Differential longevity and redistribution: theoretical and empirical issues.

Summary In this article, we study the impact of differences in longevity on the design of public policies, particularly those related to retirement. First, we show that even though life expectancy has increased dramatically over the last century, large disparities remain. Second, we study from a normative point of view how differences in longevity are generally taken into account in life-cycle models and show that certain assumptions can have strong implications in terms of intragenerational redistribution. We identify at least three arguments in favor of redistribution to agents with low longevity: multi-period inequality aversion, mortality risk aversion, and compensation for characteristics for which agents are not responsible. We then extend our analysis to take into account the fact that individuals may be partly responsible for their longevity.Finally, we link these results to current debates on pension system reform. We show that, in general, because pensions are conditional on the survival of beneficiaries, public pension systems will redistribute resources from short-lived agents to long-lived ones. We provide suggestions for reforms that would aim to take better account of these differences in longevity and, in particular, those relating to the creation of a "longevity annuity" as desired by the Comité d'Amours and to the development of independent insurance, whether private or public.