State Failure

State Failure

Robert H. Bates
Department of Government, Harvard University, Cambridge, Massachusetts 02138;
email: robert.bates.harvard.edu@gmail.com


ABSTRACT

State failure is characterized by government predation and the militarization of civic society. Drawing on data from the study of civil war, state failure, and violence, this article explores the roles of per capita income, ethnicity, and democratization. It argues that public revenues are more relevant to state failure than are private incomes; that several configurations of ethnic groups can lead to violence; and that state failure is far more likely in intermediate democracies than in full democracies or aristocracies.


INTRODUCTION

“State failure” refers to the implosion of the state, by which is meant two things. The first is the transformation of the state into an instrument of predation. As states fail, politicians employ political power to levy resources from those who lack it. Rather than deploying the power of the state to enhance security, those in power use the state to promote their own interests, rendering others insecure.

Were the abuse of power the sole distinctive characteristic of state failure, then we would be unable to differentiate it from authoritarianism (see, e.g., Findlay 1991, Wintrobe 1998). The difference between the two arises from the second key characteristic of state failure: a loss of the monopoly over the means of coercion. When states fail, political competition takes place between groups bearing arms. Political parties become political militias as elites transform them into military bands. Private firms seeking protection and private citizens seeking security affiliate with these militias as they search for sources of the security that the state no longer provides.

Note the distinction between state failure and revolution. Both lead to violence; but revolution creates a new order, whereas state failure yields disorder. As stressed Marx & Engels (1994 [1848]), revolutions result in new ways of organizing the economy and polity. Thus did Tocqueville (1996 [1866]) see in the French Revolution the emergence of the bureaucratic state, freed from the constraints of the patrimonial order of l'ancien regime. Thus too did Skocpol (1979) view revolutions as political attempts to refashion underlying modes of production, thereby enriching states hitherto impoverished by the unproductive nature of their economies.

It is also necessary to distinguish between state failure and civil war. Not all civil wars originate in failed states. The civil war in the 1860s in North America pitted a rapidly modernizing central government against a highly militarized secessionist movement (Moore 1966). Vietnam's civil war in the 1960s resulted from conflict between a coherent and capable government in the North and an incoherent and corrupt regime in the South, propped up by foreign allies. In neither case did civil war follow the implosion of a state; rather, in both instances, the state was strong—and becoming stronger.

Sierra Leone offers an apt illustration of state failure. As documented by Reno (1995, 1998)—one of the first to explore the phenomenon in depth—the economy of Sierra Leone had been looted by Siaka Stevens and his henchmen, who despoiled the diamond and mineral industries and looted the earnings of those who exported timber and agricultural products. The result, Reno argues, was the collapse of political order and the rise of political violence, as citizens and firms backed political movements that offered to safeguard life and property. The process of disintegration in Sierra Leone has been paralleled in Somalia, Yugoslavia, Afghanistan, and elsewhere, albeit with obvious and important variations (for a list of failed states, see Bates et al. 2006).


THE INCIDENCE OF STATE FAILURE


The Political Instability Task Force (Bates et al. 2006) counts as state failure the outbreak of ethnic war, revolutionary war, and genocide, as well as “adverse regime changes.” The latter term means the collapse of the state, as in Somalia; the revolutionary displacement of an elite, as in Iran; the dissolution of a state, as in Yugoslavia; or the shift from a democratic to an authoritarian regime, as in Haiti. Reviewing the history of all independent countries with populations of 500,000 or more, the Task Force records 139 cases of state failure. Their data indicate that the rate of state failure has remained remarkably stable at two to three cases per year, rising abruptly (to roughly five) once in the early 1960s and again in the early 1990s. Because states that fail are often slow to recover, the stock of failed states doubled from the late 1950s to the mid-1960s and then rose sharply again in the 1990s. Since then, the stock has declined: Failed states now constitute <10% of the global population of states.

The match between the Task Force's measure and the characterization of state failure offered above is imperfect. In addition, the Task Force's analysis focuses on prediction rather than explanation. They locate risk factors that enable them accurately to forecast, with a two-year time lag, the breakdown of states. Although the factors that they employ—such as life expectancy at birth—may yield reliable forecasts of state failure, they offer few insights into the causes of state failure. To probe the origins of state failure, we therefore search for other sources.

