The real reason DOGE failed isn’t what you think

Elon Musk will step back from his Trump administration work with a trail of wreckage — and failure — behind him. Musk said last Wednesday that he’d scale back his White House work to one or two days each week soon, likely in May. But he’d already had his power reined in, becoming far less powerful in Washington as his grand ambitions hit a wall. During his White House service, Musk’s “Department of Government Efficiency” dismantled a few agencies, fired tens of thousands of federal employees, canceled lots of contracts, and caused a whole lot of chaos. But beyond making sure that far less lifesaving aid goes to people in foreign countries, it’s difficult to see what he’s accomplished. The story of DOGE’s failure on spending is simple enough: Its huge ambitions to cut $1 trillion never seemed even faintly realistic, and Musk indeed never got anywhere near that target. Yet DOGE was also, effectively, an attempt at a new way of running the federal government — an effort to have Musk wield power like a CEO of the civil service, ordering layoffs and making career civil servants dance to his tune, while allies burrowed in every agency carried out his agenda. And this failed too. For about the first six weeks of President Donald Trump’s administration, Musk really did seem to have something approximate to CEO powers — thrilling the tech right, the Silicon Valley executives who’ve embraced Trump, and dreamed that what they viewed as a sclerotic, inefficient, and untrustworthy federal government could be run like one of their businesses. Then, in early March, things suddenly changed. In what was, in retrospect, a crucial turning point, much of Trump’s Cabinet revolted against Musk’s dictates, and Trump reined him in, decreeing that his agenda would have to be approved by Cabinet secretaries, rather than imposed on them. Much of Musk’s agenda — centralization of power under the president, coercion, and firing of the civil service — is shared by Trump and right-wing activists, and will continue after Musk leaves. What turned out to be unsustainable was Musk’s own role as the “decider,” and the pace at which he tried to ram those changes through. Musk’s status as a special government employee always came with a legal time limit — he can only serve 130 days in a 365-day period. But it’s hard not to think that, if Trump and Musk really wanted him to stick around longer, they’d figure out some workaround. Instead, there seems to be mutual agreement that the time has come for Musk to, mostly, depart. The shock and awe phase: Musk as CEO of the federal workforce When Trump said he’d appoint Elon Musk to head a so-called Department of Government Efficiency focused on government spending, most of Washington yawned. They’d seen such toothless efforts before. Almost no one was prepared for what Musk did — for how aggressively he’d move and for the specifics of what exactly he’d try to pull off. In its first phase, DOGE represented nothing less than a new model of how to run the United States federal government. In it, Musk, a White House adviser empowered by the president, had the power to order sweeping changes and have them carried out rapidly. He acted as the de facto CEO of the federal workforce, as if he got to decide who gets fired, who gets promoted, and how money is spent. This was in keeping with the grand ambitions of leading figures on the tech right. They dreamed that Trump and Musk could, working together, lead an overhaul of the federal government that was something akin to Franklin D. Roosevelt’s epochal New Deal, but “in reverse,” as the venture capitalist Marc Andreessen put it. They wanted the government to be run more like one of their startups — like Musk’s own takeover of Twitter — and to use the Silicon Valley playbook of disruption to make it happen. Trump and many of his hardline advisers loved this idea. They viewed the federal bureaucracy — the “deep state” — as fundamentally opposed to Trump, and as having worked to obstruct his agenda in his first term. An unleashed Musk, they hoped, could change the very basics of how the government functioned. They wanted to put federal bureaucrats, as Office of Management and Budget director Russ Vought said, “in trauma.” And for a while, it seemed to be happening, because Musk had grabbed hold of the levers of power. One lever was firing power. Musk had the Office of Personnel Management (OPM), the government hiring and HR office, stacked with his allies, and repurposed the sleepy office as an instrument of control over the federal workforce. OPM then sent out directives to federal employees — laying off some, offering buyouts to others, while ordering yet others placed on paid administrative leave. (Early on, it became clear that civil servants who defied DOGE requests would swiftly be placed on leave, getting them out of the way.) Another lever was spending power. DOGE ran rampant through government agencies, rapidly canceling contracts it claimed were wasteful, or even, in the case of USAID, dismantling an entire agency (putting it “into the wood chipper,” as Musk said). Musk had even greater ambitions of getting access to far more government data, centralizing information split across databases in many different agencies. He had a belief, The Atlantic reported, that “by controlling the computers, one could control the entire federal bureaucracy.” How DOGE got leashed: The Cabinet struck back It turned out, though, that there were some flaws in Musk’s plan. In his zeal to move fast and break things, Musk paid little heed to the question of whether what he was doing was legal. But much of it seemed quite clearly not to be. And after a few weeks, judges started to say so, ordering the restriction of DOGE’s access to some key government systems and the reversal of some of its spending cuts. There is also a larger question of whether Musk’s appointment and apparent powers themselves are legal — a judge overseeing the case sounded skeptical. His “cut first, ask questions later” approach also led to political problems, as Musk’s popularity dropped, and the administration was dogged with embarrassing headlines about cuts to veterans’ health and other popular causes. His political acumen also came into question when his big public effort to swing a Wisconsin Supreme Court election flopped. What