Many have speculated that drought set off the Syrian Civil War. We presented some evidence from Mexican history that supports the hypothesis ... but we also showed that the Mexican results were highly sensitive to how you defined “revolutionary activity.”
Is there more evidence that drought causes unrest?
Does the same pattern hold in modern Francophone West Africa? No! First, they find some correlation for rainfall but none for temperature. In fact, for temperature the effect is sometimes reversed. Second, it no longer matters whether the area was part of a 19th-century kingdom; that part of the past really does seem to be past.
I am not sure if the results have a direct application to Syria. The Syrian drought does appear to have been record-breaking; the West African shocks are not of the same magnitude. But they do imply that the correlation between unrest and drought is not one-to-one. They also provide a useful corrective to the recent spate of articles that in effect argue that history matters once: history matters, a lot, but its effects are not permanent.
* The baseline level of violent resistance was higher in areas that had been under a 19th-century African kingdom, but the nature of the excluded category is not clear, since stateless areas also exhibited more violence.