Roundtable 11-3 on Ethnic Politics and State Power in Africa: The Logic of the Coup–Civil War Trap60 min read

Since the periods of decolonization and the Cold War, Africa has been the site of numerous protracted conflicts. Some countries have experienced repeated cycles of violence and civil war, while other countries have headed off major conflict and maintained relative peace. What factors account for these differences? In a clear, compelling, theoretically innovative study, Philip Roessler argues that civil wars often emerge from power struggles among political elites. In fragile, ethnically divided states, powerholders have greater fear of coups d’état staged by rivals at the core than civil wars waged by excluded minorities on the periphery. As a result, powerholders in weak states are likely to incorporate potential rivals at the center in a system of ethnic power sharing and to exclude regional and ethnic minorities on the margins. While such tactics may work in the short term, ongoing rivalry at the center, compounded by political exclusion elsewhere, may eventually lead to the coup–civil war trap. In such cases, rival power aspirants may mobilize marginalized groups, with civil war as an outcome.

Philip Roessler.  Ethnic Politics and State Power in Africa:  The Logic of the Coup–Civil War Trap.  Cambridge:  Cambridge University Press, 2016.  ISBN:  9781107176072 (hardback, $110.00); 9781316628218 (paperback, $36.99).

7 October 2019 |
Editors: Thomas Maddux and Diane Labrosse | Production Editor: George Fujii


PDF Version

Introduction by Elizabeth Schmidt, Loyola University Maryland

Since the periods of decolonization and the Cold War, Africa has been the site of numerous protracted conflicts. Some countries have experienced repeated cycles of violence and civil war, while other countries have headed off major conflict and maintained relative peace. What factors account for these differences? In a clear, compelling, theoretically innovative study, Philip Roessler argues that civil wars often emerge from power struggles among political elites. In fragile, ethnically divided states, powerholders have greater fear of coups d’état staged by rivals at the core than civil wars waged by excluded minorities on the periphery. As a result, powerholders in weak states are likely to incorporate potential rivals at the center in a system of ethnic power sharing and to exclude regional and ethnic minorities on the margins. While such tactics may work in the short term, ongoing rivalry at the center, compounded by political exclusion elsewhere, may eventually lead to the coup–civil war trap. In such cases, rival power aspirants may mobilize marginalized groups, with civil war as an outcome.

Until now, most scholars of coups and civil wars in Africa have analyzed them separately. Roessler, in contrast, offers a theory that considers both phenomena as part of a whole, which he tests by analyzing qualitative case studies and quantitative datasets. The two primary case studies (Darfur in western Sudan and the Democratic Republic of Congo), which involved intensive fieldwork, are supplemented by several subsidiary cases (notably, Burundi, Chad, Liberia, Nigeria, Uganda). With its wide reach, innovative approach, and sophisticated methods, the findings of Roessler’s book are significant. They have broad implications for our understanding of multiple conflicts in Africa and in the Global South more generally.

Three scholars with expertise in African security, conflict, and governance issues contribute to this discussion of Ethnic Politics and State Power in Africa. All three praise Roessler’s work as path-breaking in terms of methodological and theoretical innovation and in the presentation of new evidence from across the continent. They also suggest areas that might be strengthened as well as topics for further research.

Hilary Matfess, whose work focuses on gender, security, and governance, describes the study as “a tour de force that brings together extensive qualitative fieldwork, theory-building and theory-testing case studies, and quantitative analysis” that gives greater nuance to our understanding of ethnicity in Africa and makes a significant contribution to the literature on African institutions, incentives, and violence. She notes that Roessler’s creative new theory and firm evidentiary base demonstrate “why ethnicity is the dominant socio-political cleavage in Africa.” She concludes that his findings offer important insights that are applicable to other regions.

Matfess also points to areas that might be strengthened or extended. When leaders are faced with challenges to their authority, what options do they have besides the co-optation of rivals or civil war? She suggests that Roessler might consider the prospects for consociationalism, whereby ethnic or regional groups are granted collective rights enforced by international bodies, which would lead to “international preferences shap[ing] domestic politics.” (The two other reviewers warn that foreign involvement may have a negative impact, especially when the intervening powers are also interested parties.) Finally, Matfess recommends further investigation of “the dynamics of in-group power jockeying” and how “multiple or cross-cutting cleavages” such as political and religious affiliations can increase the potential for violence.

Jason Lyall, whose research investigates the dynamics, effects, and effectiveness of political violence, calls Roessler’s study “path-breaking in design and execution,” and one that successfully uses multiple methods to illuminate the delicate relationship between powersharing and political violence in Africa. Offering “a compelling mix of theoretical rigor, methodological pluralism, and new empirical results,” the book in his opinion “deserves a wide audience among scholars of political violence.”

Lyall also suggests topic for further investigation. He notes that the strength of Roessler’s theory lies in its predictive power, yet Roessler does not delve into the nature of the opposition’s “threat capabilities” and the ways in which they are mobilized. For instance, how do power seekers from the same ethnic group bargain amongst themselves? While Roessler gives a great deal of attention to “violence specialists” in the ruling regime and in the opposition, he pays less attention to African militaries, which are significant perpetrators of violence on the continent. International factors are also neglected, although they have important consequences for conflict outcomes, especially when neighboring states are interested parties. In contrast to Matfess, however, Lyall considers foreign involvement to be a double-edged sword—as likely to be detrimental as it is helpful to the cause of regional peace.

Kai M. Thaler, whose work examines violence, civil wars, and regime stability and change, refers to Roessler’s study as “a tour de force of mixed-methods research [that uses] rich original qualitative data and quantitative evidence” to develop a nuanced understanding of ethnic politics and conflict in Africa. The book’s “powerful and compelling” argument, built from a solid evidentiary base, “advances the study of civil conflict not only in Africa, but globally.” Like Lyall, Thaler would like more information on the role of foreign intervention in the coup–civil war trap. He would like insight into cases where politics are “characterized primarily by non-ethnic political cleavages” and others where ideology has replaced ethnic identification with a “shared national, multiethnic identity” or other sociopolitical identities. Do such cases signal prospects for overcoming the coup–civil war trap?

Responding to the reviewers’ comments, Philip Roessler recapitulates “how the balance of threat capabilities between [rival groups] mediates bargaining outcomes.” He also clarifies the role that ethnicity plays in the coup–civil war trap, arguing that while “ethnic divisions compound the …problem …they do not cause it.” At the heart of the matter is a political problem: the struggle for power that can also emerge among players of the same ethnic group. However, Roessler notes, intra-ethnic civil war is far less common than intra-ethnic coups, perhaps as a result of self-regulation. While the reviewers raise other important issues, Roessler concludes, those are topics for future book-length studies. He completes his response by highlighting some of the new scholarship that addresses these critical topics.


Philip Roessler is Associate Professor of Government and Co-Director of the Center for African Development at William & Mary. His research focuses on conflict, state-building, and development in Africa. Drawing on extensive field research across the region, he has written two books: Ethnic Politics and State Power in Africa: The Logic of the Coup-Civil War Trap (Cambridge University Press, 2016) and Why Comrades Go to War: Liberation Politics and the Outbreak of Africa’s Deadliest Conflict (Hurst Publishers/Oxford University Press, 2016) with Harry Verhoeven. His articles have been published in the American Journal of Political Science, International Organization, Journal of Politics, Nature, World Politics, and other outlets. He received his B.A. from Indiana University and his Ph.D. from the University of Maryland.

Elizabeth Schmidt is Professor Emeritus of History at Loyola University Maryland. She received her Ph.D. from the University of Wisconsin and has written extensively about U.S. involvement in apartheid South Africa, women under colonialism in Zimbabwe, the nationalist movement in Guinea, and foreign intervention in Africa from the Cold War to the war on terror. Her books include: Foreign Intervention in Africa after the Cold War: Sovereignty, Responsibility, and the War on Terror (Athens: Ohio University Press, 2018); Foreign Intervention in Africa: From the Cold War to the War on Terror (Cambridge University Press, 2013); Cold War and Decolonization in Guinea, 1946-1958 (Ohio University Press, 2007); Mobilizing the Masses: Gender, Ethnicity, and Class in the Nationalist Movement in Guinea, 1939-1958 (Portsmouth, NH: Heinemann, 2005); Peasants, Traders, and Wives: Shona Women in the History of Zimbabwe, 1870-1939 (Heinemann, 1992); and Decoding Corporate Camouflage: U.S. Business Support for Apartheid (Institute for Policy Studies, 1980).

