August 2015 ABSTRACT

Transcription

1 THE SOCIETAL ORIGINS OF POWERSHARING ETHNIC GEOGRAPHY, THREAT CAPABILITIES AND HORIZONTAL ACCOUNTABILITY IN WEAK STATES Philip Roessler Department of Government College of William and Mary Dave Ohls School of International Service American University August 2015 ABSTRACT Why are some weak states trapped in devastating cycles of ethnic exclusion and civil war while others experience durable peace, secured by informal powersharing between rival ethnic elites? We argue these two equilibriums are rooted in the distribution of societal power and rival groups capabilities to credibly threaten to recapture state power if excluded from the central government. Only when both the ruling group and opposition possess strong threat capabilities does self-enforcing powersharing emerge. A strong opposition induces the ruler to commit to powersharing and to reluctantly accept coup risk over civil war risk. The ruling group s own threat capabilities, in turn, constrain the opposition from trying to convert its share of power into absolute power. Supported by extensive quantitative and qualitative evidence with particular reference to weak states in sub-saharan Africa, the paper advances a novel understanding of the informal sources of horizontal accountability and powersharing. 1

2 1. INTRODUCTION Violent ethnic conflict has been one of the key causes of mass killing, economic underdevelopment and societal displacement in the world since World War II (Gurr 2000; Harff 2003). Such conflicts are particularly concentrated in weak states, which lack effective institutions to regulate the distribution of power and broadcast authority (Fearon and Laitin 2003; Hironaka 2005). Yet not all weak states are plagued by ethnic-based civil wars. In some, largescale political violence has been averted or managed through powersharing, in which rival groups agree to share control of the state (Lijphart 1977; Horowitz 1985; Rothchild 1997; Wimmer et al. 2009). Understanding the sources of durable powersharing is important to address the scourge of civil war. Existing research tends to focus on the importance of either formal institutions (Lijphart 1977; Magaloni 2008; Norris 2008) or external intervention (Walter 2002; Hartzell and Hoddie 2007). While both factors can serve as key levers to reduce uncertainty, most powersharing regimes since World War II have emerged without external intervention, and in most weak states informal institutions have greater influence than formal ones (Helmke 2004; Reno 1998). Under such conditions, in which societal groups cannot rely on strong institutions or external actors to hold rulers accountable, the distribution of power is ultimately determined by the threat of force (North et al. 2009; Svolik 2012). Rivals must be able to credibly threaten a rebellion that would oust the ruler should he reject or violate powersharing (Svolik 2012; Boix and Svolik 2013; see also Acemoglu and Robinson 2006). But herein lies a paradox: the threat of force is necessary to hold rulers accountable and sustain powersharing but, as it can also be abused by rivals to grab and monopolize power for themselves, the shadow of violence can incite the very exclusion it seeks to deter. When does the threat of force lead to a peaceful and productive 2

3 equilibrium underwritten by powersharing and when does it merely reproduce the exclusionconflict cycle? Addressing this question is critical to understanding the foundations of political order. We argue the roots of peaceful powersharing lie in the distribution of societal power. The link between the two has been largely undertheorized. We argue that the threat of force acts as a credible deterrent on the ruler s violation of powersharing when it is backed up by strong societal power, in which a given group possesses the mobilizational potential to credibly threaten to recapture state power from its societal base if it is excluded from the central government what we describe as a group s threat capabilities. Unless a group possesses strong threat capabilities, the ruler is weakly constrained from excluding that group and appropriating its share of power for himself and his co-ethnics. However, for powersharing to be self-enforcing, the ruling group similarly must possess strong threat capabilities otherwise the opposition will face few constraints from exploiting access to the central government to appropriate the ruling group s share of power. In the absence of strong institutions to regulate self-enforcing powersharing, mutually strong threat capabilities transform the choices made in response to the coup-civil war trap that plagues weak, ethnically-divided states. Strong rivals commit to powersharing to avoid mutually costly civil wars, but in turn such arrangements open the door to future coups d état. Under such conditions, interethnic powersharing is more likely, and violent contestation for power when it does occur takes the form of coups, which lead to a change in executive authority but do not significantly disrupt the relative distribution of political power. We empirically test these theoretical expectations both qualitatively and quantitatively, focusing on sub-saharan Africa from independence to We expect the distribution of societal power to have a particularly important effect on powersharing in post-colonial Africa 3

