African Election Seasonality: 145 Executive Results and 100,000 Surveys

Brian Engelsma (Harvard Dataverse) produced this dataset to study agricultural seasonality's effect on 145 executive-election results across 32 African countries. It combines satellite-derived harvest metrics with over 100,000 Afrobarometer survey responses to evaluate retrospective voting patterns. The data includes quasi-experimental evidence from two Zambian by-elections to account for election timing endogeneity.

Use Cases

• Analyzing the correlation between harvest timing and incumbent vote shares in African executive elections
• Evaluating sociotropic versus pocketbook voting mechanisms using Afrobarometer survey responses
• Modeling the curvilinear relationship between time-from-harvest and voter support

Strengths

• Covers 145 executive-election results across 32 African countries
• Includes over 100,000 individual Afrobarometer survey responses
• Incorporates objective agricultural measures derived from satellite imagery

Limitations

• Geographic coverage is restricted to 32 countries within Africa
• Relies on the Afrobarometer survey framework which may have its own sampling biases
• Potential for endogeneity in election timing, though partially addressed by by-election cases

Provenance

Source

Brian Engelsma, Harvard Dataverse

Collection Method

Compiled from satellite imagery, election records, and Afrobarometer surveys

Freshness

Last updated March 2026 according to source metadata.

Geography

32 African countries