On the afternoon of Friday, February 7, as staff members were getting ready to leave the headquarters of the National Institutes of Health, just outside Washington, D.C., officials in the Office of Extramural Research received an unexpected memo. It came from the Department of Health and Human Services, which oversees the NIH, and arrived with clear instructions: Post this announcement on your website immediately.

The memo announced a new policy that, for many universities and other institutions, would hamstring scientific research. It said that the NIH planned to cap so-called indirect costs funded by grants—overhead that covers the day-to-day administrative and logistical duties of research. Some NIH-grant recipients had negotiated rates as high as 75 percent; going forward, the memo said, they would now be limited to just 15 percent. And this new cap would apply even to grants that had already been awarded.

The announcement was written as if it had come from the NIH Office of the Director. It also directed all inquiries to the Office of Extramural Research’s policy branch. And yet, no one at the NIH had seen the text until that Friday afternoon, several current and former NIH officials with knowledge of the situation told me. “None of us had anything to do with that document,” one of them said. But the memo was dressed up in a way clearly intended to make it look like a homegrown NIH initiative. (Everyone I spoke with for this story requested anonymity out of fear of reprisal from the Trump administration. HHS did not respond to requests for comment.)

Over the next several days, the memo sparked confusion and chaos at the NIH, and across American universities and hospitals, as researchers tried to reckon with the likely upshot—that many of them would have to shut down their laboratories or fire administrative staff. A federal judge has since temporarily blocked the cap on indirect costs. But the memo’s abrupt arrival at the NIH, and the way it bulldozed through the agency, underscores how aggressively the Trump administration is exercising its authority and demanding compliance. “Their approach seems to be We go in; we bully; we say, ‘Do this; you have no choice,’” and shows little regard for the people or research affected, one former official told me.

Typically, a memo communicating a major decision related to grants would take months or years to put together, sometimes with public input, and released six months to a year before being implemented, one current NIH official told me—earlier, even, “if the impact will be more substantial.” In this case, though, Stefanie Spear, the HHS principal deputy chief of staff, told officials in the Office of Extramural Research, which oversees the awarding of grants, that this new memo needed to be posted to the NIH website no later than 5 p.m. that afternoon—within about an hour of the agency receiving it. Soon, the timeline tightened: The memo had to be published within 15 minutes. “It was designed to minimize the chance that anyone within an agency could even have time to respond,” another former NIH official told me.

Substantial changes are generally vetted through HHS leadership, and NIH officials have always “very much abided by the directives of the department,” the former official said. But in the past, drafting those sorts of directives has been collaborative, a former NIH official told me. If NIH officials disagreed with a policy that HHS proposed, a respectful discussion would ensue. Indirect-cost rates are controversial: The proportion of NIH funding that has gone to them has grown over time, and proponents of trimming overhead argue that doing so would make research more efficient. A cut this deep and sudden, though, would upend research nationwide. And to grant recipients and NIH officials, it seemed less an attempt to reform or improve the current system, and more an effort to blow it up entirely. Either way, a unilateral demand to publish unfamiliar content under the NIH’s byline was unprecedented in the experience of the NIH officials I spoke with. “It was completely inappropriate,” the former official told me.

But Spear and Heather Flick Melanson, the HHS chief of staff, insisted that the memo was to go live that evening. Officials immediately began to scramble to post the notice on the agency’s grants website, but they quickly hit some technical snares. Fifteen minutes passed, then 15 more. The two HHS officials began to badger NIH staff, contacting them as often as every five minutes, demanding an explanation for why the memo was still offline. The notice went live just before 5:45 p.m., and finally, the phone calls from HHS stopped.

Almost immediately, the academic world erupted in panic and rage. At the same time, the news was blazing through the NIH; staff members felt blindsided by the memo, which appeared to have come from within the agency but which they’d known nothing about. The notice’s formatting, tone, and abruptness also led many within the agency to suspect that it had not originated there or been vetted by NIH officials. “I’ve never seen anything so sloppy,” the current NIH official, who has written several NIH notices, told me. “We also don’t publish announcements after 5 p.m. on Friday, ever … I checked multiple times to be sure it was real.”

The NIH had already been caught in the Trump administration’s first salvo of initiatives. On January 27, a memo from the Office of Management and Budget froze the agency’s ability to fund grants. (In the following week, multiple federal judges issued orders that should have unpaused the funding halt, but many grants remained in limbo.) And in 2017, during Donald Trump’s first term, his administration went after indirect costs, proposing to cap them at 10 percent. That prompted the House and Senate Appropriations Committees to introduce a new provision that blocked the administration from altering those rates; Congress has since included language in its annual spending bills that prevents changes to indirect costs without legislative approval. On February 10 of this year—the Monday after the memo restricting those rates went up—yet another federal judge issued yet another temporary restraining order that again instructed the NIH to thaw its funding freeze.

Last week, the NIH told its staff to resume awarding grants, with prior indirect-cost rates intact. But “the damage is done,” the former NIH official said. Scientists across the nation have had their funding disrupted; many have had to halt studies. And at the NIH—where roughly 1,000 staff members recently received termination notices, amid a mass layoff of federal workers that stretched across HHS—those who remain fear for their job and the future of the agency. The nation’s leaders, NIH officials told me, seem entirely unwilling to consult the NIH about its own business. If the administration remains uninterested in maintaining the agency’s basic functions, the NIH’s purpose—supporting medical research in the United States—will crumble, or at least deteriorate past the point at which it resembles anything that the people who make up the agency can still recognize.

This is an edition of The Atlantic Daily, a newsletter that guides you through the biggest stories of the day, helps you discover new ideas, and recommends the best in culture. Sign up for it here.