Africa is one of the globe's most violence-prone and war-torn regions, and in their study of the political economy of postindependence, Ndulu et al. (2007) use a measure of state failure that more closely corresponds to the notion that frames this review. Ndulu et al. count as failed states those in which a central government has disintegrated and political order collapsed, as marked by the mobilization of militias and the spread of violence. However, they include only African cases, and so we return once again to the search for useful data.

In doing so, we pause to note a suggestive finding (Table 1). In 60% of the country-years in which there were failed states in Africa between 1970 and 1995, there were also civil wars; and in 70% of the country-years in which there were civil wars, there were failed states (Ndulu et al. 2007).



Table 1 Relationship between civil war and state failure (Ndulu et al. 2007)a



The close correspondence between civil wars and state failure in Africa directs our attention toward the possibility of using data on the first to study the second. Further encouragement comes from Fearon & Laitin (2003), whose data on civil wars (Figure 1) trace a time profile that closely parallels that of the Task Force's data on state failure (Figure 2). The close correspondence is important; whereas much is known about civil wars, little systematic research has been conducted on the causes of state failure. When addressing three major themes—the roles of economic factors, ethnicity, and democracy in the process of state failure—we therefore take frequent recourse to the data on civil war.


ECONOMIC FACTORS

Those who study the impact of economic factors on political disorder focus on the impacts of poverty, public revenues, and natural resources.
Poverty

One of the most robust findings is the relationship between poverty and political disorder (Sambanis & Hegre 2006). The lower the per capita income, the greater the likelihood of civil war (Collier & Hoeffler 2004); and, as reported by the Political Instability Task Force (Goldstone et al. 2003), the lower the life expectancy at birth—a close correlate of poverty—the higher the likelihood of state failure.

Collier & Hoeffler (2004) explore competing interpretations of their finding. One, which they reject, links conflict to anger over economic inequality and political injustice and thus stresses the importance of grievance. The second, which they endorse, highlights the role of financial incentives for rebellion. In poor societies, they argue, armed forces can be recruited cheaply; the returns to redistribution compare favorably with the returns from production, and labor commands a higher price in military activity. “Greed,” they conclude, thus provides the more compelling explanation of the relationship between per capita income and civil war.

Sambanis & Hegre (2006) run a dizzying number of cross-national regressions and conclude that income and population size constitute (the only) robust correlates of civil war. Nonetheless, the finding remains contestable, if only because other factors could be at play. Gross domestic product per capita may stand as a proxy for government capabilities (Fearon & Laitin 2003); were other measures of these capabilities available, then the relationship between income and political disorder would presumably weaken in equations that included them. Alexander (2007) reports that when per capita income and civil war were entered in equations that included country-specific fixed effects, no relationship could be found—a finding that underscores the importance of time-invariant variables such as political history or the structure of political institutions.
Fiscal Dearth

Recall that the opening discussion identified two aspects of state failure: the self-interested use of the power of the state by those who occupy positions within it and the state's loss of a monopoly on violence. In probing further the role of economic forces, we return to these themes.

It may not be the poverty of the citizens but rather the poverty of the state that precipitates political disorder (Skocpol 1979, Bates 2008). Without public revenues, governments cannot pay their civil servants (who may therefore turn to corruption) or their armed forces (which may then use their weapons to pay themselves). When those in power lack the resources to induce influential groups—military officers or regional elites, for example—to remain politically faithful, then the state begins to disintegrate (see also Azam 2007, Azam & Mesnard 2003).

These findings help to account for the temporal distribution of the data reported above, which indicate that both the rate and stock of civil wars and state failures peaked at the time of the global recession of the late twentieth century. In the 1970s, higher oil prices sent the industrial economies reeling, and exports from the poorer nations therefore declined; as the governments in the developing world drew their taxes from trade, their public revenues also declined. Governments borrowed to make up for this shortfall; but when in the early 1980s the Federal Reserve raised interest rates in the United States, these governments incurred an increased burden of debt. Pressured by their creditors to retrench their spending, governments froze salaries. Bureaucrats then turned to extracting bribes and security services to preying on shop keepers or travelers on the roads. When those with public power turned predatory, the process of state failure began.

Measuring private incomes on the one hand and the likelihood of state failure on the other, observers find that a decline in the first correlates with an increase in the second. But the causal path between economic decline and political implosion runs, the evidence suggests, through the public treasury rather than through the private economy.
Resource Wealth

In an article drafted in the mid 1990s, Collier & Hoeffler (2004) argue that economies that contain valuable natural resources are more likely to become violence torn (see also Collier 2000, Mueller 2004). In interpreting this finding, they argue that rebel groups seek wealth and seize control over resources to finance their operations.

