U.S. science is in chaos

Last year Christopher Reynolds started to worry that his space telescope was going to be killed.

The mission had started taking shape nine years earlier, a billion-dollar orbiting observatory that would look back in time into the early universe to study the first black holes, the formation of galaxies, and more. Eight teams of researchers pitched NASA their ideas; Reynolds, an astronomer at the University of Maryland, was part of a group that wanted to deploy a new technology: x-ray mirrors made of single-crystal silicon. It sounded promising enough that in October 2024 Reynolds’s group got a $5-million grant from the agency to refine the idea—the Advanced X-ray Imaging Satellite, or AXIS. The scientists teamed up with spacecraft builders at the nasa Goddard Space Flight Center. “Everything seemed to be going pretty well,” Reynolds says. “And then we started to get hit by programmatic chaos.”

Last June the budget hawks in the Department of Government Efficiency (DOGE) pushed NASA into offering a broad package of buyouts, paid leave and early retirement. Over the next few weeks nearly 4,000 NASA employees—about a fifth of the workforce—took the deal. Reynolds’s AXIS team lost 20 people. The engineer designing the heaters to keep the x-ray mirror at a constant temperature: gone. The lead project manager: gone. William Zhang, the astrophysicist who invented the telescope’s mirror technology: gone. “We were literally left with their PowerPoints, trying to figure out what they’d done and where we were with aspects of the design,” Reynolds says.

On supporting science journalism

If you're enjoying this article, consider supporting our award-winning journalism by subscribing. By purchasing a subscription you are helping to ensure the future of impactful stories about the discoveries and ideas shaping our world today.

Around the same time President Donald Trump’s budget proposal came out—with massive cuts to science funding. In the U.S., private money funds vast amounts of scientific development research, and philanthropy contributes a bit, but something like 40 percent of all the funding for basic, blue-sky, exploratory research comes from the federal government. The program that would have funded AXIS was zeroed out entirely.

That was just the request, Reynolds figured at the time; Congress still has to do the actual appropriation. “In any normal year, that’s what would have happened,” he says. “But the center leadership started quite quickly aligning their priorities to the president’s budget request.”

Goddard reassigned engineers to projects that would be funded if Congress approved the budget as written. Reynolds’s team lost its systems engineers, which in turn delayed sharing of AXIS’s proposed design with Goddard’s cost analysts and schedule specialists. “We got our very first cost estimate in the middle of September 2025,” Reynolds says. “We were 10 percent over budget.” He started trying to find things to cut. But then, in October, the federal government shut down. “The whole center just stopped,” he says. “Everything stopped.”

When the shutdown ended in mid-November, Reynolds’s team had just two weeks to get on budget. It failed. The plan the group submitted would cost too much and take too long. “Our last hope was that NASA headquarters would understand what had gone on and give us some leeway,” Reynolds says. NASA did not. After nearly 10 years of work, AXIS was dead.

Now, Reynolds says, he’s fine, mostly. He’s a tenured professor and has other research to work on. “The jobs that are lost are the future jobs,” he says. “And there’s an entire field of study in which U.S. leadership is at stake.” The hardest part, though, is how it happened. DOGE’s cuts sliced through American research grants like a thresher, “but this was much murkier,” Reynolds says. “We were never canceled. We were just starved to death.”

Countless scientists around the country are going through the same thing. Thousands of federal grants have been frozen or canceled, with perhaps 2,600 still in limbo—about $1.4 billion worth. The National Science Foundation and the National Institutes of Health are awarding three quarters of their usual number of grants. Fewer people are entering graduate programs. Nearly 95,000 scientists have left federal government employment. The NIH used to issue as many as 850 “Notices of Funding Opportunity” every year—requests for proposals that sought specific kinds of research. In 2025 the agency issued 120. By mid-March of 2026, the NIH had sent 14.

What’s going on is nothing short of a generational change in how the U.S. organizes its scientific enterprise. More than that, science feels different. Its purpose, its existential vibe, seems to have shifted. The cultural status of the people who do it has changed. And they don’t understand why.

The prevailing emotions among scientists right now are rage and shock. A survey conducted by science news website STAT found that more than half of researchers with grants from the NIH—once a reliable source of $40 billion a year—reported some level of disruption to their funding: a total freeze, a delay in disbursement or a reduction in amount. And 81 percent of researchers in tenure-track positions said they were concerned that funding disruptions could affect their productivity enough to jeopardize their chances of getting tenure.

