Theme 17 of 43
Agriculture, Land Structure & Agribusiness
Rural properties larger than 15 fiscal modules are 1.5% of the Rural Environmental Registry and hold 45% of the declared area. In the 2024/25 crop year, large producers took 53% of rural credit with 6% of the contracts.
The Rural Environmental Registry (CAR), run by the Brazilian Forest Service, holds the declared boundaries of 7.6 million active rural properties. The Central Bank's SICOR records every rural credit contract, with the borrower's size and the program. Comex Stat shows what leaves through the ports, and IBGE's Municipal Livestock Survey counts herds municipality by municipality. Together, these sources show the same concentration in three layers: land, the credit that finances land, and exports, which sell the product almost unprocessed to a single buyer.
Land: 1.5% of properties declare 45% of the area
In the October 2025 snapshot of the CAR, 7.58 million active rural properties declared 470.6 million hectares. The 112,870 large properties, above 15 fiscal modules, are 1.5% of the total and add up to 212.5 million hectares, 45.2% of the area. The 7.13 million small properties (up to 4 modules) are 94% of the registry and cover a third. Ranking properties by size, the largest 1% hold 42.9% of the area and the largest 10% hold 77%.
| Size (fiscal modules) | Properties | % of properties | Area (million ha) | % of area |
|---|---|---|---|---|
| Small (up to 4) | 7,130,822 | 94.0% | 156.9 | 33.4% |
| Medium (4 to 15) | 338,736 | 4.5% | 101.1 | 21.5% |
| Large (over 15) | 112,870 | 1.5% | 212.5 | 45.2% |
One rural property in a hundred declares almost half of the registered land. The CAR is self-declared and boundaries overlap, but the concentration shows up the same way under any cut.
Rural credit: large producers take 53% with 6% of the contracts
In the 2024/25 crop year rural credit totaled R$ 383.3 billion across 2.33 million contracts. Contracts with large producers were 5.7% of the total and took R$ 204.4 billion (53.3%), with a median value of R$ 608 thousand. Small producers signed 80.8% of the contracts and received 20.1% of the money, with a median of R$ 12 thousand. Pronaf, the family farming line, accounted for 17.1% of the value, and 57.9% was contracted outside any program, generally at market rates or funded by tax-exempt Agribusiness Credit Notes (LCA).
| Size | Contracts | % of contracts | Value (R$ bn) | % of value | Median contract |
|---|---|---|---|---|---|
| Small producer | 1,878,089 | 80.8% | 77.0 | 20.1% | R$ 12 thousand |
| Medium producer | 314,020 | 13.5% | 101.9 | 26.6% | R$ 198 thousand |
| Large producer | 133,333 | 5.7% | 204.4 | 53.3% | R$ 608 thousand |
A large producer's median contract is worth 50 times a small producer's. Rural credit repeats, in money, the concentration of land.
Soy: 108 million tonnes, 79% to China and almost none processed
In 2025 Brazil exported 108.2 million tonnes of soybeans for US$ 43.8 billion; 79.4% of the value went to China. Soybean meal, the first stage of processing, brought in US$ 8.1 billion and had no Chinese buyers, because China imports the bean to crush in its own plants. Soybean oil, worth US$ 1,061 per tonne, was only US$ 1.5 billion. Of every five dollars of the soy complex, four left as raw beans.
| Product | US$ billion | Million tonnes | US$ per tonne | % to China |
|---|---|---|---|---|
| Soybeans | 43.8 | 108.2 | 404 | 79.4% |
| Soybean meal | 8.1 | 23.9 | 340 | 0% |
| Soybean oil | 1.5 | 1.4 | 1,061 | 6.9% |
China buys the bean and keeps the crushing, and the industrial jobs that come with it, on its side of the Pacific.
