The climate-ag grab bag

The framing does most of the work, so it is worth stating precisely. Three things have to happen at once: produce far more calories, on no additional land, with far fewer emissions. Grunwald’s numbers are large. Nearly 10 billion people by 2050 need roughly six quadrillion more calories, which he renders as a dozen extra Olive Garden breadsticks per person per day. Food and land are about a third of total emissions and have to fall 75% to 80%. And on the current trajectory, even assuming yields keep improving at the rate they did through the green revolution, feeding that population takes about 1.5 billion additional acres, which he calls two more Indias and which in practice means the Amazon and the Congo. If yields stopped improving entirely it would be seven more Indias, and since yield growth has actually tapered over the past couple of decades, the recent rate implies three. What makes this a land problem rather than a hunger problem is a point he is careful about: the world is reasonably good at meeting food demand, because if people will pay for food, someone grows it. Almost a billion people are food insecure and that is terrible, but the binding failure mode is not famine. It is that the extra food gets grown by clearing the forests and wetlands that are storing the carbon.

Once land is the constraint, the usual environmental ranking flips. Grunwald does not soft-pedal the case against industrial agriculture and lists it himself: animals treated badly, people treated badly, too many antibiotics, water and air pollution, real greenhouse gas emissions, bad politics. His defense is narrow and functional. Factories are good at manufacturing, and agriculture’s job for the next 25 years is to manufacture a great deal of food without expanding its footprint. His evidence for what happens when you subtract the inputs is Sri Lanka, where an agrochemical ban was followed by collapsing yields, a country that could no longer feed itself, and production and pollution exported elsewhere. The same logic demotes regenerative agriculture, and the demotion is more qualified than it is usually reported. He grants that regenerative practices do good things for soil and can help on-farm biodiversity in moderate ways, and says some regenerative practices are consistent with excellent yields and he is all for those. His objection is specifically to the yield-reducing kind, which he thinks is most of it: agriculture that is less intensive has to be more extensive, and the extra acres come out of peatland and forest. A practice can lower emissions on the field it is applied to and raise them globally.

The same test then ranks food itself, and this is where he puts most of the available leverage. Beans, lentils and other pulses are the most land-efficient calories, with corn and soy close behind; chicken and pork are worse; beef and other ruminants are, in his word, spectacularly inefficient. It takes about nine calories of grain to make one calorie of chicken, while beef runs 30, 50, or by some accounts 100 times less efficient than eating the grain directly, and dairy is worse than chicken. That reframes the cattle problem in a way worth holding onto. The methane in the burps is real, and he is enthusiastic rather than dismissive about feed additives, which some people think cut emissions by as much as 80% while the animal is being fed them. But cattle spend only a small fraction of their lives in feedlots, and the burps are a small part of what makes beef a climate problem. The larger prize is producing beef on less land. He points to degraded Brazilian cattle ranches, hundreds of millions of acres carrying almost no cows, where better grass and water infrastructure can triple or sometimes quadruple output. His framing is the Willie Sutton one: go where the emissions are.

What keeps the argument honest is that he refuses to let any single lever carry the load, including his own. Even a 50% cut in rich-country beef consumption combined with a 50% cut in global food waste, both very large wins, would still leave the need to raise yields significantly. And the supply side is harder than it was, because the green revolution already spread and the easy gains are taken: only 5% to 7% of the world’s irrigable farmland still needs irrigating, roughly half the fertilizer already applied is wasted, so more fertilizer is not the answer, and climate change is itself a threat to yields, with some scientists putting potential wheat losses as high as a third over the next few decades. What he asks for is not more green revolution but a greener one: fertilizers that get nitrogen into plants instead of into the Gulf of Mexico, where the dead zone is the size of Connecticut, and without the nitrous oxide emissions that come with it. That is the shape of the conclusion, which is less a conclusion than a refusal to accept a small one. We need to do all the things.