If deterrence works,
the rest is a subsidy
You do not have to be against the military to ask this. You only have to take their own explanation seriously enough to follow where it goes.
“They got money for wars, but can’t feed the poor.”
Tupac Shakur, Keep Ya Head Up, 1993
Everything below is a longer way of saying that, and it arrives from the opposite direction. This page grants the strategists their theory and follows it until the budget looks absurd. He never needed the theory. You do not have to hold any position on deterrence to notice that the money exists when it is for one thing and cannot be found when it is for another; that is visible from a street corner without a single citation.
His version fit in a breath and was set to music, which is why it travelled further than every white paper ever written on military spending — and why the front door of this museum is a music game. The argument that moves is the one somebody can carry.
And he was not first. Forty years before him, from as far up the other end of the world as it is possible to stand:
“Every gun that is made, every warship launched, every rocket fired signifies, in the final sense, a theft from those who hunger and are not fed, those who are cold and are not clothed. This world in arms is not spending money alone. It is spending the sweat of its laborers, the genius of its scientists, the hopes of its children.”
Dwight D. Eisenhower, “The Chance for Peace,” 16 April 1953
That is a five-star general who commanded the invasion of Europe, standing up as a Republican president, calling military spending a theft from the hungry. Eight years later he named the machinery on his way out the door, in his farewell address: the military-industrial complex, and the warning was his.
Three voices, three impossible directions, one conclusion.
The Supreme Allied Commander in 1953. A rapper in 1993. And below, the strategists’ own premise followed all the way to the end. This is not a partisan page and it cannot be made into one — the man who ran D-Day got there first, and he said it in office, out loud, to the country that had just made him president.
Grant the premise. All of it.
The standard defence of the enormous postwar military establishment rests on a claim that is genuinely serious and genuinely argued by serious people: that nuclear deterrence is what prevented a third great-power war. Historians call the stretch since 1945 the Long Peace. The mechanism proposed for it is that the cost of a war between major powers became, for the first time, unmistakably higher than any prize it could win.
This page does not dispute that. It concedes it, completely and without hedging, because the argument is stronger with the concession than without it, and because an argument that has to deny its opponent's best point was never much of an argument.
So: deterrence worked. Now follow it.
Then what is everything else for?
If the warheads bought the peace, they bought it. The conventional establishment underneath them — the carriers, the bases, the aircraft programs, the standing forces — did not buy that peace, because by the hawks’ own account the deterrent already did.
So it is buying something else. And the record of the last eighty years says plainly what: the wars that deterrence did not prevent. Korea. Vietnam. Iraq. Afghanistan. The interventions, the occupations, the campaigns conducted at a distance from anything resembling a threat to the country paying for them.
Not one of those was compulsory. Every one of them was chosen.
That is the whole argument in one line: if the deterrent already works, the rest of the budget is not buying safety. It is buying the ability to fight the optional ones. Which is a subsidy for a choice, and a choice can be made differently next year without anybody becoming less safe.
“Nuclear weapons solved war.” They did not, and this page will not say they did, because the sentence collapses the moment anyone presses it.
War did not stop in 1945; it changed venue. Korea, Vietnam, Cambodia, Iraq, Afghanistan, Syria, Yemen, Ukraine, and a long list besides. Millions of people have died in war since the year war was allegedly solved. The claim is true only if you quietly mean great-power war in Europe — and once you have been made to say that out loud, the sentence has shrunk to a footnote and the budget argument goes down with it.
Say the narrow thing that survives instead: the deterrent appears to have prevented one specific catastrophe. That is enough. It was always enough.
And “solved” is the wrong word for what happened
It nearly ended everything by accident, more than once. October 1962, off Cuba. November 1983, when a NATO exercise was reportedly read on the other side as preparation for a first move.
And the one that belongs on this site. On 26 September 1983, a Soviet early-warning officer named Stanislav Petrov watched his system report American missiles inbound. Procedure was to pass it up the chain. He judged it a malfunction — the count was too small to be a real first strike, and the satellites had been unreliable — and he reported a false alarm instead.
He was right. It was sunlight on high-altitude clouds.
