Andrew Bauer
Uyanga Gankhuyag
Local Public Finance and Capacity Building in Asia
Issues and Challenges
OECD Fiscal Federalism Studies
Chapter 6. Natural resource taxation and revenue sharing in Asia
Abstract
More than 30 countries have put in place natural resource revenue-sharing systems – systems that allocate revenues from natural resources to subnational governments separately from other fiscal revenues. The main aim of such systems is to enable natural resource-producing regions of a country to benefit more from natural resource extraction in these regions and mitigate conflicts between national and subnational governments. Natural resource revenue-sharing systems can be derivation-based, in which case revenues are allocated predominantly by origin, or indicator-based, in which revenues are allocated typically by the equalisation principle. Derivation-based systems aim to compensate producing regions for the costs of resource extraction more directly, but they also bring significant equity challenges, volatility and public financial management challenges at the subnational level. They also may encourage accelerated resource exploitation. In contrast, indicator-based systems aim for greater equalisation but may suffer from excessive complexity. They are largely used to counterbalance the downsides of derivation-based systems and may not be politically acceptable to resource-producing regions. To work smoothly and enhance trust between national and subnational authorities, natural resource revenue-sharing systems need to be overseen by an adequate governance mechanism and be transparent. If well designed and implemented, resource revenue sharing can help drive economic transition and poverty reduction in resource-rich regions.
Introduction
In nearly every country, subnational governments receive public funds through a combination of direct tax collection and transfers from the national government. In most, non-renewable natural resource revenues are apportioned no differently than all other revenues. However, in more than 30 countries – most of them resource-rich – distribution of non-renewable natural resource revenues is governed by a set of rules that are distinct from those governing distribution of general revenues. Many of these systems were introduced or reformed during the 2000s “commodities super-cycle” – the sustained period of high prices of minerals and hydrocarbons – resulting in resource-rich countries seeing increases in their fiscal revenues from these resources. In most cases, the revenues in question are generated from the extraction of non-renewable resources such as minerals, oil and gas, but there are cases of countries where revenues generated by renewable resources, such as hydropower, forestry and fisheries, are apportioned through a separate system too.
In several countries, some revenues from the oil, gas and mineral sectors are collected by the national government and transferred back to their area of origin or adjacent areas. Angola, Bolivia, Brazil, Canada (some regions), Chad, People’s Republic of China (hereafter “China”), Colombia, Democratic Republic of the Congo (hereafter “DRC”), Ecuador, Ethiopia, Ghana, Guatemala, Guinea, India, Indonesia, Iraq, Kyrgyzstan, Madagascar, Malaysia, Mexico, Mongolia, Nepal, Niger, Nigeria, Papua New Guinea, Peru, the Philippines, South Sudan, Uganda, the United States (some regions) and Venezuela each have enacted a “derivation-based” intergovernmental transfer system for all or part of their mineral, oil or gas revenues.
Some resource-rich subnational governments are extremely dependent on these transfers. In Nigeria and Peru, for instance, more than 80% of the budgets of some subnational governments depend on resource revenue transfers from the central government.
A few countries also transfer some of their natural resource revenues to subnational governments using an “indicator-based” formula. In these countries, the national government distributes natural resource revenues to subnational authorities based on a set of objective indicators – such as population, revenue generation, poverty level or geographic characteristics (e.g. remoteness) – irrespective of where the natural resources are extracted. Ecuador, Mongolia, Mexico and Uganda are examples of countries that use indicator-based resource revenue-sharing formulas.
In another set of countries – including Argentina, Australia, Canada, China, India, the United Arab Emirates and the United States – subnational governments collect substantial revenues directly from oil, gas or mining companies. Direct tax collection can constitute a significant proportion of local budgets. For example, from 2012 to 2014, more than 25% of all fiscal revenues collected in Alberta, Canada came from direct petroleum taxation. In the United States, severance taxes from the oil sector in 2014 constituted 72% of total fiscal revenues in Alaska, 54% in North Dakota, and 39% in Wyoming.
Governments establish resource revenue‑sharing arrangements to address several, sometimes competing objectives. These are different from the objectives that can justify more general subnational tax assignments or general intergovernmental transfer programmes covering all fiscal revenues, such as improving public service delivery at the local level, fiscal risk sharing, or the equalisation of opportunities across the country.
Communities in the vicinity of mining, oil and gas projects or other large-scale natural resource developments often bear the most significant environmental and social impacts of resource exploitation, but sometimes benefit less than other regions. For example, extractive industries may attract migrants to the region, causing added congestion in public utilities (e.g. clogging transportation networks such as roads and railroads or putting a strain on water delivery systems). The presence of oil or mining companies in a region may also raise housing rents and costs of everyday non-tradeable services such as taxis and restaurants. Local governments can use resource revenue sharing as compensation or to fund efforts to mitigate the social and environmental losses associated with extraction, not just at the production site but also across all affected areas. Ecuador, for instance, levies USD 1 per barrel of oil produced in the Amazon region, the implicit assumption being that associated environmental damage is directly linked to the number of barrels that a company produces (Viale and Cruzado, 2012[1]).
In the most pressing cases, natural resource revenue-sharing arrangements may be introduced to mitigate or prevent violent conflict. The perception that revenues are not being distributed fairly can be a source of conflict and instability, especially in societies divided along ethnic, religious or other fault lines. Establishment of natural resource-sharing arrangements can signal the recognition of claims to a specific territory or a share of subsoil assets. Resource revenue sharing can also help build peace by encouraging dialogue between national authorities and local leaders, and generating a “peace dividend” for local populations (Bailey et al., 2015[2]). Thus, national governments will sometimes transfer a share of resource revenues to local governments in resource-rich regions to preserve or create harmony between the central government and the regions, as has been the case in Indonesia, southern Iraq, Nigeria and Papua New Guinea. In Indonesia, special resource revenue-sharing agreements with the regions of Aceh and West Papua helped end years of violent conflict. In Nigeria, the 1999 Constitution establishing a revenue-sharing system, followed by a 2002 court case reinforcing entitlements of oil-producing states, has contributed to greater peace and security in the Niger Delta. That said, resource revenue sharing does not always prevent conflict and may exacerbate it. Poorly designed revenue-sharing systems can incentivise groups to seize control of extractive sites to access a higher share of revenues. These revenues can then be used to finance violent actions.
