Key Takeaways
A peer-reviewed Lancet study projects more than 14 million additional deaths by 2030 from USAID cuts, including about 4.5 million children under 5.
The estimate is highly sensitive to assumptions, and other models put the toll far lower, from several hundred thousand to a few million.
All the figures are modeled projections, not a verified death count, which is the core of the dispute.
A contested death-toll estimate sits at the center of a viral fight over the USAID cuts. Representative Ro Khanna has cited a peer-reviewed study to argue the cuts could kill millions, while Elon Musk, who led the DOGE review that gutted the agency, calls the claim false and has threatened to sue. What the study actually says is more careful than either side's slogans.
The study, published in The Lancet in July 2025 by a team at the Barcelona Institute for Global Health, evaluated two decades of USAID programs across 133 countries. It forecast that steep cuts could cause more than 14 million additional deaths by 2030, including about 4.5 million children under age 5, per France 24 and UCLA.
That forecast is a projection, not a body count, and it rests on assumptions. The model compares a world where aid keeps flowing with one where the cuts hold through 2030, and it depends on how fast programs collapse and whether other donors fill the gap. Change those inputs and the number moves. A February 2026 follow-up from the same group projected 9.4 million deaths rather than 14 million, per CNN, and the Center for Global Development estimated roughly 500,000 to 1.6 million lives lost using different methods. No one has a registry of deaths caused by the cuts, so every figure is a modeled estimate with wide uncertainty.
Critics of the high-end number say it assumes the cuts stay frozen and discounts a State Department waiver for lifesaving aid and programs Congress later refunded. Defenders counter that those fixes have been partial and slow, and that broken supply chains harm people even when funding returns.
The deeper problem is one WYDE's coverage keeps hitting. Aid is hard to measure, and when nobody can point to a verified count of who was helped or harmed, the gap fills with politics, the same vacuum visible in the fight over whether voluntary giving reaches people and in the 59% collapse in global hunger funding the UN flagged this month. The most notable part is not the death toll. It is that the world's largest aid program ran for two decades without a measurement system anyone fully trusts.
People Also Ask
What does the Lancet study on USAID cuts say?
It projects more than 14 million additional deaths by 2030 from USAID cuts, including about 4.5 million children under 5, based on two decades of program data across 133 countries.
Why do estimates of USAID-cut deaths vary so much?
The figures are modeled projections that depend on how fast programs collapse and whether other funders fill the gap. Different assumptions produce estimates from several hundred thousand to more than 14 million.
Did 4.5 million children die from the USAID cuts?
No. The 4.5 million figure is a projection of children under 5 who could die by 2030 if the cuts persist, not a count of deaths that have occurred.
What are critics and defenders of the projection disputing?
Whether the high-end number overstates the toll by ignoring waivers and refunded programs, or whether those partial fixes still leave millions at risk.
Sources
The Lancet, France 24, UCLA, CNN, Center for Global Development.