The underreported death toll of wars: a probabilistic reassessment from a survey with UCDP coders

Event datasets, such as those provided by the Uppsala Conflict Data Program (UCDP), provide high-quality data on conflict fatalities. However, such data are likely to suffer from an unknown extent of bias and uncertainties in the reports they are based on. Although a substantial literature documents reporting bias in conflict research, analyses that quantify this bias are mostly limited to single countries. Here, we combine a survey with UCDP coders and statistical modeling to derive a distribution of plausible number of fatalities given the number of battle-related deaths and the type of violence documented by the UCDP. We provide a generalizable, cross-national measure of uncertainty around UCDP reported fatalities that is more robust and realistic than UCDP’s documented low and high estimates, countering UCDP’s intrinsic tendency to under-estimate fatalities, and we make available a dataset and R package that can be applied to future releases of the UCDP data.