Report No. 91343-GLB Intergovernmental Fiscal Management in Natural Resource–Rich Settings Lorena Viñuela, Kai Kaiser, and Monali Chowdhurie-Aziz Table of Contents I. Special Features of Non-Renewable Resources and Policy Issues .................................................................... 2 II. Revenue Sharing Instruments .................................................................................................................................. 5 III. Fiscal Principles on Natural Resource Management ....................................................................................... 7 IV. International Experience .................................................................................................................................... 10 V. Determinants of Revenue Sharing ........................................................................................................................ 14 VI. Macro-economic Considerations ...................................................................................................................... 17 VII. Conclusion............................................................................................................................................................ 22 References .......................................................................................................................................................................... 35 Intergovernmental Fiscal Management in Natural Resource –Rich Settings 1 1. In resource-dependent countries, natural resources constitute one of the main assets available for financing local governments because the economy is not greatly diversified. The goal of this note is to highlight different critical dimensions of intergovernmental fiscal relations in these settings, present a survey of the range of arrangements used for managing resource rents across multiple levels of government, and synthesize basic principles or considerations in the implementation of revenue-sharing systems across different contexts. 2. The design and implementation of measures to improve intergovernmental management of the oil, gas, and mining sector must consider the core policy objectives, fiscal context, and overall political structure. Paying attention to the constraints and political economy drivers that shape intergovernmental relations is critical to identify the feasible reforms and alternatives to improve performance that are available in a given country. I. Special Features of Non-Renewable Resources and Policy Issues 3. The inherently complex design of intergovernmental fiscal systems becomes even more challenging in the extractive sector owing to the distinctive technical and economic characteristics of oil, gas, and mining, and the interactions between these and institutional and political factors. The first characteristic is exhaustibility. There is a finite amount of these resources in the ground since they are formed by extended geological processes and cannot be easily replenished. The extraction of a resource in the present time reduces the amount of the ore body or field available in the future and as a result there is an associated cost, known as user cost. This exhaustibility introduces issues of inter-generational equity and optimality of the extraction profile as well. 4. At the same time, extractive industries demand long term planning for both government and companies. Exploiting mineral resources requires high frontloading of investments, which 1 This paper is co-authored by Lorena Viñuela, Monali Chowdhurie-Aziz, and Kai Kaiser are irreversible and highly specific to the industry. Significant exploration expenditures and risks precede startup, exploration expenses occur long before taxable income is available or even before a decision to mine or extract oil is made. It is also characterized by high economic and technological complexity and economic and geological risks for investors and governments that cannot be fully foreseen during the time contracts are being negotiated. 5. Notably, commodity prices often highly volatile. Producers are price takers and reach to changes in international prices. The more progressive a country’s fiscal regime is, the more vulnerable it is to price changes. On the other hand, progressivity allows the government to capture windfalls profits during boom periods. It is not uncommon that high prices trigger revisions of fiscal terms if they are regressive, which undermines the country’s credibility and the long term prospects for the sector. Depending on the overall soundness of macro-economic management (and whether spending is smoothed or not) and the design of the intergovernmental fiscal system this volatility could be directly to subnational governments with significant implications for their ability to plan and finance service delivery. 6. At the same time, there is a large diversity of mineral types (oil, sand, coal, base metals, gold, and diamonds) with diverse scales of operations and potential value added. Intergovernmental fiscal systems have to grapple with a multitude of revenue sources and variation in their volume and location. 7. Important differences exist between the mining and petroleum sectors. The life cycle of a mining project is considerably longer than projects in the petroleum sector. On the other hand, oil production generally generates higher greater rents than mining. Extraction costs per barrel may vary significant, however, depending on the prevailing geology and transport costs to market. Oil production tends to be more enclaved than mining. The footprint of mineral extraction in local communities is much more visible as mining tends to generate more environmental and social externalities, but it also has more positive spillovers in the local economy (Otto 2001). 8. Mineral and oil resources are generally concentrated in a small number of subnational units. In many countries producing regions are sparsely populated. For example, the autonomous okrugs of Yamal-Nenets and Khanty-Mansi account for 90 percent of the gas production and 65 of oil production in the Russia Federation, but have less than 2 percent of the national population (see Figure 1). Similar instances can be found in numerous countries. In other cases, most of the production takes place off-shore, but on the coast of few states or municipalities, such as in the case of Brazil. Figure 1: Major Oil and Gas Producing Regions in Russian Federation (2007) percentage 0 10 20 30 40 50 60 70 80 90 Yamal-Nenets Autonomous Okrug Khanty-Mansi Autonomous Okrug Orenburg Oblast Republic of Tatarstan Tomsk Oblast Astrakhan oblast Republic of Komi Republic of Bashkortostan Samara Oblast Perm Oblast Udmurtia Republic Nenets Autonomous Okrug Sakhalin Oblast Continental Shelf National population Proved reserves Oil production Gas production Source: Kurlyandskayai et al. 2012. 9. Such asymmetric distribution introduces concerns around horizontal equity and economic efficiency, in particular whether to earmark resource revenues and expenditures to the areas of production and what are the potential distortions introduced by such rules. In countries where a portion of royalties are earmarked to the producing regions, these tend to be concentrated in a small number of localities. In Colombia, until a recent reform, 48 percent of the royalties went to the producing departments and 13 to producing municipalities. In practice, this rule meant that two thirds of royalty revenues went to 5 (out of 32) departments (see Figure 2). To date, there has been practically no correlation between the amount of royalties received and departmental performance in reducing poverty or providing public services, despite the fact that some departments received several times the national average of transfers (World Bank 2011). 10. New finds and increases in production or prices can also exacerbate horizontal imbalances, unless there are considered within the overall intergovernmental transfer system and combined with other tools to equalize transfers. In Peru, where 20 percent of mining taxes are distributed to producing regions and 30 percent to local governments, the increase in prices in the second part of the 2000s coupled with rising production volumes meant that producing localities received significant windfalls, in many cases well beyond their absorptive capacity. In Brazil, new discoveries off of the coast of Rio de Janeiro has meant that state went from receiving half of royalties in 1997 to over 75 percent a decade later (see Figure 3). During the same period, the amount that was distributed to states and municipalities increased from 150 US$ million to 7 US$ billion, which meant that Rio de Janeiro’s share grew from 75 US$ million to 5.6 US$ billion. In Colombia, royalty payments almost triple (going from 1.8 COP trillion in 2000 to 5.2 COP trillion in 2009). Figure 2: Distribution of royalties by Figure 3: Distribution of oil revenues by department in Colombia (percent), 2008. region in Brazil (percent), 1997 and 2008. 25 80 70 20 60 50 percentage percentage 15 40 10 30 20 5 10 0 0 SAN ANDRES SUCRE NARIÑO BOYACA CESAR CORDOBA ANTIOQUIA CASANARE N. DE SANTANDER GUANIA AMAZONAS VICHADA META ARAUCA LA GUAJIRA SANTANDER BOLIVAR CALDAS VALLE DEL CAUCA HULA ATLANTICO RISARALDA VAUPES PUTUMAYO CUNDINAMARCA GUAVARE TOLIMA CAUCA CAQUETA MAGDALENA CHOCO Rio Grande do Sul Pernambuco Rondônia Acre Roraima Pará Tocantins Distrito Federal São Paulo Ceará Mato Grosso do Sul Amazonas Paraíba Minas Gerais Espírito Santo Alagoas Santa Catarina Maranhão Bahia Paraná Goiás Piauí Rio Grande do Norte Mato Grosso Amapá Sergipe Rio de Janeiro 1997 2008 Source: World Bank 2011. Source: ANP, Gobetti et al. 2012. 11. These features of natural resource revenues need to be considered when assessing the potential benefits and costs of the various intergovernmental fiscal arrangements. II. Revenue Sharing Instruments 12. Countries generally use a mix fiscal and nonfiscal instruments to mobilize revenues from extractive industries, each with its own benefits and disadvantages along economic, administrative, and revenue-enhancing dimensions.2 Nonfiscal alternatives include auctioning 2 For a more extensive discussion see Barma et al. 2012 (chapter 4) and Le and Viñuela 2012. exploration and extraction rights, production sharing, and equity participation. Fiscal instruments comprise royalty (specific and ad valorem), corporate income tax, presumptive income tax, resource rent tax, and property tax, as well as other taxes such as value added tax, and import and export duties (Otto and Andrews 2006, Sunley et al. 2003). 13. In turn, petroleum and mineral revenues can be shared vertically across levels of government using a variety of arrangements, which are summarized in Table 1. Nonfiscal options imply sharing part of the resources and revenues received as part of production sharing agreements or equity participation. They can also include in-kind revenue such as capital assets received as part of resource-for-infrastructure deals (Foster 2009). Fiscal arrangements range from separation of tax bases to intergovernmental transfers (Broscio 2006). Table 1: Instruments for sharing rents from natural resources Method Separation of Tax Concurrence of Sharing of Sharing of Intergovernmental Bases (own-source Taxes (sharing of Revenue Revenue In-Kind Transfers out of taxes) tax bases) Revenue from Natural Resources Determination Subnational National National Mostly National National of the tax base Determination Subnational Subnational National Mostly national National of the tax rates (within limits) Administration Subnational Mostly national National By the producing Mostly national firm Criterion for Origin Origin Origin Origin Need, equity, or beneficiary other jurisdiction Source: Brosio 2006, p. 441. 