Climate and Conflict: A Comment on Hsiang et al.’s Reply to Buhaug et al.

The authors argue that a fixed effect-based meta-analysis is inappropriate for studying the climate-conflict relationship due to extreme causal heterogeneity across different climatic conditions, conflict types, and spatial scales. While acknowledging a minor coding error, they maintain that the distribution of effects across studies contradicts the claim of a robust, singular aggregate relationship between climate and conflict.

• The authors argue that the fixed effect model used by Hsiang et al. (2013) is inappropriate because it assumes causal homogeneity, which is violated by the diversity of the data. The sample includes a wide array of social phenomena (from urban riots to civil war), climatic events (from heat waves to ENSO cycles), and spatial scales (from municipalities to the entire world).
• The authors contend that there is no single 'true' effect connecting various climatic conditions and forms of conflict across different spatiotemporal scales. Consequently, they argue that the estimated average effect reported by Hsiang et al. lacks substantive meaning.
• The authors acknowledge a coding error identified by HBM where one result from O’Loughlin et al. (2012) was omitted when estimating the aggregate temperature effect. However, they state that correcting this error has little influence on the overall interpretation of the temperature-conflict pattern.
• The authors caution that a tendency toward positive effects in current literature does not prove that sustained warming will increase future civil conflict risk. They note that the surveyed literature focuses on climatic variability rather than climate change, and it is not certain that future societies will respond to normal weather in the same way they currently respond to anomalies.
• The authors suggest that selection bias may be present in climate-conflict research, potentially driven by 'norms of conformity' that reward specific results. They argue that sampling based on reported results rather than original datasets may accentuate this bias.

APA

Nordkvelle, J., & Buhaug, H. (2014). Climate and Conflict: A Comment on Hsiang et al.’s Reply to Buhaug et al. Peace Research Institute Oslo. https://cdn.cloud.prio.org/files/eb3b5662-7997-446a-806d-1ae0b65cf305/Buhaug and Nordkvelle - Climate and conflict PRIO Paper.pdf?inline=true

Chicago

Nordkvelle, Jonas, and Halvard Buhaug. Climate and Conflict: A Comment on Hsiang et al.’s Reply to Buhaug et al. Peace Research Institute Oslo, 2014. https://cdn.cloud.prio.org/files/eb3b5662-7997-446a-806d-1ae0b65cf305/Buhaug and Nordkvelle - Climate and conflict PRIO Paper.pdf?inline=true.

Wikipedia

{{cite report |last1=Nordkvelle |first1=Jonas |last2=Buhaug |first2=Halvard |title=Climate and Conflict: A Comment on Hsiang et al.’s Reply to Buhaug et al. |publisher=Peace Research Institute Oslo |date=2014 |url=https://cdn.cloud.prio.org/files/eb3b5662-7997-446a-806d-1ae0b65cf305/Buhaug and Nordkvelle - Climate and conflict PRIO Paper.pdf?inline=true |access-date=23 October 2026 |via=Climate Insights Directory}}

BibTeX

@techreport{nordkvelle2014climate, author = {Nordkvelle, Jonas and Buhaug, Halvard}, title = {{Climate and Conflict: A Comment on Hsiang et al.’s Reply to Buhaug et al.}}, institution = {Peace Research Institute Oslo}, year = {2014}, url = {https://cdn.cloud.prio.org/files/eb3b5662-7997-446a-806d-1ae0b65cf305/Buhaug and Nordkvelle - Climate and conflict PRIO Paper.pdf?inline=true}, urldate = {2026-10-23}, note = {Indexed by Climate Insights Directory} }