Why this session matters to us
SOFI is the closest thing the UN has to an official scoreboard for hunger, and it is built almost entirely out of measurement decisions. The report tracks the Prevalence of Undernourishment (SDG indicator 2.1.1) and the Food Insecurity Experience Scale (2.1.2), and since 2020 it has added a third lens: the cost and affordability of a healthy diet. This year’s edition zooms all the way in on the first half of that lens, the cost. The full report launches in Rome on 21 July; what happened in Conference Room 6 was a preview, two FAO economists explaining what the number is, and a room of governments testing what it is for.
That makes it our kind of session. The headline that will travel from this report is a single figure with a single verb attached: a healthy diet costs more than a third of the world can pay. The interesting part is underneath, in what the metric deliberately does and does not count.
Cost is not affordability is not consumption
The Chief Economist opened by drawing a line most food-policy conversation blurs. There are three different questions, he argued, and this report answers only the first. What does a healthy diet cost is a question about agri-food systems, how food is produced and moved. Whether people can afford it is a question about income. Whether they actually consume it is a question about habits and choice.
The discipline of measuring only cost is the point. It isolates the one driver a ministry of agriculture can actually engineer. And the number it produces is stark: the FAO presenters put the global average cost of a healthy diet at 4.28 purchasing-power-parity dollars per person per day in 2025, up about 25% in five years, and already higher than the extreme-poverty line of roughly three dollars a day. By FAO’s reckoning that prices out around a third of the global population, some 2.6 billion people, from ever reaching a healthy diet.
The host, ECOSOC’s president, made the same move in a different register: the price of food is not a coefficient of production so much as a coefficient of access, affordability and equity, and therefore a measure of whether development is reaching anyone at all.
Calories are cheap, nutrition is expensive
The single most useful thing the metric reveals is a paradox hidden inside the average. Staple foods supply about half the calories a person needs but only around 16% of the cost of a healthy diet. Animal-source foods provide roughly 13% of the energy but more than a quarter of the cost. Fruits and vegetables give only about 12% of the calories, yet they account for some 40% of the cost worldwide, and closer to half in high-income countries. The foods a body most needs are the ones the market makes hardest to reach.
The finding that reframes the policy question is where that cost is generated. Between 70 and 75% of the final price a consumer pays is added after food leaves the farm, in storage, transport, processing and markets. That turns hunger from a problem of growing more into a problem of moving what already exists, and it is why economic growth alone barely shifts the number while roads, logistics and lower energy costs consistently do. The presenters’ modelling put a figure on it: a 10% gain in agri-food value-chain productivity would cut the overall cost of a healthy diet by about 9%.
The regional picture cuts against intuition. The highest cost is in Latin America and the Caribbean, at about 4.91 PPP dollars a day, roughly 20% above the world average, despite the region being a major food exporter. High-income countries, counterintuitively, pay less, because efficient systems and low food loss deliver nutrition cheaply. In Africa it is animal-source foods that are dearest, and the presenters cited an estimate that 45% of fruits and vegetables harvested there are lost before anyone eats them. Several governments recognised their own reality in that. Tanzania’s ambassador reframed basic infrastructure as an agricultural input.
The presenters also warned that the obvious policy lever can backfire. Subsidising staples, they argued, does not just make calories cheap; it pulls land and labour toward starches and raises the relative cost of the nutrient-dense foods that are already the expensive part. The prescription is to repurpose that support toward fruits, vegetables, legumes and animal-source foods, and to sequence it, building supply before stoking demand. The cautionary tale was quinoa: an export boom that raised producers’ incomes yet left the growing regions less able to afford their own nutrition, and, the presenters said, more anaemic than before.
What the benchmark can’t see
To their credit, the presenters spent as long on the metric’s limits as on its findings. The cost of a healthy diet is a least-cost benchmark: it prices a basket of six food groups against local items (apples in one country, bananas in another) using the international price-comparison programme, and it is built for a single reference person.
It is a benchmark, not a diet plan, and it counts only what carries a market price. It leaves out the food a farming household grows for itself, the cost of the time it takes to prepare raw ingredients, and unpriced costs like water. Read carelessly, a clean number stands in for a much messier reality.
That is exactly where the room pushed back. Interventions from the representative of Israel and from the NGO Plant-Based Treaty questioned the basket itself: if the reference diet includes animal-source foods and is built for one average adult male, does the choice of assumptions pre-decide the answer, and could a scenario built on legumes and other plant proteins come out cheaper and lower-emission? It was the sharpest measurement point of the session.
The FAO response drew the line carefully. The basket is not FAO’s preferred diet, it said; the structure is the average of more than thirty national food-based dietary guidelines, and the report is a statement about cost, not a nutrition recommendation.
Both things are true, and the tension between them is the honest heart of the exercise. A cost metric can only ever be as neutral as the diet it is told to price, and the diet it was told to price is itself a negotiated object.
Our read
The cost of a healthy diet earns its place precisely because it refuses to be one number for everything. By separating the supply-chain driver from income and from behaviour, it names something a government can act on rather than lament, and the “70 to 75% after the farm gate” finding is the rare measurement result that actually redirects a policy. That same discipline is its exposure. The headline figure will travel as a fact, but it rests on a benchmark basket that the room was already contesting, and a metric this load-bearing should expect that scrutiny, not deflect it.
The most candid moment came at the end, when a question about how to convince finance ministries pulled the session onto the same ground the UN’s Beyond GDP debate keeps hitting. The last mile, the presenters conceded, is not statistical at all. It runs through the cost of a country’s debt, and today the system that prices that debt gives no credit for investing in tomorrow’s nutrition.
Why it matters for the SDGs
This sits at the centre of SDG 2 (end hunger and all forms of malnutrition), where the cost and affordability of a healthy diet now complements the two long-standing yardsticks, the Prevalence of Undernourishment and the Food Insecurity Experience Scale. It reaches directly into SDG 3, with the presenters attributing nearly 12% of deaths a year to poor diets, and into SDG 1 and SDG 10, since a diet that costs more than the poverty line is a measure of exclusion, not appetite. The warnings about degrading soil and water to shave the price tag put SDG 12 in play, and the closing argument lands on SDG 17: whether a better hunger metric ever changes a decision depends on whether debt and development finance are made to price the future. If the cost of a healthy diet becomes something lenders and budgets actually weigh, it reshapes what countries are rewarded for doing. If not, it joins the well-built numbers that describe the problem without moving it.
Watch & read
- Event page, FAO Liaison Office in New York, and the HLPF 2026 programme listing.
- Full recording (UN Web TV, 1h51m).
- The State of Food Security and Nutrition in the World, the FAO flagship page; SOFI 2026 launches in Rome on 21 July 2026.
- Beyond GDP, the sister debate on whether a better metric can change a financing decision · Full HLPF 2026 coverage.
Quotations are lightly edited from an automated (Otter.ai) transcript of the UN Web TV recording and should be read as close paraphrase. This event previewed SOFI 2026 before the report’s public launch (Rome, 21 July 2026), so figures are as the FAO presenters stated them and were not verified against the published report. The Spanish-language intervention is summarised rather than quoted, and some floor speakers are cited by role where a name could not be confirmed.