Global Food CrisisChapter 01

An open, evidence-based data story

The World Has Food. Why Does Hunger Persist?

The world grows more calories than its people need — yet hundreds of millions still go hungry. This is not primarily a story of scarcity. It is a story of access: of conflict, poverty, climate shocks and prices that stand between food and the people who need it.

645 million people

faced hunger in 2025 — 7.8% of the world's population

Three consecutive years of improvement, and still roughly one person in thirteen. Nearly 14 million fewer people than in 2024; 43 million fewer than at the 2022 peak.

SourceFAO, IFAD, UNICEF, WFP & WHO — The State of Food Security and Nutrition in the World 2026 (SOFI 2026) — reference year: 2025

3,016 kcal

per person, per day

average dietary energy available in the world's food supply in 2023 — more than enough, on paper, for everyone

SourceFAO Food Balance Sheets — reference year: 2023

Food exists. Access does not.

Dietary energy supply measures what reaches a country's markets and households, before waste and before the question of who can afford it. Averaged across the planet it has passed 3,000 kilocalories per person per day. Yet averages hide distribution, and availability is not affordability. A market can be full while a family's pockets are empty.

Hunger in the 2020s is rarely caused by the world running out of food. It is caused by war that blocks farms and aid, by incomes too small to buy what is on sale, by droughts and floods that erase harvests, and by price spikes that price the poorest out first. Those causes can be measured, understood and addressed — which is why this story exists.

This is Chapter 01 of an open project. It walks through four questions: how large hunger is, where it lives, why it persists, and whether the world is making progress — ending with what one person can realistically do. Every statistic on this page carries its source, its reference year and its caveats.

Chapter 2 — The Scale

How many people are affected — and what are we counting?

'Hunger' is one word hiding several different measurements. Confusing them leads to double counting and bad comparisons. These four indicators answer different questions, use different methods and cover different populations.

Hunger / undernourishment

Not enough calories, habitually, for an active and healthy life.

645 million

people, 2025 (7.8% of the world population)

Show methodology and limitations
Definition
The formal UN measure of 'hunger' is the prevalence of undernourishment (SDG indicator 2.1.1): the share of people whose habitual food consumption does not provide enough dietary energy for a normal, active, healthy life.
How it is measured
A modeled estimate. FAO combines national food supply data (Food Balance Sheets) with information on food distribution and population to estimate the distribution of dietary energy intake. Because it depends on annual supply statistics, the estimate lags by about a year and is revised each year as new data arrive.
Limitations
Counts dietary energy only — not the quality or diversity of diets. A country can show low undernourishment while micronutrient deficiencies ('hidden hunger') remain widespread. For many countries with very low prevalence, FAO simply reports 'below 2.5%'.

SourceSOFI 2026 (FAO et al.)

Moderate or severe food insecurity

People reporting limited, uncertain or inadequate access to food.

2.1 billion

people, 2025 (about one in four worldwide)

Show methodology and limitations
Definition
Measured with the Food Insecurity Experience Scale (SDG indicator 2.1.2), which asks people directly about their access to food over the previous 12 months — from worrying about running out of food (moderate), to actually going without eating for a whole day (severe).
How it is measured
Derived from survey interviews (via the Gallup World Poll and national surveys) in more than 150 countries, then scaled into national prevalence estimates using statistical models.
Limitations
Self-reported experience, so it captures dimensions that supply data miss — but it is a survey estimate with sampling uncertainty, and the 'severe' end overlaps, without being identical to, undernourishment. Roughly 86 million fewer people than in 2024.

SourceSOFI 2026 (FIES-based estimate)

Acute food insecurity (IPC/CH Phase 3+)

People in food crisis right now, in the places assessed.