14. Whereas most alternatives allocate revenues according to the principle of origin, each of these distributes authority over the tax base, rate, and administration in a different manner. In the case of the separated tax base system, national and subnational governments are entitled to levy separate taxes on mineral production using different instruments (e.g. the national government collecting income tax and state government collecting royalties). The national and subnational governments separately administer their own instruments. In a tax-base-sharing arrangement, two or more levels of governments could tax the same base using the same instrument with the same or different rates (e.g. each level collecting different royalties). Tax revenue sharing normally implies that tax bases, rates and the percentage accrued to producing regions are determined by the central government. 15. The complexity and administrative costs associated with the various tax instruments limit the options available for subnational governments in the cases where the tax base is shared. Resource rent taxes, which are more progressive, impose considerable administrative costs and require greater technical capacity (Brosio 2006). As a result, state and provincial governments generally prefer to directly levy royalties (Otto 2001) because that system not only is simpler but also reduces delays and variability in revenues (McLure 2003). Levying royalties at both levels of governments, however, potentially can lead to vertical externalities by increasing the overall burden of the tax. 16. Intergovernmental transfers, which are a grant from the central government that raised the funds to lower tiers of government, can vertically channel resources on the basis of origin or using other criteria such as equity. In most cases, transfers systems combine grants which have equalizing objectives with those that separately compensate producing regions. For example, Nigeria distributes the funds in the Federation Account (which is almost entirely financed by oil receipts) dividing 40 percent equally among all states, and the rest according to population, land mass and terrain, social needs, and internal revenue efforts. However, oil producing states receive an additional 13 percent of oil revenues generated in their territory. III. Fiscal Principles on Natural Resource Management 17. The general literature on intergovernmental fiscal relations and decentralization recommends that a function is assigned to the level of government that would be able to conduct it with the greatest possible efficiency. If the functions require adapting to different needs, local governments would be in a better position to elicit information from citizens and deliver the mix of policies that better adapt to their preferences and economic conditions (Oates 1972). Conversely, functions and policy areas in which there are clear economies of scale or spillovers would be better served by the central government. Yet even in the areas where responsibilities have been devolved to lower tiers of governments, central governments retain significant roles in setting standards, regulation, and financing. 18. This normative literature is largely concerned with the potential efficiency and equity gains of decentralization. If properly implemented, devolution should allow local governments to choose different tax-expenditure mixes that best accommodate to heterogeneous local preferences and circumstances (Brennan and Buchanan 1965) and it should lead to gains in efficiency through local informational advantages, increased accountability, and competition and experimentation among local governments (Oates 1972). 19. Students of decentralization are also preoccupied with addressing vertical and horizontal fiscal imbalances arising from the gap between the distribution of functions and revenues (Schroeder 2001, Shah 2007). Central governments are generally more efficient at collecting taxes than subnational governments. As a result, the more decentralized functions are in a given country the greater the fiscal gap would be. If some revenue collection responsibilities are transferred to the lower levels, the gap could be smaller. But the gap ultimately depends on how effective tax administration is at the subnational level and the size of the revenue base that exist at that level. Whatever the case may be, transfers are likely to be an important part of subnational revenues in all countries. 20. Transfers and revenue sharing systems could be designed to solve these gaps and counteract some of the negative incentives associated with transfer. At the same time, transfers to subnational governments can be used to compensate for differences in needs and fiscal capacity by redistributing resources across jurisdictions. 21. Following this logic, literature on the assignment of revenue from natural resources recommends that subsoil natural resources are managed at the national level. Federal or central governments are better placed to collect revenues from extractive industries, which are complex and difficult to implement. Assigning tax collection to a single level of government has the additional benefit of preventing vertical externalities and overtaxation of the sector (Brosio 2006). A central system of natural resource rent collection also introduces efficiency by reducing administration and compliance costs and by allowing ring-fencing of projects across jurisdictions (McLure 2003; Mieszkowski 1983). 22. Revenue sharing or transfers are preferred over the assignment of own-source taxes to subnational governments and sharing tax bases (Boadway and Shah 1994; McLure 1983). As a result of the uneven geographic distribution of natural resources and population, assigning rents exclusively to the state or regional level could lead to considerable horizontal imbalances, and even have political and economic destabilizing effects. In addition, there is uncertainty over the amount and location of resources and where future resources may be found, so it is in the interests of subnational governments to allow for some redistribution. 23. A key concern is not just the amount but the predictability of fiscal transfers that subnational governments receive. It is likely that subnational governments are less well placed than national governments to cope with significant revenue volatility. Resource revenues are subject to volatility arising from these sources, including variable rates of extraction over time, payments from producing companies, and prices. Subnational governments need some degree of medium-term revenue predictability in support of sound budgeting and execution. If subnationals depend on central transfers, they likely will be quite vulnerable to adjustments by the central government. Even if transfers are based on rules-based criteria, transfers based on resource revenues can be subject to significant volatility (for example, due to price fluctuations, a fixed royalty share for subnationals may vary significantly). 24. The theory of federalism offers economic arguments to guide decisions on the amount of resources to be distributed to subnational governments and how they should be allocated across localities. First, subnational governments should be refunded for the additional costs and investments on local infrastructure that exploiting nonrenewables require before the rents are distributed (Bahl and Tumennasan 2004). Because resources tend to be concentrated in few subnational jurisdictions, such excess cost is imposed unevenly and would be frontloaded to a large extent. 25. Second, negative environmental externalities associated with the exploration and exploitation of mineral resources should be internalized in taxes and fees and be used to compensate the producing subnational units that bear these (Ahmad and Mottu 2002, Brosio 2006). Similarly, there may be compelling grounds to compensate regions with extractive industries for additional costs incurred to provide additional services for in-migrants associated with the industry, in the absence of the ability to raise these revenues directly (Bahl and Tumennasan 2004; McLure 1983). This requires some measure of what these costs are, and if resource revenues are earmarked for these purposes. 26. However, there are other arguments made in favor giving additional revenue allocations to the originating regions on the basis of heritage. If a country’s constitution has given regions ownership rights to subnational governments, these should be compensated for the user cost. Ideally the funds collected from the extraction of a mineral would be reinvested in capital goods that can replace the depleted natural wealth (Hartwick 1977) or preserved for future generations. The underlying rationality is that nonrenewable are part of a country’s assets, and thus, consumption of revenues resulting from sales should more accurately be classified as consumption of capital instead of consumption of income (Dabán and Hélis 2009; Humphreys, Sachs, and Stiglitz 2007). 27. Nevertheless, if all or a large portion of resource rents are earmarked to region of origin, the central government and nonproducing localities might not have adequate revenues to fulfill their functions, while producing regions are able to provide more public services and investment than the rest of the country. In this case, there will not only be vertical and horizontal disparities, but also inefficiencies given that the social return rate of investment would likely be lower in resource-rich but sparsely populated regions than in other jurisdictions (McLure 2003). IV. International Experience 28. In practice, the degree and nature of the participation of subnational governments in the management of subsoil resources depends on how ownership, regulation and expenditure responsibilities have been distributed across levels of governments. This distribution of functions is the result of a complex process of institutional development and political bargains that do not necessarily focus on equity or efficiency of public expenditures. 29. There is a significant diversity in how countries have addressed the extractive sector in their intergovernmental fiscal system. Table 2 provides a stylized overview of intergovernmental management across resource-dependent settings. Countries are classified by the extent of vertical claims over subsoil assets (rows) as well as if ownership is vested in the national or provincial/state government (columns). It is important to note that countries could be further classified along a number of other dimensions. Table 2: Tax Assignment Instruments and Ownership of Sub-soil Resources Ownership Tax Assignment National Ownership Subnational Ownership National budget/full Angola Norway centralization Algeria Oman Azerbaijan Qatar Bahrain Saudi Arabia Chile Timor-Leste Iran, Islamic Rep. of United Kingdom Kuwait Yemen, Rep. of Libya b Downward revenue Ecuador Peru Argentina c sharing Ghana Brazil Australia d Indonesia Colombia India Kazakhstan Iraq Mexico Nigeria Malaysia Venezuela, R. B. de Papua New Guinea a e Concurrence of taxes Russian Federation Canada f (sharing of tax bases) United States Upward Revenue-sharing United Arab Emirates Arrangement Source: Author. Note: a. Joint ownership of onshore resources between the Russian Federation and subnational entities. The continental shelf is owned exclusively by the Russian Federation. b. Provinces own subsoil resources and control exploration licenses and exploitation concessions. c. States have ownership and control of resources onshore and the federal rules apply to coastal waters and territorial sea. d. States own resources onshore, but the federal government regulates and controls exploration licenses and exploitation concessions. e. Provinces own and manage onshore resources. The federal government owns offshore resources and shares control and revenues with coastal provinces. f. States own and regulate subsoil resource management, except in federal land and offshore. 