What cutting junk foods from SNAP could mean for millions of recipients

• A ban would require action through Congress.
• Late last month, House Republicans voted to pass a budget resolution that sets the stage for $230 billion or more in cuts to agriculture programs, with a large chunk expected to come from SNAP.

State of play: The Food and Nutrition Act defines food for SNAP purposes as any food or food product for home consumption, with some limited exceptions like alcoholic beverages or hot foods for immediate consumption.

• In order to narrow that definition, either Congress would need to change the law or a state would need to propose and get approval for a demonstration project to test that, Katie Bergh, senior policy analyst at the Center on Budget and Policy Priorities, told Axios.
• "This is something that we've seen a handful of states request in the past, where they essentially are asking the Department of Agriculture to approve a request to restrict the foods that SNAP participants within their state can purchase in some form," she said.
• But no such requests has ever been approved under either Republican or Democratic presidents, including under the first Trump administration.

Catch up quick: Newly-confirmed Health and Human Services Secretary Robert F. Kennedy Jr. and Agriculture Secretary Brooke Rollins have indicated they're in favor of removing sugary drinks and processed foods from SNAP.

• "The one place that I would say that we need to really change policy is the SNAP program and food stamps and in school lunches," Kennedy said on Fox News last month. "There, the federal government in many cases is paying for it. And we shouldn't be subsidizing people to eat poison."
• Rollins echoed the sentiment, telling reporters at the White House, "When a taxpayer is putting money into SNAP, are they OK with us using their tax dollars to feed really bad food and sugary drinks to children who perhaps need something more nutritious?"

By the numbers: More than 41 million people in the U.S. received SNAP benefits in 2024.

• The average SNAP benefit per person in fiscal year 2025 is $6.16 per day, according to estimates from the Center on Budget and Policy Priorities.

Zoom in: There are a number of risks with the proposal to cut foods from SNAP, including logistical and cost concerns, access in food deserts, equity issues and questions over how to measure success and behavioral changes among consumers.

• Anything that increases administrative burden affects other parts of the program at the state level, Gina Plata-Nino, SNAP Deputy Director at the Food Research & Action Center, told Axios.

Case in point: There would be a large impact on retailers. Those that are not large scale like Walmart, with resources to change markings on SNAP-approved foods, could decide not to operate the program at their stores at all.

• Deciding which foods to cut also presents questions on how to define and measure junk foods, Plata-Nino said, questioning whether it'll be through sodium or sugar content.

• Orange juice, for example, has a lot of sugar but is important for diabetics having medical issues, she noted. Cheese has a higher content of sodium than some chips, and milk has a higher fat content than other drinks.
• "Are we going to ban milk and cheese?" she questioned.

Food deserts are areas where residents have limited access to affordable and healthy food, particularly fresh fruits and vegetables.

• In such areas, individuals may be driving two hours each way to a grocery store and so they're unable to go as often, Gina Plata-Nino said.
• Because fresh produce doesn't last a month until they're able to go again, people there may buy food in bulk, processed items that last longer or opt for frozen options — foods that could potentially be cut under the proposal.

Between the lines: The diets of Americans across income levels are falling short of what experts recommend.

• Bergh said aside from feasibility and cost concerns, the premise of the junk food cutting effort is "fundamentally misguided."
• "Contrary to some of the claims that proponents of these efforts have been making, there's actually pretty extensive research linking SNAP participation to better health outcomes and lower medical costs," Bergh said, noting limited data on what SNAP participants buy.

What she's saying: The data available shows there's no meaningful difference in the types of foods people are purchasing with SNAP versus other payment methods.

• "So it's pretty troubling that the solutions being proposed here are ones that really only single out the lowest income Americans in a way that's really stigmatizing and burdensome for them," Bergh said.
• "Just as a basic principle, everyone should have the same ability to choose the foods that best meet their needs regardless of how I'm paying at the checkout line."

Go deeper: America's food aid gap, mapped