Job market paper
30 Years of Welfare-to-Work: Intergenerational Impacts of Welfare Reform
Nearly 30 years ago, President Bill Clinton signed the Personal Responsibility and Work Opportunity Reconciliation Act (PRWORA), abolishing federal entitlement to cash assistance. Using confidential Census Bureau data, we document the long-run impacts of PRWORA on individuals who were children at the time of reform. Our research design compares siblings born to the same mother who were different ages at the time of reform and contrasts these within-family differences between families more and less likely to have received welfare. We find that exposure to welfare reform decreased childhood welfare receipt and increased parental employment. While on average these counteracting effects left net family income unchanged, the average conceals considerable variation, with Black families facing decreases in net income and white families facing increases. In adulthood, exposure to welfare reform increased both earnings and employment. Earnings effects follow a life-cycle profile, appearing as early as age 18 and fading entirely by age 24, while employment gains persist through at least age 30 and come from lower-paying, less stable jobs. We also find that welfare reform reduced teen births, teen arrests, and receipt of housing assistance, but had no discernible impact on mortality and participation in the Supplemental Nutrition Assistance Program or Medicaid. Effects are more pronounced for girls than boys and proportional to parents' time spent working, which suggests that our results are driven by children emulating their parents.
Work in progress
The Take-Up Gap: Sorting versus Place Effects in Medicaid
Take-up of safety net benefits varies widely across places, but it is unclear whether this reflects the effect of place itself or the sorting of different types of people into different places. We study this question for Medicaid among low-income, childless adults in states that expanded eligibility under the Affordable Care Act, where eligibility rules are fixed across place. Using internal Census Bureau data, we estimate a two-way fixed effects model to decompose person and county effects in Medicaid take-up. We then follow individuals who move across counties in an event-study design, asking whether their take-up shifts toward the level of their destination effect, which would indicate a causal effect of place. We relate the estimated county effects to state policies and local characteristics, including administrative burden, information, and stigma, to understand what drives low take-up.
Sharing a Home, Sharing the Cost: Impacts of the Informal Safety Net
Low-income families rely not only on support from the formal safety net but also on informal support from family and friends. Unlike publicly provided aid, the informal safety net generates private costs for the individuals providing assistance. We study these costs in the context of a particularly prevalent form of informal support: doubling up, in which multiple families share a single residence. Using linked Census Bureau data, we estimate the prevalence of doubling up and its impact on the social and economic well-being of host families. Our empirical strategy exploits quasi-random variation in a household's likelihood of hosting, combining variation in the timing of a relative’s receipt of federal rental assistance and distance to that relative. The project will produce national estimates of the prevalence of doubling-up and causal evidence on the effects of co-residence for host families.
Why Is Welfare Hard to Get? Using AFDC to Estimate Revealed Welfare Weights
Why do governments make welfare hard to obtain? Standard models attribute eligibility rules and administrative burden to fiscal constraints and optimal targeting of aid, but governments may also hold preferences over which families deserve aid. This paper tests for such preferences by estimating states’ revealed welfare weights. Identification comes from how states adjusted policy instruments (acceptance rates, case closures, and payment levels) after federal actions in the 1960s required them to readmit Black families to welfare. Historical changes to the federal matching formulas, which altered states’ costs without changing caseload composition, provide a race-neutral benchmark for budget effects.
Resting papers
Does Welfare Reduce Mortality? Evidence from Supplemental Security Income
In this paper, I test whether transfer payments from the Supplemental Security Income (SSI) program reduce infant mortality in U.S. counties. In its 1990 Sullivan v. Zebley decision, the Supreme Court relaxed medical eligibility criteria for children on SSI, resulting in a dramatic rise in child enrollment and a large influx of income into some of the poorest areas in the U.S. I use an instrumental variable strategy, exploiting geographic variation in the expected treatment effect of this policy change, to test whether the resulting increase in child SSI enrollment had a causal impact on infant mortality. I find that a 1 percentage point increase in child SSI enrollment leads to a 1.45 percentage point reduction in infant mortality rates.