An independent project ·Why this exists ·How this is made
Asheville · Western North Carolina
Step Up AVL
Step Up AVL Food security · The SNAP rewrite · Part 2 of 5

Who loses SNAP now?

The 2025 budget law is already pushing people off food aid, faster in North Carolina than almost anywhere. Who is losing it and going hungry, why the work rules do not do what they promise, and why no official count is coming.

Picture a server in West Asheville whose winter hours slipped under the line for two slow months. The renewal notice went to an old address. The proof-of-hours form came due the Friday she worked a double. The case closed the week after. She is a composite, not one person, but every step of that is how the new rules work.

A law that moves this much money decides who eats. That part is not a prediction, and it is not years away. It is already in the numbers.

The money half of the story comes next in this series. It asks who now pays for what the program costs: who pays for SNAP now. This piece follows the people. New to SNAP? Start with what it actually is and how little it gives →

Already happening

The rolls are already shrinking

Nationwide, SNAP enrollment fell to about 37.8 million people by February 2026, roughly 4.3 million fewer than a year before. That is a drop of about 10 percent. The Food Research & Action Center counts more than 5.5 million gone between January 2025 and March 2026. North Carolina's drop is among the steepest anywhere, from about 1.5 million people in January 2025 to under 1.2 million by March 2026. That is down about 22 percent. The state health department expects the new work rules alone to cost about 90,000 more North Carolina adults their benefits. The drop is not from any single change. It comes from the wider work rules, from renewals that now come more often and trip people up on the paperwork, and from new limits on which noncitizens can get help, all landing at once.

One honest caution on the timing. Much of that drop came before the new work rules even took effect on December 1, 2025. So the 2025 law did not cause all of it. The swollen pandemic-era rolls had been unwinding for a while on their own. What the law does is take that drift and make it policy. The roughly 90,000 the state expects to lose their benefits under the work rules are coming off now, through the recertifications that run into the summer of 2026, after the window these early numbers cover. So the sharpest part of the law's own effect is still ahead of the count, not behind it.

Supporters of the law read that drop differently, as the program dropping people who never should have been on it. Some of it is exactly that. Ineligible cases and paperwork errors get cleared out, and the U.S. Department of Agriculture says its removals include fraud, like benefits drawn on dead people's Social Security numbers. But the department has not shown how many of the people who left were actually ineligible, and researchers who went looking for a happier explanation did not find one. There is little sign the economy improved enough to explain a drop this size. The more likely story is much simpler. Eligible people are losing benefits because the rules got harder to satisfy, not because they stopped needing food.

Supporters make one promise about the work rules that can be checked against evidence: that they move people from aid into jobs. Economists have studied earlier versions of the same rule. The most careful of that work, from the National Bureau of Economic Research, found the requirement cut SNAP enrollment by about half among the adults it covered while producing no measurable rise in employment. A fresh 2026 study reached the same conclusion: tougher requirements thinned the rolls and left employment flat. Even the studies most favorable to the rules find at most a small bump in work, and a shaky one. People left the rolls. On the whole, they did not end up in jobs. That is the gap between what the rule is sold as and what the record shows it does.

-22%
NC SNAP enrollment, Jan 2025 to Mar 2026 (among the steepest state drops)
~90,000
NC adults expected to lose benefits under the new work rules (NCDHHS)
~776,000
Fewer children on SNAP across the 12 age-reporting states (ProPublica)
Hit hardest

The rule falls heaviest on people with no address

The work rules do not fall evenly. They fall hardest on one group in particular: people without a stable home. Many of them already work, in day-labor and gig jobs that pay in cash and leave no clean record for a caseworker to see. The paperwork can drop them even when the work is real. They are also the people the old law had tried to shield: Congress added an exemption for homeless adults in 2023, and the 2025 law repealed that exemption before it had fully taken hold.

That is a large enough story to stand on its own, and we tell it separately in who we call homeless, which follows what happens when a work rule is aimed at people whose work was never the problem.

