The world promised zero hunger by 2030. Its official gauge can't report any value below 2.5% — and 62 of the 165 economies it scored for 2023 are recorded at that floor

In 2015 every UN member state adopted Sustainable Development Goal 2 — "Zero Hunger" — and its first target: end hunger by 2030. The official yardstick for that target, SDG indicator 2.1.1, is the Food and Agriculture Organization's Prevalence of Undernourishment (PoU), the modeled share of a population without enough dietary energy for an active life. We pulled the series that carries it — the World Bank's SN.ITK.DEFC.ZS, sourced from FAO, updated 13 July 2026, latest data year 2023 — and kept every economy with a 2023 value. That is 165 in all (of 212 records carrying a 2023 value, 43 regional aggregates plus four income-group aggregates are set aside). These are modeled estimates: most countries cannot observe caloric shortfall directly. ^mmodeled Sixty-two of the 165 — 37.6% — are recorded at exactly 2.5%. Not near it: at it, the identical number. No economy in the file sits between zero and 2.5%; the next value up is 2.7%. That is because FAO does not report a PoU below 2.5%: at that level the margin of error is expected to exceed the estimate itself, so FAO publishes the result only as "below 2.5%" rather than a point figure. The World Bank series that carries the indicator encodes that "below 2.5%" result as the number 2.5, and reports a genuine estimate that lands at 2.5% the same way — so an exact 2.5 in the data marks the floor without saying whether the true figure sits at it or below it. The floor pools economies that share nothing else. China and Brazil sit on it beside Norway and Switzerland. By income group the 62 break down as 45 high-income, 16 upper-middle, and one lower-middle (Uzbekistan). Only nine high-income economies register above it at all — led by Macao and Trinidad and Tobago at 11.2%, then Oman (5.9%) and Panama (5.7%). Across all 165, the median economy is at 5.2% and the highest is Haiti at 54.2%.

The following is analysis, not fact. The single vantage most desks take is the global rate: 8.5% in 2023 on this series, up from a 2017 low of 7.4% and roughly flat since the pandemic. (FAO's 2025 flagship estimates a slight easing to 8.2% in 2024 — still above the pre-pandemic mark, and it projects 512 million people chronically undernourished in 2030.) ^mmodeled That number tells you the world is off pace. It does not tell you the gauge cannot see the finish line. Read all 165 filings at once and the limit is structural. The goal is zero. The smallest positive value the primary indicator can report is 2.5%. A true prevalence at or below that — down to zero — can surface only as 2.5%, and a figure shown as 2.5% may mean an estimate sitting at the floor or a censored "below 2.5%"; the series does not say which. So indicator 2.1.1 cannot register a country arriving at the goal, and for the 62 economies recorded at 2.5% it has run out of resolution: it cannot separate a country near zero from one at 2.4%. Against a target of zero, the official measure can certify, at most, "at or below 2.5% everywhere" — never zero. A metric that stamps China with the same 2.5% as a Nordic state is not distinguishing them — by construction, it cannot.

The floor is defensible, and FAO says so plainly: below 2.5% the model's uncertainty swamps the signal, so reporting a precise sub-2.5% figure would be false precision. That is candor, not concealment. Read charitably, the floor is an honest admission that the last stretch to zero is unmeasurable with this method — which is also the point. And a floor value does not mean hunger is solved. SDG 2.1 carries a second indicator, 2.1.2, the survey-based prevalence of moderate-or-severe food insecurity. On that measure the same 2023 vintage puts the United States at 10.3% and Brazil at 13.5% — both at the 2.5% floor on undernourishment — with the world at 28.4%. PoU counts chronic caloric shortfall as a national-average model; food insecurity counts people reporting constrained access to food — from worrying about running out to going a whole day without eating — for lack of money or other resources. The two indicators for one target disagree by design, and the harsher one has no floor.

**View from the statisticians.** To a national statistics office at the floor, 2.5% is the correct entry — the responsible refusal to publish a number the data can't support. The complaint is not with FAO's method but with the mismatch between a political headline ("zero") and the tool chosen to score it: a gauge built to track hunger where it is widespread, asked to certify its absence where it is rare.