Information about Asia and the Pacific Asia y el Pacífico

13 Issues in Natural Resource Taxation

Ehtisham Ahmad, Vito Tanzi, and Qiang Gao
Published Date:
September 1995
  • ShareShare
Information about Asia and the Pacific Asia y el Pacífico
Show Summary Details
Tamar Manuelyan Atinc and Bert Hofman* 

Resource taxation plays a small role in China compared with other forms of taxation. In 1991, it constituted 0.6 percent of total tax revenues. Introduced at the time of enterprise income taxation reforms in 1984, the resource tax differs from the tax in many other countries in that it is paid by the users of natural resources rather than the producers; it could thus be classified as a product tax rather than a natural resource tax. This is not to say that resource taxation has been absent: the pricing mechanism, which has maintained relatively low prices for raw materials, has acted as an implicit taxation on natural resource exploitation, combined with a subsidy for the user of these resources (Table 1). In fact, the implicit marginal tax rate on production of coal mining and oil drilling has contributed to the Y 12 billion in losses in that sector, or about one third of all losses by state-owned enterprises in 1991. In 1991, about 50 percent of total coal production was traded at planned prices, amounting to an implicit subsidy (or forgone tax revenue) of about Y 16 billion, or 1 percent of GNP.1

Table 1.Chinese and International Prices, 1991
Beijing Retail1
Price (Y/ton)InternationalBorder Price2
PlanNon-Plan(U.S. dollar/ton)(Y/ton)
Steam coal75–80120–12553186
Diesel oil500–700900–1300182965
Fuel oil350–40070080–94424–498
Source: World Bank, (1992).

Prices quoted by fuel supply bureau, November 1991.

Converted with official exchange rate of Y 5.3/US51.

Source: World Bank, (1992).

Prices quoted by fuel supply bureau, November 1991.

Converted with official exchange rate of Y 5.3/US51.

Planned prices are rapidly adjusted to market levels,2 and during ongoing price reforms, a further alignment of resource prices with world market prices can be expected; rents from resource exploitation will accrue to the exploiting firms, if no further measures are taken; and natural resource taxation will play an increasingly important role in guaranteeing to the government, as the owner of the resource, an appropriate reward for the depletion of the resource.

China’s government recently announced that a resource tax would be levied on the recovery of petroleum and metallic and nonmetallic mineral resources,3 a decision that should contribute to the optimal management of China’s resource base. Changing from implicit to explicit taxation of natural resources raises a number of policy issues, to be discussed in this paper:4 (1) the valuation of natural resource rents that should accrue to the government as owner; (2) the determination of the optimal depletion rate and the impact of tax policy on depletion; (3) the treatment of externalities of resource exploitation; and (4) the appropriate distribution of natural resource revenues among the different levels of government.5

In view of the pervasiveness of state ownership of natural resources in both developed and developing countries, the government has a dual role to play. As resource owner, it has to ensure that it receives a competitive return on its natural resource endowment. As tax authority, it has to be concerned with the overall allocative effects of tax policy and has to ensure that different sectors of economic activity receive uniform treatment under the tax code.6 For China, a third role can be distinguished: the role of owner of the resource-exploiting enterprise. There are many reasons to separate the role as enterprise owner from that of resource owner, and China is currently developing mechanisms by which state-owned enterprises work at arm’s length from government.7 Moreover, in the future, use of foreign or Chinese nonstate enterprises may be used for natural resource exploration. Therefore, we assume that the exploiting agent behaves as if it were a private sector agent, with profit maximization as its only goal.

Government as Owner of Resources

Economic Rents and Resource Rents

If resources are extracted, the government as owner of the resource needs compensation for the loss of resource reserves. Whether resources should be extracted depends on the value of the resource in the ground compared with the value of the resource exploited. The value of the resource exploited is the market value minus all costs necessary for exploitation, which is defined as the economic rent.

The value of resources in the ground equals the shadow price (or user costs) of the resource, which is determined by the theoretical price that should be charged to induce a program of exploitation that maximizes social benefits from exploitation. In an optimal program, the resource should in the end be completely exploited, and the returns made on exploiting the resource should equal those of leaving them in the ground. That revenues made with exploitation could be invested at the market interest rate implies that the shadow price of the resource should increase with the interest rate in an optimal extraction program.8 If the rate of price increase is lower, the government would have an interest in increasing the extraction and investing the ensuing income flow in other assets with higher rates of return. If it is higher, the resources are better left in the ground and exploitation postponed until such time as the rate of return on the resource flow is higher than the (social) discount rate. Note that alternative uses of the land, such as agriculture, may yield higher returns than resource exploitation. In this case, the shadow price is too high to validate exploitation.

