivalent to the rate of CT per ton direct to forestry businesses that could provide an acceptable measure of additional tons of CO2 sunk, since this would be easier to administer and less open to fraud. The basic problem, however, would be how to establish additionality. It has been suggested that if potential qualifying companies were limited to large paper and forest product companies, who were required to measure sunk CO2 by reference to their entire land stock, it might make additionality easier to establish. But there would be a host of other complex technical issues in measuring CO2 sunk in forestry, even if additionality could be established. These issues are far too complex to do justice to in this short chapter. Measurement, if it did become feasible, would be a highly technical operation, and, as for CCS, administration would probably have to be carried out by a specialist technical agency either within the IRS or entirely separate from it (since the required expertise might already exist in other government departments).11

Extending CT to non-fossil fuel emissions

With non-fossil fuel emissions it is again normally possible to identify different stages at which a CT could theoretically be imposed, but it is generally harder to identify anything equivalent to an upstream taxing point; and a large proportion of these emissions are fugitive, not measured for normal business purposes, and in some cases very difficult to measure at all in relation to individual taxpayers. This means that administration is likely to be more complex, and coverage in some cases may be impractical. A factor to take into account is the relative cost or difficulty of reducing the emissions. If the prime purpose of a CT is seen as reducing emissions, it may be felt that if CT would result in a relatively large reduction in GHG (particularly those with high GWPs), it may be worth the administrative effort even if the emission is relatively small; and conversely that there is little point in applying CT to a substantial but hard-to-administer emission if no one can reduce it anyway. (An alternative point of view is that governments should apply CT to significant emissions even if they are difficult to reduce, primarily as a means of raising revenue.)

Non-fossil fuel CO2

CO2 emissions from industrial processes (over and above their energy input) were 2.8 percent of total GHG emissions in 2010. Natural gas systems accounted for 0.5 percent. CT on combusted fossil fuels would increase the competitiveness of natural gas against other fossil fuels, which might provide a special reason for trying to ensure that all emissions from natural gas industrial processes were fully taxed, so that it did not also enjoy an unfair advantage. In view of the numbers, small producers might need to be excluded. Iron and steel producers and cement producers between them accounted for 1.3 percent, and a further four industries – lime production, waste incineration, limestone and dolomite use, and ammonia production – for 0.6 percent. The numbers of companies engaged in those businesses are not large – for example, there are around 116 cement plants owned by 39 companies, and 23 steel mills (per Metcalf and Weisbach 2009) – and estimating the CO2 output from those industrial processes should on the whole be feasible, though it would require special technical measurements and procedures, adding administrative complication. This would bring a further 2.4 percent of GHG emissions within CT (or around 85 percent of the CO2 from industrial processes). Most of those businesses are, however, EITE, so leakage and international competitiveness issues would be particularly important with some of these taxpayers. There might therefore be strong political pressure not to tax CO2 emissions from industrial processes, or alternatively to protect them from competition by border tax adjustments on imports and exports (as discussed later).

Non-CO2 GHG emissions

Non-CO2 GHG emissions accounted for 16.4 percent of 2010 emissions, methane being the most significant. These gases have high, sometimes very high, GWPs. A CT would have a very high impact on their price. They may therefore be a relatively low cost source of emissions reductions. Metcalf and Weisbach (2009) estimate that they could possibly account for one-third of the emissions reductions that might arise from carbon pricing, presenting a strong policy case for bringing them within CT as far as possible.


Methane accounted for 9.8 percent of 2010 GHG emissions. The main sources were natural gas systems (3.2 percent), enteric fermentation12 (2.1), landfills (1.6), coal mining (1.1), and manure management (0.7).

  • On natural gas systems, Metcalf and Weisbach (2009) estimate that around one-quarter of emissions come from field operations and are not practical to measure and tax, but suggest that it would be practicable to tax the remaining three-quarters emitted from processing, transmission and storage, and distribution, by monitoring inputs and outputs and taxing the difference. These proportions may have changed, since it is claimed that shale gas development produces higher emissions from field operations than conventional gas. For the reasons discussed earlier, there may be a special case for taxing standard estimated methane emissions from gas field operations, since failure to do so would unfairly increase the advantage CT gave to natural gas over other fossil fuels.

