The agriculture and food sector accounts for between 26 and 34 percent of global greenhouse gas emissions, depending on which boundaries the accountant draws around the system. That range is wide because food is enmeshed in every part of the economy: land use, transportation, refrigeration, packaging, retail energy, and household waste. The widely-quoted figure of one-third of human emissions is roughly correct, and it has remained roughly stable for three decades despite enormous improvements in the efficiency of production.
Where the emissions actually are
The single largest contributor inside the food system is land-use change — the conversion of forests, grasslands, and wetlands into agricultural production. Once that one-time emission is accounted for, the next largest contributor is enteric fermentation in ruminant livestock, principally cattle. Methane from rice paddies, nitrous oxide from synthetic fertilizer application, and energy use in food processing follow.
Transportation, which dominates consumer intuitions about food carbon (“food miles”), is small in the global picture — typically under ten percent of food-related emissions, and concentrated in air-freighted perishables and refrigerated supply chains rather than in routine truck shipments.
The leverage points
The single largest individual-level lever, by a substantial margin, is shifting consumption away from beef and lamb toward plant proteins, poultry, or pork. The carbon intensity per gram of protein varies by roughly an order of magnitude across these categories, and substitution at any margin reduces aggregate demand for the most land- and methane-intensive production.
The second largest is reducing food waste at the consumer level. Roughly thirty percent of food calories produced for human consumption are wasted somewhere between the field and the plate. The portion that occurs in households — primarily through over-purchasing and miscalibrated portion sizes — is several times larger, by mass, than the portion lost in retail.
What does not help much
“Buying local”, in temperate climates and outside of perishables, has a very small effect compared with composition changes in the diet. Heated greenhouses producing out-of-season tomatoes locally can have a higher carbon intensity than tomatoes shipped by sea from a sunnier climate. The correct intuition is that energy-intensity of production usually dominates distance.
The food system is one of the most globally entangled in the economy. The carbon math reflects that entanglement: levers that look obvious from the consumer side often produce smaller results than less-visible upstream levers. The decisions that move the needle most are the ones that change what gets grown and how much of it ends up in the trash.