Are Non-Inclusive Peace Agreements Effective Counter-Insurgency Strategies?

ABSTRACTCivil wars are complex in ways that challenge effective resolution. Civil war actors tend to be dynamic in nature and often splinter then coalesce over time potentially evolving into multiple dyads pitted against their government. Previous work has demonstrated that when multiple rebel factions emerge, civil wars tend to be longer in duration as satisfying multiple factions tends to be more challenging. However, governments may choose to pursue dyadic agreements hoping to end the conflict either through subsequent dyadic negotiations or through more targeted military tactics focusing intently on the remaining faction(s). This study suggests that which dyads will be excluded from the peace is predictable, but also seeks to better understand what happens to those excluded factions that hedge their bets to continue the fight. Propositions are tested on all civil war peace agreements between 1945-2013. Key findings indicate the importance of political power-sharing provision and third-party assistance in bringing excluded groups into the fold.KEYWORDS: civil warpeace agreementsrebel dynamicspower-sharingmediationforeign military interventioninclusivityUN peacekeeping Disclosure StatementNo potential conflict of interest was reported by the author(s).Notes1 Small and Singer, A Resort to Arms: International and Civil Wars, 1816-1980; Gleditsch et al., “Armed Conflict 1946-2001: A New Dataset.”2 Olson Lounsbery and Pearson, Civil Wars: Internal Struggles, Global Consequences; Olson Lounsbery and Cook, “Rebellion, Mediation and Group Splintering: An Empirical Investigation of Competing Hypotheses”; Christia, Alliance Formation in Civil Wars; Nilsson, “Partial Peace: Rebel Groups Inside and Outside of Civil War Settlements.”3 Cunningham, “Veto Players and Civil War Duration.”4 Hampson, Nurturing Peace: Why Peace Settlements Succeed or Fail; Licklider, “The Consequences of Negotiated Settlements in Civil Wars, 1945-1993”; Ohlson and Söderberg, “From Intra-State War to Democratic Peace in Weak States”; Rubin, “The Actors in Negotiation.”5 Stedman, “Spoiler Problems in Peace Processes.”6 Johnston, “Negotiated Settlements and Government Strategy in Civil War: Evidence from Darfur.”7 Nilsson, “Partial Peace: Rebel Groups Inside and Outside of Civil War Settlements.”8 Zartman, “The Timing of Peace Initiatives: Hurting Stalemates and Ripe Moments”; Zartman and de Soto, Timing Mediation Initiatives.; Zartman, Ripe for Resolution: Conflict and Intervention in Africa.9 Walter, “Designing Transitions from Civil Wars – Demobilization, Democratization, and Commitments to Peace.”10 Hartzell and Hoddie, Crafting Peace: Power-Sharing Institutions and the Negotiated Settlement of Civil Wars; Pearson et al., “Replicating and Extending Theories of Civil War Settlement”; DeRouen, Jr., Lea, and Wallensteen, “The Duration of Civil War Peace Agreements”; Glassmyer and Sambanis, “Rebel-Military Integration and Civil War Termination.”11 Krause, Krause, and Bränfors, “Women’s Participation in Peace Negotiations and the Durability of Peace”; Gizelis, “Gender Empowerment and United Nations Peacebuilding”; Hudson, “Peacebuilding through a Gender Lens and the Challenges of Implementation in Rwanda and Co^te d’Ivoire”; Kew and Wanis-St.John, “Civil Society and Peace Negotiations: Confronting Exclusion”; Paffenholz, “Civil Society and Peace Negotiations: Beyond the Inclusion-Exclusion Dichotomy.”12 Wallensteen, Understanding Conflict Resolution: War, Peace, and the Global System; Fearon, “Why Are Some Civil Wars Last So Much Longer than Others?”; DeRouen, Jr. et al., “Civil War Peace Agreement Implementation and State Capacity”; Joshi and Quinn, “Implementing the Peace: The Aggregate Implementation of Comprehensive Peace Agreements and Peace Duration after Intrastate Armed Conflict”; Brancanti and Snyder, “Time to Kill: The Impact of Election Timing on Post-Conflict Stability.”13 Licklider, “Obstacles to Peace Settlement”; Ohlson and Söderberg, “From Intra-State War to Democratic Peace in Weak States”; Rubin, “The Actors in Negotiation.”14 United Nations, “‘Building Peace Can Take a Generation, but