8 The Assignment of Oil Tax Revenue
- Jeffrey Davis, Annalisa Fedelino, and Rolando Ossowski
- Published Date:
- August 2003
This paper provides a conceptual statement of the considerations that might lead one to conclude that revenues from the taxation of oil should be assigned to one level of government or the other in a multilayer system of government and discusses techniques that might be used to implement assignment of revenues to a subnational level of government. The assignment of oil tax revenues is one aspect of the theory of tax assignment, the question of who (which level of government) should tax what, and how.1 In this case what to tax (oil) is given, and we need only ask who should tax it, and how. Revenues from oil could be assigned to only one level of government or to both central and subnational governments. The paper focuses on whether subnational governments should have the power to tax oil, why, and (if so) how.
The analysis presented here is best applied before oil is discovered. In that context, decisions on revenue assignment can be made behind the “veil of ignorance,” not knowing how much revenue will be at stake or which will be the oil-rich jurisdictions. Regional vested interests will not yet have arisen and a nationally oriented view of costs of compliance and administration, of distributional equity, and of allocative efficiency is possible. Once oil has been discovered, the political dynamic is likely to be very different. If reserves are extremely valuable, and especially if they are highly concentrated geographically, regional interests will come to the fore and a different view of distributional issues—the distribution of revenues between the central and subnational governments and among subnational governments—is likely to dominate all other issues in political discussions. When the stakes are high, such discussions may become especially acrimonious.2
The paper concentrates on the division of taxing powers between the central government and the highest subnational tier of government. Though written in the context of federal systems, for the most part and with obvious modifications, the ideas presented here are equally applicable to unitary systems. Also, it is assumed that enough revenue is at stake to make the question important. Finally, as in most other normative literature on fiscal federalism, it is assumed that subnational governments respond to the wishes of their constituents.3
I. Forms of Revenue Assignment
Before turning to the question of who should tax oil, it is useful to ask how subnational governments might tax oil and how oil revenues might otherwise be channeled to subnational governments, if there is a desire to do so. The aspects of revenue assignment to be considered, aside from the crucial question of who gets the money and under what conditions, are who defines the tax base(s), who sets the tax rate(s), and who administers the tax(es). Together these determine how much fiscal autonomy subnational governments have in raising revenue from oil.
The following five alternatives are distinguished, ranked in order of subnational control over the power to raise revenues from oil.4 Except for the first, each of these may merge into the one below, depending on specific details.
Royalties from subnational ownership of resources. If subnational governments own natural resources, revenues from exploiting the resources accrue to them as owners, in the form of royalties. They have no need to use taxation to gain these revenues, although they may also receive oil revenues in one or more of the following four ways. This describes the situation in Canada and to a limited but important extent in Alaska, but, for the most part, not in the rest of the United States.5 There is no reason, in principle, that national taxation could not reduce the amount subnational governments can realize from the ownership of oil reserves. This case is not discussed at length, but many of the same concerns that are relevant for tax assignment (e.g., vertical fiscal imbalance and horizontal fiscal disparities) are as relevant for assigning ownership of natural resources as for tax assignment.
Subnational legislation and implementation of taxes. Subnational governments may have the legal right to legislate and collect taxes on natural resources. This is the situation in the United States and Canada. Subnational taxing powers may be exclusive or concurrent with national powers, and constitutions or national laws may limit such legal rights. For example, there may be limits on the types of taxes subnational governments can impose on oil, on tax rates, and even on the definition of the tax base. For example, it might be required that key provisions such as depletion allowances follow national law, or the required conformity might be more far-reaching. Conformity facilitates compliance and administration, without seriously compromising subnational fiscal autonomy. Conformity is not required in the United States, where the lack of uniformity across jurisdictions (for example, in corporation income taxation) complicates compliance and administration, as well as creating undesirable gaps and overlaps in state tax bases. Unless severely restricted, this approach provides maximum subnational autonomy in taxation of oil.
Subnational surcharges on national taxes. Conformity may be so comprehensive that subnational governments utilize the same tax base as the national government. This may occur voluntarily, mandatorily, or as a requirement for tax administration by the national government. The result, either de jure or de facto, is a subnational surcharge on the national tax. While fiscal autonomy of subnational governments may be compromised in the last two situations, the most important aspect of autonomy, the power to set subnational tax rates and obtain revenues from oil, is retained in all three situations.
