# U.S. Public School Food System: Supply Chains, Procurement, Operations, Nutrition Outcomes, and Local Sourcing ## Executive summary **Scope.** This report examines the U.S. public K–12 school-food system, centered on the National School Lunch Program (NSLP) and School Breakfast Program (SBP). No state, school district, or school-food authority (SFA) was specified, so federal rules are treated as the baseline; state reimbursement supplements, procurement thresholds, universal-meal laws, health codes, labor rules, and local contracting requirements can materially alter the economics and operating model in a particular jurisdiction. Federal regulations explicitly layer school-meal procurement requirements on top of applicable state and local law. citeturn3search3turn20view3 The central finding is that **school food is not primarily a “food purchasing” problem; it is a tightly coupled procurement, logistics, labor, kitchen-capacity, reimbursement, and regulatory system.** A district can buy nutritionally excellent ingredients yet fail operationally if it lacks trained labor, refrigeration, preparation equipment, receiving capacity, delivery frequency, forecasting accuracy, or enough student participation to absorb fixed costs. USDA itself frames purchasing as a process beginning with menu planning and continuing through forecasting, sourcing, bidding, documentation, inventory, preparation, and meal service. citeturn18search6 The system has three major food channels. Commercial products generally move from agricultural producers through processors/manufacturers and foodservice distributors into district warehouses or kitchens. USDA Foods—historically called “commodity foods”—provide an additional federally financed channel tied to NSLP lunches. Local procurement can shorten the chain through farms, cooperatives, food hubs, local processors, and direct-to-school transactions. In practice, successful districts mix these channels rather than choosing one exclusively. citeturn19search5turn18search6 **Procurement is unusually rule-bound.** Purchases made with the nonprofit school-food-service account are governed by NSLP regulations and the Uniform Guidance. Federal rules distinguish informal micro-purchases and simplified acquisitions, formal sealed bids and competitive proposals, and narrowly authorized noncompetitive procurements. Micro-purchases can avoid competitive quotations when prices are documented as reasonable; simplified acquisitions require quotations from an adequate number of qualified sources; formal procurements require public competition. States and districts may impose lower thresholds or additional procedures. citeturn20view3turn3search3 Cooperative purchasing therefore is best understood not as a separate federal procurement method, but as a way to aggregate volume while the underlying solicitation and contract still comply with applicable procurement rules. **The financial picture is structurally tight.** The last completed nationally representative comprehensive cost study, for SY 2014–15, estimated an average reported cost of **$3.81 per lunch and $2.72 per breakfast**, while estimated “full” costs—including costs borne outside the foodservice account—were **$6.02 and $4.19**, respectively. Food and labor each represented about **45%** of reported cost, leaving only about 10% for other costs. citeturn11view0turn12view0turn12view1 Those figures are now old and should not be treated as current prices. USDA's 2024–25 National School Foods Study is designed specifically to refresh national evidence on costs, revenues, purchasing, nutrition, plate waste, and dietary intake; project information current in 2026 still describes the study and its follow-up cost collection rather than published final national estimates. citeturn29search1turn29search9 For SY 2026–27 in the contiguous states and District of Columbia, federal lunch support includes a Section 4 payment of roughly **$0.45–$0.47 per lunch depending on SFA characteristics**, a **$4.31 Section 11 free-lunch payment**, and an additional **$0.09 performance-based amount** for certified meals; the minimum national value of USDA Foods is **$0.32 per NSLP lunch**. Standard free-breakfast reimbursement is **$2.54**, rising to **$3.05** for severe-need schools. These are federal components rather than a statement of actual district cost or financial adequacy. citeturn2view0turn2view2turn16search0 **Nutrition evidence is substantially stronger than a decade ago.** After implementation of the Healthy, Hunger-Free Kids Act standards, national observational evidence found materially better lunch dietary quality among NSLP participants relative to nonparticipants; one JAMA analysis estimated a difference-in-differences improvement of 8.1 Healthy Eating Index points. citeturn24search0 USDA's nationally representative work similarly finds foods consumed from school sources to have substantially higher dietary-quality scores than foods consumed from non-school sources. citeturn7search3 The evidence does **not** support a simple claim that school meals drive childhood obesity. A 2024 systematic review of universal free meals found higher meal participation and possible benefits for attendance, obesity, and disciplinary outcomes, but certainty was moderate only for lunch participation and generally low or very low for the other outcomes. citeturn24search1 Newer quasi-experimental studies are generally reassuring on weight and suggest potential attendance benefits, but they remain an evolving evidence base. citeturn24search3turn7search4 **Farm-to-school has moved from a niche experiment toward a material purchasing channel.