Institutionalizing Peace: Power Sharing and Post-Civil War Conflict Management

We seek to answer the question, What effect does international war participation have on the ability of political leaders to survive in office? We develop a model of political reliability and derive seven related hypotheses from it that anticipate variation in the time a national political leader will survive in office after the onset of a war. Drawing upon a broadly based data set on state involvement in international war between 1816 and 1975, our expectations are tested through censored Weibull regression. Four of the hypotheses are tested, and all are supported by the analysis. We find that those leaders who engage their nation in war subject themselves to a domestic political hazard that threatens the very essence of the office-holdinghomo politicus, the retention of political power. The hazard is mitigated by longstanding experience for authoritarian elites, an effect that is muted for democratic leaders, while the hazard is militated by defeat and high costs from war for all types of leaders. Additionally, we find that authoritarian leaders are inclined to war longer after they come to power than democratic leaders. Further, democratic leaders select wars with a lower risk of defeat than do their authoritarian counterparts.

We know very little about how civil wars end. Harrison Wagner has argued that negotiated settlements of civil wars are likely to break down because segments of power-sharing governments retain the capacity for resorting to civil war while victory destroys the losers' organization, making it very difficult to resume the war. An analysis of a data set of 91 post-1945 civil wars generally supports this hypothesis but only in wars over identity issues. Moreover, while military victories may be less likely to break down than negotiated settlements of identity civil wars, they are also more likely to be followed by acts of genocide. Outsiders concerned with minimizing violence thus face a dilemma.