This project analyzes spatial patterns of violence in civil wars. Civil wars are more than military clashes between rebels and a government. Instead, multiple forms of violence in which a myriad of actors are involved take place in the context of a civil war. Rebel groups fight each other, civilians are targeted, militias do the killing for someone else, ethnic groups clash, and gangs fight over illicit business. While this complexity is widely recognized, quantitative conflict research mostly studies different forms of political violence in isolation, rather than in their interrelationship.
In this project I develop and test a methodological approach that permits the quantitative study of relationships and shifts between different forms of violence in war. The approach solves a violence attribution problem: How do we know that a violent event is related to (i.e., takes place because of) a war if the perpetrators are not civil war combatants? After all, much collective violence also takes place outside civil wars. Moreover, some countries have multiple simultaneous rebellions, and an act of violence could be related to any of them. The problem is magnified if actors try to avoid attribution, as is the case for instance in unclaimed terror attacks.
A promising option is to attribute violence by its location rather than its actors: Violence is related to a civil war if it happens in the conflict zone of that war. While this spatial approach has the potential to facilitate more integrated research on different forms and actors of political violence during and after civil wars, it rests on a critical assumption, namely that most war-related violence indeed takes place in the conflict zone. The empirical part of this project scrutinizes this assumption.
As an important part of the project I collaborate with Mihai Croicu and Ralph Sundberg at my department to update and further develop the UCDP GED Polygons – conflict zone data (shapefiles) that describe the primary geographical area of armed activity for each conflict covered by the UCDP.
Bara, Corinne; Mihai Croicu and Ralph Sundberg. Mapping Conflicts: The new UCDP GED Polygon Dataset [work in progress]