** USDA reports approximately **$1.8 billion in local-food purchases in SY 2022–23** and widespread farm-to-school activity. citeturn15view0 Yet “local” does not automatically mean cheaper, healthier, or operationally easier. Local procurement frequently shifts costs away from the ingredient invoice and into staff time, aggregation, delivery, receiving, processing, menu redesign, and supplier development. Michigan's 10 Cents a Meal program is especially revealing: after permitting part of grant awards to cover transportation and labor, grantees reported that those funds expanded their ability to buy and serve local foods, with most of the reimbursed ancillary cost going to labor rather than transportation. citeturn27view1 New York uses a different mechanism—higher per-lunch reimbursement for reaching a 30% New York-food purchasing threshold—which shows how policy can make local sourcing an economically rewarded operating target rather than a voluntary add-on. citeturn25search6 The strongest operating model emerging from the evidence is therefore **hybrid**: use large-scale competitive contracts and USDA Foods for predictable staples; aggregate demand through cooperative arrangements where doing so improves purchasing power; retain sufficient district culinary capacity to prepare and finish foods; use local suppliers selectively where seasonality, volume, processing form, delivery, and price work; and explicitly finance the additional labor and logistics required for scratch cooking and local sourcing. Small and rural SFAs are especially exposed to diseconomies of scale: USDA economic research found significant relationships between purchasing volume and produce prices, including substantially higher fruit-and-vegetable unit costs among low-volume buyers. citeturn7search7 ## Scope and system architecture The institution that matters operationally is usually the **school food authority**, the entity responsible for operating the federal meal programs and maintaining the nonprofit school-food-service account. An SFA may correspond to one district, multiple schools, or another approved organizational structure. It can self-operate foodservice or contract with a food service management company, but outsourcing does not transfer away the SFA's federal responsibilities: the SFA must retain control over meal-service quality and general character, prices, claims, and oversight of the contractor. citeturn19search2 A useful way to understand school food is as **several parallel supply chains converging at the cafeteria**: ```mermaid flowchart LR A["Farms, ranches, fisheries
and other producers"] --> B["Processors / manufacturers
cutting, freezing, baking,
packaging, formulation"] B --> C["Broadline / specialty
foodservice distributors"] C --> D["SFA warehouse,
commissary or school kitchen"] A --> E["Local aggregator,
co-op or food hub"] E --> D A -->|"Direct local purchase"| D F["USDA procurement /
USDA Foods entitlement"] --> G["State distribution,
commercial processors
and distribution channels"] G --> D D --> H["Central production kitchen"] H --> I["Satellite / finishing kitchens"] D --> J["On-site production kitchen"] I --> K["Cafeteria service"] J --> K K --> L["Students"] ``` This diagram is an analytical synthesis of the purchasing and distribution paths recognized in USDA school-food guidance and program regulations. USDA describes purchasing as extending from menu planning and forecasting through sourcing, bidding, USDA Foods, local purchasing, documentation, equipment, and final meal service. citeturn18search6turn19search5 The **commercial path** is usually optimized for standardization and scale. A district may specify thousands of cases of a standardized product, receive bids from large distributors, and schedule regular deliveries. The product itself may contain ingredients from multiple agricultural producers and may have undergone substantial processing before it reaches the district. This model lowers transaction cost per unit and can improve year-round availability, but it can also reduce transparency into origin and make smaller suppliers unable to meet case-pack, insurance, delivery, electronic-ordering, and volume requirements. Those tradeoffs are one reason USDA encourages procurement planning before specifications and solicitations are finalized. citeturn18search6turn20view3 The **USDA Foods path** partially substitutes federal commodity purchasing for district cash purchasing. Schools participating in NSLP receive an entitlement value tied to lunches served; for SY 2026–27 the statutory national-average minimum value is $0.32 per lunch. USDA Foods can arrive in forms suitable for direct use or be incorporated into processed end products under the USDA Foods processing framework. citeturn16search0turn28search0 Economically, the important point is that the entitlement reduces a district's need to buy an equivalent amount of food with cash, but it does not eliminate distribution, storage, processing, handling, or kitchen labor. The **local path** can be shorter physically yet more complicated institutionally. A farm might sell directly to a district, but many districts need intermediaries that consolidate products from multiple farms, provide cold storage, carry insurance, perform minimal processing, and make school-compatible deliveries. Michigan's creation of a searchable Farm to Program supplier directory illustrates the information problem: districts need to know not only who grows a product but its availability window, delivery service, processing capacity, and food-safety characteristics. citeturn26view0 The governing policy environment has changed substantially since the Healthy, Hunger-Free Kids Act: ```mermaid timeline title Major policy changes shaping today's school-food system 2010 : Healthy, Hunger-Free Kids Act enacted : Stronger school-meal nutrition and program standards 2012 : Updated NSLP meal standards implemented nationally 2020-2022 : Pandemic-era nationwide flexibilities greatly expanded free meal access : Supply-chain and operating waivers changed normal service models 2023 : Community Eligibility Provision minimum identified-student threshold lowered to 25% 2024 : USDA final rule updated meal patterns toward 2020-2025 Dietary Guidelines : Geographic-preference and Buy American provisions updated 2026 : SY 2026-27 reimbursement rates in effect 2027 : Added sugars limit and next sodium reductions take effect ``` The 2010 law and subsequent 2012 implementation were associated with a measurable improvement in school-lunch dietary quality. citeturn24search0 The 2023 CEP rule broadened the set of schools able to elect community eligibility by lowering the minimum identified-student percentage. citeturn21search1 The 2024 rule set the current trajectory for added sugars, sodium, whole grains, milk, local procurement, and Buy American implementation. citeturn22search9turn21search2 ## Procurement, financing, and regulation A district does not simply ask, “Which vendor has the lowest price?” It first has to determine the legally permissible procurement method, write specifications that are not improperly restrictive, estimate quantities, decide contract duration and renewal structure, establish evaluation criteria, address domestic-purchasing requirements, plan for substitutions and shortages, and document the transaction. USDA's own purchasing guide