Resource revenue-sharing systems have also helped to increase prosperity in resource-rich regions and address conflicts. For example, in Canada, an oil revenue-sharing accord of the federal government with Newfoundland province contributed in large part to the restoration of the economic prosperity of the province in the mid-2000s, following years of recession due to the collapse of fisheries.
Resource revenue sharing ought not to be confused with the issue of resource ownership (Haysom and Kane, 2009[3]). Ownership of natural resources per se does not necessarily translate to better prosperity for communities and regions with natural resources unless it is backed by arrangements on management and control of resources and associated revenue sharing. Shifting the discussion from ownership of natural resources to natural resource benefit sharing can help address the sticking points of disagreement between national and subnational leadership on who benefits from natural resources. Natural resource revenue sharing is one way of sharing benefits from natural resources.
This chapter summarises a study carried out by the Natural Resource Governance Institute (NRGI) and the United Nations Development Programme (UNDP) that reviewed countries’ experiences in natural resource revenue sharing, with some updates. It emphasises examples from Asia-Pacific but also includes salient examples from other regions. The chapter is organised as follows: the next section provides key definitions, distinguishing natural resource revenue-sharing systems from other related concepts; the third section discusses details about the design of natural resource revenue-sharing systems; the penultimate section reviews challenges with natural resource revenue-sharing systems and ways of addressing them; and the final section presents concluding remarks.
Natural resource revenue-sharing systems: Definitions and distinctions
A natural resource revenue-sharing system is a system whereby fiscal revenues from natural resources are allocated to subnational governments – differently and separately from fiscal revenues generated from other sources. Natural resource revenue-sharing systems can be classified based on: 1) the principles used to distribute resource revenues to subnational governments; and 2) the methods of sharing. Generally speaking, two principles of distributing resource revenues to subnational governments can be distinguished: distribution by derivation (by origin) and distribution by other principles (based on indicators) ‑ the latter called “the indicator-based principle”. In derivation-based principle, a portion of natural resource revenues collected, extracted in a given region, is returned to that region. In the indicator-based principle of distribution, objective indicators are used to determine allocations, such as population, levels of education or revenue-generating capacity by region (see Figure 6.1).
Figure 6.1. Types of natural resource revenue-sharing systems
Source: Authors’ illustration based on NRGI-UNDP (2016[4]).
Derivation-based systems are more directly related to some of the justifications for natural resource revenue sharing - they are used to compensate producing regions for the costs of extraction and depletion of a finite asset. However, derivation-based systems also bring significant public financial management challenges at the subnational level. For instance, they can generate fiscal volatility that subnational governments are often ill-equipped to manage. They can also exacerbate local Dutch disease effects and can encourage further conflict. For example, between 2005 and 2008, the increase in global mineral prices and the consequent increase in fiscal transfers to mining regions incentivised local leaders in Peru to instigate violent protests in order to extract additional transfers from the central government and gain jurisdiction over mine sites (Arellano-Yanguas, 2010[5]).
In terms of methods, the natural resource revenues may be shared by tax assignments to subnational governments, or by fiscal transfers from central to subnational governments. By definition, the tax assignment method allocates natural resource revenues based on the derivation principle, since the subnational government of the natural resource-producing and taxing province gets to keep all or some of these revenues, while non-producing subnational governments do not get any share. This type of natural resource revenue sharing is prevalent in countries such as Argentina, Canada, India, the United Arab Emirates and the United States. Often subnational governments collect a large share of resource revenues directly in federal states, but there are exceptions. For example, Brazil, Malaysia, Mexico, Nigeria and the Russian Federation are federal states, yet almost all resource revenues are collected by their respective national governments. China is a unitary state, yet the provinces collect a large proportion of the royalties.
The fiscal transfers method can be designed so that it allocates natural resource revenues either on a derivation basis or an indicator basis. The derivation-based allocation comprises the majority of countries with natural resource revenue-sharing systems. In the Asia-Pacific region, this includes China, Indonesia, Kyrgyzstan, Papua New Guinea and the Philippines.1 In addition, India, Malaysia and Mongolia have a derivation-based fiscal transfer system for at least a portion of their natural resource revenues.
Indicator-based systems can use, for instance, subnational indicators of population, poverty levels, and geographical characteristics such as remoteness, and generally seek to meet the fiscal needs of provinces lagging behind. They treat natural resource-producing and non-producing provinces in the same way. Fewer countries use such indicator-based systems for natural resource revenue sharing; Ecuador, Mongolia, Mexico and Uganda are among them. Indicator-based systems have several advantages. They can be more effective at addressing poverty and horizontal inequality by targeting less developed and more affected regions. They do not exhibit the same volatility, and “shock-concentrating” effects of derivation-based systems and public financial management challenges are less pronounced.
A key issue with indicator-based systems is that the objective of such a system is usually to equalise, but so is the objective of general equalisation or other types of inter-governmental transfers. The primary economic justification for separating non-renewable resource revenues, and then distributing them based on a special indicator-based formula is to counterbalance the downsides of a derivation-based system. For instance, in Canada, provinces collect royalties and provincial corporate income taxes, while the federal government collects national corporate income taxes. As a result, fiscal revenues per capita are disproportionately large in oil-rich provinces such as Alberta and Saskatchewan. Canada’s provincial equalisation formula helps to rectify this by calculating the revenue-generating capacity of each province on a per capita basis and allocating equalisation payments to provinces with a below-average capacity to generate own revenues. All provincial corporate income taxes from extractive industries and 50% of royalties are included in this formula. Through this formula, Canada has reduced inequalities in the fiscal capacities of provinces while allowing resource-rich regions to retain most of their resource revenues.
