Climate Change Could Push 1.36 Billion More People Into Hunger by 2050

Somewhere between now and 2050, if global emissions stay on their current track, the world could find itself trying to feed billions more people with significantly less food than it produces today. That's not a doomsday scenario invented by activists — it's the central finding of a peer-reviewed study published in Scientific Reports by economists at the University of Melbourne and the Australian National University, who used one of the most detailed food-and-climate models ever built to run the numbers on what heat and water stress will do to global food production by mid-century.

The Number That Stops You in Your Tracks

Start with the headline figure. Under the worst-case emissions scenario tested — a high-population-growth pathway paired with aggressive warming — an additional 1.36 billion people could face severe food insecurity by 2050 compared to 2020. To put that in perspective: that's roughly the entire current population of India added to the global hunger count. Even the most optimistic scenario the researchers modeled, which assumed meaningful climate mitigation, still produces 556 million newly food-insecure people. Not a rounding error. Not an asterisk. More than the current combined population of the United States and Europe. This study, led by Professor Tom Kompas and co-authored by Tuong Nhu Che and R. Quentin Grafton, covered 141 countries and 30 commodity sectors, making it among the most comprehensive analyses of its kind.

Why Earlier Climate Models Kept Getting It Wrong

Most of the big integrated assessment models used in climate research are what's called "recursive" — they simulate the future one step at a time, using the outcome of one period as the starting point for the next. That sounds reasonable, but it means farmers, governments, and markets in the model are essentially blind to what's coming. Real economic actors aren't. They plan ahead. They invest or don't invest based on expectations about the future. The GTAP-DynW model, by contrast, is forward-looking. Producers in the model can see what's coming across all time steps simultaneously and optimize accordingly. It also directly embeds climate damages — including heat stress on agricultural labor productivity, something most models quietly skip — into the economic structure rather than bolting them on afterward. That transparency matters if you actually want to understand what's driving the results.

How Do Heat and Water Stress Actually Destroy Crops?

The two stressors work differently, and the study tracks both. Water stress is calculated basin by basin — all 15,006 of them globally — using WRI's Aqueduct projections. As water withdrawals rise relative to available freshwater, the land becomes less efficient. Less water means less output, especially for irrigated agriculture, which currently supplies up to 40% of the calories humans consume. Heat stress operates through a different channel: it directly damages labor productivity in the agricultural sector, something that gets worse as temperatures climb. In the model, each additional degree of warming degrades both the physical yield of crops and the ability of agricultural workers to actually harvest them. In China, the combined effect under the worst scenario is a projected 22.4% drop in food production by 2050. For India, 16.1%. For the United States, 12.6%.

Who Gets Hit the Hardest — and Why That's Not a Surprise

Africa. That's the short answer, and it's not especially surprising given what the continent faces on multiple fronts simultaneously. Agricultural output there is already vulnerable to heat and drought. Population is projected to grow substantially by 2050. And unlike net food exporters — Australia, France, Russia, the United States — African nations can't simply offset domestic production losses by pulling from trade surpluses. In the worst scenario, domestic food production in many African countries would cover less than half of local food demand. The Middle East, South Asia, and Central America follow closely behind. There's also an interesting trade dynamic buried in the results: as water-stressed regions produce less food, agricultural commodity flows increasingly move from low-stress to high-stress countries — and food exports into China grow across all three scenarios, a sign of how water scarcity reshapes global trade geography even before political pressures enter the picture.

What the Model Still Can't Tell Us

The authors are candid about the gaps. The biggest one: the model only accounts for water stress on irrigated cropland. Rainfed agriculture — which covers enormous areas of sub-Saharan Africa and South Asia — isn't yet incorporated into the water stress component, because the necessary basin-level data for that interaction isn't available in the GTAP framework. So the projections are almost certainly conservative. Climate change also reshapes rainfall patterns, altering how much "green water" (rainfall captured in soil) is available to rainfed crops — and both blue and green water resources are under increasing threat, particularly in South and East Asia. What comes next for the research team is filling in those gaps. A version of this model that captures rainfed cropland water stress too would likely produce numbers that are harder to sit with.

• Irrigation is both solution and problem. — Irrigated agriculture reduces water stress risks compared to rainfed land, but expanding irrigation in already water-stressed regions accelerates groundwater depletion and makes the long-term situation worse.
• Trade won't save everyone. — Wealthier nations can offset domestic food losses by importing more, but poorer net-importing countries, especially in Africa, have no such buffer, which is why severe food insecurity concentrates there.
• The projections are probably too low. — Because the model currently excludes water stress impacts on rainfed cropland, the actual number of additionally food-insecure people by 2050 could exceed even the 1.36 billion figure given here.

"Substantial declines in global food production of some 6%, 10%, and 14% to 2050 — and the number of additional people with severe food insecurity correspondingly increases by 556 million, 935 million, and 1.36 billion compared to the 2020 model baseline." — Kompas, Che & Grafton, Scientific Reports, 2024.

Advertisement