POVERTY THE WORLD BANK REDUCTION AND ECONOMIC MANAGEMENT NETWORK (PREM) Economic Premise MAY 2012 • Number 83 JUN 010 • Numbe 18 69260 Natural Capital and the Resource Curse Otaviano Canuto and Matheus Cavallari An abundance of natural resources is intuitively expected to be a blessing. Nonetheless, it has been argued for some decades that large endowments of natural resources—oil, gas, and minerals in particular—may actually become more of a curse, often leading to slow economic growth and redistributive struggles (including armed conflict). Over the years, vast empirical literature has addressed this “paradox.� The literature has had to rely on proxies for natural resource abundance because of the lack of appropriate data, generating doubt on whether results would be similar if direct measures of natural wealth were available. This gap is now starting to be filled with the data series released by the World Bank (1997, 2006, 2011) on natu- ral capital and other forms of countries’ wealth. This note presents an analysis of these data to revisit some of the conclusions reached in the literature on the relationship between natural resource abundance and economic growth. The findings are in alignment with the view that there is no clear deterministic evidence of natural resource abundance as a curse or a blessing; therefore, the effect on a country depends on other determinants. Natural Capital and Income Levels ests, minerals, and energy); and intangible capital. The latter is clearly a wealth component requiring much further work. In- How to measure development progress? “While precise defini- tangible capital is still measured as a residual, the difference tions may vary, development is, at heart, a process of building between estimates of total wealth and the sum of natural and wealth—the produced, natural, human, and institutional capi- produced capitals:1 “It implicitly includes measures of human, tal which is the source of income and wellbeing� (Andersen social, and institutional capital, which includes factors such as and Canuto 2011, xi). However one defines development, it the rule of law and governance that contribute to an efficient supposes rising income levels and an underlying process of economy� (World Bank 2011, 4–5).2 building and managing a portfolio of assets. At least in an ac- Table 1 reproduces the aggregate figures of wealth and per counting sense, such wealth accumulation spans the wide capita wealth by type of capital and income group in 1995 and range of capital types mentioned above. 2005 (World Bank 2011, 7). From an accountant’s perspective, The World Bank (1997, 2006, 2011) has started a break- there are some striking features regarding an archetype of prog- through in national accounts toward capturing the span of as- ress up the income-wealth ladder. First, the share of produced sets. A set of wealth accounts covering a 10-year period, 1995 capital in wealth moves upward from low levels in low-income to 2005, for more than 120 countries is now available. This countries, but remains reasonably modest thereon. Figures for includes produced capital (machinery, structures, and equip- lower-middle-income countries are heavily influenced by the ment); natural capital (agricultural land, protected areas, for- weight of China and by its extraordinarily fast pace of produced 1 POVERTY REDUCTION AND ECONOMIC MANAGEMENT (PREM) NETWORK www.worldbank.org/economicpremise capital accumulation, with the ex-China subgroup displaying waiting to be exploited. In contrast, the aver- numbers closer to a smoother evolution. It appears that cumu- age kilometer of Africa had only $25,000 of lative savings used for investments in physical assets accompany known subsoil assets. Most likely, this re- and support the rise in income levels, but typically at a propor- flects a massive failure in the discovery pro- tionate speed. cess in Africa: the scope for resource extrac- Secondly, intangible wealth is the largest single component tion may be five times what it currently appears to be. of total wealth at all levels of income, but increasingly so as it Collier remarks that the discovery process depends upon moves to the upper-middle- and high-income levels. Increased “good governance,� a missing component of intangible wealth educational attainments, as well as improvements in institu- in many low-income countries. tions, governance, and other intangible forms