consequently places forecasting and menu planning before bidding. citeturn18search6 ### Procurement models compared | Procurement model | How it works | Primary economic advantage | Main constraints and risks | |---|---|---|---| | **Micro-purchase / direct purchase** | Purchase below the SFA's applicable micro-purchase threshold. Competitive quotations are not required when the price is documented as reasonable; purchases should be distributed equitably among qualified suppliers to the extent practicable. citeturn20view3 | Very low transaction cost; useful for small local buys, emergency fill-ins, or testing a supplier. | Threshold is governed by federal plus state/local rules. Splitting transactions to evade competition is inappropriate; price-reasonableness documentation still matters. | | **Simplified acquisition / small purchase** | Above the micro threshold but at or below the entity's simplified-acquisition threshold; quotations must be obtained from an adequate number of qualified sources. citeturn20view3 | Lighter process than formal bidding while preserving price competition. | Supplier market may be thin in rural areas; state/local thresholds can be lower than federal ceilings. | | **Invitation for bids / sealed bid** | Public solicitation with adequate specifications; award generally goes to the lowest responsive bid from a responsible bidder. citeturn20view3 | Strong price transparency; works well for high-volume standardized products. | Specifications must be sufficiently precise; a price-first structure can make it harder to value service, flexibility, local attributes, or innovation unless those are built legitimately into requirements. | | **Request for proposals / competitive proposals** | Publicly solicited proposals evaluated using disclosed factors and relative importance; award goes to the responsible offeror providing the most advantageous proposal considering price and other factors. citeturn20view3 | Better suited to complex services and situations where logistics, menu capability, technology, or service quality matter alongside price. | More evaluation effort and documentation; criteria must be objective and applied as advertised. | | **Cooperative / consortium purchasing** | Multiple public entities aggregate demand and use a shared solicitation or contract. Under federal rules, this is an organizational arrangement rather than a distinct procurement category; the underlying purchase still must satisfy applicable competition and contracting rules. citeturn20view3turn3search3 | Larger volume, fewer solicitations, potentially lower distributor margins and better service coverage. | Can favor large suppliers, reduce local flexibility, or create contract-access problems if the cooperative procurement was not structured for participating entities' legal requirements. | | **Food service management company** | SFA competitively contracts with a private firm to manage some or all foodservice operations. The SFA must retain program control and oversight. Cost-plus-a-percentage-of-cost and cost-plus-a-percentage-of-income arrangements are prohibited. citeturn19search2 | Access to purchasing scale, management systems, labor expertise, and standardized operating procedures. | Management fees and contract incentives can alter behavior; the district remains accountable for federal compliance and must actively administer the contract. | | **USDA Foods** | Federal food assistance tied to NSLP lunch participation supplements commercial procurement; products may move through state distribution and approved processing channels. citeturn16search0turn28search0 | Converts part of federal support into food rather than cash and can lower cash food cost. | Menu usefulness, storage, delivery timing, processing charges, inventory planning, and entitlement management determine the real value captured by the district. | | **Local / farm-to-school procurement** | Uses the same lawful procurement methods but sources geographically proximate products; USDA permits geographic-preference approaches within federal rules, subject to state/local law. citeturn18search3turn23search6 | Fresher seasonal products, supplier diversification, agricultural-market development, educational linkage. | Seasonality, small supplier volumes, delivery frequency, processing, labor, food-safety documentation, and transaction costs can offset ingredient-price benefits. | | **Noncompetitive procurement** | Allowed only in specified circumstances such as sole source, qualifying emergency/public exigency, written authorization, or inadequate competition after solicitation. citeturn20view3 | Enables continuity when true competition is unavailable. | High audit/compliance risk if used merely for convenience; requires strong documentation. | Federal school-meal regulations specifically require procurement conducted with the nonprofit school-food-service account to comply with the relevant federal procurement standards and applicable state and local requirements. SFAs must maintain written procurement procedures, manage contracts, and follow conflict-of-interest/code-of-conduct requirements. citeturn3search3turn20view3 This is why the correct procurement method in one district cannot be assumed to apply unchanged in another. The **Buy American provision** adds another constraint: SFAs operating NSLP/SBP generally must purchase domestic commodities and products to the maximum extent practicable, using documented exceptions consistent with program rules. citeturn23search9turn22search9 This can interact with seasonal produce availability, processed-product formulations, distributor substitutions, and price shocks. ### How the system is financed The school-meal enterprise is usually designed as a nonprofit foodservice operation whose recurring economics depend on a combination of federal reimbursement, USDA Foods, student payments for paid meals where applicable, state/local subsidies, nonreimbursable or à-la-carte sales, and occasionally grants or district support. The exact mix has changed substantially in jurisdictions that have adopted universal free meals. citeturn19search5turn2view0 CEP changes the revenue model in eligible schools by allowing breakfast and lunch to be offered at no charge to every student without individual household applications. USDA's 2023 final rule reduced the minimum identified-student threshold to 25%, making more schools eligible to consider the model. Whether CEP is financially attractive still depends on each school's directly certified