Jason Lyall is Associate Professor of Political Science and the Director of the Political Violence FieldLab at Yale University. He is the author of Divided Armies: Inequality and Battlefield Performance in Modern War (Princeton University Press, 2020). His research on the dynamics, effects, and effectiveness of political violence has also been published in American Political Science Review, American Journal of Political Science, International Organization, Journal of Conflict Resolution, Journal of Peace Research, Journal of Politics, and World Politics, among others.

Hilary Matfess is a Ph.D. student in Yale University’s Political Science department. She is the author of Women and the War on Boko Haram: Wives, Weapons, Witnesses (Zed Books, 2017). Her current work focuses on gender, security, and governance in sub-Saharan Africa.

Kai M. Thaler is Assistant Professor of Global Studies at the University of California, Santa Barbara and from 2018-2019 a Postdoctoral Fellow at the Sié Chéou-Kang Center for International Security and Diplomacy at the University of Denver’s Josef Korbel School of International Studies. His work on civil wars, violence, regime stability and change, and methods for studying violent conflict has been published in journals including Afrique Contemporaine, Civil Wars, Genocide Studies and Prevention, Journal of Democracy, and Journal of Mixed Methods Research. His current book project examines how rebel organizational ideas, goals, and institutions established while fighting the state persist to shape statebuilding and service provision efforts after rebel victory in civil wars.



Review by Jason Lyall, Yale University

Path-breaking in design and execution, Philip Roessler’s Ethnic Politics and State Power in Africa offers a master-class in the careful use of a mixed-method approach to investigate the fraught relationship between power-sharing and political violence in sub-Saharan Africa. The book charts an ambitious course: to demonstrate that civil wars and coups, which are typically studied in isolation, actually share a common taproot in African rulers’ decisions about how (and how much) political power to share with rivals. Drawing inductively on hard-won insights from fieldwork in Sudan, Roessler devises a strategic exclusion theory that he subsequently tests using cross-national quantitative data and within-case qualitative evidence. The result is a compelling mix of theoretical rigor, methodological pluralism, and new empirical results that deserves a wide audience among scholars of political violence.

The book’s central argument can be summarized quickly. Governing against a backdrop of weak institutions and low state capacity, Africa’s rulers are faced with a basic dilemma: they cannot credibly commit not to exploit their office to privilege their own co-ethnics over other ethnic groups. Power-sharing, however, has its own risks. Incorporating non-coethnics raises the prospect that they, in turn, will use their positions to challenge, or exclude completely, the ruler and his ethnic group. Thus the commitment problem has a second side: mistrustful rulers cannot be sure that these ethnic groups are supporting them in a bid to better position themselves for a future challenge.[1] As a result, rulers engage in strategic exclusion, enacting defensive safeguards such as military purges to tamp down rivals’ coup-making capabilities. Doing so, however, substitutes the immediate risk of a coup for that of potential future civil war, for two reasons. First, exclusion creates grievances, increasing the pool of ‘violence specialists’ who could take up arms against the ruler. Second, these coup-proofing measures degrade the military’s counterinsurgency capabilities by dismantling interethnic networks, making it much more difficult for soldiers to acquire information from excluded ethnic groups.[2] Without access to local support or information, militaries resort to indiscriminate violence to blunt insurgent challenges against the capital, though often at the cost of further grievances and counter mobilization. The coup-civil war trap is thus born of interethnic bargaining failure. As Roessler notes, “one of the devastating implications that follows from this theoretical framework is that civil war represents the consequences of a strategic choice by rulers, backed by their co-ethnics, to coup-proof their regimes from their ethnic rivals’’ (xvi).

Roessler employs a nested research design to build and then test his proposed strategic exclusion argument. The empirical investigation opens with a rich theory-building case study of why large-scale civil war did not occur in Sudan’s Darfur during the early 1990s but eventually did a decade later. Over the course of three chapters, Roessler documents how the Sudanese government’s interethnic political networks were initially central to defeating (though not deterring) armed rebellion in Darfur in 1991. Elite interviews reveal, however, that the condominium between Omar al-Bashir, the president of Sudan, and Hassan al-Turabi, the leader of Darfur’s Islamic Movement, subsequently collapsed under the weight of al-Bashir’s fears of an Islamic Movement-led coup against him. Purging his onetime allies from his regime, Bashir succumbed to the logic of the coup-civil war trap, replacing the risk of a coup with that of a civil war. Indeed, Darfur quickly descended into war, fueled by anti-regime grievances and the clumsy brutality of Sudan’s counterinsurgency campaign.

To test the coup-civil war trap’s generalizability, Roessler next turns to quantitative tests of power-sharing’s effects on coup attempts and civil war onset. At the heart of these efforts is the Ethnic Power Relations (EPR) dataset, which measures the ethnic distribution of access to executive political power across a subset of countries in which ethnicity was politically salient during 1946-2005.[3] Several findings emerge. As predicted, ethnic groups included in the government via power-sharing arrangement are more likely to be involved in coups and less likely to initiate civil wars. Excluded groups, by contrast, are less likely to launch coups but are more likely to become embroiled in civil wars, though the differences here are less stark than for included groups (217). The balance of threat capabilities between the government and the opposition (which is treated as a unitary actor in these tests) is thought to drive these results. If the government and opposition are both strong, then power-sharing occurs; if both are weak, however, then regimes are unstable and exclusionary. Power imbalances between the government and the opposition in either direction torpedo power-sharing arrangements as credible commitment problems reemerge. Some stability can arise, then, but only under a narrow set of circumstances. A final model-testing chapter uses process-tracing to explore how these same commitment problems led co-conspirators to junk a power-sharing agreement in the Democratic Republic of the Congo (DRC), sparking successive rounds of preemptive and preventative exclusion in 1997-1998 that caused Africa’s deadliest conflict.

All ambitious books invite further debate, and Ethnic Politics is no exception. I offer four comments here, each centering around the issue of how dynamics are treated in the coup-civil war trap.

First, much of the theory’s dynamism and predictive power arises from what Roessler terms the opposition’s “threat capabilities.’’ Fear of a rival’s capacity for armed action is what justifies strategic exclusion and sharpens the credible commitment dilemma. Despite their centrality, however, we catch only glimpses of the nature of these capabilities and how they are mobilized. Threat capabilities, for example, are narrowly cast as “the mobilizational capacity to capture the capital city if excluded from state power’’ (272). Two fairly crude proxies, namely, the relative size of an ethnic group and the distance from its homeland to the capital, are used in the cross-national analysis to measure a group’s capacity to seize the capital (275). These measures appear best suited for would-be coup plotters seeking political power via capital capture rather than a secessionist-seeking ethnic group, where distance from the capital increases the likelihood of insurgent success. More generally, much of the mobilization process itself occurs off-stage. Ethnic leaders appear to have little difficulty in raising new armies when necessary (94), for example, while followers are implicitly assumed to rally to their leaders in times of trouble. Yet, as the Darfur case study demonstrates, it took nearly eleven years for the DLF’s mobilization capacity to reach levels that might challenge the government. Many insurgent groups also simply die out without ever posing a credible threat.[4] At a minimum, there is considerable variation in the ability of opponents to launch coups and civil wars that is not captured by ethnic demography or spatial concentration.

How rulers assess opponents’ threat capabilities is also somewhat unclear. At times, the theoretical discussion implies that leaders can accurately assess the intentions and capabilities of would-be challengers, especially if threats can be mapped precisely from basic features of each ethnic group’s size and location. At other times, most notably in the theory-building case, leaders are jumping at shadows, forced to make quick decisions under a high degree of uncertainty. Bashir himself is characterized as “paranoid’’ (159, 171), often guided more by rumors than actual threats. Understanding how leaders assess risk and threat is therefore central to determining whether the coup-civil war trap is the tragic but unavoidable consequence of clear-sighted rational leaders making decisions under uncertainty or, alternatively, whether the “trap’’ is the product of misperceptions and contingency and thus potentially avoidable. Whether the trap is born of rational decision or miscalculation in turn shapes our expectations about sequencing of events. While the theory implies that the opposition takes up arms first, in practice many of these civil wars and rebellions appear to be initiated, and sustained, by government decisions to strike first, before threat has actually materialized. Indeed, Bashir’s decision to launch a brutal counterinsurgency campaign in Darfur led to mobilization among victimized groups. Bashir, in other words, was not responding to the risk of civil war; he was creating it (181, 184, 190). In practice, the level of credible threat necessary to justify a leader’s fears of a coup or civil war is left unspecified by strategic exclusion theory.

