Association for Public Policy Analysis & Management

Government transfers in the United States are frequently provided as in-kind rather than cash benefits to induce consumption of specific goods, such as food, housing, and medical care. However, if recipients of in-kind transfers consume more of the good than the value of the in-kind transfer, economic theory implies that a marginal change in the benefit will have the same effect on consumption of the good regardless of whether the transfer is made in kind or in cash.

The Supplemental Nutrition Assistance Program (SNAP; formerly named the Food Stamp Program) is one of the largest in-kind transfer programs in the United States, providing average monthly benefits of $255 to almost 22 million households in 2016. The primary objective of the program is to allow low-income households to achieve a more nutritious diet by providing an in-kind benefit to be used to purchase food.

A large body of research has examined whether SNAP benefits have the same impact on food spending as would an equivalent cash transfer. A central issue is whether SNAP recipients are “inframarginal”, meaning that their desired food spending is greater than their SNAP benefit, and therefore would be expected to treat their SNAP benefits as they would cash. Conversely, SNAP recipients could be “extramarginal”, where desired food spending is less than or equal to SNAP benefits.

We use PSID data from 1977 to 2013, which allows us to examine the food spending of SNAP recipients over a period of almost four decades. The core survey allows us to identify families that participated in SNAP and their benefit amount in the month prior to the survey. PSID respondents provide information on their spending on food at home, food away from home, and food delivered to the home. The survey includes a direct question about whether the family spent more than their SNAP benefit on food at home, which allows for a more precise estimate of the proportion of SNAP recipients that are extramarginal than other data used in prior studies.

We find that roughly 30% of SNAP participants are extramarginal, which is much larger than previous estimates and implies that in-kind benefits provided through SNAP promote food consumption. We also illustrate the importance of the correct identification of extramarginal recipients, which depends on a number of factors, including the ability to distinguish between food at home (which can be purchased with SNAP benefits) and food from restaurants (which cannot), and to adequately identify spending on food at home beyond what is purchased with SNAP benefits. This information could inform future data collection efforts on the food spending of SNAP recipients. Finally, we identify the factors that are associated with the probability that a SNAP recipient is extramarginal. We find that two in five SNAP families with annual incomes below $10,000 rely solely on SNAP for their food at home purchases. This suggests that changes in SNAP benefits will have the greatest effect on the poorest SNAP households.