Farming, the millennia-old practice fundamental to human survival, has degraded land to the point that today, 1.7 billion people live in areas where crop yields are falling. Without government intervention – and no, not short-term handouts that do nothing to address larger structural issues – the world could face a farmland deficit twice the size of India by 2050.
Subsidies only make the global farm structure more lopsided. The largest 1% of farms – those over 1,000 hectares – already control more than half of all farmland. 70% of government support worldwide is tied to production, rewarding high yields through intensive use of fertilisers, agrochemicals, and water. It also encourages producers to plant the same crop repeatedly on the same land, which harms soil health, and to clear forests to expand farmland, depleting soils and ecosystems. All the while, it expands big farm landholdings.
And because of their scale and influence, large-scale commercial farms that have exhausted soils and stripped forests hold the key to reversing the damage. Concentrated mainly in Latin America, Europe, and the United States, they produce more than half of global crop-derived nutrients and dominate key trade commodities (such as cereals, pulses, sugars, and oil crops).
At the other end of the spectrum, smallholders working on two acres or less make up 85% of the world’s 570 million farms but cultivate only 9% of the land, mostly across Asia and Africa. While smallholders are not blameless in land degradation, the disparity in land distribution means governments must focus their efforts on large farms. From fertiliser use and irrigation to tillage and land clearing, farm size determines both how land is managed and the capacity to act at scale.
Unsurprisingly, profit is the dominant force shaping farmers’ land-use choices. Large operations run on thin margins. Many are locked into business models and machinery built for intensive production, making a shift to more sustainable practices expensive and complex. Rebuilding the health and productivity of land requires farmers to bear upfront costs, including sometimes taking the land out of production for restoration, while the negative externalities of polluted water, carbon and biodiversity losses are felt by all and not immediately.
This mismatch between private costs and public gains is why land degradation endures despite technical solutions.
Even well-intentioned state efforts to protect land often fail because they don’t address the deeper economic drivers of degradation, such as distorted subsidies or commodity market pressures. China’s Grain-for-Green program has paid farmers to convert degraded cropland into forest to reduce erosion. But participation drops when food prices rise, as cultivation becomes more profitable than the state compensation to leave land idle. Similarly, in the US, enrollment in the Conservation Reserve Program falls in years of high commodity prices. In the EU, agri-environmental schemes under the Common Agricultural Policy lose participants when market returns outweigh restoration payments.
What to do? Governments can encourage producers to sustainably get more food out of existing farmland using technology. Precision fertiliser and water management, drought- and heat-tolerant crop varieties, and integrated pest management can help farmers produce more on existing cropland with less environmental harm. AI is already strengthening these approaches through machine-learning models that analyse satellite and sensor data to optimise input use, detect pest risks early, and guide regenerative practices, like cover cropping, where they have the greatest impact.
Countries should also coordinate more by aligning incentives, sharing data, and negotiating trade-offs, since policies in one country can unintentionally drive degradation in another. Large farms operating across borders respond to differences in land, trade, and subsidy policies by shifting production rather than changing practices.
In Brazil, the Amazon Soy Moratorium and Forest Code pushed major agribusinesses to stop sourcing from recently deforested land in the Amazon. While this reduced deforestation by about 80% between 2006 and 2014 in affected areas, pressure later shifted to the Cerrado, Brazil’s soy and cattle frontier in the east of the Amazon, where deforestation persists and enforcement remains weak.
To be sure, unintended consequences are almost unavoidable. For instance, the EUDR regulation – which requires proof that crops were not grown on recently deforested land – has raised concerns that companies may scale back purchases from smallholder farmers in Africa and elsewhere, where supply chains are difficult to trace, inadvertently risking deeper rural poverty and higher consumer prices.
The fact is that intensive production through heavy fertiliser and chemical use eventually erodes productivity. This reduces yields and lowers the nutritional quality of crops as soil becomes exhausted, biodiversity is lost, and water becomes scarce or polluted. In over-farmed regions such as Europe and North America, large-scale operations clearly illustrate this link between past land degradation and current yield losses.
For some, putting large farms at the centre of sustainability efforts only deepens inequality and invites greenwashing. But it’s also possible to design policies that foster both effectiveness and equity. Policies aiming to cut poverty can have spillover impacts on deforestation.
Reversing just 10% of degraded cropland could produce enough food to feed 154 million more people annually. Land degradation is the result of farmers’ choices shaped by government policy, not an inevitable outcome of agriculture. Agriculture can become a driver of regeneration. So can big farmers.
Maximo Torero is chief economist of the United Nations’ Food and Agriculture Organization in Rome.
