Why Food Systems Are Still Missing from Climate Policy

At COP30 in Belém this November, in the country that grows more soy and slaughters more cattle for export than any other, the formal negotiating track on agriculture collapsed. The Sharm el-Sheikh joint work on agriculture and food security ended abruptly in the first week, with delegates agreeing only to pick the conversation back up at COP31. Food was left out of the summit's final agreement entirely. Raj Patel of the International Panel of Experts on Sustainable Food Systems put it starkly: food had been "erased" from the negotiations, less than two years after 160 countries signed a sustainable agriculture declaration in Dubai. The setting made the omission almost theatrical. But it was not an aberration. It is the pattern, and understanding why requires looking past the summit itself to the accounting, the politics and the lobbying that keep reproducing it year after year.

Where the Money and the Molecules Don't Match

Food systems, from fertilizer manufacture through farming, land clearing, processing, transport and waste, generate roughly a third of global greenhouse gas emissions. The Food and Agriculture Organization put the 2015 figure at 18 billion tonnes of CO2 equivalent, about 34 percent of the anthropogenic total, with methane from livestock and rice paddies accounting for over a third of that share. No other sector comes close to matching energy and industry as a source of warming while attracting so little of the policy machinery built to address it.

The numbers illustrate the mismatch cleanly. An analysis by the investor network FAIRR found that among G20 countries, only Indonesia has set an emissions reduction target specifically for agriculture. Sixty-five percent of G20 members have quantified sectoral targets somewhere in their climate plans, but those targets cluster almost entirely around energy, transport, housing and black carbon. Not one G20 country has a quantified target for methane from livestock, even though livestock alone accounts for an estimated 14.5 percent of global emissions and 15 of the 20 G20 nations have signed the Global Methane Pledge. The World Resources Institute's ongoing review of nationally determined contributions under the Paris Agreement finds the same story repeated across the developing world: agriculture appears in NDC narratives as a vulnerability to be adapted to, far more often than as a source of emissions to be cut.

Finance tells a parallel story. The Climate Policy Initiative's most recent landscape analysis found that agrifood systems receive about 7.2 percent of global climate finance, roughly 95 billion dollars a year, against an estimated 1.1 trillion dollars needed annually by 2030 to bring the sector in line with climate goals. That funding has grown sharply since 2019, when it stood near 28.5 billion dollars, but 90 percent of that growth came from just the European Union and China. Strip those two out and the picture for sub-Saharan Africa, South Asia and Latin America, the regions where smallholder farmers face the sharpest combination of climate exposure and limited adaptive capacity, looks far bleaker. A sector responsible for a third of emissions is being financed at a fraction of that share, and what finance does arrive is concentrated where it is least urgently needed.

Why a Cow Is Harder to Audit Than a Coal Plant

Several forces explain the gap, and they compound rather than operate in isolation. The first is measurement. Energy emissions come from a relatively small number of point sources, power plants, refineries, factories, that can be metered and audited with reasonable confidence. Agricultural emissions are diffuse, biological and dependent on soil type, livestock breed, feed composition and land management practices that vary field by field. Building an accounting system precise enough to anchor a binding national target is a genuinely harder technical problem than doing the same for a coal plant.

The second is political economy, and nowhere is the calculation clearer than in India, which holds the largest stake in how this argument gets resolved. Agriculture employs nearly half of India's workforce, including hundreds of millions of small and marginal farmers already contending with groundwater depletion, heat stress and rising debt. At COP30, Indian negotiators pushed back against language that would have moved toward binding methane reduction commitments in agriculture, arguing instead for what officials described as policy space for development and flexibility in how agricultural emissions are reported. Proposals to shrink livestock herds, in this reading, collapse the difference between a Punjabi dairy smallholder and an industrial feedlot. India did succeed in writing food security and rural livelihoods into the COP30 text, but secured no comparable breakthrough on the 300 to 400 billion dollars in annual adaptation finance it says the developing world will need by 2030. The same asymmetry recurs across the Global South: framing agriculture through adaptation is politically survivable in a way that framing it through mitigation is not, and NDCs reflect that choice systematically rather than incidentally.

Brussels' Own Contradiction

The reluctance is not confined to the developing world. The European Union, which presents itself as the standard-bearer of ambitious climate policy, runs one of the starkest versions of the same contradiction through its own budget. The Common Agricultural Policy consumes close to a third of the EU's seven-year budget, roughly 387 billion euros for 2021 to 2027, and Brussels has counted a substantial share of that spending toward its climate targets. The European Court of Auditors examined that accounting in a 2021 special report and found it did not hold up: CAP payments made up close to half of all EU climate-related spending on paper, yet agricultural emissions across the bloc had barely moved in more than a decade. Direct payments to farmers, the policy's largest line item, are disbursed largely on the basis of land farmed rather than emissions avoided, meaning a farm can collect climate-labeled subsidy while its emissions profile stays flat. The mismatch matters beyond Europe because the CAP is routinely cited internationally as a model for large-scale agricultural support, and its climate accounting has been effectively hollow for years without denting either its reputation or its share of the EU budget.

