1. A World That Never Settles

The Articles of Agreement signed at the International Monetary Fund in 1944 contain one sentence of striking plainness: the par value of every member's currency was to be expressed in terms of gold as a common denominator, or in terms of the weight and fineness of the United States dollar as of July 1, 1944.

Common denominator.

Those two words say everything about what the project actually was. Forty-four countries, each with its own sharply drawn borders, its own wildly uneven creditworthiness, each just emerging from a war that had left a great tangle of unsettled accounts running between them — and the task this arrangement set for itself was to press all of their currencies onto a single scale, so that debts, surpluses, exchange rates, and reserves could all be translated into one common ledger.

To feel the size of that difficulty, consider what exactly was being asked to become commensurable. A nation's currency is not a commodity; standing behind it is that country's capacity to tax, its central bank's willingness to show restraint, whether its parliament will still be in the same hands next year, whether it can win a war it might have to fight. None of these things convert into one another by any natural rate. And yet that was precisely the demand packed into the phrase common denominator: to give them a rate of conversion where none existed.

Essay Fourteen told the story of an earlier version of the same ambition. That earlier arrangement had defined each currency as a fixed weight of gold, so that dividing one legal definition by another yielded an exchange rate precise to the seventh decimal place. And that whole structure had come apart between 1931 and 1933.

Eleven years later, the same undertaking was attempted again — and this time, more carefully.

The care showed above all in one thing: this time the system did not pretend to be automatic.

The postwar order did not revive the nineteenth-century classical gold standard, in which any private citizen could walk up to a counter and exchange paper for metal. An ordinary holder of dollars could not convert them into gold.

That difference was not a small one, though it rarely survives into the textbook formula. What had made the gold standard so persuasive was precisely that its promise held for every single person carrying paper money: if you don't believe it, go and exchange it. That sentence was the whole source of the old system's credibility, even though almost nobody ever actually walked up and did it. Under the postwar arrangement, that sentence no longer held for individuals at all. Convertibility had turned from a right available to anyone into an arrangement conducted strictly between states.

The party actually bound to honor the promise of gold was the United States government, and the counterparties it owed that promise to were, above all, foreign official holders and foreign central banks. The Articles of Agreement, in the same breath, explicitly permitted member states to control capital movements, so long as such controls did not restrict payments for current transactions.

In other words, that seemingly objective metallic constraint was, from day one, hedged in on every side by sovereign judgment, central-bank cooperation, diplomatic bargaining, and administrative discretion.

This is what determines what it actually was that got cut in 1971.

What was severed was not some metal chain that had never been touched by anything else. What was severed was a rope that credit, agreements, and politics had already knotted and reknotted many times over.

That distinction decides how the whole story ought to be told. If the chain had truly been pure metal, then 1971 was a rupture — a passage from constraint into its absence. If it had always been a knotted rope, then 1971 was something else entirely: it was the moment the knots were made visible.

Across the sixteen essays before this one, nine distinct methods have appeared by which a construct deals with its remainder. What follows here is not a tenth method. It is something else: the moment the construct admitted, in its own voice, the very thing it had refused to admit all along.

2. Two Plans for One Ledger

The conference ran from July 1 to July 22, 1944, at a hotel in Bretton Woods, New Hampshire, drawing forty-four nations and roughly seven hundred and thirty delegates. The United States Secretary of the Treasury presided. Britain's delegate was John Maynard Keynes; the principal architect on the American side was Harry Dexter White of the Treasury.

On its surface this was a multilateral conference. In truth there were only two draft plans on the table, and they were rivals.

Keynes's plan was called the International Clearing Union. It is best remembered for proposing a supranational unit of account, to which he gave a name. But the real stakes were never in the name. They were in what he wanted international payments to become.

In his own words, the aim was to generalize the essential principles of banking to the international plane.

In practice this meant every nation's central bank would hold an account within the Union. Surplus countries would appear as creditors, deficit countries as debtors. The Union would run on bookkeeping entries and overdrafts, so that no country would be forced, merely because it was short of gold or dollars at a given moment, to compress its imports and compress its employment immediately. Each country would receive a quota set at roughly three-quarters of its average trade turnover over the prior three years, and could run its account into overdraft up to the limit of that quota.

And for countries that built up enormous surpluses over the long run, Keynes proposed that pressure be applied to them as well.