population, meal participation, state supplements, and operating costs. citeturn21search1 The strongest nationally representative benchmark remains USDA's SY 2014–15 School Nutrition and Meal Cost Study, but it must be treated as a **structural benchmark, not a current price list**. ### Cost comparison | Cost or funding benchmark | Lunch | Breakfast | Interpretation | |---|---:|---:|---| | **SY 2014–15 reported cost, average SFA** | **$3.81** | **$2.72** | Costs recorded by the foodservice operation. Nationally representative SNMCS estimate. citeturn11view0turn12view0 | | **SY 2014–15 full cost, average SFA** | **$6.02** | **$4.19** | Adds costs supporting meals that were borne outside the foodservice account; shows why accounting cost can exceed cafeteria-account expenditure. citeturn11view0turn12view0 | | **SY 2014–15 meal-weighted reported cost** | **$3.66** | **$2.34** | Gives greater weight to larger-volume SFAs, illustrating scale effects. citeturn11view0 | | **SY 2014–15 meal-weighted full cost** | **$5.55** | **$3.50** | Again lower than the unweighted SFA average, consistent with economies of scale. citeturn11view0 | | **SY 2026–27 Section 11 free-meal reimbursement, contiguous U.S.** | **$4.31** | — | Lunch component before the separate Section 4 and performance-based payments. citeturn2view0turn2view2 | | **SY 2026–27 Section 4 lunch component** | **about $0.45–$0.47** | — | Rate varies with SFA characteristics; qualifying certified lunches also receive an additional $0.09 performance-based amount. citeturn2view0turn2view2 | | **SY 2026–27 free SBP reimbursement** | — | **$2.54 standard; $3.05 severe need** | Federal cash reimbursement; severe-need status materially changes breakfast economics. citeturn2view0 | | **SY 2026–27 USDA Foods minimum value** | **$0.32 per NSLP lunch** | — | In-kind food entitlement value, not cash and not necessarily equivalent to the district's realized net savings after handling/processing. citeturn16search0 | The 2014–15 and 2026–27 figures should **not** be compared as though they were prices from the same year. Inflation, wages, food prices, standards, participation, universal-meal policies, supply-chain conditions, and operating practices have changed. The value of putting them side by side is institutional: it shows that federal reimbursement is a per-meal revenue stream, while actual cost includes labor, food, utilities, management, equipment, logistics, and support services. USDA has commissioned the 2024–25 National School Foods Study precisely because the old national cost baseline needs updating. citeturn29search1turn29search9 The old cost structure nevertheless reveals an important operational fact: ```mermaid pie showData title Reported school foodservice cost structure, SY 2014-15 "Food, including USDA Foods" : 45 "Labor" : 45 "Other" : 10 ``` In SNMCS, food and labor each accounted for about 45% of reported costs and other costs roughly 10%. citeturn12view1 The practical implication is substantial: a program that focuses only on getting cheaper ingredients is targeting roughly half the controllable cost base. An ingredient that is inexpensive per pound but requires significantly more trimming, batch preparation, transport, washing, or service labor may not lower total meal cost. Scale compounds the problem. USDA Economic Research Service analysis found that produce purchasing volume was strongly related to prices: SFAs buying about one-quarter of average fruit-and-vegetable volume had costs about 16% higher than SFAs buying roughly four times average volume. The same analysis found regional and size-related cost disadvantages for small SFAs. citeturn7search7 This helps explain why cooperative purchasing, distributor aggregation, shared central kitchens, and food hubs can be economically important even when districts want more local and minimally processed food. ## Food preparation and cafeteria operations There is no single “school kitchen.” A useful operational classification is **on-site production, central/commissary production, satellite/finishing kitchens, hybrids, and externally managed or vended service**. USDA purchasing guidance explicitly recognizes that operators range from single-site programs to large districts in which purchasing and operational tasks are divided among many staff. citeturn18search6 An **on-site production kitchen** receives ingredients and prepares most meals at the school where they are served. Its advantage is menu flexibility and the ability to prepare closer to service time. Its disadvantages are duplicated equipment, management, skilled labor, storage, and food-safety controls across many buildings. A small elementary school may therefore have higher unit costs than a district's large production kitchen even when both serve the same menu; national cost evidence is consistent with these economies-of-scale effects. citeturn7search7turn11view0 A **central kitchen/commissary** concentrates production in one facility and transports finished or partially finished foods to satellite schools. This can justify higher-capacity combi ovens, kettles, tilt skillets, blast chilling, automated packaging, warehouse space, and specialized culinary labor. The tradeoff is a more demanding logistics system: production scheduling, route design, time-and-temperature control, holding quality, packaging, delivery vehicles, receiving space, and school-level reheating or finishing become critical. These are analytic consequences of centralization rather than separate federal program rules; federal food-safety and meal-standard requirements apply regardless of production architecture. citeturn18search0turn19search5 A **hybrid model** often makes the most operational sense. High-volume, labor-intensive preparation can be centralized, while salads, fruit, final baking, reheating, seasoning, or assembly occur at school sites. This preserves some economies of scale without turning every meal into a fully transported finished product. Contracting with a **food service management company** changes who performs many tasks but does not eliminate district accountability. Federal rules require the SFA to retain control over the quality and general nature of foodservice, monitor operations, retain key signature and claim authority, and procure the management contract competitively. citeturn19search2 ### Labor, equipment, and menu planning School foodservice depends on a combination of directors, procurement/business staff, dietitians or menu planners where