The shadow price of the resource is often labeled “resource rent,” or “royalty.” It is the minimum payment the government needs to be compensated for the loss of its natural resource and is similar to any factor payment. Resource rent and economic rent may differ, just like the market wage rate may be higher than the minimum wage an individual worker would accept for his labor. Resource rents and economic rents will also differ from project to project, since the costs of extraction, the quality of the resource, and the location of the mine or well differ, which makes economic rents per mine differ. The marginal mine yields just enough revenues to compensate for exploitation costs and resource rents.

In the absence of an asset market for trading mineral rights, how can the government determine the royalty price of natural resources? The royalty price, or the resource rent, is the opportunity cost of the resource in the ground, or the amount by which the present value of the stock declines when the unit is removed. Under competitive conditions, this would be given either by present value of the net benefit stream associated with exploitation, adjusted for risk, or by the alternative uses of the land, whichever is higher. The unit value of the resource at the minegate can be determined with reference to the world market price. If the opportunity cost of exploitation is higher than the valuation of the resource at the mine, the resource is better left unexploited, until such time that world market prices are sufficiently high to cover the opportunity costs.

Extracting Rents

The government needs to ensure that it will at least receive the opportunity costs of the resource. The way this is done, however, could influence the behavior of the exploiting enterprise, and with that the total amount of economic benefits generated. The government thus needs to take account of the incentives it offers for the exploiting enterprise. Before undertaking an investment, the exploiting enterprise has to determine whether the income generated by the project yields a competitive rate of return on physical capital employed, adjusted for risk, after payment to third parties, including remuneration of other factors (labor, resources) and taxes. Its evaluation will be based on the objective that maximizes the net present value of the project, which will involve judgment about the optimal extraction rate given the expected evolution of prices, costs, and government tax policy. Each tax diminishes returns to the enterprise and increases it for the government, but if the enterprise is taxed too high, the project will no longer be financially viable to the firm. The government and the exploiting enterprise engage in an explicit or implicit contract that includes tax payments, regulatory policies, and possible government equity participation; each party to the arrangement needs to be interested in the arrangement over the duration of the project. The ideal arrangement would induce the enterprise to exploit the resource such that it maximizes social benefits from exploitation.

What Tax?

Ideally, the government should tax the resource rents with a tax that takes economic rents as a base. A pure rent tax does not distort investment and production decisions: thus it carries no excess burden. Unfortunately, it is virtually impossible to calculate economic rents, because this requires information on accrued costs, including imputed capital costs to the firm that are not directly observable, such as real depreciation of assets, depletion of resources, real financing costs, and risk.9 Moreover, these costs may differ substantially among firms and among resource bases. Cash flow taxes are an attractive alternative, both economically and administratively. They are simple to calculate because they do not involve imputation of capital costs; capital costs are taken into account as they occur and deductions are taken in full up front for depreciable assets. Economically, cash flow taxes are equivalent in present value to rent taxes calculated on accrued net income.10

In practice, cash flow and rent taxes are rarely used to capture resource rents. Cash flow taxes reduce tax liabilities up front, which might be problematic for some governments. Moreover, both types of tax are susceptible to manipulation by international, vertically integrated companies through transfer pricing practices or debt-equity choice, and both taxes imply a high degree of risk for the government. The cash flow tax can be modified, however, and the development costs may be capitalized and depreciated once the cash flow from the natural resource exploitation becomes positive. This smooths out the tax obligation of the resource-exploiting enterprise. If interest over the full book value of the capitalized development costs can be deduced, the system approaches an ideal cash flow tax. The modified cash flow taxation differs from a profit taxation in that income taxation usually does not allow capitalization of all costs.

Royalties (severance taxes, production taxes, stumpage fees) that are levied on each unit extracted can theoretically be constructed such that they capture only rents of exploitation. The ideal (maximum) royalty would then be the difference between the mine gate price and the costs of extraction.11 Actual royalties or severance taxes hardly meet the requirements that would make them a rent tax. They can be levied either as ad valorem or per unit taxes on production. Both types of tax reduce economically recoverable reserves and create incentives for “highgrading” of deposits, leaving economically profitable lower-grade deposits unexploited. Unit taxes do not take account of either the value of the resource or the extraction costs. They induce producers to change the quality selection or extraction profile in order to reduce the present value of taxes. Flat unit taxes postpone extraction, while the effect of progressive unit taxes depends on the rate of increase of the tax compared with the interest rate used by the producer to discount future income flows. Ad valorem taxes do take account of the value of the resource but they induce the producer to allocate extraction to periods with lower prices in present value terms, and thus future resource prices determine acceleration or deceleration of extraction. In addition, ad valorem taxes create risk sharing between government and producer as tax revenues are dependent on market prices. Production sharing is analogous to an ad valorem tax when the government simply appropriates a certain share of total output. As such, production-sharing arrangements have an element of risk sharing.

