Essays on Policy, Consumer Behavior, and Health

Abstract

My dissertation consists of three chapters that examine how public policies and local markets influence consumer behavior and health. The first chapter examines the effectiveness of e-cigarette flavor restrictions and the role of geographic spillovers in policy evaluation. Using retail scanner data from 2018 to 2019 and a staggered difference-in-differences design, we analyze how state size influences policy evasion costs and the impact of flavor bans. The results show that flavor preferences play a greater role than e-cigarette loyalty in product substitution. Consumers’ distance to the borders affects their substitution toward menthol cigarettes but not toward tobacco-flavored (unflavored) e-cigarettes. Cross-border sales decrease by 4.6% for every 10-mile increase in population-weighted average distance from borders. Our findings suggest that evidence from small geographic areas, where evasion is easier, may underestimate harmful substitution and overstate the effectiveness of broader policy implementation. Building on these findings, the second chapter examines the heterogeneity in consumer responses across demographic groups. Using a structural demand model and retail scanner data with demographics from four U.S. states, this chapter identifies preferences for e-cigarette and cigarette products across demographic groups. Our results indicate that areas with higher poverty rates or larger populations of young adults exhibit more inelastic demand for both cigarettes and e-cigarettes than areas with lower rates, suggesting that tax policies may be less effective in these populations. Our counterfactual analysis of bans on different tobacco products suggests that some consumers use e-cigarettes primarily for their flavors, while others do so because of the device itself or due to nicotine dependence. The third chapter examines how pricing decisions among stores influence one another, as well as how these patterns differ between healthy and unhealthy foods for different store types. I use spatial panel regressions to account for the stores' geographic proximity. The results reveal significant local price interactions among small stores, whereas grocery stores and supermarkets show weak effects. Local price interactions among small stores could exacerbate disparities in access to nutritious foods, particularly in areas with limited supermarket availability and a high concentration of small stores.