Together We Can Get It Right,’ Says Secretary-General in Security Council Debate on Mending War-Torn States.”15 Nilsson, “Turning Weakness into Strength: Military Capabilities, Multiple Rebel Groups and Negotiated Settlements.”16 Zartman, Ripe for Resolution: Conflict and Intervention in Africa.17 Cunningham, “Divide and Conquer and Concede: How Do States Respond to Internally Divided Separatists?”18 Nilsson, “Partial Peace: Rebel Groups Inside and Outside of Civil War Settlements.”19 Stedman, “Spoiler Problems in Peace Processes.”20 Stedman.21 Pearlman, “Spoiling Inside and Out: Internal Politial Contestation and the Middle East Peace Process.”22 Christia, Alliance Formation in Civil Wars.23 Atlas and Licklider, “Conflict Among Former Allies After Civil War Settlement: Sudan, Zimbabwe, Chad, and Lebanon.”24 Stedman, “Spoiler Problems in Peace Processes”; Blaydes and DeMaio, “Spoiling the Peace? Peace Process Exclusivity and Political Violence in North-Central Africa.”25 Greenhill and Major, “The Perils of Profiling Civil War Spoilers and the Collapse of Intrastate Peace Accords.”26 Most and Starr, Inquiry, Logic, and International Politics.27 Most and Starr; Siverson and Starr, “Opportunity, Willingness, and the Diffusion of War.”28 Cunningham, Gleditsch, and Salehyan, “Non-State Actors in Civil Wars: A New Dataset.”29 Cunningham, Gleditsch, and Salehyan.30 Cook and Olson Lounsbery, Conflict Dynamics: Civil Wars, Armed Actors, and Their Tactics.31 Olson Lounsbery and Cook, “What’s on the Table? Bargaining Progression Amid Civil War.”32 “Philippines Signs Long Awaited Peace Deal with Rebels.”33 “’Fight Was Leading Us Nowhere’: Former Abu Sayyaf Militants Speak after Surrender to Philippines Forces.”34 Olson Lounsbery and DeRouen, Jr., “The Roles of Design and Third Parties on Civil War Peace Agreement Outcomes.”35 Cil and Prorok, “Selling Out or Standing Firm? Explaining the Design of Civil War Peace Agreements.”36 Melin and Svensson, “Incentives for Talking: Accepting Mediation in International and Civil Wars”; Johnston, “Negotiated Settlements and Government Strategy in Civil War: Evidence from Darfur.”37 Lintner, Burma in Revolt: Opium and Insurgency Since 1948.38 Bauzon, “The Philippines: The 1996 Peace Agreement for the Southern Philippines: An Assessment.”39 Gurr, Why Men Rebel; Regan, Sixteen Million One: Understanding Civil War.40 Doyle and Sambanis, Making War and Building Peace: United Nations Peace Operations.41 Regan, Civil Wars and Foreign Powers: Outside Intervention in Intrastate Conflict; Regan and Aydin, “Diplomacy and Other Forms of Intervention.”42 Cunningham, “Blocking Resolution: How External States Can Prolong Civil Wars.”43 Greig and Diehl, International Mediation; Olson Lounsbery and DeRouen, Jr., “The Viability of Civil War Peace Agreements.”44 Greig and Diehl, International Mediation.45 Clayton and Dorussen, “The Effectiveness of Mediation and Peacekeeping for Ending Conflict.”46 DeRouen, Jr. and Chowdhury, “Mediation, Peacekeeping and Civil War Peace Agreements.”47 Feith, “The Aceh Peace Process: Nothing Less than Success.” Special Report 184.”48 There were 15 UN peacekeeping missions initiated between 1948 and 1988. Comparatively, there have been 56 missions initiated between 1989-2017, according to the UN.49 Doyle and Sambanis, Making War and Building Peace: United Nations Peace Operations; Fortna, Does Peacekeeping Work? Shaping Belligerents’ Choices After Civil War. Examples of current missions in rather challenging civil wars include those in the Democratic Republic of Congo and the Darfur region of Sudan.50 Doyle and Sambanis, Making War and Building Peace: United Nations Peace Operations.51 Högbladh, “Peace Agreements 1975-2011 – Updating the UCDP Peace Agreement Dataset.”52 Davies, Pettersson, and Öberg, “Organized Violence 1989-2021 and Drone Warfare”; Gleditsch et al., “Armed Conflict 1946-2001: A New Dataset.”53 Kreutz, “How and When Armed Conflicts End: Introducing the UCDP Conflict Termination Dataset.”54 Findings using this refined coding approach are presented in the Appendix (Table 7). Table 6 in the same Appendix reports findings that predict termination through Peace Agreement or Abatement. Both