Pierre E. Biscaye (UC Berkeley)

Agricultural shocks and long-term conflict risk: Evidence from desert locust swarms

Pierre E. Biscaye
UC Berkeley

Résumé 

How do transient agricultural shocks affect the risk of conflict over time? I study this question for the case of exposure to a desert locust swarm, effectively an agriculture-specific natural disaster, across 0.25 degree grid cells in Africa and the Arabian peninsula from 1997-2018. Using difference-in-differences and event study approaches I find that locust swarms lead to consistent increases in the probability of violent conflict over the following 14 years. Average long-term effects are large: an increase of 0.8 percentage points (43%) in the annual likelihood of experiencing any violent conflict. I find similar patterns for impacts on measures of conflict over output and over factors of production and on non-violent protests. Cereal yields are lower after cells are affected by a locust swarm, consistent with a wealth mechanism decreasing permanent income and agricultural productivity and lowering the opportunity cost of fighting. Effects of swarm exposure are concentrated in years with groups actively engaged in violent conflict in neighboring cells: when low opportunity cost of fighting is combined with opportunities to fight. Patterns of long-term impacts on violent conflict are similar for severe droughts indicating the mechanisms are not specific to locust shocks. The wealth mechanism is a potentially important channel relating economic shocks to the long-term incidence of violent conflict.

Lire le papier