When the country is a price taker, export taxes have the same effects on the producer as do production taxes, discussed above. Export taxes, however, have the additional disadvantage of diverting resources for domestic use, representing an implicit subsidy from the government to processing industries (e. g., Cameroon for timber). In China, where the internal price differs from the world market price, the imposition of an export tax may be necessary in order not to distort against domestic supply. However, the first-best solution to stimulate domestic processing would be to liberalize prices and subsidize processing directly.

Modified Royalties

The output taxes discussed above result in suboptimal extraction profiles, because typically they ignore some extraction costs. Pure rent taxes should be levied on net revenues after deduction of all operating and capital expenses. Some production taxes and production-sharing arrangements allow for the deduction of some costs (typically, operating costs) and thus approximate a rent tax. Notwithstanding the disadvantages of simple royalties or production taxes, they are often used by government to capture part of the rents. The main advantage is that they are easy to monitor, and no assessment needs to be made on the costs incurred by the exploiting enterprise. However, more complex royalties that take some cost factors into account could be based on relatively simple, transparent formulas that limit monitoring difficulties.

Income-based taxation is sometimes used for rent taxation. Resource properties are usually subject to general income taxation, but taxes specific to the resource industry are sometimes based on an income concept. If some allowance for the capitalization of the development costs is made, this approaches the cash flow taxation. Colombia uses an income-based tax for coal and Indonesia, for hard minerals, the United Kingdom for oil, and in Canada and the United States, the corporate income tax makes special allowances for resource-exploiting firms. In some of these countries, however, the rate of return on equity in resource exploitation firms has been very high, and the rents have apparently not been captured very effectively.12

Nontax Instruments

An auction or competitive bidding for the sale of exploitation rights offers a mechanism for the government to capture rents. If there are sufficient bidders to prevent collusion and information is uniformly available, the bids reflect the present value of the future net benefit stream, adjusted for risk. This is exactly how resource rents were defined earlier. Thus, a properly designed and implemented auction system can capture 100 percent of resource rents for the government and would solve the government’s problem of assessing the rents of a certain resource project. Leasing is sometimes used by governments to allocate exploitation rights, for example, in the petroleum sector. Leasing has the same effects as an auction and therefore a rent tax if the leasing period is indefinite. Shorter periods will distort incentives for extraction rates, but may be chosen for political reasons, if nonstate ownership of the resource is considered not acceptable. For an investor, however, leases and auctions may not be the ideal way to pay the rents that are due to the government. Because the auction or lease price is paid up front, the exploiting firm becomes vulnerable to expropriation of the profits after the investment has been made. This would reduce the price the enterprise is willing to pay for the right of exploitation.

Direct equity participation in production is another way for the government to appropriate resource rents. The price at which the government purchases equity relative to its market price determines the level of rents captured by the government. The government’s equity participation can be structured and priced such that it is equivalent in effect to a pure rent tax,13 but direct participation exposes the government to the risks associated with the exploitation of natural resources. It is in this aspect the same as a cash flow or rent tax. The notable difference is that equity participation confers an active role on the government as shareholder, while taxation relegates government to a passive role typical for a tax authority in a mixed economy.


Proper remuneration for the exploitation of resources may involve more than payment of the royalty price, reflecting the scarcity value of the resource. The opportunity cost of extraction may also involve externalities generated by the extraction process (e. g., land erosion, environmental degradation due to open pit mining) that reduce the value of other assets. These economic costs should be taken into account, and tax policy in the mineral sector should be modified to induce socially optimal adjustments to extraction profiles.

