This outstanding book casts new light on the relationship between power sharing, coups, and civil wars in sub-Saharan Africa. In a piece of scholarship that is remarkable for its analytic clarity, methodological transparency, and commitment to comparative testing, Philip Roessler provides an exemplar of how inductive theorizing from a single case can be used to develop theory with much broader applicability.

The book’s central argument is that African rulers face a perilous trade-off when deciding how much power to share with other ethnic groups. Building on his 2011 article, Roessler argues that rulers face a coup–civil war trap (“The Enemy Within: Personal Rule, Coups, and Civil War in Africa,” World Politics, 63(2), 2011). If they share power with a particular ethnic group, rulers are at greater risk of being thrown out of office via a coup orchestrated by group members. But if they exclude the group from power, they are at greater risk of the group’s successful launch of a civil war. Given the lower and less immediate probability of their losing office via civil war, Roessler argues that African rulers have chosen to mitigate the first risk at the expense of the second in most places: “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 coethnics, to coup-proof their regimes from their ethnic rivals” (p. xvi).

The book also makes an original secondary argument about the mechanism by which ethnic exclusion leads to increased civil war risk. Ethnic exclusion does not only instigate grievances; it also leads to weaker counterinsurgency efforts as the government cannot tap into local networks to provide critical information on rebels. In contrast, ethnic inclusion facilitates counterinsurgency but also increases the ability of ethnic leaders to mobilize military force against the incumbent ruler in a coup d’état.

The theoretical framework also makes predictions about the circumstances under which rulers trade off the risk of civil war for the risk of a coup and power sharing emerges. If a group is well positioned to capture the capital city in a civil war, its chance of seizing power via rebellion begins to approximate its chance of doing so via a coup. In these instances, rulers may be willing to share power with the group, especially if their own group also has high threat capacity and is therefore likely to be included in any future power-sharing agreement if a coup d’état does materialize.

Ethnic Politics and State Power in Africa begins by outlining its argument at length in the first section of the book. Then the author provides empirical evidence drawing on multiple methods, including a qualitative study of Darfur that was used to build the theory, cross-national quantitative tests, and qualitative tests using the case of Africa’s “Great War” in the Democratic Republic of Congo.

The study of Darfur in the second section is particularly remarkable. It draws on elite interviews in Khartoum, Darfur, Asmara, N’Djamena, Abuja, Europe, and the United States to explain why a full-scale civil war did not break out in the region in the early 1990s but did in the early 2000s. Roessler shows both that Sudan’s Islamic Movement was critical in keeping the peace in Darfur and that Omar al-Bashir recognized this even as he decided to purge Hassan al-Turabi and his supporters from the government after December 1999 to avoid a coup.

The author then tests the effect of power sharing on coups and civil wars in the book’s third section using the Ethnic Power Relations (EPR) data set developed by Andreas Wimmer, Lars-Erik Cederman, and Brian Min based on expert surveys (“Ethnic Politics and Armed Conflict: A Configurational Analysis of a New Global Dataset,” American Sociological Review, 74(2), 2009). Utilizing data on all ethnic groups in sub-Saharan Africa included in the data set, from independence to 2005, Roessler shows that ethnic groups that are included in the central government are significantly more likely to be involved in coup attempts and significantly less likely to be involved in rebellion. Because he places great theoretical emphasis on power sharing as a strategic decision based on groups’ threat capabilities, it is surprising that his empirical models do not account for this. However, if his theoretical predictions about when rulers share power with groups are correct (i.e., power sharing occurs when civil war threat is high or coup threat is low), he is biasing his empirical results against his predictions about the effects of power sharing by not accounting for the selection problem.

A bigger issue is that Roessler’s empirical evidence does not actually demonstrate that power sharing decreases the security of rulers. His analysis shows that the exclusion of an ethnic group from power reduces the likelihood of the group producing any coup perpetrators, not that it decreases the overall likelihood of a coup against a ruler. Ultimately, what matters for a ruler is not the number of ethnic groups who participate in a coup against him but whether a coup materializes at all in his country, and it is not evident that this risk can be decreased by excluding members of non-coethnic groups. Indeed, rulers in Africa appear subject to coups from opportunistic individuals regardless of whether they share power across ethnic lines or not: To this point, 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.

The third section of the book also includes more direct evidence of the commitment problems bedeviling power sharing. Roessler shows that coconspirators—groups that helped the ruler come to power in a violent transition—frequently launch rebellions after being excluded from power. He also draws on qualitative evidence from Africa’s Great War, collected in partnership with Harry Verhoeven, to test the importance of coup-proofing in civil war outbreaks.

The test of when power sharing emerges is left to the final section of the book. Here, Roessler draws on joint work with David Ohls to argue that leaders substitute civil-war risk for coup risk by engaging in power sharing in cases where both the ruling group and its rival are in strong positions to seize power in a civil war. This hypothesis is a little bit difficult to square with the author’s hypothesis that coconspirators—many of whom have demonstrated capabilities to help win civil wars—frequently launch rebellions after being excluded from power. Instead, he operationalizes threat capabilities by averaging groups’ relative size and absolute distance to the capital, and shows that power sharing is significantly more likely in the case of strong opposition and ruling groups.

The book provides a powerful theoretical framework with broad implications, for Africa and beyond. But despite my overall admiration for the book, I disagree with Roessler’s claims about the inherent instability of power sharing in sub-Saharan Africa due to the pervasiveness of ethnic security dilemmas.

The theory and evidence in my own book also provide a framework for understanding the prevalence of inclusive governance in Africa. Focusing on an important example of “ethnic” big men in Africa—traditional chiefs—I show that chiefs play critical roles in mobilizing local development projects, and that their relationships with politicians are relevant for the flow of resources to communities. As a result, politicians have incentives to share power with these “development brokers” so that they can tap into their networks to ensure more effective delivery of goods and services. In this view, inclusive governance is beneficial because it allows the state to do more with the same amount of government resources. This framework helps us understand the pervasiveness of power sharing and ethnically inclusive governance across sub-Saharan Africa.

My main difference with Roessler is in the extent to which I think that ethnic security dilemmas undermine power sharing. As mentioned, I do not think the evidence shows that power sharing increases political instability across sub-Saharan Africa, and I am not convinced that commitment problems between rulers and potential challengers are especially great when rulers share power with non-coethnics. Ultimately, Roessler and I make fundamentally different arguments about the importance of ethnicity in organizing politics in our respective books. He argues that “ethnic followers” almost always line up behind “ethnic patrons” (p. 95). Instead, I view citizens as primarily concerned with the performance of their rulers. I provide evidence that citizens may consider the opinions of their local chiefs insofar as they provide information on politicians’ performance, but neither ethnics nor non-coethnics of chiefs blindly follow them. In this framework, politics need not break down along ethnic lines.

Inevitably, the extent to which citizens weigh performance versus ethnic alignment in deciding whether to support politicians varies from setting to setting, with some places looking more like Zambia and others more like Sudan. I found Roessler’s discussion of ethnic commitment problems most compelling in countries that have already experienced ethnic conflict—as in Sudan, the Democratic Republic of Congo, and several of the cases of coconspirator civil wars. But this begs the question of how we get on the path of ethnic polarization in the first place, something that is not sufficiently explained by either Roessler’s theoretical framework or my own. Understanding these inflection points is an important research agenda for future scholarship.