Essays on Long-Term Care Insurance: Risk Misperceptions, Suboptimal Timing, and Racial Disparities

This dissertation focuses on long-term care (LTC) in old age and the low ownership of private long-term care insurance (LTCI). Older adults in the US face significant risk of needing long-term care services. Despite the risk, most Americans never attempt to insure against it, and those who do tend to apply late in life when they are likely to be rejected. The first essay in this dissertation builds a quantitative equilibrium model to study consumers’ decision about when to buy LTCI in the presence of uninsurability risk. The model evaluates how much of the delay in purchase can be explained by risk misperceptions. I formulate the problem as a three-period model where consumers may buy insurance in a competitive market based on their observable risk characteristics during the first two periods in order to insure against nursing home risk in the third period. I estimate the model using the Health and Retirement Study (HRS). The model successfully reproduces the low ownership rates and timing of purchase across all income quintiles. I find that risk misperceptions lead to 42% lower than optimal purchase rate, and cause 29% of early buyers to delay their purchase. The sub-optimal behavior is particularly pronounced among the middle-income consumers. The second essay studies the patterns of LTC risk and LTCI purchase across race and ethnicity in the US. By 2050, the share of Blacks and Hispanics in the over-85 population is expected to increase to 40%. A better understanding of risk and insurance across different segments of the aging population is particularly important, given this increasing racial heterogeneity in older adults. Using the HRS, this chapter studies racial disparities in late-in-life health outcomes and estimates the risk of having a nursing home stay over lifetime across race and ethnicity. It finds that Blacks and Hispanics have a lower risk of entering a nursing home than Whites. The paper also examines differences in the likelihood of LTCI purchase across different racial and ethnic groups. The findings suggest that once demographic, socioeconomic, and family characteristics are accounted for, non-Whites are not less likely to buy insurance.