4 given the strong societies and weak states political order that emerged after decolonization (Migdal 1988). Unable to rely on strong formal institutions to regulate society, political order and societal peace hinges on a ruler s ability to strike alliances and share power with Big Men embedded in rival ethnic groups. Quantitatively, we use data from the Ethnic Power Relations (EPR) 3.0 dataset (Wimmer et al. 2009) and a compatible geo-referenced dataset, GeoEPR-ETH (Wucherpfennig et al. 2011), to develop a parsimonious operationalization of threat capabilities derived from a group s ethnic geography that is, its location and size as a proportion of the population with the expectation that the larger a group and the closer to the capital city it is, the more likely its members can credibly threaten to take power if excluded from the central government. Statistical analysis of more than 200 politically-relevant ethnic groups in 35 sub-saharan African countries demonstrates that ethnic powersharing is most likely when both the ruling group and opposition possess strong threat capabilities. Strikingly, this is the case despite an elevated risk of coups. In contrast, when either the ruler s or opposition s threat capabilities are weak, rulers are more likely to reject powersharing. This strategy is effective at coup-proofing the regime from ethnic rivals but at an increased risk of civil war. Qualitatively, we employ a diverse case selection to explore the dynamics of ethno-political bargaining, coups and civil war across different distributions of societal power. The logic of selfenforcing powersharing accounts for the puzzling historical pattern seen in Ghana and Benin, in which ethnic rivals traded executive power via coups yet no one group sought to monopolize power and coup-proof their regimes using ethno-political exclusion. This stands in stark contrast to the policies of ethno-political exclusion rulers and their co-ethnics in Sudan, Liberia and South Africa implemented to monopolize their hold on power, leading to civil war. 4

5 This paper breaks new ground in the study of powersharing and conflict. First, it offers a coherent theory of how the distribution of societal power can lead to horizontal accountability and powersharing in the absence of strong political institutions or a third-party enforcer. Second, it incorporates the study of civil war into a broader framework that considers the competing risks rulers face to their hold on power. Civil war studies, especially the study of ethnic-based conflict, continue to be dominated by a focus on grievances or opportunity structure (Collier 2009; Gleditsch and Ruggeri 2010; Cederman et al. 2013), often failing to consider how rulers assess competing threats and why they are unable to simply increase concessions in the face of an aggrieved population and weakened capacity to repress (Fearon 2010). A bargaining framework is essential to address these questions. 1 Third, this bargaining approach leads us to challenge the conventional view of civil war as a form of effective resistance (Sambanis 2004), in which a guerilla movement wins by not losing (Kissinger 1969). We conceive of groups mounting such peripheral insurgencies holed up in mountainous terrain as possessing low threat capabilities; they are tolerated exactly because they are weak. The fact that most theorization of civil war draws from this stylized version of a guerilla army is problematic because it suggests a potential selection problem in the study of civil war inferences are drawn disproportionately from observable (i.e., weak) insurgencies, as those groups with the greatest potential to wage violent opposition are compensated to not do so. 2 Reconceptualizing civil war as a contest for state power, rather than as simply conflicts in which the opposition can avoid defeat at the hands of the government, shows that (in contrast to the conventional wisdom) strong threat capabilities underwrite peace, not conflict, as they constrain the ruler from reneging on powersharing. 1 For earlier applications of the bargaining approach to ethnic-based conflict, see (Cetinyan 2002; Fearon 1995a). 2 This is a key insight that Cetinyan makes (Cetinyan 2002). 5

6 Finally, this theoretical framework accounts for Africa s large state problem (Clapham 1996; Clapham 2006; Green 2012) that is, why Africa s largest countries, such as Sudan, Democratic Republic of Congo, Angola, Chad and Ethiopia, tend to be continually plagued by ethno-political exclusion and civil war. Big states tend to contain more groups with low threat capabilities, which lack the ability to hold the ruler accountable for violating powersharing. The paper proceeds as follows. Section two analyzes bargaining in weak states as confronting rulers with a coup-civil war trap. Understanding this dynamic is essential to developing a model of self-enforcing powersharing. Section three explores how the distribution of societal power and the balance of threat capabilities structure ethnic bargaining, and the conditions under which this produces durable powersharing. The fourth and fifth sections test the threat capabilities theory of powersharing, drawing on quantitative and qualitative data, respectively. Section six concludes with the theoretical implications and identifies several avenues for future research. 2. THE COUP-CIVIL WAR TRAP IN WEAK STATES One of the defining characteristics of weak states is that locally, spatially-concentrated social communities, such as kinship or ethnic groups, serve as more durable and binding political organizations than large, socially-complex cross-cutting organizations that can enforce rules impersonally (Horowitz 1985; Migdal 1988; North et al. 2009). In the modern state this strong societies and weak states institutional dichotomy (Migdal 1988) gives rise to a coup-civil war trap political inclusion is necessary to extend the reach of the regime and prevent societal rebellion, but bringing one s rivals to the center of power is risky as it lowers the costs they face to seize power for themselves in a coup d état. 3 In this section we elucidate the dynamics of the 3 We distinguish between coups and rebellions by the organizational basis of the 6

