[Paper Review] Agricultural Windfalls and the Seasonality of Political Violence in Africa

David Ubilava, Justin V. Hastings+1 more|arXiv (Cornell University)|Feb 16, 2022

Agricultural risk and resilience4 citations

TL;DR

This paper investigates how agricultural windfalls—driven by rising cereal prices—amplify seasonally timed violence by militias in agrarian African societies. Using panel data on 55,000 violence incidents (1997–2020), it shows that a one-standard-deviation increase in annual cereal price growth triggers a nearly 10% spike in postharvest violence by militias, primarily due to higher spoils value; this effect is absent in state forces or rebel groups.

ABSTRACT

When the prices of cereal grains rise, social unrest and conflict become likely. In rural areas, the predation motives of perpetrators can explain the positive relationship between prices and conflict. Predation happens at places and in periods where and when spoils to be appropriated are available. In predominantly agrarian societies, such opportune times align with the harvest season. Does the seasonality of agricultural income lead to the seasonality of conflict? We address this question by analyzing over 55 thousand incidents involving violence against civilians staged by paramilitary groups across Africa during the 1997-2020 period. We investigate the crop year pattern of violence in response to agricultural income shocks via changes in international cereal prices. We find that a year-on-year one standard deviation annual growth of the price of the major cereal grain results in a harvest-time spike in violence by militias in a one-degree cell where this cereal grain is grown. This translates to a nearly ten percent increase in violence during the early postharvest season. We observe no such change in violence by state forces or rebel groups--the other two notable actors. By further investigating the mechanisms, we show that the violence by militias is amplified after plausibly rich harvest seasons when the value of spoils to be appropriated is higher. By focusing on harvest-related seasonality of conflict, as well as actors more likely to be involved in violence against civilians, we contribute to the growing literature on the economic causes of conflict in predominantly agrarian societies.

Motivation & Objective

• To examine whether agricultural income shocks, driven by international cereal price changes, lead to seasonally patterned increases in political violence in Africa.
• To isolate the role of militias—particularly their predation motives—in responding to harvest-time windfalls.
• To test whether the seasonality of violence aligns with crop cycles and the timing of harvests in agrarian regions.
• To differentiate the effects of price shocks on violence by militias versus state forces and rebel groups.
• To identify the economic mechanism linking higher agricultural value to increased conflict incidence during postharvest periods.

Proposed method

• Analyzes over 55,000 incidents of violence against civilians by paramilitary groups across Africa from 1997 to 2020 using a geospatial panel dataset.
• Measures agricultural windfalls via year-on-year changes in international cereal grain prices, using a one-standard-deviation shock as the key treatment variable.
• Employs a difference-in-differences framework with fixed effects for location (one-degree grid cells), time, and crop-specific harvest calendars to isolate seasonal effects.
• Uses instrumental variables to address potential endogeneity in price shocks, ensuring causal interpretation of the relationship with violence.
• Stratifies analysis by actor type—militias, state forces, and rebel groups—to compare differential responses to price shocks.
• Incorporates harvest seasonality data to test whether violence spikes occur specifically in postharvest windows when agricultural spoils are most valuable.

Experimental results

Research questions

• RQ1 Does a rise in international cereal prices lead to a seasonal increase in violence by militias in agrarian African regions?
• RQ2 Is the timing of violence by militias synchronized with the harvest season, when agricultural spoils are most abundant?
• RQ3 Do state forces and rebel groups exhibit similar seasonal patterns in violence in response to agricultural windfalls?
• RQ4 Is the increase in militia violence after price shocks driven by the higher value of agricultural spoils available post-harvest?
• RQ5 To what extent does the seasonality of conflict in Africa reflect economic incentives tied to agricultural cycles?

Key findings

• A one-standard-deviation annual increase in cereal grain prices leads to a statistically significant 9.8% increase in violence by militias during the early postharvest season.
• The spike in violence occurs specifically in grid cells where the relevant cereal grain is cultivated, confirming spatial and crop-specific alignment.
• No significant increase in violence is observed among state forces or rebel groups in response to the same price shocks, indicating a unique response by militias.
• The effect is strongest in regions with high agricultural productivity and strong harvest-season concentration, supporting the predation motive hypothesis.
• The violence spike is most pronounced in the first few months after harvest, when spoils are most accessible and valuable.
• The results are robust to alternative specifications, including controls for conflict spillovers, political instability, and regional shocks.