Problems of Attributing Civil War Violence: Evaluating Data Quality

• In Event: Qualitative Field Research, Research Quality & Research Transparency

Abstract

Despite growing research on civil wars, still relatively little attention has been given to some of the most basic building blocks of armed conflict: who are the perpetrators of violence? This question is often more complex than it appears. Armed groups masquerade as Boko Haram in Nigeria, in Mali it remains unclear whether Al-Qaeda or local militias stand behind some attacks, and splintering factions of the Sudan Peoples Liberation Army obscure attribution in South Sudan. Available datasets, theories of armed conflict, and policy interventions hinge on the accurate attribution of violence. And yet, scholars have rarely matched the methodology of studying armed conflict with the reality of this complexity. This paper develops a framework to evaluate the evidentiary basis of data from conflict zones in order to improve measurement procedures. Methodologically, it examines the types of evidence used to draw inferences on the actors, timing, and events of war. What constitutes good data in conflict zones? What shared standards can be used in evaluating evidentiary quality and rigor behind sources? Political Science typically privileges causal analysis, but research demands equally rigorous standards to draw these data-level and descriptive inferences. The paper draws on original data sources including internal rebel records unearthed in the field, qualitative interviews, newspapers, and other supporting material documentation. Empirically, it applies the framework to a set of mass killings that erupted in northeast Democratic Republic of the Congo in 2014 that have held considerable costs for civilian lives and international resources. It evaluates 245 original interviews and direct documentary evidence, and compares the results against standard international analysis of the killings. It demonstrates that different evidentiary protocols hold consequences for what perpetrators are identified behind attacks, the mechanisms of how violence is produced, and the policies recommended to deescalate violence.

Author

• Rachel Sweet, University of Notre Dame