This article examines heterogeneity in Militarized Interstate Disputes data from 1816-2001.

Data & Methods

The Correlates of War dataset uses two coding logics to define disputes. One captures state-to-state militarized actions, while the other includes government protests against private citizen targeting by states. The study shows that these are distinct types and that using only protest-dependent cases distorts findings.

Key Findings

The authors demonstrate that omitting protest-dependent disputes provides more accurate estimates for key conflict predictors. Existing controls for dispute heterogeneity, particularly focusing on fatal conflicts, significantly underestimated the impact of contiguity (geographic proximity) between states and overestimated the pacifying effects of regime similarity.

Implications

Governments rarely escalate to militarized interstate disputes solely to protect private citizens in these specific cases. The study provides a detailed list of all identified protest-dependent cases, offering more precise data for future research.