A system that requires one man to overrule a confident machine, correctly, on the worst night of his life, is not solved. It is a coin that keeps landing on its edge.
Which is this house’s entire thesis at a scale nobody wants to think about. The machine reported. It reported with total confidence, and it was wrong — the same tell, in a bunker, with the world on the table. Only humans score, and one of them scored for all of us, and almost nobody knows his name.
How I learned to stop worrying and love the invoice
Kubrick got here in 1964, from the other end. Dr. Strangelove or: How I Learned to Stop Worrying and Love the Bomb does exactly what this page does — it grants the logic of deterrence completely, refuses to argue with a single step of it, and then follows the logic until it arrives somewhere insane.
The object at the centre of that film is the strongest companion Petrov could have. The doomsday machine is automated on purpose, so that no human can call it off. That is stated as its entire virtue: take the human out, and you take the hesitation out, and a deterrent that cannot hesitate is finally credible. Then it triggers by accident, and nobody can stop it, because they built it so that nobody could.
Strangelove’s nightmare was a system with no Petrov in it.
Nineteen years later a man sat in front of the real version and hesitated, and hesitating was correct. The film’s joke and the history landed on the same point from opposite directions: the safety was never in the hardware. It was in the one part everybody kept proposing to remove.
There is a second joke in there that is also a real argument. The machine fails to deter anything because the Soviets had not announced it yet — they were saving the reveal. A deterrent nobody knows about deters nobody, which is the funniest available way to say that the entire doctrine runs on communication rather than on tonnage. What deters is the other side knowing. That is a much cheaper product than the one being sold.
So take the title at its word and go all the way. Stop worrying. Love the bomb. Concede that the arsenal did the job, exactly as advertised.
Now let’s talk about the invoice.
Because the bill that arrives every year is not itemised for the thing you just agreed worked. It is itemised for everything else — and “everything else” has a record, and the record is Korea, Vietnam, Iraq, Afghanistan. You already paid for the peace. This is a separate charge.
Deterrence is not free. Warheads need maintenance, delivery systems, early warning, command and control. “Stop funding the arsenal” would defund the very thing being credited, so the argument has to be about the conventional budget specifically or it eats itself.
Conventional forces do things a deterrent cannot. They answer aggression below the nuclear threshold, keep shipping lanes open, honour alliance commitments, and move faster than anything else on earth when an earthquake happens. A country with warheads and nothing else has one option and it is unusable.
Ukraine. A nuclear power invaded a neighbour that had given up its inherited Soviet warheads in the 1990s in exchange for assurances. Deterrence did not prevent that war; some argue it enabled it, by making outside intervention too risky. That is the hardest single fact for the thesis above and it belongs on the page rather than off it.
The Long Peace may have other parents. Trade dependence, the spread of democracies, international institutions, or plain luck. Crediting the warheads is a hypothesis, not a measurement.
The ruling
The objections narrow the claim; they do not answer it. Grant every one. Fund the deterrent, fund the disaster response, fund the alliance obligations, fund the ships that keep the sea lanes open. Nothing above argues otherwise, and a page that pretended those were worthless would deserve to lose.
What remains after all of that is still enormous, and still, by the hawks’ own accounting, not the thing that bought the peace. The residue is the war-fighting capacity for conflicts of choice, and the last eighty years are the receipt.
On Ukraine specifically: it is the strongest objection and it cuts a direction people do not expect. A country that gave up its warheads was invaded by one that kept them — which is an argument about deterrence, and not an argument for a larger conventional budget somewhere else. It made the case for the arsenal and said nothing at all about the rest.
“Then how do you pay for it?”
It arrives on schedule. Grant everything above, agree the residue is optional, propose spending a piece of it on something that keeps people alive, and the answer is always the same four words. So this section is the answer, and it starts by throwing away the version of it that loses.
“The money is fake anyway; it’s all fiat currency.” No. It is the fastest way to lose the room, and it is not true. A dollar is a claim on real goods and real labour, and those are not fake at all — the welder is real, the concrete is real, the shift you worked last week was real. Say the money is fake and you have just said the paycheque is fake, in front of people who earned theirs. The other side does not even have to answer you; you did it for them.