Subnational governments may collect surcharges on the national tax, or the national government may collect surcharges for them. Federal administration of provincial income tax surcharges is available on a voluntary basis to Canadian provinces that choose to adopt the federal income tax base. Other provinces, while administering their own taxes, rely heavily on the federal definition of taxable income.
Subnational surcharges may be imposed on either the national tax base or the national tax liability. Subnational autonomy is greater if the surcharge is levied on the national tax base than if it is levied on the amount of the national tax, as in Canada until recently. In the latter case the structure of subnational taxes depends on the structure of the national tax.
Tax sharing. The term tax sharing is used in this paper to describe the practice of returning revenues from particular taxes to subnational jurisdictions on a derivation basis, that is, to the jurisdictions where the revenues originate (or are deemed to originate). Conceptually this system can be viewed in either of two ways: as a surcharge system in which the subnational government lacks the power to choose the tax rate or, because subnational governments lack all aspects of fiscal autonomy in the raising of revenue, as tantamount to a grant system. Sharing rates can be the same in all jurisdictions, or they can vary across jurisdictions, as in the former Soviet Union. In the latter case tax sharing shades into revenue sharing and the similarity of tax sharing and grants is clear.
Revenue sharing. In revenue sharing, revenues are shared with subnational jurisdictions on the basis of a formula, rather than being channeled to their jurisdictions of origin, as in the techniques considered above. As with tax sharing, revenue sharing allows no subnational fiscal autonomy.6 Revenue sharing is one way to offset vertical fiscal imbalance. The central government can also share its oil revenues with subnational governments in a way that offsets horizontal fiscal disparities created by the geographical concentration of oil.
Tables 8.1 and 8.2 summarize the key features of the last four methods of assigning revenues to subnational governments. The key points are (i) that only separate legislation and administration and subnational surcharges provide subnational autonomy over tax rates, the key to subnational fiscal autonomy; (ii) that independent legislation and administration may entail an unacceptable level of duplication and cost of compliance and administration and is thus not likely to be appropriate for many countries, especially developing countries; (iii) that, except in the case of revenue sharing, revenues flow to jurisdictions of origin, which may create horizontal fiscal disparities; and (iv) that only revenue sharing (and horizontal equalization) can be used to offset horizontal fiscal disparities.
|Method of Revenue Assignment|
|Choice of tax base||Subnational||National||National||National|
|Choice of tax rate||Subnational||Subnational||National||National|
|Method of Revenue Assignment|
|Subnational Fiscal Autonomy over|
|Duplication/Costs of Compliance|
|and Administration||Potentially high||Low||Low||Low|
|Reduction of Fiscal Disparities||None||None||None||Possible|
II. Economic Considerations in Revenue Assignment
Several types of economic considerations are relevant for revenue assignment.
Benefit taxation. Subnational governments should have access to taxes (or fees or other charges) that compensate them for benefits provided to (and other costs imposed by) those who exploit their natural resources. Examples of costs that may be incurred by subnational governments or their citizens include the costs of roads needed to carry heavy machinery, costs of hospitals needed to care for those injured on the job, and costs of health problems caused by damage to the environment. Taxes (and fees and charges) of this type should be levied, to the extent possible, in a way that reflects costs incurred by subnational governments or their residents. This implies that taxes on resource rents or the value of production are not likely to be optimally designed for this purpose; taxes on the volume of production are likely to be much more appropriate. This type of tax is not examined further in what follows.
Subsidiarity. National governments ordinarily have greater capacity to raise revenue than do subnational governments. This, combined with the advantages of decentralized government, suggests that subnational governments should have access to revenues from any tax that is not inappropriate for them, for example, because it interferes with international trade. This principle, commonly called subsidiarity, implies examination of other considerations to see what kinds of taxes are inappropriate for use by subnational governments. The implications of geographic concentration of oil revenues are first ignored and then these implications are considered. Legal and political issues are addressed in the next section.
