Leadership Changes and Civil War Agreements: Exploring Preliminary Links

1 Introduction

Leadership changes among state and non-state actors pose opportunities for policy change, in particular with respect to decisions regarding peace and war. Studies focusing on interstate conflicts suggest that leadership overturns in states often lead to shifts in UN voting (Mattes et al. 2015). Furthermore, the coming to power of new leaders is suggested under certain conditions to render states engaged in conflicts more conducive to negotiated settlements (e.g. Bercovitch and Lutmar 2010; MacGillivay and Smith 2008; Quiroz-Flores 2012). However, within the context of intrastate disputes we argue that this relationship is more complex and multilayered due to factors such as multiplicity of groups, competing interests, legitimacy issues, and revolutionary goals, frequently found among rebel groups. We suggest that changes in leaders of rebel groups is likely to lead to more, rather than less, hardline conflict positions, at least in the short term. Subsequently, when it comes to leadership changes among sides engaged in civil war, these complexities will lend to a relationship that is less straightforward and pronounced between leadership change and conflict resolution than that which appears to exist in interstate conflicts, particularly when the leadership change is on the side of a rebel group. In this article we explore the impact of leadership change in rebel groups on their conflict behavior and address possible processes precipitated by the overturn of leaderships in civil conflicts. We test our hypothesis on a dataset of leadership changes and agreements ending civil wars in Africa, 1975–2007. A discussion of the results is followed by directions for further research.

2 Leadership changes during conflict

Research focusing on interstate conflict has produced a growing body of leader-specific research that explores the influence of leadership changes on foreign policy decisions. Numerous studies suggest that negotiation processes might be prompted by domestic shifts such as the rise of new leaders, the emergence of a divided leadership, or a split in a government previously unified in its war aims (Mitchell 2000, 89; Stedman 1991). Several explanations are proposed to account for this effect. First, internal political changes expose strategies and ways out of a situation which might have been overlooked or ignored by the previous leadership (Greig 2001; Stedman 1991). Secondly, even cases in which the current leadership supports negotiations, a change in leadership might be needed to pull away from failed conflict policies (Lieberfeld 1999; Mitchell 2000; Stedman 1991). Thirdly, new leaders are not necessarily committed to the policies of their predecessors or held accountable to them. Therefore, they will find it easier – and politically less costly – to “change course” (Mitchell 2000, 89–90). In fact, the very advent of a new leader will often produce, or result from, an expectation for a policy shift, thus creating an environment in which new strategies are easier to pursue. Finally, adversaries are less likely to harbor distrust against new leaders than against their predecessors, making it easier to engage in negotiations. Thus, periods of leadership change may signal the potential for the development of a peaceful conflict resolution process.^[1]Bercovitch and Lutmar (2010) find statistical support for this argument when tested against a dataset covering 221 interstate conflicts that took place in the 1945–2000 period. They find that conflicts in which one or both of the participating countries experienced leadership changes are likely to be conducive to the initiation of negotiations and mediation, and that this relationship is stronger in democracies than in non-democracies.

Yet whereas research on leaderships and leadership changes has flourished in the context of interstate relations, a similar shift to leadership-specific research has not occurred in the scholarship on intrastate war. This neglect is particularly puzzling considering the fact that the majority of conflicts taking place in the international system since the end of the Cold War constitute civil wars. A sizable body of literature has emerged on the outbreak, severity and duration of intrastate disputes (e.g. Fearon 2004; Fearon and Laitin 2003; Hegre et al. 2001; Metternich 2011; Regan 2002; Sambanis 2001; Walter 2015). A similarly impressive body of literature has focused on the conditions under which settlements ending civil war are reached, and the conditions that promote the durability of these settlements (e.g. Badran 2014; DeRouen et al. 2009; Fortna 2004; Greig and Regan 2008; Hartzell and Hoddie 2003; Hartzell et al. 2001; Licklider 1995; Walter 2002). However, the role of leaders and leadership shifts in these intrastate processes has drawn limited attention, and to date has not been put to statistical testing.^[2]