There is also some ambiguity about whether absolute or relative power capabilities are most important for determining bargaining outcomes. Roessler argues that power-sharing is most likely to occur when both the government and opposition possess `“strong’’ threat capabilities (274). Yet when both sides have “weak’’ threat capabilities, power-sharing is ruled out despite the fact the relative distribution of capabilities is identical across these two scenarios. Similarly, the risks of a coup or civil war are exactly opposite under these two scenarios: a strong/strong power balance is thought to generate a high risk of coup but low risk of civil war, while a weak/weak balance is associated with low coup risk and high civil war risk. Complicating matters are situations of power asymmetry, where strong/weak and weak/strong pairings of government and opposition threat capabilities generate identical predictions of no power-sharing, low coup risk, and high civil war risk. Why absolute power matters in some situations, while relative power concerns predominant in others, is left unclear. Similarly, if civil wars have a higher barrier to entry than coups, as Roessler argues, then it is unclear why the opposition prefers to launch coups, not civil wars, when it has “strong’’ threat capabilities.

Second, the dynamics of intra-ethnic bargaining are mostly absent from the theoretical framework. Surprisingly, a large share of the coup attempts recorded in the book’s dataset (see Table A3.2 in the appendix) were initiated by co-ethnics who enjoyed high political status, not members of excluded or marginalized groups. Roessler does acknowledge the need to deal with intra-ethnic dynamics in the book’s conclusion, suggesting that “coethnics face the same commitment problem as they do when bargaining with ethnic rivals’’ (300). This raises the question of what role ethnicity is actually playing in the theory. Co-ethnics share dense network ties that facilitate information-sharing and reinforce shared preferences over strategies and outcomes. Co-ethnic coups are thus puzzling since they suggest heterogeneous preferences among group members. Ironically, such coup attempts should be especially likely to be detected early and snuffed out since the government would have extensive ties among its co-ethnics. In short, credible commitment problems should be far less severe due to co-ethnic in-group policing that is absent in interethnic bargaining. Given the frequent purging of co-conspirators, however, it appears that co-ethnicity is an imperfect tool for assessing threats, especially if lust for power trumps co-ethnic bonds. Rulers, in other words, likely use other criteria than ethnicity for crafting and implementing their strategies.

Third, the military as an institution is curiously neglected in both the theory and case studies. Frequent reference is made to “violence specialists,’’ but most of the action takes place between (and within) the ruling regime and its opposition. The military itself, however, is pushed to the margins of the strategic exclusion theory, an odd position given that nearly every coup originated from the military (or ex-military officers). In these environments, militaries are themselves engines of uncertainty that can complicate leaders’ assessment of relative risk. Problems below the coup horizon may plague militaries, including poor relations between officers and rank-and-file that in turn complicate specific counterinsurgency operations or raise broader questions of political reliability. It is telling, for example, that one of the most important developments in West and Central African militaries is the rise of mutinies over unequal pay and poor treatment.[5] These mutinies now outnumber coups and civil wars. Similarly, militaries may rely on militia such as the Janjaweed in Sudan to prosecute these wars, creating another set of potential principal-agent problems. The efficacy of certain branches within the military may also help explain why some incumbents are able to safeguard against coup attempts or defeat them when they arise. And, perhaps most importantly, militaries still often retain large numbers of ethnic groups that have been excluded politically, creating opportunities for coup attempts as well as trained recruits for a potential civil war.[6] Adding the military to the institutional context of decision-making would render explicit many of the sources of uncertainty that inform the strategic exclusion model while also helping to explain behavioral outcomes.

Finally, Roessler’s account of the coup-civil war trap focuses almost exclusively on within-country dynamics. This makes sense as an initial starting point, but results in the omission of international factors, especially neighboring states, that can shape the severity of credible commitment problems and the dynamics of escalation. As Roessler himself acknowledges (302), the internal dynamics of the coup-civil war trap in both Sudan and the DRC were heavily influenced by the clandestine, and then overt, intervention of neighboring states. Chad, for example, provided material assistance and a safe rear area for rebels, contributing to their decision to attack Sudanese forces in April 2003. Rwanda’s role in Kabila’s cabinet (and subsequent purge), along with its participation in the Kitona Raid (with Uganda), ensured that the Second Congo War would embroil its neighbors. More generally, former colonial powers such as France likely shape incentives and rewards for would-be challengers as well as the durability of existing regimes (as in Mali during 2012) in ways not captured by the book’s theoretical framework. The United States, too, is increasingly active, with military-to-military partner arrangements throughout Africa, including Niger, Djibouti, and Cameroon. Future work could build upon the strategic exclusion framework to incorporate how external actors, including regional institutions, shape leaders’ assessments of risk, the relative balance of capabilities, the anticipated costs of fighting, and the nature of the postwar settlement.

Taken together, the book’s central argument and sweeping empirical evidence provide a compelling framework for integrating the study of coups and civil wars. While coups and civil wars have actually been in decline since the mid-1990s,[7] Roessler is skeptical that the coup-civil war trap can be transcended entirely, whether in sub-Saharan Africa or elsewhere. As a result, further investigating these dynamics, and possible policy responses that might mitigate them, will remain an important task for scholars and policy-makers in the future.



Review by Hilary Matfess, Yale University

Defying the implicit antiquity requirement for such a classification, Philip Roessler’s 2016 Ethnic Politics and State Power in Africa: The Logic of the Coup-Civil War Trap has already secured its position within the political violence canon. The book asserts that although power-sharing between “rival Big Men in the central government” is a critical aspect of governing, doing so raises the risk of a coup (5). The result is that rulers face a trade-off, in which “excluding rivals weakens their coup-making capabilities but at the cost of increasing the risk of civil war…” (5). Roessler’s identification and description of this trade-off is a powerful rejection of both the weak-state hypothesis and the ‘ancient hatreds’ accounting for civil conflict. The book is a tour de force that brings together extensive qualitative fieldwork, theory-building and theory-testing case studies, and quantitative analysis to bring forth a novel description of governance and conflict in Africa.

Roessler asserts that the book is an exercise in developing and testing five theories: that “ethnic power-sharing is a key source of societal peace in postcolonial Africa;” that the same power-sharing is accompanied by a risk of coup as it raises “rivals’ coup making capabilities;”  that “ethnopolitical exclusion mitigates coup risk but at the cost of increasing the risk of civil war,” which is the natural synthesis of the previous two claims; that “interethnic alliances increase a set of groups’ capabilities to capture power but then confront the co-conspirators with a severe commitment problem after victory that often ends with regime factionalization and civil war;” and that peace will be the result of strategic balancing and that durable power-sharing emerges when the threat of mobilization against one another prompts groups to share govern together (20-21).

From the five broad claims Roessler sets out to test, a variety of additional observations and sub-theories emerge. Among these is a concept that Roessler refers to as “cooperative counter-insurgency.” He notes that “the denser and more extensive political networks” that emerge from power-sharing “reduce the feasibility of armed rebellion” by making the government less dependent on indiscriminate violence and undermines support for the insurgency, leaving “the rebels like fish out of water” (146).  Regarding coups, Roessler delineates the logic of mounting a coup in the face of potential exclusion from a power-sharing arrangement. Roessler notes that a pre-emptive coup is a logical response for those fearful of exclusion from political power, as they recognize that the collective mobilization costs of a coup are much lower than the strain of organizing and sustaining the sort of force necessary to mount a civil war. He also notes that, in the aftermath of a successful coup that removed the previous regime, “rulers are significantly more likely to purge their co-conspirators from the central government than other power-sharing holders,” helping to shed light on the governance characteristics of regimes that come to power through coups (242).