In a reanalysis of Collier & Hoeffler's data, Fearon (2005) criticizes their measure of resource wealth (the value of primary products as a percentage of exports), noting that it omits the value of exports of gem stones and ferrous metals. He also notes that the relationship between Collier & Hoeffler's measure and the likelihood of civil war largely disappears once allowance is made for missing data. Fearon & Laitin (2003) find that among all natural resources, only petroleum exports significantly correlate with the likelihood of internal conflict. Bates (2008) replicates this finding in his analysis of African cases.

Fearon & Laitin (2003) and Bates (2008) thus call into question the generality of Collier & Hoeffler's (2004) findings. They also challenge their account of the relationship between natural resource wealth and political insurrection. As stressed by Haber et al. (2004), rebel armies cannot loot petroleum fields; they lack the capital and technical skills required to operate them. Moreover, in Africa at least, a large portion of the major oil fields lie off-shore and therefore out of the reach of armed militias, and yet Africa's oil states remain significantly more prone to state failure.

Because revenues from petroleum production flow into the coffers of states and not into the bank accounts of militias, the link between oil revenues and state failure must run through the impact of oil revenues on governments rather than on rebels. Relevant is the literature on the “rentier state” (Mahdavy 1970, Beblawi & Giacomi 1987, Chaudry 1994, Shambayati 1994, Yates 1996; see also Karl 1997), which argues that because petroleum states do not depend on their people for public revenues or the payment of taxes, they tend to rule in an arbitrary and repressive manner. Though drawn largely from the study of the Middle East, this argument finds confirmation in accounts of the Sudan, Chad, and Angola, whose governments engage in high levels of repression and often attack their own people. And as illustrated by recent events in Nigeria and Venezuela, even when the governments of oil states are subject to election, they possess the resources to corrupt the electoral process and thereby elude institutions designed to induce political restraint.


ETHNICITY


One line of research thus focuses on the interaction between economic forces and state failure. A second focuses on the nature of society and, in particular, on its ethnic composition. Following the breakup of the Soviet Union, ethnic rivalries animated civil wars in Eastern Europe, the Balkans, and the Caucasus. Clashes between Hutu and Tutsi, Kalenjin and Kikuyu, and Krahn and Gio lay close to the core of civil strife in Rwanda, Kenya, and Liberia, respectively. So prominent a role have ethnic groups played in recent outbreaks of political disorder that Sambanis (2001), Goldstone et al. (2003), and Fearon (2004) have treated “ethnic wars” as a special kind of conflict, seeking to determine whether they are more intense and longer lived than other forms of conflict.

There can be little doubt that ethnic groups quarrel, be it over the control of trade or the distribution of wealth and power (the classic on this subject remains Horowitz 1985). Upon closer inspection, however, ethnicity proves to bear a less straightforward relationship to the likelihood of state failure than might be expected.
Diversity and Polarization

Many scholars focus on ethnic diversity. Using ELF (Taylor & Hudson 1972), a frequently employed index of ethnolinguistic fractionalization, Fearon & Laitin (1996, 2003) painstakingly address Collier & Hoeffler's (2004) claim that higher levels of ethnic diversity yield a greater likelihood of civil war. Using annual data rather than five-year panels and multiple imputation rather than case-wise deletion, they are unable to replicate Collier & Hoeffler's findings. In his analysis of cross-national data from Africa, Bates (2008) explores the relationship between a variety of measures of ethnic diversity1 and the likelihood of state failure; even in this most culturally diverse and conflict-ridden of continents, he finds no evidence of such a relationship.

Others, such as Reynal-Querol (2002), focus on ethnic polarization. Measures of polarization capture the simplification rather than the diversity of the composition of populations; Reynal-Querol's measure achieves its maximum when the population sorts into two groups, each ethnically distinct. Bates (2008) finds that this measure too fails to correlate with the likelihood of state failure in his sample of African cases.

Those who focus on ethnic diversity and ethnic polarization do so because they believe that there is a relationship between the distribution of political preferences and the likelihood of cooperation. In the context of this essay, the implication would be that ethnically diverse or polarized societies could be expected to fail to reach agreements; when attempting to choose public policies or to finance public goods, they would find it difficult to prioritize. Therefore, if poor, their bureaucracies would be ineffective; if wealthy, they would be sprawling, with each group receiving its preferred package of public services. The ineffective bureaucracies would result in a weak state; the oversized bureaucracies would result in an expensive and thus potentially extractive one.

