Ethanol/Grunwald 1: Michael Grunwald Is Right About Ethanol. But What Actually Spares the Land?

What Does an Acre Produce? Farming, ethanol and the economics of living landscapes — essay 1 of 4

Sep 29, 2026 · @Thorsten Arnold

Michael Grunwald’s We Are Eating the Earth [1] makes the strongest mainstream case yet against burning cropland in fuel tanks, and the National Farmers Union agrees with nearly all of its assessment. As farming practitioners, we disagree with some of the policy implications and offer our insights into agricultural markets that are driving farmexpansion. Grunwald shows that higher yields make it possible to use less land. He does not show what makes it happen. For farmers and for the land, that missing step is the whole question.

A good book about land

Grunwald has followed policy and money for three decades — at The Washington Post, Time and Politico — and won the George Polk Award for national reporting. In 2008 he wrote Time‘s cover story “The Clean Energy Myth,” linking U.S. ethanol to clearing in the Amazon. His new book is built around the researcher behind that story: Tim Searchinger, a lawyer turned land-use scientist at the Environmental Defense Fund, the World Resources Institute and Princeton.

The argument starts from simple splits regarding how we use land. Fossil fuels cause about two-thirds of climate change; the remaining third is mostly land. Farms and pastures cover two of every five acres on Earth, and agriculture is the leading driver of deforestation and biodiversity loss and wetland destruction and water pollution. The central concept is carbon opportunity cost: land is never free, because every acre could store carbon as forest, wetland or grassland. From there, three moves:

1. Bioenergy is a trap. Ethanol, biodiesel and wood-pellet power counted as climate solutions only because the accounting left out the land they consume.
2. Feeding the world is the harder problem. Roughly 50% more food by 2050, without further farmland expansion — through less beef in rich countries, less waste, less bioenergy, and higher yields on existing land.
3. Most popular fixes disappoint — plant-based meat, cultivated meat and, above all, what he calls “the soil fantasy” of regenerative carbon farming. His sympathies lie with high-yield, high-tech agriculture.

It is a rigorous and brave book that compiled excellent recent data and insights. Grunwald names the environmental organizations that defended corn ethanol as a “bridge fuel” to protect their coalitions. On ethanol he has been right since 2008.

Where we agree: ethanol

On the technical substance, Grunwald’s evidence and the NFU’s work agree point for point:

1. Land is never free. Conventional lifecycle accounting omitted indirect land-use change. Divert corn, prices rise, and pasture and forest are converted elsewhere [2].
2. “Marginal” land is rarely marginal. Early bioenergy estimates counted the Amazon, the Congo and the Serengeti as “potential cropland” [1] — the same boundary error that relabels recovering ecosystems as “idle.”
3. The bridge that never ends. Environmental groups defended corn ethanol as a bridge to cellulosic; cellulosic never arrived; the same industry is now morphing toward “sustainable aviation fuel” [1]. Policy-promise lock-in, documented from the inside — the pattern I traced for Canada in The Bridge That Never Ends.
4. Demand is an artefact of political choice.S. ethanol “plateaued at 15 billion gallons, unable to compete beyond its mandate” [1].
5. The models flatter ethanol. They assumed generous yield responses, and credited ethanol with emission cuts because higher grain prices made poor people eat less [1].

Grunwald offers the strongest recent empirical confirmation. The U.S. Renewable Fuel Standard raised corn prices by 30%, expanded U.S. cropland by 2.1 million hectares, raised water-quality degradants by 3–5%, and left corn ethanol’s carbon intensity “no less than gasoline and likely at least 24% higher” [3].

The ethanol industry disputes these findings [8], and official carbon-intensity scores still credit corn ethanol with large emission cuts — largely because they assign indirect land-use change a small value. The industry reminds us that about a third of the corn’s weight returns to livestock as distillers’ grains [9], which somewhat softens the land impact . Yet my argument mainly rests on what ethanol does to farm economics and to land.

And the costs travel via global market signals. Brazil lost a Rhode Island-sized area of forest in six months during the 2007–08 grain price spike [1]. Livestock farmers around the world are cast as the destroyers of nature; part of the conversion attributed to them is demand displaced from our fuel tanks. And if taking into account the economic system, they are merely doing what banks require them to do.

Where we agree: expansion must stop

Living landscapes need trees and wetlands and hedgerows and riparian corridors, and fenceline-to-fenceline cropping provides none of them. Forests, wetlands and native grasslands cannot remain the reserve land bank for agricultural demand. Southern Ontario had lost roughly 72% of its pre-settlement wetlands by 2002, mostly to agriculture [4]. Further expansion of farmland has to stop, and in places be reversed. Grunwald is right that protection must be explicit. It does not happen as a by-product of higher yields.

The missing acre

So here is the question Grunwald’s book raises and does not answer:

If a farmer raises corn yield by 10%, what mechanism causes 10% of that farm to become forest, wetland or grassland?

There isn’t one.

