Projects

Current research projects and working papers.

Working papers

The Family Spillovers of Social Security: Evidence from Survivors Benefit Eligibility
  • Applies Regression Kink Design (RKD) with SIPP microdata to examine whether Social Security survivor benefit eligibility at age 60 reshapes intergenerational financial transfers between widowed parents and their adult children, and downstream effects on children's labor supply.
  • Data wrangling and analysis conducted in Python.
When Income Support Hurts and Helps: Health Effects of Survivors Benefits
  • Uses a Fuzzy RD at the age-60 Social Security eligibility threshold and HRS panel data (1998–2018) to identify causal effects on health outcomes.
  • Finds that eligibility worsens short-run self-reported health and raises diabetes rates (driven by financially dependent widows), while a 1965 reform expanding eligibility improved survival to age 70 with mortality declines in cardiovascular disease, tuberculosis, and suicide.
  • Analysis conducted in Stata and Python.
Effects of the ACA's Dependent Coverage Provision on Young Adults
  • 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, structural life-cycle 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
  • 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)