Natural resource revenue sharing as part of the intra-governmental system, which is discussed in this chapter, should be distinguished from revenue sharing between companies and governments, which would be referred to as tax and royalty collection, and benefit sharing. Benefit sharing expresses the overall idea of the extractive industry sharing its benefits with the host country’s residents. Benefits can be shared through different mechanisms – either through the payment of taxes that can then trickle down to host communities through the government, or through voluntary financial contributions, or development-oriented projects directly benefitting host communities (Wall and Pelon, 2011[6]) (see Figure 6.2).
Figure 6.2. Revenue sharing versus benefit sharing
Source: Authors’ illustration based on NRGI-UNDP (2016[4]).
Natural resource revenue-sharing systems
This section reviews the key characteristics of natural resource revenue-sharing systems:
Which revenues are shared?
Between whom and how much is shared?
How are natural revenue-sharing systems established?
Which revenues are shared?
Countries can share all or some streams of revenues generated from natural resources.
Revenues collected only from – and specific to – natural resource companies include royalties, signature bonuses and fees for permits for natural resource exploitation. Royalties are rents paid to the owner of natural resources – usually the government – in exchange for the right to exploit the resource. In many countries, royalties, along with corporate income taxes, constitute the largest portion of revenues from the extractive industries. Royalties may be assessed on the volume of mineral or hydrocarbon production (unit-based) or the value (ad valorem), and less commonly, they can be imposed in fixed amounts. Royalties can also be set on a sliding scale – set at a variable rate depending on the market price of the resource, as in China with Special Petroleum Proceeds, a royalty-like instrument levied with rates varying depending on the international price of crude oil. Signature bonuses are one-time payments made to host governments upon signing a contract for exploitation of natural resources. License fees for utilisation of natural resources are also specific to natural resources. Royalties and license fees are usually imposed not only for actual extraction but also for exploration. Production entitlements are also collected from governments from the oil and gas sector – in-kind payments instead of taxes or rents. Some countries also have introduced mandatory contributions from mining, oil and gas companies towards local development, education or other purposes. Kyrgyzstan has a 2% levy on mining companies’ revenues, called “payment for development and maintenance of local infrastructure”.
Revenues collected from all companies irrespective of their business sector include corporate income taxes, withholding taxes, customs duties, value-added taxes (VAT), property taxes, land taxes, natural resource utilisation fees, pollution taxes, and other fees and fines. In addition, the government can also receive dividends from government equity held in natural resource companies.
Countries’ experiences show that it is more common to include royalties and property taxes in natural resource revenue-sharing schemes, which reflect the relative ease with which these taxes and payments are assessed. Less commonly included are corporate income taxes, goods and services taxes or dividends from government equity. In the same country, the revenue-sharing arrangement can differ between mining on the one hand, and oil and gas on the other.
For example, Papua New Guinea shares only royalties from oil and gas. Indonesia has a different arrangement for mining versus oil and gas. Whereas for mining, only mineral royalties are included in the resource revenue-sharing arrangement, for oil and gas, all streams of revenues are included. In China, only a few revenue streams are included, of which the most important is the Mineral Resources Compensation, a royalty imposed on mining (excluding coal). This fee is shared 50-50 between the central government and the producing provinces.
Several issues matter in deciding to include certain revenue streams in the natural resource revenue-sharing scheme. First, the decision on which revenue streams to include has implications on the magnitude, the volatility and timing of revenues received. Royalties, corporate income taxes, and goods and services taxes are usually much larger than property taxes or license fees.2 In terms of timing, whereas some revenues such as license fees or property taxes do not vary significantly depending on the volume of the resource exploited, royalties are dependent on the volume. But royalties are still more predictable compared to corporate income taxes. In addition, different revenues flow to the government at different times in the life cycle of a mine or oil and gas field. Thus, signature bonuses are collected upfront, royalties are usually collected as soon as production begins, while corporate income taxes peak several years into production, after costs have been recovered. Dividends may not be collected until much later. In general, less volatile and more predictable revenue streams should be included in natural resource revenue-sharing schemes.
Second, some revenue streams are easier to calculate, to attribute and to collect. Taxes – or revenue streams – that are easier to administer tend to be assigned or transferred to subnational governments, because their administrative capacity tends to be less than that of national governments. Land and property taxes are relatively simple to calculate. Royalties are more complicated but are still manageable for subnational governments with a reasonable level of capacity. To calculate royalties, one needs information on production volume, quality (grade) and market prices. Royalties can also be more easily attributed to specific mines or oil and gas fields in given territories.
In contrast, corporate income taxes are more complicated. They require more information – not just revenues of the company, but also its costs. They may also include tax credits and tax deductions and may be subject to “creative accounting” to avoid taxes. Corporate income taxes are also more difficult to attribute. This is because some companies may operate in different parts of the country, whereas their corporate income tax accounting does not apportion common costs to their operations in different locations. Many companies are headquartered in capital cities and thus, most of the revenues may be attributed according to their headquarters, rather than field operations – where the actual natural resource exploitation happens. Subnational governments should have access to sufficient information and adequate administrative and analytical capacity before they are assigned such complex taxes. For this reason, corporate income tax administration generally rests with the national government, whereas property and land taxes are more likely assigned to subnational governments. With royalties, the practice varies: in some countries, they are assigned to the national government and in others, to subnational governments. Table 6.1 shows the assignment of these three types of revenues – between national and subnational governments in selected countries in Asia-Pacific.
Offshore resources are usually under the jurisdiction of national governments. Arguably, exploitation of offshore resources has a less direct, visible impact on adjacent communities, although it can disrupt fisheries or cause oil spills. Offshore resources are also more difficult for armed groups to occupy, and thereby less likely to become the target of demands for local leaders for a share of the revenues. As a result, offshore resource revenues are often not included in resource revenue-sharing schemes. Nevertheless, there are countries – Australia, Brazil, Canada, Italy, Malaysia and the United States – where adjacent subnational governments receive a share from offshore resource exploitation (Brosio, 2006[7]).
With regard to revenues from common minerals such as construction materials (e.g. sand, gravel), many countries assign revenues from them to subnational governments, as magnitude of these revenues is smaller, they are less prone to volatility and, geographically, they are more evenly distributed.
