A principal–agent theory and network analysis of high-end cooperation among militant groups

Abstract This study develops a principal–agent theory of why armed groups engage in high-end cooperation involving training or intelligence. While such cooperation enhances capacity, it generates downstream risks. Principals mitigate these risks by selecting agents who are skill-wise complementary but politically non-substitutable. Agents accept high-end support when doing so does not leave them politically overshadowed by a dominant actor. Using original network data on 53 groups in Northeast India (1981–2021) and temporal exponential random graph models, I show that high-end cooperation is more likely in dyads characterized by organizational capacity differentials, complementary attack portfolios, and distinct constituencies. The findings recast militant alliances as products of competitive threat perception and carry broader implications for the study of multiparty conflicts.