Living and dying at older ages: Essays on the Hispanic mortality paradox and the annuity puzzle in the United States
The four chapters comprising this dissertation share a concern to explore issues of health and mortality at older ages in the U.S. The studies deal with three puzzles posited in the literature. First, we examine the Hispanic mortality paradox. Using mortality estimates based on Medicare records, we find that conventionally constructed Hispanic death rates are underestimated, although the results confirm the mortality advantage for Hispanics relative to non-Hispanic whites. We find that foreign-born Hispanics have lower mortality than U.S.-born Hispanics and non-Hispanic whites, and this advantage remains unaffected up to age 100. This result contradicts previous theoretical simulations for both the healthy migrant effect and the assimilation effect, which show mortality of foreign-born Hispanics converging toward that of U.S.-born Hispanic at late adult ages. The second puzzle examined is the unusually low mortality of U.S. whites at older ages, when compared to other low-mortality populations. We find surprisingly similar age patterns of mortality for several population subgroups in the U.S., suggesting that the high proportion of immigrants and people of different ethnic groups among whites does not explain the unusual age pattern of mortality for this population. In the last two essays, we shift attention to the relation of health and mortality at older ages with the “Annuity Puzzle”. We provide evidence of the existence of adverse selection in annuities. Using a multistate model, we show that the mortality of annuitants resembles that of very healthy and high educated people in the population, and that disability rates among annuitants are twice as low as in the population. Finally, we address how differences in health status at retirement influence the decision to purchase a life annuity. We propose a life cycle model in which an elderly person considers the effects of both uncertain future medical expenses and uncertain survival when choosing an optimal level of annuitization. We predict that adverse selection and precautionary motives reduce the utility value of a life annuity by almost 50 percent for people in poor health, and that full annuitization is not the optimal solution for retirees, regardless of health status. Our analysis suggests that offering higher payouts for consumers in case of medical shock can make annuities more attractive for most of the elderly.
Demographics|Minority & ethnic groups|Sociology|Economics
Turra, Cassio Maldonado, "Living and dying at older ages: Essays on the Hispanic mortality paradox and the annuity puzzle in the United States" (2004). Dissertations available from ProQuest. AAI3138082.