The Consequences of Negotiated Settlements in Civil Wars, 1945-1993

Although the title of the article sounds that this is an examination of consequences of negotiated settlements, the article actually is a review and test of several arguments about civil war termination. He challenges some widely-accepted arguments in the literature. He constructs a dataset consisting of 91 civil wars between 1945 and 1993. He constructed the data set by using various data sets constructed earlier. By using data he gathered, he challenges those widely-accepted notions about civil war termination and negotiated settlements. 
He opposes the view that civil wars last long and they end in negotiated settlements rather than victory of one party. According to data presented by Small and Singer, out of 44 civil war between 1940 and 1980, only 4 (10%) are listed as in progress. Therefore, civil wars tend to end in one way or another. Licklider’s data also show that 84 civil war cases (27%) ended during the same calender year in which the civil war began. 
Another widely-accepted argument about negotiated settlements is that they minimize the casualty because without a negotiated settlemetns belligerents would kill more people. Licklider argues that this theoretical argument is false because we cannot know what would have happened if the war have continued. Moreover, his data show that casualty patterns are similar for wars that end in negotiated settlements and military victories. Therefore, his data prove that the argument that negotiated settlements reduce casualty is a myth. 
After a civil war ends, people who attempted to kill each other have to live together within the same political entity. Therefore, negotiated settlements envisage a sort of power sharing agreement. Such power sharing is expected to force former belligerents to cooperate and work in harmony. However, Licklider argues that this theoretical argument is not true because negotiated settlements with power sharing agreements will create “internal balance-of-power situations that make it difficult for the new government function effectively. Military victory, on the other hand, will destroy the organizational structure of one side, making a resumption of the civil war much more difficult.” Hence, civil wars ending with negotiated settlement including a power sharing agreement is likely to resume in the future. He further argues that post-civil war environments require significant structural changes so that grievances that caused the war in the first place can be redressed. Power sharing agreements, however, will create veto group that can block such reforms. He also notes that victory by one side does not necessarily mean that the winner will definitely pursue structural reforms but that if intended, it will be easier to implement reforms. 
The literature on civil wars suggest that wars over ethnic/religious identities are more difficult to solve, bloodier and longer than civil wars fought over ideologies or revolutionary ones. Threfore, we should expect that identity civil wars are less likely to end in negotiated settlements, last longer and cause more casualty. However, Licklider’s data show that civil wars fought over identity do not last longer. He finds that “of the 57 civil wars which had ended, 5% of the 37 concerned with identity lasted longer than 10 yearsö as compared to 15% of the 20 political-economic wars”. He concludes that civil wars that are concerned with identity are not more intense than civil wars concerned with political/economic motives. 
In addition to these findings, the article makes two more important points. First, the most favorable option for an enduring peace in a civil war fought over identities is military victory, not a negotiated settlement. Second, military victories may be followed by mass atrocities, massacres or genocides because the winner may wish to annihilate the opposition so that the