30. In the majority of countries ownership of natural resources and related revenues are vested in the central government. Yet in a number of federal settings—including Argentina, Australia, Canada, India, the United States and the United Arab Emirates—state or provincial governments have rights over subsoil assets. In most cases, the federal government retains rights over off- shore assets and those found in federal land. There is also a larger set of countries where subnational governments receive a share of revenues and royalties from extractive activities carried out within their geographical area. 31. A wide range of disparity exists in the extent to which revenues are spread across levels and in the overall degree fiscal and administrative decentralization. There is no strict correspondence between the form of government, either federal or unitary, and the more or less decentralizing character of revenue-sharing systems. However, only in federal countries do subnational units have ownership of nonrenewable resources. 32. Numerous countries use derivation-based revenue sharing with producing subnational units. For example, in Brazil a formula determines the amount of transfers according to population size and per capita income, while states and municipalities receive additional shares of royalties according to the level of onshore or offshore production. Table 4 provides a more detailed synthesis of derivation-based natural resource fiscal transfers in a selected number of more decentralized settings, including Indonesia, Nigeria, Papua New Guinea, Peru, Sudan, and Venezuela. 33. Moreover, some countries have implemented asymmetric or mixed models relative to resource extraction subnational entities. In addition to basic revenue derivation-based sharing, Indonesia granted additional allocations to its special autonomy regions of Aceh and Papua. Historically, these regions had been associated with significant secessionist tendencies and poor developmental outcomes. In an analogous manner, the Malaysian oil-producing provinces have a degree of special autonomy in addition to derivation-based revenue sharing in the context of a highly centralized federation. 34. A comparable variation is observed based on which level of government is responsible for setting tax bases and rates and carrying on the collection of rents from nonrenewable resources. In the vast majority of countries, taxes are set and administered by the national government and then are shared with the lower tiers. In several cases, however, federal and second-tier governments impose their own taxes on a shared basis as in Canada and the United States. The United Arab Emirates is the only country in which rents are collected solely by subnational governments and then shared with the central government in cash and in-kind. The emirates have full ownership over subsoil resources and collect royalties, company profit transfers, and income tax receipts. The transfers are negotiated yearly to smooth the budget (Ahmad and Mottu 2002). 35. Intergovernmental systems also distribute petroleum and mineral revenues that are raised through nontax mechanisms like production sharing and equity participation. Subnational governments also received revenues from production sharing in Indonesia, Malaysia, and Sudan and from equity participation in Papua New Guinea. 36. The degree to which countries have been able to implement predictable and rules-based transfer systems varies significantly. Although many countries have relatively transparent rules, in practice transfers are often subject to manipulation and delays. Transfers need to be transparent and predictable, based on need and fiscal capacity, and need to provide sufficient resources to undertake the devolved functions to yield results (Bahl and Linn 1994; Bird 2001; Ma 1997; Shah 2004). In the case of resource rent, transparency and predictability have additional implications given that uncertainty about the division of resource rents has been cited repeatedly as a central contributor to conflict and political volatility in resource-dependent countries (Herbst 2001). Given conceptions of ownership, subnationals may have a valid claim to receive a significant part of rents from a nonrenewable resource based on heritage arguments. Figure 4: Decentralization and Resource Figure 5: Decentralization and Revenue Rents (percent), 2008. Dependence (percent), 2008. Source: World Bank Decentralization Indicators 2012, Source: World Bank Decentralization Indicators 2012, Wealth of Nations Database 2010. Barma et al. 2012. 37. Overall, there is no apparent relationship between the types of intergovernmental distribution of functions and resources and the form of government or degree of decentralization. As well, resource abundance is not, positively or negatively, associated with the overall degree of decentralization, measured as the share of public expenditures carried out by subnational governments (see Figure 4). If only resource dependent countries are considered, or those that derive more than 25 percent of GDP or public revenues from extractive industries, it appears decentralization is negatively correlated with resource dependence for this set of countries. However, this association should be considered with caution as this is a truncated sample. V. Determinants of Revenue Sharing 38. The features of intergovernmental systems are largely dependent on how governments prioritize often-competing aims of national cohesion and conflict avoidance, local service delivery and effective macro-management. Whether more centralized or decentralized arrangements for managing oil and mining resource and their rents are more or less appropriate to a particular circumstance depends both on these objectives and the underlying political economy underpinnings of intergovernmental relations. 39. In opposition to the recommendations of the normative literature on fiscal federalism, most countries share revenues with subnational governments and have earmarking arrangements for producing regions. Such arrangements are the result of political bargains and historical legacies that are generally difficult to change. A recent stream of positive literature on federalism offers a more useful insight on why this may is the case. Recognition is growing that performance of federal and decentralized governments depend on their design, representative institutions, and parties (Eaton 2004), which in turn emerge as a response to underlying social, cultural, or demographic factors (Rodden 2007). 40. The so-called second generation literature has emphasized that fiscal federalism is frequently introduced as a means to protect the wealth of rich regions against threats of expropriation by other regions. Asset specificity and capital mobility, as well as horizontal inequality among regions shape the locus of tax power and expenditure responsibilities (Boix 2003). In settings with unevenly distributed endowments and pronounced income differences between regions, the redistribution preferences of the median voter in poorer regions can create pressures for tax centralization, which rich provinces will oppose. Without a strong center that is capable of redistributing rents across regions, federations risk dismembering. In the long run, if federal institutions are maintained, they contribute to shaping interregional inequality (Beramendi 2007). 41. The allocation of rents from natural resources is frequently a vexing political question in developing countries where rents from natural resources account for a large share of total government revenues. The uneven distribution of natural resource rents creates additional strains on national unity and contestation of central-local government relations as it generally entails that small resource-rich (and often sparsely populated) regions share rents with larger areas. 42. Particularly in low income countries strong subnational claims on natural resources or historical neglect can be a source of conflict. If the people residing in the resource-rich regions are dissatisfied with the returns (mostly the resulting increased incomes) from the presence of natural resources because it does not meet their expectations, this may lead to political and social unrest (Ross 2007). Thus, citizens residing in a peripheral region may favor independence due to the appeal that ownership rights of the natural resource holds (Collier and Hoeffler 2002; Ross 2007). The recent attention to the impact of extractive industries on local communities and rights of indigenous people over such resources has generated calls for greater sharing (Otto 2001). Such claims, however, often are intertwined with a sense that central authorities have historically neglected resource-rich regions, and hence these regions need resources to provide these services. 43. In this context, federalism is often viewed as a means to balance centripetal and centrifugal forces in large and ethnically or culturally diverse societies (Filippov, Ordeshook, and Shvetsova 2004; Treisman 1999). By increasing development at the local level (Bakke and Wibbels 2006), reducing discretion and ambiguity of natural resource rents (Bird and Ebel 2006; Herbst 2001), and providing citizens with opportunities to shape policies (Brancati 2009), revenue-sharing mechanisms have the potential to solidify national unity and diffuse separatist sentiments. In addition, decentralization can enhance local autonomy and offers a setting in which minorities can enjoy self-rule at the regional level (Jeong 2002). Alternative regional arenas favor intragroup competition, which in turn reduces power struggles at the national level (Horowitz 2002). 44. In contrast, others warn about the potential negative incentives that sharing resource revenues can generate. When ethnic or linguistic minorities with historical grievances are concentrated in the same jurisdictions in which natural wealth is found, sharing revenues can strengthen their position to contest national unity. Thus, overlapping ethnic cleavages and distribution of resources can encourage disintegration (McLure 1994). Decentralization can increase ethnic conflict and secessionism indirectly by reinforcing ethnic and regional identities and promoting growth of regional parties (Brancati 2009). In the absence of political competition and adequate downward accountability mechanisms at the subnational level, decentralization can lead to elite capture, corruption, and weak government performance that reinforce regional inequalities (Jeong 2002). In addition, if no national consensus is reached and subnational actors view revenue-sharing arrangements as inequitable, conflict is likely to persist. What is equitable is largely going to be defined by the particular political context in which allocations across different regions are made and also is subject to changes over time (Herbst 1989, 2001). 45. Ultimately, the prospective outcome and options depend on governance thresholds and on overall characteristics of the political system, especially the degree of nationalization of parties and the availability of political and judicial channels to solve disputes over resources (Collier and Venables 2009). Bakke and Wibbels (2006) test for the determinants of intranational conflict in 22 federal settings. Their results suggest that “peace preserving merits of federalism depend not only on the design of institutions, but how these respond to the characteristics they govern.� Regional inequality and ethnic diversity are two key variables to consider in this regard. Copartisanship between the center and the regions averts conflict and holds the state together only when the national governments are inclusive to ethnic minority and majority regions. 