  • Metcalf and Weisbach (2009) suggest that it could be cost effective to bring enteric fermentation within CT, by taxing cattle per head, with varied rates for five different cattle types (which emit different amounts of methane), and reduced rates where there was proof of emissions-reducing cattle diets. Although this may be feasible, there is no doubt that it would be complex to administer – the US has many farms and cattle herds, whose composition and diets might change regularly. (Other countries such as Australia have in general concluded that agricultural emissions are too difficult to tax and simply exempted them from CT.)

  • There are around 1,800 landfills, which currently capture and combust 50 percent of their methane emissions (turning them into less harmful CO2). It would be feasible to measure and tax those emissions, providing a worthwhile incentive to increase their capture.

  • For coal bed methane it would be relatively straightforward to capture and tax the two-thirds emitted from underground mines, since companies have to control those emissions for safety reasons. But it would be less easy to capture and measure the one-third emitted by surface mining. It could be argued that it was inappropriate to tax emissions that were difficult to measure and control, but again it could be countered that there is a special case for taxing all methane emissions from production of fuels, even if this necessitates taxing on the basis of a standard estimate.

  • Manure management, the least important source discussed here, is not practical to tax.

Nitrous oxide

This gas accounted for 4.5 percent of 2010 US GHG emissions. But three-quarters of this comes from agricultural management (soil, fertilizer practices, and so on) and is not practical to tax. (Alternative approaches such as taxing fertilizer could have unintended consequences, such as more cows and enteric fermentation.) The remaining quarter comes mainly from combustion emissions from cars and other machinery. Metcalf and Weisbach (2009) suggest that mobile combustion emissions could be taxed downstream by means of annual vehicle tests and calculations of emissions per gallon, miles per gallon, and annual mileage. The sheer number of vehicles and machines involved would make taxing this source a costly and difficult procedure. An alternative option would be to identify a limited number of businesses that emit nitrous oxide above a certain limit, and restrict CT to those. An alternative or additional approach might be to keep tightening nitrous oxide emissions standards for new vehicles and machinery.

HFCs (and PFCs) are synthetic chemicals that substitute for ozone-depleting substances in the manufacture of products such as refrigerators and air conditioning systems. They have very high GWPs, but are hard to tax because emissions are fugitive (resulting from leakage or incorrect disposal). Metcalf and Weisbach (2009) suggest use of a downstream deposit/refund mechanism (with refunds also for “banked” gases in products already produced). The biggest problem would be how to charge the appropriate deposits on imports of HFCs and of goods containing HFCs, which would be difficult to quantify.


Altogether non-fossil fuel emissions accounted for 19.2 percent of 2010 US GHG emissions (Table 3.3). It might be possible to tax nearly half of those at reasonable administrative cost (though at considerably greater cost than upstream taxation of fossil fuel emissions). This would include most CO2 emissions from industrial processes, and CO2e from methane emitted by natural gas systems, landfills, and coal mining. Methane from enteric fermentation, nitrous oxide from mobile sources, HFC from selected products, and possibly various other minor emissions might also be considered, though are more difficult. Taxing those sources would require setting up new taxation procedures and mechanisms, and the development of new technical skills. It would probably be best to introduce this gradually over time.

Table 3.3Potential US carbon tax coverage
Source% US GHGPracticable to tax?Comments
Fossil fuel emissions79.079.0Assumes comprehensive upstream taxation, no exemptions.
Fossil fuel non-energy1.81.8As above.
Industrial processes2.82.5Develop special measurement systems for main industries concerned. Will probably need de minimis limits. Opposition likely.
Methane9.86.0Mainly fossil fuel producers (possibly only large ones) and landfill.
Nitrous oxide, other GHG6.60.7Very difficult to tax – would probably need special fees, deposits, etc., outside mainstream CT system.