This was the true heart of the whole proposal. The goal he set down in writing was to bring an expansionary rather than a contractionary pressure to bear on world trade.

At the time this was a genuinely new proposition. For centuries before it, the ledger of international payments had only ever pushed in one direction: whoever owed money had to tighten its own belt. What Keynes wanted to add to that ledger was a counterpressure running the other way, so that when the accounts failed to balance, the only available remedy would not be one that dragged everyone down together.

This was not abstract rhetoric. Behind it lay the weight of memory from the years between the two world wars: deflation, unemployment, collapsing incomes, and the Britain of Essay Fourteen, which had been able to defend its parity only by grinding down its own wages.

To see clearly what Keynes was really doing here, one has to return to a very plain question: when a nation's balance-of-payments account refuses to balance, whose job is it to set it right.

The traditional answer pointed in only one direction. The deficit country did the forcing. It raised interest rates, suppressed demand, restricted imports, drove its own prices and wages downward, until the books balanced again. The surplus country had to do nothing at all, because it had done nothing wrong.

This was exactly what Keynes wanted to change. His argument was that an unbalanced account is a condition the two sides produce together, not a fault belonging to one side alone. Loading the entire burden of adjustment onto the deficit country might work out on paper, but in its consequences it simply pushed the whole cost onto the people within that country with the least bargaining power of all.

White's plan was a different animal. It looked much more like a finite pool of resources than like a world central bank.

Under his design, the Fund would not conjure international money out of nothing; instead, member countries would pay in gold, their own currencies, and securities in advance, forming a common pool. His draft specified that the initial subscriptions should total no less than five billion dollars, with a quarter of each nation's contribution paid in cash, a quarter in interest-bearing government securities, and at least half of the cash portion paid in gold.

What White wanted was a controllable, bounded stabilization mechanism resting on national subscriptions. It could supply short-term liquidity. What it would not do was hand over the power to create world reserves to a supranational clearing union that could run enormous overdrafts and, worse, turn around and constrain the surplus countries themselves.

What was finally agreed upon leaned far closer to White, carrying only fragments of Keynes.

Britain gave up the unit-of-account currency, gave up the Clearing Union, gave up the more generous overdraft arrangements, and accepted a structure built around the dollar and gold.

This outcome is not hard to understand, and it needs no conspiracy to explain it. The war had just ended, the gold sat in American vaults, the dollar was what everyone needed, and Britain was deep in debt. For any plan to be adopted, the side that actually held the money had to say yes. What Keynes was trying to do had its own logic — but logic was not the only thing operating at that table.

Of the traces of Keynes that did survive, the most significant was the so-called scarce currency clause: if a surplus country's currency became persistently scarce, the Fund could ration it, and other countries would be permitted to impose limited import restrictions in response.

In the years that followed, this clause was almost never actually invoked.

Its importance never lay in its being invoked. Its importance lies in the fact that it recorded, in writing, a disagreement from the conference floor that had never been entirely erased: the question of exactly who should bear the burden of adjusting for global imbalances.

A clause written into the text of an agreement and then almost never used is a peculiar kind of object. It is a place the construct leaves inside its own text, marking the spot where another road had once been possible.

Essay Nine described something close to this: a construct can decide what one generation reads, but it cannot decide forever, because whatever gets cut and whoever did the cutting end up bound inside the same cover. What happens here goes further still. Nothing was cut at all. It sits printed in the middle of the official text, available for anyone to turn to; it is simply that no one ever has. A clause that is never used and a clause that never existed produce almost the same effect in practice — and an entirely different one in the record.

3. Fixed, But Adjustable

What the conference produced was not a return to the gold standard. What it built was a hierarchy.

At the top sat gold. In the middle sat the dollar, pegged to gold at the official price of thirty-five dollars an ounce. Below that sat every other national currency, each assigned a fixed but adjustable parity against the dollar, held in the market within a narrow band of roughly one percent above or below that parity.

The whole difficulty of the system was packed into that phrase, fixed but adjustable. Make it wholly fixed, and there is no escape hatch — a country that cannot hold the line has nowhere to turn but to squeeze its own domestic economy. Make it wholly adjustable, and fixity means nothing — nobody can say what tomorrow's rate will be. So the system chose the middle path: fixed in ordinary times, open to negotiation in the event of a fundamental disequilibrium. And as for who was to judge whether a disequilibrium counted as fundamental, the Articles of Agreement supplied no formula anyone could simply apply.