Dutch Disease

An economic cost reminiscent to externalities is the “Dutch disease,” the negative impact of mineral production on the development prospects of the rest of the economy.14 The problem arises when the increase in wealth experienced by the discovery of natural resources induces an appreciation of the real exchange rate. The change in the country’s comparative advantage and the increase in the real exchange rate reduce the competitiveness of traditional exports, potentially leading to unemployment and idle capacity in this sector. If the traditional export sector shows increasing returns in production, a temporary reduction in production due to the resource boom may permanently affect competitiveness. The relative importance of Dutch disease depends on the extent of the economy’s dependence on natural resources. A diversified economy with a relatively small share of its output and exports derived from natural resources is unlikely to experience serious problems, while for smaller resource-dependent economies, it may wreak havoc with their agriculture sector in particular.15 While a country will be unable to avoid the problems generated by a sudden increase in wealth, the government can compensate for the associated costs in the traditional sectors and prepare the economy for the transition away from a resource-dependent production structure.

Production taxes will induce producers to take externalities into account. While these taxes increase costs and therefore influence investment and extraction decisions, the modification in behavior is desirable in the presence of externalities, and thus welfare improving. The cleanup costs of environmental damage require the accumulation of production taxes, or the posting of a bond by the exploiting enterprise; otherwise the enterprise will be tempted to walk away from those costs by going bankrupt.


The government as owner of the resource is exposed to risk arising from the uncertainty on, inter alia, the size and quality of the stock and the future evolution of natural resource prices and extraction costs. Various possibilities exist for risk-sharing arrangements between the owner and the exploiter of the resource. The government’s willingness to share risks depends on the level of diversification of the economy and on access to international capital markets.16 More diversified economies will experience smaller income fluctuations because of resource price fluctuation. Economies with greater access to capital markets are better able to sustain greater exposure to the natural resource sector than less creditworthy economies, because income fluctuations can be smoothed out. The risk attitude of the government will have an impact on the design of the contractual relationship with the producer, as well as on the acceptable rate of return on resources. With uncertainty, the government maximizes the expected rate of return on its assets. The government equity participation is one way of risk assumption, while ex post rent taxes and ad valorem and ad rem output taxes have different implications for the extent of risk sharing.

In conclusion, government as owner of the resource has to be concerned with putting in place a policy framework that will provide adequate renumeration for the loss of an asset and compensate for any economic costs associated with mineral development. The rate of return that the government is willing to accept on its natural resource endowment will be related to the rate of return on other assets, adjusted for risk. Proper evaluation of natural resource rents and economic costs will lead to the optimal rate and timing of extraction to maximize economic welfare.

Government as Tax Authority

In its role as tax authority, the government has to be concerned with the overall allocative effects of tax policy. As discussed above in the section on externalities, tax policy can have an efficiency-enhancing impact when used to correct for market imperfections. This important governmental role notwithstanding, tax policy is typically concerned with devising a system for generating sufficient revenues to finance public expenditures in a manner that minimizes distortions in decisions of economic agents regarding investment, output, and production techniques (relative use of inputs). The previous section discussed the problems caused by certain taxes levied for capturing resource rents on incentives within the resource industry. In this section, the treatment of the resource sector in the economy-wide tax system is discussed, underlining the importance of uniform treatment across sectors.

A typical feature of the system of direct taxes in most economies is taxing capital income in addition to personal income. As part of the net of economy-wide taxes, natural resource industries are subject to the corporate income tax, which represents a tax on capital income earned in corporations. To preserve the neutrality of economic incentives for the allocation of investment capital among the different sectors, industries should be taxed in a homogeneous fashion. In practice, however, resource industries face a reduced rate of effective taxation, as they benefit from special privileges including rapid expensing of exploration and development expenses when the rest of the economy is operating on the accrual method of determining taxable income. Additional provisions including depletion allowances and interest deductibility reduce further the rate at which equity income is taxed relative to other industries. This has implications for investments across sectors.

Other Goals

Governments often seek to achieve other goals than rent appropriation with natural resource taxes. The taxes on resources are often set differently for different uses, most notably a lower tax for domestically processed resources. Local employment goals are often pursued, and resource-exploiting enterprises may be “taxed” by forcing more people than necessary onto their payroll. Also, taxes may be set lower than the opportunity costs of the resource, in order to develop resources that are not economically viable, but may secure supply of a crucial resource. In general, taxation is less suited to achieve such goals, as they only indirectly influence the behavior of the exploiting enterprise, and more direct measures, such as direct subsidization of the desired activity, are often the better policy.

Use of Revenues

The government should realize that, in using the natural resource in exploitation, it loses an asset, which cannot be used in the future, and in an economic sense this is equal to accumulating debt. If no alternative assets are bought for the revenues, which are instead used for government consumption, the net worth of government will fall, and future generations will have fewer assets than the present one. The distribution over time of the benefits to some extent is a moral issue, but a good rule of thumb is that the revenues from natural resources should be spent such that a constant stream of future income is generated. This would, for instance, imply that most of the revenues should be invested or international debt repaid with it. The decision on spending the revenues should not be restricted by earmarking them for a certain purpose, but should be decided upon within the overall budgetary process. Appropriate spending of the revenues would also reduce Dutch disease effects.

