Do we really understand the age–VSL relationship?

Proposals to apply lower values to risk reductions for older adults have drawn attention to the theoretical and empirical findings describing the relationship between the value of a statistical life (VSL) and age. While the theoretical results are ambiguous and the empirical results mixed, the conclusion that VSL declines with age seems to have gained a status close to one of a professional consensus. This paper questions this rush to consensus by raising analytical and empirical concerns. We focus on the way in which individuals perceive multiple mortality risks and present both analytical and empirical results that are inconsistent with VSL declining with age.