Evidence of the former comes largely from the United States. In their study of municipal governments, Alesina et al. (1999) find that the more ethnically diverse the jurisdiction, the lower the level of public good provision. Evidence of the latter comes from studies of Africa, where, in order to satisfy the demands of ethnic groups, the state expands the level of public employment and the size of the bureaucracy (e.g., Kasfir 1976). When the level of public revenues subsequently declines, the bureaucracy then becomes “a force for itself,” using public power to secure the private incomes of its members. Overgrown and underfed, the public services turn predatory.
Concentration

While some scholars address the significance of ethnic diversity and polarization, others stress ethnic concentration. Groups that constitute a majority of the population can reasonably aspire to dominate politically; even in states in which elections do not take place, numbers count. In the presence of a dominant group, others might reasonably fear the “privatization” of the state and the prospect, therefore, of future exclusion from the benefits that power can provide. Should the majority be large, then little could be done to alter the political consequences; but should it be narrow, its dominance might be contested. Others might prefer to pay the costs of challenging the suzerainty of the majority, with the possibility of victory, rather than to bear the costs of permanent exclusion.

In light of this reasoning, it is notable that Bates (2000) finds a curvilinear relationship between the size of the largest ethnic group and the likelihood of violence in an African set of countries, and that Collier & Hoeffler (2004) report a similar finding for a global set. Others, though failing to find a relationship between measures of ethnicity and the likelihood of civil war, report a positive relationship between a measure of ethnic concentration and the duration of civil wars, particularly ethnic civil wars (Collier et al. 1998, Elbadawi & Sambanis 2000, Fearon 2004).
Rule by Minorities

In a surprising number of cases, minorities rule. When they do so, they often prepare the ground for state failure.

Miquel (2005) argues that minorities retain power by rendering its loss costly. If those who hold power use it to render the lives of those who lack it miserable and insecure, then members of the elite could reasonably expect their downfall to result in retaliation by those whom they had previously ruled. By inspiring hatred, incumbents can thereby lock in the loyalty of their supporters, discouraging defection and eliciting political vigilance.

Bueno de Mesquita & Smith (2003) and Adam & O'Connell (1999) argue that minority regimes not only repress those they rule but also impoverish them. When a minority employs the state to redistribute income, it taxes everyone: It seizes a portion of the wealth of all members of the polity. But it consumes the benefits itself. Paying only a small portion of the costs while consuming all of the benefits, minority regimes will choose a higher level of resource extraction than would larger regimes. In an effort to retain power, minority regimes choose to repress; in order to prosper, they secure high rates of redistribution.

Evidence for the first argument appears in the findings of Collier et al. (2007). Using data from Africa, 1960–2002, they find a relationship between high levels of ethnic fractionalization and what they call “rule by fear,” i.e., cases in which governments employ high levels of force and violence (see Figure 3).
View larger version (30K)
Figure 3 Relationship between ethnic fractionalization and rule by fear (Collier et al. 2007). Data are from Africa, 1960–2002.
Information about Downloading


Reanalyzing their data yields evidence in support of the second argument (Table 2). The higher the level of ethnic diversity or rule by fear, the coefficients suggest, the greater the likelihood that governments will use public power to engage in regional or ethnic redistribution (Equation 1) or looting (Equation 2).2



Table 2 Ethnic diversity, rule by fear, and the behavior of African governmentsa



At first sight, the data in Equation 3 tend to counter, rather than to support, the thrust of this analysis. They suggest that higher levels of ethnic diversity and the use of fear to govern associate with a lower likelihood of state failure. Attending to the coefficient on the interaction term alters this initial assessment, however. The coefficient suggests that when ethnic minorities employ rule by fear—i.e., when the measures of fractionalization and rule by fear interact—then state failure does indeed become more likely.3


DEMOCRATIZATION


In addition to economic and cultural factors, scholars stress the role of democratization in state failure. Most employ the well-known Polity index, which assigns low numbers to autocracies, high numbers to democracies, and intermediate values to “partial” or “unconsolidated” democracies. Virtually every researcher who explores this relationship reports that civil wars and state failures disproportionately take place in states that occupy this intermediate category.4

As is well known, many of the intermediate democracies are young. Some were created during the process of decolonization and others at the end of the Cold War. Should new states be politically less stable than old, the relationship between regime type and conflict could therefore be the result of the age of intermediate democracies rather than their political characteristics. In a carefully crafted paper, Hegre et al. (2001) address this possibility and find that the “inverted U” relationship between regime type and civil war persists after controlling for the date of independence.