of wealth are im- Values of natural capital may also change as a result of het- perative if a country is to overcome “middle-income traps� (Ca- erogeneity among natural resources. The value of existing as- nuto 2011). The nexus with savings and investments is not as sets may rise if increases of global production have to be based straightforward as in the case of produced capital, because the on less efficient sources at the margin. Such “rents� tend to be dynamics of intangible wealth accumulation—or depletion— reflected in the value of natural capital in countries well en- depend to a large extent on factors of another nature (quality of dowed with high-quality resources. education; institutional evolution; collective knowledge tacitly One may then guess that abundance of natural capital—as embedded in routines of firms, the public sector, organizations, measured by per capita natural wealth—is in principle favorable and other social groups; and non-research- and development- to raising per capita income levels. Furthermore, the average derived technical progress).3 archetype of wealth-cum-income progression depicted in table It follows that natural assets comprise a substantial por- 1 may take place with different shares of natural wealth in dif- tion of total wealth at low-income levels, decreasing in rele- ferent countries. This is illustrated by the different wealth com- vance—particularly compared to intangible wealth—if the econ- positions as of 2005 among high-income countries (the United omy succeeds in moving up the income ladder. However, no States, Japan, Norway, Canada, and Australia), as well as among matter how large or small that natural capital is, it combines middle- and low-income countries (table 2). with unskilled labor and existing intangible wealth to generate Differences in the quality of natural resources may also ex- income, and the subsequent corresponding savings and invest- plain why countries with different levels of technological ments create new produced capital and intangible wealth. achievement can have similar income levels. The World Bank As one might expect, levels and compositions of natural (2007) developed country indexes of technological achieve- capital vary widely between countries (World Bank 2011, ap- ments and showed their (nonlinear) relationship with income pendix C). In all cases, however, if the use of nonrenewable natural capital—extractive resources such as oil, gas, and minerals—does not lead to the ac- Table 1. Wealth and Per Capita Wealth by Type of Capital and Income Group, 1995 and 2005 cumulation of other forms of productive 1995 wealth, but instead is used to support con- sumption, there will be no income-generat- Total wealth Per capita Intangible Produced Natural (US$ wealth capital capital capital ing assets to replace it when it is exhausted. Income group billions) (US$) (%) (%) (%) As for the renewable part of natural capital, Low income 2,447 5,290 48 12 41 such as forest land, inappropriate manage- ment regimes and property rights may also Lower-middle income 33,950 11,330 45 21 34 lead to wealth depletion. 4 Upper-middle income 36,794 73,540 68 17 15 Natural capital also varies over time for High-income OECD 421,641 478,445 80 18 2 reasons other than its use. The fact that it di- World 504,548 103,311 76 18 6 minishes in relative terms as a wealth compo- 2005 nent when the economy moves up the in- Low income 3,597 6,138 57 13 30 come ladder does not preclude it to rise in absolute terms as a result of technological Lower-middle-income 58,023 16,903 51 24 25 changes or new discoveries. As Collier (2009, Upper-middle income 47,183 81,354 69 16 15 2) has highlighted, this partially explains High-income OECD 551,964 588,315 81 17 2 why paradoxically: World 673,593 120,475 77 18 5 As of 2000, the typical square Source: World Bank 2011, 7. kilometer of the OECD had Note: Figures are based on the set of countries for which wealth accounts are available from 1995 to 2005. Data in this table do not include high-income oil exporters. OECD = Organisation for Economic Co-operation $125,000 of known subsoil assets and Development. 