Roessler’s work proceeds in four parts. After describing the puzzle and addressing the extant literature on ethno-political exclusion and civil war in Africa, he then uses a Darfur case study to build his theory. The Darfur case study itself could well be regarded as two case studies, as Roessler leverages temporal differences to analyze an instance in which the threat of civil war was successfully mitigated, as well as the conditions under which the Darfur Civil War emerged roughly a decade later, after Omar al-Bashir, the President of Sudan, dismantled the political network of the National Islamic Front (NIF) as part of a coup-proofing strategy. The third section of the book tests the model, first by using the Ethnic Power Relations Dataset and then through a test of the model using Africa’s Great War in the Great Lakes Region.  The book’s final section discusses how some countries have avoided the “vicious exclusion-conflict cycle” (269), delineates continent wide normative changes (such as the African Union’s [AU] rejection of coup-based transitions of power), and discusses how this theory can travel outside of Africa.

Despite the layers of analysis and myriad insights conveyed, Roessler’s thesis is almost electric in its clarity. Though the qualitative case studies dive into the messy entanglements and fluid loyalties of elite power brokers, Roessler’s narrative is deft and ably carries along even readers who are un-initiated in the socio-political landscapes of the countries being discussed. Roessler’s careful parsing of why ethnicity is the dominant socio-political cleavage in Africa and willingness to consider the implications of his thesis for other regions defined by different fault lines allows the framework to travel convincingly and widens the book’s audience.

While undoubtedly one of the most compelling and grounded contributions to the field today, Roessler’s work could benefit from a more thorough probing of the extent to which the ‘coup-civil war trap’ characterizes the options that African leaders face. Given the vast ground already covered by this book, many of the following criticisms are perhaps better considered avenues for future research than shortcomings of the contribution at hand.

I identify four areas where the argument could have been furthered or strengthened: through the discussion of alternatives versions of coup-proofing, an unpacking of the role of consociationalism in reducing uncertainty in multi-ethnic governments, an examination of the role of cross-cutting cleavages in reducing ethnic salience and the existence of non-ethnically motivated coups, and more attention to the organizational hurdles facing would-be coup plotters or insurgent leaders.

Perhaps most obviously, Roessler’s work leaves under-addressed the other options facing leaders who fear ascendant power-brokers or challenges to their power, including the assassination of individual challengers or the imprisonment of factional leadership. Such behavior would reduce both the risk of a successfully organized coup and make it more difficult for the would-be-coup plotters to organize themselves into a force capable of exerting a military challenge to the state. There are a number of examples of such behavior by African autocrats; consider the prosecution on charges of corruption of former members of the ruling Ethiopian People’s Revolutionary Democratic Front (EPRDF) who challenged then-Prime Minister Meles Zenawi in the early 2000s or Paul Biya, the President of Cameroon, who has imprisoned Anglophone separatist leadership in response to their bid for independence in recent years.[8] These two forms of regime fortification are largely overlooked in Roessler’s account, despite characterizing a number of African political landscapes.

Additionally, Roessler is conspicuously quiet about consociationalism, despite an analytical emphasis on the role of power-sharing. It would be interesting to see to what extent Roessler believes that consociational institutions or arrangements can create the conditions for durable power-sharing by reducing “the uncertainty that comes from incorporating ethnic rivals” (98). It is possible that with enough information, there are incentives that produce durable power-sharing arrangements that are not predicated on the threat of civil war or undermined by the potential for a coup. Recent decades have seen a number of multi-laterally supported constitutional reform processes and peace settlements resulting in the adoption of consociational constitutional designs.[9] A discussion of the nature and viability of such internationally enforced, formal power-sharing institutions would complement Roessler’s discussion of the AU’s anti-coup norm (310) and would contribute to a line of research in which international preferences shape domestic politics.

Relatedly, another aspect of Roessler’s analysis that seems especially ripe for greater nuance is in the earnestness of power-sharing and the possibility for ethnic exclusion from non-political, but politicized realms. It is possible, for example, for there to be equitable ethnic power-sharing in a toothless legislature, or for an executive to have an ethnically balanced cabinet in which ethnic rivals are given inconsequential portfolios or scant budgets.[10] The book would have also benefited from a discussion of the challenges levied by co-ethnics and the dynamics of in-group power jockeying. As Kate Baldwin noted in her review of this book, “more than 40% of successful coups in sub-Saharan Africa are orchestrated in part by monopoly or dominant groups (see Table A3.2 in the book’s appendix), and many of the ethnically homogenous countries excluded from Roessler’s analysis (Burkina Faso, Djibouti, Lesotho) have fallen into coup traps.”[11] Even in multi-ethnic coalitions, this dynamic is at play; the power-shuffling within the Tigrayan wing of the EPRDF in Ethiopia under Meles Zenawi is a testament to the difficulty of managing co-ethnic challenges to power.

Relatedly, it would have been interesting for Roessler to speak to the ways in which multiple or cross-cutting cleavages can contribute to the likelihood of violence. He is careful to describe ethnicity as the prominent, not the pre-determined nor the only, mechanism of identification and organization; he asserts that “…upon independence, ethnic institutions were much stronger than other coordinating mechanisms, such as political parties and became the basis of political competition, crowding out other institutions” (51), while recognizing the fluidity of ethnic identification (65) and potential for the patterns of identification to shift in the future. Despite the recognition that ethnicity is one of many potential identities, the book does not unpack the consequences of cross-cutting cleavages on the process of power-sharing or, in the face of a failed power-sharing endeavor, the implications that such cleavages have on the collective action required for mobilization into war. Under Roessler’s logic, it may be easier to “strategic[ally] exclude” (28, passim) a rival ethnic group that is beset by cross-cutting cleavages (or for whom the elites are otherwise less able to utilize the “politics of fear” (96) to secure support for their agenda), but this possibility is not given due attention in this text. In his discussion of the Islamic Movement in Darfur Roessler discusses how the cross-cutting cleavage of ethnicity contributed to the decline of a religiously mobilized organization (126). Accordingly, despite the colonial, structural roots of ethnic mobilization’s strength, ethno-regional mobilization could be undermined by alternative social identities. Discussing the prospects of this emerging pattern, particularly in light of the increasing relevance of armed mobilization into multi-ethnic terrorist organizations, could have contributed to a more nuanced understanding of the future of violence and governance in Africa.  Furthermore, this discussion would have complemented Roessler’s discussion of the role of information asymmetry in shaping inter-ethnic relations and power-sharing.

Finally, on a point that is unrelated to a greater nuancing of the ubiquity of the coup-civil war trap, I would have appreciated a discussion by Roessler of the organizational challenges of waging a civil conflict or organizing a coup. While he recognizes the difficulty of mobilizing and maintaining an armed challenge to the state and briefly alludes to the use of the “politics of fear” to generate and maintain support among co-ethnics (96), the motivations and preferences of those on the front lines of civil conflict in Africa are largely treated as a black box. Similarly, he cites Naunihal Singh’s work on what determines a successful coup, without much discussion of how potential coup-plotters assess their odds of a successful endeavor.[12]

It is a testament to the quality of Roessler’s work that the most forceful critique that I can muster is that I would like to read more of his work. The book is already brimming with asides, allusions to relevant cases not fully delineated in the book, and ‘pop out boxes’ elaborating on concepts that do not necessarily further the analysis but complement the reader’s understanding, suggesting that this is a rich line of inquiry for Roessler and others in the field to develop. Ethnic Politics and State Power in Africa: The Logic of the Coup-Civil War Trap is a powerful updating of the literature on ethnicity in Africa and a valuable contribution to the literature on institutions, incentives, and violence.



Review by Kai M. Thaler, University of California, Santa Barbara

Why, in the postcolonial period, has there been such wide variation in levels of conflict, peace, and stability across Africa’s independent countries? Some African states have been bedeviled by large-scale civil wars and coups (e.g. Democratic Republic of the Congo, Liberia, Sudan, Uganda); some are more prone to civil wars alone (e.g. Angola, Eritrea, Mozambique); some have primarily been afflicted by coups and coup attempts (e.g. Egypt, Equatorial Guinea, Ghana); and some have been more reliably free of armed conflict, albeit with periodic crises or separatist tensions (e.g. Namibia, Senegal, Zambia). To explain these divergent outcomes, Philip Roessler argues that we must look at the ethnic geography of African countries and how it shapes competition between different ‘Big Men’ seeking power and resources. In Ethnic Politics and State Power in Africa: The Logic of the Coup-Civil War Trap, Roessler uses interviews, field observations, primary and secondary sources from Sudan, South Sudan, and the Democratic Republic of the Congo (DRC), and results from cross-national statistical analyses to build and test a theory. He argues that the size and political-military capabilities of ethnic groups shape the competition for state power, leading to ethno-political exclusion and civil wars at one extreme, or power-sharing and an increased coup risk at the other.

