Session 4 Damages


a) The aggregate reduced-form approach to measuring damages. The idea here is to look at aggregate indicators, such as GDP, in order to see how climate affects them: simply run regressions with an aggregate like GDP (or some transformation of it) on the LHS and climate (or some form of it) on the RHS. GDP focuses on the damages to an economy’s productive capacity. Obviously, there are other aggregate measures, such as mortality.

Another important distinction is whether the variation in the RHS variable is over time or across countries/regions. Your task is to discuss both. A good reference for the former is the 2005 National Academy of Sciences piece by Nordhaus, “Geography and macroeconomics: new data and new findings”. A good reference for the latter is the set of recent papers by Dell, Olken, and Jones (some published, some not).

As for mortality regressions, there are several recent papers here as well, among them our IIES paper by Kudamatsu, Persson, and Strömberg and a paper by Burgess and Greenstone; Dell, Olken, and Jones also look at this outcome variable in their AER Papers and Proceedings paper.

b)The bottom-up approach to measuring damages The idea here is to add up damages estimates that are of a microeconomic nature, and this approach normally includes attempts to measure the value of lost lives, less straightforwardly defined damages such as a decrease in biodiversity, etc. The best source here is likely Nordhaus’s book (with Boyer), “Warming the world”, though it would be important to search for updates and other papers as well. One additional source is the Stern report, which also attempts to measure even less obvious implications of climate such as an increased prevalence of conflicts (due to, say, migration pressures in damage-ridden areas) both within countries and between countries. You don’t need to read the entire report but you should look at the overview: the listing of damages of different sorts and how this list compares to Nordhaus’s.

c) Damages Damages due to climate change can affect the economy in different ways, examples are damages to output, to capital stocks and effects directly on utility and life expectancy. Also political instaility and similar more indirect effects may be important. Make a list of how damages have been modelled and construct a simple dynamic model where we can introduce these different damages. Show how different damages are affected by growth and what discount rate to use for different damages. You may for simplicity use a simple two-period model. Examples of damages can be found in Nordhaus books, the Stern report. For the discounting discussion, you might want to look at Sterner&Persson “An even Sterner Review”