Modeling Household Food Choice Behavior Using Store-Based and Household-Based Scanner Data

Paper Session

Friday, Jan. 6, 2017 2:30 PM – 4:30 PM

Swissotel Chicago, Montreux 1
Hosted By: Agricultural and Applied Economics Association
  • Chair: Mary Muth, RTI International

Place, Price, and Poverty: Understanding Factors Shaping Household Food Behavior

Scott W. Allard
,
University of Washington
Pat Ruggles
,
University of Chicago

Abstract

Rising rates of food insecurity have led researchers to examine how the local retail food environment affects household food purchases, consumption, and food security. Particular attention has been given to identifying the presence of "food deserts," areas with low or no spatial access to retail stores that sell fresh food and groceries. Very few data sources, however, can link the local food retail environment and food pricing to household food purchases and food insecurity in space. As a result, research often is limited in its ability assess the relationship between access to food retailers, food pricing, and food security, especially for race and ethnic minorities, the poor, and other vulnerable households. To address these critical gaps in the literature, this paper explores the relationships between household food security, food purchases, food pricing, and the geography of the local retail food infrastructure for low-income households, using unique public and restricted use data files from the National Household Food Acquisition and Purchase Survey (FoodAPS). In addition to detailed household characteristics from the survey, we use household food purchase data from FoodAPS and linked IRI commercial scanner data to calculate local food price indices. We also use linked store location information from TDLinx to calculate food retailer spatial access measures. Household shopping decisions are modeled as a function of spatial access to retailers and the spatial contours of food pricing. We find that many population sub-groups identified in the literature as being vulnerable to low food resource access, such as racial minorities or urban residents, have greater or comparable spatial access to several different types of food resources compared to less vulnerable population sub-groups.

Store Choice and Consumer Behavior in Food Deserts: An Empirical Application of the Distance Metric Method

Lauren Chenarides
,
Pennsylvania State University
Edward C. Jaenicke
,
Pennsylvania State University

Abstract

While food access is an increasingly studied component of research related to diet and health, consumer behavior and store choice have been relatively overlooked in understanding the dietary health-food access relationship. Especially in areas with high poverty rates, where the proportion of low access and low income population persists over time, consumers are faced with shopping at non-traditional stores, which may augment the negative welfare impacts. Using detailed scanner data, namely IRI’s Consumer Network Panel and IRI’s InfoScan, along with Nielsen’s TDLinx store characteristics data, this paper develops a structural model of store choice that frames Pinkse, Slade, and Brett’s (2002) distance metric (DM) method inside a demand system to model what behaviors drive consumers’ store choice decisions, highlighting underserved communities. We frame the DM method within the Exact Affine Stone Index (EASI) demand system framework (Lewbel and Pendakur, 2009). While the DM method has been used previously to model brand choice, this paper is the first to use it to investigate store choice. Because our store-choice model is based on demand for store attributes (such as relative prices, product assortment measures, store services, and distance between stores), it reveals consumer preferences for store types and provides insight into policy prescriptions that attempt to improve food access. Despite the importance of this topic, little research exists that documents how the distribution of consumer types and geographic patterns are associated with store choice. The use of the data sources listed above supports a more complete picture of both the food environment and consumer behavior, and it is our hope that our methods and results will generate significant interest and discussion with applications in marketing, health, and food policies related to food access.

Does a Nutritious Diet Cost More in Food Deserts?

Linlin Fan
,
University of Illinois-Urbana-Champaign
Kathy Baylis
,
University of Illinois-Urbana-Champaign
Craig Gundersen
,
University of Illinois-Urbana-Champaign
Michele Ver Ploeg
,
USDA Economic Research Service

Abstract

Food deserts and their potential effects on diet and nutrition have received much attention from policymakers. While some research has found a correlation between food deserts and consumer outcomes, it is unclear whether food deserts truly affect consumption behavior. In this article, we compare food prices in food deserts, defined as low-income, low-access census tracts, and non-food deserts to observe whether and to what extent consumers face higher prices for a complete diet in food deserts. If a nutritionally complete diet costs significantly more in food deserts, resident consumers may be constrained from consuming healthier foods. We use IRI InfoScan, a nationwide store-level sales dataset, and calculate a census-tract level Exact Price Index (EPI) based on a food basket defined by the Thrifty Food Plan (TFP). The EPI addresses potential biases from both product heterogeneity and variety availability. We have three central findings. First, prices for common foods are not significantly different between food deserts and non-food deserts. Second, after controlling for differential access to food variety, we find that the EPI in food deserts is 3% to 8% higher than similar census tracts with more store access and 3% to 6% higher than low-access census tracts with higher income. Third, the higher EPI in food deserts is driven by the lack of supermarkets nearby. In sum, residents of food deserts do not face substantially higher prices for common goods that are available in both food deserts and non-food deserts; instead, the higher prices in food deserts are driven by the smaller variety of foods available in food deserts. This finding has implications for the mechanisms through which policymakers may wish to enhance access to food in low-income, low-access areas.

Is the Focus on Food Deserts Fruitless? Retail Access and Food Purchases Across the Socioeconomic Spectrum

Jessie Handbury
,
University of Pennsylvania
Ilya Rahkovsky
,
USDA Economic Research Service
Molly Schnell
,
Princeton University

Abstract

Despite an absence of causal evidence showing that limited access to healthy foods is to blame for unhealthful consumption, policies aimed at improving poor diets by improving access are ubiquitous. In this paper, we use Nielsen household-based and store-based scanner data, TDLinx store characteristics data, and government data sources to examine the healthfulness of household food purchases and the retail landscapes facing consumers. Our goal is to measure the role that access plays in explaining why some individuals in the United States eat more nutritious foods than others. We first confirm that households with lower income and education purchase less healthful foods. We then measure the spatial variation in the average nutritional quality of available food products across local markets, revealing that healthy foods are less likely to be available in low-income neighborhoods. Though significant, spatial differences in access are small relative to the spatial differences in store sales and explain only a fraction of the variation that we observe in the nutritional content of household purchases. Systematic socioeconomic disparities in household purchases persist after access is equated: even in the same store, wealthier and more educated households purchase more healthful foods. Consistent with this result, we further find that the nutritional quality of household purchases responds very little to changes in their retail environment, especially among households with low levels of income and education. Together, our results indicate that even if spatial disparities in access are entirely resolved, over two-thirds of the existing socioeconomic disparities in consumption would remain.
JEL Classifications
  • Q0 - General