US federal research spending remains at about 200 billion dollars, but the National Institutes of Health (NIH) has lost a fifth of its staff, the National Science Foundation a third, and defense laboratories 14%, a New York Times investigation finds

US federal research spending remains at about 200 billion dollars, but the National Institutes of Health (NIH) has lost a fifth of its staff, the National Science Foundation a third, and defense laboratories 14%, a New York Times investigation finds

The New York Times published the first article in its “Science Under Siege” series on October 31. The White House disputes its account, saying that Trump is modernizing scientific institutions and that most scientists who left federal service did so voluntarily.

A Pentagon report released over the summer says that hiring freezes and federal workforce cuts cost Army, Navy and Air Force laboratories about 14% of their staff in fiscal year 2025. Laboratory directors had no control over who left. Replacing those employees has proved difficult: staffing caps and new hiring committees have slowed recruitment. Unless these policies are reversed promptly and for the long term, the report warns, “the engine of innovation risks stalling.”

The newspaper attributes the losses to three factors. The Department of Government Efficiency (DOGE), led by Elon Musk, fired employees and canceled grants. The White House used funding cuts to pressure universities. The third factor was a campaign by Office of Management and Budget Director Russell Vought to assert political control over funds appropriated by Congress. Vought requested a 41% cut to NIH research funding and a 55% cut to the National Science Foundation, but Congress rejected those cuts, leaving spending largely unchanged. In May, Vought’s office and 41 agencies proposed a rule under which senior agency appointees would review grants before they were awarded, exercising their own judgment, while expert reviews would remain advisory. A temporary funding law prohibits finalizing the rule through December 11.

NIH received 1% more funding in fiscal year 2026 and spent its entire grant budget, but awarded about 56 600 grants, compared with 64 000 in 2024. By the summer of 2025, it had fallen behind in awarding grants, and Vought’s office instructed it to allocate half the remaining funds to projects funded upfront for their full duration. This spreads the budget across fewer projects. The Association of American Medical Colleges considers this a partial explanation for the decline: the share of these projects among competitively awarded research grants rose from 10% in 2024 to 25% in 2025. The number of competitive grants awarded fell by 19% in 2025 compared with 2024, and by 2% in 2026, but the total grant count remained lower. According to R&D World’s count, that total includes annual payments for ongoing projects, which have become less numerous; projects funded upfront receive a single payment.

Ann Cohen of the University of Pittsburgh studied how earlier stress and health problems affect the risk of Alzheimer’s disease. In August, NIH terminated her grant on the grounds that the project focused on structural racism, according to the newspaper. Cohen says that rationale reflects a misunderstanding of her work.

Basic research often takes decades to bear fruit, the newspaper writes: a 1976 discovery about the origins of cancer-causing genes earned a Nobel Prize and led to cancer drugs two decades later. “Science is a long game,” says Marcia McNutt, who led the US National Academy of Sciences until this summer. “By the time we stop winning Nobel Prizes, we will already have lost the ability ever to recover.” Former White House official May Mailman, who helped lead the pressure campaign against universities, offers a different assessment: “There is a sense that we are spending more on research than we are getting out of it, that the system is a little broken.”