Temperature and civil conflict

Am. J. Poli. Sci.: https://doi.org/10.1111/ajps.12425

As recognition of the dangers posed by climate change increases, scholars have sought to estimate the effects of temperature anomalies on the risk of civil conflict. The logic is that temperature shocks increase the risk of political violence through mechanisms such as crop failures, resource scarcity and increased interpersonal aggression.

superclic / Alamy Stock Photo

A new study by Kenneth Schultz of Stanford University and Justin Mankin of Dartmouth College complicates this narrative by arguing that temperature cannot be seen as unaffected by conflict. The authors examine historical patterns of weather monitoring in sub-Saharan Africa and find that civil conflict is associated with the loss of weather observation stations, which increases measurement error in temperature records. For one commonly used meteorological dataset, conflict-associated gaps in coverage have created increasing downward bias in estimates of temperature anomalies over time. Reconsidering the relationship between temperature shocks and conflict using methods that reduce measurement error results in a significantly larger effect of temperature on conflict than uncorrected models imply.

These results highlight the need to consider the ways in which conflict affects the collection of information about the climate. Conflict cannot change the weather, but it does change what is known about it.