Research
My research studies how health risks, survival, and the design of health insurance shape household saving and the wealth distribution. Combining household data with structural models, I examine why health-related precautionary saving can persist among wealthy households and how means-tested insurance can encourage asset decumulation near eligibility thresholds.
Job Market Paper
Marginal Utility Shocks and the Precautionary Saving Puzzle
Abstract: This paper asks whether precautionary saving induced by health shocks helps explain saving at the top of the wealth distribution. Using the Health and Retirement Study, it documents that pre-retirement couples in the top wealth decile save about $1.1 million over two years, whereas standard income-fluctuation models predict roughly $2 million of dissaving. It then shows that a luxury-health framework, in which survival remains valuable at all wealth levels, can resolve this puzzle. The paper embeds this framework in an income-fluctuation model and derives a sufficient condition under which the precautionary saving motive remains active at arbitrarily high wealth. Given that saving remains positive beyond this threshold and mortality generates random separation across cohorts, it proves that the induced stationary wealth distribution has Pareto tails, the first such result from precautionary saving alone. After estimating the health production function and calibrating the model, the paper shows that the mechanism predicts saving of $121,960 against $1.1 million observed in the top decile, accounting for roughly 11 percent of observed top-decile saving, whereas the benchmark model predicts dissaving.
Working Papers
Means-Tested Insurance and Pareto-Tailed Wealth Distribution
Summary: This paper examines how means-tested health insurance can shape saving incentives near the bottom of the wealth distribution. Among single-member households aged 55–64 in the Health and Retirement Study (HRS), medical spending falls in the second and third wealth deciles. The pattern persists when total medical expenditure is imputed using the Medical Expenditure Panel Survey (MEPS). The paper develops conditions for consistency of the imputation and of OLS estimates using the imputed dependent variable. A stylized model shows how an asset threshold for means-tested insurance can generate strategic asset decumulation, linking insurance eligibility to household saving behavior.
Work in Progress
Online Innovation, Market Entry and Competition in Remote Indigenous Communities