What Cariacica agreed to buy

Mozzarella, lettuce, carrots and cassava appear in a school-food contract signed in Cariacica in March 2026.

The list connects two everyday needs: schools need food, and farms need buyers. Making that connection last requires deliveries, invoices, payment and meals children can eat. Cariacica's public records allow several of those steps to be followed.

The records are unusually useful because they extend beyond the contract itself. A reader can examine amounts committed, claims recognised and money paid to named organisations.

Contract 042/2026 was signed on 3 March with APRODER, the Associação dos Produtores Rurais de Roda D'Água e Região, for R$1,238,598.55, about €210,000. Alongside mozzarella, lettuce, carrots and cassava, the list includes plantain, beetroot, scallions, chayote, kale, cauliflower, yam, oranges, fruit pulp, yoghurt, butter, cornmeal, organic bananas and zucchini. Plantain is a starchy cooking banana; cassava is a starchy root.

Chayote is a pale-green relative of squash and melon, often shaped like a pear. Usually cooked as a vegetable, it has crisp flesh and a mild, faintly sweet flavour. The contract's detail gives the purchasing programme a more tangible form than its total value alone.

That total is spread across several budget commitments. Different school categories or funding lines can appear under the same contract. To avoid counting the same obligation twice, the records need to be followed line by line.

Follow the money further than the signature

Three Portuguese accounting terms help explain the trail. An empenho reserves money for an obligation. A liquidação recognises an amount due after the claim and documents have been checked. Payment records the transfer that follows.

On 3 September, one APRODER line showed R$277,817.65, about €47,100, committed, liquidated and paid in full. A separate record under the same school-food call shows a payment of R$102,285.10, about €17,400, on 31 August. Its linked accounting record identifies invoice NF 000.000.606.

Contract 041/2026 concerns APROVISTA, the Associação de Produtores Rurais de Boa Vista. In the 3 September snapshot, one line had R$196,072.26, about €33,300, committed and R$182,981.08, about €31,000, liquidated and paid. Payment followed on 13 July, three days after the accounting check citing invoice 000.000.199.

Update, 13 September: another payment completes one line

The register checked for the 13 September update records a further R$13,091.18 paid on 4 September as payment 0964/2026. Added to the July payment, it completes the R$196,072.26 committed under line 0955/2026. That line is fully liquidated and paid; it is not a total for all APROVISTA business or the whole programme. APROVISTA payment register

The remaining balance on this line has therefore been paid. How the association distributed the money and what schools received still require other evidence. Comparisons below retain the original 3 September date.

The trail establishes that these purchases moved beyond a signed agreement. Named organisations received municipal money.

What that payment meant for farmers and children lies further along the story.

Supplier records show different stages

The portal also preserves differences between suppliers. At the same point in time, their records can show different stages of work.

On 3 September, most of the APROVISTA line had been paid. ASCOPRUM, the Associação Comunitária de Produtores Rurais de Munguba, had R$77,479.13, about €13,100, committed on one line, of which R$58,069.46, about €9,900, was liquidated and paid. A CAFC ES cooperative line showed R$35,815.58, about €6,100, committed, with nothing yet liquidated or paid.

The CAFC ES entry does not explain the timing. Delivery schedules, documents or other circumstances could account for its position. The evidence supports a difference in recorded stage, without establishing a failure by the supplier.

What reaches the farmer after the association is paid?

After payment reaches an association, the public ledger tells us less.

Brazil's education-development fund, FNDE, says acquisition prices should reflect local supply conditions and the costs required to deliver the food, including transport and packaging. Those costs affect what a sale is worth to a producer.

The amount on a contract is consequently different from the income a farmer keeps.

To understand that income, we would need the association's distribution of payments and the costs of transport, administration and lost or rejected products. We would also need to know whether smaller farms can take part on workable terms.

A municipal report of 2 February 2026 adds detail about transport. It says CAFC received a box truck and APRODER a dump truck. CAFC president Davi Dutra described dependence on outside transport firms. APRODER representative Thiago Degasperi expected its vehicle to carry soil-preparation inputs and reduce costs. These are attributed accounts in the municipality's report, not newsroom interviews or measured savings. municipal report dated 2 February 2026

Research elsewhere in Espírito Santo helps identify what to ask. A 2025 study across 20 municipalities found purchasing and sales difficulties, problems with public calls and limited technical capacity. Qualitative research in the state also describes how associations can open markets while facing demanding logistics, planning and administration.

Those findings guide the inquiry. Establishing whether the same difficulties affected Cariacica in 2026 would require local evidence.

What reached the school kitchen?

At the school end, the municipality reported in 2025 that its network served more than 50,000 pupils. It described weekly family-farm deliveries of vegetables, fruit, breads and other foods for menus planned by nutritionists.

That account supplies background, rather than a review of 2026 deliveries. The payment records cannot show quantities and condition on arrival, how much food was used, whether menus matched seasonal supply or what changed in children's nutrition.

