本文檢視政府與武裝部隊的關係對內戰爆發的影響。政府所掌握的壟斷性武力是嚇阻與鎮壓叛亂的主要工具。武裝部隊是否願意服從政府指示，為政府奮力作戰，可用於衡量政府鎮壓內亂的能力，也正是潛在叛亂份子推斷起義是否能成功的重要考量。當武裝部隊擁有的自主性越高，或是部隊內部派系傾軋，都會削弱國家鎮壓叛亂的能力，潛在的叛亂份子若推斷政府無力號令武裝部隊，便可能趁機發動內戰。本文分析1946至2014年各國政府與其武裝部隊之間的關係，結果發現，若武裝部隊曾發生抗命、曾發動軍事政變、或內部有對立的情形，國內較容易發生武裝叛亂。本文結論指出，嚇阻武裝叛亂有賴於武裝部隊與政府的緊密結合，潛在叛亂份子對於政府動員武裝部隊的能力較為敏感，而非武裝部隊的數量或裝備。 This paper examines the relationship between a state's control over its security forces and the onset of civil war. A government-controlled force represents a state's capacity to deter and repress armed rebellion. Whether armed forces are willing to follow orders and fight for the government or not represents its capacity to repress rebellion. This is a main factor that is taken into account when rebels plan to initiate an armed conflict. More autonomous armed forces, or an internal split within the forces, could weaken a state's capacity for repression. Rebels are more inclined to begin rebellions under these conditions. Using the UCDP data between 1946 and 2014, this paper analyzes the relations between governments and their armed forces. The findings suggest that records of defection in the security sector, experiencing a coup d’état, or having divided security institutions increases the probability of civil war. The paper concludes that successful deterrence against armed rebellion depends on strong cohesion between a government and its armed forces. Rebels are more sensitive to the information of their government's capacity to mobilize rather than the size and equipment of the government forces.