In conclusion, the government as tax authority should ensure that different industries in the economy receive uniform treatment under the corporate income tax. There is no economic justification for providing preferential treatment to resource industries. As discussed in the first section of the paper, however, there are good economic reasons why resource industries should be afforded special treatment in the tax system. This has to do with the fact that natural resources generate rents that should accrue to the government as owner, requiring fiscal instruments that tax resources over and above the levies implicit in general income taxes.17 The important public policy issue is to devise a system for the extraction of resource rents that takes into account all economic costs as well as the relative risk profile of governments and private sector agents.

Distribution of Resource Taxes Among Levels of Government

Once it is decided how to tax natural resource rents, the question is what level of government should control natural resource taxation. The feasibility of optimal taxation of resource rents is not independent of this issue. The regional dimension of taxation is highly relevant to the Chinese context. The revenues from China’s existing resource tax belong to the taxes that are shared according to the tax contracts. In the present tax experiments, the tax is shared, with 50 percent to local government, or 70 percent for local governments of minority areas, bringing in a redistributive element in the resource taxation.

China’s fiscal system is de jure a unitary system, with both rates and base decided by central government. However, because of its decentralized control of tax collection, the system operates akin to a federal system. In addition, the recent tax experiments are moving away from the traditional tax contracting system and in the direction of a tax assignment system, with only limited sharing of taxes.

Moreover, central government has recognized the need for increased local control over a number of taxes. New taxes, like the natural resource tax, are proposed, and this broadening of the tax base may break the present deadlock between central and local governments. In this context, it should be decided what level of government will obtain control over the tax, and aspects of efficiency, stabilization, and equity need to be considered.


The accrual of natural resource rents to a subnational government would induce movements of production factors and migration into that jurisdiction, in order to capture part of the rents. From a national perspective, the costs involved in such relocation, including the crowding of the locality with the resource rents, are wasteful, as the rents could be distributed by other means, among which are equalization schemes administered by the central government. Moreover, production efficiency would be violated, if local government could lower enterprise taxation, because of natural resource revenues. In China, migration is strongly discouraged by existing regulation, but this has not stopped many people from moving out of the inland rural areas to the coastal cities in order to benefit from the prevailing higher wages. A similar movement could be expected to localities that generate substantial resource rents. However, empirical evidence from the United States suggests that the efficiency losses due to rent-induced factor movements remain limited.18

Local governments may be less inclined to levy the optimal resource tax than the central government, as local governments have a stronger interest in such goals as reducing unemployment than in the optimal use of the natural resource.

Contrarily, natural resources would be the ideal local tax base from the perspective of excess burden: because the tax base cannot move, no excess burden from interjurisdictional movement of the tax base would arise. Moreover, the local government may be better able to determine the optimal tax, which, as shown above, depends on the productivity of the extracting firm, which in turn depends strongly on local conditions of extraction. Better information on local conditions is the standard argument for the devolution of government,19 and in this context this could imply that at least part of the authority over the tax rate could be left to local governments.

The development of natural resources can raise considerable demand for local public goods. Negative externalities from natural resource development in the form of pollution are usually strongly spatially concentrated. The cleanup costs of at least part of this would fall on local budgets. Moreover, the rents attached to the natural resource may depend on local government input, such as roads and waterway maintenance; thus giving the local government a share in these rents may align its interest with those of the central government.20 However, this compensation for the provision of local public goods to develop natural resource extraction need not take the form of taxes: the central government as an owner could provide fees or grants for local governments to achieve this goal.

In China, the present tax administration would argue against central assignment of the natural resource taxation. The double subordination of the tax bureaus means that they operate to a large extent as a local government institution. Since the local government would have only little interest in the collection of a natural resource tax that accrues to the central government, the effective tax rate would differ from the optimal tax, which would induce too high a rate of depletion of the natural resources. If the exploiting firm is locally owned, the natural resource tax would even be strictly against local interests, and exemption of the tax may occur. Thus, in present circumstances, a local share in the tax seems unavoidable from a collection point of view. If the central government succeeds in its efforts to establish an effective national tax administration,21 centralizing the revenue of a natural resource tax may become more feasible.