3 Civil wars and rebel group leadership change – research hypothesis

A key motivation for researching conflict processes is to gain a better understanding into the conditions that might lead to the de-escalation of conflict once violence has erupted. Within this context, internal conflicts pose considerable challenges for peaceful conflict resolution. Civil conflicts carry potentially grave consequences – for domestic populations, neighboring states, and regional or even global stability. Not only do a high number of civil conflicts result in complex humanitarian crises, but rarely do they remain contained within the borders of the conflict state. In many cases, either massive refugee flows are prompted by the violence and/or the conflict spills over national boundaries. In other cases, allies are drawn into the conflict transforming the conflict into a multiparty internationalized confrontation. The high costs and potential for spread involved in intrastate disputes render such conflicts events in which internal and external parties might be expected to have a high interest in resolving, or containing at the very least. Yet, at the same time, intrastate disputes are particularly difficult to resolve peacefully. They often concur with a revolutionary struggle and involve individuals who place principles before interests, thereby rendering a compromise particularly difficult to negotiate (Olson and Pearson 2002; Stedman 1991). This lends to conflicts that tend to be more complex, drawn out and intractable than interstate wars, consequently entailing particularly high negotiation costs and risks.

Thus, there is a need to gain a better understanding of the dynamics involved in civil conflicts and of the conditions that might promote their resolution by peaceful means. Within this context the role played by leadership changes among rebels groups and the government side constitute important information both for decision makers among the disputing parties and for third parties considering the prospects of intervention.

Because civil conflicts often spawn a multiplicity of groups, alliances, and sub-groups, some of the complexities entailed in ending civil conflicts as compared to interstate disputes are linked to the number of groups involved and competing interests among their leaders. Instead of clearly defined and legitimate representatives of governments that partake in interstate negotiations, leaders of these groups are not always officially recognized, are often reluctant to engage in any formal negotiations and are motivated themselves by complex and often contradicting interests. New leaders, might not only be reluctant to break away from such fluid circumstances but might also seek to bolster support of the group members and strengthen cohesion among them by emphasizing intransigent positions both vis-à-vis the government and vis-à-vis competing groups, rather than flexibility. When these groups constitute rejectionist movements, splinter groups, or spoilers, intent on derailing any progress made in negotiations new leaders are even more unlikely to shift to positions that allow for compromise. Although rendering negotiations a particularly difficult endeavor such actors cannot be ignored. ^[3] Thus we expect leadership changes under conditions of multiplicity of competing groups and sub-groups to lead to greater inflexibility, at least in the short-term. Another leader-specific factor that may impinge on a peaceful impact of leadership change is that in many intrastate conflict situations, rebel groups, especially their leaders, profit personally from the ongoing conflict especially when natural resources are used to fuel the dispute (Collier 2000; Collier and Hoeffler 2004). This implies that neither the current nor the succeeding leader would have a strong incentive to reach a settlement. While governments involved in interstate conflict may as well benefit from conflict as well, such benefits are typically in the realm of public support (such as in a rally around the flag effect), do not last long and are not likely to survive over leadership shifts.

The above suggests that when it comes to leadership changes among sides engaged in civil war, these complexities will lend to a relationship that is less straightforward and pronounced between leadership change and conflict resolution than that which appears to exist in interstate conflicts, particularly when the leadership change is on the side of a rebel group. Even in cases where the new leaders identify an interest in ending the conflict peacefully, new rebel leaders will find it more difficult and politically risky to change course than new government leaders. One explanation for this is that new rebel leaders often lack the necessary legitimacy-base, political apparatus and resources that new government leaderships – both in democratic and non-democratic regimes – possess and that are needed to implement policy shifts without destabilizing their power base. Thus, the complexities involved in civil war will lead to different, and more sophisticated, causal mechanisms between leadership change and negotiations processes.