available, managers, cooks, prep workers, servers/cashiers, warehouse staff, drivers, and maintenance/support functions. The federal professional-standards framework requires school nutrition personnel to meet role-appropriate training and professional requirements; USDA operates a dedicated training and tracking system for those requirements. citeturn18search1 Labor capacity has become particularly important as policy pushes menus toward lower sodium, lower added sugar, greater whole-grain content, fruits, vegetables, and scratch or speed-scratch preparation. Michigan's local-food experience provides a concrete example: when transportation and labor became eligible grant expenses, nearly $136,000 of roughly $160,000 reported ancillary expenses was labor for functions including food preparation, menu work, delivery staff time, and procurement—far more than the reported transportation amount. citeturn27view1 Equipment therefore acts as a constraint on nutrition policy. A district cannot easily convert from individually packaged heat-and-serve entrées to fresh vegetables and scratch-cooked proteins without adequate receiving space, refrigeration, sinks, prep surfaces, knives and food processors, ovens/steamers, kettles or skillets, hot/cold holding, dishwashing capacity, and ventilation. USDA's Food Buying Guide explicitly places equipment requirements alongside forecasting, sourcing, USDA Foods, local purchasing, and menu documentation. citeturn18search6 Menu planning is simultaneously a **nutrition task, demand-forecasting task, and purchasing specification**. The menu determines which ingredients are needed, in what creditable quantities, at what frequency, with which package sizes and processing forms. USDA guidance therefore starts the purchasing cycle with the menu rather than with a vendor catalog. citeturn18search6 As of SY 2026–27, NSLP lunches operate within calorie ranges of **550–650 calories for grades K–5, 600–700 for grades 6–8, and 750–850 for grades 9–12**, with required weekly quantities of fruits, vegetables and vegetable subgroups, grains, meats/meat alternates, and milk. At least 80% of grains offered weekly must be whole-grain rich. Saturated fat must average below 10% of calories. citeturn21search8 The current standards are also in the middle of a planned transition. Through June 30, 2027, NSLP lunch sodium limits are 1,110 mg for K–5, 1,225 mg for 6–8, and 1,280 mg for 9–12. Beginning July 1, 2027, those limits fall to 935, 1,035, and 1,080 mg, respectively, and an added-sugars limit of less than 10% of calories takes effect. Breakfast sodium limits similarly tighten in 2027, while breakfast continues to require one cup of fruit daily and five cups per week. citeturn21search2turn21search8 Those standards influence the entire upstream market. A manufacturer selling pizza, breakfast pastries, sauces, bread, deli meats, or entrées to schools must formulate products that let a district build compliant menus. A district switching to a lower-sodium specification may therefore affect supplier formulation and bid competition rather than merely changing a recipe at the cafeteria. This is an inference from how menu requirements and procurement specifications interact. citeturn18search6turn21search8 ### Food safety Participating schools must operate food-safety systems based on hazard-analysis and critical-control-point principles, and federal school-meal rules require food-safety inspections. USDA's implementation guidance specifies a school food-safety program designed to prevent or minimize foodborne-illness risk and the requirement for two food-safety inspections per school year. citeturn18search0turn18search4 Food safety can complicate local procurement because school buyers may need documentation concerning supplier handling, refrigeration, washing/processing, delivery temperature, traceability, insurance, and recall capability. Michigan's Farm to Program directory explicitly lets institutional buyers state food-safety requirements and identify suppliers offering delivery or minimal processing—an example of local food infrastructure evolving toward institutional requirements rather than expecting schools to buy in a farmers'-market model. citeturn26view0 ## Nutrition, health, and academic outcomes The most defensible conclusion from the research is **not simply that “school lunch is healthy,” but that meals produced under the post-2012 federal standards are generally nutritionally superior to the alternative foods children consume outside the school-meal program, while evidence on downstream health and academic outcomes varies in causal strength**. A JAMA study using national NHANES data compared periods before and after implementation of the Healthy, Hunger-Free Kids Act. Among students presumed to participate in NSLP, adjusted lunch Healthy Eating Index scores improved substantially; the estimated change relative to nonparticipants was **8.1 HEI points**. Because the study used serial cross-sectional data and estimated participation rather than randomly assigning meals, it supports a strong policy association rather than proving individual-level causality. citeturn24search0 USDA analysis of the nationally representative School Nutrition and Meal Cost Study reached a complementary conclusion. Foods obtained from school sources had substantially higher overall dietary-quality scores than foods consumed from nonschool sources, including among students experiencing food insecurity, with especially strong performance for fruits, dairy, and whole grains. citeturn7search3 These findings matter because the relevant counterfactual is often not an ideal home-cooked meal; it is whatever a child would actually eat if the school meal were unavailable. The evidence also undercuts a persistent concern that improved school-meal standards necessarily cost more. In the 2014–15 national study, schools in the highest lunch nutrition-quality quartile did not have statistically significantly higher reported lunch costs than schools in the lowest quartile; estimated average costs were about $3.90 versus $3.85. citeturn12view2 That does not mean every healthier ingredient is cost-neutral. Rather, it suggests that overall menu quality can be improved through product mix, procurement, portioning, and menu design without a mechanically higher total foodservice cost. ### Universal free meals The evidence for universal free school meals is strongest for **participation**. A 2024 systematic review in *JAMA Network Open* concluded that universal