The Erasing of American Science

For decades in the United States, scientists and government officials have coexisted in a mostly peaceable and productive symbiosis. The government has funded science and then largely left well enough alone. Scientific agencies have been staffed by scientists; scientists have set scientific priorities; scientists have ensured the integrity of the science that is done, on the theory that scientists know their own complicated, technical, sometimes arcane work best. Under that system, science has flourished, turning the government’s investment into technological innovation and economic growth. Every dollar invested in research and development has been estimated to return at least $5 on average—billions annually.

Recently, that “we pay; you do” mutualism has grown shakier and, since January, fractured into all-out antagonism. In less than a month, the Trump administration has frozen research funds, halted health communications and publications, vanished decades of health and behavior data from its websites, terminated federally funded studies, and prompted researchers to scrub extensive lists of terms from manuscripts and grant proposals. Those changes are, by official accounts, in compliance with Donald Trump’s recent executive orders, which are intended to derail “wasteful” DEI programs and purge any references to “gender ideology” from content funded or published by the federal government.

The administration’s actions have also affected scientific pursuits in ways that go beyond those orders. The dismantling of USAID has halted clinical trials abroad, leaving participants with experimental drugs and devices still in their bodies. Last week, NIH announced that it would slash the amount its grants would pay for administrative costs—a move that has since been blocked by a federal judge but that would substantially hamper entire institutions from carrying out the day-to-day activities of research. The administration is reportedly planning to cut the budget for the National Science Foundation. Mass layoffs of federal workers have also begun, and two NIH scientists (who asked not to be identified for fear of professional repercussions) told me they participated in a meeting this morning in which it was announced that thousands of staff across the Department of Health and Human Services would be let go starting today. Robert F. Kennedy Jr. has now become the head of that department, after two confirmation hearings in which he showed a lack of basic understanding of the U.S. health system and a flagrant disregard for data that support the safety and effectiveness of various lifesaving vaccines. (The White House did not return repeated requests for comment.)

The mood among scientists is, naturally, grim. “There is so little clarity about what to do,” Elizabeth Wrigley-Field, a sociologist at the University of Minnesota, told me. Many researchers I spoke with feel immobilized by the recent changes. “Everyone is in a panic right now,” Mati Hlatshwayo Davis, the director of health for the city of St. Louis, Missouri, told me. “And when researchers don't know what they’re allowed to do, science is not going to get done.”

Science can be polarizing, especially in times of crisis: During the worst days of the coronavirus pandemic, people fought intensely over the validity of data, and who counted as an expert. And new presidents frequently shift federal priorities, including scientific ones, as is their prerogative—George W. Bush, for instance, defunded certain types of stem-cell research. But generally, those changes have been “much smaller in scale and much more targeted,” Alexander Furnas, a policy researcher at Northwestern University who studies how the government has funded science, told me. Funding for science has also, historically, had strong bipartisan support: Furnas and his colleagues have found that Republicans have appropriated more money to science than Democrats have in the past 40 years. Now the government is experimenting with just how much it can renegotiate its relationship with science.

The new administration seems unlikely to abandon science in its entirety—research into space exploration or artificial intelligence may well continue without friction and even flourish under Trump’s leadership. In an executive order released yesterday, the White House reaffirmed its commitment to tackling chronic disease, a priority of Kennedy’s. But the new administration can pursue certain sectors of science, and talk up scientific values, while still diminishing the research enterprise as a whole. Science and government are now weeks into what will likely be a prolonged battle over how research can and will be done in the United States.

Although some of the government’s actions against science have since been blocked or undone, the suddenness, frequency, and extent of the changes so far have left researchers dreading what might come next.

Nearly every scientist I spoke with for this story told me that they expect national health to decline under the new administration’s leadership. Attempts to defund science on the whole could affect work across fields, throttling drug discovery, clinical trials, climate adaptation, and more. Many also expect that the moratorium on DEI-focused programming will have severe impacts on who is able to do the work of science—further impeding women, people of color, and other groups underrepresented in the field from entering and staying in it. By deleting data and imposing restrictions on the ways in which new data can be collected, the government has also set a worrying new standard for its reach into American research.

The government already wields enormous power over science: It has for decades been the biggest funder of basic research in this country. It employs some of the country’s best scientists. And it’s a crucial source of scientific data, aggregating from across sources to paint portraits of national trends in cancer rates, infectious disease, energy use, air quality, food consumption, and more—essentially, every quantifiable aspect of what it means to exist as an American. Researchers rely on those banks of intel.

Political interference has obstructed American science before. In the 1990s, after a study linked keeping guns at home to higher homicide rates at home, the National Rifle Association lobbied to eliminate the entire National Center for Injury Prevention and Control at the CDC, which had funded that research; in response, Congress passed the Dickey Amendment, which forbade the CDC from using its funds to “advocate or promote gun control.” The agency essentially halted its funding of any firearm-related research until 2019, when Congress again began allocating funds for research on gun violence. By then, many experts had been scared off the topic, Megan Ranney, the dean of Yale’s School of Public Health, told me: Other sources of funding were scarce, and researchers worried that any professional connection to this topic could hurt their chances of getting government grants, even for unrelated projects. When Ranney chose to study firearm violence in the early 2000s, she said, “I was explicitly told not to, because it would doom my career as an academic scientist.” The CDC-funding gap also left a chasm in data sets tracking gun ownership and violence. Only now, Ranney told me, is a critical mass of well-trained researchers starting to make up for those decades of lost evidence.

