Has DOGE Actually Cut $55 Billion?
A Preamble fact check with the actual receipts
“When you look at some of the waste, fraud, and abuse that’s being uncovered by DOGE and the people at DOGE — these are young — often young but super brilliant people, incredible computer scientists and other things. And they’re finding levels of fraud and waste and abuse like, I think, nobody ever thought possible.”
–President Trump, Feb 18
According to DOGE’s own calculations, in the weeks since Trump has taken office, those young workers have already saved American tax payers $55 billion. DOGE has a website to track the cuts, which says they’ve saved the money through “fraud detection/deletion, contract/lease cancellations, contract/lease renegotiations, asset sales, grant cancellations, workforce reductions, programmatic changes, and regulatory savings.”
The website includes “receipts.”
But have they actually saved $55 billion? My team and I have gone line by line through DOGE’s website, matching up the receipts with the actual government contracts.
A couple of things stood out to us. First, some of the contracts that DOGE lists haven't actually been canceled as they claim, so no money has been saved.
And some of the numbers that supposedly show savings just don’t add up. In fact, it’s difficult to even know where some of these numbers are coming from.
DOGE has in no way come close to showing receipts for $55 billion in savings, even if all of their numbers are correct, which they aren’t. Our calculations show somewhere closer to $7 billion (again, that is assuming every single dollar DOGE is claiming has been saved is accurate).
Have the contracts actually been canceled?
Some of the line items on DOGE’s website show massive cost savings, in the hundreds of millions. But not all of them have even been canceled, meaning no money has been saved.
Let’s start with this one, which falls under the Social Security Administration, and suggests savings of nearly $232 million.
When you click on the DOGE receipt, you see the image below. The last time this contract was modified was in October, during the Biden administration. There is no indication here that it has been closed or canceled.
Here’s another one. DOGE says they’ve saved $73 million from ending this contract:
But clicking on the receipt shows the last time this file was updated was in November.
And these are just a handful of several examples.
Some of the line items show a $0 savings amount, and it’s not clear how or why these would be added to the list. And to reiterate, you can’t save money on contracts that weren’t canceled or that we are legally obligated to pay. Either this information is being compiled in error, or the public is being lied to.
Calculated Savings Are Wrong
Let’s look at some of the math DOGE is using to figure out how much they’ve saved.
But first, let me show you how we got this information, and then you can have the tools to examine it yourself. To start, go to the search feature on USASpending.gov, which is a website that tracks government spending.
Click on the receipt on the DOGE website. (Make sure you toggle on savings instead of total contract value at the top). Then copy the “Procurement Identifier” number at the top of the receipt. Put that number into the keyword search on the USA Spending website, and it will pull up the contract.
Let’s look at this one, which shows over $62 million in savings on the DOGE website.
From the receipt on the DOGE website, you’ll see the Procurement ID is 7200AA22N00007.
When you plug that number into USASpending.gov, it shows you this:
This is a $99 million potenial contract, but over $30 million has already been paid – that’s the “outlayed amount.” Then, $37.6 million is shown as the obligated amount, which is the minimum amount the US government is legally obligated to pay.
The potential award amount here is the total amount that could be paid, if all of the options are executed. For example, some contracts might have the option to extend for a few years, which would up the bill.
The current award amount shows the amount the government has already committed to the contract.
But the potential amount is only if the contract is extended, so any savings calculated should be on the current award amount. In this case, that would be about $10 million, the difference between the current award and the obligated amount.
The math does not math.
And that’s not the only example.
Here’s another one:
DOGE says they’ve saved nearly $37 million dollars on this contract. But is that true? A search of USASpending.gov shows this:
So, just under $18 million has been already given out, slightly more than the $17.7 million is legally required to be paid.
And this receipt was last modified in May, so again, it doesn’t seem to have been canceled by DOGE at all. So really, the savings are apparently $0.
There are so many examples of this, I could do it all day, and you can too.
One after another of these “receipts” are wrong.
But wait, there’s more
When you go through the stated savings on the DOGE website, you’ll come across some that are round numbers. They jump out because most of them are similar, for example, $25,000,000 or $20,000,000.
Many of these are Indefinite Delivery Contracts (IDC). These are contracts where the government doesn’t know exactly how much of something they need when the contract is signed. So they allow the government to go back and reorder as needed, up to the contracted dollar amount. This prevents them from having to put together a new purchasing proposal, which takes time and costs taxpayer money, every time they need a handful of something.
For example: the government knows that over the next year it’s going to be doing some major computer system upgrades, and it’s going to need to install malware detection software on those machines. It also doesn’t know how many computers will break and need to be replaced over the next year, so they might sign an IDC contract for their ongoing software needs.
In this screenshot, you can see a blanket purchase agreement (a type of IDC contract) for $20,000,000:
Pulling it up on USASpending.gov shows this:
Blanket purchase agreements are not legally binding contracts, so if nothing is ever ordered, no money is obligated to be spent.
