Date of Award

June 2020

Degree Type


Degree Name

Doctor of Philosophy (PhD)


Public Administration


Sarah E. Hamersma


Children, Food Stamps, National School Lunch Program, Nutrition, Supplemental Nutrition Assistance Program, Welfare

Subject Categories

Social and Behavioral Sciences


This dissertation uses experimental and quasi-experimental methods to evaluate policy instruments available to state and local officials administering food assistance programs for improving children’s utilization of services, nutritional intake, and food security. More specifically, this dissertation consists of three chapters pertaining to the National School Lunch Program (NSLP) and the Supplemental Nutrition Assistance Program (SNAP).

In the first chapter, co-authored with Amy Ellen Schwartz, I use large, uniquely detailed longitudinal administrative data on New York City students and schools, including the different lunch menus they offer over time, to present the first plausibly causal evidence on the link between school menus and participation in the NSLP, and their implications for disparities in program utilization across students from diverse backgrounds. Using student and school fixed-effects models, I find that the introduction of new menus increases the share of students participating on both the extensive and intensive margins, and can help to close racial, gender, and socioeconomic gaps in the utilization of school lunch. In one extension, I find evidence that increases in participation are driven by the adoption of newer, more innovative menus. I find no evidence of changes in attendance or adverse weight outcomes. These findings provide evidence in support of the efforts that many school districts are taking to boost lunch participation by reformulating their menus and help to inform the decisions of those considering similar changes.

While students stand to gain from participating in school lunch, it is also important that they make the right dietary choices as they make their way through the lunch line. In the second chapter, I use primary data collected through a month-long field and survey experiment I designed and conducted to investigate the efficacy of using cheap material rewards to induce better dietary choices among low-income Black and Hispanic children—who are more likely to be obese than their high-income or white peers—in a school lunch setting. While existing studies have shown material rewards to be effective in the short term and when introduced intermittently, this study shows that their effect can dissipate quickly over time when offered daily. I find no evidence that the introduction of extrinsic incentives crowded out intrinsic dietary behaviors. These findings have implications for the design and implementation interventions using material rewards for improving dietary habits among school-aged children.

In the third chapter, I broaden the scope of my research to include SNAP, which helps to safeguard the food security of millions of children. The politicization and racialization of the program have made it a target of reforms that effectively limit its coverage and efficacy, and contribute to its disparate implementation across states and counties in ways that exacerbate social inequity. I designed a large survey experiment evaluating the efficacy of highlighting the child beneficiaries of SNAP for inducing greater public support for the program. I find that emphasizing its child beneficiaries can increase support overall and across key political constituencies, though more so when those children are characterized as White than Black. As an extension, I also examine the generalizability of these findings to the Temporary Assistance for Needy Families program, and again find that highlighting child beneficiaries leads to increases in support, though the estimated effects are less pronounced and even more so dependent on the children’s race. These findings can help to inform the outreach efforts of program administrators.

This dissertation adds to the existing literature by providing new insights and expanding on previous work. The results of the first chapter corroborate the many anecdotal accounts suggesting that school menus influence student lunch participation, but also show that other barriers, such as the price of meals or stigma, may be more important. The second chapter extends previous work on using material rewards to induce better dietary behaviors by presenting evidence that they may not be as cost-effective or easy to implement as previously thought. The findings of the third chapter show that providing information on the child beneficiaries of welfare programs can change public attitudes towards them and offer fresh evidence that public opinion of government policy is often based on the demographic groups perceived to benefit from them. This dissertation also highlights the implications that policy instruments can have for social equity and economic equality by focusing on disparities in program utilization, access, and outcomes across race, gender, age, and socioeconomic status. Lastly, it offers guidance for policy makers and program administrators by providing new evidence about the efficacy of various policy instruments available to them for administering food assistance programs.


Open Access