Local assignment of natural resource taxation may complicate economic stabilization. World market prices for natural resources are highly volatile, and so are the rents that should accrue to the government under a rent taxation arrangement. Moreover, natural resource prices tend to be high in cyclical upturns and low in cyclical downturns. Since China is opening up more and more to the outside world, the economy will fluctuate more and more in step with the international business cycle.

Local government in China does not have legal access to capital markets, and is thus more likely to base expenditures on short-term revenue considerations. The procyclical pattern of natural resource revenues is therefore likely to lead to procyclical government expenditures. To avoid this, centralization of at least a substantial portion of the revenues seems advisable. However, some of the procyclical spending pattern could be avoided, if regulated access of local government to capital markets was developed. Apart from stabilization considerations, local governments’ regulated access to borrowing would also enhance intertemporal efficiency. Besides, the local government has considerable unregulated access to borrowing in any case, owing to the intertwining of state-owned enterprises and the government. As an alternative to borrowing for mitigating stabilization problems, China’s authorities could regulate the accumulation and decumulation of reserves for local governments, and in this way avoid procyclical spending of natural resource revenues. Finally, a national windfall profit taxation, like the one in the United States, could reduce the procyclical effect of locally accruing natural resource taxation. Of course, the auctioning of mineral rights may completely isolate local governments from price fluctuation.


The standard literature on the allocation of revenue sources maintains that tax bases that are highly unevenly distributed among regions should be taxed centrally,22 because fiscal equalization would become impossible if fiscal capacities of subnational governments are highly unequally distributed.

The inequality of fiscal capacity, however, is not unique to natural resource taxation alone; thus, solely centralizing this tax for that reason may give rise to distortions. Moreover, in China, fiscal equalization is only weakly developed, and “quota grants” have been eroded by inflation over the last five years. This has substantially reduced the poor provinces’ capacity to provide an efficient package of public goods. At the same time, the internal provinces that are the main suppliers of natural resources in China belong to the poorest states. Introduction of a local natural resource tax or giving local governments a substantial share in the revenues of a central resource tax would enhance the fiscal capacity of these provinces substantially.

To illustrate the size of possible changes in tax capacity, let us assume an efficiency price of coal of Y 85 at the mine, the coal-producing provinces that export half their production are giving the rest of the country a subsidy of about Y 8 billion (1988 data). The main losers of taxing coal up to its efficiency price are Beijing, Liaoning, Jilin, and Tianjin, who together received an implicit subsidy of Y 3 billion, or about 1.5 percent of provincial GNP in 1988. The main benefactors of a tax, if revenues were to be assigned at the provincial level, would be Shanxi, with Y 4.3 billion (14.5 percent of provincial GNP); Ninxia, with Y 0.2 billion (4 percent); and Henan, with Y 0.6 billion (0.8 percent).

The first-best solution for the equalization problem is to establish an equalization scheme that takes into account total fiscal capacity and then redivides this among provinces according to objective criteria. The natural resource tax could then be one of the taxes taken into account in local fiscal capacity. Alternatively, with a central natural resource tax, the receipts could be used to replenish an equalization fund, which is disbursed according to needs criteria. The development of natural resources could then be seen as a needs factor.

International Practice

International practice on the assignment of natural resource taxation shows a variety of approaches, but in many at least part of the revenues, and sometimes the control of natural resource taxes, is assigned to subnational governments.

The United States allows states to tax natural resources, with severance taxes (production taxes, license taxes, conservation taxes), which amounted to 4.3 percent of total state revenues in 1981. The average rate for oil severance tax is 5 percent, with a range of zero to 12.5 percent. The states also benefit from natural resource exploitation through mineral rights leasing. For instance, in Alaska, the collection of mineral lease revenues amounted to $1,119 million, almost as high as the severance tax collection of $1,170 million. Further resource-related revenues for states are the corporate income tax as part of the general enterprise taxation; finally, the local government collects property taxes on the assets of natural resource enterprises.

Canada has constitutionally assigned the ownership of subsurface mineral rights to the provinces and collects royalties from them. The federal government as regulator of interprovincial trade can also tax natural resources, but traditionally, the corporate income tax is the main means by which the federal government benefits from natural resource exploitation. Alberta is the main benefactor of natural resource revenues.

Australia assigns to the states all rights not explicitly mentioned in the constitution (“residual rights”), and the ownership of natural resources is understood to be among these rights. The federal government has the rights to offshore drilling, on which it collects a royalty of 10 percent, 60 percent of which is passed to the states. In 1979, 20 percent of total revenues of the states came from natural resource taxation, as compared with 9 percent of federal revenues. The strong regional differences due to resource taxation are redressed by strong equalization components in federal grant schemes.

