FAT-TAILED UNCERTAINTY, LEARNING, AND CLIMATE POLICY

Abstract

Low-probability, high-impact risks are critical features of climate change economics; however, there are many unanswered policy and modeling questions about the implications of fat-tailed uncertainty. This paper examines the impact of fat-tailed uncertainty about the climate sensitivity on abatement decisions using a sequential decision-making framework. The results demonstrate how policy prescriptions from integrated assessment models are sensitive to the specifications of uncertainty, learning, and damages. Fat tails alone do not merit immediate and stringent mitigation but require strongly convex damages and slow learning. The analysis illustrates the potential value of midcourse corrections on reducing consumption risks imposed by uncertain damages from climate change and focuses attention on the dynamics of learning.