7 coup-civil trap, before explaining how the distribution of societal power structures ethnopolitical bargaining and the way the coup-civil war trap plays out. The coup-civil war trap is rooted in two fundamental political problems that plague weak states. The first is the predicament that arises from strong social institutions and weak crosscutting political ones. Dense networks, strong norms of reciprocity and spatial clustering among co-ethnics facilitate intra-group cooperation (Fearon and Laitin 1996; Habyarimana et al. 2009; Weidmann 2009), but serve as barriers to inter-group cooperation. Excluded from rival ethnic groups social ties, rulers are not able to directly leverage the technology mechanisms of shared language, reachability, and periodicity and thus face much higher costs of monitoring and sanctioning non-co-ethnics (Habyarimana et al. 2009). This not only makes it more difficult to collect taxes and produce public goods (Easterly and Levine 1997; Kasara 2007; Habyarimana et al. 2009) but also to effectively accommodate and repress potential dissidents (Lyall 2010). Mistrust due to the lack of norms of reciprocity also makes non-co-ethnics reluctant to share information or cooperate with the ruler, especially if they perceive that it will harm members of their own group (Nunn and Wantchekon 2009). To overcome these barriers and extend their authority beyond their own ethnic groups, rulers turn to informal institutions of powersharing (Rothchild 1986; Azam 2001; Roessler 2011). In exchange for political support from other elites (and their co-ethnics), rulers agree to share access to the state and the rents and privileges that come from controlling its key levers (e.g., collection of taxes, use of repression, issue of business licenses, trade, disbursement of foreign aid) (Jackson and Rosberg 1982b; Rothchild 1986; Reno 1998; North et al. 2009). By giving anti-regime technique. Coup conspirators leverage partial control of the state (and the resources and materiel that come with access to the state) in their bid to capture political power, whereas rebels or insurgents lack such access and have to build a private military organization to challenge the central government and its military. 7

8 rivals a stake in their regimes, rulers seek to win their loyalty and reduce their relative benefits from trying to capture power. But such informal powersharing arrangements give rise to a second political problem: neither side is able to credibly commit not to exploit its access to the state to lock-in a larger share of power via a coup d état an attempt by a given political faction or group to use extant state structures or institutional bases of power within the political system (Huntington 1968), especially the military, to unseat the ruler by force or other unconstitutional means (Luttwak 1968; Powell and Thyne 2011). 4 While rival factions within the government have incentives to refrain from using force to avoid the destruction of shared economic rents (Fearon 1995b; North et al. 2009), a coup or its analogue, a purge which significantly diminishes a rival s capability to seize power can permanently alter the strategic balance of power (Blattman and Miguel 2010) 5 and bring about long-term, compounding rewards (Garfinkel and Skaperdas 2000) for the winning side. Exacerbating this commitment problem in the post-world War II era has been the de facto capital city rule, in which the international system uses control of the capital to decide which group to recognize as the sovereign authority of a given state no matter how the group comes to power or how much control it has outside the capital (Herbst 2000). 6 Given the legitimacy and rents that come from international recognition, this rule increases incentives to capture central executive authority by whatever means possible. This commitment problem arising from the shadow of the coup d état undermines powersharing as it prevents rulers from giving real power to rivals lest they exploit it to gain 4 We consider three primary channels by which powersharing increases a rival group s ability to internally lock-in a larger share of power: 1.) included groups gain direct control of a critical segment of the military and coercive apparatus; 2.) control of rents and positions of political authority give rivals bargaining power with members of the military, which they can leverage to displace the incumbent; and 3.) inclusion gives rivals the discretionary political authority and influence to severely undercut the powers of the president. 5 See also (Fearon 1995b; Powell 2006). 6 See also (Jackson and Rosberg 1982a) 8

9 sovereign authority for themselves. 7 In rejecting powersharing, however, rulers in weak states alienate rival groups, increasing the likelihood they will withdraw their support for the regime and mobilize against it. Without access to the state, these excluded groups must build their own army via a rebellion or insurgency to try to coerce the ruler to restore (or commence) real powersharing. But in mobilizing against the state, the excluded groups benefit from the strong societies, weak states institutional environment discussed above strong social networks and territorial control reduce the costs of rebellion, while increasing the costs of counter-insurgency (Lichbach 1995; Kalyvas 2006; Weidmann 2009). This commitment problem that arises as strong groups in weak states bargain for power in the shadow of the coup d état can be seen leading to the outbreak of a number of devastating civil wars, such as the Biafran War (Luckham 1971), the Rwandan genocide (Prunier 1995) and successive civil wars in Afghanistan (Rubin 2002), as well as more contemporary conflicts in South Sudan (Roessler 2013), Syria (Robinson 2012) and Iraq (Hirsch and Smith 2014). The latter cases provide direct examples of rulers (Salva Kiir in South Sudan, Bashar al-assad in Syria and Nouri al-maliki in Iraq) choosing to violate or reject powersharing with an important rival social group as a means to protect their own group s hold on power, at the price of largescale political violence. Not all weak states are characterized by this devastating exclusion-civil war equilibrium, however. In others, powersharing has proven more durable and the key to societal peace (Lijphart 1977; Wimmer et al. 2009; Cederman et al. 2010). What accounts for this variation? What conditions produce durable powersharing and allow weak states to avoid vicious cycles of social exclusion and large-scale civil war? 7 Some such cases do feature symbolic sharing of formal power (e.g. through a representative cabinet), but nonetheless keep real power in the hands of co-ethnics in a shadow-state (Reno 1998) 9

