In November, when Donald Trump first announced his plan to place Elon Musk in charge of a new Department of Government Efficiency, the idea was widely written off as a joke. Then Trump took office, and DOGE began its very real stampede through the government. As an effort to meaningfully reduce federal spending, however, DOGE remains wholly unserious.
Musk initially promised that he would eliminate $2 trillion of the $7 trillion federal budget, before scaling back his ambitions to $1 trillion, and then $150 billion. Even that revised target is highly improbable.
Precisely measuring the budgetary effects of the Musk experiment remains difficult, but we can begin by looking at the claims made by DOGE itself. In late February, its website claimed to have achieved $55 billion in annual-spending reductions. However, its “wall of receipts” detailed only $16.5 billion of this total. Half of that figure came from a typo claiming $8 billion in savings from terminating an $8 million contract. As The New York Times has reported, that was far from the only accounting error. Once such mistakes as false contract cancellations, triple counts of the same reform, and the inclusion of contracts that expired decades ago were fixed, verified budget savings stood at just $2 billion.
[Brian Klaas: DOGE is courting catastrophic risk]
The DOGE website now claims $165 billion in savings. However, it still details only a fraction of the supposed cuts, and earlier accounting errors have given way to new ones. A common sleight of hand is canceling a “blanket purchase agreement”—in which the recipient had been given the equivalent of a credit limit to incur necessary costs on a project—and then claiming savings of the full credit limit rather than the (in many cases substantially lower) amount that was actually spent. Even assuming that the website’s stated savings have become twice as accurate as they were in February, annual savings would reach perhaps $15 billion, or 0.2 percent of federal spending.
Fortunately, more reliable sources than DOGE’s self-reported figures exist. The best is the Treasury Department’s monthly accounting of spending by agency and program. Any true DOGE spending reductions should show up in these budget totals, as should the results of other White House initiatives, including cuts to public-health spending and the ongoing efforts to eliminate USAID and the Department of Education.
These spending data do not flatter the Musk project. Total federal outlays in February and March were $86 billion (or 7 percent) higher than the levels from the same months a year ago, when adjusted for timing shifts. This spending growth—approximately $500 billion at an annualized rate—continues to be driven by the three-quarters of federal spending allocated to Social Security, Medicare, Medicaid, defense, veterans’ benefits, and interest costs. These massive expenses have been untouched by DOGE’s focus on small but controversial targets such as DEI contracts and Politico subscriptions.
We can see this by looking at Treasury’s breakdowns of monthly spending by agency. Short-term program spending can fluctuate greatly, and sustained trends might not be fully apparent for several months, but the early data are nonetheless revealing. Perhaps the highest-profile cuts under the Trump administration so far have been to public-health spending and foreign aid. And yet, even here, the numbers are rounding errors in the context of the federal budget. Public-health spending, previously about $8.2 billion monthly, fell to $7.1 billion in March, led by cuts to the National Institutes of Health and the Health Resources and Services Administration, the latter of which funds state and local health grants to serve underprivileged families.
Monthly spending on targeted foreign-assistance programs has fallen from $2.4 billion to $1.4 billion. This includes spending on “Global Health and Child Survival” programs—which includes highly effective funding to combat HIV, malaria, tuberculosis, and other illnesses in less developed countries—falling in half to $400 million a month. Payments to “International Organizations and Conferences,” such as the United Nations, have fallen to zero. And monthly USAID spending has fluctuated wildly but overall declined by one-third in the first quarter of 2025.
These cuts have already been highly disruptive to beneficiaries, contractors, and employees, and they threaten immense long-term harm. And yet, their total monthly savings have totaled just $2.1 billion. At the Department of Education, another shutdown target, spending has remained steady aside from the early termination of post-pandemic funding that was already scheduled to phase out over the next year.
Cost reductions from laying off federal employees have been too small to show up in the data. This is not surprising, because even laying off one quarter of the 2.3 million federal civilian employees would shave off just 1 percent of federal spending. To be fair to DOGE, more savings will materialize in October, when the salaries of the 75,000 federal employees who took a buyout come off the books. That should save Washington $10 billion a year, or 0.1 percent of federal spending—except even that is an overestimate, because Washington will surely end up hiring contractors to perform at least some of the work previously handled by those civil servants, and many contractors cost more than employees.
[Stephen Macekura: The government waste DOGE should be cutting]
Moving forward, identifying politically acceptable savings will become harder. Trump and Musk have already hit their easiest targets that do not directly burden most MAGA voters, such as government employees, foreigners, academics, and recipients of contracts with some kind of DEI component. More recent moves to slash Social Security customer-service and veterans’-health personnel have faced a backlash from affected Republican voters. Congress has shown little interest in passing legislation to ratify the executive branch’s cuts, meaning many of them will likely be reversed in court. This year’s appropriations bills—which require seven Senate Democratic votes to break a filibuster—will probably continue to finance and mandate the existence of the Department of Education, USAID, and traditional public-health spending.
That, by the way, is the good news for DOGE. The bad news is that the project seems quite likely to expand long-term budget deficits. Slashing IRS enforcement will embolden tax evasion and reduce revenues by hundreds of billions of dollars over the decade. Laying off Department of Education employees who ensure collection of student-loan repayments will increase the deficit. Illegally terminated federal employees are already being reinstated with full back pay, leaving the government with little to show for its trouble besides mounting legal fees.
Even if DOGE somehow manages to end up in the black, any modest savings it achieves will be completely overwhelmed by the GOP’s push to expand the 2017 tax cut at a cost of roughly $500 billion annually. Claims that Washington can no longer afford to spend 0.1 percent of its budget providing lifesaving HIV treatments to 20 million impoverished Africans cannot be taken seriously when the administration and Congress are preparing to cut taxes and expand other spending by trillions of dollars.
None of this is to say that DOGE has failed. Musk might not have followed through on his unfocused and evolving promises to eliminate payment errors, balance the entire budget, and implement regulatory reform. But he has successfully given the White House cover to purge and intimidate the civil service, helped Congress justify exorbitant tax cuts, rewarded MAGA voters with revenge against their perceived enemies, and granted himself the ability to access sensitive government data and possibly ensure his companies’ continued government contracts. Sure, annual budget deficits remain on track to double over the next decade. But if you thought DOGE was really about cutting costs, you were never in on the joke.