4 Leadership change and peace agreements in Africa – data and empirical results

In this section we empirically examine the impact of leadership shifts on intrastate postwar dispute resolution outcomes in Africa, 1975–2007. We chose to focus on Africa because of the considerable number of civil wars that have taken place in that region, and the multiple peace agreements and ceasefires that have been reached in those civil wars along the years. The data on rebel leader changes was taken from a new dataset on rebel leader changes in civil wars, compiled by the authors. For government leader-change data we rely on the Archigos data base on leaders (Goemans et al. 2009). Data on civil war peace agreements and cease-fires is taken from the UCDP Peace Agreement Dataset (V2).

Our analysis focuses on whether a peace agreement was reached between the government and rebel group(s) following a leadership change on either the government side or within any of the rebel groups involved in the conflict. In order to capture both short and longer term influence, we look at three different timeframes: peace agreements reached within 6 months of the leader change, within 12 months, and within 18 months. Our subset contains 57 civil wars. During the period covered 90 peace agreements were reached, suggesting the recurrence of war in many cases leading to subsequent agreements. Rebel group leaders changed only 39 times as compared to 59 government leadership shifts.

During the first 6 months following a leadership change on either the government or rebel side, five peace agreements were reached, and in the 6 months after that an additional five agreements were signed. In the last 6 months examined one additional settlement was reached. Thus 11 peace agreements were reached within 18 months of a leadership change on either one or both sides, which is 12.2% of the total peace agreements reached. In other words, the majority of civil war settlements (87.8%) were reached more than 18 months after a leadership change, indicating a weak relationship, if any at all, between leadership change in civil wars and peace settlements.

When we look at the impact of leadership changes that occur on the rebel side only, this trend is even more pronounced. Whereas government leadership change prompted the signing of five peace agreements within 6 months of the leader change, rebel leader changes did not prompt peace agreements at all in the short term. No peace settlements were reached within 6 months of a leadership change on the rebel side. This lends support to our claim that rebel leader changes might lead to less flexible strategies, at least in the short term.

We then ran a multinomial logit to test the influence of leadership change on the type of agreement reached at the end of civil war. Adopted form the UCDP dataset, the dependent variable has three values: 1=peace agreement, 2=cease-fire, 3=victory and other possible outcomes such as victory of one side, and missing data. We include as control variables regime type (given in the state’s W score) ^[4] and GDP ^[5] We collapsed the dataset such that the observation is the conflict year (rather than the conflict month used above), recording the highest ranking agreement in cases in which more than one agreement was reached during the course of a single conflict year. A total of 32 peace agreements and cease fires are included in the dataset. Tables 1 and 2 present the results of the multinomial logit model.

Table 3 presents the marginal effects of the leadership change variable.

As Table 3 demonstrates, the average marginal effect of leadership change on reaching a peace agreement is significant but very small (0.0040174). The probability of reaching a peace agreement is on average about 0.4 percent higher when there is a leadership change on either the government side or the rebel side than for cases when there is no leadership change. Similarly, we find that the probability of reaching a ceasefire agreement is on average about 0.3 percent higher when there is a leadership change on either the government side or the rebel side than for cases when there is no leadership change.

5 Conclusions

In contrast to theorizing and findings based in the interstate dispute literature, the preliminary results presented in this paper suggest that in civil wars the relationship between leadership shifts and ending conflict is more complex than it is found to be in interstate disputes, and may discourage, rather than promote, peace. In this sense, leadership changes, especially when they take place on the side of the rebels, may lead, at least initially, to more astringent, rather than moderate, conflict positions. An examination of the more long term effects of leadership change in rebel groups may shed further light on this process.

Also, a better understanding of other factors involved in rebel groups and leadership changes such as the effects of different types of leadership changes (violent or non-violent) or whether the rebel group was motivated by territorial or political objectives may provide further insight into the different ways by which leadership changes in rebel groups may impact a conflict’s course. Clearly, these issues warrant further theoretical and empirical attention. A deeper understanding of the impact of leader shifts in civil war will provide important information for policy makers among the conflicting sides regarding the conditions under which the prospects for successful conflict resolution are highest in civil wars. The findings will also provide important information for third parties considering the option of triggering or supporting a leadership change.