10 3. THE SOCIETAL BASIS OF SELF-ENFORCING POWERSHARING 3.1 The Limitations of Conventional Approaches to Powersharing Powersharing entails the joint participation of elites of two or more rival groups in a governing coalition, in which both sides agree to parcel out the rents that come from control of the state and each refrains from using force to grab a larger share of power. 8 This implies two necessary conditions. First, the leader in power must choose to bring the opposition into the government. Rather than attempt to rule the country unilaterally, keep potential opposition at arms-length, and extract all of the rents accompanying dominant control of the state apparatus, the ruler must prefer to accommodate the opposition and accept the risk of giving rivals positions of some power. Second, the opposition group, once invited to participate in the central government, must accept its share of power and refrain from leveraging its privileged position to gain absolute power. In accounting for such incidences of powersharing, extant research tends to focus on the role of formal institutions and external third-party mediators. The institutional approach emphasizes the importance that formal rules, such as consociational arrangements (Lijphart 1977; McGarry and O'Leary 1993), and political organizations, such as parties and legislatures (Gandhi and Przeworski 2006; Magaloni 2008; Boix and Svolik 2013), play in helping to regulate the distribution of power. Such institutions guarantee minorities representation in government, constrain rulers, and increase transparency, which reduces fear and misperceptions about the 8 We primarily employ a dyadic framework to analyze ethno-political bargaining. This is not to imply that the broader strategic environment is irrelevant (see (Walter 2009) for a useful analysis). Instead we believe this model best captures the primacy of bilateral ties in weak states, which arise due to underlying institutional and structural conditions (e.g., weak cross-cutting cleavages, state s control over the economy) that enable rulers to use divideand-rule policies and offering of selective incentives to divide society and inhibit the emergence of reliable interethnic coalitions (Acemoglu et al. 2004; Arriola 2012). In the Online Appendix we relax this assumption and analyze alternative specifications that model cross-group and coalition dynamics. 10

11 intentions of the other side. External approaches, on the other hand, view third-party enforcers, which can both help to forge a political agreement and coerce or sanction rivals if they renege on the political agreement, as necessary to overcome the commitment problem and produce peaceful powersharing (Walter 2002; Hartzell and Hoddie 2007). These research streams have made important contributions to the study of peacemaking in ethnically-divided states, but they leave several puzzles unresolved. One is that most powersharing regimes since World War II have emerged without external intervention. A second is why rulers accept formal institutions in the first place, especially if it significantly constrains their power (Migdal 1988; Reno 1998). Accordingly, in many weak states politics is defined by limited formal institutionalization and the dominance of personal or informal rule (Jackson and Rosberg 1982b; Helmke 2004). In the absence of third-party intervention and strong institutions, what sustains powersharing? Under such conditions, one of the few levers groups can employ to hold each other accountable is the threat of violence, and only when this threat is credible on both sides will groups share power. In the next section we advance a theory on the societal origins of selfenforcing powersharing. 3.2 Societal Power and Self-enforcing Powersharing Political exclusion is an instrument for the consolidation of power. In denying rivals access to the central government and increasing the costs they face to seize power, exclusion is a means for rulers to strengthen regime security 9 and maximize their control of rents. But the stronger a 9 Exclusion from the central government does not by definition negate the possibility of orchestrating a coup one could try to execute a coup remotely by secretly allying with those in the military and others who control other key strategic points within the government. But it does make it much more difficult by increasing the coordination costs and principal-agent problems that arise from needing to strike such alliances. 11

12 potentially excluded group s mobilizational capacity to recapture state power from its societal base, the less appealing such a policy becomes. For example, though ethnic dominance enables the ruling group to monopolize power and rents that come from controlling the state (Bates 1983), if pursuing such benefits provokes a costly civil war that necessitates an expensive counter-insurgency campaign, especially one with no guarantee of success, then the benefits of exclusion dissipate and may no longer outweigh the costs. In other words, a group s societal power and threat capabilities the coercive capacity it has independent of that it gains from powersharing itself represent the key levers the opposition has to hold the ruler accountable if he attempts to reject or violate powersharing and lock-in a larger share of power for himself and his co-ethnics. The ruler s commitment to powersharing, however, is not just a function of the opposition s threat capabilities but also of his own group s societal power. Unless the opposition faces the same constraints as the ruling group, such that it too would face a strategically costly civil war if it tried to monopolize power in a coup, there is little preventing it from exploiting access to the central government to appropriate the ruling group s share of power. Thus, for powersharing to be self-enforcing, such that neither group has incentives to exclude the other, the costs of reneging on powersharing must constrain not just the incumbent, but constrain in expectation any actor that may seize power in the future. Only when both sides see little strategic benefit to choosing exclusion will neither group try to permanently exclude the other, and will strong rivals reluctantly accept powersharing. It is important to note the presence of mutually strong civil war capabilities does not resolve (initially at least) the commitment problem at the heart of the coup-civil war trap. Without agreed upon rules or institutions regulating the distribution and transfer of sovereign power, elites 12

