Conflict Deaths and Events
How many people were killed? It’s one of the first questions asked in the wake of military action, and it’s often surprisingly hard to answer. During active fighting, it’s impossible to document each death as it happens, especially civilian casualties in besieged areas. During long-running conflicts, official record keeping systems fail as society breaks down. Mass graves may be uncovered years after an event, if at all, and bodies are difficult to exhume and identify. Well-known numbers associated with historical conflicts tend to be historians’ best guesses, rough estimates, or politically motivated fabrications.
Today, Carnegie Mellon statisticians are working to develop rigorous methods to estimate conflict deaths and other conflict events. The work involves both matching records from disparate sources and estimating the number of unrecorded deaths. Much of this work is being carried out through a partnership between CHRS, the Human Rights Data Analysis Group (a nonprofit think tank based in San Francisco, CA) and the Carter Center’s Conflict Resolution Program.
Program manager Robin Mejia is engaged in an ongoing effort to estimate the number of child kidnappings during El Salvador’s civil war, partnering with Asociación Pro-Búsqueda de Niñas y Niños Desaparecidos.
In 2017, CMU PhD student Niccolo Dalmasso developed software to enhance the categorization and matching of name strings in Arabic, to support the work of the Human Rights Data Analysis Group, which is estimating the death toll from that conflict.
A 2009 workshop held at CMU highlighted areas of controversy and agreement in this domain and was the subject of the center’s first book project, Counting Civilian Casualties: An Introduction to Recording and Estimating Non-Military Deaths in Conflict (Oxford University Press, 2013).