Date of Award

Spring 4-30-2018

Degree Type

Dissertation

Degree Name

Doctor of Philosophy (PhD)

Department

Economics

First Advisor

Dr. Rusty Tchernis

Second Advisor

Dr. Charles Courtemanche

Third Advisor

Dr. Ian McCarthy

Fourth Advisor

Dr. Nguedia Pierre Nguimkeu

Abstract

This dissertation examines the causal impacts of the Supplemental Nutrition Assistance Program (SNAP) on adult weight and nutrition outcomes using novel approaches to address nonrandom selection into the program and survey data quality issues. The overarching objective of this study is to provide credible estimates of the effects of SNAP to help policy makers and administrators engage in meaningful debates and prescribe nutrition assistance policies that promote the overall well-being of low-income Americans as well as mitigate any unintended consequences of the largest nutrition assistance program in the United States.

The first chapter proposes a model to estimate treatment effects when program participation is endogenously misreported. This chapter shows that failure to account for endogenous misreporting can result in the estimate of the treatment effect having an opposite sign from the true effect. Expressions for the asymptotic bias of the ordinary least squares (OLS) and instrumental variable (IV) estimators are provided and the conditions under which sign reversal may occur are discussed. This chapter then develops a method for eliminating this bias when researchers have access to information related to both participation and misreporting. The root-n consistency and asymptotic normality of the proposed estimator are established after which Monte Carlo simulations are used to demonstrate the remarkable performance of the estimator in small samples.

The second chapter estimates the effect of SNAP on adult obesity addressing self-selection and endogenous misreporting of participation. Using the methodology developed in the first chapter of this dissertation, the second chapter estimates the causal impact of SNAP on obesity using data from the National Longitudinal Survey of Youth – 1979 cohort. From a simple partial observability model of participation and misreporting, I predict probabilities of participation which are used to consistently estimate the average effect of SNAP on body mass index (BMI). The estimated misreporting model confirms some prior findings in the literature regarding the correlates of reporting error. However, contrary to most previous studies, I do not find any evidence of a statistically significant effect of SNAP on BMI.

The third chapter studies the potential problems with administrative records and their implications for econometric estimates using the National Household Food Acquisition and Purchase Survey (FoodAPS) data set, which contains two different administrative measures of SNAP participation as well as a survey-based measure. This chapter first documents substantial ambiguity in the two administrative participation variables and show that they disagree with each other almost as often as they disagree with self-reported participation. Estimated participation and misreporting rates can be meaningfully sensitive to choices made to resolve this ambiguity and disagreement. Finally, this chapter documents similar sensitivity in regression estimates of the associations between SNAP and food insecurity, obesity, and the Healthy Eating Index. These results serve as a cautionary tale about uncritically relying on linked administrative records when conducting program evaluation research.

Share

COinS