Segment 1 — The Cold Open

From a small plane over the interior of São Paulo state, the ground looks almost oceanic: kilometer after kilometer of sugarcane, grown less for the table than for the tank. A day’s travel north, along the ragged arc of settlement that rings the Amazon, the view inverts. Pale rectangles of new pasture appear where forest stood last year, trees burned or stacked in windrows, grass seeded for cattle that used to live much farther south. This was designed to wean Brazil off imported oil and to give its sugar industry a domestic future. Instead it helped set off an indirect land-use cascade — cane onto pasture, soy into the gap, cattle into the trees — that tied a celebrated climate fuel to some of the fastest forest loss on Earth.

Segment 2 — The Good Intention

In the mid-1970s, Brazil’s economy ran on oil it did not have. The first oil shock had turned a manageable import into a national emergency: tankers and dollar bills leaving the country, inflation already high, a military government watching a strategic weakness become a fiscal one. What Brazil did have was sugarcane, a crop stitched into the coast since the colonial mills, and a sugar industry that had just ridden a brief world-price boom and then watched it collapse. The National Alcohol Program — Proálcool — braided those two problems into a single industrial policy. Ferment cane into ethanol. Blend it into gasoline or burn it neat. Cut the oil bill. Give mill owners a home market that foreign sugar buyers could not snatch away. President Ernesto Geisel’s government was not groping in the dark. Brazil already knew how to grow cane, already had mills, already had a car culture that could be told what to put in the tank. Ethanol looked like energy independence with rural jobs attached, a patriotic substitute for a commodity OPEC could reprice overnight. Planners were not aiming at the Amazon. Cane is a crop of the Center-South — São Paulo and its neighbors — of good soils, a dry harvest season, and roads that already existed. The rainforest was a different map, a different climate, a different political problem. Substituting a field of grass in the agricultural heartland for a barrel of imported crude looked, with the information on the desk, like a clean swap.

Segment 3 — The Implementation

The program arrived as a stack of ordinary tools: blending mandates, subsidized credit for distilleries, guaranteed purchases, and pumps that made ethanol a daily liquid rather than a laboratory curiosity. Through the late 1970s and into the 1980s, mills multiplied in the traditional cane belt, blending shares rose, and automakers sold cars built to run on hydrous alcohol. For a stretch the results looked like the textbook promised. A large share of the light-vehicle fleet was drinking cane instead of gasoline. The oil-import needle moved. Then petroleum prices fell, sugar prices occasionally recovered enough to pull cane back toward the export crystal, and drivers of dedicated-ethanol cars found themselves stranded between empty alcohol pumps and a policy that had lost its urgency. The industry did not vanish. It waited. In the early 2000s it returned in a more elegant machine: flex-fuel engines that could swallow any mixture of ethanol and gasoline, so the driver no longer had to bet on a single fuel. Oil was expensive again. Climate politics had begun to treat biofuels as a way to decarbonize driving without waiting for a new fleet. Brazilian officials and millers could say, with some justice, that they had built the world’s most sophisticated cane-ethanol system — high field yields, bagasse burned for mill power, a greenhouse-gas story at the tailpipe that beat gasoline. Environmental campaigns that had learned to fear palm oil and American corn often treated Brazilian cane as the exception that proved the rule. It grew far from the Amazon. It often grew on land that was already open. From Brasília and from São Paulo, the map still looked like a success that had merely been interrupted and then improved.

Segment 4 — The Unintended Consequences

The exception was less exceptional than it looked, because land is a market, not a photograph of one field. As flex-fuel cars took over showrooms and ethanol demand recovered, then surged, cane did what a profitable crop does: it expanded. It expanded overwhelmingly in the Center-South, and overwhelmingly onto pasture. On a single property that can look like progress. Degraded cattle grass becoming a high-yield energy crop is the definition of intensification, and mill agronomists could point to it with genuine pride. The trouble started on the next property, and the one after that. Ranching in Brazil is a land-hungry business with thin returns per hectare, low stocking rates on much of the extensive herd, and a long habit of moving when nearby land gets dear. Cattle are not just meat. In a frontier economy they are a way to occupy ground, to claim it, to hold it as a store of value while grass does the slow work of justifying the title. When millers and cane growers bid up pasture in São Paulo, Minas Gerais, and Goiás, a rational rancher sold, took the capital, and bought a much larger tract where forest was still being priced as if it were empty — in Mato Grosso, Pará, Rondônia, along the roads that had opened the southern and eastern Amazon. One hectare released in the cane belt did not become one spare hectare for the country. It became several hectares of new clearing at the forest edge, because frontier land was cheaper and because the national herd was still growing.