266 million

people, 2025 — in 47 monitored food-crisis countries (22.9% of their analysed populations)

Show methodology and limitations
Definition
A different lens: not a person's chronic condition but a population's current crisis. Phase 3 (Crisis) means households face large food consumption gaps and sell assets to survive; Phase 4 (Emergency) means even larger gaps and acute malnutrition; Phase 5 is Catastrophe/Famine.
How it is measured
Consensus country assessments under the IPC / Cadre Harmonisé frameworks, compiled by 18 partner organizations. In 2025 this included 39.1 million people in Phase 4 (Emergency) and 1.4 million in Phase 5 (Catastrophe).
Limitations
Only food-crisis countries are assessed — so this number is NOT a world total and must never be added to the 645 million chronic measure. Coverage changes year to year: 47 countries had comparable data in 2025, the lowest in a decade, partly because the crises themselves disrupt data collection.

SourceGlobal Report on Food Crises 2026 (GNAFC)

Famine (IPC Phase 5)

The most severe classification — starvation and death already evident.

2

famine classifications in 2025 (the Sudan and the Gaza Strip) — a first for two areas in one year

Show methodology and limitations
Definition
Famine is a rigorous technical classification, not a metaphor. It requires evidence that at least 20% of households face extreme food consumption gaps, that acute malnutrition in children exceeds 30%, and that the crude death rate has passed 2 per 10,000 people per day — with starvation, death and destitution already visible.
How it is measured
A Famine Review Committee of independent experts must validate the evidence before Famine is declared. Since 2010, famine has been confirmed on six occasions: Somalia (2011), South Sudan (2017 and 2020), the Sudan (2024 and 2025) and the Gaza Strip (2025).
Limitations
The bar is deliberately high, which means recognition often comes late — and the places where evidence is hardest to gather (active war zones) are precisely where famine is most likely to occur.

SourceIPC Famine Facts

Why these numbers must never be added together. The 645 million, the 2.1 billion and the 266 million measure different things, in different places, with different methods — and real people are counted in more than one of them. The 2.1 billion already includes most of the 645 million; the 266 million is a subset of food-crisis countries only. Adding any of them together double-counts people and mixes incompatible definitions. Each number answers its own question — and each question matters.

Two different yardsticks

Chronic — how people live persistently

Undernourishment and FIES-based food insecurity track long-run conditions, for the whole world, every year.

Acute — how populations are hit by shocks

IPC/CH phases track current crisis severity, only in assessed food-crisis countries, and change with each shock.

Chapter 3 — The Geography

Hunger is not equally distributed

The global average of 7.8% conceals a world of extremes. Africa now carries more undernourished people than any other continent: about 309 million in 2025, surpassing Asia's 292 million for the first time. One continent's crisis is not another's trend.

Prevalence of undernourishment, share of population

Latest available country assessment — 2023 for most countries (FAO via Our World in Data)

Loading map…

<2.5%2.5–5%5–10%10–20%20–35%35%+No recent assessment

SourceFAO (SDG 2.1.1) · Our World in Data — reference year: 2023

The regional picture (2025)

Africa

309 million people — 20.0% of the population

Prevalence fell for the first time since 2017 (from 20.3% in 2024), but population growth keeps the absolute number near record levels.

Asia

292 million people — 6.7% of the population (2024 series value)

Asia drove most of the world's recent improvement, especially Southern Asia; its prevalence is now well under half of Africa's.

World

645 million people — 7.8%

SOFI 2026 (FAO et al.)

All countries with data — the accessible alternative to the map

Prevalence of undernourishment, share of population — Latest available country assessment — 2023 for most countries (FAO via Our World in Data)
CountryUndernourished (%)Year

Reading this map honestly

  • Map area is not people. A large country shaded lightly can hold more undernourished people than a small one shaded dark. The table above shows country values — the 2025 regional totals come from SOFI 2026.
  • Some countries are too small to draw at this map resolution (Barbados, Macao, Malta, Seychelles). They carry data and appear in the table, just not as visible shapes.
  • Grey means 'no recent assessment', not 'no hunger'. The Sudan's latest chronic estimate (2021) predates the 2023 war; its current crisis appears in the acute-food-insecurity figures of Chapters 2 and 4.
  • Values for most countries refer to 2023, the latest full FAO assessment round. The world and regional figures quoted elsewhere on this page refer to 2025 (SOFI 2026) — different vintages are labeled, never mixed.

