The part of SNAP we can actually fix

So the realistic place to start is not a bill. The 2025 law cut about $187 billion from federal nutrition spending over ten years, by the Congressional Budget Office's estimate, and analysts call it the largest cut in the program's history. If that record holds, part of what was cut will eventually come back. It will come back through Washington, years from now, and only after the damage has been counted. Betting this piece on a quick reversal in Congress would be a poor plan.

There is a second reason to set the legislative road aside. It needs saying only once. As of mid-2026 the same party holds the North Carolina General Assembly, both chambers of Congress, and the White House, and that governing majority is the one that enacted the 2025 cut. The cut itself can only be undone in Washington, where a near-term reversal would mean that majority voting down its own law; a rescue from the General Assembly would mean the same majority spending state money to blunt a law its own party passed. Neither is a claim about which party is right. It is the arithmetic of who currently holds the votes, and it means the legislative path is the blocked one for now.

Which is why this piece looks past the legislature, toward the things that do not need a new law. The useful question at the end of a series like this one is narrower and closer to home: while the federal pendulum takes its years, what can actually move here, and who moves it?

For most of SNAP's history the deal was steady: Washington paid for the food and split the cost of running the program, and in North Carolina, one of ten states where counties run SNAP, the counties carried the bulk of the state-and-local share. The 2025 law changed both halves. Starting in October 2026, the federal government drops from paying half of the administrative cost to a quarter. That pushes about $69 million more a year onto North Carolina's counties, and roughly $16 million onto the state. For Buncombe that is about $2 million from October 2026 through June 2027. After that it runs about $2.8 million a year. The county budget it comes out of is broken down in our Buncombe County budget brief. Then, starting in 2028, the same law makes the state pay a share of the benefits themselves if the error rate stays too high. North Carolina's 2026 budget passed that bill straight down to the counties: if the state owes an error-rate penalty, it covers the tab by withholding it from county sales-tax revenue. The county was already paying to run SNAP; now it pays more, and it is on the hook for a share of the error-rate bill too.

Those two rising bills work differently. The county pays most of what it costs to run SNAP, and its budget controls that spending directly. The error-rate penalty is set by a statewide score instead, and nothing one county does moves that score much. Either way, real county money is now on the table, which makes the county budget the first place to look.

Fund the caseworkers. County caseworkers enter the data that becomes the state's error score, and the county now carries most of a nonfederal share that just rose to three-quarters of the cost of running that office. A well-run Department of Social Services is squarely in the county's interest: it keeps eligible families from being dropped over a missed form, and it holds down the churn and backlogs that waste the administrative dollars the county now largely pays. The new error-rate penalty is a different matter, and mostly a shared one. It is triggered by the statewide rate, which no single county's accuracy is likely to move. What a county owes on that penalty depends more on the size of its sales-tax base than on how well its own office runs. Part of the bill is a flat charge against every county's sales tax. The rest falls on counties with attributed errors, and even that share is weighted by sales-tax base as well as by errors. For a high-sales-tax county like Buncombe, the bill lands hard either way. Fund the office for those reasons, not to save money on the penalty (the mechanics are in Who Pays for SNAP Now). Starving the office is a false economy. One line item, one vote the commission controls.

Back the navigation layer. The near-term harm is paperwork, so support the groups that fight it directly. MANNA FoodBank's food helpline walks neighbors through applications and renewals. Pisgah Legal Services takes on wrongful denials and appeals. Both run on donations and volunteers. A gift to either needs no vote and nobody's permission, which makes it the easiest of these to actually do this year.

Widen the produce match. The Appalachian Sustainable Agriculture Project, or ASAP, runs Double SNAP at about five Buncombe farmers markets, where it matches benefit dollars spent on fruits and vegetables dollar for dollar. Funding comes from outside SNAP, through a federal nutrition-incentive grant and local philanthropy. The shopper's dollar buys twice the produce and the farmer keeps the whole amount. A local foundation or a single large donor could widen the match this season without touching the SNAP statute at all.