Every member, upon joining the Fund, was assigned a quota, which fixed both its subscription obligation and how much it could in turn draw from the Fund's resources. The total quotas finally agreed upon came to roughly eight billion eight hundred million dollars, with the general rule that a quarter be paid in gold and three-quarters in the member's own currency.

On paper the mechanism was elegant. A deficit country would first draw on its own reserves and adjust its domestic policy; if short-term difficulties persisted, it could draw foreign exchange from the Fund; and if the imbalance proved fundamental, it could apply to adjust its parity.

And yet, within this arrangement built to pursue a common scale, room for national policy discretion was left in on purpose. Article VI of the Articles of Agreement states in plain language that a member may exercise such controls as are necessary to regulate international capital movements.

This clause was a lesson learned from the 1930s, and it is the single most honest piece of the entire design. It amounts to an admission that no country can have all three things at once — a fixed exchange rate, capital free to move in and out, and a domestic policy of its own choosing. One of the three has to be given up. In the 1930s, what got given up was domestic policy, and Essay Fourteen has already counted the cost of that choice. This time, what was given up was the free movement of capital.

What was being given up was clear enough to the people making the choice at the time. Locking down capital meant that moving money abroad now required a reason, an application, a wait; it meant somebody earned less, and some deals simply did not get made. Those costs were real. What was bought in exchange was that a government could decide, in a given year, to spend a little more to protect employment, without waking up the next morning to find its own currency under attack. Which side of that trade was worth more was argued over at the time, and the argument has never entirely stopped.

So this was never a pure market order. It was a market order carefully managed by states and central banks: exchange rates stable enough for trade, so that prices could be compared across borders; capital flows that could be locked down, so that governments could pursue employment and growth without instantly triggering a currency crisis.

This design also carried an asymmetry that is often overlooked.

Postwar paper currencies were not, as a rule, convertible into gold; the only party truly committed to official gold convertibility was the United States Treasury. Every other currency's capacity for anonymous settlement had, in the end, to pass through the dollar — and the dollar, in turn, had to pass through a single nation on its way to gold.

At the war's end, the United States held most of the world's official gold, while Europe and Japan desperately needed dollars to import, rebuild, settle accounts, and repay debts. And so, in those first years, the common scale and trust in the United States were almost perfectly identical.

It was precisely because the two coincided that so many people later came to mistake the system's foundation for pure metal.

It had, in truth, always been a mixture of metal and credit. It is only that, in those early years, no one had any reason to work out the proportions.

When two different things give the same answer for long enough, people come to believe they are one and the same thing. Telling them apart requires waiting for the day they give different answers — and by the time that day arrives, it is usually too late to rearrange anything.

4. The Riddle in the Vault

The real trouble only began to take shape once the system had actually succeeded in working.

Western Europe restored current-account convertibility in 1958; Japan achieved convertibility in 1964. Only at that point could Bretton Woods be said to have entered its fully operational phase.

It was exactly at this moment that one economist's warning began to sound sharp.

He testified before Congress in 1959 and published a book in 1960, and the logic he set out was simple. If the United States stopped exporting dollars, the world would face deflation and contracting trade for want of reserves; if the United States kept exporting dollars, then sooner or later foreign official claims on dollars would exceed the American gold stock, the market would come to doubt that the thirty-five-dollar gold price could hold, and a run on gold would follow.

Both roads led to trouble.

The shape of this dilemma is worth holding onto, because it does not belong to money alone. Whenever a single thing is made to serve two purposes at once, it falls into exactly this structure: the dollar had to be the world's liquidity and, simultaneously, America's liability. The first role demanded there be more and more of it; the second demanded there be less and less. Whenever a construct assigns itself two roles whose demands point in opposite directions, the more successfully it performs them, the closer it comes to failure.

This was no longer a problem confined to the seminar room, because the dates involved are exact and verifiable.

By August 1960, total outstanding American dollar liabilities to foreigners already exceeded the U.S. gold stock. By December 1965, even dollar liabilities held solely by foreign official institutions had exceeded the gold stock. By 1966, this was simply established fact.

America's monetary gold stock fell from roughly twenty billion six hundred million dollars in January 1947 to roughly eleven billion one hundred million dollars in December 1970.

