Ending Civil Wars: A Case for Rebel Victory

This article is a good survey of civil war termination and how civil war termination affects the prospects for enduring peace and democratization. The main finding of the article is that, contrary to the conventional wisdom, civil wars ending with negotiated settlements are more likely to recur and that civil wars ending with a rebel victory are more likely to produce enduring peace. 
Scholars and practitioners of peace tend to believe that the best way to end a civil war is through a negotiated settlement. It is believed that through negotiations, former rebels will be given a say in the reconstruction of the state. This will lead to a more democractic polity, and therefore more prosperity in the long run. Is this true? Toft, tests this common assumption by examining how civil wars ended so far, how they ended and what they resulted in. Toft believes that stable environment and flourishing society in the early years of a civil war torn country can be misleading and that we should analyze the impact of the way civil war ends in the long run. 
The authors questions a popular belief about negotiated settlements. In defense of negotiated settlements, the most common argument is that negotiated settlements stop violence and reduce the number of death compared to ongoing fighting until victory is achieved. Toft believes that there are two problems with this logic. First, negotiated settlements are used by rebels to reorganize, recover and rearm. Then, negotiated settlements decrease the number of death in the short run but increase in the long run, costing even more death and damage. Second, she argues that measureing the cost of fighting only in terms of human life is misleading. What about condemning the `saved` population to desultory and miserable living standards. Is this not part of the cost? Therefore, she argues that we should pay at least equal attention to what civil war outcome brings about in the short and long run. 
The author analyzes all civil wars fought between 1940 and 2007. She sets some criteria to select cases for the dataset and 137 cases are qualified. Of these cases 118 cases experienced no violence for at least five years and considered ended civil wars. According to her dataset, the recurrence of wars from 1940 to 2000 ending in military victories was 12 percent (10 of 81); 22 percent (5 of 23) in negotiated settlements; and 31 percent (4 of 13) in cease-fires/stalemates. 
According to her findings, civil wars ending in negotiated settlements are twice as likely to recur as civil wars ending in victory. Civil wars ending in rebel victories lead to more stable post-conflict environment than government victories. She also finds that 17 percent (8 of 48) of civil wars ending in government victory recur, whereas only 6 percent (2 of 33) of civil wars ending in rebel victory recur. In short, whether a civil war ends in a victory matters as much as who achieved the victory. 
The author also finds that victory is more preferable to a negotiated settlement because the former reduces the risk of civil war recurrence, while the latter increases the probability. `Victory reduces the likelihood of civil war recurrence by 24 percent, relative to all other types of civil war termination`. On the other hand, `negotiated settlements increase the chances of recurrence by 27 percent, relative to all other types`. 
Is a civil war ending in a negotiated settlement more likely to result in more death in the longer term than if the combatants fight until one side achieves victory? To assess this argument, she uses four measures: total deaths, total deaths per capita, total battle deaths, and total battle deaths per capita. Her results show that civil wars ended through negotiated settlement resulted in significantly more deaths (as measured by total deaths per capita). This finding is statistically significant. 
Findings of the article show that civil wars ending in negotiated settlements are more likely to recur and to last longer. Moreover, failed negotiated settlements ignite more deadly wars. Therefore, normative argument common in the literature that negotiated settlements prevent loss of lives is not supported. On the contrary, her results favor rebel victories for enduring peace and avoiding civil war recurrence. 
The author also examines the relationship between civil war outcome and post-war politics, and the relationship between negotiated settlement and regime type. Her results show that there is no difference between democracies and authoritarian regimes with regard to signing negotiated settlements. When she disaggregates the data she finds that civil wars ending in rebel victory occurred within more authoritarian states. 
What about post-war period? Which type of civil war outcome promises more democracy? The results show that a state that ended its civil war through a negotiated settlement is likely to observe increase in democratization in the early years of postwar period. However, negotiated settlements are associated with greater levels of authoritarianism as the postwar period extends, after five years. Interestingly, her results show that states whose civil wars ended with rebel victories perform better in democracy. Therefore, negotiated settlements may be a very democratic process but it does not lead to more democratic politics. 
Finally, she examines the relationship between type of civil war settlement and economic prosperity. She finds that how a civil war ends is not correlated with economic growth. 
In short, this is a good article presenting a compelling argument against advantages of negotiated settlement.