Screen for it at the clinic. Not everyone who qualifies for SNAP signs up. Nationally, about 88 percent of eligible people were enrolled in 2022, the highest share on record. Among eligible seniors, only about half were enrolled. A health system that asks about food at intake and refers patients to the navigation layer reaches eligible people no form ever finds. Food is also becoming a second, better-funded line of care. Doctors can write prescriptions for fruits and vegetables. Some patients get meals designed around a medical condition. Medicaid, hospitals, and insurers pay for both, and none of it touches SNAP. MAHEC and Dogwood Health Trust already work this ground locally. North Carolina's Healthy Opportunities Pilots showed the Medicaid version can pay for food, but the state paused that program in mid-2025 for lack of funding. While it is on hold, a clinic working with area farmers is the better place to start a smaller produce-prescription program.

There is a plainer way to say why closing that gap is worth it. The benefit dollars are almost entirely federal, so every eligible resident who enrolls brings outside money into local grocery stores. In a weak economy, USDA estimates that each SNAP dollar generates about $1.50 in economic activity. That money lands with local businesses, not the county treasury, so it is economic development for the community rather than a rebate to the county budget, on top of the help it gives the household that enrolls.

Get the error rate under 6 percent the right way. North Carolina's error rate is 7.36 percent, down from 10.21 percent. That lands it in the first tier of the benefit cost-share that starts in 2028. At that rate the General Assembly's fiscal staff puts the bill at about $150 million a year, and the state's 2026 budget passes it on to counties by withholding their sales-tax revenue. Under 6 percent, that bill is zero. The rate is an annual score, not a number the state can flip in a quarter. The federal year that ends September 30, 2026 is the one to watch. For the first bill the state can use that rate in place of its 2025 rate, whichever is lower. The same 2026 rate sets the 2029 bill outright. More than half of payment errors trace to income changes. USDA's own simplification options cut the number of reports where those errors happen without changing who qualifies: simpler reporting, longer certification for stable households, and the streamlined path for seniors and people with disabilities. The error rate counts bookkeeping mistakes, overpayments and underpayments, not fraud. The state can work toward the target by fixing paperwork, and it should say so by publishing churn and wrongful-denial counts next to the error rate, so the score shows the paperwork improved rather than families getting dropped.

Cut the renewal paperwork. SNAP does not allow automatic renewal the way Medicaid does. But a state can make fuller use of what SNAP already allows: the Elderly Simplified Application Project and data matching, which use records the state holds to waive the interview and extend certification periods for stable, fixed-income households. That keeps seniors and people with disabilities from being dropped over a form. This is the administrative side of the enrollment story in Who Loses SNAP Now. Enrollment has already fallen from about 1.47 million in March 2025 to about 1.25 million in June 2026. Across North Carolina, about 90,000 adults are projected to lose SNAP under the expanded work rules.

Open restaurants to people without a kitchen. A state can elect the Restaurant Meals Program, letting homeless, elderly, and disabled recipients buy hot meals at participating restaurants. Choosing it takes no change in eligibility law, though the state still has to recruit restaurants, which must each be authorized by the USDA, and set up the payments. It reaches the homeless recipients described in Who We Call Homeless, along with the elderly and disabled, people who often have no way to cook what SNAP buys.

Get the card working where people are. The state can do what California, Oklahoma, and Alabama have already started: add a chip to every EBT card, which blunts the magnetic-stripe skimming that drains benefits by the hundreds of millions. Those states acted without waiting for a new federal law, working through their card vendors and the USDA. The national chip standard has been set since 2024. What is missing is the money and the store card readers to finish the switch everywhere. Online SNAP purchasing already runs statewide, and only the USDA can authorize a retailer. But the state can recruit rural grocers, walk them through that federal step, and use non-SNAP money to cover the delivery fees that keep online ordering out of reach in the mountains.

Count the hungry. Washington ended its annual hunger report the same season it cut the aid. A state agency or a university can publish that count instead. Every restoration in the record above followed the harm being counted, so keeping count is one of the most useful things the state can do on its own.

The federal fixes are real, and they belong on a longer horizon. There are three. Restore the national hunger count Washington discontinued. Fund the chip-card switch so every state can finish it. And win back the work-rule exemptions the next time the farm bill is open. A farm bill is moving through Congress now, but on the record above, a restoration of what was just cut tends to arrive years later, after the harm is counted, not in the same cycle that made the cut. None of it is within a Buncombe reader's near-term reach. Worth tracking. Not worth building the plan around.