Reserve status is not a prize. Robert Triffin named the obligation in 1960 and it has not changed since: to supply the world with its reserve asset you must supply it with your paper, which means running deficits permanently, and the standing demand for that paper is a bid for your currency that has nothing whatever to do with what you sell.
That bid holds the dollar dearer than trade alone would set it, and the people who pay for the difference are the ones who have to sell something abroad. Manufacturing employment is about 12.6 million, roughly 8 per cent of the workforce against 13 per cent in 2000, with between four and a half and five and a half million jobs gone. Stephen Miran, who now chairs the Council of Economic Advisers, put the cause plainly before he held the office: the imbalance comes from persistent dollar overvaluation driven by “inelastic demand for reserve assets”, and “the manufacturing and tradeable sectors bear the brunt”.
The privilege is fifty basis points on the federal borrowing rate. The bill went to the towns that used to make things, and it was never itemised.
Measure the privilege honestly and it is small. Reserve demand takes an estimated fifty to sixty basis points off federal borrowing costs, worth something like 90 billion dollars a year. That is nothing to a machinist in Ohio and everything to a Treasury that can place its paper into an automatic bid and therefore never has to hold a vote first. The campaign against Iran cost about 72 billion in two months, and 67 billion more was requested in July.
So do not accept an accord as the remedy. Century bonds, coordinated devaluations, tariffs dressed up as monetary policy: every one of them keeps the discretion and merely renames it. Remove the automatic bid and a war becomes a budget vote, which is what it was always supposed to be. And no rival will end this. Creditors will, and they have started: 289 tonnes of gold in a single quarter, and gold now past both the euro and US Treasuries in the world's official reserves.
Robert Triffin described the trap in 1960 and it has not been escaped since. If your currency is what the world holds, the world must be able to get hold of it, which means you have to send out more than you take in, permanently. Reserve status is therefore not a bonus paid to a strong economy. It is an obligation that requires you to run deficits whether or not you want to, and the country that accepted it has spent sixty years describing the consequences as bad luck.
The mechanism is a price, and prices are the part of this argument the mainstream keeps skipping. Foreign central banks, sovereign funds and reserve managers must hold dollar assets for reasons that have nothing to do with buying American goods. That demand is a standing bid under the currency. It holds the exchange rate above where the trade account alone would put it, and every exporter competing on price pays the difference. Nobody legislated that transfer, nobody votes on it, and no line in any budget records it.
The privilege is fifty basis points on the federal borrowing rate. The bill went to the towns that used to make things, and it was never itemised.
Stephen Miran wrote the clearest statement of it, in 2024, before he was appointed to chair the Council of Economic Advisers. The imbalance, he argued, has its root in persistent dollar overvaluation driven by “inelastic demand for reserve assets”, and where the cost falls is not in doubt: “the manufacturing and tradeable sectors bear the brunt”. His account reverses the direction the textbooks teach. America does not run current account deficits because it imports too much; it imports too much because it must export Treasury securities to supply the world with reserve assets.
The size of the bill is a matter of record. Manufacturing employment stands at around 12.6 million, roughly 8 per cent of the workforce against 13 per cent in 2000. Between four and a half and five and a half million of those jobs have gone since the turn of the century. Those are not abstractions and they were not distributed evenly: they were concentrated in the places least able to replace them, and the compensation promised at the time was never delivered.
The reply is that real manufacturing output is higher now than in 2000, so the losses must be productivity rather than the exchange rate. Take it seriously and it still does not carry. Output measured in value added rises fastest exactly where automation is cheapest, which tells you about margins rather than about industrial capacity, and a country can post record output while losing the ability to build a thing at scale in a hurry. And the honest form of the productivity argument is a share, not a verdict. Answering that some of the loss was automation does not establish that none of it was the currency.
The other reply is the accounting identity: the current account balance equals investment minus saving, so capital comes in first and goods follow. That is true and it is not an explanation. Saving and investment are not fixed quantities that the world hands you; they respond to prices, and the reserve bid is one of the prices. Naming an identity is naming the arithmetic in which the cause has to be found, and then declining to look for it.
The strongest thing said against this reading is the recent evidence, and it deserves a straight answer. In the first half of 2026 the American trade deficit fell by 189.3 billion dollars, 33.8 per cent against a year earlier, exports up 11.7 per cent and imports up 0.4, while the dollar's reserve role did not shift at all. Correct, and it settles nothing. Nobody argues that the reserve bid sets the trade balance quarter by quarter; a tariff wall and an oil shock will do that in six months. The claim is about the level of the exchange rate over decades, and decades is the period in which the factories left.
Now measure what is bought with the burden. Reserve demand takes an estimated fifty to sixty basis points off federal borrowing costs, on the order of 90 billion dollars a year. That is the exorbitant privilege, in full. Michael Pettis, who has argued this longer than most, reduced it to a sentence: “Being able to take on debt is not a privilege.”
What the cheap borrowing actually purchased is the point. A government that can place unlimited paper into an automatic bid does not have to win an argument before it acts. The campaign against Iran ran to about 72 billion dollars in two months, better than 1.2 billion a day, with a further 67 billion requested in July. None of it required the country to decide it was worth the money, because the money never presented itself as a constraint.
And now the constraint has arrived from the other side anyway. Net interest on the federal debt reaches roughly a trillion dollars this financial year, against 885 billion for national defence, and it is growing faster than any other line in the budget. Set that against the 90 billion the reserve role is worth. The privilege was never large enough to pay for what it made possible, and the interest bill is what the difference looks like when it finally shows up.
This is where this reading divides from itself, and the division should be stated rather than smoothed over. Miran's remedy is an accord: century bonds, swap lines extended as leverage, tariffs used to persuade creditors to term out their holdings. That is a serious proposal and it is the wrong one. It keeps every discretionary power that caused the problem and adds a negotiation in which they are exercised. Accords, coordinated devaluations and tariffs dressed as monetary policy all leave the same officials with the same instruments, and merely rename them.
The weaponisation is not a separate error either. It is what a state does when it has an instrument it never has to justify. Frank Giustra stated the consequence exactly: “Financial systems are built on trust and, if they are weaponized, they lose the trust necessary to retain their dominance.” Every use is cheap at the time and every use is paid for later, by someone who was not consulted.
Miran concedes the strongest point the other side has, and it is worth conceding with him: the United States is far from a tipping point, in part because there are no meaningful alternatives to the dollar or to Treasury securities. That is right, and it is beside the point. Creditors do not need a rival currency in order to leave. They need an asset with no counterparty, and one exists.
Which is what the tonnage is. Central banks bought 289 tonnes of gold in the second quarter, sixty-two per cent above the year before, and gold has passed both the euro and US Treasuries in official reserves for the first time since 1996. Poland, an ally in no danger from anybody, led the buying. Nobody will announce the end of this arrangement and nobody needs to. The correct answer is not to negotiate a better version of it but to stop requiring one: spend what you tax, borrow what a market will lend at a price it sets, and let the wars be voted on like everything else.