International considerations

If the policy objective of a US CT was to tax only US GHG emissions, and it applied upstream to fossil fuels (for simplicity let us say at the processing stage), administration of international transactions would be relatively straightforward. Where fossil fuels were imported in unprocessed form, they would be taxed at the processing stage rather than on import. Where they were imported as processed products, they would be subject to a border tax adjustment (BTA). This would impose CT in exactly the same way as on domestically produced fuels, using the same simple volume measurements and product classifications. Fuel imports occur at a relatively small number of points, subject to regulation. Taxing imports at those points would not greatly complicate administration. The imposition of tax on imported refined fuels should not cause WTO problems because the taxes imposed would be calculated on exactly the same basis as on domestically produced products. Where fossil fuels were exported in unprocessed form, CT would not apply. Where they were exported in processed form, they would be exempted. This should not impose major administrative burdens – again the same taxpayers and measurement systems would be involved as for CT imposition. The exemption might in theory cause WTO problems, but it could be argued that CT on fossil fuels was essentially a destination-based tax on consumption, so that exemption should be allowed. It should be noted that WTO rules do not prevent zero-rating of exports for VAT purposes or exemption of exports from excise taxes, which would provide clear parallels for CT exemption of exported fuels. The taxation of emissions from imported fossil fuels and the exemption of exports would mean that US producers did not suffer competitive disadvantages relative to foreign producers in the same markets (though of course CT would affect competition between different fuel types within the US).

Countries could in theory not apply CT to imported refined fuels and not exempt exported ones, effectively applying an origin-based system. But it is hard to see why this would make sense on theoretical grounds, since it would mean that the country that collected CT was one where emissions were neither emitted nor consumed. Apart from that, realpolitik is likely to influence countries’ decisions on how CT taxing rights should be allocated internationally. It is hard to see why the US would support the principle of allocating them primarily to fuel-exporting countries (which would include Saudi Arabia, Russia, and Iran, to mention a few at random).

Imports of other non-CO2 sources of GHG would likewise be taxed and exports exempted. In practice the main issue would arise with imported HFCs, including those incorporated in manufactured goods, which, as discussed earlier, could be difficult to measure. Imports and exports of other items would, generally speaking, have no impact on US GHG emissions, though of course CT would apply in the normal way where US emissions occurred later as a result of importation (for example, if cattle were imported and enteric fermentation subsequently taxed).

With a CT based on emissions a problem arises with “bunker” fuels used for international shipping or air transport. These would not clearly fall within any particular country’s jurisdiction. If an upstream CT were applied to fossil fuels, shipping and air companies might be able to avoid it by fuelling overseas. EU proposals to charge aviation fuel emissions on flights using EU airports have been controversial, and international agreement may be required to determine how CT on such fuels should be applied and administered.13

If there was a clear and simple policy to tax GHG emissions in the US, then international transactions would be unlikely to cause major problems (other than the problem of bunker fuels just discussed). The law would be easy to apply and interpret; disputes would be unlikely; avoidance opportunities would be limited. This policy would be consistent with the general international approach, which is to tax domestic GHG emissions and not consumption. There would be no need to distinguish the tax treatment of exports and imports on the basis of whether the other countries concerned were members of the carbon pricing “club.”

But the position would change radically if the aim were to tax GHG consumption rather than GHG emissions. There are theoretical arguments for this, but practical arguments against. The theoretical arguments include:

  • The demand of US consumers causes GHG to be produced.

  • Foreign GHG resulting from US demand harm the US (and everyone else) just as much as US emissions.

  • The US government is best placed to influence US consumer behavior through its tax system.

  • If it does not tax GHG consumption, there may be leakage of GHG emissions from mobile businesses, because US firms that would suffer from CT might re-locate activities to countries that did not tax GHG, or might lose business to non-US competitors, negating the effect of CT either way (and possibly even resulting in an increase in global emissions).14

  • Another way of putting this is that US firms might avoid CT by continuing to emit GHG but re-locating their emissions abroad.

  • And if they did not they might be competitively disadvantaged in both domestic and export markets against firms from those countries.

Taxing US GHG consumption would mean applying CT to US imports of products whose production had involved untaxed foreign GHG emissions; and also exempting US exports from CT that they would otherwise, directly or indirectly, have suffered. Assuming effective implementation, this would have the following advantages:

  • It would meet the theoretical argument for using GHG consumption as the CT tax base.

  • It would prevent leakage – there would be no advantage in foreign re-location of production intended for consumption in the US.

  • It would prevent avoidance – again there would be no tax saving from foreign re-location of production intended for the US market.

  • It would not harm the competitiveness of EITE US firms. Imports would have no CT advantage over domestic production. Exports would be exempted from CT.