Finally, most natural resource revenue-sharing systems that have been introduced in recent decades include revenue streams from non-renewable mineral and hydrocarbon resources, given that these resources generate substantial amounts of revenues. There are cases, however, when revenues from renewable resources are also covered in the revenue-sharing scheme. In the Philippines, in addition to mining, revenues from fishery and forestry are also included in the “National Wealth”, 40% of which is shared with subnational governments. In Norway, local governments received a share of revenues from hydropower produced in their territories.
With whom and how much is shared?
Vertical distribution describes the allocation of revenues between different levels of government. For simplicity, there is the allocation between the national government on the one hand, and all levels of subnational governments on the other. Vertical distribution can range from highly centralised, with minimal natural resource revenues shared with subnational governments, to highly decentralised, with most resource revenues collected by subnational governments. The degree of vertical distribution is determined by underlying fiscal arrangements between levels of government – tax assignments to subnational governments provided in law or transfers from national to subnational governments.
In highly centralised systems, natural resource revenue sharing also tends to be centralised. For example, in Afghanistan and Myanmar, most resource revenues are collected by the central government, while subnational governments collect only minor land taxes and fees.
Table 6.1. Mineral and petroleum tax collection by level of government, selected countries in Asia-Pacific
Notes:
1. Only offshore.
2. Though legally it is under national jurisdiction, subnational governments sometimes collect land taxes.
3. Local governments at the aiyl aimak level collect “payments for development and maintenance of local infrastructure”, which are royalty-like payments.
4. Royalties are only assessed and collected by indigenous groups and some local government units.
Source: Authors’ compilation.
In other countries, most natural resource revenues are still collected by national governments, but they transfer significant shares of these revenues to subnational governments. For instance, in Indonesia, 80% of mineral royalties, 30% of revenues from gas and 15% of revenues from oil is allocated to subnational governments of producing provinces. In the Philippines, 40% of “national wealth” – revenues from mining, forestry and fisheries – is allocated to producing provinces, municipalities and barangays.
The vertical distribution of natural resource revenues also depends on the size of the province or district, and the administrative capacity of subnational governments. For instance, a third-tier subnational government in Indonesia (Bojonegoro regency) governs a population of 1 million people, whereas a typical third-tier subnational government in Mongolia (soum government) oversees a population of 3 000; so their administrative capacities would differ correspondingly. The distribution of natural resource revenues in relation to administrative capacities is further discussed below.
Horizontal distribution describes the allocation of revenues to subnational governments at the same level of government. In several cases, of the subnational portion of the natural resource revenues, fixed shares are allocated to resource-producing provinces (2nd tier of government), but also to districts (3rd tier) within the producing provinces. Among these, the examples of Indonesia and the Philippines are useful. For instance, in Indonesia, the share of oil, gas and mineral revenues going to subnational governments is further split between producing provinces, as well as producing and non-producing districts within these provinces. In contrast, in the Philippines, the share of mineral revenues is split between producing provinces, producing districts (municipalities), and further to producing barangays (the 4th tier of government). In other words, while in Indonesia, non-producing districts within producing provinces receive a share of natural resource revenues, in the Philippines, they do not. Instead, the remaining share goes further to the smaller administrative unit from where the resource is produced. Table 6.2 provides examples of vertical and horizontal resource revenue distribution in selected countries.
Table 6.2. De jure derivation-based inter-governmental transfer formulas in selected countries
Note: Some listed countries also have other types of inter-governmental transfer systems in addition to the derivation-based inter-governmental transfer system.
Source: National legislation or mining codes; Agustina et al. (2012[8]), “Political economy of natural resource revenue sharing in Indonesia”, Asia Research Centre Working Paper 55, ARC, London; Morgandi (2008[9]), “Extractive Industries Revenues Distribution at the Sub-National Level,” Revenue Watch Institute; Viale (2015[10]), Distribución de la renta de las industrias extractivas a los gobiernos subnacionales en América Latina: Análisis comparative y de tendencias, INTE-PUCO, Lima.
One issue with horizontal distribution is that the area of impact of resource extraction often does not overlap with the administrative division in the country. In other words, such allocation based on administrative boundaries may result in a situation where areas – and people living in these areas – in the impact zone of extractive activities may be left out of fiscal revenue allocation, while areas not impacted may benefit from it. For example, in Indonesia, Blora and Bojonegoro regencies (districts) sit above the Cepu block, one of Indonesia’s most lucrative oil fields. But because Blora is in the Central Java province, while Bojonegoro is in the East Java province (the latter is the home of most oil wells), Blora receives significantly fewer resource transfers (see Figure 6.3).
Figure 6.3. Blora and Bojonegoro, Indonesia
Source: NRGI-UNDP (2016[4]).
While laws setting natural resource revenue-sharing schemes stipulate allocation to “producing”, “non-producing” and “adjacent” regions, these terms are often not clearly defined. For example, laws often do not specify whether “producing” means where the resource is located or where the production facilities are located. In cases of oil and gas, the field may lie beneath several districts, but oil wells may be located in one of them. Similarly, in underground mining, the deposits may span the territories of several districts, but the mine mouth can be located in one district. Alternatively, a mine can be located in one district, but mine facilities and waste storage can be located in the adjacent one. In Kyrgyzstan, for example, 20% of a 2% extra non-tax payment by mining companies (royalty on mining) is allocated to “producing” villages. There are already cases of conflict between villages where a mine crosses village boundaries. One option is to specify how revenues would be split between subnational jurisdictions in such special cases. In the Philippines, this was made clear by specifying that in such cases, revenues will be split between jurisdictions where natural resources cross jurisdictional boundaries, using population size (weighted at 70%) and land area (weighted at 30%). The term “producing” can be defined as the volume or value of production, and this can mean different outcomes for subnational jurisdictions.
The area of impact of resource extraction may also extend beyond resource-producing areas. The impact zone can include areas stretching hundreds of kilometres, such as downstream rivers impacted by extraction, or along the transportation routes of minerals. Brazil and Colombia addressed this issue by allocating royalties from oil and gas not only to producing municipalities but also to municipalities through which oil and gas are transported (Brosio and Jimenez, 2012[11]).
