Science Figures Interpreted and Analyzed by RealClearScience
It is widely believed that climate change will lead to violent conflict, for instance, over dwindling water and food supplies. However, a recent paper in PNAS casts some doubt on this.
Using a complicated statistical analysis, researchers analyzed the impact of precipitation and temperature on conflict in east Africa. (See below.)
The above panels measure conflict as a function of precipitation (Panel A) or temperature (Panel B). The dashed line represents long-term average (mean) precipitation or temperature, and the X-axis shows changes in these variables (as reported by standard deviations from the mean). The Y-axis represents conflict.
Panel A, therefore, shows that as precipitation increases, conflict decreases. The authors note that when rainfall is 2 standard deviations above average, conflict is reduced by about 30%. However, drier conditions did not result in a statistically significant change in conflict.
Panel B shows even more mixed results. When temperature is 1 standard deviation above average, conflict drops by about 12%, but increases by about 30% when the temperature is 2 standard deviations above average. There was no change in conflict for cooler weather.
Though the authors found a statistically significant connection between rainfall/temperature and conflict, they warned that other factors (e.g., politics, economics, current events) are usually more important and better at predicting conflict. Thus, while climate change is likely to have real effects on human behavior, its contribution to global conflict should not be exaggerated.
Source: John O’Loughlin, Frank D. W. Witmer, Andrew M. Linke, Arlene Laing, Andrew Gettelman, and Jimy Dudhia. "Climate variability and conflict risk in East Africa, 1990–2009." PNAS October 22, 2012, doi: 10.1073/pnas.1205130109.