The Continuity axiom of expected utility theory implies that no catastrophe is so bad that one should not be willing to accept a gamble that might result in that catastrophe, as long as the gamble is sufficiently likely to instead result in some improvement on the status quo. This is implied by most normative decision theories, including risk-averse alternatives to expected utility theory. In this seminar, Joe Roussos argues that when there is significant uncertainty about the probabilities, these trade-offs are not warranted. Rather than opting for an ambiguity decision theory, he suggests that one can achieve the required caution by rejecting Continuity and using something like Steel, Bartha and DesRoches's lexical utility theory. However, unlike Bartha and DesRoches (2021) and Steel and Bartha (2023), Roussos argues that there are policy situations with catastrophic outcomes in which one should use expected utility theory: those with reliable probabilities.

Joe Roussos is a Pro Futura Scientia Fellow at Stockholm University and a researcher at the Institute for Futures Studies, where he leads the project Expertise in crises. Joe’s work spans decision theory, social epistemology and philosophy of science. His research focuses on the role of experts in policymaking, the management of scientific uncertainty, and modelling as a methodology. Joe received his PhD from the London School of Economics, with a thesis titled Policymaking under scientific uncertainty. His work has been published in the Australasian Journal of Philosophy, Philosophy of Science, WIREs Climate Change, and Ethical Theory and Moral Practice.