Moreover, natural resource allocations to subnational governments are usually oblivious of changes in population in the provinces or districts driven by mining or oil and gas extraction projects – so-called boomtowns. Such population changes can be sudden and substantial, with subnational governments struggling to meet the public service needs of the enlarged populations. For example, during the mining boom in the 2000s, the population of Zaamar soum in Tuv aimag, Mongolia, quadrupled from 5 000 to 20 000 people, putting increased demand on the local government to provide health and education services. In Lao People’s Democratic Republic, commissioning of a large mine in Sepon, Savannakhet province in 2005 led to in-migration to the area and additional demand for public services, in response to which the government relocated 200 public servants to the area to help deal with the demand (ICMM, 2011[12]).
In some cases, governments use asymmetric arrangements with resource-rich provinces, whereby a larger share of natural resource revenues is allocated to resource-rich provinces with special status. For example, in China, the Tibet Autonomous Region and the Inner Mongolia Autonomous Region receive 60% of the Compensation Fee levied on mineral resource extraction activities, whereas other provinces receive 50%. In Indonesia, the Aceh province also has an asymmetric arrangement, ending 30 years of conflict: a peace agreement (Memorandum of Understanding) signed in 2005 stipulated, among other things, that the Aceh province is to receive 70% of revenues from oil and gas for 8 years, and 50% thereafter – compared with 15% of oil revenues and 30% gas revenues received by other provinces (UNEP and UNDPA, 2015[13]; Keating and Brown, 2015[14]).
Countries with revenue allocation by indicator-based inter-governmental transfers use indicators measuring the population, poverty rate, or the economic activity of provinces (e.g. regional gross output), or allocate revenues expressly to provinces lagging in terms of development. For example, the formula for petroleum revenue allocation in Mexico includes population, revenue generation, as well as a third variable, which benefits states with low populations and high revenue generation (Castanada and Pardinas, 2012[15]; Courchene and Diaz-Cayeros, 2004[16]). In Bolivia, 1% of the value of gross sales of petroleum is allocated to Beni and Pando, the two poorest municipalities at the time the resource revenue-sharing system was established.
In mixed systems, a part of natural resource revenues is allocated based on derivation, while another part is allocated based on indicators. For example, in Uganda, 6% of petroleum royalties are allocated to local governments “located within the petroleum exploration and production areas”. Of this amount, 50% is allocated based on the level of production or size of the area affected, and the other 50% is allocated based on “population size, geographic area and terrain” (Government of Uganda, 2015[17]). Likewise, Mongolia has a system which aggregates 10% of VAT, local budget surpluses, 30% of petroleum royalties, 5% of mining royalties, 30% of royalties from large (strategic mines) and 50% of mineral exploration license fees (Shotton, 2017[18]; Government of Mongolia, 2011[19]). Of these, the first three sources of funding are allocated on an indicator basis, and the last three are allocated by origin.
Especially with indicator-based systems (or the indicator-based part of a mixed system), the allocation formulas can get unnecessarily complicated. To avoid this, indicators to be used for allocation should be kept to a minimum; they should be easily measurable and regularly updated. Importantly, indicators and formulas should be understandable to subnational governments and local stakeholders, which is part of the transparency of the system.
In some countries with natural resource revenue-sharing systems, there are “clawback” provisions built into the general inter-governmental transfer systems, cancelling out or moderating the effect of a derivation-based system. These “clawbacks” result in less general transfer allocations to subnational governments that receive more allocations through natural resource revenue-sharing schemes. In Canada, the Northwest Territories retain a share of mineral, oil, gas and water-related revenues, of which 25% is passed on to aboriginal governments in these territories. However, in the allocation of general unconditional grant per the Territorial Financing Formula from the federal government to Northwest Territories, for each dollar raised by the province in taxes, approximately 70 cents is deducted. Thus, much of the resource revenues raised in taxes by the territory is clawed back (Bauer, 2014[20]). Natural resource revenue-sharing systems – and the receipts by subnational governments – should be analysed not only separately, but also along with all other flows of revenues – including own revenues and general transfers. In general, rather than having a derivation-based natural resource revenue-sharing system together with clawback arrangements that effectively negate this system, it is more straightforward not to have a natural resource revenue-sharing system in the first place.
Formulas for allocation of natural resource revenues ought to derive from the objectives of the natural resource revenue-sharing system. Often, these objectives are implied; but it is worth it to make them explicit. For example, if the objective is to compensate the regions for the loss of livelihoods and environmental damage of resource-exploitation activities, then affected areas should be defined and indicators can be used, such as the volume or value of mineral production; area of land under mining contracts; length of mineral transportation routes, etc. If the objective is to improve development in poor regions that are resource-rich, indicators such as population size, poverty rate, remoteness and access social services can be used.
The legal basis for resource revenue sharing
Systems for natural resource revenue sharing are usually established in legislation, including the formulas for allocation. In rare cases, natural resource revenue-sharing systems are referenced in constitutions (Brazil, Canada, Iraq, Nigeria, South Sudan).
At the other extreme, there are also ad hoc resource revenue-sharing systems. Kazakhstan allocates disproportionately large per capita transfers to the oil-rich and conflict-affected regions Atyrau and Mangistau, as part of annual allocations. This fiscal arrangement started from a political agreement that set a precedent for such transfers to these regions, but it is not specified in the law.
Formula-based allocations result in more stable and predictable financing flows, are more conducive to good subnational budget planning, and are thus clearly superior to ad hoc allocations. However, the implementation of the rule can be weak, and lack of information on fiscal transfer can prevent verification, and thus a formula-based system may not work in practice. Alternatively, the laws (and the formulas) can be changed so often as to render the system essentially ad hoc. This is the case in Mongolia, where the (General) Local Development Fund, which represents a mixed system, has been amended at least three times since its introduction in 2013 (Shotton, 2017[18]).
