Death as a Measure of Duration of Conflict

Abstract

This paper introduces a new measure of conflict duration and argues that the number of deaths in a conflict can serve as such a measure. The paper demonstrates that there are information gains to this approach. The well-known conflict database of the International Peace Research Institute is compared with the database of the Center for Systemic Peace, which includes data on the number of deaths in addition to length of conflict. The number and distribution of conflicts vary; however, duration analysis based on the conventional measure of duration yields results that are robust over the above-mentioned datasets. We also show that the number of deaths, as a measure of duration, challenges some of the results based on the number of years as a measure of duration. In the 1990s, the duration of conflicts is significantly different from before when the duration measure is the number of years - we do not find a statistically significant distinction when the number of deaths is used. Ethnic conflicts have a longer survival time in terms of time but not in terms of number of fatalities.