46. There is no blueprint for intergovernmental institutions in these settings, but the nature of regional disparities (including natural resources), as well as party representation and political power should be taken into consideration. Care must be taken that intergovernmental design does not become too narrowly focused on distributional issues, but rather includes the range of complementary institutions that allow for credible and sustainable multilevel government and governance. These include the presence of checks and balances mechanisms (for example, through judicial institutions and inter- or intra-political party bargaining). In the absence of such institutional mechanisms, intergovernmental fiscal design could degenerate to short-term contestation. Given significant price volatility associated with natural resource revenues, the combination of weak complementary institutions (including revenue management) and shifting demands can make for a precarious mix. 47. An important aspect of addressing these issues is to see this debate as being about significantly more than simply granting more resources. Subnational regions may be concerned about a number of other issues, rather than maximizing transfers in the short run. These issues may be about greater autonomy in certain areas. Such concerns, for example, can be addressed with asymmetric approaches to different regions, including those richly endowed with natural resources (Bird and Ebel 2007). Above all, when spending capacity of subnational governments are weak, alternative spending mechanisms to provide benefits to the population of resource-rich regions could be pursued. These could include unconditional or conditional cash transfer programs, linked to the existing resource wealth. Particular measures could be taken to encourage performance on the part of resource-rich regions. VI. Macro-economic Considerations 48. A central issue is how to balance in the design of intergovernmental fiscal systems macroeconomic sustainability objectives and political pressures. Resource-dependent countries that were best able to weather the commodity price shocks of 2008/09, whether decentralized or not, had put in place effective medium-term revenue management measures. Numerous countries have formal saving or stabilization funds. Yet the rate that is saved during high commodity price and production periods varies considerably, as does the success of ensuring medium-term stability in government revenues during downturns (see Table 5). On the other hand, some countries like Bolivia and Iraq saved substantial amounts of resources because of their low budget execution, whereas others have built up foreign reserves, as in Russia and Kazakhstan. Although accumulated savings have been significant in the recent commodity boom period, they quickly can become vulnerable to rapid annual drawdowns outside of some established institutional criteria and mechanisms. 49. How windfall revenues are used and allocated across levels of government during commodity booms determines how much fiscal space countries had during the price bust to smooth spending. If windfalls are used to finance current expenditures and wages and expand nonoil deficits, the social adjustment cost likely would be higher than if spending is delinked from commodity revenues. All political systems face intense pressures to spend during boom times, especially those in developing democracies where social and infrastructure needs are great. At the same time, investing or hedging oil savings implies high political risk if oil prices continue to increase and savings are lost. Recent experiences in Chile, Indonesia, Mexico, and Nigeria show that policy lessons from previous commodity cycles have motivated national authorities to follow prudent fiscal policies and avoid procyclical spending and excessive borrowing, both at the national and subnational levels. Box 1: Mexico’s Windfall Sharing and Hedging Experience Mexico’s federal government remains relatively dependent on o il revenues. Mexico is the seventh-largest oil producer in the world. The sector generates approximately 16 percent of export earnings but more than 35 percent of total government revenues. Although oil revenue in Mexico represents a significant share of revenue for the federal budget, state and local governments have been shielded from price volatility through several mechanisms. First, transfers are made after deducting energy subsidies. Second, only a limited part of petroleum rents are redistributed; 20 percent of ordinary extraction rights are incorporated into a general fund and distributed to the states based on a formula taking into account their characteristics. Producing municipalities received an extra 3.17 percent to compensate for the environmental damage that resulted from oil extraction. After the reference price (US$70 per barrel) is surpassed, however, states receive an additional share of the windfalls, which diffuses pressures for redistribution during boom times. Third, a stabilization fund receives 10 percent crude oil production when the price is more than US$32 per barrel and is on a sliding scale that starts at 1 percent when the price is below US$22 per barrel. Part of the saved revenues have been use to diversify risk. Both in 1990 –91 and in 2007–08, Mexico sold oil futures and used oil swaps to hedge price risks with highly successful results. These measures have made shared revenues more stable and have allowed for countercyclical public spending at the national and subnational levels. Revenue-sharing rules are highly transparent and predictable and are accompanied with a high degree of consensus over the need for fiscal discipline. In 2009, the distribution formula was changed to encourage local governments to mobilize additional resources. States are guaranteed a threshold equivalent to the transfers received in 2007, but they will get additional transfers according to the relative weight of their own tax collection. Mexico has successfully diversified its economy. Nevertheless, it faces major challenges to diversify the sources of fiscal revenues and to increase the competitiveness of its oil industry. The monopoly held by the state-owned Petroleos Mexicanos (PEMEX), which suffers from chronic underinvestment and has made little exploration efforts in the past two decades, is likely to lead to a decrease in the levels of production in the near future that can compromise fiscal stability. In 2008, the Mexican Congress passed a law reforming the internal governance and accountability of the company to reverse the situation. Macrofiscal Management: Aligning Incentives across the Intergovernmental System 50. High fiscal dependence on extractive sector creates vulnerability to price cycle and can lead to fiscal volatility. However, volatility can be addressed by smoothing oil revenues, most commonly achieved via saving funds and hedging of oil revenues that prevent the transmission of oil price volatility to the budget (Davis, Ossowski, and Fedelino 2003). Nevertheless, given various spending pressures, uncertainty about oil price, and political risk, stabilization institutions are difficult to implement credibly. Revenue diversification by broadening the overall tax base is another key strategy. 51. The theory of federalism explains that central governments are in a better financial and political position to smooth spending across boom cycles and offset the volatility associated with natural resources, which tend to be significantly higher than other sources of fiscal revenues. National governments have larger budgets and access to nonresource revenues and financial markets (McLure 1994). Optimally, when governments face price fluctuations, they would impose stabilizing rules to the intergovernmental transfers to protect subnational governments from the boom-bust cycle and smoothing will be done before the division of rents across levels of government. During boom periods, these rules would prevent current expenditures from being increased beyond a sustainable point, whereas in downturns, the central or federal governments would be able to use the saved funds to finance subnational deficits and resort to financial markets if necessary to secure additional resources (Gonzalez, Rosenblatt, and Webb 2002). 52. On the other hand, diversifying subnational sources of revenues could reduce vulnerability to price shocks of second- and third-tier governments. Local governments can raise a number of taxes without introducing major distortions on the extractive sector that discourage investment, such as property taxes and user fees (Otto 2001). These taxes are not linked to profitability and price fluctuations and are applied throughout the project life cycle, thus providing a more stable source of revenue. In addition, improving the overall capacity of subnational governments to tax nonextractive activities should be pursued concurrently. 53. Ideally, intergovernmental fiscal arrangements in resource-dependent settings will be consistent with overall revenue management and stabilization, and with the criteria by which revenues are assigned to subnational governments. Intergovernmental fiscal systems might distribute the burden of stabilization across levels of governments in a different way and central governments might need to resort to bargaining or to other incentives to have discipline at the subnational level. Indonesia offers an example of a system that has combined overall revenue management strategies with a significant degree of fiscal decentralization (Box 2). 54. A key policy challenge in the presence of significant subnational assignments of resource revenues is to design some type of credible coordination device that encompasses both national and central levels of governments. The criterion of subnational revenue predictability suggests that particular emphasis should be placed on ensuring that subnational governments have some degree of medium-term predictability. Simple fiscal rules with conservative resource prices are typical strategies. Subnational governments may feel that national governments could use conservative pricing to withhold resource revenues from them. As a result of poorly designed sharing schemes, subnationals could begin to lobby for overly optimistic prices. A more promising strategy would be to ensure that subnationals see that it is in their interest to adhere to common fiscal rules. This could be achieved, for example, by providing them with access to stabilization mechanisms if prices in the future are less buoyant. Box 2: Indonesia: Revenue Sharing, Equalization, and Special Autonomy Indonesia’s 2001 “Big Bang� decentralization saw the introduction of significant derivation -based natural-resource revenue sharing. The overall decentralization and the introduction of resource revenue sharing were driven largely by the desire to keep the country together, particularly after a history of decades of highly centralized authoritarian rules. Resource-rich provinces, in particular, had a sense that they did not adequately benefit from their resource wealth (Bahl and Tumennasan 2004). Before the far-reaching decentralization reform, all petroleum and mining revenues accrued to the central government and were not subject to any revenue-sharing arrangement with provincial and district-level governments. Since 2001, the central government distributes at least 25 percent of central government revenue to lower level governments through a general transfer mechanism through an equalization grant pool. In addition, producing districts and provinces were granted 15 percent of onshore oil revenue and 30 percent of onshore gas, whereas 64 percent of mineral royalties went to the regions and 16 to the provincial governments. Concurrently, to reduce centrifugal pressures, Aceh was granted 55 percent of oil revenues and 40 percent of rents from gas, and Papua was granted 70 percent of both. Derivation-based sharing of oil revenues has not been without its challenges. Derivation-based sharing was associated with significant revenue disparities. In many cases, beneficiaries were not able to spend the windfall, and instead built up significant cash reserves. The