In summary, there are many considerations for deciding on the allocation of natural resource revenues. Once such systems are in place, they are difficult to change. Provisions for revisiting resource revenue-sharing systems once every few years can be embedded in laws. In addition, any changes to resource revenue-sharing systems should be analysed and considered carefully. For this purpose, simulation models should be used to assess implications of changes in revenue allocation formulas on provinces and districts, which can help to take actions to minimise the transition impact on “losing” subnational governments.
Revenues from natural resources can change substantially over time. Therefore, their magnitude – both nationally and at the subnational level – should be considered not only at a point in time when the system is designed but also into the future. The models above can be used to apply different scenarios of revenue levels under different allocation formulas.
Addressing challenges with resource revenue-sharing systems
This section discusses challenges with resource revenue-sharing systems, specifying whether they are relevant to one or the other type of system. It also discusses ways to address these challenges.
Inequality. Natural resource revenue-sharing systems can aggravate regional inequalities when they are derivation-based.3 If resource-producing regions are relatively wealthy to start with, and a derivation-based system is applied, these regions will receive disproportionately larger revenues. In Brazil, for example, the state of Rio de Janeiro is the third wealthiest state by gross domestic product (GDP) per capita and is one of Brazil’s largest offshore oil producers. It disproportionately benefits from the revenue-sharing regime. Until 2013, producing states in Brazil were allocated 52.5% of royalties and 40% of “special participation” earnings; since the 2013 reforms, these shares were reduced to 20% and 34%, respectively (Viale, 2015[10]).
Incentives for resource exploitation. Resource revenue-sharing systems by derivation may encourage accelerated resource exploitation. This is particularly the case if the resource-hosting provinces and districts are poor and natural resource revenues account for a significant share in subnational revenues.
Volatility. Since the underlying revenues are volatile, derivation-based resource revenue-sharing systems are pro-cyclical, exacerbating the natural resource boom-bust cycles. The resulting volatility incentivises over-spending on wasteful projects or increasing government wages unsustainably during boom times while resulting in painful public spending cuts or ratcheting debt in bust times. In Brazil, large oil royalty windfalls to municipalities were associated with increases in spending on government housing and urban infrastructure projects, all the while the efficiency of public service provision deteriorated, resulting in a decrease in access to piped water, connections to sewage networks and garbage collection (Ardanaz, 2014[21]). Whereas at the national level, the shocks can be absorbed and their effects smoothed out, at the subnational level, there is less possibility to do so.
There are several ways to reduce the volatility of revenues at the subnational level. One option is to assign or transfer to subnational governments those streams of revenues that are more stable, with the caveat that the magnitude of more stable streams of revenues tends to be smaller. Alternatively, an indicator-based formula can be used instead of a derivation-based formula. However, both these options may not be politically acceptable to resource-rich regions.
The second option is to enable subnational governments to save resource windfalls in stabilisation funds, to use when resource revenues decline and replenish when they increase. Such a mechanism can smooth spending, protecting it from the volatility on the revenue side [see NRGI, (n.d.[22])]. There are a few cases of subnational jurisdictions establishing such funds. Among the more established funds are Abu Dhabi (United Arab Emirates), Alabama (United States), Wyoming (United States), Alberta (Canada) and Quebec (Canada) (Bauer, 2014[23]). The problem with such funds is that they are prone to patronage and corruption; subnational governments may lack the capacity to manage such funds and, apart from a few exceptions, the magnitude of funds may be too small to justify setting up of a stand-alone fund. The pooling of funds from several resource-rich regions in a country could justify establishing a fund where each jurisdiction’s resource revenues are too small to warrant one.
Another way to help stabilise the inflow of revenues is to enable subnational governments to borrow when revenues decline and pay back when revenues increase. Although this option avoids revenue management challenges with stabilisation funds, it has problems with debt management and over-borrowing. There have been many cases of debt crises of subnational governments due to such attempts to manage subnational-level volatility. In some cases, governments have bailed out the subnational governments – such as Chile, Colombia, Indonesia, Mexico and Russia. In other cases, they were allowed to default on their debts, such as in Bolivia, Nigeria and Peru (Liu and Waibel, 2008[24]).
Another solution involves national governments stepping in to smooth revenue volatility. For example, resource revenue allocation can be determined based on a moving average of resource revenues over several years, rather than a single year, or allocated based on non-production-based indicators.
Public financial management problems. The very nature of a derivation-based mechanism may undermine a fundamental principle of sound financial management, the “finance follows function” rule since derivation-based fiscal transfers are neither linked to subnational needs for financing nor the expenditure responsibilities of subnational governments (Boadway and Shah, 2007[25]; Martinez-Vazquez, 2015[26]). During booms, local governments end up receiving revenues that exceed their absorptive capacity or needs, so these revenues are likely to be wasted on high-visibility vanity projects or expenditures that can be increased quickly, such as government salaries, without improvements in service delivery. Again, while all systems of natural resource revenue-sharing systems may be affected by such public financial management challenges, in derivation-based systems, these challenges are more pronounced since derivation-based systems do not distribute revenue based on subnational government needs.
Measures can be taken to improve subnational financial management of resource revenues. Some countries put conditions on the spending of resource revenues, or earmark them for certain uses, primarily for spending on human or physical capital investments. For example, in Indonesia, 0.5% of revenues must be spent on education by provinces and regencies, while in Kyrgyzstan, Regional Development Fund allocations financed by royalty-like payments from mining companies, are required to be spent on small-scale infrastructure and economic development. In Ecuador, severance tax payments must be spent on environmental restoration, among others (Viale and Cruzado, 2012[1]). Although such conditions and earmarks can protect spending on underfunded expenditure items, they can undermine budgetary autonomy and flexibility without necessarily improving service delivery. More fundamentally, earmarking does not guarantee that more money is available for an expenditure item due to the fungibility of money. Subnational governments can add resource revenues to a budget line and subtract revenues from other sources, leading to a net impact of zero.