13 embedded in each group are still vying to control the executive (and gain the international recognition and rents the come with it) and anticipate their rivals have the same intentions. This can lead to political instability and actually increase coup risk (Horowitz 1985; Harkness 2014). Under such conditions, however, rival groups prefer to share power and reluctantly trade executive authority via coups which do not significantly alter the relative distribution of power than accept the high mutual costs associated with a war for exclusive control of the state. This is a striking and counter-intuitive implication of the theory: mutually strong threat capabilities induce rulers to accept coup risk over civil war risk. Why would authoritarian rulers adopt a strategy that brings about such high individual political risks? This seemingly goes against a number of seminal studies in comparative politics, which suggest the opposite: authoritarian rulers, in an effort to secure their political and physical survival, have few qualms about using exclusion and other strategies that may be in their personal interest but which bring about devastating costs for their citizens and the state, such as economic crises, international conflict, civil war and state collapse (Reno 1998; Bueno de Mesquita et al. 2003; Bates 2008; Chiozza and Goemans 2011; Roessler 2011). To account for this, it is important to distinguish between the interests of the ruler and interests of the ruling group in which the ruler is embedded and on whose support the ruler is dependent upon to stay in power (North et al. 2009). While rulers may prefer to use exclusion to substitute coup risk for civil war risk to protect their personal hold on power, under conditions of mutually strong threat capabilities, this offers little political or strategic advantage for the ruling group as a whole, whose relative share of power is secured by its societal power, but also constrained by the strength of its rival. Consequently, in the face of a devastating civil war the 13

14 costs of which would be borne by the group s members the ruling group is unlikely to support the ruler s rejection of powersharing in a bid to consolidate his personal hold on power. The same does not hold for weak groups, however, which lack the capabilities to reclaim power once displaced from the central government. For these groups, the ruler s and the group s interests are better aligned. A coup from their stronger rivals not only threatens the individual ruler s hold on power, but also the group s relative share of power, since it lacks the societal strength to deter the new ruling group from monopolizing power. Under these conditions, both the ruler and the ruling group favor exclusion, and are more willing to endure the costs, such as civil war, that come from it. 3.3 Rethinking Civil War In conceiving of civil war as a potential lever of political accountability, we diverge significantly from existing scholarship on the subject. Most studies equate civil war with insurgency, which Fearon and Laitin define as a technology of military conflict characterized by small, lightly armed bands practicing guerrilla warfare from rural base areas (Fearon and Laitin 2003). The success of an insurgency is often measured based on a group s ability to effectively resist the government that is, inflict a minimal level of costs on the government while avoiding defeat (Sambanis 2004; Collier 2009). As Henry Kissinger famously quipped, the guerrilla wins if he does not lose (Kissinger 1969). Equating civil war with armed insurgency has led scholars to focus on those conditions that render insurgency more feasible and attractive (Fearon and Laitin 2003) and make it hard for the government to effectively defeat the insurgents. Such factors include distance to the capital, sanctuary in neighboring countries, mountainous terrain, 14

15 and ethnic group concentration (Fearon and Laitin 2003; Buhaug et al. 2008; Weidmann 2009; Salehyan 2009). We instead conceive of civil war as a contest for state power between the central government and an organized, armed opposition movement. Armed rebellion is the lever the opposition employs to gain a share of power and rents. Civil war capabilities, thus, represent a group s capacity to credibly obtain its fair share (and given that bargaining is a dynamic process, capturing state power for oneself is the only credible way to guarantee one s share of power). This approach suggests a different set of predictors than is conventionally used in the civil war literature. We would expect groups located in remote, mountainous areas far from the capital to have weaker, not stronger, civil war capabilities. In contrast we expect large groups, located closer to the capital, and that control valuable resources and dominate economic markets to have the greatest capacity to credibly threaten to bring down the central government and thus lock-in a share of power. One reason that threat capabilities has been largely marginal to the study of civil war is the political contests underlying civil war and potential civil wars have been largely missing from existing analyses (Blattman and Miguel 2010). The dominant theoretical frameworks that focus on grievances or opportunity structure tend to gloss over the political bargaining that is inherent to civil war onset (Fearon 2010). 3.4 Balance of Threat Capabilities and Powersharing: Hypotheses We argue that strong threat capabilities produce durable, but sometimes fluid, powersharing, in which rivals prefer inclusion (even when it risks trading power via coups) to engaging in mutually costly civil wars for absolute power. This leads to our first hypothesis: 15

16 H1: When both the ruling group and opposition group have strong threat capabilities, the opposition is included in the central government, reducing civil war risk but increasing coup risk. When groups have asymmetric threat capabilities, however, self-enforcing powersharing is significantly less likely. While both sides may wish to share access to the central government to avoid costly conflict, the power differential undermines such an agreement. Most obviously, a weak group bargaining with a strong ruling group lacks the threat capabilities to hold the strong group accountable if it violates the terms of a deal. The stronger group s inability to credibly commit itself not to exploit its greater bargaining leverage induces the weaker group to stay on a war footing or face political irrelevance. This increases the likelihood of protracted civil war (Fearon 1995a; Walter 2002), particularly given the challenges a weak central government faces to extending its authority over rivals without support from trusted intermediaries inside that group. Additionally, because the weak group s low civil war capabilities render them unlikely to be able to reclaim power if they lose access to the central government, they are more likely to approach political bargaining as a one-shot game and adopt extreme policies (e.g., coup attempts followed by consolidating power in an ethnocracy, using high-levels of repression) to try to hold on to power at all costs. This further drives a strong group to hedge its bets on exclusion as it sees the marginal costs of a relatively less-threatening civil war as being significantly lower than the marginal costs of a coup. This leads to our second hypothesis: H2: When the ruling group s threat capabilities are high and the opposition group s are low, the opposition is excluded from the central government, reducing coup risk but increasing the risk of civil war. Similar dynamics lead a weak ruling group to universally reject powersharing with either a strong or a weak opposition. Again, the breakdown of powersharing arises from the weak groups weak threat capabilities and thus their political vulnerability. When bargaining with 16