aggregate buildup of reserves fueled the perceptions that provinces and districts were receiving too many transfers relative to their expenditure responsibilities (Lewis 2005). The allocation of natural resource was to some extent offset by the introduction of a formula-based equalization grant (the DAU), which now accounts for more than 25 percent of total national revenues (net of shared revenues). The formula incorporates measures for fiscal capacity, which mean that regions benefiting for resource revenues receive fewer equalizing transfers (Hofman et al. 2006). The government has addressed the issue of fuel subsidies. In the past decade, Indonesia has been a borderline net energy exporter and importer. The fact that energy prices were pegged below international prices resulted in the rapid growth of fiscal liabilities in this area. During the oil boom, nearly one-third of central government expenditures were claimed by fuel subsidies, while derivation-based transfer liabilities were increasing. Although the government was able to institute political unpopular price increases (in part through the introduction of a large compensation package, including conditional transfer schemes to households), energy subsidies continue to be a reality in Indonesia. Starting 2009, the government reduced the divisible DAU pool under a burden-sharing program to account for energy subsidy liabilities. Fuel subsidies are deducted from the rents before sharing. Moreover, the government has earmarked petroleum revenues, and in 2009, an additional 5 percent of oil and gas revenues were transferred to regional governments for spending on primary education. This policy was pursued as part of a prudent fiscal policy and countercyclical macroeconomic policies for stabilizing the external and internal balances. Natural Resources and Subnational Governance: Linking Natural Resources to Outcomes 55. Fiscal autonomy may create incentives for local government to expand the local tax base by promoting economic growth and moderating tax rates. These incentives are weakened by the presence of significant natural resource revenues. By reducing the need to mobilize resources locally to finance public services, subnational governments have less need to take into account the demands of taxpayers. In this manner, resource-rent dependence at the subnational level poses accountability challenges that negatively affect the quality of service delivery in the localities in which these revenues are concentrated. Historical evidence from the United States points to poorer outcomes in the resource-dependent states, notably Texas and Louisiana (Goldberg et al. 2008). 56. Emerging evidence from developing countries shows that large transfers to producing states and municipalities have not yielded the social and growth outcomes that were anticipated. For example, in Brazil, 60 percent of royalties are paid by the national oil company Petrobras to states and municipalities on whose territory the oil is extracted. Although per capita public spending on education and health is higher in these communities, local populations have not experienced improvements in the quality and quantity of public services (Caselli and Gennaioli 2007; Caselli and Michaels 2009). Most of the windfall revenues have been used to increase government wages. At the same time, households have reported less welfare income than their counterparts in nonproducing districts and no visible improvements on the quality of housing and infrastructure have been achieved. Moreover, municipal gross domestic product growth in these localities fell after the introduction of royalties in 1997 (Slaibe Postali 2009). 57. Linking the intergovernmental transfers to performance and downward accountability mechanisms is crucial to prevent shared rents from being captured by elites or being used to enlarge subnational governments with no real impact on the quality and quantity of public services. Efforts to build local capacity and mobilize stakeholders for good governance should be an integral component of decentralization policies. Simultaneously, enhancing the ability of the states to raise their own revenues can increase the accountability of subnational governments for their fiscal performance (Boadway and Shah 2009). VII. Conclusion 58. The design and implementation of measures to improve intergovernmental management of the oil, gas, and mining sector must consider the core policy objectives, fiscal context, and overall political structure. Paying attention to the political economy constraints and drivers that shape intergovernmental relations is critical to identify the feasible reforms and alternatives to improve performance that are available in a given country. Most of the attention in this area has been focused on benefit sharing or distribution. Intergovernmental relations, however, are shaped by the underlying distribution of political, economic, and demographic endowments and as such are highly politicized. Policy makers need to address both the aggregate challenge of effective management of natural resource extraction and managing its fiscal benefits across various policy objectives. In many settings, this entails managing across time and across levels of government and territory. 59. Practitioners need to be able to draw on the normative literature and actual international experiences to address the concerns of their specific country context. This survey of different experiences highlights the importance of achieving a consensus simultaneously around stabilization and intergovernmental fiscal rules. Choices and trade-offs need to be considered in light of the particular historical political-economy circumstances of the country of interest. Intergovernmental bargaining not only should not be centered on benefit sharing, but also should consider the systemic effects that these rents have on fiscal sustainability. Although constitutional design (for example, federal or unitary) can play a critical role in framing policy options, de facto implementation ultimately will determine how successful resource-rich countries with significant intergovernmental dynamics are in harnessing natural resources for development. 60. Decentralization in these settings should not expect inherently better or worse outcomes, but rather understand that results will depend on the institutional context of each setting. These findings echo the thrust of the recent literature on natural resources (Dunning 2008, Haber and Menaldo 2011). Natural resources are not intrinsically good or bad for development, but developmental impacts depend on how countries take advantage of them. In some cases, intergovernmental dynamics can provide useful checks and balances (or spaces for subnational innovation on central authorities), whereas in others, they can weaken effective intergovernmental coordination. In the extreme, poor management and weak institutions may undermine national cohesion or even increase the risk of conflict and civil war. 61. Countries have resorted to different intergovernmental arrangements to achieve diverging objectives. The cross-sectional survey reveals that intergovernmental fiscal design choices are mostly determined by how governments prioritize the competing objectives of territorial cohesion, economic stability, equalization of service delivery standards, and maximization of the wealth derived from natural resources across space and time. Institutional alternatives, therefore, should be evaluated in light of the core objectives that each country is pursuing and should consider how well they are able to juggle these with the other goals. The underlying political economy conditions will determine whether objectives are conflicting and how successful countries will be in using intergovernmental systems to solve them. 62. The variation is reflected in differences in the vertical share of natural rents allocated to subnational entities, the criteria used to distribute them across second- and third-tier governments, and the instruments through which the rents are shared. The size of the pool of resources that are to be shared across cases ranges widely. Some central governments allocate resources to producing and nonproducing localities; however, in other countries, resource-rich states and provinces enjoy especial autonomy and receive a larger part of the rents. Rents are divided by giving the different levels of government tax prerogatives over a shared base or through vertical transfers. At the same time, transfers, which can be automatic or linked to performance, are allocated through derivation, cost reimbursement, or on the basis of an equalization formula, or are earmarked to a particular public service. 63. The amount of rents being shared is as important as the transparency and credibility of the intergovernmental sharing arrangements. To fulfill their functions effectively, subnational governments need a stable and predictable institutional and fiscal environment. Transparency and predictability will contribute to ease intergovernmental tensions and facilitate policy dialogue around possible reforms. Smoothing resources available to subnational governments across commodity price cycles is an important factor in this regard. Lastly, linking financing to better sectoral and subnational priorities and improved performance is critical to avoid some of the negative consequences that excessive dependence on natural resources may generate for governance and accountability. 64. The choice of intergovernmental systems not only is the result of different goals, but also is influenced by the resource profile and the overall degree of diversification of the economy of a country. Whether a country is a mineral or a petroleum producer, how much known reserves it possesses, where they are located, and the level of fiscal dependence will determine the range of intergovernmental arrangements that are fiscally and politically feasible. Countries that are highly dependent on resource rents will be able to distribute a smaller share of total resource rents, and their ability to compensate the producing regions for the extraction of nonrenewable resources will be more limited than countries with a more developed and diversified economy. A key concern will be whether prevailing legal and regulatory frameworks, policy-making mechanisms, and most important, de facto implementation are immune (or flexible) to the volatility in natural-resource prices and territorial production patterns. Policy makers need to strategically assess the extent to which prevailing arrangements are consistent with the effective management of the sector (for example, in terms of exploration and exploitation patterns), national cohesion and stability, macrofiscal sustainability, and broad-based development. At the same time, policy makers and international development partners need to remain continually aware that feasible options will be conditioned by the political-economy setting. 65. Yet in any context it would be critical to: (i) align incentives across the intergovernmental systems, reducing moral hazard; (ii) achieve consistency of intergovernmental fiscal systems and macroeconomic stability (iii) smooth revenue volatility before sharing the benefits; (iv) diversify subnational revenues; (v) link natural resource finance transfers to outcomes (transparency in the transfers and downward accountability mechanisms); (vi) secure national unity without introducing large horizontal inequalities; and (vii) enhance predictability of transfers and clarity on the rules (over quantity). Table 3. Basic Characteristics of Resource Dependent Countries Country Population Area GDP per Capita EI Rents per EI Share of Total EI Revenue Oil and Gas Oil and Gas Mining Share of Mining Revenue National Oil/ Extractive Stabilization Subnational (millions, 2010) (1,000 sq km) Capita Exports (%, Share of Total Share of Total Revenue Share Total Exports Share of Total Mining Industry Fund Expenditures as (US$, 2010) Public Revenues Exports (%, (%, 2007-2009) Company % of Total (%, 2007-2009) 2007-2009) Public Expenditures Algeria 34.4 2,382 4,460 2,111.5 98.5 68.6 98.5 68.6 .. .. Yes No No n.a. Angola 18.0 1,247 3,960 3,138.0 95.6 83.6 95.6 83.6 ¾ ¾ No No No n.a. Azerbaijan 8.7 87 5,180 3,661.1 94.7 53.4 94.7 53.4 .. .. Yes Compl. No 24.5 11,407.9 .. No Bahrain 0.8 1 25,420 79.8 85.5 79.8 85.5 .. No No 2.8 Bolivia 9.7 1,099 1,790 606.8 63.5 70.0 44.4 26.0 19.1 44.0 Yes No No 30.6 362.4 64.2 No Botswana 1.9 582 6,890 64.2 13.9 .. .. 