Research

Abstract: State-level daycare regulations represent a significant policy lever for improving early childhood environments, yet their long-run causal effects remain largely unknown. This paper examines stricter child-to-staff ratios and director education requirements implemented across states between 1984-1994. My identification strategy exploits variation in policy location, timing, and children's age at exposure, using difference-in-differences to estimate causal impacts on adult outcomes measured 25-40 years later. I find that exposure to stricter regulations increases adult annual income by approximately $1,400 (1.5-2%). The regulations operated through intensive margin human capital accumulation: significant increases in graduate degree attainment (0.5 pp) but no effects on college attendance or employment rates. Treatment effects are substantially smaller for children from states with weaker economic conditions and families with lower parental education. Short-run analysis reveals regulations reduced daycare supply by 5-10%, yet maternal labor supply remained unchanged due to substitution toward unregulated family daycare. These patterns suggest regulations created a quality-access tradeoff that primarily benefited middle-class children while potentially harming those vulnerable to losing access. These findings highlight that regulations alone, without addressing access and affordability, may increase inequality in early childhood investments and long-run outcomes.

Abstract: A fundamental question in public economics concerns the desirability of providing government services directly vs. contracting out. Given the preponderence of in-kind benefits in the US and other developed countries, this is a question of great importance. Yet, there is limited evidence.  Using a difference-in-differences design, we study the transition of Mississippi’s Women, Infants, and Children program from state-run warehouses to retail-based food distribution. Retail distribution reduced participation by 12–14%, with larger declines among children and disadvantaged groups, while lowering per-participant costs by 32% through warehouse elimination. Higher retail prices were offset by lower redemption rates. Evidence suggests increased shopping burden, rather than other take-up frictions, drove the decline in participation. The findings underscore trade-offs in privatizing benefit delivery while preserving service quality and access.

Abstract: Despite the benefits of taking a college entrance exam, a large fraction of low-income students who register for the ACT with a fee waiver fail to sit for the test on their registered date. Fee waiver students comprise 30% of all ACT registrations, yet represent 60% of all absentees on national test days. With the aim of improving test day attendance among this low-income student population, we designed and implemented three nationwide natural field experiments in the U.S. Two experiments tested the effect of reminders and information sent to the fee waiver students directly. The third tested whether incentivizing key school staff with a lottery that was linked to achieving 90% attendance for their fee waiver students could improve attendance. For field experiments one and two, we find no evidence that our treatments affected a student's likelihood of attending the test. In experiment three, we document treatment effects up to a 6.6 percentage point increase in attendance. The marginal value of public funds of this intervention is estimated to be infinite if at least one student is induced to enroll in a 4-year college, which is likely. Stated another way, estimated after-tax benefits to society are positive and government revenues also increase, which suggests the investment is worthwhile. We also find that school leaders and management have a large impact on fee waiver test attendance. A one standard deviation in the school fixed effects is associated with a 19 percentage change in fee waiver test attendance.