The youngest

Children are leaving fastest of all

Children are leaving in outsized numbers. ProPublica looked at the twelve states that report enrollment by age and found about 776,000 fewer children on SNAP, about 46 percent of those states' total drop. The Center on Budget and Policy Priorities put the figure for those same twelve states at more than 700,000, and noted the real national number is higher because most states do not report by age. Children are not subject to any work rule. They lose SNAP when their whole household's case closes, and more households are closing, mostly under the weight of the new paperwork rather than because their incomes rose.

Off the books

And the country stopped measuring

The main way the government measured hunger is also gone. In September 2025, the USDA ended the survey behind its annual food security report, the food questions the Census Bureau had asked households every December since the 1990s. The department called it redundant. The edition covering 2024 was the last one. It is the whole measurement that stopped, not just one report. The other hunger estimates people might point to are built on that same survey. Whatever the reason, there is no plan to count hunger for the years these changes take effect. Anyone waiting for the official numbers to show what happened will be waiting for a report that is not coming.

Here is a simple test of whether the law's new error-rate penalty is really about accuracy. It scores a state the same for underpaying a hungry family as for overpaying one, and the eligible people a state wrongly turns away are not counted at all. The bill lands on the state's books, but the cheapest way to shrink it is to have fewer families on the rolls. The full mechanics are in who pays for SNAP now.

The bottom line

What it means for Buncombe

Here is what all of this means for people at home.

If you get SNAP in Buncombe County: your monthly benefit is set by a federal formula, and none of these changes cut that dollar amount. What they change is who qualifies. The new work rules reach more people, adults up to age 64 and parents of teenagers, so some people who get help today will lose it. The state expects about 90,000 North Carolina adults to fall off. Buncombe's slice of that is not public yet, but the county is not exempt. If your case is still open, expect renewals to come around more often, and keep your proof of income and rent current, because a missing form is now one of the most common ways a case closes.

If you do not get SNAP: this still reaches you. About 29,000 of your Buncombe neighbors are on it, roughly one in ten people in the county. The roughly $60 million a year they spend with it runs through the same checkout lines you stand in, from Ingles, the grocer headquartered up the road in Black Mountain, to the tailgate markets. When the rolls shrink, that spending shrinks with them. And you help pay for the changes: the county and state cost of all this is the next piece in this series, who pays for SNAP now →

The people who need the help did not change this year. What changed is how many of them the law is willing to feed.

The takeaway

For thirty years, the government at least counted the hunger it was preventing. Now it has stopped even measuring, in the same season it made food aid harder to get and keep. Enrollment is falling faster in North Carolina than almost anywhere, with the steepest losses among children who never had a work rule to fail. The people who need help did not change this year. What changed is how many of them get fed, and whether anyone will be left counting.

For more information see: www.stepupavl.org

The server in the opening is a composite, not one person: every step in her story is drawn from how the new work-hour and renewal rules operate, documented below. We never present a composite as a real individual.