In the coming years, HIV/AIDS research—to take just one example—could be similarly stifled, Amy Fairchild, a historian at Syracuse University, told me. Epidemiologists who study sexually transmitted infections in general tend to diligently track gender, because patterns of transmission can differ so greatly along that axis. But under Trump’s leadership, that scientific rigorousness has turned into a potential vulnerability. Already, the administration has issued guidance limiting PEPFAR-funded pre-exposure prophylaxis to only “pregnant and breastfeeding women,” excluding by omission other populations extremely vulnerable to infection, including both men who have sex with men as well as transgender people. And several sexual-health researchers told me that the Trump administration recently issued a termination order for their large, CDC-funded study that focused on reducing health disparities among populations affected by multiple STIs. (A judge has since issued a temporary restraining order allowing the study to resume.)

But the new administration’s approach to science bleeds past the bounds of any single field. One cancer researcher at George Washington University, for instance, had to halt a project investigating the best ways to collect information about gender and sexual orientation from patients. The terms the Trump administration has flagged as “gender ideology”—a list that includes pregnant people, transgender, binary, non-binary, assigned at birth, cisgender, and queer—have already been politicized. But by targeting gender as a category as well as the entire concept of DEI, the government has now granted itself a way to directly affect just about any scientific discipline in which human identity and behavior are relevant, or in which people of diverse backgrounds are involved in the work—which is to say, any scientific discipline. The National Science Foundation, for instance, has been hunting for hundreds of DEI-related terms in grants, leaving researchers fearful that merely attempting to include diverse populations in their work will tank their chances of success.

Many scientists told me that they expect the list of terms the government is flagging or excising to expand. The National Institutes of Health is reportedly scouring existing grants that mention COVID, and the Commerce Department has asked the National Oceanic and Atmospheric Administration to scan grants for climate-related words, though the upshot of those searches isn’t yet clear. Kate Brown, a science historian at MIT, told me that some of her colleagues have been passing around lists of words that they think “should not appear in grant applications,” including environment, climate, and race. David Ho, a climate scientist at the University of Hawaiʻi at Mānoa, told me that a dean at his school recommended removing even the term biodiversity from public-facing documents, out of concern that it would be flagged by searches for DEI-related terminology. (A spokesperson at the university told me that they couldn’t find any official message from the school or the university.)

Navigating these new restrictions isn’t as simple as “find and replace.” Deshira Wallace, a health-behavior researcher at the UNC Gillings School of Global Public Health, has been modifying her grant proposals to, for instance, describe the impacts of heat exposure on certain groups overrepresented among farmworkers, rather than referring directly to climate change or Black and Latino people. But she cannot hide that diversity and health equity are core to her work. Other researchers are concluding that the simpler solution may be to step away from projects that, for instance, examine the impacts of air pollution on racially segregated communities (impossible to study without acknowledging diversity) or firearm deaths among American kids (which vary by gender and other demographics).

Scientists are used to pressure from institutions and academic journals to tweak the language that they use, in order to increase the chances that their research might be funded. But to be barred from using certain words—certain concepts—that are firmly grounded in fact represents a new level of government interference, seemingly driven by a skepticism that scientists are approaching their own work correctly. Asking about gender “is established science,” helping researchers identify populations with different needs, wants, behaviors, and risks, Hilary Reno, the medical director of the St. Louis County Sexual Health Clinic, told me. (She herself runs a CDC-funded community survey about sexual-health care that asks about gender, and recently received notice from the agency that she needed to terminate all activity “promoting or inculcating gender ideology.”)

Dismissing established scientific methodology as politically transgressive is a dangerous precedent to set. In the current political environment, Republicans have become generally less trusting of scientists than Democrats are; right-wing groups are openly treating some scientists as political enemies, publishing “watch lists” filled with federal workers, many of them people of color, who work on government programs aimed at reducing health inequities. Scientific expertise itself is now being billed as a political liability, which opens the door to “a populist approach to what counts as valid scientific knowledge,” Fairchild told me. That creates the opportunity, too, for future leaders of any political affiliation to conflate popular opinion with data-driven truths. Assaults on facts and evidence become that much more permissible if facts can be massaged or dismissed by people in power.

Science, like government, has its flaws, inefficiencies, and scandals. The ways in which research has been structured and funded in the U.S. have also been subject to plenty of valid criticism and scrutiny. But the internal standards American science has set for itself have also largely held up. Reputable research must contain data that have been collected with the correct tools under reasonable conditions, are repeatable under different settings by different researchers, and have been thoroughly vetted by independent experts in the same field. These standards can make science seem tedious and opaquely technical, but empiricism serves research’s ultimate goal: to reveal what’s currently true.

