Projects

Current research projects and working papers.

Working papers

When Income Support Hurts and Helps: Health Effects of Survivors Benefits Job Market Paper, Reject & Resubmit, Journal of Health Economics; Coauthor: Qilu Pan

This paper explores how income support after widowhood affects health over the life cycle, using Social Security Survivors Benefits as our setting. Using a fuzzy regression discontinuity design at the age-60 eligibility threshold and panel data from the Health and Retirement Study. Found that eligibility worsens self-reported health and increases diabetes in the short run. These effects are concentrated among widows who retire before age 60 and are most financially dependent on benefits. Mechanism analyses point to a consumption channel: the arrival of cash at eligibility relaxes a binding liquidity constraint, inducing a transitory increase in food expenditures that temporarily elevates metabolic risk. However, over the years following eligibility, depression scores improve and hospital admissions increase, suggesting that sustained benefit receipt reduces financial stress and expands access to care. Exploiting the 1965 reform that lowered the earliest eligibility age from 65 to 60, we find consistent evidence that expanded eligibility improves survival to age 70, with mortality declines concentrated in cardiovascular disease, tuberculosis, and suicide. Income support thus generates short-run health risks while improving longer-run survival.

Survivors Benefits and Intergenerational Transfers: Evidence from Adult Children’s Labor Supply Working paper

Applied Regression Kink Design (RKD) using SIPP and ATUS microdata to examine how Social Security survivor benefit eligibility at age 60 reshapes resource transfers within families. Find that eligibility significantly increases intergenerational financial transfers from widowed parents(aged 55-65) to adult children, but leaves time-based caregiving transfers unchanged — suggesting income effects drive the reallocation. Downstream effects on children's labor supply are also identified: benefit eligibility reduces children's working hours. Heterogeneity analysis by parental health status reveals differential responses across health groups. Additionally, exploit the Medicare eligibility kink at age 65 to examine how the interaction between Social Security and Medicare jointly shapes these family spillovers.

Effects of the ACA's Dependent Coverage Provision on Young Adults Working paper

Uses Difference-in-Differences and a triple-difference framework on CPS microdata (2008–2013) to estimate causal effects of the ACA's dependent coverage provision on insurance uptake and labor-market outcomes.
Produces parallel-trend graphs and subgroup analyses in Stata.

Methods & data

My empirical toolkit spans quasi-experimental designs, VaR models, and applied machine learning. I work with large longitudinal microdata (more than five million observations) and frequently merge administrative records with survey data.

Tools

Programming: Python, Stata, MATLAB, SQL, Tableau
Econometric methods: Regression Discontinuity (RDD/Fuzzy RD), Regression Kink Design (RKD), Difference-in-Differences (DID/DDD), Instrumental Variables (IV/2SLS), Panel data models, Life-cycle models, Welfare models
Machine learning: Regression, classification, clustering
Datasets: Health and Retirement Study (HRS), Survey of Income and Program Participation (SIPP), Current Population Survey (CPS), Consumer Expenditure Survey (CES), Human Mortality Database (HMD)