The common scale did not snap all at once. It simply came to resemble, more and more, an international bank operating on fractional reserves: the convertibility it had promised on its books had grown larger than the metal sitting in its vault.

It should be said clearly that fractional reserves are not, in themselves, a flaw. Every bank operates this way, and it is exactly this arrangement that lets deposits become loans and lets money actually move. Its only requirement is that not everyone shows up demanding payment on the same day. This was precisely the position the whole arrangement found itself in after 1960: there was no single day on which the arithmetic failed outright; there was, from that point forward, simply every single day depending on nobody deciding to ask for payment all at once.

No country simply sat waiting for a run to happen. Patch followed patch.

In November 1961, eight Western central banks established the London Gold Pool, jointly buying and selling gold on the London market to hold its price to thirty-five dollars an ounce; the initial selling limit was two hundred and seventy million dollars, and it was expanded again and again thereafter.

In the same period, starting in 1962, the American central bank greatly expanded its network of currency swap arrangements, using temporary swaps and short-term credit lines to cover the positions of foreign central banks that preferred, for the moment, not to convert their dollars into gold. By the time the gold window closed in 1971, the total swap capacity had grown from an initial nine hundred million dollars to eleven billion two hundred million dollars.

Both of these facts point to the same underlying reality. On its surface the system appeared to run on anonymous, impersonal rules; underneath, it was leaning more and more heavily on mutual guarantees among central banks, on unspoken understanding, and on improvised firefighting.

And in the midst of all this, one phenomenon deserves to be singled out.

Later researchers found that the postwar central banks did not, as some of the system's designers had expected, gradually let go of gold. On the contrary, they went on steadily rebuilding and holding onto their gold reserves.

On the books, this made no financial sense. Gold pays no interest, and holding it carries an opportunity cost, whereas dollars deposited in the United States earn interest. And yet gold went on being treated as a cleaner backstop asset than any other country's credit.

Researchers came to call this the gold puzzle of Bretton Woods.

In name, the system had already made the dollar, not gold, the primary medium of international reserves. In practice, official institutions everywhere still refused to hand their final insurance policy over to the dollar entirely.

This is the first place in this particular stretch of history where the remainder shows its face — and it shows up in a spot almost nobody thinks to look: not on the excluded side of the ledger, but on the balance sheet of the construct's own people.

Those central banks settled in dollars every single day, quoted at parity every single day, publicly defended the system every single day. And at the very same time, they kept something in reserve down in their vaults, against the day this common scale might stop counting for anything.

In their public statements they accepted the scale; with their own hands they kept a second option quietly aside. Essay Six described one form the remainder can take: a second scale that the market keeps hidden away, silent in ordinary times, surfacing only once the construct's credit fails. What is different this time is that the people keeping that second scale in reserve were exactly the same people using the first scale every single day.

5. Two Prices for One Metal

France was the one to make that suspicion public.

On February 4, 1965, Charles de Gaulle argued at a press conference that the world's monetary system ought to return, by stages, to a gold basis, and that nations bearing international financial responsibilities ought to settle their balance-of-payments debts in gold rather than in paper claims.

Around the same time, France's finance minister summed up in a single phrase the advantage that the dollar's central position conferred on the United States: an exorbitant privilege. What he meant was that America could settle its accounts using liabilities of its own issuance, of the very kind that every other country had to earn through real goods, real services, or genuinely high-quality assets.

France's position was not necessarily driven by principle alone; geopolitics and national standing were mixed in as well. But it did put its finger squarely on the most sensitive point in the entire system: what passed for anonymous international money was, in truth, one particular nation's promissory note.

Between 1967 and 1968, the cracks went from visible to unstoppable.

On November 18, 1967, Britain moved the pound's parity against the dollar from 2.80 to 2.40, a devaluation of roughly fourteen percent. Pressure on both the dollar and gold rose sharply in its wake.

In March 1968, the London Gold Pool collapsed.

After the Pool disbanded, the central banks of seven nations agreed to establish a two-tier gold market: official transactions among governments would continue at the fixed official price, while the private market would be left to float freely, and the central banks pledged that they would no longer sell monetary gold into that private market.

On the eighteenth of that same month, the United States formally repealed the statutory requirement that Federal Reserve notes be backed by a twenty-five percent gold cover.

This step deserves to be looked at closely.

A system originally meant to tie the price of gold, exchange rates, reserves, and payments to one single scale had, for the first time, been split apart by official hands themselves into a world of two prices.

