The creation of durable peace following armed conflicts has been widely researched from a variety of perspectives. There is much less research, however, concerning when and why mediation can produce durable peace because most mediation research focuses on achieving a short-term success as indicated by the creation of a new peace agreement. This is an exploratory study which examines several factors considered to be important for the creation of durable peace. This study finds that the two most important factors are the power dynamics between the parties and the management of spoilers. Moreover, this study finds that these two factors are interlinked inasmuch that changes to the parties' levels of power can facilitate the emergence of spoilers. These findings are based on the systematic examination of mediation in four cases of armed conflict by utilizing a modified contingency model of mediation which is tested against the mediations conducted in the 1973 Egyptian-Israeli war, the Bosnian war, the third Angolan war, and the first Chechen war. This study argues that a well-designed agreement can shift the power dynamics between the parties so that their struggle for power will not take violent forms, and it can help prevent the emergence of new spoilers because it does not favor one party more than the other. Well designed agreements can be created even when the balance of power between the parties is unequal, and efforts to further weaken the already weaker party should be avoided because it can contribute to the emergence of spoilers from within the disputing parties. An original model for durable peace which accounts for these new findings is then developed. This model argues that to create durable peace mediators must produce good agreements that are balanced and channel the struggle for power into nonviolent mechanisms and processes, and manage the spoilers who threaten the peace.

Thèse--Genève. / At head of title: Université de Genève, Institut universitaire de hautes études internationales. Bibliography: p. [177]-180.

Abstract Obviously, inside mediation through individuals or institutions emerged from the conflict situations becomes an alternative way to end armed conflict as it is regarded to have more local legitimacy with a higher sense of ownership. Surprisingly, researchers in peace and conflict studies have shown more interest to mediation processes dominated by outside mediators, while little has been known about the conditions that contribute to success of inside mediators. This thesis aims to contribute to this understudied research field by answering the question; under what conditions do mediators contribute to the ending of armed conflict in fragile states? In this thesis, I focus on the role of customary leaders, a form of inside mediators. The point of departure of this research question will be the distinction between the mediator’s involvements (Inside vs outside) and how their presence contribute to different outcomes. The variances will be explained with reference to two contending concepts of success and failure of mediation process. Success of mediators’ involvement can be measured by focusing on the following three areas; the initiation of peace process, conclusion of peace process and the sustainability of peace outcomes. The causal argument suggest that mediation processes dominated by inside mediators are more likely to end armed conflicts, as they have  more local legitimacy that can  engender a higher level of ownership, and sustain peace agreements as the actors feel engaged to the peace process. By testing this theory, this thesis applies the structured focused comparison method by selecting three cases of Southern, northwest and northeast regions in Somalia that have developed differently. The empirical findings of this thesis supports the hypothesis testes, as mediation processes dominated by inside mediators in Northeast and Northwest regions displayed a higher level of local ownership and legitimacy than the Southern regions. Finally, further researches on inside mediators success in a different countries is suggested in the future in order to know the level of inside mediators’ effectiveness in ending armed conflict.