Hegre (2003) finds that low-income democracies are 15–20 times more likely to experience civil war than are high-income democracies; he finds no significant difference in the likelihood of civil war when he compares low- and high-income autocracies, however. A similar pattern emerges when he employs literacy as a measure of development, although the difference between under- and fully developed democracies is then far less pronounced than when the level of development is proxied by income. Similar findings emerge from the Political Instability Task Force (Goldstone et al. 2005), which finds that partial democracies and autocracies with some political competition—i.e., intermediate regimes—are up to 20 times more likely to experience state failure than are full democracies or consolidated autocracies (authoritarian regimes without political competition), other things being equal. Democracy and underdevelopment thus appear to be a volatile combination (see also Mansfield & Snyder 1995, Snyder 2000).

Restricting his attention to African cases, Bates (2008) reports that the period of democratization in Africa was also the period of state failure. Using qualitative as well as quantitative data, he argues that forces favoring democratization originated at the global level and were especially strong at the end of the Cold War. Had incumbent autocracies prepared for a transition to democracy, they would have behaved with great restraint, he argues; having failed to do so, they had reason to fear political reprisals following the loss of power. In response to the increase in insecurity inspired by the transition to democracy, incumbents turned predatory, increasing the level of corruption, looting public assets, and suppressing those who attempted to displace them. Facing threats from their putative guardians, people began to provide their own security. They formed militias and states broke down.


CONCLUSION


The study of state failure focuses on the implosion of political regimes. State failure is marked by political predation; when states fail, those with power employ it to extract resources from those without power. The latter flock to those who offer them security, albeit often for a price: the obligation to contribute to their new political community, in some cases by bearing arms. Political predation from the top is thus accompanied by the militarization of civic society below. The state no longer possesses a monopoly on the use of force, and society is plunged into political disorder.

As noted above (Figure 2), the Political Instability Task Force reports that the stock of failed states has declined in recent years. Research by Doyle & Sambanis (2006) suggests one reason: International intervention is generally effective, leading to the cessation of conflict and a return to political order. Studies by Collier et al. (2006), however, caution against too optimistic an assessment. Past conflict, they find, associates with a higher likelihood of new outbreaks of violence; roughly 40% of civil wars resume after their termination. The currents of violence may have abated but not died out.

Some view the breakdown of order in Somalia as the first modern instance of state failure; others, the collapse of political order in the Balkans. Whichever benchmark is chosen, attempts to investigate the phenomenon systematically are but two decades old. Although the extent and depth of this research speaks well of those who have undertaken it, more remains to be done.

Early efforts have focused on the impact of variables commonly measured by international institutions, most of which produce economic data. But as the Political Instability Task Force reports (Goldstone et al. 2003), the factors that best predict state failure are political, not economic. The nature of political regimes associates far more strongly with subsequent political instability than do indicators of economic wellbeing. Progress in understanding the political origins of state collapse is presently constrained, however, by its reliance on “formalistic measures,” i.e., measures that capture the structure of political institutions. Needed too is systematic information regarding the strategic properties of the national political game. It is these that shape political choices, including the choice of whether to compete at the ballot box or in the field of battle (see the conclusion to Bates et al. 1998).

Although recent research has tended to emphasize the political wellsprings of state failure, future research needs to employ new kinds of data. In addition to incorporating information concerning deeper political forces, it needs to make systematic use of subnational data. The origins of political disorder lie in conflicts whose own origins are, to a great degree, internal to the nation-state: regional inequality, conflicting partisan preferences, religious differences, and so on. Aggregate, national-level data offer the wrong optic by which to view within-country conflict.

