The Effects of SNAP on Nonfood Expenditure: An Instrumental Variables Approach

Year: 2016

Research Center: Institute for Research on Poverty, University of Wisconsin-Madison

Investigator: Almada, Lorenzo N. and Jaehyun Nam

Institution: Columbia University

Project Contact:
Lorenzo N. Almada
Columbia University
Columbia Population Research Center
730 School of Social Work
1255 Amsterdam Avenue
New York, NY 10027
Phone: 706-566-4548


The Supplemental Nutrition Assistance Program (SNAP), formerly known as the Food Stamp Program, is the Nation’s largest food assistance program. It aims to reduce food insecurity and provide adequate levels of nutrition to low-income Americans. A vast body of literature has documented SNAP’s effectiveness in increasing food expenditures, reducing food insecurity, and to a lesser extent, improving the nutritional quality of diets and overall health of participants. However, relatively little is known about the potential effects of SNAP participation on expenditures on nonfood goods and services. Theoretically, SNAP participation could free up income for households to spend on other goods and services, yet little is currently known about whether, or in what ways, SNAP participation affects nonfood consumption behaviors.

This study fills this research gap by investigating the causal effects of SNAP participation on household expenditures, with a central focus on nonfood-related expenditures. The study examines total nonfood spending as well as six subcategories of nonfood expenditures using the 2000-11 waves of the U.S. Bureau of Labor Statistics’ Consumer Expenditure Survey (CE). Selection into SNAP is addressed by employing an instrumental variables (IV) approach that exploits variation in State-level policies and program administration to instrument for SNAP participation, conditional on household and State-level characteristics. Respondent misreporting is addressed by adopting an approach, based on parametric methods for misclassified binary dependent variables, that produces consistent estimates when using instrumental variables.

Findings of this study establish that SNAP-reporting households are more disadvantaged than income-eligible, non-SNAP households, both in demographic characteristics (for example, income, education, homeownership, and other welfare receipt) and annualized household expenditures. This study also determines that SNAP misreporting is prevalent among low-income households in the CE, particularly among households without children. The findings indicate that households are primarily underreporting SNAP receipt by between 5 to 40 percent, implying that conventional IV estimates are overstated by nearly 50 percent, on average.

The analyses of this study reveal that, after adjusting for misreporting, SNAP participation among low-income households increases expenditures on nonfood by 41 percent, or approximately $5,400 per year, while increasing expenditures on food by nearly 60 percent, or approximately $2,900 per year. Across nonfood subcategories, the results indicate that SNAP participation increases education and enrichment expenditures by approximately 50 percent, housing expenditures by roughly 60 percent, utilities expenditures by 115 percent, and transportation expenditures by 45 percent. The results are generally similar but measured less precisely when separately examining households with and without children. Overall, the findings of this study indicate that SNAP is serving its intended purpose of increasing household expenditures on food, while also allowing households to allocate some of their income to various nonfood spending categories.