Annuity markets and retirement security

Abstract

This paper describes the role of adverse selection in annuity markets, and sketches some of the public policy implications of the existence of such selection effects. It shows that a substantial fraction of the difference between the expected value of the payouts on both voluntary and compulsory annuity products in the U.K., and the premium cost of those products, is attributable to adverse selection. This is simply the fact that the individuals who currently choose to purchase annuities are on average longer-lived than randomly-selected individuals in the U.K. population. Adverse selection is most pronounced in the voluntary annuity market, but there is also some evidence of selection in the compulsory market, where individuals can choose which type of annuity to purchase and how much of their defined contribution balance to annuitize.
Requiring all persons to annuitize their retirement account balances at a specified age is one way to substantially reduce the degree of adverse selection in the annuity market. More generally, however, any policy that encourages a large fraction of the population to participate in the annuity market is likely to have a similar effect. Doyle, Mitchell, and Piggott (2001) compare the annuity markets in Australia and Singapore, and they find a greater degree of adverse selection in the former than the latter. They attribute this difference to the relatively generous government old-age safety net in Australia, which reduces the fraction of households that find it attractive to purchase private annuities.
Whether government policy should compel participants in defined contribution retirement saving programs to annuitize all or part of their accumulated balance, and whether there should be restrictions on the set of annuity policies that individuals can purchase, is an important and controversial policy issue. There are key tradeoffs associated with any policy choice. Compelling annuity purchase reduces the chance that some elderly households will outlive their resources, and it reduces the degree of adverse selection in the annuity market. Compulsion may also limit the choices available to retirees, and force some retirees to purchase annuity products that are substantially different than the products they would choose in an unconstrained market. Developing models that can evaluate the welfare costs of adverse selection and of limiting consumer choice in a world with heterogeneous consumer preferences should be a central goal for future work.