Anticipated and Actual Bequests

This paper uses data on anticipated bequests from two waves of the Health and Retirement Study and the Asset and Health Dynamics of the Oldest Old (AHEAD), and on actual bequests from AHEAD. Actual bequests were measured in exit interviews given by proxy respondents for 774 AHEAD respondents who died between waves 1 and 2. Because the exit interview is representative of the elderly population, the distribution of estate values is quite different from that obtained from estate records, which represent just a wealthy subset of the population. Anticipated bequests were measured by the subjective probability of leaving bequests. Between waves 1 and 2, increases in bequest probabilities were associated with increases in the subjective probability of surviving, increments in household wealth, and widowing while out-of-pocket medical expenses reduced the likelihood of a bequest. By comparing bequest probabilities with baseline wealth we were able to test a main prediction of the life-cycle model, that individuals will dissave at advanced old-age. The AHEAD respondents anticipate substantial dissaving before they die.