Heating Up the Dry Debate

The study evaluates the relationship between three drought indicators and communal conflict in Sub-Saharan Africa (1989-2014) using multilevel logistic regression. It finds that while all three measures are positively associated with conflict, the strength of the relationship increases as the measure moves from meteorological (SPEI) to agricultural (NDVI) and finally to socio-economic (EM-DAT) impacts. Specifically, socio-economic droughts approximately double the odds of experiencing violent conflict. However, the author notes that the NDVI effect may be driven by spatial correlation—where conflicts and vegetation loss occur in the same land types—rather than a direct temporal causal link. The study concludes that preventing meteorological droughts from escalating into socio-economic disasters could potentially reduce communal conflict risks.

• All three drought indicators (SPEI, NDVI, and EM-DAT) are positively associated with a higher likelihood of communal conflict across Sub-Saharan Africa, supporting the hypothesis that droughts generally increase conflict risk.
• Socio-economic drought, as measured by EM-DAT, has the most significant effect, approximately doubling the odds of experiencing violent conflict.
• Agricultural drought (NDVI) is significant in all main models, with odds ratios for conflict ranging between 1.39 and 1.92.
• Meteorological drought (SPEI) shows a weak and inconsistent relationship with conflict, becoming significant primarily when considering droughts in neighbouring areas.
• The relationship between NDVI drought and communal conflict is likely based on spatial correlation (occurring in the same types of land) rather than temporal correlation (occurring in the same years).
• There is a sequential relationship between the three drought types: lack of precipitation (SPEI) increases the likelihood of reduced vegetation quality (NDVI), which in turn increases the likelihood of a socio-economic disaster (EM-DAT).
• The level of democracy is a strong negative predictor of communal conflict, with higher democracy scores significantly decreasing the odds of conflict.
• Previous conflict in a cell is the strongest predictor of current conflict, increasing the odds of a conflict in year t by four to six times if a conflict occurred in year t-1.
• Ethnic exclusion, as measured by the GeoEPR 2014 dataset, was not found to have a significant effect on the likelihood of communal conflict in this study.

APA

Methi, F. (2020). Heating Up the Dry Debate. Peace Research Institute Oslo. https://cdn.cloud.prio.org/files/538d4bcc-c0ec-42a4-9d94-42831b945386/MA_Fredrik_Methi.pdf?inline=true

Chicago

Methi, Fredrik. Heating Up the Dry Debate. Peace Research Institute Oslo, 2020. https://cdn.cloud.prio.org/files/538d4bcc-c0ec-42a4-9d94-42831b945386/MA_Fredrik_Methi.pdf?inline=true.

Wikipedia

{{cite report |last1=Methi |first1=Fredrik |title=Heating Up the Dry Debate |publisher=Peace Research Institute Oslo |date=2020 |url=https://cdn.cloud.prio.org/files/538d4bcc-c0ec-42a4-9d94-42831b945386/MA_Fredrik_Methi.pdf?inline=true |access-date=23 October 2026 |via=Climate Insights Directory}}

BibTeX

@techreport{methi2020heating, author = {Methi, Fredrik}, title = {{Heating Up the Dry Debate}}, institution = {Peace Research Institute Oslo}, year = {2020}, url = {https://cdn.cloud.prio.org/files/538d4bcc-c0ec-42a4-9d94-42831b945386/MA_Fredrik_Methi.pdf?inline=true}, urldate = {2026-10-23}, note = {Indexed by Climate Insights Directory} }