Roessler departs from the assumption that the withdrawal of colonial regimes, whether under violent or more peaceful circumstances, left a power void at the center, since colonizers had “hegemonic control” (17) of the central state and had created few formal institutions. In the absence of institutional ties to bind the citizens of newly-independent nations, political competition tended to fall along ethnic lines, as ethnic ‘violence specialists’[13] sought to advance the fortunes of themselves and their social groups. The Organization of African Unity (OAU) failed to undercut this anarchic situation, instead adopting the principle that control of the capital city was sufficient for international sovereign recognition, an approach that was mirrored globally.[14] This increased the stakes for controlling the capital and the resources and rents it made available.

For rulers of ethnically divided states, this leads to a dilemma. Including powerful ethnic rivals in the government might help maintain peace and stability and widens the government’s networks to monitor and sanction the population, but it creates a commitment problem in which the ruler cannot be sure that the included rivals would not use their influence to stage a coup and seize more power for themselves. This commitment problem even afflicts ‘co-conspirators’ who seize power together through coups or civil war, but then fall prey to the coup-civil war trap themselves. Excluding rivals, however, both builds resentment over marginalization and undercuts the government’s monitoring and enforcement capabilities among the excluded group,[15] limiting prospects for cooperative efforts to stop nascent insurgencies using selective violence and increasing the risk of civil war. Exclusion under uncertainty is only the preferable option for rulers up to a point. The greater the “threat capabilities” (19) of a group to reclaim power through rebellion if it is excluded, the more likely it should be to be included in a power-sharing arrangement, leading to a reluctant acceptance that power may occasionally be traded between groups through coups. Individual leaders might not like this arrangement, as a coup would likely sideline them politically or potentially kill them, but they are constrained because they are operating in a two-level game (75): they are bargaining with ethnic rivals, yet also accountable to their own ethnic constituency, whose members would bear the costs of a civil war (285-286, fn.19).

The argument and the evidence mustered in support of it by Roessler are powerful and compelling, and the book advances the study of civil conflict not only in Africa, but globally, by showing how civil wars and coups may be two sides of the same coin, though their examination is rarely integrated in this manner.[16] Roessler developed the theory inductively through a fieldwork-based case study of Sudan’s civil wars in Darfur from 1989-1991 and in the early 2000s, demonstrating how a purge in Khartoum in the 1990s undercut the government’s networks in Darfur. This meant that the cooperative counterinsurgency that in the first civil war limited violence and led to relatively quick resolution was impossible in the second civil war, in which the government used mass violence against the Fur and their allies. Drawing on the theoretical insights from this thorough within-case comparison, Roessler conducts a nested analysis (see 22), engaging in large-N theory testing of his contentions about how ethnic powersharing or exclusion affect coup and civil war risk. Roessler uses the Ethnic Power Relations (EPR) dataset (version 1.0),[17] finding support for his argument.

Roessler then returns to case study analysis for a fieldwork-based case study of the civil wars in the Democratic Republic of the Congo from the 1990s through 2000s, arguing that they can be explained as products of the coup-civil war trap, and that the Second Congo War was a co-conspirator civil war following the breakdown of relations between rebel leader-turned-president Laurent-Désiré Kabila and his Rwandan sponsors. Finally, Roessler returns to the EPR data to quantitatively test the threat capabilities theory of ethnic power-sharing to understand how and why some states have been able to avoid large-scale civil war. He finds that where two or more ethnic groups exist with durably high threat capabilities, power-sharing is more likely than exclusion, potentially leading to more coups, but not civil wars. This is a tour de force of mixed-methods research, using rich original qualitative data and quantitative evidence to inform each other throughout the research sequence in order to generate new insights and a more complete picture of the social reality of ethnic politics and conflict in Africa.

As with any study of this scope and ambition, there are some missed opportunities and slight missteps. First, while it does well in situating Roessler’s argument within the realist international relations literature, the contention that the end of colonialism produced anarchy in former colonies requires further testing and justification. Roessler does discuss Donald Horowitz’s idea of “split domination” (222-226), in which “one ethnic group controls the armed forces and another dominates the civilian regime” as a result of colonial policies.[18] Roessler also includes French colonial history as a control variable in his final set of models in Chapter 10. Colonial powers, however, exercised far different forms and strength of rule in comparison to one another and across their own sets of colonies, and while in some cases the end of colonialism brought relative anarchy, in others, “the colonial state, where skillfully managed, could still retain the capacity to orchestrate the process of transferring power.”[19] Since variation in colonial power may have enduring effects on postcolonial governance and ethnic conflict,[20] more attention to the different colonial backgrounds of states would have strengthened Roessler’s cross-national analyses.

Another alternative explanation that receives little attention from Roessler is ideology. While I am admittedly predisposed to interest in how ideology may supersede or interact with structural factors, it seems that ideology could overcome the coup-civil war trap in Africa by rendering ethnicity less salient and restructuring politics around a shared national, multiethnic identity. Scott Straus, for example, discusses how in some cases, post-independence African leaders constructed a multiethnic ideology of shared nationhood, preventing the ethnic mass violence that plagued other countries on the continent.[21] Left-wing liberation movements sought to create multiethnic, multiracial crosscutting coalitions that rejected ethnic politics in favor of programmatic socioeconomic goals. The Frente de Libertação de Moçambique (Frelimo) in Mozambique, for example, purged early leaders who advocated more ethnocentric goals.[22] While the northern origins of these purged leaders may have contributed over time to a view of Frelimo as biased in favor of southern groups’ interests, civil war came to independent Mozambique not because of ethnic competition for state power at the center, but instead due to international meddling by Rhodesia and South Africa.

Coup leaders, too, such as Thomas Sankara in Burkina Faso and Jerry Rawlings in Ghana, have sought to supplant ethnicity with programmatic national ideologies. Though a coup cut short Sankara’s rule, Rawlings had relative success containing ethnic tensions and Ghana’s history of coups to shepherd the country towards democracy.[23] This also suggests that the emergence of co-conspirator conflict in Liberia after the country’s 1980 coup, a shadow case for Roessler (238-241), was not inevitable. I am largely sympathetic to Roessler’s presentation of the reasons for the coup attempt by Thomas Quiwonkpa in 1985 against his co-conspirator, Samuel Doe, which helped spark Liberia’s subsequent civil war.[24] Many Liberians thought and hoped that the 1980 coup would end the marginalization of all of Liberia’s indigenous ethnic groups, and early on, in fact, “the regime vacillated between an attempt to establish a populist program of development, on the one hand, and a retaliatory indigenous hegemony with a Krahn core, on the other.”[25] If Doe had curbed his personal ambitions and favoritism toward his own Krahn ethnic group, he could have worked to construct a non-ethnic nationalist ideology, as Rawlings and Sankara soon did. Absent ethnic cleavages at the center, however, it is of course quite possible that ideological or partisan identities may harden, leading to a similar coup-civil war trap with intergroup competition viewed as zero-sum and these non-ethnic identities justifying violence.[26]

Related to ideology, the goals of members of a rebel or coup coalition may affect the likelihood of a subsequent co-conspirator conflict. It may be a rare case, but where one coalition member seeks central state control and another seeks secession or autonomy, civil conflict can be avoided if the first group accepts the goals of the second. This occurred with the Eritrean People’s Liberation Front and the Tigrayan People’s Liberation Front, who cooperated in toppling Ethiopia’s Derg regime and gaining Eritrea’s independence[27]—though the subsequent interstate war between Eritrea and Ethiopia could be viewed through the co-conspirator conflict lens.

Finally, the case study of the Second Congo War, or ‘Africa’s Great War,’ provides important and convincing new evidence about the origins of the war and the decision-making of Kabila government. The case, however, is not an ideal choice for testing Roessler’s stated theory. First, the Alliance des Forces Démocratiques pour la Libération du Congo-Zaïre (AFDL) was, as Roessler describes it, an especially fragmented, coalitional organization, “with little structure or political program” (250), and so intra-organizational tensions may be overdetermined in comparison to more cohesive, established organizations. Second, the role of the Rwandan security establishment as Laurent-Désiré Kabila’s co-conspirators, rather than as a purely domestic partnership, makes this a good test case for a coup-civil war trap in cases of foreign intervention leading to regime change, but not necessarily with regard to rebel victory or coup regimes more broadly. One can argue that in an era of increasingly internationalized civil wars,[28] it is all the more important to examine a case with the Second Congo War’s dynamics, but Roessler does not take this tack.

