The conclusion of the Cold War has had contradictory effects on on-going conflicts in the former superpower client states of Ethiopia, Somalia and Angola. The argument proposed in this dissertation seeks to explain these variations in conflict. Two conditions, distribution of power and governing agreement, are considered to be fundamental to this explanation. In many cases, disputants have had access to arms in the past and their ability to pursue violent means to their objectives in the future is contingent on their continued access to weapons or other resources. These patterns are reflected in the distribution of power. The second condition, governing agreement, reflects the fact that in the immediate post-Cold War era, disputants have frequently sought to resolve their differences through negotiations. What is relevant here is whether any subsequent political pact exists between disputants and whether it involves power-sharing or a winner-take-all process. These two conditions, military viability and governing agreement, are combined in a matrix to suggest how their interaction leads to divergent outcomes. Depending on the mix of conditions, the outcomes range from "stable peace" to "tense peace or low-level conflict" to "civil war." The argument is applied to the three cases from 1985 to 1995. The author concludes that winner-take-all elections, though often the preference of African leaders, can be a recipe for disaster in tightly contested states with a history of violent conflict. On the other hand, while power-sharing agreements are difficult to arrive at (and even more difficult to implement), agreements leading to governments which are inclusive of the main disputants are an important step in the mitigation of conflict. By building security and creating the sense that both parties have a stake in the continued functioning of a political system, armed movements may be more willing to relinquish their weapons and embrace peace.

No description available.