So as promised, I read the piece by Burke et al. They claim to have found a correlation between temperature increases and the onset of civil conflict in most of Africa. The mechanism is that hot weather messes up crop yields and therefore increases the likelihood of conflict (especially in places where people depend on rain-fed agriculture). This conclusion is based on the findings of a tight correlation between economic underdevelopment and civil wars. Nice and dandy, if you believe that people fight because they are poor. Sure, the opportunity costs are much lower for the poor aggrieved who oftentimes than not choose the conflict route to settling disputes. But state capacity, in my view, has a much greater influence on whether people choose to fight or not.

The paper’s policy prescriptions are even dodgier. The authors recommend that foreign aid be conditioned on projected adverse effects of climate change. Firstly, this “solution” is based on the premise that greater proportions of Africans will continue to depend on agriculture into the foreseeable future. This might be true, but shouldn’t we be trying to expand African economies and reducing dependence on agriculture (which necessarily forces us to deal with issues of governance)? Secondly, the idea that foreign aid should be conditioned on climate change is just, well, silly. Many a failed development initiative on the Continent can be blamed on the erratic nature of foreign aid. Adding more variance by pegging aid flows to climate changes will only make things worse.

For a more refined critique see Chris Blattman’s Blog.