  • Double taxation could be avoided by exempting imports from fellow members of the international carbon pricing “club,” and not exempting exports to them (this would be administratively simpler than taxing imports and exempting exports).

  • It would provide an incentive for other countries to join the carbon pricing “club” – they would collect CT on their exports rather than letting the US collect it (though this would be counter-balanced if staying out of the “club” meant that imports were taxed that would otherwise be exempt).

A variant on this approach would be to tax carbon-intensive imports from countries not in the “club” but not to exempt carbon-intensive exports to those countries. EITE US industry would suffer competitive disadvantage in export markets, but would be protected from unfair competition in its home market. This might be sufficient to reduce leakage to acceptable levels. It would have the advantage of maximizing US government CT revenues and creating a clearer incentive for other countries to join the “club.”

This is all fine in theory, but there are major practical and administrative problems in taxing GHG consumption.

  • Measuring the GHG “content” of imports and exports. As discussed earlier, this can be extremely difficult to measure, particularly for manufactured products – and the cumulative effect of CT cannot be measured in the same way as with a tax like VAT. Even if it were possible to measure GHG inputs with all the relevant information to hand, that information would be difficult to get hold of for imports. BTAs could be possibly be simplified by basing them on standard input/output tables reflecting average energy input per unit and average emissions intensity of energy production in the US. This would not be an accurate measurement of GHG input, since foreign GHG input might be higher or lower than US GHG input (because different fuels and/or processes were used), but arguably it would create a level playing field in the US and might therefore be enough to prevent leakage and avoidance. But even this would be very complex. Any application of CT to imports and exemption of exports would inevitably be incomplete and based on somewhat arbitrary distinctions and measurements. It would be onerous to apply, open to avoidance, and probably subject to frequent dispute.

  • Reconciling BTAs with WTO/GATT legal rules. Since it is hard to measure GHG content and the amount of CT imposed, it would be hard in many cases to square taxation of imports and exemption of exports with WTO rules. Where CT was passed on to consumers as a price increase and not explicitly as a tax (for example, on manufactured goods), then applying CT to such goods on import or exempting them on export would appear discriminatory. It might, however, be possible to apply BTAs in limited circumstances – for example, if CT applied downstream to CO2 emissions from industrial processes like steel making, it might be possible under WTO rules to apply similar BTAs to tax imports and exempt exports of those products. (In general it might be easier to conform with WTO rules where CT was applied downstream.) But again this would inevitably be an incomplete and somewhat arbitrary solution to the problem.

  • Identifying which countries should be treated as carbon pricing club members. This would involve numerous difficulties. For example:

    • How to define the coverage level and carbon pricing rate required for club membership.

    • How to define an acceptable limit of “fiscal cushioning” (by subsidies, CT exemptions or reduced rates, output-based rebates, etc.) and establish whether a country was within that limit at any particular stage.

    • If particular industries were favored in the other country, whether to exclude them from preferential treatment.

    • Whether to allow countries with alternative regulation-based approaches to join the club.

    • How to establish whether club members were implementing carbon pricing effectively in practice (failure could result from administrative inefficiency, lax enforcement, even corruption).

    • If other club members taxed emissions and not consumption (as at present), how to ensure that imports from them were not products they had in turn imported from non-club members.

    • Who would administer club membership rules – an international body would be necessary to avoid duplication and apply common standards, but the US might not be happy to rely on it.

These may not quite amount to three “impossible things before breakfast,” but they would undoubtedly complicate CT administration very significantly. Instead of a clear underlying policy objective (CT applies to defined US GHG emissions), there would be a confused underlying policy objective (CT applies to defined US GHG emissions, except sometimes it applies to US GHG consumption). Disputes and uncertainties would have to be resolved according to complex and detailed rules rather than according to a clear principle. Rather than complicating CT administration in this way (contrary to normal international practice), it might be better to find simpler ways of addressing the policy and commercial objections to a CT based on US emissions – for example, by temporary grants or subsidies to affected industries. Obviously these should not completely cancel the effect of CT (which would make CT administration pointless altogether) but they might temporarily cushion the blow and reduce (though not eliminate) competitive disadvantages and the potential resulting leakage and avoidance.

