This thesis is made up of three main essays that aim to develop a deeper understanding of issues involving the
public insurance programs for the elderly, and the risks they insure against.
In the first essay (Chapter 2), using data from the Health and Retirement Study linked to administrative Medicare
and Medicaid records, along with the Medical Expenditure Panel Survey, we estimate the stochastic process for
total and out-of-pocket medical spending. By focussing on dynamics, we consider not only the risk of catastrophic
expenses in a single year, but also the risk of moderate but persistent expenses that accumulate into a catastrophic
lifetime cost. We also assess the reduction in out-of-pocket medical spending provided by public insurance schemes
such as Medicare or Medicaid. We find that although Medicare and Medicaid pay the majority of medical expenses,
households at age 65 will on average incur $59,000 in out-of-pocket costs with 10 percent of households incurring
more than $121,000 in out-of-pocket expenses over their remaining lives.
In the second essay (Chapter 3), we compare dementia prevalence and how it varies by socioeconomic status (SES)
in the United States and England. We compare between country differences in age-gender standardized dementia
prevalence, across the SES gradient. Dementia prevalence was estimated in each country using an algorithm based
on an identical battery of demographic, cognitive, and functional measures. Dementia prevalence is higher among
the disadvantaged in both countries, with the United States being more unequal according to four measures of SES.
Once past health factors and education were controlled for, most of the within country inequalities disappeared;
however, the cross-country difference in prevalence for those in the lowest income decile remained disproportionately
high. This provides evidence that disadvantage in the United States is a disproportionately high risk factor for
In the final essay (Chapter 4), we assess the optimal structure the U.S. Social Security system, taking into account
the current system’s unfunded liabilities, transition dynamics and political feasibility constraints. We base the
assessment on an estimated overlapping generations general equilibrium model that features both aggregate and
idiosyncratic uncertainty. The quantitative analysis establishes that although transition costs greatly restrict the
U.S. government’s ability to move away from the current Social Security system, ignoring the political feasibility
constraints allows the government to increase welfare by transitioning to a more progressive and less costly to
operate system. However, taking into account the political feasibility constraints overturns this result, as no reform
is simultaneously welfare increasing and politically feasible.