Here is the version that survives contact, in three moves, none of which require anybody to believe anything strange about money.
One: a currency issuer is not a household. A household has to get dollars before it can spend them. The United States issues the dollar. It does not collect the thing it alone creates in order to have some; Congress appropriates, and the money exists. Taxes do real work — they drive demand for the currency, they pull spending power back out, they decide who carries the load — but “we ran out” is not among the jobs they do.
Two: the constraint is real resources, not dollars. This is the part that matters, and it is a limit rather than a licence. The question is never whether the account has the number in it; the question is whether the turbines, the transformers, the concrete, the licensed welders and the grid interconnection exist. If they do, the funding is an act of Congress. If they don’t, no appropriation on earth conjures a welder.
Three, and this is the test: notice where the question shows up. Nobody stood at the appropriations desk during a war and asked how it would be paid for. Nobody asked during the bailouts. Nobody asks about the arsenal. The question appears, reliably, in front of one category of spending, and it is the category where the beneficiary is a person.
A rule that only gets applied to one team is a preference wearing a rule’s clothes.
Inflation is the real ceiling, and it is not a technicality. Bid for resources that do not exist and you do not get the thing; you get higher prices for it, and the people the spending was supposed to help pay that first, at the till. The constraint is entirely real. It is simply not the bank balance.
This argument does not apply to Michigan. States cannot issue currency. A state, a county and a city genuinely are households: they tax, they borrow, they receive, and they can run out. Anyone who carries this reasoning into a council meeting about a pool deserves to lose that meeting, and will. Use it on the national invoice; leave it at the state line.
Mainstream economics disputes the framing. The conventional account says sustained deficits raise interest rates and crowd out private investment, and that the issuer’s freedom is bounded by what lenders and markets tolerate. That is a serious position held by serious people, and this page is standing on the contested side of a live argument rather than on settled ground.
Reserve status is a condition, not a law of nature. Much of the room the United States has comes from the dollar being what the world holds. Treating that as permanent is a bet, and bets get called.
And move two cuts both ways. Nuclear construction is precisely where the real-resource limit bites — heavy forgings, specialised trades, regulatory throughput. Saying “dollars are not the constraint” does not make a reactor arrive sooner. It only tells you which problem to actually work on.
The ruling: this does not unlock infinite spending, and any page claiming it does is selling something. What it does is move the question one step, from can we afford it to what do we want the real resources doing — which is the same question this entire page has been asking about the budget. The country is not short of money. It has been short of a decision.
Different lane, same invoice: fund the electricity out of the stockpile
Which brings the argument somewhere concrete, because there is a bill already sitting in the drawer for exactly this and it has been paid once before.
Electricity demand is climbing, the data centres are being built either way, and the country is arguing about where the next generation comes from. Meanwhile there is a stockpile that was built at enormous public expense to be used exactly never, and a body of experience showing that the material in it burns perfectly well in a power station.
It already happened, at industrial scale, for twenty years. From 1993 to 2013, uranium out of dismantled Soviet warheads was downblended into ordinary reactor fuel and burned in American plants; a large share of this country’s nuclear fuel came out of the arsenal that had been pointed at it, and nobody put a sticker on the bill. When that programme ended it did not end because the physics failed. The plutonium half died on a spreadsheet, and the whole story is next door.
For twenty years, the invoice ran backward.
So the two halves of this page meet here. The residue is optional; the funding question is a decision rather than an arithmetic; and the raw material is already bought and sitting in a bunker. Turning a warhead into a decade of somebody’s lights is not a peace slogan. It is the single most literal available answer to “how would you pay for it”: with the thing you already paid for.
And this is the pro-life argument
A budget is a moral document; it is the only campaign promise that arrives with receipts. So the test is not what a person says about life. It is what they fund.
Pro-life all the way down means the womb and the war budget measured with the same ruler. A bomb is spent once; a book is spent a thousand times. And the arithmetic here is not a slogan:
You cannot be pro-life and fund the optional ones.
Not the deterrent. Not the lifeboats. The optional ones — the wars that were chosen, by people who could have chosen otherwise, and paid for out of the same document that could have built something instead.