Grunwald’s land arithmetic is sound: higher yield means fewer hectares technically required for a given amount of food. But there is no automatic economic arrow from fewer hectares required to fewer hectares farmed. Across countries, rising yields were generally not accompanied by declining cropland [5]; in integrated global markets, productivity gains can even increase cropland use and emissions [6]. Grunwald half-concedes the point through Searchinger’s “wicked problem”: farmers who grow more per acre “can usually make more money by clearing more acres” [1]. He treats it as the exception. Farmers know it as the rule.

I surmise that the reason behind this is Grunwald’s epistemic distance. Grunwald sees agriculture from the air and from Washington. From the air, a farm is land use — acres that could be forest. From Washington, farmers are a lobby chasing subsidies. What he rarely sees is the farm balance sheet, where most farmers are price-takers carrying land and equipment debt that must be serviced every year. On that balance sheet, productivity gains drive three powerful economic feedbacks:

1. Technology forces farmers to produce more, more cheaply, or exit — Willard Cochrane’s treadmill.
2. Land turns higher returns into higher land prices, rents and mortgages.
3. Machinery turns capital investment into pressure to farm more acres.

All three make the supposedly “spared” acre financially valuable. The second essay in this series takes them apart.

There are, in other words, two ledgers. Grunwald counts one opportunity cost of land: the carbon a forest or wetland would have stored. The farmer counts another: the mortgage payment a hedgerow does not earn. The first appears in climate models and on no farm ledger. The second appears on every farm ledger and in no climate model. Land sparing will not happen until something connects them.

Ethanol makes Grunwald’s problem worse

This is where ethanol stops being one bad fuel among many and becomes central. North American agriculture already produces more than its markets absorb at remunerative prices; the treadmills guarantee it. Ethanol was the political answer: manufacture another source of demand. Fuel markets can soak up volumes of corn that food markets never could.

But ethanol does not relieve the pressure on land. It deliberately adds demand for cropland at exactly the moment the ecological objective requires reducing it. Higher corn prices call forth more production, raise what land can earn, and raise what every non-crop acre costs its owner in foregone revenue. The new supply erodes the scarcity that propped up prices, and the pressure to manufacture demand returns — higher blends, low-carbon fuel credits, aviation fuel. The bridge loops.

Ontario is on that loop now. O. Reg. 663/20 ratchets renewable content in gasoline to 13% in 2028 and 15% from 2030, with no sunset [7]. Every percentage point is a standing claim on farmland — here, and wherever displaced production goes.

Critics will say Ontario has no prairie left to plough, and its ethanol runs on Ontario corn. That misses how land markets work: displaced demand travels through global commodity prices, and the conversion happens wherever land is cheapest to clear. The local signal is there too — southwestern Ontario farmland now averages about $33,700 an acre [10], a price every remaining woodlot and wet corner has to be measured against. What Ontario lacks is a systematic record of hedgerow, woodlot and wetland loss on farmland since the mid-2000s. The NFU should call for one.

And to those who answer that the world needs more food: if grain were scarce, no government would need to mandate burning it. The fourth essay in this series takes that objection apart.

Where we start: a moratorium

The NFU’s ask is concrete. Ontario should place a moratorium on further ethanol expansion in gasoline under O. Reg. 663/20, freezing renewable content at current levels. It is the minimum consistent response to evidence Grunwald and the NFU share, and it requires no agreement on anything else.

The moratorium is where Grunwald and farmers already agree. The harder conversation is what agricultural policy must do after it: how to take the pressure off land without taking the income out of farming. That conversation needs the view from the balance sheet. Grunwald should have it with farmers.

Next in the series: “Three Treadmills and a Fuel Tank” — why higher yields do not automatically spare land.

References

1. Grunwald, M. (2025). We Are Eating the Earth: The Race to Fix Our Food System and Save Our Climate. Simon & Schuster.
2. Searchinger, T. et al. (2008). Use of U.S. croplands for biofuels increases greenhouse gases through emissions from land-use change. Science 319: 1238–1240.
3. Lark, T. J. et al. (2022). Environmental outcomes of the US Renewable Fuel Standard. PNAS 119(9): e2101084119.
4. Ducks Unlimited Canada (2010). Southern Ontario Wetland Conversion Analysis.
5. Rudel, T. K. et al. (2009). Agricultural intensification and changes in cultivated areas, 1970–2005. PNAS 106(49): 20675–20680.
6. Hertel, T. W., Ramankutty, N. & Baldos, U. L. C. (2014). Global market integration increases likelihood that a future African Green Revolution could increase crop land use and CO2 emissions. PNAS 111(38): 13799–13804.
7. Ontario Regulation 663/20 under the Environmental Protection Act — renewable content rising to 13% (2028) and 15% (2030 onward). Cleaner Transportation Fuels: Renewable Content Requirements for Gasoline and Diesel Fuels.
8. Renewable Fuels Association (2022). Rebuttal to the Lark et al. report.
9. Hoffman, L. & Baker, A. (2011). Estimating the Substitution of Distillers’ Grains for Corn and Soybean Meal in the U.S. Feed Complex. USDA Economic Research Service.
10. Farmtario (2026). Farmland value stagnates in Ontario in 2025. Reporting Farm Credit Canada’s 2025 Farmland Values Report.