Tables report findings like to the original coding scheme as reported in the findings section of the paper.55 In fact, in our sample of 65 excluded factions, only two groups were able to terminate through their own victory.56 Cunningham, Gleditsch, and Salehyan, “Non-State Actors in Civil Wars: A New Dataset.”57 We recognize that political power-sharing mechanisms are more likely in conflicts over government, whereas territorial power-sharing mechanisms are more appropriate for conflicts over territory (i.e., they are inversely related to one another). As a result, we examine these variables in separate models to avoid collinearity issues.58 Högbladh, “Peace Agreements 1975-2011 – Updating the UCDP Peace Agreement Dataset.”59 Högbladh.60 Olson Lounsbery and DeRouen, Jr., “The Viability of Civil War Peace Agreements.”61 While we do not have any specific expectations regarding time, we considered the possibility that time since a dyad’s exclusion may influence outcome, particularly whether a dyad remains ongoing. To account for this possibility, we controlled for time since exclusion (i.e., the peace agreement). Findings were consistent with those presented. Table 8 in the Appendix includes models with this time variable, which itself is not statistically significant in any of the models indicating that other factors (e.g., power-sharing mechanisms and mediation presence) are more important predictors than time.62 Davies, Pettersson, and Öberg, “Organized Violence 1989-2021 and Drone Warfare.”63 Olson Lounsbery and DeRouen, Jr., “The Viability of Civil War Peace Agreements.”64 Meier et al., “External Support in Armed Conflicts: Introducing the UCDP External Support Dataset (ESD), 1975-2017.”65 Guo and Fraser, Propensity Score Analysis: Statistical Methods and Applications. As a robustness check, the results from Heckman selection models are provided in Table 4 of the online Appendix. Our findings using logit do vary somewhat from those of the Heckman selection models. Nonetheless, we present the separate logit models as our findings because doing so allows us to compare outcomes across excluded versus included dyads. Further, we were unable to reject the null hypothesis that p = 0 (the level of correlation between the two error terms) and must conclude that utilizing the selection model is not appropriate.66 We recognize that multinomial logistic regression is the preferred approach for predicting dyad outcomes given the categorical nature of the variable structure. Using such an approach involves the estimation of several models simultaneously, which depends on the number of categories of the outcome variable. As a result, a large sample size is required (see “Multinomial Logistic Regression: Stata Data Analysis Examples,”) which has been noted as a minimum of 10 cases per independent variable (Starkweather and Moske, “Multinomial Logistic Regression.”) Our sample not only falls below that minimum requirement, but fails to produce model statistics confirming a degrees of freedom problem. Subsequently, we test our hypotheses we use logistic regression to model particular outcome categories.67 All analyses were completed using Stata 17. The replication files can be found at icpsr.umich.edu.68 Cunningham, “Veto Players and Civil War Duration.”69 Findings that estimate multinomial logistic regression do suggest that weaker groups are more likely to end in government military victory than to end via a negotiated outcome (see Table 8 in the online Appendix) but given the limited number of cases and resulting degrees of freedom problem noted above, we rely on the logistic output presented here.70 Territorial provisions did not achieve statistical significance in any of the models. These findings are reported in Table 5 of the Appendix.71 These findings are confirmed in the multinomial logistic estimations noted in the online Appendix (Table 8).72 Full predicted probability shifts are included in the Appendix (Tables 2b-d).73 Sisk, Power Sharing and International Mediation in Ethnic Conflicts; Papagianni, “Power Sharing, Transitional Governments and the Role of Mediation”; McColloch and McEvoy, “The International Mediation of Power-Sharing Settlements.”