Essays on food and consumer economics

Abstract

This dissertation studies the U.S. food market and examines economic questions closely related to the producer, consumer, and retailer. In particular, the three essays in this dissertation attempt to answer the following questions: Why organic production growth has remained slow? Can consumers in different regions purchase the same basket of food with an equal amount of food stamp allotment? Would temperature variation lead to changes in retail food sales? The first essay investigates whether price pass-through differs between organic and conventional fresh fruit and vegetable markets and sheds light on structural differences between markets. Using retail and wholesale price data from five metropolitan statistical areas with terminal markets in the United States, we estimate a rolling-window retail pricing model. We find that pass-through rates are 10 to 15 percentage points lower in the organic market, and differences are statistically significant. We also find that the gap between pass-through rates narrows as the organic market share increases. Overall, our results suggest that the organic market is relatively less competitive than the conventional market. The implication is that farmers may have less incentive to convert to organic farming as they may not be able to capture the full retail price premia paid by consumers. The second essay estimates the real value of state-level Supplemental Nutrition Assistant Program (SNAP) benefits by constructing panel price indices for the Thrifty Food Plan (TFP) using retail scanner data. Subsequently, the essay examines the extent to which a change in the purchasing power of SNAP affects program participation. We show that without correcting for population bias and outlet bias, the regional disparities in real SNAP benefits could be overestimated. The estimated difference in real SNAP benefits for a household of four ranges between 4 and 8 percentage points for the period 2006–2016. This inequality of real SNAP benefits across states is persistent during the same period. Examining the relationship between SNAP benefits and participation, we find that a ten-percent increase in the real SNAP benefits leads to a 1.2 percentage point increase in SNAP caseload per capita. Themagnitude raises to 8.1 percentage points when we examine the effect on caseload per eligible individuals. We also show that these effects would be neglected if we use the unadjusted TFP prices to estimate real SNAP benefits. The third essay examines the relationship between temperature and retail food sales using storescanner data from 42,543 stores over 11 years in the United States. We find that high temperatures increase total food sales in retail stores while extremely low temperatures decrease them. There are substantial variations in the temperature-sales relationship across food categories, and the effects are in opposite directions for drinks and non-drinks. Furthermore, we show that the effects of temperature on retail food sales differ across retail channels and income levels. Our findings have direct managerial implications and provide insight into how weather conditions may affect health-related outcomes.