Considerations That Are Independent of Geographical Concentration
Unstable revenues. Oil revenues are notoriously unstable. National governments are in a better position to offset this instability than are subnational governments, for example, because they have larger budgets (in relation to GDP), more sources of non-oil revenues, greater access to credit markets, and the power to engage in monetary policy. They may also be better equipped, technically as well as politically, to withstand the pressure to increase spending when current revenues are high (and even to borrow against future revenues) than are subnational governments, but this is not inevitable.
Taxation of resource rents. Taxation of resource rents is (aside from benefit taxation) the preferred form of oil taxation, as it minimizes distortion of economic decisions. It is also fairly complicated to implement.7 Thus national governments are likely to be more able to implement rent taxes than are subnational governments, which are more likely to rely on production taxes—taxes that encourage inefficient “high-grading” (cessation of production once revenues no longer cover costs, including taxes).8 Finally, rent taxes are likely to be more unstable than other forms of taxation of natural resources. This feature also makes them inappropriate for subnational governments.
Ring-fencing. If a tax on natural resources is to be economically neutral, it is important to allow losses incurred in one period to be carried forward or back to offset income earned in another period. The question of whether to allow taxpayers also to offset expenses incurred in one area against revenues gained in another raises interesting issues that cannot be addressed here. Subnational governments are unlikely to allow a deduction for expenses incurred in another jurisdiction. National governments can either allow it or not, via ring-fencing.
Locational mobility. Subnational taxation of economic activities that are geographically mobile can induce inefficient tax-motivated locational decisions. Subnational taxation of oil has the advantage, compared to many other types of taxes, that oil resources are geographically immobile.
Tax exporting. Taxes are exported when they reduce the real incomes of nonresidents of the taxing jurisdiction (perhaps by reducing the income of companies owned by nonresidents). Tax exporting is unfair, unless the exported taxes compensate for benefits received by (or costs due to) those who bear the burden of taxation. Also, if a subnational jurisdiction can export its nonbenefit taxes, it will have an incentive to expand public services beyond the point where marginal social benefits of public services equal marginal social costs of taxation, because only taxes paid by residents will be considered in making decisions on the proper level of taxation and spending.9 This suggests that subnational governments should not levy taxes that would be exported.10
Tax exporting to consumers of oil is probably less pervasive than sometimes thought. The fact that oil is exported does not imply that oil taxes can be exported. That depends on the degree to which the taxing jurisdiction dominates the world market. With few exceptions the requisite market domination is lacking.11
Exporting to nonresident owners of oil may occur if oil reserves are privately owned. Exporting to owners of concessions to exploit oil reserves may occur if unanticipated tax increases are not contractually precluded.12
Duplication of administration and compliance. Concurrent national and subnational taxation can create duplication of costs of compliance and administration. Compliance and administration can be further complicated if subnational jurisdictions do not utilize the same taxes, define tax bases the same way, or cooperate in tax administration. Tax surcharges, tax sharing, and revenue sharing minimize duplication and complexity.
Vertical fiscal imbalance. If oil revenues were to be assigned entirely to subnational governments, an unusual form of vertical fiscal imbalance might develop, in which subnational governments have adequate revenues, but the national government does not. Under such circumstances it might be appropriate to assign part of revenues to the central government.
Considerations That Depend on Geographical Concentration
Horizontal fiscal disparities. Assignment of oil revenues to the jurisdictions where they originate allows resource-rich jurisdictions to provide a given level of public services with non-oil taxes that are lower than those levied elsewhere or to provide higher levels of public services than elsewhere with a given level of non-oil taxes. This result can be challenged on grounds of both distributional equity and economic efficiency.
Distributional inequity. The implications of horizontal disparities for distributional equity are fairly apparent;13 it may seem unfair that residents of oil-rich jurisdictions receive more public services or pay lower taxes (or both) than residents of other jurisdictions.
Uneconomic locational decisions. The lower taxes and greater public spending made possible by oil revenues may cause more labor and capital to be attracted to oil-rich jurisdictions than is optimal.14 Empirical evidence suggests that these distortions may not be a significant problem, except where extraction of oil is highly concentrated in jurisdictions with extremely low population density, such as Alaska.15
Wasteful government expenditures. Another implication of horizontal fiscal disparities is that oil revenues may be spent on government programs and projects that have lower social value than if spent elsewhere. A particularly popular form of waste is local investment of oil revenues, particularly in uneconomic “downstream” processing activities.16
Equalization: Responding to Horizontal Fiscal Disparities
Where assignment of oil revenues to subnational jurisdictions creates unacceptable horizontal fiscal disparities, equalization is likely to be advocated. This is where revenue sharing comes into its own. All of the other forms of revenue assignment examined here channel revenues to the jurisdiction where they originate and are thus likely to create horizontal fiscal disparities. By comparison, revenue sharing can be designed to channel revenues to “have-not” jurisdictions and thus reduce horizontal disparities.
