Conflict and Structural Change

Conflict and Structural Change: Research Project Plan

1. Motivation and Research Question
This project seeks to understand the two-way relationship between violent conflict and structural transformation. Our main question is whether conflict prevents economies from undergoing structural change and sustained growth, or whether the absence of growth itself breeds conflict. Beyond establishing associations, we aim to provide a quantitative assessment of the importance of conflict for structural change and growth, disciplined by empirical findings.

2. Building on the Gravity of Violence
We build directly on The Gravity of Violence (Couttenier, Marcoux, Mayer & Thoenig, 2025), which develops a structural spatial framework linking trade, productivity, and conflict. The paper uses ACLED data on African conflicts (1997–2023), aggregated to ethnic homelands, and innovates by modeling violence as bilateral “flows” from group recruitment bases to attack destinations. The framework delivers a gravity equation for violence, estimated with PPML, showing sharp spatial decay and strong effects of ethnic and political borders. The model also allows for counterfactual analysis of how economic shocks alter the incidence and geography of violence.

This provides a natural starting point for extending the analysis to structural transformation, where violence may directly alter occupational and sectoral choices, and where uneven development may in turn fuel conflict.

3. Linking Conflict to Structural Transformation
The empirical task is to merge conflict data with measures of structural change. We will begin with IPUMS International, which provides microdata on employment shares across agriculture, manufacturing, and services. These can be aggregated to the country-year level, and—crucially—aligned with the regional breakdown used in the conflict data. The goal is to test whether regions exposed to higher levels of violence experience slower reallocation of labor out of agriculture and into higher-productivity sectors.

We will complement IPUMS with other indicators of local economic activity. These include nighttime lights, recently harmonized local GDP measures, and FAO agro-ecological suitability data. Together, they will provide multiple windows onto the economic costs of violence and the extent to which conflict traps regions in low-productivity equilibria.

4. Empirical and Quantitative Strategy
Our empirical analysis will combine panel regressions of sectoral employment and local economic activity with measures of violence intensity. Identification will draw on the spatial frictions estimated in the gravity model, as well as exogenous shifters such as crop price shocks. This approach follows the spirit of Gravity of Violence, but with outcomes extended to structural change.

On the modeling side, we will extend the gravity framework to incorporate sectoral occupational choice. Individuals may become farmers, fighters, or workers in non-agriculture, with relative productivities and security conditions determining the allocation. This will allow us to quantify how much violence delays structural change, and conversely, how accelerated transformation may reduce conflict.

5. Research Assistant Role
The RA will play a central role in data work: becoming familiar with the ACLED conflict data and its spatial aggregation, applying for and working with IPUMS microdata, and testing the feasibility of linking the two sources. Additional tasks include preparing simple empirical exercises, such as correlating violence exposure with structural change indicators, and exploring complementary data sources like nighttime lights and local GDP.