17 another weak group, neither side possesses the mobilizational potential to hold the other to account if they renege on powersharing. Consequently, both have strong incentives to eliminate the other from state power before they are eliminated, leading to ethnic exclusion and civil war. This leads to our third hypothesis. H3: When the ruling group s threat capabilities are low and the opposition group s are also low, the opposition is excluded from the central government, reducing coup risk but increasing the risk of civil war. When the opposition is strong, a weak group faces a more acute dilemma. Powersharing puts the stronger rival in a position to usurp power in a coup, likely leading to the permanent exclusion of the weak group. But employing exclusion to prevent such an outcome provokes a strategically costly civil war. Both are bad outcomes, but losing power via civil war is seen as more uncertain than in a coup in which its rival already controls a significant share of the state. This leads to our fourth hypothesis. H4: When the ruling group s threat capabilities are low and the opposition group s are high, the opposition is excluded from the central government, reducing coup risk but increasing the risk of civil war. Figure 1 summarizes the theoretical predictions and how they align with H1-H4. Figure 1: Threat Capabilities, Self-Enforcing Powersharing and Coup and Civil War Risk Strong Opposition Weak Opposition Strong Ruling Group H1: Powersharing and societal peace (despite high coup risk) Powersharing: Yes Coup risk: High Civil war risk: Low H2: War-prone ethnocracies Powersharing: No Coup risk: Low Civil war risk: High Weak Ruling Group H4: Repressive minority rule (or unstable powersharing) H3: Unstable, violent, exclusionary regimes 17

18 Powersharing: No Coup risk: Low Civil war risk: High Powersharing: No Coup risk: Low Civil war risk: High 4. QUANTITATIVE EMPIRICAL EVIDENCE To test the generalizability of the threat capabilities theory of powersharing, we focus our analysis on a subset of states in sub-saharan Africa. While we expect the argument to apply to all weak states with strong, spatially-concentrated societal groups, we focus on Africa primarily because many states in sub-saharan Africa meet these criteria, and because we can obtain societal data not only on access to state power (Wimmer et al. 2009) but also on the ethnicity of those groups launching rebellions and those executing coups (Roessler 2011). There are two primary reasons we would expect the theory to be particularly relevant to African states. First, the strong societies and weak states problem is a defining characteristic of political order that emerged after decolonization (Migdal 1988; Herbst 2000; Englebert 2002). Unable to rely on strong state structures to broadcast power, rulers have needed to strike powersharing deals with ethnic rivals (Rothchild 1986; Rothchild 1997). Second, politics postindependence Africa has centered almost completely on control of the central government. The decision by African heads of states at the founding of the Organization of African Unity in the early 1960s to accept the inviolability of colonial state borders has contributed to strong norms and institutions against state partition, reducing the viability of exit as a policy response to marginalization or discrimination compared to other regions in the world (Englebert and Hummel 2005). Consequently, even the most peripheral groups have had to target control of the central government as a means of addressing their material grievances. 18

19 To test our main hypotheses, we estimate models of interethnic powersharing, coups, and civil war using the Ethnic Power Relations dataset of ethnic groups in 35 African states from 1946 to 2009 (Wimmer et al. 2009; Cederman et al. 2010). 10 The EPR dataset is especially useful to test this paper s central argument because, drawing from a survey of country-experts, it provides information on the inclusion or exclusion of politically relevant ethnic groups in central governments across countries in Africa (and other states in which ethnicity is politically salient). Moreover, because it does not restrict its analysis to only cabinet posts, but also to representation in the presidency and senior posts in the administration, including the army, depending on their de facto power in a given country, it can better capture actual power distribution than what a formal analysis of the ethnic composition of the cabinet would provide. Since the theory is about the interaction between the ruling group and a given rival group, the unit of analysis is the ethnic-dyad-year, with models including all politically-relevant ethnic groups in relation to the ruling group. Ruling groups are those identified as being politically-dominant, or having the most politically powerful status, based on the EPR dataset (most of the time, though not always, this corresponds to the ethnic group of the executive). 11 As we are interested in whether particular binary outcomes do or do not occur, we use a logistic regression specification with standard errors clustered by country to account for nonindependence of ongoing political relationships within states, and with the cubic polynomial approximation of the hazard to account for temporal dependence (Carter and Signorino 2010). 4.1 Variables 10 We use EPR 3.0 available at An alternative would be to use EPR-ETH version 2.0 or EPR-ETH All stem from the original EPR 1.1 dataset and do not include major differences in their codings of access to state power in Africa. 11 When there are multiple senior partners in government, we code whichever group controls the executive as the ruling group. 19

