2017 Research Grants

2017 Healthy Food Behavior Research Grants: Using Behavioral Economics to Promote Healthier, Economical Food Choice

We recently funded 2 proposals presenting projects related to the Supplemental Nutrition Assistance Program (SNAP) that employ behavioral economic strategies to promote purchase of healthy, economical foods in a variety of food retail environments. The funding period lasts from September 1, 2017 to August 30, 2018.

Innovatively Using eCommerce to Promote Healthy Grocery Purchases

PI: Lizzy Pope, University of Vermont

Background: eCommerce grocery is a technology that is poised to fundamentally disrupt the grocery business and have increasing influence on the way consumers purchase food. Unlike a traditional brick and mortar grocery store, there is much potential in an online space to use behavioral economic techniques to promote healthier purchasing for all consumers including Supplemental Nutrition Assistance Program (SNAP) beneficiaries. Thoughtfully thinking through the choice architecture of the online space, and arranging default displays to motivate healthier purchasing are techniques that could have a significant impact on a population level. With the United States Department of Agriculture’s interest in making sure that SNAP beneficiaries can use their benefits in the online space, it is increasingly important for health researchers to understand how eCommerce grocery environments could be subtly manipulated to support healthy purchasing for SNAP beneficiaries as well as all consumers.

Specific Aims: The main objective of this proposal is to use choice architecture and default display manipulations to encourage healthy purchasing in an eCommerce grocery environment. The first specific aim will determine the impact of displaying targeted healthier items on the first webpage of a grocery category on food purchasing. Aim two will examine the effect of displaying nutrition information in green or red font on consumer purchases, and aim three will track the impact of displaying calories and serving sizes in a larger font on Nutrition Facts Panels.

Research Design and Methods: Over the course of one year, an eCommerce grocery site will host a series of three interventions. Each intervention will take place for one month and be followed by a one-month washout period, where the site will return to its baseline design before the next intervention begins. Baseline data will be collected for one month at the beginning of the study period, and supplemented by data from the one-month washout periods. It is expected that food purchasing data from 10,000 orders/month will be collected. For aim one, the “Cereal and Breakfast,” “Bread,” and “Snacks” categories will feature thirty healthy items on the first page of the category. The rate of purchasing for these targeted items when they’re listed first versus baseline will indicate whether putting healthy items first had a positive impact on purchasing. During aim two, items in the “Produce” and “Snacks” sections will have their Nutrition Facts Panels displayed in either red or green font. We will compare the proportion of green to red items purchased during the intervention and baseline periods to determine whether color on Nutrition Facts Panels can impact purchasing. For aim three, all items in the “Snacks” category will display a Nutrition Facts Panel that uses a large font for calories and serving sizes. Purchasing rate of products during the intervention will be compared with baseline purchasing rate to determine whether label font size impacts purchasing. Each outcome measure will be analyzed for all consumers and SNAP beneficiaries.

A Default Option for Health: Improving Nutrition within the Financial and Geographic Constraints of Food Insecurity

PI: Julia Hormes, University at Albany, State University of New York

Background: Food insecurity is positively associated with obesity risk and weight-related health problems. Participation in the Supplemental Nutrition Assistance Program (SNAP) exacerbates the link between food insecurity and excess body mass. The long-term impact of psychoeducational interventions targeting nutrition in SNAP (“SNAP-Ed”) is uncertain, and implementation is costly. Food-insecure individuals face financial and geographic barriers to access to healthy foods, and there is an urgent need for interventions that effectively improve nutrition within these unique constraints. The proposed study will evaluate the efficacy of a novel, non-monetary “default option” intervention targeting food choice in food-insecure adults shopping for groceries online on a typical SNAP budget. The default intervention draws on research in behavioral economics to improve food choice behavior by manipulating choice architecture. The proposed project leverages ongoing research activities and partnerships, with pilot studies at local food pantries providing preliminary evidence for the efficacy of the default approach in the target population. Several states are currently piloting programs making online shopping available to SNAP participants. The planned project represents an important step in evaluating the use of online platforms in disseminating interventions targeting food choice behavior in low-income consumers in a cost-effective, sustainable, and easily scalable manner.

Specific Aim: We will systematically evaluate the efficacy of a default intervention in improving the nutritional quality of foods purchased online and within the financial constraints of SNAP, compared to psychoeducation based on SNAP-Ed. We hypothesize that the default approach enhances nutritional quality in food-insecure individuals more effectively than psychoeducation.

Research Design and Methods: We will execute a prospective randomized controlled trial enrolling food-insecure adults at food pantries in Albany, New York. Following a baseline visit (time 1), participants will be randomized to a psychoeducation or default condition and invited to complete four weekly study visits (times 2 – 5). At each study visit, participants will purchase groceries using the online shopping and home delivery services of a local grocery store and a $48.50 budget (maximum weekly SNAP benefits for a single adult in New York State). At times 2 to 5, the psychoeducation group will review SNAP-Ed materials prior to purchasing groceries. The default group will be presented with a pre-filled online shopping cart containing a nutritionally balanced selection of groceries to which they can make any changes as they wish. We will evaluate between-group differences in the healthfulness of grocery purchases using Healthy Eating Index scores and a repeated measures design. Minimum sample sizes of 34 and 24 are needed for adequate power (≥.80) to detect medium-sized effects in mixed analyses of variance with two groups and two (i.e., time 1 vs. 2, to evaluate the immediate impact of the intervention) or four assessments (i.e., times 2 – 5, to capture learning effects over time). We conservatively account for ~20% attrition and will randomize 21 participants each into the psychoeducation and default groups. Physical and mental health, eating behaviors, and weight will be captured at baseline, end of the intervention, and three-month follow-up.