Now, to be sure, the end product of science is supposed to be science, not grants or tenure. Applying for highly competitive grants with limited funding is what scientists have always had to do to carry out the science—a flawed process with few alternatives. But arbitrary cancellations and delayed disbursements are unprecedented. And justifying them on the basis of politics—prohibiting, for instance, grants that include language referencing diversity, equity and inclusion (DEI)—was unheard of until now.

When Jenna Norton, a program director at the National Institute of Diabetes and Digestive and Kidney Diseases (NIDDKD), first got to the NIH 12 years ago, she wanted to increase research into the social determinants of health—structural racism in home-loan practices meant that nonwhite people got iced out of home ownership and generational wealth, which forced them to live in neighborhoods closer to toxic sites such as factories and highways, without sidewalks and amenities. “It’s a challenging field to quantify, but we’re getting to a place in science where we can start asking these questions,” Norton says. Now the topic is verboten in U.S. grants. “That whole line of research has been shut off and censored because some people find the words ‘structural racism’ offensive.”

Political operatives at the NIH passed around lists of words that grants weren’t allowed to use—in either applications or existing, funded projects. Program managers across the NIH and the NSF were told to ask affected researchers whether they’d care to change the language in their research descriptions or risk losing their funding. Some researchers whose grants Norton managed at the NIDDKD called her to say they wanted to preemptively change the language in their grant applications—before they’d been dinged. Norton complained so much that she was placed on administrative leave, although she has since been reinstated.

Of course, not all lost science had obvious political implications. As Reynolds, the AXIS lead researcher, puts it, “whether there are black holes at a redshift of 10 or not is not a partisan issue.”

These kinds of obstacles are a new experience for most researchers. Getting into a career in science was already hard—students often undertake intellectually taxing and physically grueling academic work lasting years longer than most people spend in school, with limited remuneration. The people who do it tend to be mission-driven: they want to help others, learn something about the universe or invent something new. If they consider the political implications, it’s because they’re intrinsic to the work. “It’s not just that people feel their career is under attack,” says one longtime public health researcher. “They feel they personally are under attack.”

DEI associations aren’t the only topics that get captured by the new political filters. Now, for the first time, grant recipients aren’t allowed to subcontract to collaborators on projects overseas. “That’s obviously a problem when you study nasty diseases such as Lassa fever and Ebola, because they’re not in this country,” says Kristian Andersen, an evolutionary biologist at Scripps Research in La Jolla, Calif. “That’s my whole career. This is why I came to the United States.”

Most years, when Andersen advertises a postdoctoral research opportunity in his laboratory, he gets up to 200 applicants with perhaps a third of them from Europe. This year he had 100 applicants and none from Europe. Typically his lab would apply for two or three so-called center grants every year. This past year there were none in virology, immunology or viral immunology to apply for. So what’s next? Andersen, who’s Danish, says that “for people like myself, I think the best option is probably to leave and do science elsewhere.” And he isn’t the only one thinking of getting out. Of about 1,650 scientists who responded to a poll by the journal Nature, 75 percent said they were considering it.

“The most passionate and creative scientists are very intuitive and very driven by emotion and curiosity,” says Gregory Feist, a psychologist at San José State University who studies scientists. “Until Trump, they’d been able to keep political questions out of mind.” Their work was, if not above politics, at least outside it—essential to everyone, regardless of where they were on the political spectrum.

Now they see things differently. “The big eye-opener for me this past year is how quickly things can change,” a NASA climate scientist says. This shock at the ease with which the government can rewrite the system came up in multiple interviews. “Is your grant going to be frozen? Is it going to be terminated? Is it going to be reinstated? Is it going to be delayed because you’re required to change the wording?” asks Scott Delaney, a former Harvard University epidemiologist who co-created the watchdog group Grant Witness. “The reality is, because of what happened and what’s happening now, the trust between researchers and the federal government is completely broken.”

Without that trust, the entire system could blow apart. “Laboratories are going to close. Trainees are going to go to other countries or pursue nonscience careers,” says Carole LaBonne, a developmental biologist at Northwestern University. “This compact that has existed since World War II, that made the U.S. the successful, prosperous nation that it is, is being dismantled.”

What broke the compact? Several researchers identified the response to the COVID pandemic as a flash point. Public health guidance flailed initially on questions of masking, school closures and frontline drugs. It also produced a good vaccine in under a year, an unheard-of success. Ultimately around a million people died of the disease within the first two years.

