Project:
The Impact of the Monthly SNAP Issuance Cycle on Supermarket Purchasing and Implications for Policy Change

Year: 2019

Research Center: Tufts University/University of Connecticut (UConn) RIDGE Program

Investigator: Eric B. Rimm and Rebecca L. Franckle

Institution: Harvard University

Project Contact:
Eric B. Rimm
Harvard T.H. Chan School of Public Health
665 Huntington Avenue
Boston, MA 02115
Phone: 617-432-1843
Email: eric.rimm@channing.harvard.edu

Summary:

The Supplemental Nutrition Assistance Program (SNAP) provides financial assistance for food purchases to one in seven Americans. SNAP benefits are issued once monthly, which may negatively impact SNAP participants because most households spend the majority of benefits in the first 2 weeks after issuance and may not have sufficient money for groceries at the end of the month. A number of potential issues related to single-day benefit distribution have been identified in the literature, including a decline in energy intake over the benefit cycle and increased hospital admissions over the course of the month. Prior research has also demonstrated that food spending peaks sharply during and immediately following SNAP benefit issuance; thus grocers may be inclined to target SNAP customers on these days by changing prices or promoting certain high-demand items such as sugar-sweetened beverages. However, evidence is lacking on cyclical purchasing patterns for specific foods by SNAP participants.

Furthermore, policy changes to SNAP to promote healthier purchases are currently being debated, but current public perspectives on these program changes are not well understood. A better sense of the public pulse on key policy alternatives for the program will be critical to inform future intervention studies and SNAP policy discussions. This mixed-methods project addresses three critical aims related to SNAP issuance and grocery purchases using the following methodology:

Aim 1: Supermarket sales data from a panel of 951 individually tracked shoppers over 6-8 months were analyzed to understand the impact of SNAP issuance on purchasing patterns. Difference-in-difference models were used to examine purchasing patterns of selected food groups by SNAP participants versus SNAP nonparticipants in the first 2 weeks following receipt of benefits compared to the last 2 weeks of the SNAP cycle, adjusting for covariates (race/ethnicity, sex, household size, season, and intervention arm). Analysis was completed using SAS 9.4 software.

Aim 2: Adult supermarket shoppers, including both SNAP participants and nonparticipants, were recruited through existing relationships with a retail partner in Maine. In December 2017, focus groups gauged shopper awareness of the impact of SNAP policy on shopping behaviors and explored perceptions of the benefits and barriers of potential SNAP policy changes. In accordance with grounded theory, focus group data were collected and analyzed iteratively to identify common themes. All focus groups were transcribed verbatim, and analysis was completed using Nvivo 12 Pro Software.

Aim 3: Questions about six possible SNAP policy changes (e.g., restricting the purchase of sugary drinks) were inserted into a national telephone survey, the Social Science Research Solutions (SSRS) Omnibus, May 8-31, 2018. Responses were weighted for demographic variables to be nationally representative of U.S. adults.

Aim 1 Findings: Sales were analyzed for a total of 951 shoppers in Maine (248 SNAP participants and 703 nonparticipants), with an average of 27.4 transactions per shopper available for analysis. Shoppers were predominantly female (84 percent) and non-Hispanic White (94 percent). For the 248 SNAP participants, adjusted mean spending significantly decreased across all examined food groups from the first 2 weeks to the last 2 weeks of the monthly SNAP benefit cycle (p<0.05). The relative decline in spending by SNAP participants varied across the examined food categories (sugar-sweetened beverages: -29 percent, fruits: -25 percent, vegetables: -24 percent, red meat: -38 percent, poultry: -53 percent, convenience foods: -40 percent). Difference-in-difference estimators comparing SNAP participants to nonparticipants were statistically significant (p<0.05) for all food groups examined.

Aim 2 Findings: A total of 35 participants took part in six 1-hour focus groups, about a quarter of whom were current SNAP participants. Four main themes were identified surrounding potential changes to SNAP policy: (1) participant welfare, (2) retailer impact, (3) administration, and (4) political beliefs. The participant welfare theme discussed budgeting, autonomy, transportation, health, and stigma. The retailer impact explored inventory, staffing, sales and marketing, and store crowding. The administration theme included cost, implementation, timing, and fraud. Political beliefs considered distrust and the role of government. These findings highlight a range of barriers, beliefs, and opportunities relevant to potential modifications to SNAP policy.

Aim 3 Findings: A total of 1,073 adults, including 387 SNAP participants, took part in the national telephone survey. Overall, support for SNAP policy changes was high. Of the proposed policy changes, support for removing sugary drinks from the list of allowable products for purchase using SNAP was lowest (63 percent), while support for additional benefits for fruits and vegetables was highest (83 percent). Support for removing sugary drinks was higher among nonparticipants than SNAP participants (64 percent versus 48 percent, p<0.0001) as was removing candy (69 percent versus 48 percent, p<0.0001). Support for removing sugary drinks was supported by a majority of each political party but varied significantly (71 percent Republican, 56 percent Democrat, 65 percent Independent, p=0.02). Support for additional SNAP benefits based on the amount of fruits and vegetables purchased was stronger among SNAP participants (90 versus 77 percent for nonparticipants). Among SNAP participants (N=387), 35 percent preferred the current policy of receiving benefits once a month, 37 percent preferred changing the policy to twice a month, and 28 percent had no preference. Despite debate among policy makers, respondents supported proposals to improve the nutritional impact of the SNAP program, regardless of participation in the program or political ideology.