Soy entered the same cascade from the side, which is why the story is not only about alcohol. The soybean frontier — pushing out of Paraná and Rio Grande do Sul into the Cerrado and then toward the Amazon — was driven by export demand, by breeding that taught a temperate crop to live in the tropics, and by a world hungry for protein meal. Cane and soy both prefer land that is already cleared and reasonably close to silos, crushers, and mills. Both, therefore, preferred pasture. Where soy took the grass in Mato Grosso, the cows did not disappear. They moved again, another step north and west, into the places where the only cheap land left was forest. Researchers who later tried to follow this chain called it indirect land-use change. The feedstock never had to touch the trees. The displacement did. Direct conversion of Amazon forest to sugarcane remained rare. Humidity and rain in the harvest season make the crop a poor fit there anyway, which is why the satellite picture of the cane belt could look so virtuous while the satellite picture of the forest did not. In the 2000s, as ethanol production climbed to meet a flex-fuel fleet and as soy filled trucks bound for the coast, Amazon clearing ran at some of the highest rates ever recorded. Newly deforested land went, by and large, to cattle pasture. The fuel in a tank in Campinas did not fell a specific tree in São Félix do Xingu in a way a prosecutor could photograph. It changed relative prices in a land market that already rewarded clearing.

Second-order effects piled onto the first. The Cerrado — a savanna woodland of enormous biodiversity and deep carbon stocks — absorbed much of the direct expansion of both soy and cane and received a fraction of the Amazon’s international grief. Millions of hectares of grass became cropland in the agricultural heartland, which raised land prices, favored larger holdings, and made it still more attractive to speculate in frontier claims. Labor followed the mills: hard harvest work in the cane, then increasing mechanization that cut jobs as it cut costs. Water followed too, in dry pockets of the Center-South where a thirsty crop met a thirsty climate. Third-order, the climate arithmetic of ethanol itself became a fight. Count only the mill, the tractor, and the tailpipe, and cane alcohol can look like a real improvement on gasoline. Count the forest that would not have been cleared if the cattle had stayed on the old grass, and a fuel adopted to reduce emissions can spend years paying back a carbon debt incurred somewhere the mill never planted. The deepest irony was spatial. A policy designed to substitute for oil — a commodity drawn from underground — ended up rearranging a commodity produced on the surface: grass. And grass, in Brazil, is the leading edge of the forest.

Segment 5 — The Aftermath

By the second half of the 2000s the connection was no longer a rumor among agronomists. Brazilian and international researchers were mapping the pasture-to-cane and pasture-to-soy displacements and asking what they implied for the forest. The government’s response, and the industry’s, went first after the direct footprint, which was the footprint you could regulate with a map. Agroecological zoning restricted where sugarcane could expand and kept mills out of the Amazon biome — a real constraint on a distillery, and a weaker one on the rancher who sold to a mill and bought land somewhere the zoning did not reach. Soy traders, under campaign pressure, agreed not to buy beans grown on Amazon land cleared after a cutoff date. Direct soy-driven deforestation in the forest biome fell sharply. Cattle, the residual user of freshly felled land, were not in that agreement. Later tools — enforcement against unauthorized clearing, restrictions on rural credit, a rewritten forest code — brought Amazon rates down for a time, which showed that policy could bite, and then those rates proved politically fragile, which showed that the underlying land market had not been repealed. Ethanol itself was not reversed. Flex-fuel remains the everyday architecture of Brazilian driving. Cane remains a pillar of the rural economy. Blending mandates still exist in countries that learned, from Brazil and from their own programs, to worry about leakage. California, the European Union, and others began writing indirect land-use factors into biofuel rules — contested multipliers, fought over by millers and environmental economists, but an admission that the tank is connected to the frontier. The meta-irony is familiar. The cleaner the certified hectare of cane, the more the remaining damage hides in the hectare that was never certified because it was never cane.

Segment 6 — The Lesson

Land-use leakage is what happens when you fix a field and ignore a market. If a policy makes one use of land more valuable, the uses it displaces do not vanish; they look for the next cheapest acre, and in a country with a forest frontier that acre is often trees. Measuring only the thing you subsidize — the ethanol in the pump, the soy in the silo — will flatter the intervention, because the harm has been asked to relocate. Brazil’s planners in the 1970s were solving the problem in front of them with the crop they had, and they did cut an oil dependence that was real. The pattern their success revealed is now the pattern of a dozen climate policies: carbon offsets that protect one forest while logging shifts to the next; renewable-fuel rules that treat a crop as carbon-neutral so long as it did not personally bulldoze a biome; even the siting of “clean” infrastructure that assumes the map is closed. The question left on the table is the one every energy transition has to answer in public, not just in a lifecycle spreadsheet: when we count the benefit in the tank, whose land are we leaving out of the ledger?

💬 Reply to this email — Patrick reads every one.

Share: X · LinkedIn · WhatsApp

Forwarded this email? Subscribe here — it's free.