Chapter 4 — The Causes

Why does hunger persist?

No single cause explains hunger. Households are pushed into food insecurity by overlapping shocks — war, weather, prices, poverty — sitting on structural weaknesses: thin markets, weak infrastructure, few safety nets. The Global Report on Food Crises attributes one primary driver per country; in reality, drivers compound each other.

What drove acute food insecurity in 2025?

Loading data…

One primary driver attributed per country — 147.4 + 87.5 + 29.8 million nearly spans the 266 million total without overlapping. In reality drivers co-occur: 31 countries reported economic shocks as an additional driver. Primary-driver populations must not be added to other indicators.

SourceGlobal Report on Food Crises 2026 (GNAFC) — reference year: 2025

Conflict and insecurity

147.4 million people — the primary driver in 19 countries

Conflict remained the largest single driver of acute food insecurity in 2025: 19 countries and territories, home to 147.4 million acutely food-insecure people — more than half the global total. Six of the ten largest food crises were conflict-driven, and conflict was the primary driver for almost every population in Catastrophe (IPC Phase 5).

War causes hunger without food disappearing: farms are abandoned, markets are cut off, currencies collapse, aid convoys are blocked, and people flee. 'Hunger is increasingly being used as a weapon of war,' the UN Secretary-General wrote in the foreword to the 2026 Global Report on Food Crises — which recorded, for the first time, outright famine emerging in two conflict-affected areas in the same year.

Case: Sudan — famine amid a forgotten war

Since civil war broke out in April 2023, Sudan has become one of the largest food crises in the world. By May 2025, more than half the population — around 25 million people — faced Crisis or worse (IPC Phase 3+). In September 2025 the IPC confirmed Famine in the besieged towns of El Fasher and Kadugli, with a risk of further spread across Darfur and Kordofan.

SourceGRFC 2026; IPC Famine Review Committee

Case: Gaza — famine confirmed, then pushed back

In August 2025, after 22 months of siege and displacement, the IPC confirmed Famine in the Gaza Governorate. A ceasefire and scaled-up aid allowed that classification to be pushed back by December 2025 — a reminder that famine is man-made and reversible, but also how quickly it returns when access is denied.

SourceIPC Global Initiative; WHO (Dec 2025)

※ Correlation is not causation: not every conflict produces famine, and the mechanism runs through measurable channels — displacement, destroyed supply routes, blockaded aid. Where those channels are absent, food usually still reaches markets. Where they operate, no amount of local production protects a population.

GRFC 2026 — full report (PDF) ↗IPC — Famine Facts ↗

Poverty and food affordability

About 2.7 billion people could not afford a healthy diet in 2025

The world does not merely need calories to exist — people need to be able to buy them. Nearly one in three people on Earth could not afford a healthy diet in 2025: the fruits, vegetables, dairy, eggs and protein that make up adequate nutrition, not just enough staple calories.

A healthy diet costs far more than its raw staples because costs accumulate after the farm gate: storage, cold chains, processing, transport and retail all add margin. Where infrastructure is weak — unreliable electricity, bad roads, fragmented markets — nutritious food becomes a luxury precisely where incomes are lowest. This is why SOFI 2026 is subtitled 'Understanding and addressing the high cost of a healthy diet'.

The poverty trap in one paragraph

Poor households spend half or more of their income on food, so any price rise or income shock bites them first. They respond by cutting the most expensive calories — the nutritious ones — long before they cut total energy. Adults go hungry so children can eat. This is why hunger and obesity coexist in the same households, and why poverty drives malnutrition even where markets are full.

SourceSOFI 2026; World Bank Food Prices for Nutrition

※ Affordability is measured against the cost of a healthy diet basket, an estimate that varies with local prices and methodology — it indicates constrained access, not a guaranteed outcome for every individual.