13.9 No No 4.8 Brunei Darussalam 0.4 6 31,180 23,436.0 97.2 60.0 97.2 60.0 .. .. No No No n.a. Cameroon 18.9 475 1,160 152.9 33.3 38.2 33.3 38.2 .. .. Yes Yes No n.a. 392.3 .. No Chad 11.1 1,284 600 84.3 57.1 84.3 57.1 .. Yes No n.a. Chile 16.8 757 9,940 2,150.8 55.7 22.7 .. .. 55.7 22.7 Yes No Yes 10.0 Colombia 44.5 1,142 5,510 695.3 23.9 ¾ 23.9 ¾ ¾ ¾ Yes No Yes 27.7 13.9 69.6 No Congo, Dem. Rep. 64.2 2,345 180 94.6 ¾ 25.0 Yes Yes n.a. Congo, Rep. 3.6 342 2,310 2,260.8 87.7 86.0 87.7 86.0 .. .. Yes Yes No 1.9 Ecuador 13.5 284 4,510 1,268.3 ¾ 49.0 49.0 .. .. Yes No Yes n.a. Equatorial Guinea 0.7 28 14,680 17,385.9 98.9 93.5 98.9 93.5 .. .. Yes Yes No n.a. Gabon 1.5 268 7,760 4,727.1 77.3 65.7 77.3 65.7 .. .. No Yes No n.a. Ghana 23.4 239 1,240 60.5 65.4 ¾ ¾ ¾ 65.4 Yes Compl. No n.a. Guinea 9.8 246 380 29.3 89.9 24.5 89.9 24.5 .. .. No Yes No n.a. Indonesia 228.3 1,905 2,580 364.0 22.7 29.3 22.7 29.3 .. .. Yes No No 9.1 Iran, Islamic Rep. 72.0 1,745 4,530 2,351.2 75.8 69.2 75.8 69.2 .. .. Yes No Yes 2.7 Iraq 31.2 438 2,320 2,468.2 97.5 81.0 97.5 81.0 .. .. Yes No No n.a. Kazakhstan 15.7 2,725 7,440 3,870.6 54.3 44.6 54.3 44.6 .. .. Yes Yes No 31.3 Kuwait 2.7 18 43,920 35,226.7 94.4 72.1 94.4 72.1 .. .. Yes No Yes n.a. Kyrgyz Republic 5.3 200 880 14.4 59.0 ¾ .. .. 59.0 .. No Yes No 25.5 Lao PDR 6.2 237 1,000 ¾ 50.4 ¾ .. .. 50.4 .. No No No n.a. Libya 6.3 1,760 12,020 9,232.4 97.5 89.7 97.5 89.7 .. .. No No No n.a. Malaysia 27.0 330 7,900 1,593.1 47.0 13.0 47.0 13.0 .. .. Yes No No 11.2 Mauritania 3.2 1,031 1,060 451.0 34.7 11.0 11.0 34.7 .. No Yes No n.a. Mexico 106.4 1,964 9,330 1,068.3 15.8 35.6 15.8 35.6 .. .. Yes Yes 31.8 Mongolia 2.6 1,567 1,890 178.0 58.6 28.9 .. .. 58.6 28.9 No Compl. No 31.7 135.9 58.8 No Namibia 2.1 824 4,650 58.8 6.3 .. .. 6.3 No No n.a. 91.5 No Niger 14.7 1,267 360 ¾ 91.5 42.0 .. .. 42.0 Yes Yes n.a. 492.0 Yes Nigeria 151.3 924 1,180 97.5 83.7 97.5 83.7 ¾ ¾ Yes Yes n.a. Norway 4.8 324 85,380 19,903.9 49.8 39.1 49.8 39.1 .. .. Yes Yes Yes 28.3 Country Population Area GDP per Capita EI Rents per EI Share of Total EI Revenue Oil and Gas Oil and Gas Mining Share of Mining Revenue National Oil/ Extractive Stabilization Subnational (millions, 2010) (1,000 sq km) Capita Exports (%, Share of Total Share of Total Revenue Share Total Exports Share of Total Mining Industry Fund Expenditures as (US$, 2010) Public Revenues Exports (%, (%, 2007-2009) Company % of Total (%, 2007-2009) 2007-2009) Public Expenditures Oman 2.8 310 17,890 11,121.9 81.0 86.3 81.0 86.3 .. .. No No Yes n.a. 431.0 80.6 Yes Papua New Guinea 6.5 463 1,300 80.6 34.3 .. .. 34.3 No No n.a. Peru 28.8 1,285 4,710 532.4 32.9 25.0 .. .. 32.9 25.0 Yes Yes No 20.4 Qatar 1.3 11 43,484.8 88.9 64.5 88.9 64.5 .. .. Yes No No n.a. Russian Federation 141.8 17,098 9,910 3,825.9 65.8 32.8 65.8 32.8 ¾ ¾ Yes Yes Yes 40.7 Sao Tome e Principe 1,200 ¾ ¾ ¾ ¾ ¾ .. .. No No No n.a. Saudi Arabia 24.7 2,150 17,200 13,445.4 89.7 89.3 89.7 89.3 .. .. Yes No No n.a. Sierra Leone 5.6 72 340 3.0 90.3 ¾ .. .. 90.3 No Yes No n.a. South Africa 48.7 1,219 6,100 679.4 6.0 2.0 .. .. 6.0 2.0 Yes No No 37.9 371.3 .. No Sudan 41.4 2,506 1,270 90.2 55.7 90.2 55.7 .. Yes No n.a. Syrian Arab Republic 21.2 185 2,640 659.7 39.7 29.6 39.7 29.6 .. .. Yes No No n.a. .. Yes Timor Leste 1.1 15 2,220 ¾ 97.3 98.2 97.3 98.2 .. No Yes n.a. Trinidad and Tobago 1.3 5 15,380 11,484.1 87.0 57.8 87.0 57.8 .. .. Yes No No 4.6 5,555.5 .. No Turkmenistan 5.0 488 3,700 77.3 ¾ 77.3 ¾ .. No No n.a. United Arab Emirates 4.5 84 21,716.8 43.3 70.2 43.3 70.2 .. .. Yes No Yes n.a. Uzbekistan 27.3 447 1,280 824.1 36.9 17.2 22.7 ¾ 14.3 17.2 Yes No No n.a. Venezuela, RB 27.9 912 11,590 3,470.8 81.2 46.3 81.2 46.3 .. .. Yes No Yes n.a. Vietnam 86.2 329 1,100 187.4 17.5 31.0 17.5 31.0 .. .. Yes No No n.a. 375.0 .. No Yemen, Rep. 23.1 528 1,060 98.9 68.1 98.9 68.1 .. No Yes n.a. Zambia 12.6 753 1,070 221.8 66.7 ¾ .. .. 66.7 ¾ No Yes No n.a. Total 1,473.0 Average 7,952.9 5,392.0 69.4 50.6 70.1 58.0 54.0 23.6 19.9 Source: Barma et al. 2012, World Bank Decentralization Indicators 2012. Note: Population, area, and GDP per capita from WDI and Government Finance Statistics database. Subnational governments’ share of total government expenditures; that is, the percentage of total ex penditures accounted for by subnational governments, measured as the sum of local and provincial total expenditures excluding current and capital transfers to other levels of government, divided by the sum of local, provincial, and national expenditures, excluding intergovernmental transfers to lower tiers. Table 4. Intergovernmental Transfer Design in Resource-Rich Settings Country Primary Goal Main Transfer Schemes Allocation Criteria/Formula AFRICA (AFR) Congo, Dem. Rep. Equalization Derivation-based revenue sharing and 40 percent of revenue collected is to be transferred to the provinces and an of equalization fund. Broad outlines set by additional 10 percent is to finance an equalization fund for all the 2006 Constitution, and implementation, provinces.a Distribution of mineral royalties: 40 percent of mining pending further regulations, set out in royalties, as well as 10 percent of surface rents should be conceded to the annual budget laws. provinces, including 25 percent for the provincial administration and 15 percent for the area where mining activities were conducted. Nigeria Equalization of service Extensive revenue sharing. All oil revenues are received into the Federation Account. The revenue- delivery capacity; General and derivation-based sharing to sharing formulas are set by an act of the National Assembly. strong focus on rules-based state and local governments. The central government distributes the pool according to a formula that block transfers to states and Automatic transfer combined with takes into account population and state size. local government authorities stabilization mechanism (oil-price-based Nigeria shares 45 percent of federal revenues with states and (LGAs) fiscal rule by transferring oil revenues to the municipalities. budget in accordance with a reference price, For natural resource revenues, the constitution provides that oil-producing together with a ceiling on the nonoil states receive 13 percent up front of the net oil revenue as derivation grants; deficit). the federal government receives 52.7 percent, states 26.7 percent, and local Production sharing. governments 20.6 percent.b Direct mechanisms of partial revenue New revenue-sharing rule adopted in 2007.c redistribution to nonproducing regions. Sudan Stabilization objective: Derivation-based revenue sharing with The wealth-sharing protocol requires that a significant portion of the (prior to 2011) excess oil revenues are South and producing provinces based on additional oil revenue is automatically transferred to the states if not saved transferred into a fund for realizations. Main beneficiary South Sudan. in the ORSA. At least 2 percent of oil revenue shall be allocated to the oil- future generations Production sharing. producing states and regions in proportion to output produced in such states Excess oil revenues are saved in the Oil and regions. Revenue Stabilization Account (ORSA) After the payment to the ORSA and to the oil-producing states and regions, whenever the production reaches more than 50percent of net oil revenue derived from oil-producing wells in Southern 2 million barrels a day. Sudan shall be allocated to the Government of Southern Sudan (GOSS). d A Future Generation Fund shall be established once national oil production reaches 2 million barrels per day. Eastern Europe/Central Asia (ECA) Russian Federation Equalization of fiscal Provides both shared revenues and transfers The government provides transfers based on a nonderivation formula basis, capacity and expenditure to states. which benefits the poorer and non-natural-resource-endowed regions. Tax revenue sharing. Intergovernmental Transfer Formula: Natural resource taxes are collected mainly (1) 40 percent to federation by subnational governments as an important (2) 30 percent to producing oblasts source of own-source revenue in oil- (3) 30 percent to rayon and cities producing regions. East Asia/Pacific (EAP) Indonesiae Achieve the equal provision Some derivation-based revenue sharing, Producing provinces receive fixed proportions of revenues: of minimum standards in including from oil and gas. Aceh: 55 percent of oil revenues and 40 percent of gas revenues services; equalizing block Straight automatic transfer with no Papua: 70 percent of oil revenues and 70 percent of gas revenues grants for local governments stabilization mechanism. In addition to own source, districts and provinces enjoy 15 percent of Production sharing. onshore oil revenue and 30 percent of onshore gas revenue transferred to Additional earmarked grants may be the originating provinces as well as an equalization block grant (DAU). f distributed for specific purposes based on a The equalization grant pool is a fixed percentage (25 percent) of central formula that takes into account both resources after tax sharing with the regions: the central government revenue capacity and expenditure needs. distributes at least 25 percent of central government revenue to be In 2009, an additional 5 percent of oil and transferred to all lower level governments through a general transfer gas revenues was be transferred to regional mechanism. governments earmarked for spending on Distribution of mineral royalties: 80 percent of mineral royalties go to the primary education. regions (64 percent to regencies and 16 percent to provincial government) Direct mechanisms of partial revenue and 20 percent to central government. redistribution to nonproducing regions. Malaysia Fiscal equalization Annual equalization grants to equalize the Revenues from all oil and gas industries are “pooled� together. objectives difference between fiscal capacities and Derivation is used as the primary basis for revenue allocation. Most grants fiscal needs of the local authority. are based on objective criteria except for deficit grants that are granted Production sharing (some direct sharing of exceptionally. oil revenues to states). Formula: population (50 percent), GDP per capita (50 percent). The government collects oil royalties of which 5 percent are passed to the states of Sabah, Sarawak, and Terengganu and the rest are retained by the federal government. Papua New Guinea By law,g revenues distributed at the No formula-based redistribution according to specific characteristics. subnational level cannot be greater than 20 Depends on the provincial government's share of the project. Derivation to percent of the net total revenues. the province owning equity in the mining or oil project: the 2 percent Dividends from equity in mining and oil mining royalty is shared between provincial government (80 percent) and projects.h landowners (20 percent). Development levy (2 percent of value) Different shares of royalties are transferred to provincial governments. shared among local and provincial These are set out in the Memorandum of Agreement for the project. governments. Derivation to the province in which the project is located. Mining and oil royalties (2 percent of Special support grant (1 percent royalty) is paid on a derivation basis to the value). provincial government (20 percent goes to the producing local No redistribution mechanism to benefit