Governments also establish special-purpose funds that are intended to benefit resource-rich provinces and districts, such as Nigeria’s Niger Delta Development Commission or Kyrgyzstan’s Regional Development Funds. These funds may be set up simply as separate accounts of the general budget, or as entirely separate institutions. An example of the former is the General Local Development Fund in Mongolia, which is managed by the Ministry of Finance and administered by it together with local authorities. An example of the latter is Nigeria’s Niger Delta Development Commission, a federal commission controlled primarily by representatives of oil-producing states, along with the federal government and some companies, that receives 15% of inter-governmental transfers due to states and 3% of oil companies’ annual budget, and is supposed to spend funds on economic development in the Niger Delta. Whether national or subnational authorities should manage a fund is context-specific.
Complexity. The problem of complexity is more common to indicator-based resource revenue-sharing systems. Indicator-based allocation requires the collection of data to track indicators at the subnational level. These data requirements can become extensive. At one time, the Canadian allocation formula contained 37 indicators. Currently, the indicator-based part of Mongolia’s resource revenue allocation scheme includes 3 stand-alone indicators and another composite indicator (Local Development Index), which in turn consisted of 65 indicators (Volooj et al., 2016[27]). However, an indicator-based system does not need to be so complex. At the other end of the spectrum, South Africa’s indicator-based resource revenue-sharing scheme uses only one indicator – regional GDP as a proxy for fiscal capacity (Shah, 2007[28]). Having fewer indicators not only saves time and resources when collecting data, but it also helps make the system better, e.g. less prone to data manipulation and errors, less likely to include several highly correlated indicators, and more transparent.
Transparency challenges. To enforce revenue-sharing systems and benefit from them, it is essential that the necessary information is available and that the systems are transparent. Transparency – at the subnational level, with information in disaggregated form – is necessary to align the expectations of local communities with the magnitude of revenues received at the local level and to hold subnational governments accountable. If a resource revenue-sharing system is not transparent, it will not be effective in improving trust and addressing disagreements between subnational and national governments. Subnational governments in several countries, including the Democratic Republic of the Congo (DRC) and the Philippines, have questioned whether they are receiving what they are entitled to due to lack of transparency around payments from extractive companies to the national government.
In addition to past flows of revenues, subnational governments should also be informed of prospective flows of revenues to handle their volatility and manage expenditures better. For this purpose, in addition to the above, they also need project-by-project projections on natural resource projects – sales revenues along with costs, and the applicable fiscal regime.
However, in practice, only about half of the countries reviewed disclose details of natural resource revenues collected or transferred to subnational governments. Natural resource revenue-sharing systems have significant information requirements. Under derivation-based systems, subnational governments need several pieces of information: the applicable resource revenue-sharing system with the formulas for allocation, disaggregated information on fiscal payments from natural resource projects in their locality, by type of revenue streams, information on sales as well as the costs of natural resource projects. Much of this information, for larger mining, oil and gas projects, is available through the Extractive Industries Transparency Initiative (EITI), for countries that are part of this initiative. Under indicator-based systems, in addition to the information above, the basis for making the assessment and the actual calculations to allocate revenues should also be transparent. For example, the Australian Commonwealth Commission makes such detailed information, with a breakdown per region, publicly available (Government of Australia, n.d.[29]).
Depletion. Resource-dependent regions are affected by depletion of their resources, usually to a greater extent, compared with countries as a whole. Moreover, subnational governments may be less informed than national governments about how long their resource revenues are expected to last and may be less prepared to deal with the economic and fiscal consequences of resource depletion. To mitigate this problem, subnational governments should invest their resource revenues during good times into financial, human and physical capital.
Sovereign wealth funds that allow their owners to earn interest or “permanent” income is a form of financial capital. Defined by the International Monetary Fund (IMF) as government-owned entities with a macroeconomic purpose that do not have explicit liabilities and invest at least partly in foreign assets, sovereign wealth funds must be distinguished from extra-budgetary domestic spending accounts, development banks or so-called “strategic investment funds” whose primary purpose is to channel money into domestic assets. While these domestic spending vehicles may have a role to play in driving economic growth, the international experience with these types of funds highlights several challenges. Chief among them is that they are inherently political, as asset managers choose domestic winners and losers when deciding which local assets to invest in. They can also circumvent legislative oversight and government procurement systems (Bauer, 2018[30]).
Although there are a few cases of successful subnational funds such as funds in Abu Dhabi (the United Arab Emirates), Alaska (United States) and Quebec (Canada), many national and subnational funds have also not achieved their objectives. If such funds are created, appropriate deposit, withdrawal and investment rules should be put in place and enforced, their operations should be transparent, and there should be proper oversight (Bauer, 2014[23]).
Resource revenues can also be invested in human and physical capital, to support the diversification of the local economy away from its natural resources. Again, there are not many recent examples of successful diversification of local economies away from mining, oil and gas. The better-known examples are Dubai in the United Arab Emirates and the coal-producing Appalachian region of the United States (Gelb, 2010[31]).
Conclusion
In conclusion, natural resource revenue-sharing systems can enable communities to benefit from resource exploitation in their areas. Regardless of national or subnational legal ownership, natural resource revenue-sharing systems can bring tangible benefits, can act as a response to local claims for a share of subsoil wealth, and can help mitigate conflicts between national governments and subnational authorities. By introducing a natural resource revenue-sharing system based on clear objectives, clear formula, written into law and implemented systematically, debates around benefit sharing can shift from purely political to technical and more productive discussions.
Natural resource revenue-sharing systems are also superior to direct benefit sharing from natural resource companies to individuals and communities, in that they go through public financial management systems and therefore, arguably, are more accountable. This is not the case in all contexts, however – in some countries and communities where there is a lack of trust by subnational leaders towards the national government, or systems for public service delivery are absent, direct company-to-community transfers may be the preferable or the only option. Nevertheless, where possible, natural resource revenue-sharing systems, if designed and implemented well given the local context, can be more inclusive and transparent.
However, there are many challenges concerning the design and administration of natural resource revenue-sharing systems. This chapter reviews some key challenges with these systems and experiences of countries in addressing these challenges.