The official price remained the language of the rules: it was written into the Articles of Agreement, used in settlements between central banks, and every formal account was still kept by it. The private market price, meanwhile, began expressing a different judgment — and a far more anxious one.

It is worth glancing back at Essay Six here. The users of jiaozi paper currency, in that earlier case, had kept a second scale hidden away precisely because they did not fully trust the government's paper. This time the two-tier market was not something the public smuggled into being on its own; it was drawn by official hands. The authorities themselves acknowledged that a second price existed, and set aside a lawful space for it to float in. The construct did not merely know there was another scale out there — it went so far as to mark off the ground for it to stand on.

The coexistence of two prices was not a mere technical convenience. It was the first time the construct openly admitted that the number it had set and the number the market was willing to pay were no longer the same thing.

Essay Ten observed that the easiest thing for a construct to do with whatever it cannot answer is to draw a line and put it outside the door. What happened here follows the same shape, only more nakedly: it was not a category of problem being pushed outside, but a second price for the very same thing, sent out the door to float freely, out of sight and, it was hoped, out of mind.

This solution carried one precondition: the number outside the door could never be brought back in and used inside. That is exactly why the pledge was necessary. A gate had to separate the two worlds, or arbitrage would blow the official price apart within a single day.

It was at exactly this stage that the Fund introduced a new international reserve asset, called Special Drawing Rights.

The official description is unambiguous: it was a supplementary international reserve asset, created in 1969 to supplement members' official reserves. The first allocation of three billion units took effect on January 1, 1970, and across the whole period from 1970 to 1972, a total of nine billion three hundred million units were allocated.

This was a reform heavy with symbolism. At a moment when gold was insufficient and dollars were both too abundant and too fragile, the system tried to manufacture yet another international reserve unit — one more abstract, more institutional than either.

It expressed, once again, the same underlying impulse to force value into a single unified scale. And at the same time it demonstrated that neither gold alone nor the dollar alone was any longer enough to sustain that scale.

The fate of this new unit is also worth noting. It had no metallic body, no single nation standing behind it; it was simply a number that a group of countries had produced together through joint bookkeeping. It was more abstract, and in a sense more pure — and it never grew into the world's principal currency. Whether something can serve as a scale, it seems, depends on more than how cleanly it has been defined.

There was one more thing that slipped rapidly out of control in the second half of the 1960s.

Even as the American government tried to restrict capital outflows, multinational corporations and banks increasingly found ways around the controls, moving dollars out into banking systems beyond America's borders to operate there instead. By around 1969, Germany's central bank had already come to regard this offshore dollar market as a significant factor undermining its own ability to constrain domestic credit.

The clause in the Articles of Agreement permitting capital controls remained, on paper, entirely intact. Capital, meanwhile, had already found a place that existed beyond the paper altogether.

The measurable reserves and deficits recorded on official balance sheets found themselves increasingly surrounded by financial chains that were off the books, offshore, cross-border, and far more dependent on personal relationships.

This was a textbook case of overflow. A construct draws up a rule, and that rule governs only the transactions it knows how to define; capital simply flows toward whatever the rule fails to define, not out of any deliberate intent to break the law, but only because that space is unregulated, and therefore cheap. The tighter the grip, the stronger the pull toward the outside. The boundary drawn between organizational forms in Essay Twelve reappears here in a new setting: the very same dollar was regulated if it sat in New York and unregulated if it sat in London — and it remained, the entire time, the very same dollar.

6. A Word Called Temporary

By 1971, the very nature of the crisis had changed.

The question was no longer simply whether the United States could hold the thirty-five-dollar gold price. It was whether other countries were still willing to go on playing their part in keeping the whole performance running.

By late April 1971, key figures at Germany's central bank and government were already discussing with the Managing Director of the Fund the possibility of letting the mark float. On May 6, Germany's Minister of Economics ordered the mark to float temporarily.

Once a major surplus country refused to keep defending a fixed-parity system, that system's troubles stopped being America's problem alone.

At the same time, the sheer scale of capital movement suddenly ballooned. America's net capital outflow was twelve billion nine hundred million dollars in 1970 and twenty-nine billion six hundred million dollars in 1971, of which eleven billion occurred in the third quarter alone.