really reined Musk in, though, was that Trump’s Cabinet struck back. Musk and DOGE’s initial advance came as much of Trump’s Cabinet was still awaiting Senate confirmation. Then, when Cabinet secretaries were first sworn in, most seemed to be looking on haplessly as he bulldozed over their objections. Eventually, though, some of them decided they’d actually like to run the agencies they were confirmed to lead, not have Musk do it for them. And so they started to make their objections known. The earliest sign of this came in late February, when Musk emailed the entire federal workforce asking five things they did last week, and some agency heads and top officials told their staff not to answer. The objections came from some with unimpeachable MAGA credentials, like newly confirmed FBI director Kash Patel. Agency heads emphasized that agency heads are responsible for employee review, meaning that Musk isn’t. Musk responded by fuming on X: “This mess will get sorted out this week. Lot of people in for a rude awakening and strong dose of reality. They don’t get it yet, but they will.” Then, in early March, a showdown occurred at an “explosive” closed-door Cabinet meeting, in which Secretary of State Marco Rubio and others aired their grievances with Musk. Some of their objections were about cuts they deemed foolhardy, but others were mainly about who was in charge — they didn’t want to be subordinate to Musk. It seems that, in Trump’s eyes, Musk came out the loser. The president soon announced that from now on, DOGE would have to work with the Cabinet secretaries to make cuts, and to use a “scalpel” but not a “hatchet.” “Did Trump just rein Elon in?” I asked at the time. And with hindsight, the answer clearly seems to be yes. Since then, DOGE has faded as a story, pushing far less dramatic changes. Slowed down by Cabinet secretaries, DOGE began focusing on weaker targets like the US Institute of Peace or the Social Security Administration, and even some of DOGE’s planned cuts there were called off. One illustrative example of Musk’s declining influence relative to the Cabinet came in mid-April. Musk managed to get the White House to name an ally of his as the acting IRS director, while cutting out Treasury Secretary Scott Bessent, who oversees the IRS, from the process. But the victory didn’t last; Bessent was furious, and got Trump to revoke the appointment. Cabinet secretaries, it turns out, have a lot of leverage — losing them (if they’re outraged enough to quit) is embarrassing, and replacing them is politically costly. Overall, it appears that Musk’s stint as the incredibly empowered CEO of the federal workforce ended in early March. He became “just” another White House adviser — still significant, but no longer able to make millions of federal employees hang on his every word. And now he’s reducing his work to, he says, only one or two days a week, which will make his relevance, and ability to win internal power struggles, decrease further. What will DOGE’s legacy be? Musk’s wrecking ball will have real long-term consequences. The agencies he and his allies have dismantled — USAID and the Consumer Financial Protection Bureau — will be very difficult to put back together, so the US will be doing far less foreign aid and consumer protection for the foreseeable future. Tens of thousands of other layoffs or buyouts in federal agencies could also have consequences we don’t fully understand; when you recklessly break things, problems could materialize later. And much of DOGE’s work will continue after he leaves. Trump’s agenda of reducing the federal workforce and strengthening presidential control over the civil service will also keep moving forward, now spurred more by figures like Vought. And though the Musk allies at the OPM aren’t unilaterally engineering layoffs anymore, Trump will reportedly give the office the authority to review new agencies’ hiring requests. Meanwhile, Cabinet secretaries are gradually rolling out “reorganization plans” for their agencies, which will be accompanied by many more layoffs and office closures. DOGE team members remain inside these agencies, though without Musk around to help them, their influence will likely depend on how much each Cabinet secretary chooses to empower them. Finally, part of DOGE’s mission was, in theory, not just to cut government but also to make it more efficient. At least some in Silicon Valley were enthused about building things and making them better, not just smashing them up. Occasionally, there have been glimmers that some at DOGE are trying: Joe Gebbia, a co-founder of Airbnb, has been tasked with modernizing the absurdly antiquated paper-based retirement process for federal employees. Such work is difficult and takes time, but perhaps DOGE will actually have some positive successes at some point. Still, Musk’s ambitious big picture goals for DOGE appear to have clearly ended in failure. He certainly did not succeed in changing the overall trajectory of US spending. This idea never made much sense, since federal spending is overwhelmingly on defense or entitlement programs like Social Security and Medicare, not the discretionary agency spending Trump focused on. And indeed, his ambitions on that front have only kept shrinking — a goal to cut $1 trillion became a goal to cut $150 billion, and he’ll likely fall short even of that. And he certainly did not succeed in bringing about a wholesale remaking or reboot of the federal government. Losing a couple of agencies, and reducing staff a bit at the others, is not the FDR-style revolution the tech right was looking for. And the “CEO for the federal workforce” model flopped, hurt by Musk’s sloppiness and blundering, but ultimately killed when Trump chose to back the Cabinet over him. (Ultimately, the constitution may have restrained Musk less than his loss of favor in the monarch’s court.) There can, however, be lessons in failure. A future president — say, JD Vance, who has long backed right-wing activists’ goal of purging the federal workforce, and has close ties to the tech right — could learn from what went wrong with DOGE, try a new and improved DOGE 2.0, and actually see it through. And perhaps, in a year or two, Trump will get bored and bring back a reempowered Musk to start smashing things again.

What really matters

The real reason DOGE failed isn’t what you think