free meal programs were associated with increased school-meal participation. Evidence for attendance was more modest, and findings suggesting lower obesity prevalence and fewer suspensions came from a limited number of studies with low or very low certainty. The review explicitly identified diet quality and food security as important outcomes for which evidence remained insufficient. citeturn24search1 More recent studies strengthen, but do not eliminate, that uncertainty. A 2025 study linking electronic health records for 149,052 lower-income children across 1,085 schools in 12 states found that children in schools adopting the Community Eligibility Provision had a small relative reduction in BMI z-score compared with children in eligible schools that did not adopt, though results varied in some specifications. citeturn24search3 This is inconsistent with the simple hypothesis that universal access causes widespread excess weight gain. A newer large quasi-experimental analysis comparing states with and without state-level universal free meal policies reported an estimated **3-percentage-point reduction in chronic absenteeism**, with larger estimated effects in schools serving higher shares of Black students, as well as an improvement in English-language proficiency; effects on some other academic outcomes were not statistically significant. citeturn7search4 Because state universal-meal laws are not randomized, such estimates should still be interpreted as quasi-experimental rather than experimental evidence. The academic literature is similarly nuanced. A 2022 review found mixed evidence that breakfast directly improves standardized test scores or grades but more consistent indications that easier access to free breakfast may improve attendance. citeturn24search6 A 2026 review covering 33 U.S. studies found that most examined school-nutrition policies had positive or neutral academic associations and concluded that universal free school meals and breakfast-after-the-bell approaches appear promising for improving performance and equity; it also called for additional research. citeturn24search2 Taken together, the evidence supports four conclusions. First, school meals under contemporary standards can materially raise diet quality. Second, eliminating price barriers reliably increases participation. Third, there is no strong evidence that universal free school meals worsen obesity at the population level; several newer studies point in the opposite direction. Fourth, academic effects are plausible and increasingly supported for attendance and some subgroup outcomes, but claims that meals directly raise test scores across all settings would overstate the evidence. citeturn24search0turn24search1turn24search2turn24search3 A remaining measurement issue is **offering versus consumption**. Federal standards specify what schools must offer and what constitutes a reimbursable meal; they do not guarantee that students eat every component. Plate waste, taste, lunch duration, queue time, recess scheduling, cultural fit, and presentation can therefore mediate the health effect of a technically compliant menu. The new National School Foods Study explicitly includes plate waste, student satisfaction, and dietary intake because menu compliance alone cannot answer the consumption question. citeturn29search1 ## Local sourcing and farm-to-school experiments Farm-to-school is best viewed as a portfolio of interventions rather than a single procurement model. It can include direct farm purchases, food-hub aggregation, local specifications or geographic preferences, state reimbursement incentives, grants for incremental food cost, grants for equipment and infrastructure, farmer-school matchmaking, taste tests, school gardens, and curriculum. USDA's Farm to School guidance explicitly integrates local procurement with complementary agricultural and nutrition education. citeturn18search6turn28search1 Nationally, the sector has become economically significant. USDA's analysis of the 2023 Farm to School Census reports approximately **$1.8 billion in local-food purchasing in SY 2022–23** and broad farm-to-school participation. citeturn15view0 The scale indicates that local procurement is no longer merely a collection of demonstration farms and salad-bar pilots. ### Case-study comparison | Case | Procurement / financing mechanism | Evidence and results | Analytical lesson | |---|---|---|---| | **New York 30% NYS Initiative** | An SFA qualifying with at least 30% eligible New York food purchases receives **25¢ per lunch in state reimbursement rather than 5.9¢**. The policy rewards an achieved purchasing share rather than reimbursing each individual local invoice. citeturn25search2turn25search6 | As of 2025–26, New York reported **82 participating school districts**, more than **$40.4 million in NY food purchases**, more than **$11.1 million in associated reimbursement**, and more than 750,000 K–12 students reached since 2021. citeturn25search6 | A predictable per-meal revenue incentive can make local procurement part of the SFA's recurring business model. It also creates a strong need for origin verification and purchasing records. | | **Michigan 10 Cents a Meal** | Grants support Michigan-grown foods; in 2024–25 grantees could use up to **25% of the grant for eligible labor and transportation**, explicitly recognizing costs beyond the ingredients. citeturn27view1 | Foodservice managers reported more Michigan-grown fruit and vegetables and self-reported increases in student fruit consumption (84%) and vegetable consumption (76%). These are program-manager reports, not independently measured causal dietary outcomes. citeturn27view0 | Local sourcing is often constrained by labor and logistics, not simply a local-food price premium. Financing preparation and distribution may be as important as subsidizing the commodity. | | **Michigan transportation/labor experiment** | The 2024–25 program explicitly reimbursed food preparation, menu work, procurement activity, staff delivery time, supplier delivery fees, mileage, and related expenses. citeturn27view1 | Seventy-one grantees reported using the funds; 56 said the support expanded their ability to purchase or serve local food. Nearly $136,000 of reported eligible expenditure was labor versus about $21,000 transportation. citeturn27view1 | A useful real-world indication that the hidden cost of local food can reside primarily in staff time rather than trucking. | | **Alaska Farm to School Census profile** | SFAs