FAO — Healthy diets remain unaffordable for a third of the world ↗World Bank — Food Prices for Nutrition DataHub ↗

Climate extremes and environmental shocks

87.5 million people — the fastest-growing driver

Weather extremes were the primary driver of acute food insecurity in 16 countries in 2025, affecting 87.5 million people — a roughly five-fold increase from 15.7 million people in 2020. Much of the rise traces to the 2023–24 El Niño, one of the strongest on record, which devastated harvests across Southern Africa, and to La Niña-induced drought that hit the Horn of Africa at the end of 2025.

For farming households, a failed season is not one bad year: it eats the seed stock, forces livestock sales at distressed prices and borrows against the next harvest. Droughts and floods also degrade the land itself, so each shock lowers the starting point of the next one — a slow ratchet that turns weather variability into structural poverty.

Case: The Horn of Africa's five failed rainy seasons

Between 2020 and 2023, five consecutive rainy seasons failed across the Horn of Africa — the region's worst drought in more than four decades — leaving 31.9 million people in Ethiopia, Kenya and Somalia in need of assistance. When the rains finally returned in 2023, they arrived as floods, destroying what the drought had left.

SourceWMO; Concern Worldwide; UNDRR forensic analysis (2024)

Case: Syria's 2025 harvest failure

In 2025, Syria recorded one of its worst droughts in more than 30 years: wheat and barley harvests came in about 60% below average, deepening the food crisis of a country already shattered by more than a decade of war. Climate shocks rarely act alone — they land hardest on countries whose buffers are already gone.

SourceGRFC 2026

※ Single events cannot be attributed to climate change without formal attribution studies, and the GRFC classifies 'weather extremes', not 'climate change'. The trend, not any single drought, carries the climate signal — and it compounds, rather than replaces, the other drivers.

GRFC 2026 — Box 7, Weather extremes ↗UNDRR — Horn of Africa drought, forensic analysis ↗

Economic instability and food price shocks

Food prices transmit global shocks to kitchen tables

In 2025, economic shocks were the primary driver of acute food insecurity in 12 countries, home to 29.8 million people — down from their 2022 peak, but still reported as an additional driver in 31 countries. Currency depreciation, debt distress and high import bills price food out of reach long before any shortage appears.

World food prices have swung violently twice in a generation. The 2007–08 crisis put the FAO Food Price Index at a then-record 117.7 (annual average) and triggered unrest in dozens of countries. The 2022 peak — an annual average of 144.5, monthly 160.3 in March 2022, the highest since the index began in 1990 — followed Russia's invasion of Ukraine, a major grain and fertilizer exporter. Countries that import most of their food, and households that spend most of their income on it, absorb those shocks first and recover last.

FAO Food Price Index

Annual averages · index points, 2014–2016 average = 100

The FAO Food Price Index tracks the international prices of a basket of globally traded food commodities — cereals, vegetable oils, dairy, meat and sugar, weighted by their export shares. It is not a measure of supermarket prices or of what any household spends on food, but international price swings do reach local markets through trade, exchange rates and supply chains. The 2014–2016 average is set to 100, so an index of 125 sits about 25% above that baseline — not 25% higher than the previous year.

Loading data…

Each point is the average of that year’s twelve monthly values — so the 2022 annual value of 144.5 is the highest annual average of the modern series, while the monthly index peaked higher still at 160.3 in March 2022. After easing through 2023–24, the index rose again to 127.2 in 2025 and averaged 136.0 in September 2026.

SourceFAO — World Food Situation, Food Price Index — reference year: 2025

The quiet driver: shrinking aid

Just as needs peaked, support contracted. Net official development assistance from major donors is projected to fall to its lowest level since 2014, humanitarian aid is projected down by roughly 40% between 2024 and 2026, and food aid specifically by about 44.5%. Emergency food security monitoring is being cut in the same countries where hunger is growing — which is also why some recent declines in measured hunger partly reflect lost data, not only relief.

SourceGRFC 2026 Mid-Year Update; JRC (Sept 2026)

※ Correlation is not causation: price spikes coincide with wars, energy shocks and export bans, and countries absorb them differently depending on reserves, subsidies and social protection. The index measures world export prices, not what any household pays at its local market.