The Lobby in the Room

The third force is organized influence, and it interacts with both of the others: where measurement is ambiguous and political stakes are high, a well-placed delegation has more room to shape outcomes. DeSmog's tracking of COP delegate badges found 204 agriculture-linked delegates at COP29 in Baku, including 52 from the meat and dairy sector alone, more than the entire delegation of Barbados. Nearly 40 percent of those delegates carried official country badges rather than observer credentials, up from just 5 percent three COPs earlier, giving them access to closed-door rooms that agroecological farmer groups and civil society representatives do not get. Brazil's own delegation at COP30 included representatives from JBS, BRF and Marfrig, the country's largest meatpackers, present at the same summit where the agriculture negotiating track collapsed within days. That access advantage shapes what gets proposed while a text is drafted, and just as often shapes what gets quietly dropped before it reaches plenary.

What a Real Agriculture Target Looks Like

That structural picture is not universal, which is precisely what makes Denmark's example instructive. Denmark passed the world's first tax on agricultural emissions in 2024, proof that a binding sectoral target is not simply undoable at the national level. Farmers will pay 120 kroner, about 18 dollars, per tonne of CO2 equivalent from livestock, fertilizer, forestry and disturbed carbon-rich soils starting in 2030, with the effective rate rising to 300 kroner by 2035, after a government deduction that cuts the sticker price by 60 percent. Revenue flows into a fund supporting the livestock sector's transition for at least two years after the levy takes effect. Denmark expects the measure to cut 1.8 million tonnes of CO2 equivalent by 2030, part of a broader goal of reducing agricultural emissions 55 to 65 percent by that year. Some Danish environmental groups counter that the deduction and transition subsidies let intensive livestock production off comparatively lightly rather than restructuring it. But the policy accomplished something no G20 country's NDC has managed: it put an explicit, rising, legally binding price on agricultural methane and nitrous oxide, negotiated with the farm sector rather than imposed over its objection. Whether that model travels to countries with larger, less consolidated farm sectors and thinner safety nets, including the G20 economies FAIRR examined, remains untested. But Denmark establishes that the absence of agricultural targets elsewhere reflects a political choice, not a technical ceiling.

Declarations Are Not Targets

None of this means the diplomatic architecture has stood still. The Emirates Declaration signed by 134 countries at COP28 committed signatories to integrate food into their climate plans by 2025; COP29 produced the Baku Harmoniya Climate Initiative and a billion-dollar package for farmer weather forecasting; COP30 added a hunger and poverty declaration backed by 43 countries and a Gates Foundation pledge of 1.4 billion dollars for smallholder adaptation. These commitments are real, and they matter to the farmers they reach. But declarations and side initiatives are not the same as binding targets inside the instrument that actually governs national climate commitments, and the gap between the two is where the sector's emissions keep growing.

That gap now falls to Antalya. After more than a year of deadlock between Turkey and Australia over hosting rights, the two countries struck an unusual compromise in November: Turkey will host COP31 in the Mediterranean resort city of Antalya, while Australia holds the presidency and manages the negotiations, drafts the texts and issues the final cover decision. Neither government has signaled agriculture as a headline priority the way Brazil, itself a major agricultural exporter, did before Belém, and the split arrangement adds a layer of coordination previous COPs have not had to manage. Whether the joint work on agriculture, still parked mid-conversation after collapsing in Belém, gets a genuine second attempt in Antalya or another procedural extension will be an early signal of how seriously the format intends to treat the sector.

What is at stake is not abstract. Missing the 1.5 degree pathway without addressing a third of global emissions is not a plausible route to meeting it at all, whatever else energy and industry manage to deliver. The next round of NDCs, due before COP31, is the test of whether the Emirates Declaration's promise, or Denmark's example, becomes the norm rather than the exception, or whether food systems join the long list of climate commitments that were signed loudly and implemented quietly, if at all.

Sources: FAO, "Food systems account for more than one third of global greenhouse gas emissions"; Climate Policy Initiative, "New data reveals global agrifood systems receive only 7% of total climate investment"; FAIRR, "Nationally Determined Contributions Still Lack Ambition on Agriculture"; FAIRR, "What Happened at COP29 on Food and Agriculture?"; Carbon Brief, "COP30: Key outcomes for food, forests, land and nature at the UN climate talks in Belém"; World Resources Institute, "134 Countries Sign the Emirates Declaration on Sustainable Agriculture and Put Food High on the Climate Agenda at COP28"; DeSmog, "Meat, Dairy and Pesticide Lobbyists Return in High Numbers to Climate Summit"; Inside Climate News, "Agriculture and Food Get Their Day—Again—at the Annual UN Climate Summit"; Carbon Brief, "Q&A: How Denmark plans to tax agriculture emissions to meet climate goals"; European Court of Auditors, Special Report 16/2021, "Common Agricultural Policy and climate: Half of EU climate spending but farm emissions are not decreasing"; Down To Earth, "COP30's Impact on India's Agriculture: Climate Justice and Future Challenges"; Climate Home News, "COP31 will be hosted by Türkiye but run by Australia"