III. Legal and Political Considerations
Most of the considerations examined above suggest that revenues from oil (other than those required to offset costs of providing public services and social costs imposed by exploitation of oil resources) should be reserved for national governments. But no matter how the issues examined above are resolved, there may be overriding legal and political issues.
First, a nation’s constitution may assign the power to tax oil to subnational governments. This fact may trump all the considerations discussed thus far.
Second, subnational governments may claim that natural resources are the “patrimony” of the subnational jurisdictions where they are found.17 This raises the interesting philosophical question of whether the patrimony at issue belongs to the residents of the nation or to those of the subnational jurisdiction.
Third, the very continued existence of a federation may depend on whether subnational governments are allowed to tax oil. This issue is particularly important when it arises in combination with separatist tendencies based on ethnic differences, as in the case of Tatarstan in Russia.18
Fourth, subnational governments where oil revenues originate may simply not trust the national government to deliver revenues due to them under systems based on tax surcharges and tax sharing. They may prefer to administer their own taxes, even if the definition of their tax base is severely constrained.
BoadwayRobin W. and Frank R.Flatters1982“Efficiency and Equalization Payments in a Federal System of Government: A Synthesis and Extension of Recent Results,”Canadian Journal of Economics Vol. 15 No. 4 (November) pp. 613–33.
CareagaMaite and Barry R.Weingast2001“Fiscal Federalism, Good Governance, and Economic Growth in Mexico”in In Search of Prosperity: Narratives on Economic Growthed. byDaniRodrik (Princeton: Princeton University Press) pp. 399–435.
GelbAlan and Associates1988Oil Windfalls: Blessing or Curse (New York: Oxford University Press for the World Bank).
LinkArthur A.1978“Political Constraint and North Dakota’s Coal Severance Tax,”National Tax Journal Vol. 31 No. 3 (September) pp. 263–68.
LitvackJennie I.1994“Regional Demands and Fiscal Federalism,”in Russia and the Challenge of Fiscal Federalismed. byChristine I.Wallich (Washington: World Bank) pp. 218–40.
McLureCharles E. Jr.1983“Fiscal Federalism and the Taxation of Economic Rents,”in State and Local Finance: The Pressure of the 80’sed. byGeorgeBreak (Madison: University of Wisconsin Press) pp. 133–60.
McLureCharles E. Jr.1994“The Sharing of Taxes on Natural Resources and the Future of the Russian Federation,”in Russia and the Challenge of Fiscal Federalismed. byChristineWallich (Washington: World Bank) pp. 181–217.
McLureCharles E. Jr.1998“The Revenue Assignment Problem: Ends, Means, and Constraints,”in Public Budgeting and Financial Management Vol. 9 No. 4 (Winter) pp. 652–83.
McLureCharles E. Jr.1999“Tax Assignment,”in Fiscal Transition in Kazakhstan (Manila: Asian Development Bank) pp. 273–316.
McLureCharles E. Jr.2000“Tax Assignment and Subnational Fiscal Autonomy,”Bulletin for International Fiscal Documentation Vol. 54 No. 12 (December) pp. 626–35.
McLureCharles E. Jr.Christine I.Wallich and Jennie I.Litvack1995“Special Issues in Federal Finance,”in Decentralization of the Socialist State: Intergovernmental Finance in Transition Economiesed. byRichard M.BirdRobert D.Ebel and Christine I.Wallich (Washington: World Bank) pp. 379–404.
MieszkowskiPeter and EricToder1983“Taxation of Energy Resources,”in Fiscal Federalism and the Taxation of Natural Resourcesed. byCharles E.McLure Jr. and PeterMieszkowski (Lexington, Massachusetts: Lexington Books) pp. 65–91.