Recently, scholars have begun to make use of subnational data. Vargas & Dube (2006) assess the relationship between changes in the value of primary products and the intensity of political conflict in Colombia. Others explore the relationship between the distribution of ethnic groups and the incidence of political violence (e.g., Cederman & Giradin 2007). At least as important, some have used such data to implement quasi-experimental designs (e.g., Vargas & Dube 2006), something that is difficult to accomplish with cross-national data. These innovations suggest that in the future, state failure will be viewed though the appropriate lens and analyzed with techniques that will offer deeper, more precise insights into its origins.


disclosure statement


The author is not aware of any biases that might be perceived as affecting the objectivity of this review.
Download the bibliography to reference management software


literature cited

Adam CS, O'Connell SA. 1999. Aid, taxation, and development in subsaharan Africa. Econ. Polit. 11: 225–54 [CrossRef]
Alesina A, Baqir R, Easterley W. 1999. Public goods and ethnic divisions. Q. J. Econ. 114: 1243–84 [CrossRef] [ISI]
Alesina A, La Ferrara E. 2005. Ethnic diversity and economic performance. J. Econ. Lit. 43: 762–800 [CrossRef] [ISI]
Alexander M. 2007. Is poverty to be blamed for civil wars? Dep. Gov., Harvard Univ.
Azam J-P. 2007. Chapter 6: The political geography of redistribution. See Ndulu et al. 2007
Azam J-P, Mesnard A. 2003. Civil war and the social contract. Public Choice 115: 455–75 [CrossRef] [ISI]
Bates RH. 2008. When Things Fell Apart: State Failure in Late Century Africa. New York: Cambridge Univ. Press
Bates R, Epstein D, Goldstone J. 2006. Political Instability Task Force Report, Phase IV Findings. McLean, VA: SAIC
Bates RH. 2000. Ethnicity and development in Africa. Am. Econ. Rev. 90: 131–34 [ISI]
Bates RH, Greif A, Levi M, Rosenthal J-L, Weingast B. 1998. Analytic Narratives. Princeton, NJ: Princeton Univ. Press
Beblawi H, Giacomi L, eds. 1987. The Rentier State. London: Croom Helm
Bueno de Mesquita B, Smith S. 2003. The Logic of Political Survival. Cambridge, MA: MIT Press
Cederman L-E, Giradin L. 2007. Beyond fractionalization: mapping ethnicity onto nationalist insurgencies. Am. Polit. Sci. Rev. 101: 173–85 [CrossRef] [ISI]
Chaudry KA. 1994. Economic liberalization and the lineages of the rentier state. Comp. Polit. 27: 1–24 [CrossRef]
Coe A. 2007. Civil war and state failure. In Papers Presented to Government 2728. Harvard Univ.
Collier P. 2000. Rebellion as a quasi-criminal activity. J. Confl. Resolut. 44: 83–53
Collier P, Bates R, Hoeffler A, O'Connell S. 2007. Chapter 11: Endogenizing syndromes. See Ndulu et al. 2007
Collier P, Hoeffler A. 2004. Greed and grievance in civil wars. Oxf. Econ. Pap. New Ser. 56: 563–95 [CrossRef] [ISI]
Collier P, Hoeffler A, Soderbaum M. 2006. Post conflict risk. Work. pap., Cent. Stud. Afr. Econ., Dep. Econ., Oxford Univ.
Collier P, Hoeffler A, Soderman M. 1998. On the Duration of Civil War and Postwar Peace. Work. pap., Cent. Stud. Afr. Econ., Oxford Univ.
Doyle M, Sambanis N. 2006. Making War and Building Peace. Princeton, NJ: Princeton Univ. Press
Elbadawi IA, Sambanis N. 2000. Why are there so many civil wars in Africa? J. Afr. Econ. 9: 244–69 [CrossRef] [ISI]
Fearon J, Laitin D. 2003. Ethnicity, insurgency and civil war. Am. Polit. Sci. Rev. 97: 75–90 [ISI]
Fearon JD. 2004. Why do some civil wars last so much longer than others? J. Peace Res. 41: 275–301 [CrossRef] [ISI]
Fearon JD. 2005. Primary commodities exports and civil war. J. Confl. Resolut. 49: 483–507 [CrossRef] [ISI]
Fearon JD, Laitin DD. 1996. Explaining interethnic cooperation. Am. Polit. Sci. Rev. 90: 715–35 [CrossRef] [ISI]
Findlay R. 1991. The new political economy: its explanatory power. In Politics and Policy Making in Developing Countries, ed. GM Meier, pp. 13–40. San Francisco: Int. Cent. Econ. Growth
