Long-term care and the elderly

Abstract

Long-term care expenditures represent one of the largest uninsured financial risks facing the elderly. Medicaid provides incomplete insurance against these costs: unlimited nursing home benefits with a deductible equal to the savings and income above the means-testing limits. While private insurance is available, fewer than 10 percent of the elderly are currently covered. This thesis explores how the elderly prepare for future nursing home use and the interactions between the private and public insurance systems. Chapter one exploits the state-variation in Medicaid generosity to study the financial response of the elderly to perceived future nursing home needs. I find that the elderly shift their consumption and savings decisions in response to Medicaid. Single households have lower net worth through the median of the distribution due to Medicaid policy. On the other hand, I find that married households do not lower total net worth, but they change their relative holdings of protected and non-protected assets. Chapter two explores the crowd-out effect of the public Medicaid program on demand for private long-term care insurance coverage. We estimate the impact of Medicaid program rules on private long-term care insurance coverage for the elderly. We find small but statistically significant marginal crowd-out effects.

(cont.) Our estimates imply that even a $67,000 decrease in the asset disregard for couples would only increase private long-term care insurance ownership among the elderly by 1.9 percentage points. These findings underscore that marginal reforms to the existing Medicaid program are unlikely to be an effective way of increasing private long-term care insurance coverage among the elderly. Chapter three explores individuals' expectations for future nursing home use. I compare self-reported probabilities to the statistical probability computed with a state-of-the-art model used by the long-term care insurance industry. I find that respondents tend to overestimate unlikely outcomes and underestimate likely outcomes. On average, though, the expectations are very accurate. I find that expectations for nursing home use evolve with health conditions in similar ways as the statistical probability. While I find that expectations include private information, they do not account for all information available to the individual, especially the individual's demographic characteristics.