Rainfall Variability and Violent, State-Based Conflict

The research examines how the effect of anomalous rainfall (excess and scarce) on violent, state-based conflict varies based on societal vulnerability factors. Using a causal forest model on global grid-cell data, the author finds that while rainfall variability generally has a small effect on conflict, this effect is context-specific. Scarce rainfall is more likely to increase conflict risk in communities with high agricultural employment and poor governance (low SHDI and GDP). Excess rainfall's effect is more complex, showing higher risk in areas with high population density, low globalization, low democracy, and low GDP, but contrary to expectations, it showed smaller effects in areas with high political exclusion or greater distance to the capital.

• The effect of rainfall variability on violent, state-based conflict is generally small across all studied contexts, but it is confirmed to be context-specific and dependent on the vulnerability of the community.
• Scarce rainfall is most likely to increase the probability of violent, state-based conflict in communities characterized by high agricultural employment and low levels of good governance, specifically low Subnational Human Development Index (SHDI) and GDP.
• The factors that modify the relationship between scarce rainfall and conflict differ from those that modify the relationship between excess rainfall and conflict, suggesting they operate through separate mechanisms.
• Contrary to theoretical expectations, high levels of political exclusion and greater distance to the capital were associated with smaller effects of excess rainfall on conflict.
• The analysis identified that some vulnerability factors, such as temperature and the prevalence of non-state conflict, did not have a significant impact on the treatment effect in either the scarce or excess rainfall models.
• The study encountered a significant violation of the overlap assumption, where propensity scores for receiving the treatment (SPEI3) were heavily skewed toward zero, necessitating the exclusion of a large portion of the data to avoid biased estimates.

APA

Mjelva, M. B. (2020). Rainfall Variability and Violent, State-Based Conflict. Peace Research Institute Oslo. https://cdn.cloud.prio.org/files/52a825cc-c6c8-4c4e-9d38-a3e84afd59af/MA_thesis_Mjelva_2020.pdf?inline=true

Chicago

Mjelva, Mathilde Bålsrud. Rainfall Variability and Violent, State-Based Conflict. Peace Research Institute Oslo, 2020. https://cdn.cloud.prio.org/files/52a825cc-c6c8-4c4e-9d38-a3e84afd59af/MA_thesis_Mjelva_2020.pdf?inline=true.

Wikipedia

{{cite report |last1=Mjelva |first1=Mathilde Bålsrud |title=Rainfall Variability and Violent, State-Based Conflict |publisher=Peace Research Institute Oslo |date=2020 |url=https://cdn.cloud.prio.org/files/52a825cc-c6c8-4c4e-9d38-a3e84afd59af/MA_thesis_Mjelva_2020.pdf?inline=true |access-date=23 October 2026 |via=Climate Insights Directory}}

BibTeX

@techreport{mjelva2020rainfall, author = {Mjelva, Mathilde Bålsrud}, title = {{Rainfall Variability and Violent, State-Based Conflict}}, institution = {Peace Research Institute Oslo}, year = {2020}, url = {https://cdn.cloud.prio.org/files/52a825cc-c6c8-4c4e-9d38-a3e84afd59af/MA_thesis_Mjelva_2020.pdf?inline=true}, urldate = {2026-10-23}, note = {Indexed by Climate Insights Directory} }