Hi! I am glad to have gotten an invitation from John to guest-blog here. I am an Associate Professor of economics at the Andrew Young School at Georgia State University. My research on environmental economics is in solar geoengineering, health effects of air pollution and climate change, behavioral economics, and some other stuff. I teach environmental economics at the undergraduate and PhD level.
So let me start by highlighting some recent research that I've done with co-authors Nolan Miller and David Molitor of the University of Illinois. In this article in The Conversation, we describe the results from a recent working paper in which we use claims data from Medicare combined with weather monitor data to estimate 1) the effect of temperature extremes on elderly mortality, 2) how these effects differ across different regions of the country, and 3) the extent to which this can inform us about the effects of climate change and the potential to adapt to climate change.
We summarize our findings on the first question:
Our key finding is that both heat waves and cold snaps increase mortality rates. For example, the mortality rate from a day with average temperatures between 90 and 95 degrees Fahrenheit is higher by about 1 death per 100,000 individuals than a day with an average temperature between 65 and 70 degrees. Deaths also increase, by about one-half per 100,000 individuals, on days when the average temperature is less than 20 degrees.
There is also substantial heterogeneity in these effects across the country (FYI the editors of The Conversation didn't want us to use fancy words like "heterogeneity," but I trust that readers of this blog will be OK with them):
In hot places like Miami, cold days have a very large impact on mortality, while the impact of hot days is smaller. In contrast, hot days in Fargo have a very large impact on mortality, but an additional cold day has little effect. In fact, the effect of the hottest days (90 degrees or higher) in the coldest places is about two to three times larger than the effect of the coldest days (less than 20 degrees) in the hottest places.
Finally, we use the cross-sectional heterogeneity across regions to predict the scope for future climate change adaptation. Basically, as currently-cold places like Chicago warm up and start to look more like currently-warm places like Miami, climate-wise, those currently-cold places will begin to exhibit the temperature-mortality relationship that currently-warm places currently have. Make sense?
This graph summarizes our predictions on the effects of climate change, depending on whether we allow for regional heterogeneity or not and whether we allow for future adaptation or not. Ignoring heterogeneity (blue bars) makes it look like climate change will be bad for hot places but actually good for cold places. (This incidentally is what is presented in this recent study by Solomon Hsiang and co-authors, which got quite a bit of press.) But, once you account for regional heterogeneity (green bars), that's not true anymore - climate change is bad everywhere and even worse for cold places than for hot places. Lastly, when allowing for adaptation (gray bars), climate change isn't as bad.
As we emphasize in the working paper and in the Conversation article, this doesn't mean that adaptation is a silver bullet that solves all climate change problems. For one thing, we don't model the costs of adaptation, which could be substantial. Second, we don't consider other responses to climate change like abatement or geoengineering.
The two main takeaways are that 1) we need to carefully consider regional heterogeneity when modeling climate change impacts, and 2) we need to carefully consider adaptation when modeling climate change impacts.