In 1992, Russia established a severance tax of 8–20 percent of the production value based on the domestic price, which is about one sixth of the world market price. In addition, petroleum is subject to an export tax ($50 a ton) and a “fee for the reproduction of resources” of 10 percent of production value, which is earmarked for petroleum sector investment. The export tax is a federal tax, the royalty is shared between the federation (20–40 percent), oblast (20–30 percent), and city (30 percent). The fee for reproduction of resources is assigned to cities. The substantial accrual of resource tax revenues to the local level has been primarily motivated to counter the centripetal forces in the Russian Federation.

Options for China’s Natural Resource Tax

As China’s government moves toward market-based pricing of natural resources, and resource-exploiting enterprises become more independent from government, the natural resource tax will become an important instrument for the government to appropriate resource rents.

Resource rents should ideally be taxed by pure rent taxes, but China’s low tax buoyancy may make the necessary up-front payments infeasible. Simple royalties and severance taxes would distort the production incentives for the exploiting firm, adding to an environment that is already prone to distortions due to the incomplete nature of China’s reform. The alternative of auctioning exploitation rights may not be efficient, given the limited number of Chinese bidders, the political reluctance to allow foreign firms, and the hesitation foreign firms may have in owing natural resource rights to China, due to the risk of (explicit or implicit) expropriation of the benefits.

China may want to consider a modified cash flow tax as a means to appropriate rents for the state. The concepts needed to effectuate such a tax have been developed in the context of new income tax laws, and thus natural resource tax legislation can lean heavily on the income tax law. Administration of the tax could be done fairly easily by those branches of the tax administration specializing in income taxation, and thus no separate collection units for natural resource tax need be developed. Explicit, transparent definitions of rents should be formulated and codified, which may seek a compromise between the goal of fully capturing accrued rents and limiting administrative difficulties. Although practice has shown that the adjusted cash flow tax is often unable to capture all rents, government ownership of the exploiting firm makes this less of a problem. The adjusted cash flow tax would require explicit, transparent laws and regulations.

China’s current plans seem to lean toward the introduction of royalties or severance taxes to appropriate rents. Suitable allowances for costs, both current and capital, should be made, in order to closely mimic the pure rent tax. Production taxes could be used, however, to address the negative externalities of exploitation, in which case the adjustment in exploitation profile is desired. For such taxes, no deduction of costs should be allowed.

Although the assignment of tax revenues is in the end a political decision on how much autonomy localities are allowed to have, a number of arguments favor giving the local government at least a share of the resource tax revenues. Local governments need compensation for the negative externalities natural resource exploitation brings with it and need incentives to develop local infrastructure needed for development. The negative externalities from production could best be addressed with a local production tax. However, as long as the local branches of the State Tax Bureau act, de facto, as local tax bureaus, a substantial stake for local government seems desirable for tax effort reasons. The local government’s share can be established in two ways: (1) through a share in a nationally determined tax or (2) by a locally determined rate or surcharge on a nationally determined tax base. Which should be chosen depends on the overall direction China’s tax system takes.

For rent taxation, the tax rates and bases should be determined by national tax laws, in order to avoid suboptimal rate setting by localities. However, for product taxes used to internalize externalities, local rate setting would be an option, in the context of national guidelines.

Local assignment of resource tax revenues in any form makes a sound system of fiscal equalization among localities all the more urgent. Different assignment of shares in natural resource taxation among provinces for equalization purposes seems undesirable. Rather, China should in the reform of the intergovernmental fiscal relations provide for a broad-based equalization scheme, with objective criteria for eligibility and contributions. If this policy is established, differential rates per locality as they exist now should be abandoned. To ‘counter possible procyclical effects of local resource revenues, China’s government should consider regulated local access to the capital market.

Independent of the tax chosen to appropriate rents for government, and independent of the assignment of the tax over levels of government, resource exploitation enterprises should be subject to the other taxes levied in the Chinese tax system, in order not to distort the allocation of investment among sectors. Furthermore, the arrangements made for resource taxation should be stable, transparent, and consistently applied over all projects, although actual rates may differ due to the characteristics of individual projects. The room for costly negotiations and rent seeking should, however, be minimized. Stability and predictability are essential for investments in long-gestation projects, such as natural resource exploitation. These arguments argue for a strong legal basis for resource taxation.


    BoadwayRobin andFrankFlatters“The Taxation of Natural Resources: Principles and Policy Issues,”Policy Research Working Paper WPS 1210 (Washington: World Bank1993).