Which is to say, even setting aside the entire question of whether the official gold stock was adequate, the shock of liquidity alone had already rendered the twin promise of fixed parities and official gold convertibility acutely unstable.

From August 13 to 15, 1971, President Richard Nixon gathered his core economic advisers at Camp David. Present were the Secretary of the Treasury, the Chairman of the central bank, and the officials responsible for international monetary affairs. What was to be settled that weekend was an entire new package of economic policy.

This was not a moment of the market simply evolving on its own. It was a highly politicized weekend of decision-making, mixed through with considerations of elections, trade, inflation, and international negotiating position all at once.

On the evening of August 15, Nixon announced in a nationally televised address: I have directed the Secretary of the Treasury to suspend temporarily the convertibility of the dollar into gold or other reserve assets.

The word temporarily was never, in the end, made good on.

It should be said plainly, though, that the word may not have been mere evasion at the time. The domestic measures announced alongside it really were designed as a temporary arrangement, with a ninety-day freeze spelled out explicitly; and on the international side, the thinking at the time seems mostly to have been: stop for now, in order to negotiate from a stronger position afterward. That things never went back does not mean going back was never the intention. When a construct shuts a door, it is often quite sincere in believing it has only left it ajar.

He wrapped this decision inside a whole package of domestic policy at the same time: a ninety-day freeze on wages and prices, a ten-percent import surcharge, spending cuts, tax adjustments. He framed the whole crisis as a defense against international currency speculators.

But the two most consequential sentences in that entire address were two others.

These two sentences carried the real weight of that day, and at the time they hardly stood out at all. The audience was worried about whether prices would freeze, whether imported goods would get more expensive, what their own wages would be worth the next morning. And in the midst of these practical concerns, the very definition of money was being quietly swapped out.

The first sentence: the strength of a nation's currency depends on the strength of that nation's economy.

The second sentence: he assured foreign audiences that the United States would remain a forward-looking, trustworthy trading partner.

Read those two sentences together, and what actually happened that day comes into view.

The foundation of money was openly redefined, moving from convertibility into gold to the strength of an economy and the trustworthiness of a nation.

Everything that metal had once concealed — trust in one particular country, a judgment about its policy discipline, a bet placed on its reputation — no longer hid behind the currency. The President himself named it, in so many words, as the currency's front-facing foundation.

Essay Fifteen observed that once a scale has been rewritten a single time, that fact can never be taken back. In 1933 the rewriting had moved the gold price from twenty dollars and sixty-seven cents to thirty-five dollars. This time it was not a number being changed; it was the promise that number represented being canceled outright, with a whole new account substituted in public to hold up the very same piece of paper.

There is one more passage in the speech that is often skipped over, and it should not be.

Nixon named retirees, homemakers, and wage-earning workers as the victims of inflation, and packaged the wage freeze and the tariff as measures to protect their jobs and their purchasing power.

Whether or not one accepts that framing, it demonstrates one thing: the international monetary system was never merely an abstract chessboard.

When a fixed exchange rate cannot be held, the cost lands on wages, on import prices, on employment, on budgets. And in order to defend a given parity, a deficit country typically has to raise interest rates, suppress demand, restrict imports — and what it ultimately absorbs are real, concrete losses of jobs and income.

Here, more clearly than anywhere else, one can see exactly why Keynes had so insisted on refusing to load the entire burden of adjustment onto the deficit side. Twenty-seven years later, the very question that had been voted down came back — and it came back inside a speech given by the president of the very country that had voted it down in the first place.

What that abandoned plan had set out to do was make surplus countries shoulder part of the burden of adjustment too. Twenty-seven years on, the United States itself had become the party being asked to adjust, and the remedy it reached for was to shut the window, add a ten-percent import surcharge, and push the pressure back onto everyone else. It did not adopt Keynes's remedy; it used one of its own — to solve precisely the problem Keynes had warned, all those years before, would eventually arise.

7. Three Ways to Tell the Same Story

On the subject of 1971, researchers have produced roughly three competing lines of explanation, and the disagreement among them is not a matter of taste — it runs through the question itself.

The first school holds that the collapse was structural, and unavoidable.

Its earliest and most influential formulation is the dilemma set out earlier by that same economist: a reserve currency's issuing country must keep supplying the world with liquidity, and doing so must eventually push its external liabilities past what its gold can support, triggering a crisis of confidence. Later research in this tradition has tended to trace the collapse to a double flaw in the system itself: on one side, reserve growth had become excessively dependent on the dollar; on the other, the United States, as the central country, turned after 1965 toward a more inflationary combination of fiscal and monetary policy, and failed to observe the rules that would have kept inflation low and the credibility of the gold price intact.