use multiple local purchasing channels under national farm-to-school rules. citeturn25search7 | USDA reports 74.5% of responding Alaska SFAs participating in farm-to-school, yet only 10.9% reported serving local fruit at least weekly, 21.8% local vegetables, and 3.6% local protein. citeturn25search7 | High program participation need not imply high-frequency local-food service. The pattern is consistent with major supply, seasonality, processing, and geography constraints, although the census does not by itself identify which constraint is causal. | | **National farm-to-school market** | Mix of ordinary procurement, geographic preference, direct purchasing, food hubs, state incentives, and USDA grant support. citeturn23search6turn28search8 | Local purchases reached about $1.8 billion in SY 2022–23. citeturn15view0 | Local sourcing has reached sufficient scale that aggregation, distributor participation, processing, cold-chain infrastructure, and contract design matter more than one-off farm relationships alone. | New York and Michigan illustrate two fundamentally different incentive designs. **New York pays for the outcome**—achieving a local purchasing share. **Michigan helps pay for the inputs and capabilities** required to buy locally. Neither is universally superior. An outcome incentive gives districts freedom to optimize procurement but may disproportionately reward districts that already possess storage, supplier networks, staff, and local agricultural density. A capability grant can help weaker systems build those prerequisites but creates more administrative complexity and may not guarantee sustained local purchasing once grants expire. The comparison is an analytical inference from the two program designs. citeturn25search6turn27view1 ### Seasonality and product form Seasonality is one of the most persistent structural mismatches in farm-to-school. Schools need high volumes on predictable instructional calendars; many fresh crops have short harvest windows and variable annual yields. Michigan reported that food-program managers continued to identify desired Michigan-grown foods that they could not find or buy, and administrators consequently developed a supplier directory containing information about product availability, minimal processing, and delivery services. citeturn26view0 The most scalable solution is often not “farm to cafeteria in 24 hours.” It may be **farm to processor to freezer to school**, or **farm to food hub to distributor to school**. Freezing, chopping, washing, canning, pureeing, or otherwise minimally processing a local crop can extend the school-use season and reduce cafeteria labor. The implication is that regional processing capacity can be as important to farm-to-school growth as the number of nearby farms. This is an inference from the logistical constraints documented in USDA guidance and Michigan's supplier-development experience. citeturn18search6turn26view0 ### What can be said about cost-benefit The clearest benefit measures are **purchasing volume, market access, menu variety, participation, and student exposure to foods**. New York documents tens of millions of dollars of food purchases redirected toward in-state products. Michigan grantees report broader produce variety, new farmer connections, and increased ability to serve local food. citeturn25search6turn26view0 The cost side is harder to measure. Local food may have a competitive farm-gate price while still being more expensive **delivered and ready to serve** because of smaller drops, extra invoices, irregular pack sizes, trimming, washing, labor, spoilage, storage, and staff time spent locating suppliers. Michigan's explicit reimbursement of procurement, preparation, delivery, and labor expenses is unusually useful evidence that these costs are economically meaningful. citeturn27view1 Consequently, claims that farm-to-school either “saves money” or is “necessarily more expensive” are too broad. The right metric is **total usable cost per reimbursable serving**, including edible yield, preparation labor, freight, spoilage, storage, equipment, and administrative effort—not simply the invoice price per pound. USDA's Food Buying Guide is built around this yield-and-quantity principle. citeturn18search6 ## Barriers, best practices, and research gaps The principal barriers are tightly interconnected. **Scale and supplier structure are the first barrier.** Small SFAs buy fewer cases, receive less favorable distributor economics, and spread management and equipment costs over fewer meals. USDA ERS evidence demonstrates measurable purchasing-volume effects for produce and meaningful regional cost differences. citeturn7search7 Cooperative purchasing can counter those diseconomies, but very large aggregated specifications can simultaneously make it harder for small local suppliers to compete. The best design is therefore often to aggregate genuinely commoditized items while retaining appropriately sized lots or channels for foods where supplier diversity and locality have value. **Labor is a second binding constraint.** Scratch cooking, local produce processing, recipe standardization, procurement paperwork, receiving, and food-safety controls all consume staff time. Michigan's experience is strong practical evidence: when local-food grants became usable for ancillary costs, labor accounted for the overwhelming majority of the reported transportation-and-labor expenditure. citeturn27view1 A district seeking fresher food without budgeting for additional culinary and procurement capacity can therefore create an unfunded operating mandate. **Equipment and facilities are the third constraint.** A school cannot meaningfully increase raw-product handling when the kitchen was designed primarily to heat pre-portioned food. USDA explicitly treats equipment requirements as part of menu and procurement planning. citeturn18search6 Capital investment should consequently be evaluated together with menu design: a combi oven, additional refrigeration, food processor, tilt skillet, or central produce-prep line can change the feasible supplier set and labor minutes per meal. **Procurement expertise is a fourth constraint.** Federal procurement rules allow considerably more flexibility than is sometimes assumed—micro-purchases, simplified quotations, competitive proposals, geographic preference, and carefully justified noncompetitive procurement all exist—but each carries documentation requirements. citeturn20view3turn18search3 Overly restrictive local interpretations can inadvertently eliminate small suppliers; under-documented flexibility can create audit risk. The best practice is to involve foodservice, procurement, finance, and legal staff before issuing a solicitation rather than trying to repair a poorly designed bid afterward. **Demand uncertainty is the fifth constraint.