DOGE Cost More Than It Saved

On July 4, 2026, the Department of Government Efficiency turned itself off. Its website went dark, its work declared done, its ending timed to the country's 250th birthday.

DOGE was sold as the most aggressive attack on wasteful spending in modern history. So let's do the one thing it was supposed to be about. Let's check the numbers.

The promise

On the campaign trail, Elon Musk said DOGE would cut "at least $2 trillion."

Then the target moved.

When	The number
Campaign	$2 trillion
Early 2025	$1 trillion
April 2025	$150 billion for the year
At shutdown	$215 billion claimed

A goal that falls by more than 90% before you've finished was never a plan. It was a slogan.

The receipts

Even the final $215 billion doesn't survive a look at the ledger.

DOGE's own "wall of receipts" itemized only about $35 billion of the savings it claimed. The tallies were picked apart by PolitiFact, the Wall Street Journal, the New York Times — and the conservative American Enterprise Institute. One contract was listed as an $8 billion savings. The real figure was $8 million. A clerical error, off by a thousand times, counted as a win.

When a budget expert at the Manhattan Institute actually added it up, she estimated the real savings at about $2 billion.

Two billion. Against a promise of two trillion. That is one-tenth of one percent.

It cost more than it saved

Here is the part the savings counter left out: the cutting itself had a price tag, and it was enormous.

The nonpartisan Partnership for Public Service estimated that DOGE's actions would cost taxpayers about $135 billion in a single fiscal year — the bill for putting tens of thousands of workers on paid leave, re-hiring people who were fired by mistake, and the productivity lost in the chaos.

That $135 billion doesn't even count the lawsuits. Or the revenue. Treasury and IRS officials projected more than $500 billion in lost tax collection after DOGE-driven cuts gutted the agency that collects it.

So line it up. Real, documented savings: around $2 billion. Documented cost of achieving them: $135 billion, before lawsuits and lost revenue.

DOGE didn't shrink the government's bill. It added to it.

The part that wasn't on the ledger

The savings were mostly imaginary. The cost was not.

Roughly 300,000 federal workers lost their jobs in the layoffs attributed to DOGE. Real people, real paychecks, real services behind them — cut in the name of a $2 trillion number that turned out to be a $2 billion number.

That is the exchange rate of governing by slogan. You get a headline. Someone else gets a pink slip.

What got cut, and what happened next

Cuts are abstract until they aren't. Here is what a few of them look like on the ground.

A parasite we'd beaten came back. For nearly 60 years, the United States kept the flesh-eating New World screwworm out — largely by funding a program in Central America that stops it before it moves north. In June 2026, it crossed into Texas for the first time in six decades. Fact-checkers caution you can't pin the outbreak directly on any one budget cut. But the underlying math is the whole point: the prevention program is cheap, and the USDA now estimates it saves the U.S. cattle industry $2.3 billion a year. We are now building fly factories and fighting trade bans to handle a problem we were already paying pennies to prevent. That is what false economy looks like.

An outbreak the experts saw coming. This summer, a parasite that causes weeks of explosive diarrhea tore through more than 30 states — over 2,600 cases in Michigan alone, the largest such outbreak in the state's history. Here is the part that should stop you: until July 2025, the CDC's food-surveillance system, FoodNet, actively tracked eight pathogens, including this one. After the cuts, it tracks two. The specific tripwire for this parasite was switched off — and one food-safety expert said she predicted exactly this outbreak the moment she learned FoodNet was being defunded.

Neither of those showed up on the wall of receipts. The savings were counted. The consequences were not. And they are the small stories. The largest one has its own section.

USAID, into the wood chipper

Of everything DOGE cut, the deepest damage was done overseas — and it was done with a grin.

"We spent the weekend feeding USAID into the wood chipper," Musk posted in February 2025. "Could gone to some great parties. Did that instead."

Here is what went into the chipper.

For two decades, USAID-supported programs were credited with preventing an estimated 91 million deaths — roughly a third of them children. They helped cut global HIV deaths by 74% and malaria deaths by more than half. This was not the definition of waste. It was arguably the most cost-effective thing the United States did with any money anywhere: human lives saved, per dollar, at a scale almost nothing else in the budget can match.

Then it stopped, nearly overnight. And the people it had been keeping alive began to die.

One year in, a model maintained by a Boston University epidemiologist estimated the cuts had already caused more than 762,000 deaths — over 500,000 of them children. The specific tolls read like a list of illnesses we already knew how to stop:

Cause	Additional deaths
Pneumonia (children)	164,000+
Diarrhea (children)	125,000+
Malaria	70,000+
Tuberculosis	48,000+

Those are not projections. Those are the deaths counted so far.

The projection is worse. A peer-reviewed study in The Lancet estimates that if the cuts hold, they will cause more than 14 million additional deaths by 2030 — including 4.5 million children under the age of five.

Fourteen million people. For a program that was, in the context of the federal budget DOGE was auditing, a rounding error.

There is no line on a savings counter for a child who dies of malaria the United States used to prevent. That death produces no receipt, no press release, no number to celebrate. But it is a cost — and it may be the largest one DOGE ever created. "Efficiency" did that. On a weekend. Between parties.

To be clear

Government waste is real. Duplicative software, empty office leases, programs that outlived their purpose — those are worth cutting, and cutting them well is exactly what this site is about.

The problem was never the goal. It was the execution. DOGE didn't run the careful, boring, provable work that actually saves money. It ran a marketing campaign with a savings counter, and the counter didn't add up.

The sequel

DOGE is gone. The theory behind it is not.

The same idea — that a Silicon Valley founder can walk in and fix government by force of vision — now has a second act. Joe Gebbia, a co-founder of Airbnb, joined DOGE in 2025. He is now the government's first Chief Design Officer, running a new National Design Studio created by executive order.

The pitch is familiar. Gebbia wants federal services to be "as satisfying to use as the Apple Store." The mandate is to redesign some 26,000 federal websites.

Making government easier to use is a genuinely good goal. But look at what the studio has actually shipped so far: landing pages for the administration's policies on immigration, law enforcement, and drug prices.

Not a faster tax filing. Not an easier benefits application. Polished pages that sell what the administration is already doing.

The bottom line

DOGE promised to save two trillion dollars. It saved, at best, about two billion — while running up an estimated $135 billion in costs, firing 300,000 people, and setting in motion a chain of consequences that, overseas alone, is measured in hundreds of thousands of dead. It did not shrink the government's bill. It grew it. Then, when the savings didn't materialize, it declared victory and switched off the lights on the Fourth of July.

Now the same movement returns as design: if you cannot deliver efficiency, you can at least deliver the look of it.

That is the tell. Efficiency was never the product. The appearance of efficiency was.

Spending our taxes well is not a vibe or a slogan or a beautifully designed homepage. It is a number you can check. DOGE's numbers didn't check out. We should count the next promise before we celebrate it.

Sources