On this view, the closing of the gold window certainly had its specific triggers, but the overall direction had already been set by the system's internal contradictions and by the breakdown of American policy discipline.

The second school holds that the collapse was not, in fact, inevitable — that the system could have gone on.

Already in 1966, several researchers had raised objections. They argued that the system was not bound to collapse: the United States functioned much like the world's financial intermediary, issuing short-term, highly liquid dollar liabilities while holding long-term assets abroad; so long as the United States maintained price stability, this arrangement could in principle continue indefinitely, and might even evolve into a dollar standard without gold at all. Documents that later revisited this debate put the point plainly: on this view, had American authorities maintained the price stability that a nominal anchor requires, the system could have gone on without any fixed end date.

This school did not deny that risk existed. What it denied was something else entirely: that dollar liabilities exceeding gold automatically meant the system was immediately unsustainable.

The nuance here has to be gotten exactly right. This is not an argument that liabilities exceeding gold did not matter. It is an argument that whether a promise of convertibility can hold depends on whether everyone shows up to redeem it at once; and whether they all show up at once depends on whether they believe the promise; and whether they believe it depends, above all, on whether the party making the promise has kept its own discipline. In other words, the decisive variable was never sitting in the vault. It was somewhere else entirely.

The third school places its emphasis on political choice and historical contingency.

Some researchers have objected to writing the outcome of the late 1960s as a foregone conclusion, arguing instead that the parity system was killed by mutual recrimination over exactly who should bear the burden of adjustment. Other accounts summarizing this line of argument stress further that, in 1971, American domestic anxiety over the trade deficit, over Japanese and European export competitiveness, congressional pressure, and worry over the coming year's election all combined to turn the exchange-rate question into something more than a technical monetary matter — it became a weapon in market access and domestic politics alike.

There is, in addition, a diplomatic-history approach that situates the monetary conflict within the alliance politics of the Cold War: the dollar, gold, military bases, defense spending, foreign aid, and burden-sharing among allies were never independent questions from one another.

On this view, 1971 certainly had structural conditions behind it, but the particular day the gold window closed, the manner of its closing, and the additional measures that accompanied it were never the sole possible outcome of pure technical logic.

No verdict will be rendered here. What makes these three schools of explanation matter is that they correspond to two altogether different understandings of monetary order itself.

If Bretton Woods is understood as an objective mechanism, truly grounded in gold and merely transmitted through the dollar, then 1971 looks much more like a physical rupture that was bound to arrive sooner or later.

If it is understood instead as a political and financial hybrid, using metal as its symbol while resting on central-bank cooperation and trust in American credit, then 1971 looks much more like a rebalancing of credit: gold did not evaporate overnight, and the dollar did not lose its international role overnight either. What genuinely changed was the weighting between the two, together with the world's open acknowledgment of what final settlement actually rested on.

Both readings can account for exactly the same set of facts without contradiction. And they correspond to two different worlds: in one, the construct is ultimately at the mercy of something it cannot control; in the other, the construct has always been held up by people, and metal was only ever one of the ways they chose to hold it up.

8. What Was Left After the Metal

Closing the gold window did not mean the world tidily entered an age of free floating rates the very next day. Countries first tried to save the system.

On December 17 and 18, 1971, the Group of Ten reached an agreement in Washington: the United States agreed to move the official dollar price of gold from thirty-five dollars to thirty-eight dollars, a devaluation of roughly eight and a half percent, while the other major countries revalued their own currencies accordingly and widened the permitted band of fluctuation.

The trouble was that this adjustment restored no gold convertibility, and settled nothing about who would restrain American inflation or who would bear the risk of continuing to hold dollars. The later official history summed up the whole effort in four words: too little, too late.

It survived only a short while. Around March 1973, Europe was hit by another severe foreign-exchange crisis, after which the fixed-rate system was abolished in all but name. Between February and March of 1973, the major industrial countries turned, almost universally, to floating exchange rates.

These two dates, then, need to be kept distinct. What ended on August 15, 1971, was the official convertibility of the dollar into gold. What ended in the spring of 1973 was what remained of the adjustable fixed-rate framework among the major industrial countries.