Sources. The law: the One Big Beautiful Bill Act, P.L. 119-21 (H.R. 1, signed July 4, 2025), SNAP provisions summarized by the Congressional Research Service (R48552); the money mechanics this piece refers to (the payment error-rate penalty and the county administrative cost shift) are detailed in the companion piece, "Who Pays for SNAP Now." Work rules (time limit through age 64; parents of children 14 and older; exemptions removed for veterans, people experiencing homelessness, and former foster youth, protections Congress had added in 2023): USDA Food and Nutrition Service implementation guidance and CRS R48552; effective December 1, 2025 in North Carolina. On whether the work rules move recipients into jobs, the peer-reviewed evidence finds they sharply reduce participation without measurably raising employment (National Bureau of Economic Research working papers 28877, "Employed in a SNAP?", and 32441, "Work Requirements With No Teeth Still Bite"); a fresh 2026 Brookings Institution analysis by Lauren Bauer and colleagues reached the same conclusion (reported by Stateline, "SNAP work requirements don't boost jobs, but drop participation, research finds," April 10, 2026, and Boise State Public Radio). Enrollment: about 37.8 million nationally in February 2026, down about 4.3 million year over year (ProPublica, from USDA FNS data); more than 5.5 million from January 2025 through March 2026 (Food Research & Action Center tracker); North Carolina about 1.5 million in January 2025 to about 1.3 million as of January 2026 (NC Newsline, May 12, 2026, from NCDHHS), down about 22 percent from January 2025 through March 2026 (FRAC); about 90,000 NC adults expected to lose benefits under the work rules (NCDHHS, via NC Newsline). On timing: much of that year-over-year decline predates the December 1, 2025 work-rule start and reflects the post-pandemic unwinding of the rolls (enrollment fell only about 2 percent between December 2025 and March 2026); the law's own work-rule effect, the roughly 90,000, processes through recertifications into the summer of 2026 (NC Health News, "New SNAP requirements could stress county budgets," March 17, 2026; Governing SNAP-enrollment analysis). On why enrollment fell: USDA officials attribute part of the decline to removing ineligible and fraudulent cases (Secretary Rollins cited about 700,000 cases since February 2025, including duplicate multi-state and deceased-Social-Security-number claims), but the department has not published a breakdown of how many removals were confirmed ineligible, "improper payment" is mostly over- or under-payment on eligible cases rather than fraud, and analysts found little evidence that improved economic conditions explain a drop of this size, pointing instead to administrative burden (USDA, as reported by CBS News and NPR; Center on Budget and Policy Priorities SNAP tracker). The peer-reviewed employment evidence ranges from no effect to at most a small, statistically weak positive one (American Enterprise Institute, "SNAP and Employment: What Is the Evidence?"). The North Carolina decline reflects several changes landing together: the expanded work requirements, more frequent recertification, and narrowed noncitizen eligibility, with SNAP now limited to citizens, lawful permanent residents, Cuban and Haitian entrants, and Compact of Free Association migrants, and refugees, asylees, and other humanitarian categories excluded (states implementing by November 1, 2025 per USDA FNS) (NC Newsline; NCDHHS; USDA FNS). Children: 776,134 fewer children across the 12 age-reporting states, about 46 percent of those states' decline (ProPublica, June 17, 2026; rounded to about 776,000 in the body); the Center on Budget and Policy Priorities independently estimated more than 700,000 across those same twelve states, noting the true national figure is larger and unknown because most states do not report child-level data. The payment error rate referenced in the aside measures how often a state pays the wrong amount, counting both overpayments and underpayments and excluding fraud, which is measured separately; the 2025 law ties part of a state's benefit cost to that rate, and because wrongful denials of eligible people are not counted as payment errors, budget analysts warn it pressures states to thin their rolls (USDA FNS Quality Control; Food Research and Action Center; Center on Budget and Policy Priorities); the full mechanics are in "Who Pays for SNAP Now." Local: about 29,000 Buncombe County SNAP participants with an average benefit of $171 per person per month, spring/fall 2025 figures (Buncombe County HHS; NCDHHS April 2025 county enrollment table, Buncombe 29,123); the roughly $60 million a year in local SNAP spending is derived from those two figures (about 29,000 people times $171 a month over a year); the "one in ten" framing is derived from that count against the county's population (29,123 participants against a resident population of 274,360, U.S. Census Bureau Vintage 2024 estimate, is 10.6 percent, about one in ten); Ingles Markets is headquartered in Black Mountain, North Carolina (company filings). Discontinued measure: in September 2025 the USDA terminated the Current Population Survey Food Security Supplement, the annual December question set (Census Bureau and Bureau of Labor Statistics) behind its roughly 30-year Household Food Security report, characterizing it as redundant; the edition covering 2024 (published December 30, 2025) is the last, and no measure covering 2025 or 2026 is planned. Because the leading local estimates (for example Feeding America's Map the Meal Gap) are built on that same supplement, ending it degrades the alternatives too; anti-hunger analysts dispute the redundancy rationale (USDA press release, "USDA Terminates Redundant Food Insecurity Survey," September 20, 2025; NPR; Food Research and Action Center, "Threats to Food Security Data: Why the 'Redundancy' Claim Doesn't Hold Up"; Feeding America, Map the Meal Gap methodology; DTN Progressive Farmer; Union of Concerned Scientists). National figures are labeled national; derived figures are labeled derived. This is a fast-moving policy area; figures are current as of July 2, 2026.

Found an error? Tell us and we will correct it. The framing and conclusions here are our own.

Will YOU step up?

Every group in our directory is local, vetted, and doing this work right now. Pick one and back it.