** School-food economics improve when participation is high because labor, management, equipment, and facilities contain large fixed components. Universal free-meal policies reliably increase participation, according to the 2024 systematic review. citeturn24search1 From an operating perspective, higher and more predictable participation can improve production forecasting and spread fixed cost across more meals, although actual savings depend on staffing, reimbursement, and marginal food cost. **Student acceptance remains a sixth constraint.** A meal is nutritionally valuable only to the extent it is eaten. Taste tests, menu marketing, culturally appropriate recipes, adequate time to eat, and iterative plate-waste measurement should therefore be treated as core operations rather than cosmetic communications. USDA's current National School Foods Study includes satisfaction and plate-waste measurement alongside cost and nutrition precisely because offered food and consumed food are not equivalent. citeturn29search1 The strongest evidence-supported operating practices can be summarized as a sequence rather than a checklist: **design the menu first; forecast demand from participation data; translate recipes into standardized specifications and edible yields; segment products by the procurement method and supplier market that fit them; aggregate high-volume commodities where scale produces savings; preserve smaller or flexible channels for local and specialty foods; evaluate total usable cost rather than invoice price; align kitchen equipment and labor with the intended preparation model; and measure participation, plate waste, food/labor cost, supplier reliability, and nutrition outcomes together.** That logic closely follows USDA's purchasing framework and the cost structure observed in national research. citeturn18search6turn12view1 Several important research gaps remain. First, **current national cost data are the largest gap**. The most comprehensive published national cost benchmark remains SY 2014–15, while the 2024–25 National School Foods Study was designed to generate updated nationally representative estimates of purchasing, costs, revenues, nutrition, student satisfaction, plate waste, and dietary intake. Materials current in 2026 still describe ongoing/follow-up study activities and eventual dissemination. citeturn29search1turn29search4turn29search9 Until those results are published, precise national claims about today's food-versus-labor shares or present-day full meal cost should be treated cautiously. Second, **farm-to-school evaluation is ahead on implementation data and behind on causal cost-benefit evidence**. Programs can report local dollars spent, suppliers reached, foods offered, and managers' perceptions, but robust studies that compare total delivered cost, labor minutes, plate waste, student intake, and farmer income against well-matched nonlocal procurement are uncommon. Michigan's reported consumption improvements, for example, come from food-program managers rather than direct dietary measurements. citeturn27view0 Third, **the causal pathway from universal meals to long-term health remains underdeveloped**. The 2024 systematic review found important evidence gaps for diet quality and food security, even while finding strong participation effects. citeturn24search1 Newer BMI studies are encouraging but still leave questions about heterogeneity by age, poverty, baseline food insecurity, school food quality, and duration of exposure. citeturn24search3 Fourth, **academic-outcome research is improving but still heterogeneous**. Attendance is a more consistent mechanism than standardized-test gains, and newer reviews find mostly positive or neutral academic associations rather than uniformly large effects. citeturn24search2turn24search6 Future evaluations should separate the effects of price elimination, breakfast timing, meal quality, reduced stigma, household food-budget relief, and higher participation instead of treating “school meals” as one undifferentiated treatment. Fifth, **kitchen architecture deserves more rigorous economic research**. There is surprisingly little recent nationally representative causal evidence comparing on-site kitchens, central commissaries, hybrid systems, FSMC-managed systems, and minimally equipped heat-and-serve models on total cost, food quality, labor productivity, plate waste, resilience, and student acceptance. USDA's new study should improve the descriptive base, but the operational causal question remains important. citeturn29search1 Sixth, **supply-chain resilience is insufficiently measured at the product and contract level**. Districts would benefit from research linking bid structure, number of qualified suppliers, fulfillment rates, substitutions, delivery failures, emergency purchases, origin, and final menu changes. The experience of pandemic disruptions and local sourcing suggests that supplier diversification may have resilience value that is not captured by lowest-bid price alone. USDA's purchasing framework recognizes forecasting and sourcing as linked functions, but comparative resilience metrics are still limited. citeturn18search6 The broad policy implication is that school-food performance should be assessed as a **system outcome**. Nutrition standards determine the nutritional floor; reimbursement and participation determine the revenue base; procurement design determines supplier access and price discovery; purchasing scale affects unit prices; kitchen architecture and labor determine what can actually be prepared; logistics determine reliability; and student acceptance determines whether compliant food becomes consumed nutrition. The highest-performing strategy is therefore unlikely to be maximally centralized or maximally local, maximally processed or universally scratch-cooked. The evidence points instead toward a **deliberately engineered hybrid system that captures scale where scale is valuable, preserves flexibility where freshness and supplier diversity matter, and finances the labor and infrastructure needed to turn ingredients into meals students actually eat.** citeturn7search7turn18search6turn24search0turn27view1