    • Search Google Scholar
    • Export Citation

    ConradRobertF. andZmarakShalizi“A Framework for the Ananlysis of Mineral Tax Policy in Sub-Sharan Africa,”Policy Planning and Research Working Papers WPS 90 (Washington: World Bank1988).

    • Search Google Scholar
    • Export Citation

    DasguptaPartha andG. M.HealEconomic Theory and Exhaustible Resources (Welwyn, England: J. Nisbet1979).

    HeapsTerry andJohn F.Helliwell“The Taxation of Natural Resources,”Chapter 8 in Handbook of Public Economics by Alan J.Auerbach andMartin S.Feldstein (Amsterdam; New York: North-Holland1993).

    • Search Google Scholar
    • Export Citation

    MiezowskiPeterM. andEricToder“Taxation of Energy Resourcesin Fiscal Federalism and the Taxation of Natural Resourcesed. by Charles E.McLure andPeter M.Miezkowski (Lexington, Massachusetts: Lexington Books1983).

    • Search Google Scholar
    • Export Citation

    MusgraveRichardA.“Who Should Tax, Where, and What?”in Tax Assignment in Federal Countriesed. by Charles E.McLureJr. (Canberra: Centre for Federal Financial Relations, The Australian National University1983).

    • Search Google Scholar
    • Export Citation

    NellorDavidC. L. andEmil M.Sunley“Taxation of Mineral and Petroleum Resources” (mimeograph, Washington: International Monetary Fund1993).

    • Search Google Scholar
    • Export Citation

    OatesWallaceE.Fiscal Federalism (London: Harcourt Brace Jovanovich1972).

    ScottAnthonyCentral Government Claims to Natural ResourcesOccasional Paper No.8 (Canberra: Centre for Research on Federal Financial Relations, The Australian National University1978).

    • Search Google Scholar
    • Export Citation

    TietenbergThomasH.Environmental and Natural Resource Economics (New York: Harper Collins, 3rd ed.1992).

    World BankPrice Reform in ChinaReport No.1041-CHA (Washington: World BankMay1992).


World Bank. The views expressed in this paper are preliminary, and should not be considered as those of the World Bank.


The calculation is based on a scarcity price of about 80 percent of long-run marginal costs (this takes into account potential efficiency gains in extractions), which would lead to a price of about Y 85 a ton at the mine. Table 1 quotes Beijing prices, which are higher than the mine price, due to transport costs.


In fact, often under priced transport services now have become the major obstacle to resource availability.


“Tax Reform Will Affect Every Area of Economy” China Daily, March 12, 1993, p.2.


The paper is limited to a discussion of nonrenewable resources in view of China’s endowment profile. While the economic issues raised here are valid for all natural resources, sustainability of asset use becomes a paramount consideration in the management of renewable resources, with consequent implications for government policy.


The discussion reflects the treatment of the topic in the literature on natural resource economics and is also guided by current practice in the mineral sector. For extensive treatment of the topic, see Dasgupta and Heal (1979), Conrad and Shalizi (1988), Tietenberg (1992), Heaps and Helliwell (1993), and Nellor and Sunley (1993).


Corrective tax policy to adjust for externalities or noncompetitive market structures is appropriate and does not detract from the validity of the concept of uniform tax treatment across sectors.


If the roles of enterprise and resource owner are not separate, the introduction of a resource tax makes no sense, as government as the owner of the exploiting firm will already capture all the rents.


It is assumed here that marginal costs of extraction are constant. With increasing or decreasing costs, minor variations in the analysis would apply.


This equivalence holds only when negative cash flows either are fully refundable or can be carried forward at market interest rates.


At one extreme, if the government obtained free equity with entitlement to an equivalent share in net current revenues, this arrangement would represent a tax both on resource rents and on private returns to capital. A pure rent tax would require that the government either purchase its equity at the proportional cost of capital investment (corresponding to its share in net revenues) or reimburse the private operator for an imputed return on capital investment.


See Conrad and Shalizi (1988) for the development of the arguments related to Dutch disease.


Examples of countries that have suffered from Dutch disease abound. In addition to the classical example of the Netherlands (natural gas) for which the phrase was coined, Bolivia (tin), Zaire (copper), and Indonesia (petroleum) have all been affected to varying degrees.


Where resources are privately owned, taxing resources is still an economically desirable form of taxation, because, when properly designed, rent taxes generate no distortions.


A state administration of taxation was created in 1994 to administer national and shared taxes. Full implementation of this measure is expected to take a number of years. Ed.

    Other Resources Citing This Publication