Goldstone J, Bates R, Gurr TR, Marshall M. 2005. A Global Forecasting Model of Political Instability. State Failure Task Force. McClean, VA: SAIC
Goldstone J, Marshall M, Bates R, Epstein D. 2003. State Failure Task Force Project, Phase III Report. McLean, VA: SAIC
Haber S, Razo A, Maurer N. 2004. The Politics of Property Rights: Political Instability, Credible Commitments, and Economic Growth in Mexico, 1876–1929. Cambridge, UK: Cambridge Univ. Press
Hegre H. 2003. Disentangling democracy and development as determinants of armed conflict. Presented at Annu. Meet. Int. Stud. Assoc., Portland, OR
Hegre H, Gates S, Gleditsch NP. 2001. Toward a democratic civil peace? Democracy, political change and civil war, 1816–1992. Am. Polit. Sci. Rev. 95: 33–48 [ISI]
Horowitz D. 1985. Ethic Groups in Conflict. Berkeley and Los Angeles: Univ. Calif. Press
Karl TL. 1997. The Paradox of Plenty. Berkeley and Los Angeles: Univ. Calif. Press
Kasfir N. 1976. The Shrinking Political Arena. Berkeley and Los Angeles: Univ. Calif. Press
Mahdavy H. 1970. The patterns and problems of economic development in rentier states: the case of Iran. In Studies in Economic History of the Middle East, ed. MA Cook, pp. 428–67. New York: Oxford Univ. Press
Mansfield E, Snyder J. 1995. Democratization and war. For. Aff. 74: 79–97
Marx K, Engels F. 1994 (1848). The Communist Manifesto. Berkeley, CA: Cent. Socialist Hist.
Miguel E, Gugerty MK. 2005. Ethnic diversity, social sanctions, and public goods in Kenya. J. Pub. Econ. 89: 2325–68 [CrossRef] [ISI]
Miquel GP. 2005. Control of Politicians in a Divided Society: The Politics of Fear. Milan: IGIER
Moore B. 1966. Social Origins of Dictatorship and Democracy. Boston: Beacon
Mueller J. 2004. The Remnants of War. Ithaca and London: Cornell Univ. Press
Ndulu B, Collier P, Bates R, O'Connell S. 2007. The Political Economy of Economic Growth in Africa, 1960–2000. 2 vols. Cambridge, UK: Cambridge Univ. Press
Posner D. 2004. Measuring ethnic fractionalization in Africa. Am. J. Polit. Sci. 48: 849–63 [CrossRef] [ISI]
Reno W. 1995. Corruption and State Politics in Sierra Leone. Cambridge, UK: Cambridge Univ. Press
Reno W. 1998. Warlord Politics and African States. Boulder, CO: Lynne Rienner
Reynal-Querol M. 2002. Ethnicity, political systems, and civil wars. J. Confl. Resolut. 46: 29–54 [CrossRef] [ISI]
Sambanis N. 2001. A Review of Recent Advances and Future Directions in the Quantitative Literature on Civil War. Work. pap., Dep. Polit. Sci., Yale Univ.
Sambanis N, Hegre H. 2006. Sensitivity analysis of empirical results on civil war onset. J. Confl. Resolut. 50: 508–35 [CrossRef] [ISI]
Scarritt J, Mozaffar S. 1999. The specification of ethnic cleavages and ethnopolitical groups for the analysis of democratic participation in Africa. Nationalism and Ethnic Polit. 5: 82–117
Shambayati H. 1994. The rentier state, interest groups, and the paradox of autonomy. Comp. Polit. 26: 307–31 [CrossRef] [ISI]
Skocpol T. 1979. States and Social Revolutions. Cambridge, UK: Cambridge Univ. Press
Snyder J. 2000. From Voting to Violence. New York: WW Norton
Taylor CL, Hudson MC. 1972. World Handbook of Political and Social Indicators. New Haven and London: Yale Univ. Press
Tocqueville A. 1966 (1866). The Ancien Regime and the French Revolution. New York: Collins
Vanhanen T. 1999. Domestic ethnic conflict and ethnic neoptism: a comparative analysis. J. Peace Res. 36: 55–73 [CrossRef] [ISI]
Vargas J, Dube O. 2006. Are All Resources Cursed: Coffee, Oil, and Armed Conflict in Colombia. Work. pap., Weatherhead Cent. Int. Dev., Harvard Univ.
Wintrobe R. 1998. The Political Economy of Dictatorship. Cambridge and New York: Cambridge Univ. Press
Yates D. 1996. The Rentier State in Africa: Oil Rent Dependency and Neocolonialism in the Republic of Gabon. Trenton, NJ: Afr. World Press