Which Way to Peace? Estimating Directionality in Conflict Event Data

• In Event: Networked Peace? Network Approaches to Peace and Conflict

Abstract

If conflict is the result of incomplete information, then battles reveal important insights into each warring party’s strategy, capability, and resolve. Despite that violent conflict has often been conceptualized as a back and forth between at least two actors, very few studies test the role of battle directionality on conflict outcomes. This, we argue, has largely been a measurement problem, but one that produces significant consequences for theory. In this paper we argue that prolonged shifts in the balance of power, as evidenced by a governments’ offensive strategy, explain the timing of early concessions by armed groups. Namely, when governments sustain offensive attacks against challenger groups without tipping the scales of civilian support against themselves, non-state armed challengers are more likely to surrender. We present a novel latent variable framework that parses directionality from event data via a network decomposition approach. We apply this approach to the Nigerian Civil conflict from 1997 to 2019 to investigate the relationship between directionality, civilian support, and armed group concessions.

Authors

• Cassy Dorff, Vanderbilt University

• Shahryar Minhas, Michigan State University