Conclusion – practical administrative arrangements for a US carbon tax

It should be fairly straightforward for the IRS to administer an upstream CT tax on US emissions from fossil fuels. CT would be payable by a couple of thousand taxpayers or fewer, with a relatively easy-to-measure tax base, and should not require major addition to resources. There would be no need for a major new addition to technical skills – staff currently monitor and audit the outputs of the companies concerned. CT taxpayers might need to report and analyze their output in more detail for CT than for other tax purposes, but this analysis would often already be done for commercial purposes, since fuel products with different energy outputs have different commercial values. Border tax adjustments on fuel imports and exports would be relatively straightforward to administer, since they would be similar to those applying for excise tax purposes.

CT administration should be integrated into the normal tax administration of the companies concerned – with similar self-assessment–based procedures for taxpayer service, registration, returns filing, payment, audit, and dispute resolution, and similar risk-based strategies for encouraging compliance and detecting and dealing with non-compliance. It would be unnecessary and undesirable to re-invent the wheel for CT. On the contrary the aim should be to exploit established tried and tested mechanisms as far as possible.

Staff involved in taxing natural resource companies normally need to coordinate with and obtain information from industry regulatory agencies – this would equally be needed for CT.

Extending CT to other emissions would be more complex, and in some cases involve setting up new taxing arrangements. But so far as possible it would be best to develop those within existing tax arrangements. Only where exceptional specialist technical skills were required, for example, for monitoring CCS or for quantifying the GHG effects of land use, would it be necessary to set up specialist technical offices. It would be a matter of choice whether these should be attached to the IRS – it might be better to attach them to government agencies that already had the technical knowledge required.

These conclusions would have to be revisited if the intention was to tax US GHG consumption rather than emissions. Administration of complex BTAs, applied selectively depending on membership of the carbon pricing “club,” would add significant complication. It would have a major impact on the resources and skills required by the IRS. It would also likely need the establishment of an international agency to administer the club membership rules, with whom the IRS would need to cooperate.

The IRS reports to the US Treasury, which is primarily responsible for revenue policy. But a CT would have dual revenue-raising and environmental protection objectives, and the EPA is responsible for the latter. It may therefore be appropriate for the EPA to have some oversight and input with regard to IRS administration of CT. Some technical functions relating to CT, for example, measurement of non-CO2 emissions or of CO2 sinks, might furthermore, as discussed, be better carried out by specialist agencies reporting to the EPA (or possibly some other department) rather than the IRS. There would therefore need to be close cooperation between the Treasury and IRS on the one hand and the EPA on the other, on both policy and administration. Formal arrangements for this would have to be put in place.


I am grateful to Katherine Baer and Ian Parry for their helpful comments and suggestions.

See Chapters 1 and 3 for further discussion of revenue impacts.

Under the third phase of the EU ETS, individual member states will largely auction their own emissions permits.

See seminal paper by Metcalf and Weisbach (2009) on which this chapter draws heavily (but with occasionally different conclusions for which they have no responsibility).

According to the Energy Information Administration, the US imported 32 percent of its oil in 2013, down from 60 percent in 2005.

Bluestein (2008) estimates that charges levied on processors and importers would cover 70 percent of natural gas emissions.

The BC tax does not, generally speaking, apply to emissions other than those from fossil fuel combustion.

For a discussion of the obstacles to this technology see Deutch et al. (2007), pp. 43–62.

One can imagine competitive market-based solutions being developed, with CCS providers competing to buy CO2 from emitters at a price reflecting CT per ton less their CCS costs. But if separate firms carried out capture and storage (for example, a coal plant captured CO2 which an oil producing company stored in a depleted oil reservoir), an issue to be addressed (whether CT applied upstream or downstream) would be how to apportion any CT refund or credit.

This chapter does not discuss the complicated issue of how anything equivalent to international ETS offsets might be built into a CT and administered. Again some separately administered mechanism might have to be developed.

This is a digestive process of ruminant animals, such as cows and sheep, by which microorganisms break down carbohydrates into simple molecules for absorption into the bloodstream.

See Keen et al. (2013) for further discussion.

Another form of leakage occurs if a US carbon tax reduces US energy demand and in turn reduces world energy prices, encouraging higher energy use elsewhere. It is not possible to address this through tax adjustments.


    Other Resources Citing This Publication