Last week, Representative Marjorie Taylor Greene called the national debt “one of the biggest betrayals against the American people,” suggesting that Americans’ anger about debt “gave birth to the concept of DOGE.” The idea that Elon Musk and his band of government-efficiency crusaders can bring down the debt is a tidy one. But DOGE’s current plans would hardly put a dent in the deficit.

Musk has lamented that America is “drowning” in debt, which has indeed ballooned over the past decade: As of this month, the federal debt is $36 trillion, about $13 trillion higher than it was five years ago. Debt has not been a priority of either major political party for some time, my colleague Annie Lowrey, who covers economics, told me. And despite Taylor Greene’s claims about American anger over the debt, it’s not a top-of-mind issue for people at the polls, either, Annie argued.

If Musk’s team were serious about reducing the deficit, it could explore some unpopular but effective options: reduce spending for the military and the entitlement programs that make up the bulk of the federal budget—Medicare and Social Security—or simply raise taxes, Annie suggested. Instead, what Musk and DOGE have done thus far is ravage government agencies and departments (USAID, for example, which makes up a tiny portion of the budget, and the destruction of which won’t lead to major savings). They’ve also focused on slashing the federal workforce by offering buyouts to 2 million federal workers (and, over the weekend, axing thousands more federal-agency employees); so far, salaries for the workers who have accepted the buyout offer make up a minuscule portion of the national budget in total.

Musk, Trump, and their allies have also turned to a bit of magical thinking, claiming that rooting out fraud in the government is the key to saving money. In a meandering address from the Oval Office last week, Musk claimed without evidence that USAID workers were raking in millions in kickbacks, and that people as old as 150 were claiming Social Security benefits. He wrote on X last week that “at this point, I am 100% certain that the magnitude of the fraud in federal entitlements (Social Security, Medicare, Medicaid, Welfare, Disability, etc) exceeds the combined sum of every private scam you’ve ever heard by FAR.”

Stumbling upon, and reclaiming, trillions of fraudulently spent funds would be rather convenient, and crying “fraud” is a useful way for Musk and his defenders to cast DOGE’s actions as in service of the American people. Trump has touted this same shaky logic, asserting that uncovering a bunch of fraud could mean America has less debt than previously thought. Fraud does exist in parts of the government: Some people intend to defraud government programs; others accidentally sign up for benefits they’re not actually eligible for. And the government does sometimes make payment errors—federal agencies estimated that more than $200 billion was lost in fiscal year 2023 because of such mistakes, and in past years fraud losses accounted for 3 to 7 percent of the budget. But there is no evidence that lowering the deficit is as simple as tamping down on fraud—or that fraud exists to the extent Musk claims.

Plus, by whacking the bureaucracy, Musk and his team are weakening programs that are already working to tamp down fraud. All federal programs have fraud-detection mandates. The Treasury, for example, announced in October that it had recovered or prevented $4 billion in fraud losses in the prior fiscal year, in part from employing AI machine-learning. And as he rails against what he calls fraud, Musk and his associates have effectively shut down the Consumer Financial Protection Bureau, whose mandate is to crack down on fraud in businesses (and which might have regulated Musk’s own companies).

The rhetorical trick of politicians referring to unpopular or disliked government spending as fraud isn’t new. But in an era of rampant scamming, claiming that the American government is swindling its own people hits on a salient national fear. Musk’s first few weeks running DOGE don’t bode well for his ability to solve the debt crisis. He may succeed, however, in further eroding trust in government, which could give him and his team even more leeway in their attempts to dismantle it.

Related:

The hidden costs of Musk’s Washington misadventure The government’s computing experts say they are terrified.

Here are four new stories from The Atlantic:

How MAGA is reimagining foreign policy January 6 still happened. Want to change your personality? Have a baby. Frozen food’s new wave

Today’s News

Top representatives from Russia and the U.S. met in Saudi Arabia to discuss strengthening economic and diplomatic relations between the two countries and assembling a team to start peace negotiations in the Ukraine war. The Trump administration fired thousands of probationary federal workers over the weekend in departments including the FAA, Health and Human Services, and Energy. The acting head of the Social Security Administration resigned after DOGE requested access to sensitive personal information about millions of Americans.

Dispatches

Work in Progress: Derek Thompson explains how COVID pushed a generation of young people to the right.

Explore all of our newsletters here.

Evening Read

NBC

Saturday Night Live Played the Wrong Greatest-Hits Reel

By Esther Zuckerman

Fifty years is a long time. But you wouldn’t necessarily know that from large portions of SNL50: The Anniversary Special, the much-hyped celebration of the long-running sketch show that aired in prime time last night. SNL50 was meant to commemorate the program, created and executive-produced by Lorne Michaels, for achieving five decades of cultural relevance. But the evening’s rundown suffered from a severe case of recency bias, with sketches that were more inclined to play it safe than honor the show’s extensive, complicated, and fascinating history.

Read the full article.

More From The Atlantic

Photos: “Not my President’s Day”: Thousands gather in protest against Trump. How the woke right replaced the woke left Tom Nichols: The death of government expertise How progressives broke the government

Culture Break

Marvel Studios

Watch (or skip). The newest Captain America movie (out now in theaters) seemed set to explore much-needed fresh ground for the Marvel franchise. But the movie quickly wastes any of that potential, writes David Sims.

Read. Robert Frost wasn’t always good, James Parker argues. Before he became America’s most famous poet, he wrote some real howlers.

Play our daily crossword.

Stephanie Bai contributed to this newsletter.

When you buy a book using a link in this newsletter, we receive a commission. Thank you for supporting The Atlantic.