January 2, 2026
Polling Summary: Support For Stronger Guardrails In SNAP Regarding Income & Asset Tests, Citizenship, And Work Requirements
A new national survey shows voters overwhelmingly support commonsense federal and state reforms to the Supplemental Nutrition Assistance Program (SNAP), formerly known as food stamps. A survey of 2,000 registered voters was conducted for the America First Policy Institute by RMG Research on December 12, 2025.[1]
Survey Highlights:
- 89% of voters agree SNAP benefits should only go to those most in need.
- 86% of voters favor increasing fraud protection measures in SNAP.
- 78% of voters agree SNAP should have mandatory income and asset limits.
- 77% of voters favor requiring state welfare agencies to collaborate and share SNAP data with the federal government.
- 70% of voters agree that able-bodied adults should be required to work to receive SNAP benefits.
- 64% of voters agree non-citizens should not have access to SNAP benefits.
- 62% of voters favor closing SNAP loopholes that allow households with significant assets access to SNAP benefits; 53% of voters believe states should be required to close these loopholes.
- 59% of voters favor states banning purchases of unhealthy foods with SNAP benefits.
Additional Highlights:
- Every demographic, sociographic, and party-affiliation strongly agreed with the statement: “SNAP benefits (formerly food stamps) should only be for those most in need.”
- Similar across-the-board demographic, sociographic, and party-affiliation support occurred for the following statements: “SNAP should have mandatory income and asset limits,” and “If someone is able to work, should he or she be required to work or be seeking work in order to receive SNAP benefits.”
[1] Certain quotas were applied, and the sample was lightly weighted by geography, gender, age, race, education, internet usage, and political party to reasonably reflect the nation’s population of registered voters. Other variables were reviewed to ensure that the final sample is representative of the population. The margin of sampling error for the full sample is +/- 2.2 percentage points.