High Impact, Low Probability? An Empirical Analysis of Risk in The Economics of Climate Change
To what extent does economic analysis of climate change depend on low-probability, high-impact events? This question has received a great deal of attention lately, with the contention increasingly made that climate damage could be so large that societal willingness to pay to avoid extreme outcomes should overwhelm other seemingly important assumptions, notably on time preference. This paper provides an empirical examination of some key theoretical points, using a probabilistic integrated assessment model. New, fat-tailed distributions are inputted for key parameters representing climate sensitivity and economic costs. It is found that welfare estimates do strongly depend on tail risks, but for a set of plausible assumptions time preference can still matter.