When Civil Wars Roam: Why Some African Conflicts Shift Location

Why Location of Fighting Matters

Kyle Beardsley, Kristian Skrede Gleditsch, and Nigel Lo ask why some African civil wars concentrate their violence in a few stable places while others appear to ‘roam’ across the countryside. The geographic pattern of fighting shapes humanitarian needs, civilian displacement, rebel–community relations, and the strategies available to governments and outside actors.

What the Authors Test

The authors evaluate three mechanisms that could drive conflict-zone movement: whether rebels fight primarily for a specific ethnic constituency; whether they receive outside military assistance; and whether they lack the fighting capacity to confront government forces in fixed locations. The core claim is that non-ethnic, externally backed, or militarily weak groups will be more geographically inconsistent in where they fight.

Measuring Movement With Georeferenced Event Data

Beardsley, Gleditsch, and Lo develop new measures of conflict-zone movement using shifts in conflict polygons derived from the Uppsala Conflict Data Program’s Georeferenced Event Dataset (UCDP GED). Their approach operationalizes how much the geographic footprint of violence changes over time and links those shifts to group-level characteristics.

Key Findings

• Empirical tests generally support the authors’ hypotheses: groups without strong ethnic ties and those with weaker military capacity are likelier to fight in varied locations.
• Outside military assistance is associated with more geographically mobile fighting, consistent with the idea that external support enables operations across a wider area.
• Groups that lack both strong local ethnic bases and sufficient conventional strength are especially prone to ‘roving’ behavior rather than holding stable territory.

Why This Changes Expectations

These results refine expectations about the humanitarian footprint of African armed conflicts and the interdependence between rebel groups and local populations. They also highlight a policy dilemma for counterinsurgency: mobile, non-ethnic, or externally supported groups are harder to contain and protect civilians from, complicating both military responses and relief efforts.