One of the triumphs of the twentieth century has been the extension of life. Paradoxically, living too long is one of the major concerns of the elderly. What does it mean to """"""""live too long""""""""? For the elderly, two critical concerns are health and the adequacy of financial assets. Although average life expectancy has increased dramatically, for any particular person the future remains uncertain -- both in terms of health and of the length of life itself. From this perspective, living too long means living in poor health and/or with inadequate resources to sustain prior consumption levels, the worst case being a long life of ill health and poverty. Health issues have been largely ignored in the literature on asset management and uncertain life. However, health is critical to saving and consumption decisions in several ways. First, it affects the capacity to work and earn income. Second, it affects the quality of life and the value of purchased goods. Third, although Medicare covers large portions of medical costs for the elderly, illness can still entail substantial out-of-pocket expenditures, especially for long term care at older ages. Finally, a change in current health status is a signal which carries information about future health and mortality. The main issues which concern us are: 1. Under the existing Social Security and Medicare systems, how severe are the consequences of falling into poor health or losing a spouse? 2. How well prepared are the elderly for their retirement years in terms of accumulated wealth and private and social insurance instruments? 3. What is the trade-off between private and public insurance? Has the recent decline in private pensions been compensated, in welfare terms, by changes in social security? 4. Is social insurance provided efficiently, or is it possible to increase the welfare of the elderly by changing the mix of annuities and medical coverage within a given government budget? The proposed research focuses on the impact of health on the saving and consumption decisions of retirees. The decision unit which we investigate is a husband-wife household which pools the resources of its surviving members. We use a dynamic programming approach to model the household's planning problem. The main components of the optimization model are: I) A probability transition matrix for different health survival states of the spouses; 2) A payments matrix which describes the monetary costs and gains which are associated with each transition; 3) A state dependent utility function for each living partner where the marginal utility from consumption varies with health; 4) A utility function defined over assets which are left as bequests when the last survivor dies. Given information on these components it is possible to calculate the optimal consumption saving strategy. However, some components, such as the utility parameters are not directly observable. We estimate these parameters by confronting the optimal plan with observed asset accumulation, recognizing that wealth is measured with error. For this purpose we use three different samples -- the Retirement History Survey, the Health and Retirement Survey, and the Panel Study of Income Dynamics. Having recovered the preference parameters of households, we can examine for each household the implied path of asset decumulation in the post retirement years. We can then infer the expected utility (welfare) of each household conditioned on publicly provided social security and medical insurance and privately held instruments such as pensions and life insurance and thus determine the public/private tradeoff.
| Lillard, L A; Panis, C W (1996) Marital status and mortality: the role of health. Demography 33:313-27 |
