Essays on Food and Consumer Choices

Koroles Awad

This dissertation examines decision-making at two fundamental margins of economic life: the choices consumers make about what to eat, and the choices skilled workers make about where to live and work. On the consumer side, the dissertation focuses specifically on food choices in the United States, probing how households allocate their food budgets across products that differ in nutritional quality and degree of industrial processing. On the labor side, it examines how social policy in migrants' home countries shapes the international movement of highly skilled workers. Together, the three chapters bring modern causal inference methods to bear on questions with significant implications for public health, industrial organization, and labor markets. The first chapter investigates how the adoption of Glucagon-Like Peptide-1 receptor agonist (GLP-1) medications alters household food purchasing behavior. GLP-1 drugs, originally developed to treat type 2 diabetes and subsequently approved for chronic weight management, have attracted extraordinary clinical and policy attention for their effects on appetite suppression and sustained weight loss. Yet their downstream consequences for consumer food markets remain poorly understood. This chapter uses household scanner data from Circana linked with prescription records to identify the causal effect of GLP-1 initiation on food spending, employing the staggered-adoption difference-in-differences estimator developed by Callaway and Sant'Anna. To measure how spending shifts across the nutritional spectrum, more than five million products are classified into NOVA processing categories using a hybrid approach combining expert hand-coding with a high-accuracy Random Forest machine-learning classifier trained on product descriptions and ingredient lists. The results show that GLP-1 initiation reduces total monthly food expenditures by approximately \$56 and produces a robust reallocation of the household food budget: the share devoted to ultra-processed foods (NOVA 4) declines significantly, while the share devoted to minimally processed foods (NOVA 1) rises. These findings suggest that pharmacological treatment for obesity generates meaningful nutritional spillovers beyond weight loss, with implications for public health policy and for food manufacturers facing secular demand shifts as GLP-1 adoption continues to accelerate. The second chapter turns to the demand for ultra-processed foods across the lifecycle, with particular attention to how the presence and age of children in a household shapes price responsiveness. Despite the well-documented health risks associated with ultra-processed food consumption---including higher obesity prevalence, poorer diet quality, and elevated all-cause mortality---policymakers lack reliable structural estimates of how households respond to price incentives across food processing categories. This chapter estimates an Almost Ideal Demand System using Circana scanner data from 2008 to 2012, classifying all food purchases according to the NOVA system and constructing Fisher Ideal price indices instrumented with geographically weighted store-chain prices to address endogeneity. To handle the substantial censoring that arises when households do not purchase from a given NOVA group in a given month, the demand system is estimated using the two-step Shonkwiler--Yen correction. The analysis is conducted separately for households without children and for households with children in three age brackets (0--6, 6--13, and 13--18 years). The results show that demand for ultra-processed foods is price-inelastic for all household types, with an own-price elasticity of approximately $-0.43$ in the full sample, but demand becomes markedly less elastic as children grow older. Households with teenagers exhibit the lowest price sensitivity to ultra-processed foods and the smallest cross-price substitution toward less-processed alternatives. These findings suggest that price-based policies such as excise taxes on ultra-processed foods or added sugars, while capable of encouraging some substitution toward healthier alternatives, will be least effective precisely in the household type---those with school-age and teenage children---that is most associated with elevated obesity risk. The third chapter shifts to labor economics, examining how marriage equality legislation in skilled workers' home countries affects their propensity to migrate to the United States. The chapter exploits the staggered legalization of same-sex marriage across 27 European Union countries between 2001 and 2019 to estimate how access to legal same-sex marriage in the home country alters the inflow of highly skilled European workers holding H1B visas. Using Department of Homeland Security visa admission data and the Callaway--Sant'Anna difference-in-differences estimator, the chapter finds that home-country marriage equality enactment leads to a statistically significant decline of approximately 0.077 skilled worker admissions per 1,000 labor force members annually---equivalent to roughly 600 fewer skilled workers per country per year---with effects that intensify over the five years following legalization. A complementary two-way fixed effects analysis based on the Wolfers framework further shows that U.S. federal recognition of same-sex marriage following the 2013 \textit{United States v.\ Windsor} ruling partially offsets this decline, attenuating the outflow of European skilled labor. Robustness checks using placebo visa categories, raw admission counts, and not-yet-treated control groups confirm the main findings. The results demonstrate that social policy---specifically, the legal recognition of same-sex partnerships---functions as a meaningful amenity in the international competition for skilled labor, with implications for immigration policy and for the ability of destination countries to attract and retain high-skill migrants. Taken together, the three chapters contribute to consumer economics and labor economics in complementary ways. On the consumer side, the dissertation advances understanding of how both pharmacological innovation and household structure shape food demand, with direct relevance for the design of obesity policy, nutrition interventions, and food industry strategy. The first two chapters demonstrate that consumer food choices are responsive to forces---new medications and the lifecycle of children in the home---that standard demand analyses rarely capture, and that structural methods applied to large-scale scanner data can recover credible, policy-relevant estimates of these effects. On the labor side, the dissertation contributes to the nascent but growing literature on how social amenities and legal rights influence high-skill international migration. The third chapter shows that marriage equality is not merely a domestic civil rights question but a factor that reshapes the international allocation of human capital, with measurable consequences for both sending and receiving countries. Methodologically, all three chapters employ modern difference-in-differences estimators designed for staggered treatment adoption, contributing to the growing applied econometrics literature on heterogeneity-robust causal inference. The dissertation thus offers both substantive findings and methodological tools that will be useful to researchers studying consumer behavior, food markets, and skilled labor mobility.

Essays on Food and Consumer Choices

Files and links (1)

Abstract

Metrics

Details