Peer-reviewed Publications

Davis, W.,  Xilau, J., Tchernis, R., and Gregory, C. A. (2025). A Flexible Model of Food Security: Estimation and Implications for Prediction . American Journal of Agricultural Economics, 1-27.

Abstract:  We propose a novel Bayesian Graded Response Model (BGRM) for measuring household-level food security and other latent traits. The BGRM produces continuous food security estimates along with household-level measures of estimation uncertainty. Unlike the USDA’s official model, the BGRM accommodates both binary and ordered polytomous items. We further extend the model to allow for any combination of binary, ordered polytomous, and even continuous variables. To demonstrate the model’s features, we estimate the BGRM using responses to the 10 adult core Food Security Module questions from the 2017–2018 National Health and Nutrition Examination Survey (NHANES). Results show non-trivial uncertainty in household-level food security estimates and overlap across USDA-defined food security categories. As a robustness check, we estimate the model with Current Population Survey data, finding qualitatively similar results. We illustrate an application of the continuous food security estimates by calculating Foster, Greer, and Thorbecke indices which capture the prevalence, depth, and severity of food insecurity. To demonstrate flexibility in variable selection, we also include a continuous variable, household-level monthly food spending, capturing both economic access and experiential food security information in a single latent construct. The adaptability of the BGRM positions it as a versatile tool for measuring food security and related latent traits, particularly when measures of uncertainty or a mix of different variable types are required. While the empirical application illustrates model capabilities, the primary contribution of the study is methodological.

Working Papers

Examining Heterogeneity in the Supplemental Nutrition Assistance Program Benefit Cycle:  A Finite Mixture Approach - Job Market Paper [Draft Available here.]

Abstract:   The Supplemental Nutrition Assistance Program (SNAP) is the largest nutrition assistance program in the United States. SNAP benefits are issued on a monthly basis, and recipients tend to redeem a larger share of benefits shortly after issuance, a pattern of behavior known as the SNAP benefit cycle. Previous research documenting the SNAP benefit cycle and its various impacts have often relied on cross-sectional survey data, with relatively small sample sizes, short periods of analysis, and self-reported SNAP participation and spending. In this study, I leverage a unique administrative panel dataset to investigate several aspects of the SNAP benefit cycle and uncover potential unobserved heterogeneity. The data include the universe of SNAP benefit transaction activity from 2011 to 2015 for 1.8 million unique households in Georgia, and detailed information regarding benefit receipt and redemption, including timing, dollar amount, and location of transactions. Using a panel data finite mixture model, I find evidence supporting the existence of two types of SNAP households, which I denote as slow and fast spenders. Fast spenders comprise the overwhelming majority of SNAP households, 77%, and spend approximately 70% of their benefit in the first week. In addition, fast spenders shop more frequently throughout the month, stay longer in the program, 3 years compared to 1 year and 4 months for slow spenders, and receive benefits that are almost twice as large, on average. The two groups show similar patterns of rural or urban residence, and the types of stores where they spend SNAP benefits.  Finally, I also find that shopping frequency, tenure in SNAP, household SNAP benefit, and household store preferences are not strong predictors of household spending type.   My results add to a more comprehensive  understanding of the SNAP benefit cycle and inform welfare improving policies that promote smoother spending of SNAP benefits throughout the month.

Works in Progress 

Economic Conditions and Teenage Substance Use 

Abstract: This  research examines how macroeconomic conditions affect teenage marijuana, cigarette and alcohol use. The findings are based on state and year fixed effects models, using state-level data from the Youth Risk Behavior Surveillance System, spanning 1991-2019. In contrast to the recent literature, I find no statistically significant evidence of a relationship between economic conditions and teenage marijuana use. I discover the same pattern for cigarette use, which is consistent with results in the literature. Finally, I find evidence of a pattern of counter-cyclical alcohol use among teenagers, also consistent with results in the literature. My results suggest that policymakers must consider the consequences of changes in funding of substance use programs, aimed at teenagers, during stressful economic times.  

Benefit Redemption Patterns in the Supplemental Nutrition Assistance Program in Georgia