The legal reckoning came later still. The 1976 amendment recognized that members could adopt whatever exchange-rate arrangement they judged suitable, while the Fund itself turned to surveillance over members' exchange-rate policies. The current Article IV, Section 2, states plainly that a member may adopt exchange arrangements of its choice, consistent with the purposes of the Fund. That amendment took effect on April 1, 1978.

Only at this point did the postwar constitutional ambition of using gold as a common denominator to fix exchange rates finally, in law, give way.

It is now possible to answer the question posed at the start: what, exactly, was cut in 1971.

The easiest answer is that the world passed, from that day forward, into an age of fiat money — moving from metal to credit.

That answer is inaccurate, and it obscures exactly the thing that matters most.

The more accurate answer is that money was set down, openly and more deeply than before, upon credit.

The anonymous medium of exchange did not disappear. The dollar went on serving as the core medium for international pricing, reserves, and settlement. Some scholars hold that what persisted after the gold window closed was exactly a dollar standard without gold.

What changed was that the ultimate foundation of that medium no longer disguised itself as a purely metallic constraint. It came to rest, ever more openly, on a handful of concrete things: whether America's fiscal authorities and central bank could maintain price stability, whether the American government was still considered trustworthy, whether the Fund's surveillance carried any real force, whether the major central banks still chose to cooperate with one another.

That inward-facing force, personal and relational in nature — trust, reputation, credibility — did not replace anonymous money.

It simply returned, more nakedly than before, to sit at the very foundation of anonymous money itself.

So 1971 should not be written up as some inevitable story of progress from barbaric metal to modern credit.

Gold did not immediately vanish from national balance sheets. Quite the opposite — many central banks went on treating it as their final insurance policy, a practice that had already been under way since the 1950s and did not wait for 1971 to begin. Nor did the dollar step down from the center of the international system merely because it stopped being convertible into gold; on the contrary, it went on dominating international finance even after its demetallization.

What truly changed was the position of the pendulum. Unified pricing, standardized valuation, anonymous settlement — all of it remained in place; but the final guarantee behind these mechanisms now attached itself, more visibly than before, to the credit of one nation and the reputation of a handful of institutions.

The tether that was cut was never, to begin with, a rope whose purpose was to make personal, relational factors disappear.

What it had always been was something that covered those factors over.

A cover is not a deception, and the two need to be kept separate. The system's designers were not hiding anything. What that cover did was spare everyone the daily work of judging whether some other country could be trusted — all anyone had to do was look at a single number. This was an enormous saving of effort, and for a very long stretch of time it genuinely worked. The price of that saved effort was that once the number stopped being reliable, everyone had to relearn, all at once, something they had not had to do for more than twenty years.

1971 only cut that cover open.

Once it was cut open, the forces that had been operating all along were, for the first time, put out in front where everyone could see them: which nation's word could be counted on, who was backstopping whom, which institutions were still willing to call each other in the middle of the night. None of this had ever stopped functioning; it had only, until then, been hiding behind a number that looked as though it had simply been dug up out of the ground.

The behavior of those central banks described in Section Four can now be given a fuller account.

It was not that they disbelieved the system. They used it every single day. They simply understood that any scale erected by human hands will, someday, be set down by human hands as well — and that before that day arrived, there had better be something else sitting in the vault.

Events proved them right — and proved them right roughly a decade before anyone else caught up to it.

This episode reveals something about the nature of the remainder as well. In the earlier essays, the remainder has usually sat on the side being squeezed — voiceless, unable to produce evidence, or simply left off the books entirely. This time the remainder sat in the hands of the very people with the most resources, people fully capable of speaking up, who chose instead to say nothing and quietly keep a reserve down in their own vaults. The remainder does not have to belong to the weak. It is whatever any given party keeps set aside outside the scale, on their own account — and the people with the means to keep such a reserve are, not by coincidence, exactly the ones who understand best how that scale was built in the first place.

And that scale did not disappear because of any of this. It changed foundations and kept right on measuring. The world still needed one common unit of account, still needed somebody to perform the final settlement, still needed a price set on everything, every single morning.

It is only that, from that evening in August 1971 onward, nobody could claim any longer that this price grew out of the weight of metal. It grew, instead, out of whether a group of people remained willing to keep honoring the account.

The ledger has not yet balanced. It is still being kept.