FAO — World Food Situation, Food Price Index ↗GRFC 2026 — Mid-Year Update ↗

Chapter 5 — The Trend

Is the world making progress?

The honest answer: progress happened, then reversed, then partially returned. The share of humanity that is undernourished fell steadily for fifteen years — from 12.7% in 2000 to about 7.5% by the mid-2010s — then rose again through the pandemic and the wars of the early 2020s. Since 2022 the curve has bent back down. Whether that holds depends on the causes in Chapter 4.

Global hunger over time

Loading data…

FAO revises these series every year. The line here is the FAOSTAT series as republished by Our World in Data (retrieved October 2026); the marked point is the newest SOFI 2026 estimate for 2025. Small differences between vintages are normal and are documented in the methodology notes. Europe's flat line: FAO reports any prevalence below 2.5% as “<2.5%” — the measurement floor — so the dashed European line marks that reporting floor, not an exact 2.5% every year. Number series: FAO via World Bank WDI (through 2023). The SOFI 2026 point marks the newest vintage: 645 million in 2025.

SourceFAO · Our World in Data — reference year: 2024

Percent or people — watch both

The percentage of the world that is hungry has fallen much further than the number of hungry people, because the world's population grew by about two billion over the same period. Africa illustrates the tension sharply: its prevalence fell in 2025 for the first time since 2017, yet population growth meant the continent still passed Asia in absolute numbers of hungry people. Percentage progress can coexist with record headcounts — both statements are true, and each is incomplete alone.

What moved the curve

  • 2000–2014 — the long improvement: economic growth in Asia and Latin America, expanded social protection and agricultural investment pushed prevalence down by more than a third.
  • 2015–2019 — the plateau: progress slowed to a halt; conflicts in the Sahel, Yemen and Syria began pushing numbers up again.
  • 2020–2022 — the reversal: the COVID-19 pandemic and its economic fallout, then the war in Ukraine's price shock, drove the share from 7.5% back to 8.6% at the 2022 peak.
  • 2023–2025 — three years of recovery: easing prices, post-pandemic normalization and targeted policy pushed the share down to 7.8% — still above the mid-2010s best. FAO projects about 520 million people may still face hunger in 2030, far off track for the SDG Zero Hunger target.

The acute view: food crises since 2016

People in food crisis (IPC/CH Phase 3+), 2016–2025

Loading data…

Acute food insecurity has roughly doubled in a decade — but read the coverage label under each bar: the 2026 edition assessed only 47 countries with comparable data, the lowest in a decade, and 18 countries lacked data entirely. The 2024→2025 decline partly reflects that lost coverage, not only real improvement. This series must never be compared, or added, to the chronic-hunger line above.

SourceGlobal Report on Food Crises 2026 (GNAFC) — reference year: 2025

⬇ Download the data (CSV)Contains both series shown above, with sources and reference years. Open it in any spreadsheet application.

Behind the chart: the numbers

Global hunger over time — * SOFI 2026: 7.8% (2025) · SOFI 2026: 645 million (2025)
YearWorld share (%)World number (millions)

Chapter 6 — The Action

What can one person do?

Global hunger is not solved by a single gesture — but it is not beyond individual action either. The evidence in this chapter supports realistic, non-magical choices: learn how the numbers actually work, give effectively, volunteer locally, waste less, and push for the policies the data says matter. Pick what fits your life.

What do you want to do?

Choose an interest — recommendations adapt below.

Most misinformation about hunger comes from mixing up indicators — quoting the acute-crisis number as if it were the world total, or summing measures that overlap. An hour with the primary sources makes you immune to most of it, and better at explaining it to others.

Reading is not the same as doing — but informed people are the transmission system every other action needs.

One honest sentence

645 million people facing hunger is neither fate nor fiction — it is the measurable result of choices about war, markets, infrastructure and care. Three years of improvement prove it can move. What happens in the fourth year depends on decisions not yet made — including, in small but real ways, yours.