Iran War Bill Tops $47 Billion, and the Bases Are Not in It

In July, Defense Secretary Pete Hegseth put the cost of the Iran war at $37.5 billion. ^[3] By late October, the Pentagon's own data said more than $47 billion. ^[1]

That is an increase of about $9.5 billion, more than 25 percent, in a little over three months, and the war has not ended. ^[1] My view is plain: a government that hands the public its bill in instalments owes it the whole ledger now, before the next instalment is leaked.

What the numbers say

Reuters reported on October 28 that Pentagon data put US military spending on the war above $47 billion, more than 25 percent over the Defense Department's July estimate. ^[1] The Pentagon had not shared the latest estimate publicly and did not immediately respond to Reuters' request for comment. ^[1]

A day earlier, CBS News reported a CENTCOM estimate of at least $47.1 billion as of mid-October. CENTCOM is the US military command that runs operations in the Middle East, and according to CBS the estimate had been given to lawmakers. ^[2]

So the figure exists, members of Congress have seen it, and the people paying for it learned of it from the press.

CBS gave a breakdown. About $29 billion is for replacing munitions, $6.1 billion for lost equipment and about $11.9 billion for incremental costs, the extra money the war itself has consumed. ^[2] Reuters describes the same shape: operations, more than $29 billion projected for expended munitions, and more than $6 billion for equipment and platform losses, including aircraft and drones. ^[1]

Look at the proportions. Roughly $35 billion of the $47 billion is the price of refilling the shelves and replacing what was lost. The money spent on running the war day to day is the smaller part. Those are projections of what it will cost to restore the arsenal, and replacement prices are exactly the kind of number that moves.

What the Pentagon can fairly say

Let me give the defence its best case. Nobody can price a war that is still going. Costs are real but unsettled, munitions prices are projections, and a first estimate that later rises is not proof of bad faith. An honest figure in July could be an honest underestimate by October. If every revision were a scandal, no official would ever publish a number.

That is true, and it is not the complaint. The complaint is about what the number leaves out and about who gets to see it.

The latest estimate does not clearly account for the cost of repairing US bases damaged by Iranian missile and drone attacks, according to Reuters. ^[1] CBS says the CENTCOM estimate does not include damage to US military bases in the Middle East. ^[2] A bill that omits the broken buildings is not yet a bill. It is a partial invoice, and everyone should call it that.

I spent twenty years reporting from wars, and the lesson I took from most of them is that the first figure a government offers is almost never the last. The pattern is rarely a lie. It is a habit: say the part you can defend, wait, then say the next part. Each number is true as far as it goes, and the sum of them is learned slowly.

The tip of the iceberg

Linda Bilmes, a public finance expert at Harvard Kennedy School, said in an interview published on August 21 that Hegseth's $37.5 billion represented some, but not all, upfront costs. She argued that replacement prices and long-term costs matter. ^[3] Her phrase for the July number was blunt: "The $37.5 billion is just a very tiny tip of the iceberg of cost." ^[3]

She also pointed to the long tail of war spending, including possible future health care and disability costs for service members. ^[3] That is her analysis, not a government tally of this war's total, and the $47 billion is a military-cost estimate, not the war's final fiscal, economic or human price. But her point stands against the Pentagon's own practice. If the July figure was a tip and October's is a bigger tip, no one has yet shown the public the rest of the ice.

Why the calendar matters

The timing makes this more than an accounting quarrel. The administration has asked for tens of billions of dollars in supplemental military funding, partly related to the Iran war, and that request has not passed, according to CBS. ^[2] Democrats have pressed Hegseth for a more detailed accounting. ^[2] Voters go to the polls in the midterm elections on November 3.

And the war may not be over. CBS reported that President Donald Trump has suggested he might restart hostilities after the midterms unless Iran agrees to a deal. ^[2] The Daily Inference reported on October 24 that Iran's ambassador to Pakistan said Tehran had offered to reopen the Strait of Hormuz if Washington lifts its naval blockade, and that the envoy said Trump rejected the proposal. The next estimate, in other words, may be written about a larger war than the one these numbers describe.

Congress is being asked to pay more. The public is being asked to vote.

What a full accounting looks like

I am not asking for a final number, because there is not one. I am asking for a document, published, that does five things.

It states its cutoff date, so a reader knows whether the figure covers mid-October or yesterday. It states its method, including how munitions and equipment replacement are priced. It lists its exclusions, starting with the damage to bases, and gives a range for each. It projects the costs still to come, including what happens if fighting resumes. And it says in plain words which figures are estimates and which are money already spent.

None of that is exotic. Any household that has been through a roof repair knows the difference between the quote, the invoice and the estimate for what the builder found underneath. Government gets no exemption from the question that follows: what else is in there?

The Pentagon can start by answering Reuters. It can then hand lawmakers of both parties the same breakdown it gave the ones who asked, and attach it to any supplemental request. Congress should refuse to vote on more money for this war without a published accounting of the money already spent. Candidates in the final days before November 3 can be asked, plainly, whether they will.

The estimate will rise again. The only choice left is whether the public learns the number from a leak or from a ledger.