A design for a food system that sustains — one whose operating costs are covered by earned revenue and circulating value, not by the next grant — with the evidence that it actually works, and an honest account of where it strains.
Most community food efforts start and die. They die because their fuel comes from outside the community — a grant pays the coordinator, volunteers do the work, charity is the whole model — and none of it circulates. When the grant ends, the paycheck ends. When the volunteer burns out, the garden goes to weeds. A charity consumes resources. A system circulates and regenerates them.
Grants buy capital, once. Sales buy operations, forever.
Money that keeps the lights on comes from selling to buyers who already pay; grants get spent only on one-time infrastructure that keeps working after the money is gone. Two more rules turn parts into a system: every output is another part's input, and the people whose livelihood depends on it are the ones who own it.
Food hubs are businesses, and most break even. In 2020, 91% of food hubs reported breaking even or better, up from about two-thirds in 2017 and 2019 (MSU 2021 National Food Hub Survey). The best-run quarter turned a ~4% profit, and hubs above $1.5M in sales averaged a positive margin (NSAC summary of the survey). The honest complication: the newest data shows hubs are more grant-reliant even as sales grow — over 50% rely on grant funding to enable institutional purchasing, and they cite uncertain federal funding as a top worry (MSU 2025 National Food Hub Survey). So the realistic goal isn't "zero grants" — it's grants as catalyst and capital, not grants as payroll.
The anchor demand is enormous and already paying. US institutional food service spends over $130 billion a year — 2,000 hospitals, 7,800 schools, 100 colleges are organized around redirecting it to local/sustainable food (Health Care Without Harm, Anchors in Action). Schools alone bought $789 million in local food in 2013–14, a 105% jump in two years (USDA); farm-to-school districts now average ~$114,000/year on local food, about 20% of their purchases (USDA 2019 Farm to School Census).
"Food is Medicine" turns health care into a paying buyer — and Arkansas is in. Medicaid Section 1115 waivers increasingly cover nutrition prescriptions and grocery provisions (up to 3 meals/day for up to 6 months), and the states with recent approvals include Arkansas (KFF); the number of "Food is Medicine" waivers has more than doubled since 2021 (Health Affairs Scholar). It works: a peer-reviewed evaluation of 9 produce-prescription programs, 22 sites, 3,881 people found participants ate about a serving more of fruits and vegetables a day, with lower blood pressure, BMI, and blood sugar, and less food insecurity (PubMed; Tufts summary). (The Rockefeller Foundation projects $45B in state economic activity if it scaled nationally — but that is a model, not a measured result.)
Compost pays twice. Composting operations run on tipping fees plus compost sales, and the finished compost reduces what growers spend on fertilizer and soil (US EPA) — a cost cut and a revenue line from the same waste stream.
Co-op ownership lasts. A rigorous academic review finds worker co-operatives survive at least as well as conventional firms — in France, identical death rates over two decades — and tend to be larger, not the small niche of the stereotype (Pérotin, for Co-operatives UK). A community land trust can hold farmland out of the speculative market permanently — the model the first US CLT, New Communities Inc., used for thousands of acres (CLT case study) — though that same history (a 1985 foreclosure after discriminatory lending) shows land tenure needs legal and political backing, not just a good structure.
Value-added incubators earn their keep. A shared commercial kitchen is itself a business: New Jersey's Hudson Kitchen reports its member founders generated $3.2 million in revenue and 120 jobs in 2024 alone, having supported 60+ businesses since opening (Hudson Kitchen).
1. "Self-sustaining" is a direction, not a switch. The 2025 data is clear: hubs still lean on grants. The win is moving grants from payroll (which dies) to capital and catalyst (which compounds) — not eliminating them. 2. It needs a business operator, not only a mission. Someone has to run the hub like the business it is — buyers, margins, delivery on time. 3. Anchor contracts reward reliability. A hospital or resort won't buy until you can deliver volume and consistency — the chicken-and-egg every hub crosses first, usually with a few committed early buyers. 4. Produce prescriptions need a clinical partner and a payer. The Medicaid pathway exists; a clinic and a reimbursement arrangement still have to be built. 5. This is a 3–5 year build. The flywheel is slow to start and hard to stop once it turns. Checklists light the fire; this is what keeps it burning.
County Commons doesn't run the hub. Its job is three things: put the question to the people who decide, track the loop in public on what came of it — each step dated and cited — and keep a public local-food-dollar score: how many of the county's food dollars stay in the county, moved from reconstructed to earned, year over year.
Every factual claim above links to a primary or clearly-labeled source, with verification caveats stated in the open. This is the systems design; a county's specific engine — who its anchor buyers are, what its numbers are — is built on top of it, on that county's Food as a utility page.