Roessler’s conclusion offers a strong summation of his arguments and findings, as well as a brief examination of the coup-civil war trap’s applicability in the Middle East and the prospects for overcoming the coup-civil war trap in Africa. Coups, Roessler argues, are still important issues of study despite their decline in Africa and globally since the 1990s (308), though this point can be strengthened with reference to recent coups and coup attempts outside the continent in countries like Honduras, Thailand, and, more recently, Turkey.[29]

Roessler writes that one possible sign of hope of overcoming the coup-civil war trap lies in how “the changing basis of social mobilization, urbanization, technological diffusion, and generational change are strengthening interethnic social ties and bases of trust, while leading an increasing number of individuals to reject political appeals along ethnic lines” (309). I wish I could share such optimism, but this seems like modernization theory redux, (i.e. urbanization and technological change leading greater crosscutting affinities and democratization). Technological diffusion can cut both ways, as we see in Facebook’s role in stoking ethnic tensions and violence in Germany, Myanmar, and Sri Lanka.[30] Meanwhile, in a country like South Africa, while political party affiliations are less strictly racially-divided than in the past, this has not meant the end of racial tensions among South African citizens, nor an end to interethnic violence, which has turned against African migrants. Roessler also highlights the changing stance of the OAU and its successor, the African Union (AU), regarding capital city control as a positive factor for stability since it reduces incentives for coups or rebellions. He rightly juxtaposes this shift’s limits, though: African states acquiesced to the 2013 coup in Egypt, and the norm change benefits incumbents, whom the AU has shown little willingness to sanction for abusive behavior (316-320), as the AU’s ineffectiveness in counteracting Burundian President Pierre Nkurunziza has demonstrated.[31]

For those seeking to take up Roessler’s call to further examine how intra-ethnic bargaining plays into interethnic conflict or stability dynamics, useful starts would be the ideas of in-group policing and crosscutting associational ties from the ethnic conflict literature.[32] The coup-civil war trap in states where politics are characterized primarily by non-ethnic political cleavages also deserves further study, potentially pushing Roessler’s theory to Latin America, where strong left-right ideological polarization has been common among both weaker, poorer states, and wealthier, more powerful ones. The potential effect of cross-border co-ethnics should also be further explored, since in the Darfur case, Chadian troops shirked orders to protect their co-ethnics, while cross-border ethnic ties were heavily implicated in the spillover of civil war between Rwanda and the DRC. There is also room for replicating and expanding Roessler’s quantitative tests with new data. Roessler laments the lack of adequate cross-national data on the ethnic composition of militaries (207), but new data on the ethnic makeup of militaries in Africa and the Middle East, as well as improved data on security force counterbalancing, enable more refined testing of Roessler’s theory.[33]

Roessler has done a great service to scholars of African politics, civil wars, coups and coup-proofing, ethnic politics, and regimes. The broad relevance of Ethnic Politics and State Power in Africa speaks to both its ambition and to the potential power of its theory. The impressive fieldwork that grounds Roessler’s theory and his case studies, and Roessler’s openness about his research process are exemplary. Describing in detail his inductive theory development and the ways in which he gathered qualitative data (317-320), Roessler provides a timely reminder of the important role induction can play in positivist research, as well as a useful discussion of the difficulties of interviewing key political players in contentious settings. The mixed-methods approach offers insights that would likely have been missed employing comparative cases studies or quantitative analyses alone. Roessler also makes a strong case for increased focus in weak state contexts on informal institutions and the roles of political entrepreneurs and their networks (11). The breadth and depth of the theory building, evidence, and analysis provided by Ethnic Politics and State Power in Africa should ensure that it endures as a key work for scholars of politics in Africa and beyond.



Response by Philip Roessler, College of William & Mary

I am grateful to Jason Lyall, Hilary Matfess, and Kai Thaler for reviewing my book and to Elizabeth Schmidt for introducing the roundetable. Few things are more gratifying in academia than seeing colleagues—amidst their own busy and exciting research agendas—taking the time to generously and critically engage with one’s ideas. Their thoughtful and constructive reviews raise a number of incisive questions and also lay out a research agenda for those interested in addressing some of the gaps in my model and extending it even further.

The argument at the core of Ethnic Politics and State Power in Africa arose from my having spent months in Sudan undertaking a structured comparison of two rebellions in Darfur—one the government of Sudan effectively contained in the early 1990s and the other that erupted into full-scale civil war following Khartoum’s brutal and indiscriminate counterinsurgency campaign. Process-tracing the source of these divergent conflict outcomes pointed to the role that political networks play in maintaining societal control in weak states. In the early 1990s the Sudan government leveraged its strong Islamist network in Darfur to prevent civil war. But quite strikingly, Sudan’s President, Omar al-Bashir, subsequently dismantled this network—undercutting his government’s countermobilizational capabilities. Why would Bashir choose such a costly political strategy?

Previous research, such as that of Bruce Buena de Mesquita and Alistair Smith,[34] emphasized the economic logic of building small-winning coalitions, but my research pointed to its political logic. In this instance, Bashir was responding to an escalating power-struggle with Hassan al-Turabi, the leader of Sudan’s Islamic Movement who had earlier handpicked Bashir to be president. In the face of this rising threat, Bashir purged Turabi and Turabi loyalists. Though this weakened the regime’s political support from western Sudan, especially Darfur, Bashir was willing to trade off what he perceived as the clear and present danger of a coup d’état for a future and uncertain risk of civil war.

This set of strategic interactions suggested something quite profound. In weak states, like Sudan, societal peace hinges on powerbrokers embedded in different political networks sharing power. But in the absence of strong political institutions, the challenge these powerful actors face is how to credibly commit to share regime control. Purging one’s rival manages this commitment problem but in doing so it merely displaces the threat, substituting coup risk for civil war risk. Thus, in weak, ethnically-divided states the commitment problem gives rise to a coup-civil war trap.

The reviewers raise several important queries. Here I focus on two core issues regarding how groups’ threat capabilities structure the coup-civil war trap and the dynamics of intra-ethnic bargaining before turning my attention to a broader set of issues that the reviewers advance and which others have started to address in exciting and important ways.

In the penultimate chapter, building on earlier work with Dave Ohls,[35] I show how the balance of threat capabilities between the ruling group and a given rival group mediates bargaining outcomes. Power sharing is more likely to be self-enforcing when both groups possess the capabilities (as measured by distance to the capital city and group size) to credibly threaten the capital if excluded from power. Yet, when there are asymmetrical threat capabilities or symmetrical low threat capabilities, power sharing is more likely to break down. In response, Lyall questions “whether absolute or relative power capabilities are most important for determining bargaining outcomes.” To clarify, what matters is the interaction between the two—rival groups must possess absolutely high, relatively equal capabilities. The intuition is that groups must leverage their threat capabilities to hold others accountable should their rivals move to unilaterally usurp power. Thus, though weak groups may possess relatively equal capabilities, each group knows that if it were to be displaced from the state, the possibility of returning to power would be nigh impossible. Under these conditions, ethno-political bargaining is more likely to resemble a one-shot game, and we would expect uncertainty to be high.

We might expect bargaining between weak and strong groups to be less uncertain; the strong group has the advantage and should possess the capabilities to hold to account weak groups if they try to monopolize power. While weak groups know this threat is real, they also know that their own weak capabilities leave them with little recourse to hold accountable the strong group if it tried to monopolize power. Confronted with this strategic dilemma—and fearing permanent displacement from the central government—weak groups have a strong incentive to cling to power at all cost. Minority group regimes throughout the history of South Africa, Rwanda, and Burundi represent paradigmatic examples.