Transparency does not make a programme successful

This reporting began in Albertina, Minas Gerais, where three named contracts showed the 2026 purchasing mechanism. The accessible records there allowed less detailed tracking of supplier-level execution.

That is a difference in what we could observe, rather than evidence that Albertina performed worse.

Cariacica's linked records make more of the process open to scrutiny. A resident can trace a supplier through commitments, accounting checks and payments, then ask for evidence where the record ends.

Leaving such a trail is itself a useful public capability. It helps people see what has happened and makes explanations of delay or progress easier to test.

What a dependable purchasing relationship could offer

The model's appeal lies in connecting demand that already exists. Municipal schools buy food; family farms produce it. Public purchasing can give the two a more dependable relationship.

If the arrangements work, farms may gain the confidence to plan and invest while schools receive fresh food. The conditions of that relationship determine whether the potential becomes useful in practice.

Late calls, unsuitable seasonal requirements, costly transport and slow payment can erode that predictability. Farms with little working capital may feel delays particularly sharply. Administrative work can also fall on organisations least equipped to absorb it.

Cariacica's records show contracts, processed invoices and payments to associations. Following the people on either side would tell us whether those transactions create a dependable market and a dependable food supply.

What we will watch next

Presentation revised 18 September 2026 · Two compressed relationship labels were rewritten as reader-facing questions; the evidence and claims are unchanged.

1. How does money move from the association to the farmer? What costs are deducted, and how quickly does it reach producers?
2. What happens between delivery and acceptance? Which products arrive, how often, and how are changes in quantity, quality or seasonality handled?
3. Different supplier trajectories. Why do some budget lines move quickly into payment while others remain earlier in the chain?
4. Smaller producers. Can farms with less administrative and logistical capacity participate without being pushed behind larger organisations?
5. School-side value. Do local purchases improve menu quality, reduce waste or increase resilience — and how would that be measured?
6. Ecological claims. Local or family-farm purchasing should not be treated as environmental benefit without evidence about production, transport and land use.

Following the money to farmers and the food to children

Responsibility passes through several hands: the municipality orders, producers and associations organise supply, finance staff check and pay claims, and school kitchens prepare meals.

Each part needs the others to work.

The records make some of those connections visible. The next account should follow the money into the association and the food into the school, keeping the farmers' and children's experiences alongside the municipal figures.

A paid invoice is a meaningful step. What a farmer can keep from the sale, and what a child receives at lunchtime, give that step its wider purpose.

Sources and limits

• Municipality of Cariacica, Contract 042/2026 — APRODER. Used for contracted value, dates, budget lines and food-item table.
• Municipality of Cariacica, APRODER payment 0931/2026, 31 August 2026. Used for the R$102,285.10 payment tied to the 2026 family-farm school-food call.
• Municipality of Cariacica, education liquidation records. Used for the linked APRODER liquidation and invoice NF 000.000.606.
• Municipality of Cariacica, APROVISTA payment records. Used for the July payment and the 4 September payment completing budget line 0955/2026; rechecked 13 September.
• Municipality of Cariacica, 2026 Education Secretariat budget commitments. Used to compare supplier lines at different execution stages, including ASCOPRUM and CAFC ES.
• FNDE, how family-farm food prices are defined in PNAE, updated 20 March 2026. Used for freight, packaging and other delivery costs in acquisition pricing.
• Oliveira et al., “Principais desafios para a boa execução do PNAE em municípios do Espírito Santo”, Segurança Alimentar e Nutricional, 2025. Used as an independent counter-source on execution barriers across 20 municipalities.
• Ferreira, Santos & Tosta, analysis of family farming in PNAE in northern Espírito Santo, Brazilian Journal of Production Engineering, 2025. Used as regional context, not as evidence about Cariacica specifically.
• Municipality of Cariacica, school-food system overview, 17 February 2025. Used only for background on school-network scale and the municipality’s description of weekly family-farm deliveries.
• European Central Bank, Brazilian real reference rate, 3 September 2026: €1 = R$5.8935, or R$1 ≈ €0.1697. Euro equivalents are rounded reader aids, not accounting figures.
• Produce context: University of California Agriculture and Natural Resources, “The Chayote Experiment”, and University of Kansas, “Chayote”. Used for the first-use description of chayote.

The story is based on public documents and accounting records. We have not verified the distribution of association payments to individual farmers or independently observed a 2026 delivery at a school. If later evidence changes a material claim, Experimental Newsroom will update or correct the article.

Update, 13 September 2026: Added the APROVISTA payment of 4 September and attributed transport-equipment context from the municipality’s 2 February report. Historical comparisons and euro conversions retain their stated dates. This update does not establish farmer-income or school-meal outcomes.

Correction, 4 September 2026: Euro reader-aid figures were recalculated using the European Central Bank reference rate for 3 September 2026 (€1 = R$5.8935), and direct links to the article’s source records were restored. The correction does not change the article’s conclusion.

Continue reading

• More reporting from World
• Denmark’s greener land plans depend on agreements with owners
• Japan’s flexible daycare benefit: finding hours a family can use

← Back to the Experimental Newsroom front page