To work smoothly and to enhance trust between national and subnational authorities, natural resource revenue-sharing systems need to be overseen by an adequate governance mechanism and should be transparent. They must also improve the delivery of services at the subnational level. The exploitation of non-renewable natural resources often represents a one-time chance for regions to transform their economies or escape poverty. If well designed and implemented, resource revenue sharing can help drive such an economic transition and poverty reduction.
References
[8] Agustina, C. et al. (2012), Political Economy of Natural Resource Revenue Sharing in Indonesia, ARC, London, http://www.lse.ac.uk/asiaResearchCentre/_files/ARCWP55-AgustinaAhmadNugrohoSiagian.pdf.
[26] Ahmad, E. and G. Brosio (eds.) (2015), Tax Assignments at the Regional and Local Levels, Edward Elgar Publishing, Cheltenham.
[7] Ahmed, E. and G. Brosio (eds.) (2006), The Assignment of Revenue from Natural Resources, Edward Elgar Publishing, Cheltenham.
[5] Arellano-Yanguas, J. (2010), Local politics, conflict and development in Peruvian mining regions, Institute of Development Studies, University of Sussex.
[2] Bailey, R. et al. (2015), Investing in Stability: Can Extractive-Sector Development Help Build Peace?.
[30] Bauer, A. (2018), Funds and Mechanisms for the Management of Oil and Gas Revenues to Support Sustainable Development: Insights from Country Experiences and Lessons for Senegal, United Nations Environment Programme (UNEP).
[20] Bauer, A. (2014), “All That Glitters is Not Gold: Will Expectations be Dashed in Canada’s North?”, Natural Resource Governance Institute 30 June.
[23] Bauer, A. (ed.) (2014), Natural Resource Fund Governance: The Essentials, Natural Resource Governance Institute (NRGI) and Columbia Center on Sustainable Investment (CCSI).
[28] Boadway, R. and A. Shah (eds.) (2007), A Practitioner’s Guide to Intergovernmental Fiscal Transfers, World Bank, Washington, DC.
[25] Boadway, R. and A. Shah (eds.) (2007), Intergovernmental Transfers: Principles and Practice, World Bank, Washington, DC.
[11] Brosio, G. and J. Jimenez (2012), The Intergovernmental Assignment of Revenue from Natural Resources: A Difficult Balance Between Centripetal and Centrifugal Tendencies, Edward Elgar Publishing, Northampton, MA.
[15] Castanada, L. and J. Pardinas (2012), “Sub-national Revenue Mobilization in Mexico”, IDB Working Paper Series, No. 54.
[32] Conger, B. and B. Dahlby (2015), “Policy Forum: Taxation of Machinery and Equipment and Linear Property in Alberta”, Canadian Tax Journal, Vol. 63/2, pp. 487–499.
[31] Gelb, A. (2010), Economic Diversification in Resource-Rich Countries, IMF, Washington, DC.
[29] Government of Australia (n.d.), Commonwealth Grants Commission, https://www.cgc.gov.au/ (accessed on 11 November 2019).
[19] Government of Mongolia (2011), The Law on the Budget, 23 December 2011.
[17] Government of Uganda (2015), Public Finance Management Act, 6 March 2015.
[16] Guigale, M. and S. Webb (eds.) (2004), Transfers and the Nature of the Mexican Federation, World Bank, Washington, DC.
[3] Haysom, N. and S. Kane (2009), Negotiating natural resources for peace: Ownership, control and wealth-sharing.
[12] ICMM (2011), Utilizing mining and mineral resources to foster the sustainable development of the Lao PDR, Lao PDR Ministry of Energy and Mines, National Economic Research Institute and International Council on Mining and Metals.
[14] Keating, M. and O. Brown (2015), Addressing Natural Resource Conflicts: Working Towards More Effective Resolution of National and Sub-National Resource Disputes, Chatham House.
[9] Morgandi, M. (2008), Extractive Industries Revenues Distribution at the Sub-National Level, Revenue Watch Institute.
[22] NRGI (n.d.), Natural resource fund web page, http://www.resourcegovernance.org/nrf (accessed on 11 November 2019).
[4] NRGI-UNDP (2016), Natural Resource Revenue Sharing, Natural Resource Governance Institute (NRGI) and United Nations Development Programme (UNDP) policy paper.
[24] Shah, A. (ed.) (2008), Subnational Borrowing, Insolvency and Regulation, International Bank for Reconstruction and Development and the World Bank, Washington, DC.
[18] Shotton, R. (2017), Local Development Fund allocation arrangements: Current issues and options for change, Sustainable Livelihoods Project-3.
[13] UNEP and UNDPA (2015), Natural Resources and Conflict: A Guide for Mediation Practitioners, United Nations Department of Political Affairs and United Nations Environment Programme, New York.
[10] Viale, C. (2015), Distribución de la renta de las industrias extractivas a los gobiernos subnacionales en América Latina: Análisis comparative y de tendencias, INTE-PUCO, Lima.
[1] Viale, C. and E. Cruzado (2012), Distribution of Revenue from the Extractive Industries to Sub-national Governments in Latin America, Revenue Watch Institute.
[21] Vieyra, J. and M. Masson (eds.) (2014), Fiscal Windfalls, Transparency, and the Efficiency of Public Good Provision: Evidence from Brazilian Local Government, Inter-American Development Bank, Washington, DC.
[27] Volooj, S. et al. (2016), Extractive industries Revenue Management System in Mongolia, UNDP, Ulaanbaatar.
[6] Wall, E. and R. Pelon (2011), Sharing Mining Benefits in Developing Countries: The Experience with Foundations, Trusts and Funds, World Bank, Washington, DC.
Notes
← 1. This is not an exclusive list.
← 2. One exception is property taxes on machinery, equipment or pipelines in Canada and the United States, which generate significant tax revenue for municipalities. See Conger and Dahlby (2015[32]).
← 3. However, if not designed adequately, an indicator-based system can also be inequitable. For example, in Mongolia, the ratio of allocations from the General Local Development Fund, for aimags receiving the highest and the lowest per capita allocation was 14.3 to 1, and for soums it was even higher (as of 2014, at which time the system was more indicator-based, and had only a minor derivation-based component). See Shotton (2017[18]).