Saying we are “saving” this money is much like me having a credit card in my wallet with a $20,000 spending limit. If I do not use that card and have a $0 balance, have I “saved” $20,000? No. Having the ability to buy something and not buying it is not a savings.
This is not an audit, and it’s not how audits work. Posting inaccurate screenshots on a website and simply saying the word “savings” does not mean actual savings.
So if it’s not an audit and the people working for DOGE are not auditors, what’s going on here? Is it ineptitude? Is it data manipulation? Is it government propaganda? Whatever the answer, something is not adding up.
An earlier version of this newsletter incorrectly calculated the savings on the outlayed vs obligated award amount. Our apologies for the error.
The meticulous analysis debunking DOGE's $55 billion savings claim illustrates a critical point, but it doesn’t take into account the hidden costs of all the damage being done. While it is such important work to show how DOGE's math fails even on its own terms - from phantom savings on contracts that weren't actually canceled to inflated figures that exceed the maximum possible savings - the real numbers are even worse than the $7 billion figure that Sharon’s thorough analysis has already reduced it to. Possibly less than zero would be accurate?
Let's do some basic math: Start with that $7 billion in supposed savings. Now subtract the immediate, quantifiable costs of this reckless approach:
- The administrative costs of processing mass terminations
- The legal expenses from inevitable wrongful termination suits
- The higher future costs of restarting critical programs
- The expense of rebuilding lost institutional knowledge
- The cost of repeating work that was abandoned mid-process
These alone would eliminate most if not all of the claimed savings. But the true scope of waste and damage runs far deeper: There is the economic cost of the lives lost, livelihoods ruined, and preventable diseases spread. (BTW I'm ignoring the enormous moral value of these lives and communities for the sake of staying on topic of the national budget, obviously value doesn't only come in the form of $.)
In "Veterans Left Behind, Millions Spent on Confederate Names, and Europe on Edge," Sharon revealed how DOGE's attack on the VA exemplifies this destructive approach. They're firing medical personnel who had outstanding performance reviews, after investing years in their specialized training, security clearances, and institutional knowledge. The VA claims these layoffs will save $98 million annually, but they're making these cuts while suffering from major staffing shortages - meaning they'll eventually need to spend even more to recruit and train replacements. Meanwhile, they're simultaneously spending $62.5 million to rename military bases, only to spend that amount again to rename them back.
I was just listening to the “Science Vs” podcast, a research-backed science show, which has documented an even more devastating pattern across other agencies in their recent episode “The War on Science". It uncovered thousands of scientists being terminated at the CDC and NIH, with research worth hundreds of millions being abandoned mid-stream. The University of Washington's medical school stands to lose $90-110 million in funding, much of it for clinical trials that are nearly complete. At St. Jude's Children's Research Hospital, cuts to NIH funding would cost nearly $40 million annually, with their cancer center director warning bluntly: "More children will die." The podcast identified over 30 frozen studies where volunteers were already under researcher care, including malaria treatment trials for children under 5 in Mozambique, tuberculosis treatment studies in Peru and South Africa, and critical HIV vaccine research.
The cascading institutional damage is equally severe. Science Vs revealed how scientists are desperately trying to preserve terabytes of vital health data before it disappears, including critical information that helped control the recent monkeypox outbreak. They documented how researchers knew exactly which communities were most at risk because they had detailed data on infection patterns - data that's now being fragmented or deleted under Trump’s directives. And why is this data being deleted? Somebody decided that keywords like “systemic” or “women” are too woke and any work using it should be destroyed.
The dismantling of USAID shows how these impacts extend far beyond our borders. Research participants around the world have been suddenly abandoned, destroying trust in American institutions and partnerships that took decades to build. People on experimental treatments have been left without care, and promising research into preventing future pandemics has been halted mid-stream.
Having unqualified DOGE staff - who lack both subject matter expertise and basic understanding of government operations - making these sweeping decisions compounds the damage. The combination of mathematical errors in their claimed savings and complete disregard for true costs reveals this as political theater.
This human toll has profound economic consequences too. Each preventable death represents not just a moral tragedy but the loss of economic productivity, tax revenue, and societal contribution. The CDC estimates that preventing just one premature death from cancer creates approximately $1 million in economic value. Pandemic prevention yields even greater returns - with studies showing that every dollar invested in public health preparedness saves $13-15 in avoided costs. Maintaining vaccination programs prevents billions in healthcare costs and economic disruption.
By that measure, DOGE's actions represent not savings, but an unprecedented destruction of public value and capability that will take years, if not decades, to rebuild - if we can rebuild it at all.
One of the things I struggle with in regard to Trump is his manner of speaking. The opening quote is a good example. A vague source and a lot of hyperbole. Who is “somebody” or “they” or “nobody”, “my friends” and everything is extreme or absolute, “super”, “nobody”. If only I could go through work like that. I’d be laughed out of the room if I answered questions with such little information.