government). nonproducing territories. Revenues shared at the local level are earmarked. Middle East/North Africa (MENA) Iraq No explicit natural resource sharing. Prevailing de facto arrangements: comanagement of hydrocarbon exploitation in endowed regions, and fair distribution based on population and historical damages.i United Arab Stabilization tool Oil revenues are collected by a subnational Upward sharing of revenues to federal account.j Emirates government and redistributed according to a Cash contributions are negotiated each year between the federal specific rule or formula. government and each of the seven emirates. Emirates have full ownership and control over their oil resources. Latin America (LAC) Bolivia Fiscal equalization High-level of derivation-based revenue A Departmental Compensation Fund was created to mitigate the imbalance objectives sharing. in the oil rents distribution: departments with per capita royalties below the Partial revenue redistribution from central national average will receive a compensatory transfer by the TGN. pool (allocations from equalization funds) No formula-based redistribution taking into account sociodemographic to nonproducing local jurisdictions (20 characteristics. percent of the revenues). Departmental royalties are directly transferred by YPFB to oil-producing Specific automatic transfer. departments (11 percent) and two departments as compensation (1 Tax revenue sharing, no production sharing. percent).k Revenues shared at the local level are earmarked. Brazil Fiscal equalization Revenue derived from signature bonus The states receive 21.5 percent of these tax revenues: 85 percent going to objectives accrues to the Petroleum Regulatory the north, northeast, and center-west regions, and the remaining 15 percent Agency (ANP) to fund its activities. to the south and southeast regions (which are economically better off). Equalization transfers represent a large The formula for determining state shares is complex and based on allocation of resources.l population size and per capita income. The federal government transfers a Royalties are shared according to formulas that varies if it is onshore m (52.5 specified share of certain federal taxes to a percent to producing states) or offshoren production. pool, and the council of states determines 0.5 to 1 percent of the value of total production from all land-based fields state shares using a formula. as special royalties should be paid to the landowners in Brazilian currency Offshore oil revenues are also shared. on a monthly basis.o Gross revenue on productionp is distributed according to the following formula: 40 percent to producing states, 10 percent to producing municipalities, 50 percent to federal ministries. Colombiaq Fiscal equalization Tax revenue sharing. Royalties are distributed to indigenous communities. objectives Allocation of royalty shares. Transfers from the FAEP are shared among all subnational governments Transfers from the Oil Savings and according to a formula taking into account population and tax revenue Stabilization Fund (FAEP). performance. Allocations from equalization funds to Royalties are transferred to oil-producing departments (47.5percent), nonproducing local jurisdictions. municipalities (12.5 percent), ports (8 percent), and National Royalties No production sharing. Fund (32 percent). Ecuador Stabilization tool FEIREPr was created to reduce public debt Minimal revenue sharing with regions. and to stabilize revenues (FAC), it was Specific rents are distributed to indigenous communities. replaced by CEREPS.s 80 percent of decentralized revenues go to ECORAEu that distributes the In April 2008, the CEREPS was eliminated resources to provincial municipalities in the Amazon region and then and the net revenuest transferred to the delivers the remainder to a regional fund. government to finance public investment. Mexico Stabilization tool Limited revenue sharing. Formula-based redistribution:v Minimal derivation-based revenue. 20 percent of ordinary extraction rights (incorporated into a general fund No production sharing. and distributed to the states based on a formula taking into account their Resources are channeled through a general characteristics. pool of shared resources and distributed 3.17 percent of the additional oil extraction rights (earmarked for according to a formula. municipalities in oil-producing regions to compensate for environmental Direct mechanisms of partial revenue damage). redistribution to nonproducing regions. Peru Derivation-based sharing Derivation-based revenue sharing (Canon 20 percent of mining activities taxes distributed to regional and 30 percent mechanisms Minero Mechanism). to local governments. Mining royalty. 5 out of 24 regions receive a Canony. Camisea Development Fund (FOCAM).w Mining royalties are distributed to producing areas and are used to finance Regional Compensation Fund (FONCOR).x productive investment that promotes sustainable economic development. 5 percent earmarked to research at universities. Formula-based redistribution of Canons to producing areas, used to finance investment projects with regional impacts: 20 percent are earmarked to universities and research and 20 percent to infrastructure maintenance. In 2001, the proportion of taxes to be redistributed to subnational government was increased from 20 percent to 50 percent. New formula for equalizing transfers taking into account the gap between regional unmet needs and other transfers that the region may receive (including Canons and royalties). Distribution from the FOCAM to areas crossed by the gas pipes and takes into account the pipe length, population, and unmet basic needs. Venezuela, Rep. Stabilization tool Revenue-sharing agreements. 20 to 25 percent of total central government revenues are distributed to Bol. de Transfers. regions. No production sharing. In addition, subnational governments are entitled to 20 to 30 percent of oil royalties. Source: Kaiser and Mehri 2009. Note: a. The 2006 Constitution provides for significant decentralization to provincial and local authorities and the Mining Code provides for the distribution of mineral royalties. b. Article 162 (2) of the Constitution of the Federal Republic of Nigeria, May 1999. c. The new sharing rule was adopted in 2007 by the Federation Accounts Allocation Committee (FAAC) to the effect that in any month when total revenue available for sharing fell below the budgetary target, the shortfall was taken out from the ECA and is considered as revenue sharing and not drawdowns. d. Constitution of Southern Sudan signed December 2005, Interim National Constitution ratified July 5, 2005. e. Until 2000, all oil and gas revenues accrued to the central government and were not subject to any revenue-sharing arrangement with provincial and district-level governments. f. Revenue-sharing arrangements are made according to the Indonesian Law on Fiscal Balance (Law 25/99). g. Provision in the Oil and Gas Act 1998, which regulates the exploitation of natural resources and the allocation of revenues, forbids that the aggregate revenues distributed for subnational levels be greater than 20 percent, so by law the central government receives 80 percent of oil and gas revenues. h. Transfers made by the Minerals Resource Development Company, and by Ok Tedi Sustainable Development Program in the case of Western Province. i. Article 109 of the Constitution of the Republic of Iraq, approved by Referendum on October 15, 2005. All oil revenues accrue to the New York –based foreign currency accounts of the Development Fund for Iraq (DFI) (under the 2006 arrangements). j. The United Arab Emirates Constitution made permanent in 1996. k. The four producing departments (Santa Cruz, Cochabamba, Tarija, and Chuquisaca) receive 11 percent, whereas Beni and Pando receive 1 percent of gross hydrocarbon production as compensation. l. Petroleum Law No. 9478 of 1997. m. Onshore royalty sharing formula (royalties in excess of 5 percent of production): 52.5 percent to the state where the production takes place, 15 percent to the municipalities where the production occurs, 7.5 percent to the municipalities affected by operations of landing or shipment of oil or natural gas, and 25 percent to the Ministry of Science and Technology. n. Offshore royalty-sharing formula (royalties in excess of 5 percent of production): 22.5 percent to the states fronting the production area, 22.5 percent to the municipalities fronting the production area, 15 percent to the Ministry of Defense responsible for the Navy, 7.5 percent to the municipalities affected by operations of landing or shipment of oil and gas, 7.5 percent to a special fund to be distributed to all states and municipalities in Brazil, and 25 percent to the Ministry of Science. o. Article 52 of the Brazilian Petroleum Law specifically provides that the concessionaire must set aside this special royalty. p. Minus investments, costs, taxes, and royalties. q. Until 2001, the constitution specified compulsory intergovernmental transfers to subnational governments and these reached 47 percent of the central government’s revenues in 2002. A constitutional amendment in 2001 delinked subnational transfers from current revenues of the central government after seven subnational governments had to restructure their debt. r. FEIREP = Fund for Stabilization, Social and Productive Investments, and Public Debt Reduction. s. CEREPS’ allocation of revenues: 35 percent for repurchase of public debt and productive projects, 20 percent (FAC) for contingencies and stabilization of revenues, 30 percent for education and health, 5 percent for roads repairs, 5 percent for environmental repair, and 5 percent for research and development. t. CEREPS net revenues reached US$613.2 as of December 31, 2006. No information is available for 2007. u. ECORAE: The Institute for the Ecological Development of the Amazon Region is in charge of planning the sustainable development of the Amazon region in Ecuador. v. Established by the law on fiscal coordination in which subnational governments have considerable expenditure responsibilities, and most of their revenues are in the form of conditional transfers and revenue-sharing agreements totaling about 7 percent of GDP. w. 25 percent of the gas royalties goes to the FOCAM and are distributed to areas crossed by the gas pipes. x. The allocation of the Regional Compensation Fund (FONCOR) accounted for 31 percent of transfers to regional governments in 2006 and changed in 2008 to make it more equalizing. y. Canon: royalty share according to a percentage of the gross value of oil production in the region. It is on average equivalent to 44 percent of the royalty. Table 5: Saving and Stabilization Institutions in Resource-Rich Settings, 2012 Country Fund Name Assets Inception Origin Linaburg- US$ Maduell billion Transparency Index Algeria Revenue Regulation Fund 56.7 2000 Oil 1 Australia Australian Future Fund 80 2006 Non-Commodity 10 Azerbaijan State Oil Fund 30.2 1999 Oil 10 Bahrain Mumtalakat Holding Company 9.1 2006 Non-Commodity 9 Botswana Pula Fund 6.9 1994 Diamonds and 6 Minerals Brazil Sovereign Fund of Brazil 11.3 2008 Non-Commodity 9 Brunei Brunei Investment Agency 30 1983 Oil 1 Canada Alberta’s Heritage Fund 15.1 1976 Oil 9 Chile Social and Economic Stabilization 15 2007 Copper 10 Fund Chile Pension Reserve Fund 4.4 2006 Copper 10 East Timor Timor-Leste Petroleum Fund 9.9 2005 Oil and Gas 6 Equatorial Guinea Fund for Future Generations 0.08 2002 Oil n.a. Gabon Gabon Sovereign Wealth Fund 0.4 1998 Oil n.a. Indonesia Government Investment Unit 0.3 2006 Non-Commodity n.a. Iran Oil Stabilisation Fund 23 1999 Oil 1 Ireland National Pensions Reserve Fund 30 2001 Non-Commodity 10 Kazakhstan Kazakhstan National Fund 58.2 2000 Oil 8 Kiribati Revenue Equalization Reserve Fund 0.4 1956 Phosphates 1 Kuwait Kuwait Investment Authority 296 1953 Oil 6 Libya Libyan Investment Authority 65 2006 