The experience damaged trust in science and scientists. It’s still high—the number of people saying they have a lot of trust in science has hovered around 77 percent for years. But it was 10 points higher before COVID, and it now splits hard along lines of political affiliation. “Especially in the U.S. and with social media, all of a sudden everybody was an expert on COVID. So much of it was just bullshit,” Andersen says. “And then at some point bullshit was all that was left.”

That helps to explain how a nonscientist such as Robert F. Kennedy, Jr., known for unorthodox and unproven ideas about health and medicine, became leader of the Department of Health and Human Services with oversight of the NIH. But it doesn’t explain how Elon Musk, an industrialist and the richest human to ever live, got the power to excise so much of the country’s research. It doesn’t explain why the former conservative think tanker Russell Vought could use control of the wonkish Office of Management and Budget to zero out research funding.

“I would like to see more people speaking up, but the fact is, mostly people don’t.” —Kristian Andersen Scripps Research

There’s a strain of antipathy to universities and academic truth-seeking in far-right conservatism, certainly. But other than burn-it-all-down nihilism or anti-intellectualism, why nuke the social contract between government and science? One possibility is that the deal was already dying.

In the first half of the 20th century, businesspeople, policymakers and scientists trying to figure out how airy academic research got turned into useful stuff came up with what’s now called the linear model of innovation, a theoretical (and contested) sequence that went from funding to basic research to applied research to the development of a technology or product. The best-known codification of the model came toward the end of World War II in a report called Science: The Endless Frontier, by Vannevar Bush, an engineer who had headed the wartime Office of Scientific Research and Development. Bush understood that applied science had won the war for the Allies—not only the atomic bomb but also radar, penicillin, food preservation, cryptography, and so on. Nerds saved freedom’s bacon, but Bush and others had had a hell of a time getting that nascent scientific potential onto the battlefield. So Bush proposed putting all of U.S. science on retainer.

Basic research, Bush wrote, was “performed without thought of practical ends” and “creates the fund from which the practical applications of knowledge must be drawn.” So he proposed a vast expansion of the state’s capacity to do science, via funding managed by agencies such as the NSF and the NIH. The government would give tax dollars to scientists so they could cast around in the dark doing basic research. Irregularly, some of that work would lead to new drugs or communications satellites or optimized food crops. Not every dollar of government support for science would result in a blockbuster drug or a billion-dollar technology, but a majority of blockbuster drugs and billion-dollar technologies would derive from government support. So the government promised to fund a lot. And in return, the scientists promised to jump through the government’s hoops and respond to an occasional Bat-Signal. “That’s the handshake between science and the market,” says Benjamin Jones, an economist at Northwestern, who studies innovation.

It sounds like a business- and defense-minded strategy. But as innovation researcher Benoît Godin points out, even though Bush agreed with business interests about the fact that research and the training of scientists led to industrial progress, his rationale was explicitly social. “Without scientific progress the national health would deteriorate; without scientific progress we could not hope for improvement in our standard of living or for an increased number of jobs for our citizens; and without scientific progress we could not have maintained our liberties against tyranny,” Bush wrote.

In fact, by the 1960s military and industrial interests had mostly lost patience with the ivory-tower exploratory side of the equation. The leaders of American capital and finance certainly wanted to goose scientific and technical innovation, but they thought the real problem was where the money went and how much was available. Banks didn’t want to risk loans to iffy tech start-ups with no collateral. But a special kind of investor—a venture investor—would bring high-risk dollars to research in return for partial ownership of the company doing it.

That approach seemed to stall out, too. In 1977 William Casey, future director of the Central Intelligence Agency, wrote a report for the U.S. Small Business Administration arguing that it was because venture capital didn’t have access to enough money. His new model for innovation, says M. R. Sauter, a historian of technology at the University of Maryland, brought to the center not basic research or even applied engineering but, simply, money—and the investors who had it. Casey’s report recommended changing the regulations in the Employee Retirement Income Security Act of 1974 so that institutional capital, like retirement funds, could enter the riskier venture game. In 1979 Congress did just that.

And in 1980 Congress passed the Bayh-Dole Act, moving ownership of the results of government-funded university research from the government to the universities. Now a blockbuster new drug or search algorithm could be a windfall for a university, and university administrations had common cause with venture investors. More basic discoveries started getting turned into dollars. But the alliance shifted the emphasis from state capacity to financial outcomes.