Federal data sets, in particular, are meant to be publicly accessible snapshots of reality, Rachel Hardeman, a health-equity researcher at the University of Minnesota, told me. But in any scientific endeavor, “you never manipulate data” or remove it, Niema Moshiri, a bioinformatician at UC San Diego, told me, not from your own data set, not from anyone else’s, not from the public record. The scientific process itself is iterative by design—understanding changes as knowledge accrues—but tampering with or disappearing data is, among scientists, essentially equivalent to misrepresenting evidence. “It goes against everything that we are taught as scientists,” Reno said. Researchers who have been found guilty of these acts have paid the consequences: “You get papers retracted for less than that. You get your career ruined for less than that.”

The federal government has now violated that code of conduct. In yesterday’s executive order, Trump highlighted the importance of “protecting expert recommendations from inappropriate influence and increasing transparency regarding existing data.” But that is exactly what the administration’s critics have said it is already failing to do. At the end of last month, the CDC purged its website of several decades’ worth of data and content, including an infectious-disease-surveillance tool as well as surveys tracking health-risk behaviors among youths. (On Tuesday, a federal judge ordered the government to restore, for now, these and other missing data and webpages to their pre-purge state.) And as soon as the Trump administration started pulling data sets from public view, scientists started worrying that those data would reappear in an altered form, or that future scientific publications would have to be modified. Many scientists could also easily imagine government agencies insisting on removing entire columns of information or swapping the word gender for sex—an inaccurate substitution that could muddle demographic patterns and erase entire populations of people from data sets. Some of those fears have since been realized: NIH officials have been instructed to offer only “male” and “female” as choices on clinical-data forms; CDC scientists have been ordered to retract manuscripts already submitted for publication to scrub any mention of gender-acknowledging terms; at least one published paper has been deleted from a government website, its authors told that reposting will be possible only after forbidden terms are removed.

Eventually, after a lengthy legal fight or under a different president, the Trump administration’s restrictions might be lifted. But future researchers will “have to be really careful about how we interpret these next several years,” Wallace told me: Any gaps in data sets will be permanent. So might the dent in scientists’ trust in the U.S. government. “We have really taken it for granted that that data would always be there,” David Margolius, Cleveland’s director of public health, told me. In the long term, American science may no longer look reputable to scientists abroad. With so many restrictions and caveats on how data are collected, “for me, it would feel a little bit tainted,” Wallace said.

In this more hostile environment, many scientists are weighing their next moves. Some who work for the government are expecting to be fired; some at universities told me they’d consider exiting academic science, perhaps to take a private-sector job; others told me they’d stay until they were forced out. Still others, though, said they aren’t willing to let the American government dictate the terms of their scientific integrity and success. “If these things directly impact our progress? I’m going to leave,” Keolu Fox, a geneticist at UC San Diego, told me. Science, after all, is a transferrable skill—one that plenty of other countries generously support.

For decades, the government’s relatively harmonious partnership with science has been a boon for the United States: American research brought the world a polio vaccine, the first crewed lunar landing, the internet. Since the Manhattan Project, science in America has also been explicitly recruited to the project of national security. The threat of America’s current research infrastructure buckling, though, has revealed how tenuous science’s place in the U.S. really is. “There’s no replacement at scale” for federal funding, Wrigley-Field told me. Universities don’t have the budget to foot the bill for every researcher who loses a grant; private foundations can supplement some funding, but the money they provide might come with strings. And a nation of privatized, commercialized research could invite “all kinds of compromises” and conflicts of interest, Brown told me.

Both by trying to control science and by funding less of it, the Trump administration is starting a slide toward a future where more, rather than fewer, people have reason to distrust science and its results. People whom the government is intent on disparaging may choose to stop participating in research, federally funded or not—further muddying scientists’ ability to monitor the nation’s health or assess how treatments or interventions work for different groups of people. If gender can be dissociated from health, little may stop federal leaders from burying other realities that they find unpalatable—the extent and severity of the growing H5N1 bird-flu outbreak, for example, or the costs of declining vaccine uptake.

There will undoubtedly be periods, in the coming weeks and months, when the practice of science feels normal. Many scientists are operating as they usually do until they are told otherwise. But that normalcy is flimsy at best, in part because the Trump administration has shown that it may not care what data, well collected or not, have to say. During his Senate confirmation hearings, Kennedy repeatedly refused to acknowledge that vaccines don’t cause autism, insisting that he would do so only “if the data is there.” Confronted by Senator Bill Cassidy with decades of data that were, in fact, there, he continued to equivocate, at one point attempting to counter with a discredited paper funded by an anti-vaccine group.

In all likelihood, more changes are to come—including, potentially, major budgetary cuts to research, as Congress weighs this year’s funding for the nation’s major research agencies. Trump and his administration are now deciding how deep a rift to make in America’s scientific firmament. How long it takes to repair the damage, or whether that will be possible at all, depends on the extent of the damage they inflict now.