In the book, I describe ethno-political bargaining as a two-level game (see 75-76), but most of my focus is on inter-ethnic strategic interactions (Level 2) as these are more likely to lead to civil war. Noting the prevalence of intra-ethnic coups in Africa, especially among dominant groups, Lyall and Matfess question how these dynamics fit the model and the implications for “what role ethnicity is actually playing in the theory” (Lyall). While ethnic divisions compound the commitment problem at the heart of the coup-civil war trap, they do not cause it. Ultimately, it is rooted in a fundamental political problem that can also arise among co-ethnics—how can factions or groups who have to use the threat of force, rather than formal institutional mechanisms, to gain power, credibly commit not to exploit their violent capabilities to lock-in even more power. What is striking is that while intra-ethnic bargaining also confronts factions with a commitment problem that can end in coups, intra-ethnic civil war is much less common. This divergence from the coup-civil war trap is striking, and is a fascinating topic for future research. One possibility is that while dense co-ethnic networks are not necessarily sufficient to prevent coups, they do avert the outbreak of large-scale political violence.[36] Perhaps, as Thaler notes, in-group policing helps to prevent the escalation of large-scale political violence among co-ethnics. Another striking stylized fact is that intra-ethnic coups primarily occur in ethnically-dominant regimes. In other words, infighting among co-ethnics is higher when other groups have been excluded from the central government. Why this is the case—and how bargaining between ethnic groups affects political bargaining within groups and vice versa—is an important topic for future research.

The reviewers raise a number of other important queries, such as the role of the military as an institution; the availability of other coup-proofing strategies; the importance of external dynamics; and the mediating effects of cross-cutting cleavages. Space constraints prevent me from fully addressing each; indeed, each topic could generate a book on its own. Instead, to conclude, let me highlight some of the promising scholarship that others are undertaking on each dimension.

In my model, I primarily focus on ethno-political networks that cut across different formal organizations, including the military. Lyall rightfully makes the case for more attention to be given to the military as an institution. We would expect the institutional structure of the military to be critical for how the coup-civil war trap plays out. If the military is more corporatist and professional—and thus more insulated from competition between rival political networks—we would expect ethno-political bargaining to be less explosive. Competing groups still face the commitment problem and uncertainty about potential future shifts in the distribution of power that may lead to violence but we would expect this to be less likely to lead to full-blown civil war. Kenya seems to be such a case; following the 2002 election, the breakdown of powersharing between Mwai Kibaki, the new President, and Raila Odinga, who was supposed to be made Prime Minister after constitutional reforms, led to tension and subsequent electoral violence but the corporatist structure of the military ensured this did not lead to civil war. This points to the importance of understanding variation in military institutional structures and the degree to which the military has been ethnicized as a key driver of the coup-civil war trap. Important work on this front has been undertaken by Kristen Harkness.[37] She shows how decolonization proved to be a critical juncture in the institutional structure of the military. In those states in which new rulers built ethnic armies, they undermined the potential role the military could play as a neutral third-party enforcer and keep the state out of the coup-civil war trap. Why rulers choose to build ethnic armies or not remains unanswered, however, and much more work needs to be done on this critical institution. Lyall lays out some fruitful ideas in his review.

Matfess in her review asks whether the mode of political exclusion matters for subsequent conflict escalation. Can rulers manage the commitment problem and coup-proof their regimes without provoking civil war? Many attributed the durability of Mobutu Sese Seko’s 32-year presidency of Zaire (now the Democratic Republic of the Congo) to his revolving-door style of politics, in which rivals were regularly pushed out of government to undercut their ability to accumulate too much power, but also regularly reappointed to reduce their incentives to rebel.[38] Fiona Shen-Bayh distinguishes between judicial and extrajudicial repression and purges. Whereas the latter is likely to fuel conflict escalation, she contends a judicial strategy will have the opposite effect, as it legitimizes the repression and the public nature of the punishment increases shared beliefs about the incumbent’s authority and power.[39] This points to a potential judicial solution to the commitment problem—yet, rulers tend to eschew such a strategy among regime outsiders, fueling cycles of conflict.

Matfess also rightly raises the point about the mediating effects that inter-ethnic institutions or cross-cutting cleavages may have on the coup-civil war trap, especially by helping to underwrite powersharing. In a fascinating analysis, Nils Bormann shows that cross-cutting cleavages lead rulers to build larger political coalitions than they might otherwise have done in order to internalize these cleavages and prevent rivals from leveraging them to build an alternative ruling coalition.[40]

The most prominent inter-ethnic institution post-independence African rulers sought to employ to manage ethnic divisions was a dominant political party. Many scholars attribute the initial stability of Tanzania, Kenya, and Côte d’Ivoire to their rulers’ creation and maintenance of hegemonic political parties that effectively managed political uncertainty and inter-ethnic conflicts, thus avoiding the coup-civil war trap.[41] Party institutionalization may account for the emergence of Africa’s two equilibriums that I discuss in the book. But what accounts for the fact that in some countries rulers were able to build durable inter-ethnic parties? In another fascinating analysis, Jack Paine suggests the answer may lie in colonial state-building and the legacies of pre-colonial states. In those colonial polities that subsumed pre-colonial states, like Uganda and Buganda, inter-ethnic institutionalization was less likely, as party-builders feared the future takeover of the party by the centralized group. The result was the ethnicization of politics and the onset of the coup-civil war trap.[42]

Ethno-political bargaining does not play out in a sovereign vacuum. As the Congo case in my book shows, and as Harry Verhoeven and I further developed in a book-length treatment, the coup-civil war trap in Congo in the late 1990s had significant regional implications.[43] All three reviewers call for bringing international politics into the dynamics of ethno-political bargaining—something with which I definitely agree. Andrew Boutton’s recent “Coup-proofing in the Shadow of Intervention: Alliances, Moral Hazard, and Violence in Authoritarian Regimes” is an excellent contribution on this front. He convincingly shows how defensive alliances from external states undermine power sharing by increasing rulers’ willingness to pursue costly coup-proofing strategies, such as purges of rival elites, anticipating they can draw on external support to weather the potential ensuing instability.[44]

It is heartening to see researchers build on and extend the coup-civil war theoretical framework in such promising ways. In doing so, I hope that we will move closer to more effective policy solutions to address this pernicious governance challenge. In the conclusion of my book, I offer an assessment of how urbanization, changing bases of social mobilization, technological changes and norm development may be changing the game of politics. Thaler in his review is more pessimistic—and explains how some of these factors may cut the other way and increase ethnic politicization. His point is well taken, and this strikes me as one of the most important avenues for future research: to better understand the degree to which these social, technological, and economic forces are transforming politics or reinforcing prevailing power configurations, institutions, and state-society relations across countries in Africa.

For those interested in these issues, a great starting point would be to build off of the insightful ideas and suggestions Lyall, Matfess, Thaler and Schmidt advance in this roundtable. Thanks to H-Diplo’s International Security Studies Forum (ISSF), especially Tom Maddux, for facilitating and hosting this constructive exchange.



[1] James Fearon, “Rationalist Explanations for War,” International Organization 49 (1995): 379-414.

[2] Jason Lyall, “Are Co-Ethnics More Effective Counterinsurgents? Evidence from the Second Chechen War,” American Political Science Review 104 (2010): 1-20.

[3] Andreas Wimmer, Lars-Erik Cederman and Brian Min, “Ethnic Politics and Armed Conflict: A Configurational Analysis of a New Global Data Set,” American Sociological Review 74 (2009): 316-337.

[4] Janet Lewis, “How Does Ethnic Rebellion Start?” Comparative Political Studies 50:10 (2017): 1420-1450.

[5] Maggie Dwyer, 2017. Soldiers in Revolt: Army Mutinies in Africa (Oxford: Oxford University Press, 2017).

[6] Donald Horowitz, Ethnic Groups in Conflict (Berkeley: University of California Press, 1985).

[7] Jonathan Powell and Clayton Thyne, “Global Instances of Coups from 1950-Present,” Journal of Peace Research 48:2 (2011): 249-259.

[8] Paulos Milkias, “Ethiopia, the TPLF, and the Roots of the 2001 Political Tremor,” Northeast African Studies 10:2 (2003): 13-66; “Cameroon Anglophone Leader Gets 15 Years’ Jail for ‘Terrorism’,” AFP, 26 May 2018.

[9] See, for example:  René Lemarchand, “Consociationalism and Power Sharing in Africa: Rwanda, Burundi, and the Democratic Republic of the Congo 2006.” African Affairs 106:422, (January 2007): 1-20,

[10] It’s also possible that ethnic domination is felt not through political dominance, but through the exclusion from certain economic opportunities or caches of social capital. Ethnic dominance (real or perceived) in party-owned companies, sometimes referred to as “party-statals,” can also be a source of frustration. For discussions of this phenomenon generally, see John W. Harbeson and Donald Ro. Africa in World Politics: Constructing Political and Economic Order. Routledge, 2018 and Laura Mann and Marie Berry, “Understanding the political motivations that shape Rwanda’s emergent developmental state,” New Political Economy 21:1 (2016): 119-144.

