More about that Berkeley study...

The Economist has a good article up giving some more of the background of the Berkeley study that looked at the economic effects of global warming in a new way.  Here's part of it:

A paper published this week in Nature challenges this finding. The authors—Marshall Burke, Solomon Hsiang and Edward Miguel—suspected that economists had been looking for the wrong thing: a linear relationship between temperature and growth. Instead, they looked for an optimal temperature, on the assumption that excessive cold could harm growth as much as punishing heat. That is exactly what they found: hotter-than-usual years benefit countries, rich and poor alike, up to an average annual temperature of 13°C, after which hotter weather begins to sear growth. That allowed them to draw inferences about the likely effect of climate change: for Brazil, for example, an increase in temperature of 3°C will lead to a fall in output of 3% (see chart).

The apparent heat resistance of rich countries, it turns out, is simply because some of them, such as Germany and France, lie on the colder side of the optimum, so grow faster in hotter years, whereas others, such as America and Australia, lie on the hotter side, and so wilt as temperatures rise. Within individual counties in America, for instance, every hot day (with an average temperature over 24 hours of 24-27°C) lowers the average income per person that day by 20%, according to a working paper from the National Bureau of Economic Research by Mr Hsiang and Tatyana Deryugina. Very hot days (over 30°C) lower income per person by 28%. Looking at the average impact of rising temperatures in rich countries as a group had obscured such strong responses.

 David Appel shares my skepticism of (as he puts it) an economic model atop a climate model, but his post extracting some of the material from the paper is worth reading too.  (I added a comment to his post along the lines of what I said here.)