20 There are three dependent variables of interest. The first, interethnic powersharing, is operationalized using a measure of whether a given non-ruling group was included in a governing coalition in that year. The dichotomous variable, Ethnic powersharing, takes a value of 1 if the non-ruling group was included in government (coded by EPR as monopoly, dominant, senior partner or junior partner) and a 0 if it was not (coded as regional autonomy, separatist, powerless, or discriminated). The second dependent variable, successful coup, identifies whether members of a given group were key conspirators in a successful coup attempt that year (Roessler 2011). The dichotomous variable, Successful coup, takes a value of 1 if there was at least one successful coup and 0 if there was not. The third dependent variable, group rebellion onset, identifies whether members of a given ethnic group initiated a major armed rebellion or insurgency against the central government in a given year. 12 The dichotomous variable Group rebellion takes a value of 1 if members of the group launched a rebellion in a given year and 0 if not. To proxy threat capabilities, we develop a measure based on two factors that determine a group s mobilizational potential to threaten the central government. First, the larger the size of a group, the more popular power it can wield and the larger the rebellion or uprising it can mobilize. To measure this, we use the group s proportion of the state s total population Group size (Wimmer et al. 2009; Cederman et al. 2010). Second, the more proximate a group is to the center of power, the lower the mobilizational costs necessary to seize control of the state apparatus. To measure this, we calculate the distance in kilometers between the capital city and 12 This measure captures the year a significant number of members of a given group became involved in a civil war with a minimum of 1000 battlefield deaths. (Roessler 2011) We also estimated models using measures of group rebellion onset from EPR 3.0, with substantively similar results. 20

21 the centroid of ethnic group s homeland. 13 These variables are normalized and the distance data is inverted so larger groups and closer groups have higher values. They are then combined to create a continuous Threat capabilities variable that ranges from 0 (least threatening) to 1 (most threatening). 14 The variable, Ruler threat capabilities, measures the capabilities of the ruling group for a given year. Using this coding of opposition and ruling group threat capabilities, we construct our key independent variable the Balance of threat capabilities between the ruling group and a given opposition. For ease of interpretation, and because we do not expect the joint effect to necessarily be continuous, 15 the balance of threat capabilities measure is captured through four interaction dummy variables (following Figure 1) based on whether the ruling group and the given opposition group have threat capabilities scores above or below the median for all groups in sub-saharan Africa. Not surprisingly, LH dyads are least common, while HH and HL dyads (where the ruling group is strong) are most common. High-high (HH) threat capabilities: Ruler above median for all groups; opposition above median for all groups o (N=2,024 group-years, 34% of total) High-low (HL) threat capabilities: Ruler above; opposition below o (N=2,416 group-years, 40% of total) Low-high (LH) threat capabilities: Ruler below; opposition above o (N=600 group-years, 10% of total) Low-low (LL) threat capabilities: Ruler below; opposition below o (N=951 group-years, 16% of total) 13 Results are substantively similar when using the minimum distance between the capital and the group s territory. See Online Appendix. For the geocoded EPR groups, see (Wucherpfennig et al. 2011).Though GeoEPR is currently maintained as part of the EPR-ETH family of datasets, we found GeoEPR also matches closely with EPR 3.0 as well (not surprisingly, as they are both derived from the same base dataset, EPR 1.1.) and provides valuable geospatial information on the politically-relevant ethnic groups for the African countries of interest. 14 The median value of Threat capabilities across all ethnic groups in our dataset is The median values of Group size and Centroid distance are 0.1 and 348, respectively. 15 What matters in expectation is not only the relative threat capabilities of the ruler and opposition but, when threat capabilities are asymmetric, whether it is the ruler or opposition that has the stronger threat capabilities. A pure continuous interaction term does not allow us to unpack this categorical effect, and risks inaccurately labeling asymmetric dyads with one exceptionally strong side as having mutually high threat capabilities. 21

22 As an initial plausibility check of the hypotheses, we calculate the frequencies with which each outcome occurs in each of these categories. The results, presented in Figure 3, are highly consistent with expectations. Interethnic powersharing occurred in 67.2% of high-high ethnicdyad-years, nearly twice as often as it did in other ethnic-dyad-year types. Also as predicted, coups are significantly more common, and rebellion onset significantly less common, in these cases. 22