Same soy volume, twice the deforestation: the footprint varies by trader
The Trase database traces every tonne of exported soy from its municipality of origin to the port and destination country, attaching to it a deforestation exposure — the area cleared in the municipalities feeding that flow. In 2020, US$ 34.0 billion was exported with 416,578 hectares of exposure. The revealing figure is not the total but the comparison between traders of similar size: Bunge and Cargill exported practically the same value (US$ 4.42 and 4.46 billion), yet Bunge's deforestation exposure was 2.4 times greater — 60,337 hectares against 25,655. Since volume is equivalent, the difference lies in where each sources: buying from recent agricultural frontier municipalities rather than consolidated areas is a supplier choice, not a market inevitability.
| Trader | Exports (US$ bn) | Deforestation (ha) | Hectares per US$ bn |
|---|---|---|---|
| Bunge | 4.42 | 60,337 | 13,651 |
| ADM | 4.11 | 31,196 | 7,590 |
| COFCO | 1.97 | 12,623 | 6,407 |
| Louis Dreyfus | 2.68 | 16,756 | 6,252 |
| Cargill | 4.46 | 25,655 | 5,752 |
| Amaggi | 1.63 | 9,142 | 5,608 |
Bunge clears 13,651 hectares per billion dollars exported; Cargill, 5,752 — on practically identical export volumes.
More cattle than Brazilians, and 62% of them on the agricultural frontier
The 2025 Municipal Livestock Survey counted 234.3 million head of cattle, more than the 203 million people in the 2022 Census. Mato Grosso leads with 31.6 million, followed by Pará (25.6 million) and Goiás (22.5 million). The states of the Legal Amazon plus Goiás and Mato Grosso do Sul hold 62.2% of the herd, precisely where pasture advances on the forest and the Cerrado. Chickens reached 293.3 million, 13% more than in 2023.
| Herd | Head |
|---|---|
| Chickens (laying hens) | 293,326,317 |
| Cattle | 234,251,041 |
| Pigs | 43,993,601 |
| Sheep | 22,308,942 |
| Quail | 16,787,157 |
| Goats | 13,524,707 |
| Horses | 5,563,326 |
Six in ten Brazilian cattle are in the Legal Amazon or the Center-West, where ranching is the main gateway to deforestation.
Slave labor: more than half the "Dirty List" is in the countryside
The Labor Ministry's register of employers caught subjecting workers to conditions analogous to slavery, known as the Dirty List, had 507 entries in its October 2026 version, involving 3,969 workers rescued in inspections from 2020 to 2026. Of these, 267 employers (52.7%) and 2,310 workers (58.2%) are in farming, ranching and forestry. The largest groups are in temporary crops, coffee, cattle ranching, farm support services and charcoal.
| Sector | Employers | Workers involved |
|---|---|---|
| Farming, ranching and forestry | 267 | 2,310 |
| Domestic service | 61 | 71 |
| Other sectors | 179 | 1,588 |
| Total | 507 | 3,969 |
Six in ten workers rescued from contemporary slavery on the current list were in the countryside.
Powerful crossings
- Land × size: 1.5% of CAR properties, those over 15 fiscal modules, declare 45% of the area.
- Credit × size: large producers took 53% of 2024/25 rural credit with 6% of the contracts.
- Credit × CAR: since 2019 almost every rural credit contract using public funds carries the property's CAR code, which makes it possible to cross financing with deforestation.
- Soy × China: 79% of soybeans exported in 2025 went to China, which does not buy the meal.
- Soy × trader: with similar volumes, Bunge had 2.4 times Cargill's deforestation exposure (Trase, 2020).
- Cattle × frontier: 62% of the cattle herd is in the Legal Amazon, Goiás and Mato Grosso do Sul.
- Countryside × slave labor: 58% of the workers on the current Dirty List were in agriculture.
Explanatory hypotheses
Land concentration is a colonial legacy that no national-scale land reform has reversed, and rural credit reproduces it: whoever has land has collateral, and whoever has collateral gets credit. Export specialization in raw grain responds to Chinese demand and to importers' tariff rules, which protect processing in the destination country. On the agricultural frontier, cattle occupy newly cleared land before crops arrive, which explains why the herd is moving into the Amazon and the Cerrado. Slave labor concentrates in seasonal, isolated activities such as harvesting, pasture clearing and charcoal, where inspections rarely reach.
Policy implications
The link between each rural credit contract and the property's CAR record, present in almost every operation using public funds since 2019, makes it possible to condition subsidized credit on the absence of recent deforestation, property by property. Analysis of CAR records, still pending for most properties, is a condition for that check to work. Revisiting the weight of credit outside programs and of tax-exempt Agribusiness Credit Notes would change who benefits from the implicit subsidy. In the soy and beef chains, traceability to the property of origin and due-diligence requirements on buyers, such as the European deforestation regulation, narrow the room for suppliers on newly cleared land. And the Dirty List only has an effect if banks and buyers check it against their own customer and supplier records.