[11] Kate Baldwin, “Ethnic Politics and State Power in Africa: The Logic of the Coup–Civil War Trap,” Perspectives on Politics 16:1 (2018): 156-157.

[12] Naunihal Singh, Seizing Power: The Strategic Logic of Military Coups (Baltimore: Johns Hopkins University Press, 2014).

[13] Robert Bates, Avner Greif, and Smita Singh, “Organizing Violence,” Journal of Conflict Resolution 46:5 (2002): 599-628; Douglass Cecil North, John Joseph Wallis, and Barry R. Weingast, Violence and Social Orders: A Conceptual Framework for Interpreting Recorded Human History (Cambridge: Cambridge University Press, 2009).

[14] Marika Landau-Wells, “High Stakes and Low Bars: How International Recognition Shapes the Conduct of Civil Wars,” International Security 43:1 (2018): 100-137.

[15] Roessler uses the term “emasculated” (for example, on 24, 311) to describe this loss of control. This term, even when it is used in a benign sense, as it is here, is one scholars should avoid due to its gendered connotations of masculinity as necessitating a certain form of power and domination over others.

[16] See also Jonathan M. Powell, “Trading Coups for Civil War,” African Security Review 23:4 (2014): 329-338. Powell looks not at political exclusion, as Roessler does, but at marginalization in the security forces, finding that coup-proofing efforts by rulers that fragment security forces increase the likelihood of civil war onset.

[17] Andreas Wimmer, Lars-Erik Cederman, and Brian Min, “Ethnic Politics and Armed Conflict: A Configurational Analysis of a New Global Dataset,” American Sociological Review 74:1 (2009): 316-337.

[18] Donald L. Horowitz, Ethnic Groups in Conflict (Berkeley: University of California Press, 1985), 457.

[19] Crawford Young, The African Colonial State in Comparative Perspective (New Haven: Yale University Press, 1994), 191. See also Peter Richens, “The Economic Legacies of the ‘Thin White Line’: Indirect Rule and the Comparative Development of Sub-Saharan Africa,” African Economic History 37 (2009): 33-102.

[20] Pierre Englebert, “Pre-Colonial Institutions, Post-Colonial States, and Economic Development in Tropical Africa,” Political Research Quarterly 53:1 (2000): 7-36; Robert Blanton, T. David Mason, and Brian Athow, “Colonial Style and Post-Colonial Ethnic Conflict in Africa,” Journal of Peace Research 38:4 (2001): 473-491.

[21] Scott Straus, Making and Unmaking Nations: War, Leadership, and Genocide in Modern Africa (Ithaca: Cornell University Press, 2015).

[22] Carrie Manning, The Politics of Peace in Mozambique: Post-conflict Democratization, 1992-2000 (Westport: Praeger, 2002), 44-46.

[23] Paul Nugent, Big Men, Small Boys and Politics in Ghana (London: Pinter, 1995). On coups in Ghana, see Naunihal Singh, Seizing Power: The Strategic Logic of Military Coups (Baltimore: Johns Hopkins University Press, 2014).

[24] While later rebel leader Charles Taylor owed his position in the post-coup government to Quiwonkpa, his wife’s uncle, Taylor was not one of Quiwonkpa’s “key lieutenants” as Roessler suggests (240). Instead, he acted independently, with his own ambitions for power, and Taylor fled Liberia not with Quiwonkpa, but after Taylor grew unpopular for his tight-fisted management of government vehicles and resources through the General Services Agency, criticized some of Doe’s decisions, and allegedly embezzled funds. See Mark Huband, The Liberian Civil War (New York: Frank Cass, 1998), 16-20.

[25] Amos Sawyer, The Emergence of Autocracy in Liberia: Tragedy and Challenge (San Francisco: ICS Press, 1992), 295.

[26] See e.g. Laia Balcells, Rivalry and Revenge: The Politics of Violence during Civil War (Cambridge: Cambridge University Press, 2017).

[27] Peter Krause, “Coercion by Movement: How Power Drove the Success of the Eritrean Insurgency, 1960-1993,” in Kelly Greenhill and Peter Krause, eds., Coercion: The Power to Hurt in International Politics, (Oxford: Oxford University Press, 2018), 138-159.

[28] Kristian Skrede Gleditsch, “Transnational Dimensions of Civil War,” Journal of Peace Research 44:3 (2007): 293-309; Michael Mann, “Have Wars and Violence Declined?” Theory and Society 47:1 (2018): 37-60.

[29] See Jonathan M. Powell and Clayton L. Thyne, “Global instances of coups from 1950 to 2010: A new dataset,” Journal of Peace Research 48:2 (2011): 249-259.

[30] Carsten Müller and Carlo Schwarz, “Fanning the Flames of Hate: Social Media and Hate Crime,” Social Science Research Network, 21 May 2018,; Libby Hogan and Michael Safi, “Revealed: Facebook Hate Speech Exploded in Myanmar during Rohingya Crisis,” The Guardian, 2 May 2018,

[31] International Crisis Group, “The African Union and the Burundi Crisis: Ambition versus Reality,” Crisis Group Africa Briefing 122, 28 September 2016,

[32] James D. Fearon and David D. Laitin, “Explaining Interethnic Cooperation,” American Political Science Review 90:4 (1996): 715-735; Ashutosh Varshney, Ethnic Conflict and Civic Life: Hindus and Muslims in India (New Haven: Yale University Press, 2003).

[33] On African militaries, see Kristen Harkness, When Soldiers Rebel: Ethnic Armies and Political Instability in Africa (Cambridge: Cambridge University Press, 2018). On Middle Eastern militaries, see Paul Lorenzo Johnson and Ches Thurber, “The Security-Force Ethnicity (SFE) Project: Introducing a new dataset,” Conflict Management and Peace Science, forthcoming, On security force counterbalancing, see Erica De Bruin, “Preventing Coups d’état: How Counterbalancing Works,” Journal of Conflict Resolution 62:7 (2018): 1433-1458.

[34] Bruce Bueno de Mesquita and Alastair Smith, The Dictator’s Handbook: Why Bad Behavior is Almost Always Good Politics. (New York: PublicAffairs, 2011).

[35] Philip Roessler and David Ohls, “Self-Enforcing Power Sharing in Weak States.” International Organization 72:2 (Spring 2018): 423-454.

[36] On the impact of co-ethnicity on counterinsurgency effectiveness, see Jason Lyall, “Are Coethnics More Effective Counterinsurgents? Evidence from the Second Chechen War,” American Political Science Review 104:1 (February 2010): 1-20.

[37] Kristen A. Harkness, When Soldiers Rebel: Ethnic Armies and Political Instability in Africa. (New York: Cambridge University Press, 2018.)

[38] Richard Sandbrook, and Judith Barker, The Politics of Africa’s Economic Stagnation. (New York: Cambridge University Press, 1985).

[39] Fiona Shen-Bayh, “Strategies of Repression: Judicial and Extrajudicial Methods of Autocratic Survival,” World Politics 70:3 (July 2018): 321-357.

[40] Nils-Christian Bormann, “Uncertainty, Cleavages, and Ethnic Coalitions,” The Journal of Politics 81:2 (April 2019): 471-486.

[41] See for example Aristide R. Zolberg, One-Party Government in the Ivory Coast. (Princeton: Princeton University Press, 1964); Dirk Berg-Schlosser, “Elements of Consociational Democracy in Kenya,” European Journal of Political Research 13:1 (March 1985): 95-109.

[42] Jack Paine, “Ethnic Violence in Africa: Destructive Legacies of Pre-Colonial States,” International Organization (2019): 1-39.

[43] Philip Roessler and Harry Verhoeven, Why Comrades Go To War: Liberation Politics and the Outbreak of Africa’s Deadliest Conflict (London: Hurst Publishers, 2016).

[44] Andrew Boutton, “Coup-proofing in the Shadow of Intervention: Alliances, Moral Hazard, and Violence in Authoritarian Regimes.” International Studies Quarterly 63:1 (March 2019): 43-57.