ABSTRACT

Conflict is the primary cause of hunger in many of the world’s major food crises, driving food and nutrition insecurity to unprecedented heights. Food insecurity escalates when conflict forces people to flee their homes, farms, and jobs, and when it blocks access to critical humanitarian aid. Yet the empirical studies analysing the impacts of conflict on food security have limitations in terms of their estimation methodology and the consistency of their findings. This paper seeks to fill in this gap by providing robust estimates of the impact of conflict on the likelihood of food insecurity. Using rich household-level data recently collected in Yemen and merging them spatially with detailed conflict events data, we estimate the impact of three different definitions of conflict (battles, explosions and total fatalities) on three different common measures of food insecurity. We also interact the conflict variables with a conflict exposure variable defined as the number of months over the past five years in which the household experienced the type of conflict being measured. Our empirical results show that longer exposure to conflict, rather than the contemporaneous impact of conflict incidents, leads to higher food insecurity.