Measure for Measure : How Well Do We Measure Micro-Level Conflict Intensity?

Rich measures of micro-level violent conflict intensity are key for successfully providing insight into the legacy of civil war. Yet, the debate on how exactly conflict intensity should be measured has just started. This paper aims to fuel this awakening debate. It is demonstrated how existing and widely available data - population census data - can provide the basis for a useful measure of microlevel conflict intensity, i.e. a fine Wartime Excess Mortality Index (WEMI). In contrast to measures that are based on news reports or data from transitional justice records, WEMI is relatively neutral to the cause of excess mortality, giving equal weight to victims belonging to the conquering and defeated party, to victims of large-scale massacres and dispersed killings, to victims in easily accessible locations and remote areas, and to direct and indirect victims of violence. The measure is illustrated for the case of Rwanda and it is shown that in a straightforward empirical application of the impact of armed conflict on schooling different measures for micro-level conflict intensity yield strikingly different results