2 POVERTY REDUCTION AND ECONOMIC MANAGEMENT (PREM) NETWORK www.worldbank.org/economicpremise Table 2. Shares of Natural Capital in Total Wealth Per Capita So, where can one locate a possible “natural resource (Selected Countries) curse�? Since country rankings by per capita income display Total wealth per Natural capital Share of natural several natural resource–rich countries at the top, there ap- capita per capita capital pears to be no inevitable impediment to income growth associ- Country (2005 US$) (2005 US$) (%) ated with abundance of natural resources. As one can see in Norway 861,797 110,162 12.8 figure 2, there is no clear pattern regarding gross domestic United States 734,195 13,822 1.9 product (GDP) per capita and shares of natural capital. One can find 14 countries with GDP per capita higher than Japan 548,751 2,094 0.4 US$35,000 and natural wealth ratios ranging from almost Canada 538,697 36,924 6.9 zero up to 79 percent. Australia 518,805 39,979 7.7 However, there is now enormous evidence, both system- Brazil 79,142 14,978 18.9 atic and anecdotal, of cases in which natural resource discover- Argentina 71,252 10,267 14.4 ies or appreciations, instead of being followed by a transforma- Malaysia 64,767 12,750 19.7 tion of natural capital into other forms of productive wealth through some virtuous process of savings and investment, were Botswana 58,895 5,420 9.20 accompanied by stagnation or even income regression, general- Jordan 51,454 2,690 5.2 ly in combination with political disruption.5 Nigeria 10,982 6,042 55.0 Brahmbhatt, Canuto, and Vostroknutova (2010) explain India 10,539 2,704 25.7 how certain conditions could result in situations where natural Malawi 3,471 1,170 33.7 resource booms become a curse. Weak governance and corre- Source: Authors’ calculation, from World Bank (2011, appendix C). sponding poor economic policies underlie the misallocation and mismanagement of resources. Resources shift out of pro- ductive activities into unproductive rent-seeking activity when, levels. On the other hand, many countries in Latin America for example, patronage networks are strengthened with their and the Caribbean managed to reach middle-income levels appropriation of fallen-from-heaven rents. It is not by chance with much less technological effort than others (figure 1). Pro- that resource curse cases can be primarily associated with ex- duced capital does not seem to explain such a discrepancy and, tractive industries (oil, gas, and minerals) because these are apart from non-technology-related intangible wealth, the high “concentrated ‘point source’ resources that can easily become quality of known natural resources in the region—large swaths the object of rent-seeking and redistributive struggles� (ibid, of arable land and mapped sources of oil and minerals—may 107), a point originally made by Collier and Goderis (2007). well be among the explaining factors. Consumption use of tax revenues derived from natural resource extraction through public spending is also a typical Figure 1. Technological Achievements Rise with Income Levels all countries .25 .30 developing countries only Europe and Central Asia .20 Europe and Central Asia .25 .20 .15 all countries .15 .10 index index .10 .05 Latin America and the .05 Caribbean Latin America and the Caribbean 0 0 -0.05 -0.05 -0.10 -0.10 12 0 00 00 00 00 00 0 0 0 0 0 00 00 00 00 00 00 00 00 00 00 00 00 ,0 ,0 ,0 ,0 ,0 ,0 2, 4, 6, 8, 10 14 16 18 20 0 ,0 ,0 ,0 ,0 ,0 ,0 ,0 5, 10 15 20 25 30 35 40 per capita income (PPPs) per capita income (PPPs) East Asia and Paci�c Europe and Central Asia high-income OECD countries high-income other countries Latin America and the Caribbean Middle East and North Africa South Asia Sub-Saharan Africa Source: World Bank 2007, 84. 3 POVERTY REDUCTION AND ECONOMICthe Caribbean (PREM) NETWORK her countries atin America and MANAGEMENT Middle Ea t a www.worldbank.org/economicpremise Figure 2. Natural Capital and GDP Per Capita however, one should not lose sight of the fact that the impact of a new discovery is forward looking, that is, it may cause benefits 12 or distortions in the economy only after resources are known. In this sense, it is not recommended to match wealth composi- 10 tion in a certain slice of time with past growth data. If subsoil LN (GDP PPP per capita) assets are unknown, no management decisions associated with 8 natural resource richness, either good or bad, can be made. 6 Lederman’s and Maloney’s (2007) empirical study is an- other oft-quoted source for links between natural resources and 4 economic growth. Most of their report is dedicated to showing high (>50%) the difficulties in making any safe statement about the negative 2 middle (2.2%