Forecast Series: The Mortality Impacts of USAID Cuts
A special series of five forecasts on the Trump Administration's cuts to some of USAID's most effective programs.
As of this writing, USAID, the world’s largest foreign assistance agency, appears to be effectively dead.
This does not mean the end of American foreign assistance entirely. Indeed, the Trump Administration’s FY2026 budget request still proposes $3.8 billion in global health spending — a greater than 60% reduction from years previous. But we’ve reached what is unquestionably the end of a modern era of international development. It remains to be seen in what form, under which department, and to what causes new foreign assistance will take.
In the interim, the loss of USAID programming is already being felt, from the closure of HIV centers in South Africa to the cessation of anti-malaria programs in Sierra Leone to the shuttering of nutrition clinics in Bangladesh. The impacts will continue to be felt over the coming years. We wanted to know: How big will they be?
For this special feature, we commissioned forecasts on increased mortality as the result of USAID disruptions across five of its programmatic areas:
We first assigned these forecasts during a period in which it appeared that foreign aid might be reduced to effectively zero. After substantial outcry, some funding has been reinstated. Charles Kenny and Justin Sandefur,1 of the Center for Global Development, have compiled estimates of total USAID cuts by sector and by country. Some areas, such as infrastructure and civil society, appear likely to be hollowed out entirely. Others, like maternal health and agriculture, may continue at only a tenth of their previous funding. The most critical programming, at least in terms of immediate mortality impacts, such as PEPFAR, malaria, and nutrition assistance, appear likely to persist, albeit in a heavily reduced form — around 70% of its previous funding for HIV/AIDS, 50% for TB, and 40% for nutrition.
Our goal in this series is, first and foremost, to understand the possible human impact of USAID cuts. But we also want to be transparent in the process by which we’ve arrived at these estimates. Historically, foreign aid has been among the least popular areas of government spending, but specific areas — food security, poverty relief, and healthcare — remain widely supported. We’ve published critiques of USAID programs in the past, but we believe the life-saving impacts, especially of its most cost-effective programs, remain under-appreciated. We think quantifying those impacts — transparently and probabilistically — is critical to the conversations happening now and in the future over whether these programs should be restored.
We estimate that the combined impact of the projected cuts to these five programs will result in between 483,000 and 1.14 million excess deaths over one year.
If cuts were to persist over five years, we may expect those numbers to increase to between 1.48 million and 6.24 million deaths.2 For specific details, see links to each forecast below.
Jared Leibowich estimates that PEPFAR cuts will lead to between 257,000 and 772,000 deaths in the next year, and between 1.01 million and 2.95 million deaths over the next four years.
Bruce Tsai estimates that cuts to malaria programming will result in between 43,000 and 90,000 deaths in the next year, and 250,000 to 410,000 deaths over the next five years.
Kati Conen estimates cuts to TB programming will result in 98,000 to 184,000 deaths in the next year, and 900,000 to 2.5 million deaths over five years.
Lauren Gilbert and Anna Gordon estimate cuts to nutrition programming will result in 79,000 to 89,000 deaths from severe acute malnutrition in the next year, and between 300,000 and 338,000 deaths over five years.
Erin Braid estimates cuts to WASH programming will result in 5,850 deaths in the next year, and 26,930 deaths over five years.
Also an Asterisk contributor, see PEPFAR and the Costs of Cost-Benefit Analysis.
Our forecasters each used slightly different methodologies and assumptions. Take these numbers as a high-level topline, and refer to specific forecasts for details.
Help I'm dumb and can't read. Does "We first assigned these forecasts during a period in which it appeared that foreign aid might be reduced to effectively zero." mean:
(A) Assuming PEPFAR cut to 0, there will be 257k to 772k deaths next year. In real life, it will be less, because PEPFAR has been partially reinstated.
(B) Taking into account that PEPFAR has partially be reinstated, there will be 257k to 772k deaths next year. If PEPFAR had stayed completely cut, it would be more.
Thank you for doing this work! Personally, I'd love to see these articles complemented by something like a Guesstimate visualization (https://www.getguesstimate.com/), when possible. I find that a more intuitive way to understand probabilistic predictions than longform writing.