MusgraveRichard A.1983“Who Should Tax, Where, and What?”in Tax Assignment in Federal Countriesed. byCharles E.McLure Jr. (Canberra: Centre for Research on Federal Financial Relations) pp. 2–19.
Prud’hommeRémy1995“On the Dangers of Decentralization,”World Bank Economic Review Vol. 10 No. 2 (August) pp. 201–20.
The high energy prices of the 1970s, which saw consumers in the United States paying higher prices while energy-rich jurisdictions enjoyed large increases in revenues, led to references to a “new war between the states.” References in consuming states to “blue-eyed Arabs” were countered in Texas with bumper stickers that said, “Let the bastards freeze in the dark.”
Prud’homme (1995) questions the relevance of this basic assumption for many developing countries.
There is, unfortunately, no uniformly accepted terminology in this area. These terms are used below because they clearly and accurately describe most of the alternatives.
In the United States, unlike much of the rest of the world, private ownership of mineral rights is common, except in the case of public lands.
Revenue sharing may also create less incentive than the other techniques for subnational governments to spend public funds in ways that enhance productivity. The techniques that assign revenues to the jurisdictions where they originate create incentives for subnational governments to spend public funds in ways that enhance productivity within their boundaries, as that makes it easier to raise revenues. By comparison, revenue sharing generally creates smaller incentives of this type, as the amount of funds available to subnational governments depends less on efforts to enhance productivity, and may even be negatively affected by them. See Careaga and Weingast (2001). The ease of raising revenue would seem to be a more important consideration in societies where governments do not respond to the will of their constituents than in those where they do. The incentive to spend revenues in ways that increase job opportunities, wages, the return to capital, and the value of land—considerations of prime importance to the electorate—would exist under both revenue sharing and techniques that channel revenues to subnational governments of jurisdictions where they originate.
McLure (1994) explains the difficulties of implementing three types of taxes on economic rents. These include standard problems of income taxation, such as the need to deal with transfer pricing on transactions between related parties, timing issues (in the recognition of income and expenses), and perhaps adjustment for inflation. In addition, since economic rents are the excess of returns over all costs, including the imputed cost of equity capital, it is necessary to know the normal return to equity investments.
Even worse, they may rely on property taxes, which encourage early depletion.
The fact that various jurisdictions may reciprocally export taxes to residents located in other jurisdictions within the same nation does not imply that tax exporting does not have the efficiency effects indicated. Imported taxes have only an income effect, whereas tax exporting also has a substitution effect, by affecting the tax price of public services.
By comparison, few would suggest that national governments should not levy taxes that would be exported, merely because it reduces the local cost of public services. The difference in answers to what is seemingly the same question depends on the difference in the two policy contexts. Both ask what is the best fiscal arrangement from a national point of view.
This statement assumes that the consequences of the actions of a single jurisdiction are being examined. When oil-producing jurisdictions act collusively to raise taxes (for example, under the auspices of OPEC), forward shifting is more likely. Note, however, that if the actions of OPEC lead to higher prices and higher profits of oil companies and a single country raises its taxes to capture some of the profits that result, the tax is not shifted to consumers. It is the actions of OPEC not the tax, that causes prices to be higher.
The availability of foreign tax credits (FTCs) for source-country income taxes creates a particularly common form of tax exporting, to the treasuries of home countries of multinational oil companies. This is not unique to oil taxes and thus is not considered further.
Actually, the implications for equity may be more apparent than real if factors of production (capital and labor) are mobile. As with many other instances of differential taxation, what appear on the surface to be horizontal inequities (which they are, before adjustments are made to them) may be reduced dramatically or eliminated by adjustments. That is, once capital and labor have adjusted to the differences in wage rates, returns to capital, taxes, and levels of public services, little or no inequity may remain. What remains is likely to be reflected, if at all, in land values.
See Mieszkowski and Toder (1983). The fact that Alaska has a relatively inhospitable climate presumably reduces the likelihood that locational decisions will be badly distorted.
This is common even at the national level. See Gelb and Associates (1988). The very different experiences of the state of Alaska, which followed a rational investment model, and of Alberta and the native peoples of Alaska, which followed the “development” model, is instructive; Alaska realized a far higher rate of return. See McLure (1994) and literature cited there.