Are sub-Saharan Africa’s abundant mineral and fuel resources undermining prospects for development in the region? An inﬂuential body of research asserts that natural resources curse the countries that possess them with a host of undesirable outcomes — from economic stagnation, to authoritarian rule, to violent conﬂict. Africa is no stranger to these maladies. With international commodity prices booming, its dependence on resource exports is unlikely to diminish anytime soon. Is Africa suffering from a resource curse?
Proponents of the resource curse hypothesis identify several channels through which natural resources may harm development.1 In the economic sphere, lucrative natural resources attract attention and assets from other sectors with greater long-term growth potential. In the political sphere, the effects of natural resources depend on whether the state is able to capture resource wealth. Where the state does capture resource wealth, proceeds from resource extraction
1. See Paul Collier, “Natural Resources, Development and Conﬂict: Channels of Causation and Policy Interventions,” in Economic Integration and Social Responsibility: Annual World Bank Conference on Development Economics — Europe, 2004, ed. Franc¸ois Bourguignon, Pierre Jacquet, and Boris Pleskovic (Washington: World Bank, 2007), 323–335; Jeffrey A. Frankel, “The Natural Resource Curse: A Survey,” National Bureau of Economic Research Working Papers 15836 (2010); Michael L. Ross, “The Political Economy of the Resource Curse,” World Politics 51 (1999): 297–322.
enable governments to use a mix of patronage and repression to evade meaningful accountability to the general public. Where it does not, proceeds from resource extraction enable the rebel groups that control them ﬁnance violent insurgencies.
The most inﬂuential evidence marshalled behind the resource curse hypothesis consists of cross-national statistical associations between natural resource abundance and undesirable development outcomes. For example, in a widely cited study of the impact of natural resources on economic performance, Jeffrey Sachs and Andrew Warner found that a “surprising feature of modern economic growth is that economies with abundant natural resources have tended to grow slower than economies without substantial natural resources.”2 Turning to politics, in a widely cited study of the impact of natural resources on regime type, Michael Ross found that “the antidemocratic properties of oil and minerals are substantial.”3 And in a widely cited study of the causes of violent conﬂict, Paul Collier and Anke Hoefﬂer found that “primary commodity exports substantially increase conﬂict risk.”4
2. Jeffrey D. Sachs and Andrew M. Warner, “Natural Resources and Economic Growth” (Center for International Development, Harvard University, 1997), 2.
3. Michael L. Ross, “Does Oil Hinder Democracy?,” World Politics 53, no. 3 (2001): 342.
4. Paul Collier and Anke Hoefﬂer, “Greed and Grievance in
The empirical case for the resource curse hypothesis is far from settled, though. Several studies of the past decade have used similar cross-national statistical methods to dispute it. For example, Jean-Philippe Stijns measured resource abundance using data on the “stock” of resource reserves rather than the “ﬂow” of resource exports. Unlike Sachs and Warner, he found that “natural resource abundance has not been a signiﬁcant structural determinant of economic growth.”5 Stephen Haber and Victor Menaldo investigated the within-country relationships between natural resource trends and changes in political regime type. Unlike Ross, they found that “oil and mineral reliance does not promote dictatorship.”6 And James Fearon replicated Collier and Hoefﬂer’s statistical analysis with minor changes. Unlike them, he found that “the empirical association between primary commodity exports and civil war outbreak is neither strong nor robust” and appeared to be driven mainly by oil.7 Research of the past decade has chipped away at claims of a generalized resource curse that operates throughout the world.
Surprisingly little attention has been given to how well the resource curse hypothesis ﬁts the experience of contemporary Africa. On the face of it, Africa’s record since the end of the Cold War looks like an ideal “laboratory” for investigating the relationship between natural resources and development. As of 1990 the region had experienced more than a decade of economic decline, and nearly every country was under some form of authoritarian rule. Since the mid-1990s global
Civil War,” Oxford Economic Papers 56 (2004): 588.
5. Jean-Philippe C. Stijns, “Natural Resource Abundance and Economic Growth Revisited,” Resources Policy 30 (2005):
1. Stephen Haber and Victor Menaldo, “Do Natural Resources Fuel Authoritarianism? A Reappraisal of the Resource Curse,” American Political Science Review 105 (2011): 25.
2. James D. Fearon, “Primary Commodity Exports and Civil War,” Journal of Conﬂict Resolution 49 (2005): 503.
commodity prices have risen substantially, affecting resource-rich and resource-poor countries differently. Meanwhile, divergent political trajectories emerged, with some countries making transitions to democracy and others descending into spirals of state collapse and violent conﬂict.
Comparing the records of Africa’s resource-rich and resource-poor countries during this period of political and economic ﬂuidity approximates a natural experiment, which can help clarify the effects of natural resources on development outcomes. Yet basic empirical questions nevertheless remain largely uninvestigated. Have Africa’s resource-rich countries fared any worse since 1990 than their resource-poor counterparts have? Have their economies grown more slowly, have they been less likely to democratize, and have they been more prone to violent political conﬂict? And on a regional level, has Africa’s increasing dependence on proceeds from natural resources since 1990 corresponded with greater prevalence of these unfavorable outcomes?
In this paper I present a preliminary analysis of the empirical relationship between natural resources and development outcomes in sub-Saharan Africa since 1990. The analysis is preliminary in that I aim simply to establish whether natural resources and adverse development outcomes have tended to go together in practice. I do not attempt to test causal hypotheses about the effects of natural resources. Correlation does not imply causality, but knowing what correlation exists is a good place to start investigating a causal claim. I present the evidence in a series of graphs that show regional trends over time and clarify similarities and differences between resource-rich and resource-poor African countries.
The rest of the paper proceeds as follows. In the ﬁrst section, I address the challenge of measuring the theoretically relevant properties of natural resources, settling on a measure based on the unearned “rent” component of mineral and fuel extraction. The data conﬁrm a substantial increase in Africa’s resource rents during the two decades beginning in 1990, much of which has occurred since the late 1990s. In the second section, I successively examine the empirical relationships between resource rents and three “outcome” indicators — economic growth, political regime type, and political violence. In the third section, I outline a research agenda that ﬂows from analysis — focusing on issues the sustainability of resource-led growth, local and sectoral dimensions of resource governance, and the (potential) distinctiveness of oil. The fourth section is a brief conclusion.