Oil 1 Malaysia Khazanah Nasional 36.8 1993 Non-Commodity 5 Mauritania National Fund for Hydrocarbon 0.3 2006 Oil and Gas 1 Reserves Mexico Oil Revenues Stabilization Fund of 6.0 2000 Oil n.a. Mexico Mongolia Fiscal Stability Fund n.a. 2011 Mining n.a. Nigeria Nigerian Sovereign Investment 1 2011 Oil n.a. Authority Norway Government Pension Fund – Global 611 1990 Oil 10 Oman State General Reserve Fund 8.2 1980 Oil and Gas 1 Oman Oman Investment Fund n.a. 2006 Oil n.a. Papua New Guinea Papua New Guinea Sovereign n.a. 2011 Gas n.a. Wealth Fund Qatar Qatar Investment Authority 100 2005 Oil 5 Russia National Welfare Fund 149.7 2008 Oil 5 Saudi Arabia SAMA Foreign Holdings 532.8 n.a. Oil 4 Saudi Arabia Public Investment Fund 5.3 2008 Oil 4 Trinidad & Tobago Heritage and Stabilization Fund 2.9 2000 Oil 8 UAE – Abu Dhabi Abu Dhabi Investment Authority 627 1976 Oil 5 UAE – Abu Dhabi International Petroleum Investment 65.3 1984 Oil 9 Company UAE – Abu Dhabi Mubadala Development Company 48.2 2002 Oil 10 UAE – Abu Dhabi Abu Dhabi Investment Council n.a. 2007 Oil n.a. UAE – Dubai Investment Corporation of Dubai 70 2006 Oil 4 UAE – Federal Emirates Investment Authority n.a. 2007 Oil 2 UAE – Ras Al RAK Investment Authority 1.2 2005 Oil 3 Khaimah US – Alabama Alabama Trust Fund 2.5 1985 Oil and Gas n.a. US – Alaska Alaska Permanent Fund 40.3 1976 Oil 10 US – North Dakota North Dakota Legacy Fund 0.1 2011 Oil and Gas n.a. US – Texas Texas Permanent School Fund 24.4 1854 Oil and Other n.a. US – Wyoming Permanent Wyoming Mineral Trust 4.7 1974 Minerals 9 Fund Venezuela FEM 0.8 1998 Oil 1 Source: The Sovereign Wealth Fund Institute, www.swfinstitute.org/fund-rankings/. Note: GDP = gross domestic product; GOSS = Government of Southern Sudan. References Aberbach, Joel D., Robert D. Putnam, and Bert A. Rockman. 1981. Bureaucrats and Politicians in Western Democracies. Cambridge, MA: Harvard University Press. Ahmad, Ehtisham, and Giorgio Brosio. 2008. "Political Economy of Multi-level Tax Assignments in Latin American Countries: Earmarked Revenue versus Tax Autonomy." International Monetary Fund, Fiscal Affairs Department, Washington, DC. Ahmad, Ehtisham, and Eric Mottu. 2002. "Oil Revenue Assignments: Country Experiences and Issues." IMF Working Paper, 1- 27, International Monetary Fund, Washington, DC. Bahl, Roy, and Johannes Linn. 1994. "Fiscal Decentralization and Intergovernmental Transfers in Less Developed Countries." Publius 24: 1–19. Bahl, Roy, and Bayar Tumennasan. 2004. How Should Revenues from Natural Resources be Shared in Indonesia? Reforming Intergovernmental Fiscal Relations and the Rebuilding of Indonesia: The 'Big Bang' Program and Its Economic Consequences. Northampton, MA: Edward Elgar. Bakke, Kristin M., and Erik Wibbels. 2006. "Diversity, Disparity, and Civil Conflict in Federal States." World Politics 1. Beramendi, Pablo. 2007. "Federalism." In The Oxford Handbook of Comparative Politics, ed. Carles Boix and Susan Carol Stokes. New York: Oxford University Press. Bird, Richard, and Roy Bahl. 2008a. "Subnational Taxes in Developing Countries: The Way Forward." IIB Paper No. 16, University of Toronto, Joseph L. Rotman School of Management, Toronto. Bird, Richard, and Roy Bahl. 2008b. "Tax Policy in Developing Countries: Looking Back and Forward." IIB Paper No. 13, University of Toronto, Joseph L. Rotman School of Management, Toronto. Bird, Richard, and Robert Ebel. 2006. "Fiscal Federalism and National Unity." In Handbook of Fiscal Federalism, ed. Ehtisham Ahmad and Giorgio Brosio. Northampton, MA: Edward Elgar Publishing, Inc. Bird, Richard, and Robert Ebel. 2007. Fiscal Fragmentation in Decentralized Countries: Subsidiarity, Solidarity and Asymmetry. Cheltenham: Edward Elgar Publishing, Inc. Bird, Richard M., Jorge Martinez-Vazquez, and James Alm. 2003. Public Finance in Developing and Transitional Countries: Essays in Honor of Richard Bird. Cheltenham, UK and Northampton, MA: Edward Elgar Publishing, Inc. Bird, Richard M., 2001. Public Finance in Developing and Transition Countries: A Conference in Honor of Richard Bird, April 5–6, 2001, Andrew Young School of Policy Studies, Georgia University, Evergree Hotel and Resort, Stone Mountain, Georgia. Georgia State University, Andrew Young School of Policy Studies, Atlanta, Georgia. Blanchard, Olivier, and Andrei Shleifer. 2001. "Federalism With and Without Centralization: China Versus Russia." IMF Staff Working Paper, Vol. 48, Special Issue, International Monetary Fund, Washington, DC. Boadway, Robin, and Anwar Shah. 2009. Fiscal Federalism: Principles and Practice of Multiorder Governance. New York: Cambridge University Press. Boadway, Robin, and Anwar Shah. 1994. "Fiscal Federalism in Developing/Transition Economies: Some Lessons from Industrialized Countries." National Tax Association, Tax Institute of America, Columbus, Ohio. Boadway, Robin W., and Anwar Shah. 2007. "Intergovernmental Fiscal Transfers: Principles and Practices." World Bank, Washington, DC. Boix, Carles. 2003. Democracy and Redistribution. Cambridge, UK and New York: Cambridge University Press. Bornhorst, Fabian, Sanjeev Gupta, and John Thornton. 2009. "Natural Resource Endowments and the Domestic Revenue Effort European Journal of Political Economy 25. Brancati, Dawn 2009. Peace by Design Managing Intrastate Conflict through Decentralization. New York: Oxford University Press. Brosio, Giorgio. 2006. "The Assignment of Revenue from Natural Resources." In Handbook of Fiscal Federalism, ed. Ehtisham Ahmad and Giorgio Brosio. Northampton, MA: Edward Elgar Publishing, Inc. Caselli, Francesco, and Nicola Gennaioli. 2007. "Economics and Politics of Alternative Institutional Reforms." NBER Working Paper Series, Working Paper No. 12833, National Bureau of Economic Research, Cambridge, MA. Caselli, Francesco, and Guy Michaels. 2009. "Do Oil Windfalls Improve Living Standards? Evidence from Brazil." NBER Working Paper Series, Working Paper No. 15550, National Bureau of Economic Research, Cambridge, MA. Collier, Paul, and Anthony J. Venables. 2009. "Natural Resources and State Fragility." OxCarre Research Paper, Oxford Centre for the Analysis of Resource Rich Economies, Oxford. Davis, Jeffrey M., Rolando Ossowski, and Annalisa Fedelino. 2003. "Fiscal Policy Formulation and Implementation in Oil- Producing Countries." International Monetary Fund, Washington, DC. De Figueiredo, Rui, and Barry Weingast. 2005. "Self-Enforcing Federalism." Journal of Law, Economics, and Organization. de Mello, Luiz, and Kiichiro Fukasaku. 1999. Decentralisation, Inter-governmental Fiscal Relations and Macroeconomic Governance. Paris: OECD. De Silva, Migara, 2009. Intergovernmental Reforms in the Russian Federation: One Step Forward, Two Steps Back? Washington, DC: World Bank. Eaton, Kent. 2004. Politics beyond the Capital: The Design of Subnational Institutions in South America. Stanford, CA: Stanford University Press. Eaton, Kent, Kai Kaiser, and Paul Smoke. 2010. The Political-Economy of Decentralization Reforms: Implications for Aid Effectiveness. Washington, DC. Filippov, Mikhail, Peter C. Ordeshook, and Olga Shvetsova. 2004. Designing Federalism: A Theory of Self-Sustainable Federal Institutions. New York: Cambridge University Press. Gonzalez, Christian, David Rosenblatt, and Steven Benjamin Webb. 2002. "Stabilizing Intergovernmental Transfers in Latin America: A Complement to National/Subnational Fiscal Rules?" World Bank Policy Research Working Paper, World Bank, Washington, DC. Haysom, Nicholas, and Sean Kane. 2010. "Negotiating Natural Resources for Peace: Ownership, Control, and Wealth-Sharing." Paper prepared for Henry Dunant Center for Humanitarian Dialogue, HD Briefing Paper, October, Geneva. Herbst, Jeffrey. 1989. "Creation and Maintenance of National Boundaries in Africa." International Organization 43: 673–92. Herbst, Jeffrey. 2001. "The Politics of Revenue Sharing in Resource-Dependent States." Mimeo. Hofman, Bert, Kadjatmiko, Kai Kaiser, and Bambang Suharnoko Sjahrir. 2006. "Evaluating Fiscal Equalization in Indonesia." Policy Research Working Paper No. 3911, World Bank, Washington, DC. Horowitz, David. 2002. Uncivil Wars: The Controversy over Reparations for Slavery. San Francisco: Encounter Books. International Monetary Fund. 2009. "Macro Policy Lessons for a Sound Design of Fiscal Decentralization." Fiscal Affairs Department, IMF, Washington, DC. Jeong, Ho-Won. 2002. Approaches to Peacebuilding. New York: Palgrave Macmillan. Kaiser, Kai, and Mira Merhi. 2009. "Revenue and Expenditure Management in Resource Rich Countries: Managing Post- Commodity Boom Volatility in Intergovernmental Settings." Policy Note, Washington, DC. Lane, Philip R., and Aaron Tornell. 1999. "The Voracity Effect." American Economic Review 89: 22-46. Lewis, Blane. 2005. "Indonesia Local Government Spending, Taxing and Saving: An Explanation of Pre- and Post- Decentralization Fiscal Outcomes." Asian Economic Journal 19: 291 –317. Ma, Jun, 1997. "Intergovernmental Fiscal Transfers in Nine Countries: Lessons for Developing Countries." Policy Research Working Paper. Mclure, Charles E. 1983. "Tax Assignment in Federal Countries." Centre for Research on Federal Financial Relations, Canberra. Mclure, Charles E. 1994. "Taxation of Natural Resources and the Future of the Russian Federation." National Tax Association, Tax Institute of America, Columbus, Ohio. Mclure, Charles E. 2003. "The Assignment of Oil Tax Revenue." In Fiscal Policy Formulation and Implementation in Oil- Producing Countries, ed. Jeffrey David, Rolando Ossowski, and Analissa Fedelino. Washington, DC: International Monetary Fund. Mieszkowski, Peter. 1983. "Energy Policy, Taxation of Natural Resources, and Fiscal Federalism." In Tax Assignment in Federal Countries, ed. C. E. McLure, Jr. Canberra: Centre for Research on Federal Financial Relations, Australian National University. Oates, Wallace E. 1972. Fiscal Federalism. New York: Harcourt Brace Jovanovich. Oates, Wallace E. 2005. "Toward a Second-Generation Theory of Fiscal Federalism." International Tax and Public Finance, 349–74. Available at http://www.econ.umd.edu/research/papers/363 (accessed February 2010). Otto, James. 2001. "Fiscal Decentralization and Mining Taxation." The World Bank Group Mining Department, Washington, DC. Panizza, Ugo. 1999. "On the Determinants of Fiscal Centralization: Theory and Evidence." Journal of Public Economics 74: 97– 139. Quian, Yingyi, and Barry Weingast. 1997. "Federalism as a Commitment to Preserving Market Incentives." Journal of Economic Perspectives 11: 83–92. Rodden, Johnathan. 2007. "Federalism." In The Oxford Handbook of Political Economy, ed. Barry R. Weingast and Donald A. Wittman. Oxford: Oxford University Press. Rodden, Jonathan, Gunnar Eskeland, and Jennie Litvack. 2003. Decentralization and Hard Budget Constraints. Boston, MA: MIT Press. Rodden, Jonathan, and Erik Wibbels. 2002. "Beyond the Fiction of Federalism: Macroeconomic Management in Multitiered Systems." World Politics 54 (4): 494–531. Schneider, Aaron. 2006. Governance Reform and Institutional Change in Brazil: Fiscal Responsibility and Tax. Brighton, UK: Institute of Development Studies. Shah, Anwar. 2004. "Fiscal Decentralization in Developing and Transition Economies: Progress, Problems, and the Promise." World Bank Institute, Poverty Reduction and Economic Management Division, World Bank, Washington, DC. Srinivasan, T. N., and Jessica Seddon Wallack. 2006. Federalism and Economic Reform: International Perspectives. New York: Cambridge University Press. Treisman, Daniel. 1999. After the Deluge: Regional Crises and Political Consolidation in Russia. Ann Arbor: University of Michigan Press. Treisman, Daniel. 2007. The Architecture of Government: Rethinking Political Decentralization. Cambridge and New York: Cambridge University Press. Weingast, Barry. 1995. "The Economic Role of Political Institutions: Market-Preserving Federalism and Economic Development." Journal of Law, Economics, & Organization 11: 1–31.