Association for Public Policy Analysis & Management

Supplemental Nutrition Assistance Program (SNAP) benefits are disbursed to recipients once per month. Outcomes such as food expenditures and food intake have been demonstrated to be non-constant throughout the course of the “SNAP month,” or the period of time beginning with the dates on which program participants receive their benefits (Hastings and Washington, 2000; Shapiro, 2005; Smith et al., 2016; Todd, 2015; Wilde and Ranney, 2000). Previous research has primarily evaluated the consumption of food that has been purchased with SNAP benefits or cash over the course of the SNAP month, taking advantage of variation in the timing of SNAP benefit disbursements across states. Less frequently studied is the acquisition of free food throughout the course of the SNAP month. In this paper, I investigate how the acquisition of free food – that which is obtained from sources of emergency food assistance, Meals on Wheels, family and friends, community centers, and modes of self-production – varies throughout the course of the SNAP month, hypothesizing that free food is acquired unevenly through the SNAP month.

To investigate patterns of free food acquisition over the course of the SNAP month, I use data from the United States Department of Agriculture’s National Household Food Acquisition and Purchase Survey (FoodAPS), a nationally representative survey of 4,826 U.S. households containing information about daily food acquisitions and purchases over a one-week period. FoodAPS was expressly designed to provide a more comprehensive picture of the ways in which food prices, local food environments, and participation in food and nutrition assistance programs affect the amount and types of food that households acquire, or the extent to which low-income households rely upon alternative sources of food, such as food pantries or family and friends, to supplement their food purchases (United States Department of Agriculture). FoodAPS is particularly well-suited toward answering my research question for three reasons. First, FoodAPS contains administratively-matched SNAP participation data – not often readily found in survey data – including the number of days that have transpired since a SNAP-participating household received its last SNAP benefit. Second, FoodAPS contains information on where reported food events occurred, including geocoded data, the day on which a food event occurred, and the specific food items that were obtained during a given food event. Third, FoodAPS contains data on food retail environments, such as the number of different types of food retailers in each Census block group. In addition, FoodAPS includes information on the location of food banks, food pantries, and soup kitchens, as well as the volume of food distribution and hours of operation, from Feeding America. Thus, in this study, I am able to conduct multivariate regression analyses examining the impact of the timing of SNAP benefit payments on the acquisition of free food, controlling for several important confounding factors. This study will lend insight into the take-up and provision of non-SNAP food assistance programs, and how the once-per-month SNAP benefit payment structure impacts other forms of food assistance throughout the course of the SNAP month.