23 Figure 2: Frequencies of Powersharing, Coups and Armed Rebellion Strong Opposition Weak Opposition Strong Ruling Group H1: P(Powersharing): 67.2% P(Successful coup): 1.4% P(Armed rebellion): 0.7% H2: P(Powersharing): 37.1% P(Successful coup): 0.0% P(Armed rebellion): 1.1% Weak Ruling Group H4: P(Powersharing): 43.7% P(Successful coup): 0.8% P(Armed rebellion): 1.4% H3: P(Powersharing): 33.0% P(Successful coup): 0.2% P(Armed rebellion): 1.6% To more conclusively evaluate the relationship, we estimate specifications containing a number of control variables that account for alternative explanations of elite bargaining, political violence, and governance outcomes: 16 GDP per capita: income-level and state capacity; (lagged) natural log of GDP per capita in that state-year. Log country population: population size; (lagged) natural log of the total population of the state. Log country area: territory size; natural log of the territory in square kilometers. 17 Number of ethnic groups: country s ethnic diversity; number of politically-relevant ethnic groups at independence. Institutionalized regime: institutionalization of political power and the stabilizing effects of party rule (Magaloni 2008); dummy measure of whether a given country is governed 16 Unless specified, variables come from the EPR 3.0 dataset. 17 Country area data from CIA World Factbook. 23

24 by a single-party regime, multi-party regime, or democracy versus military government (Magaloni et al. 2013). Former French colony: France s pro-active and interventionist foreign policies in their former colonies (Clapham 1996); dummy measure of a whether a given country was a former French colony. Cold War: increase in externally-brokered powersharing governments in Africa since the end of the Cold War; dummy measure of whether the year is after Years since last coup: previous successful coups (Londregan and Poole 1990); years since the last successful coup that occurred in the country (McGowan 2003). Ongoing rebellion: ongoing civil wars; whether there is an ongoing rebellion in the country in the previous year (Roessler 2011). Year: time trends; the year. 4.2 Results The statistical results provide strong support for the paper s central argument that mutually strong threat capabilities are necessary for powersharing. The models estimate the likelihood that members of a given non-ruling group are: included in the central government (Table 1); lead conspirators in a successful coup 18 (Table 2); or significant participants in initiating or joining an armed rebellion (Table 3). 19 Since HH dyads are of primary theoretical interest, we estimate models isolating this type (with all others combined as the reference category) as well as models including all other types individually (with HH omitted as the reference category) The results are virtually identical if we test for attempted coups. See Online Appendix. 19 The differences in the number of observations in the powersharing models compared to the coup and civil war models are due to the coup and civil war data only going up to Because we are interested in the balance of threat capabilities and the influence this has on the strategic relationships, rather than the individual threat capabilities of specific groups themselves, we do not include the separate, continuous threat capabilities variables in the specification. Including these individual capabilities scores for both ruling and opposition groups in robustness models does not affect the statistical or substantive results for our variables of interest. 24

25 As shown in Table 1, consistent with our first hypothesis, HH dyads are significantly more likely to feature interethnic powersharing (models EP-1, EP-3, and EP-5). Each other type of dyad individually (HL, LH and LL) is also (across most specifications) statistically significantly less likely to lead to inclusion when compared to HH dyads (models EP-2, EP-4, and EP-6). These comparisons particularly the lower likelihood of inclusion of weaker opposition groups as indicated by the coefficients for HL and LL dyads are consistent with the theorized bargaining dynamics. A group s inability to credibly threaten the center reduces horizontal accountability and an equitable distribution of power. The negative effect of the LH variable on ethnic inclusion suggests, further, that powersharing reflects not just a given opposition s threat capabilities to demand inclusion, but rather results from the strategic interactions of the ruling group and opposition group and their joint threat capabilities (though this effect appears less robust than the others and should be taken as tentative). 25

26 Table 1: Balance of Threat Capabilities and the Likelihood of Ethnic Powersharing HH threat capabilities HL threat capabilities LH threat capabilities LL threat capabilities Log GDP per capita Log country population Log country area Number of ethnic groups Institutionalized regime Former French colony Cold War Year EP-1 EP-2 EP-3 EP-4 EP-5 EP *** 0.58*** 0.54*** (0.29) (0.17) (0.18) -1.25*** -0.66*** -0.63*** (0.36) (0.19) (0.20) -0.98* (0.59) -1.44*** (0.41) -1.82*** t (0.18) t *** (0.01) t *** (0.00) constant (0.32) (0.30) (0.39) N states pseudo r *: p<0.10 **: p<0.05 ***: p< (0.36) -0.60*** (0.22) -1.82*** (0.18) 0.08*** (0.01) -0.00*** (0.00) (0.43) ** (0.20) 0.12 (0.18) (0.17) (0.06) 0.32 (0.29) 0.32 (0.39) 0.74 (0.48) 0.05** (0.02) -1.65*** (0.16) 0.07*** (0.01) -0.00*** (0.00) (38.66) (0.40) (0.46) 0.44** (0.21) 0.13 (0.18) (0.19) 0.02 (0.08) 0.36 (0.29) 0.34 (0.41) 0.73 (0.48) 0.05** (0.02) -1.65*** (0.16) 0.07*** (0.01) -0.00*** (0.00) (39.79) The second part of the argument posits that while mutually strong threat capabilities produce self-enforcing powersharing, they merely shift, rather than resolve, the coup-civil war trap that plagues weak, ethnically-divided states. Strong rivals commit to powersharing even as it opens the door to future coups d état. Relative to the prospects of a mutually costly civil war if any side was to try to rule unilaterally, trading power via coups is a preferable outcome. The ruling group knows that should the opposition successfully seize power, it will nonetheless be unlikely to risk 26