2016 Research Grants

2016 Healthy Food Behavior Research Grants: Using Behavioral Economics to Promote Healthier, Economical Food Choice

We recently funded 3 proposals that draw on behavioral economics theory to develop and test strategies for improving food choice behaviors. These projects run from October 1, 2016 to September 30, 2017.

Improving Food Choices for Low-income Families with Young Children: A Supermarket Intervention

PIs: John List, Fatemeh Momeni, and Dana Suskind, University of Chicago

Background: Despite decades of funding supporting early childhood development, disparities in health outcomes between children from low and high socioeconomic status (SES) families remain a serious public policy issue. The gap in health outcomes are in part caused by disparities in diet and nutrition, which are particularly harmful and can cause serious developmental risks early in life. Moreover, as children develop their taste preferences for food in early years and it becomes harder to affect preference for (liking) unfamiliar food through experience as they grow older, unhealthy diets during early childhood can shift food preferences away from healthier options as they age, which reinforces unhealthier food choices even further.

Specific Aims: We propose a series of interventions aimed at improving food choices of low SES families with young children. We aim to improve food choices through two channels: 1) targeting parent-child interactions and 2) taste-testing fruits and vegetables at the point of purchase. The two channels of our intervention can improve food choices through priming healthier choices. Inducing parent-child dialogue around healthy food and taste-testing fruits and vegetables can both act as stimuli that prime healthier purchases. However, the effect of our interventions goes beyond the mere priming effect. Since our main targets are families with young children, the effect will also work through guiding children’s attention to, and liking of, healthier food. Research has shown that children influence family food purchases as early as they begin to develop the necessary communication skills, and families tend to make product selections based on what their children will actually eat instead of based on nutritional value. Thus, targeting families with young children at the point of purchase can be effective in improving family food choices.

Research Design and Methods: We propose to evaluate the interventions with a natural field experiment using a 2×2 design. T-control, T-taste, and T-dialogue represent our control, taste testing, and parent-child dialogue promotion treatments, and T-comb is a combination treatment in which we open both channels. Through this combination treatment, we will test whether there are complementarities between the two channels that can amplify their effect on food choice. We will run our interventions in a SNAP grocery store in a low-income neighborhood in Chicago over three months. We will randomize our treatments at day level and will obtain purchasing data from the store. We will also explore whether our treatments have differential effects on families who shop with young children.

A Behavioral Economics Acaitlincaspi_imagepproach to Improving the Selection of Healthy Food Among Food Pantry Clients

PI: Caitlin Caspi, University of Minnesota

Background: This study will evaluate the impact of a behavioral economics intervention on the nutritional quality of food selected by clients at food pantries. In 2014, an estimated 14% of U.S. households experienced food insecurity. Along with programs like the Supplemental Nutrition Assistance Program (SNAP), food pantries are a critical component of the hunger relief system. Pantries face both supply-side issues constrain what is offered to clients, and demand-side issues impede the movement of healthy food through the system. Nationally and locally, hunger relief organizations begun to use behavioral economic strategies to encourage clients to select healthier items. These strategies have almost never been tested to evaluate their impact on client behavior. Strategies that are successful in the food pantry can be translated to the retail setting, and food habits formed at the pantry can alter client purchasing behavior outside the pantry. 

Specific Aim: This study will evaluate the impact of a behavioral economic intervention in two food pantries on the nutritional quality of food selected by clients (primary outcome). We hypothesize that, on average, the nutritional quality of foods selected by intervention clients will be higher after the intervention.

Research design: The study will take place in two food pantries in Minneapolis and St. Paul, Minnesota. Two additional pantries will be used as a control setting, to capture seasonal and secular trends in food distribution and client demand. SNAP-Ed educators at University of Minnesota Extension will help pantries adopt behavioral economic strategies including: (1) modifying food allowances at the pantry to expand choice for more healthy items (e.g., fruits/vegetables) and constrain choice for less healthy items (e.g., processed foods); (2) using item placement and promotion (e.g., bundling, end-caps, signage) to promote healthy items. The evaluation measures pre/post differences in the nutritional quality of food selected by 70 clients at baseline and 70 clients 3 months post-implementation.

Methods: We will recruit clients and perform bag checks after they have selected their food. Participants are eligible if they are ≥18 years old and capable of consent. A survey will be conducted to assess demographics and shopping intentions within the next 3 months; bag checks will capture information on each product selected. Nutritional quality will be measured by the Healthy Eating Index-2010 (HEI-2010), a valid and reliable USDA measure based on the Dietary Guidelines for Americans. Pre and post HEI-2010 scores of pantry items on the shelf will be collected to measure availability changes. Descriptive differences in pantry-level HEI scores in the intervention and control sites will be examined. Data will be entered into Nutrient Data Systems for Research (NDSR), a nutrient analysis software application. Process measures will document intervention changes (fidelity, dose). To test our primary research question, we will examine pre/post changes in average client HEI-2010 scores, controlling for demographics and any change in fruit/vegetable poundage observed at control sites. Secondary analyses will examine the HEI-2010 scores of promoted foods. We expect 80% power to detect a change in HEI-2010 client scores of 5 points between the two time points.

Testing Locally Designed Labels and Social Norm-based Incentives on Food Choice in a Native American Community

PI Christopher Gustafson, University of Nebraska-Lincoln

Background: Rates of overweight and obesity in the U.S. have increased steadily in recent decades, a trend that has affected individuals across the lifespan. The long-term consequences of overweight and obesity are severe and far-reaching, with negative impacts on health, economic outcomes, and quality of life. Significant investments in research and interventions have been made to understand and attempt to reverse the upward trend in unhealthy weights, but many of these efforts have had limited success. However, point-of-purchase strategies that seek to influence individuals while they make food decisions by prompting individuals to think about the healthfulness of their food choices and providing easy-to-process label information about nutritional quality on the shelf or on products show promise.

Many minority populations experience higher rates of overweight and obesity than the U.S. population on average. Minority households frequently face high levels of poverty and limited retail access to healthy foods, both of which have been linked to lower dietary quality and higher rates of overweight and obesity. Native American populations appear to be at particular risk for diet-related health problems. Importantly, none of the point-of-purchase strategies has tried tailoring labels to high-risk populations. Providing relevant information about others’ decisions, commonly referred to as social norms, has been shown to effectively promote healthy behaviors. Indeed, pilot data from a choice experiment conducted with residents on a Native American reservation in South Dakota suggest that building social norms into healthy food labels increases healthy choices more effectively than a generic healthy food label that had been found to be highly effective in earlier research. This result informed the development of a shelf-based food label that will highlight healthy, low-cost foods in a grocery store on the reservation. Messages prompting individuals to try to meet healthy food-related consumption goals may also be an effective way to encourage high quality diets. Research on cognitive load and working memory, however, suggests that the scope of these messages may need to be carefully considered. More narrowly defined goals may be easier to keep in mind when individuals are actively shopping than broad goals.

Specific Aims: The specific aims of the project are to 1) assess the impact of a healthy food shelf label, which was designed with the input of community members to communicate social norms, on food sales, and 2) to examine effects of non-financial incentives that are broadly or narrowly defined on food choices.

Research Design and Methods: We will evaluate the effect of the first aim of the project by examining the sales of healthier and less healthy foods before and after the implementation of the social norm label in the store on the reservation that is using the labels to highlight healthier foods and a control store in a similar community. We will investigate the second specific aim through in-store research with shoppers. Shoppers will be exposed to sign-based messages encouraging the purchase of broadly or narrowly defined healthy food goals. We will examine individual-level purchases of promoted, healthy, and less healthy foods under the broadly versus narrowly defined food goals.