Today the most influential private-sector developers of technology are in Silicon Valley, and their perspective on innovation is that it should move fast, disrupt markets and make money. That perspective is influencing government financing of science more than ever before. “Right now the [Trump] administration is very destructive and is changing its mind all the time. It has this dimmer view of science and also sort of wants to win in technology,” says Jones, the Northwestern economist. “That is fueled somewhat by the disruptive orientation of successful people in Silicon Valley who are having an influence.”

“I think that perspective is flat-out wrong,” Jones adds.

For most of this century pretty much every metric of scientific productivity—new results, new discoveries and new inventions—has appeared to be down. This idea is controversial, and the data are difficult to measure, but that’s academic because this nominal downturn opened the institutions of science to criticism that it was scientists who were failing to honor the bargain. Maybe it’s no surprise that the whole thing has turned into what Arizona State University sociologist Edward Hackett calls “academic capitalism.” Today’s investors and policymakers think all research should be economically relevant and assist in the accumulation of capital. A “knowledge-based economy,” says Lancaster University sociologist Bob Jessop, wants all scientists to be entrepreneurs. Which all sounds familiar.

This view might be why the newly reconstituted President’s Council of Advisors on Science and Technology includes just one scientist, a physicist. The other 12 members are Silicon Valley luminaries such as venture capitalist Marc Andreessen and Jensen Huang, CEO of computer chipmaker Nvidia. And in March, Trump nominated venture capital investor Jim O’Neill as director of the NSF. Companies that work on artificial intelligence, the hot tech of the moment, tout the ability of their products to take over the labor of doing science, from analyzing data to formulating hypotheses. “GPT-5.2 is kind of already intelligent enough to be a soft collaborator in many scientific inquiries,” says Sébastien Bubeck, a computer scientist at OpenAI.

That’s not the world scientists want, but it’s the one they’ve got. The problem is, subjecting science to political taste tests and a more commercial mindset almost certainly means fewer world-changing results. No one can ever know when noodling around with Gila monster saliva will yield anti-obesity GLP-1 drugs. And putting politicos atop the pyramid of grant evaluations, scientists say, will be a disaster. Researchers who manage to get grants to study health outcomes on the condition that they ignore the effects of variables such as socioeconomic status, gender and ethnicity won’t even be able to publish their findings, because peer reviewers, an NSF director says, “are not going to suddenly indulge this fantasy.” They’re going to demand that studies factor in relevant variables.

Last year a team of economists imagined what this new future might look like by creating an alternative past. In 2025 the NIH cut the amount of grant money awarded by more than 40 percent compared with years prior. What if, the team members asked, the NIH research budget had been 40 percent smaller for the past few decades? Grants in the bottom 40 percent of the priority queue, they reasoned, wouldn’t have been funded. The team tracked those grants to their outcomes—research that never happened in this parallel universe—and found that something like half of all drugs simply wouldn’t exist today. The lost therapies include imatinib, the first real treatment for chronic myeloid leukemia, and the lung cancer drug erlotinib.

What are scientists supposed to do about all this? “I would like to see more people speaking up, but the fact is, mostly people don’t,” Andersen says. “They don’t want to be a target of the federal government. Having been in that, and still being in that, [I can say] it’s not very pleasant.”

Like many other scientists, Andersen expresses disappointment in what he sees as a failure of the institutions of science—the national academies, the American Association for the Advancement of Science, universities—to mount a louder opposition. “We have seen none of that, especially from the academies,” he says.

Some scientists try to just keep their heads down and keep working. Others know they can’t. “In public health, we have a proud history of organizing, right? We were campaigners,” says Gregg Gonsalves, an epidemiologist and policy professor at Yale University. By the 21st century that had changed. “We were told it was not important, that what mattered was the number of grants and publications you had. ‘Forget all the social and political things; those are incidental.’ Turns out they were not. They’re core to it.”

Gonsalves, who was involved in the fight to care for people with HIV and AIDS in the 1980s, says that scientists now have another job: “bearing witness and putting evidence on the table. It may not be persuasive to Russell Vought or Marco Rubio, but it is for the dossier, for the truth and reconciliation commission, for the Nuremberg trials that come after,” he says. “Keep the receipts. Write down what you see. Tell them what they did. We’re very good at documenting how X leads to Y.”

That’s the thing about generational shifts. There’s always a next generation after this one.