Essays on Aging and Human Capital.

Abstract

This dissertation comprises three distinct essays on aging and human capital. The first essay focuses on the financial literacy of older women. In the U.S., women tend to have lower levels of financial literacy than men. This is consistent with a household division of labor in which men manage finances. However, women also tend to outlive their husbands, so they will eventually need to take over this task. I find that women acquire financial literacy as they approach widowhood. These results are consistent with a model in which the division of labor breaks down when a spouse dies: women have an incentive both to delay acquiring financial knowledge and also to begin learning before widowhood. This essay represents the first empirical examination of the financial literacy of both members of couples and provides a life-cycle interpretation of the gender gap in financial literacy.

The second essay, written with Robert Willis, investigates the impact of Alzheimer's disease, a different source of uncertainty, on the division of labor. We analyze the impact of information about cognitive decline on the choice of financial respondent in a large-scale survey of older Americans. We find that couples with investments that are individually controlled are much more responsive to a diagnosis of a memory-related disease than couples who passively receive their retirement income. This finding is consistent with a model of the value of information: households with the most to gain financially from preparation are most responsive to a diagnosis.

The third essay, written with Gwenith Fisher and Robert Willis, analyzes the selectivity implications of surveying an older population using the internet. We find that an internet sample selects not only on age and education but also on cognition. An internet-only sample would miss an increasing number of persons, particularly those with lower fluid and crystallized intelligence, at older ages. Because of selection, an internet-only sample would lead to an overestimate of level of intelligence and an underestimate of the rate of cognitive decline with aging.