Bretton Woods was not a stabilization agreement. It was the most expensive standing purchase order in monetary history, and the US Treasury paid the bill. Hear me out.
On July 1, 1944, 730 delegates from 44 Allied nations arrived at the Mount Washington Hotel in Bretton Woods, New Hampshire. The conference ran for twenty-two days. The principal architects were Harry Dexter White of the US Treasury and John Maynard Keynes representing the United Kingdom. The two men disagreed on nearly everything. Keynes wanted an international clearing union with a synthetic reserve currency he called the "bancor" — a supranational unit of account that would prevent any single nation from bearing the structural cost of reserve provision. White wanted a system anchored to the US dollar. White won. The resulting Articles of Agreement established a fixed exchange rate regime in which the US dollar was convertible to gold at $35 per troy ounce, and every other participating currency was pegged to the dollar within a 1% band. The International Monetary Fund and the International Bank for Reconstruction and Development were created as institutional pillars. The architecture looked permanent. Its invoice was already running.
The system is remembered as the foundation of postwar monetary stability. That framing is not wrong. But it omits the cost.
What the Bretton Woods agreement actually created was an open-ended obligation for the US Treasury to sell gold to any foreign central bank that presented dollars, at a fixed price, in unlimited quantity, for as long as the system persisted. The system persisted for twenty-seven years. The transfers were not theoretical. They were physical movements of metal from vaults beneath the Federal Reserve Bank of New York and Fort Knox to the vaults of the Banque de France, the Deutsche Bundesbank, the Bank of Japan, and the Bank of England. Every ounce moved at a price the market had not set and would eventually reject.
The $35 Peg Was a Standing Offer to Drain the US Treasury
At the end of World War II, the United States held approximately 20,205 metric tons of monetary gold — roughly two-thirds of the world's total central bank reserves. The conversion to dollar value requires one constant: one metric ton contains 32,150.7 troy ounces. The math runs as follows. 20,205 metric tons multiplied by 32,150.7 ounces per ton yields 649.8 million ounces. At the fixed conversion rate of $35 per troy ounce, this stockpile represented $22.74 billion. That was the collateral backing the entire postwar monetary order.
The mechanism was direct. Foreign governments ran trade surpluses with the United States. Those surpluses were settled in dollars. Those dollars could be presented to the US Treasury and exchanged for physical gold at the fixed rate. The system did not require foreign central banks to hold dollars. It required the US Treasury to hold gold. The distinction is the entire cost structure.
By 1950, US gold reserves had declined modestly to approximately 20,279 metric tons. By 1960, the figure had fallen to roughly 15,822 metric tons. By 1965, approximately 12,499 metric tons remained. By the morning of August 15, 1971 — the day Richard Nixon suspended gold convertibility in a televised address from Camp David — the US Treasury held 8,133 metric tons.
The net outflow over twenty-seven years: 12,072 metric tons. At the fixed price of $35 per troy ounce, that is 388.2 million ounces transferred, representing $13.59 billion in gold moved from American vaults to foreign central bank reserves. In 1971 dollars, $13.59 billion was approximately 5.6% of US GDP. That was not a modeling exercise. It was an accounts payable line.
Conceding the Strongest Argument Does Not Save the Conclusion
The standard defense of Bretton Woods is that the system delivered a generation of trade expansion, low inflation, and exchange rate predictability that enabled Western reconstruction. This is correct. Between 1948 and 1971, global trade volume grew at an average annual rate exceeding 7%. Western European economies rebuilt their industrial bases. Japan transformed from an occupied agrarian economy into the world's second-largest industrial power. The dollar's reserve status enabled the Marshall Plan's $13.3 billion in European aid, the financing of NATO, and the creation of a US-centered trade architecture that persists, in modified form, to this day.
Concede all of it. The system worked for its stated purpose.
The problem is that "it worked" is not the same as "it was sustainable" or "it was worth what it cost." The economist Robert Triffin identified the structural contradiction in 1960, in testimony before the Joint Economic Committee of the US Congress. His argument, now known as the Triffin Dilemma, is precise: for the dollar to function as the world's reserve currency, the United States must run persistent balance-of-payments deficits to supply the world with dollar liquidity. But persistent deficits erode confidence in the dollar's gold convertibility. The system requires the United States to simultaneously maintain a credible gold-backed currency and serve as a generous supplier of unbacked dollar liquidity. Both conditions cannot hold indefinitely. One breaks.
Triffin was right. By the mid-1960s, the total dollar claims held by foreign central banks exceeded the value of US gold reserves at $35 per ounce. The gap widened every year. President Charles de Gaulle of France made this arithmetic public and operational in a press conference in February 1965, announcing that France would convert its accumulated dollar reserves to gold and advocating a return to a pure gold standard. Between 1965 and 1966, France redeemed approximately $884 million in gold from the US Treasury. Other central banks followed the precedent, more quietly but on the same arithmetic.
The London Gold Pool — a consortium of eight central banks formed in 1961 to defend the $35 price on the open market — collapsed in March 1968 after sustained reserve losses. The US alone contributed an estimated 3,500 metric tons to the Pool's market interventions between 1961 and 1968. At $35 per troy ounce, that is 112.5 million ounces, or $3.94 billion in gold sold into the open market to suppress a price that the market had already decided was too low. The cost of defending the peg was not limited to the gold that left through the convertibility window. It included the gold spent fighting the market's repricing of the metal itself.
The Deepest Line on the Invoice Is Not the Gold
Strip the Bretton Woods system to its operational ledger and the numbers read as follows. The US Treasury began with 20,205 metric tons in 1945 and ended with 8,133 metric tons in 1971. Net transfer: 12,072 metric tons, or 388.2 million troy ounces, at a fixed price of $35 per ounce — $13.59 billion in nominal value. But the fixed price was itself a subsidy. On the day Nixon closed the gold window, the London free-market gold fix stood at approximately $42 per ounce. At that price, the 12,072 metric tons already transferred were worth $16.3 billion — meaning the US had sold them at a 17% discount to the price prevailing at the moment of suspension. By December 1974, when gold traded above $180 per ounce, the same tonnage would have been valued at $69.9 billion. By that measure, the United States sold $69.9 billion worth of gold for $13.59 billion over the life of the system. The delta — $56.3 billion — was the implicit subsidy to foreign central banks embedded in the fixed-price convertibility mechanism.
But gold outflow was only the visible line item. The hidden cost was policy constraint. To defend the peg, the Federal Reserve could not set interest rates purely in response to domestic economic conditions. Every rate decision carried a balance-of-payments dimension. Lower rates risked capital outflow, which accelerated gold redemptions. The Fed operated under a dual mandate before the term existed — managing domestic employment and inflation while simultaneously defending an international gold-convertibility window that moved in the opposite direction. The opportunity cost of that constraint across twenty-seven years of monetary policy decisions does not reduce to a single number. But the direction is unambiguous: US monetary flexibility was structurally impaired for the duration of the system.
This started as a reconstruction of July 1944 at the Mount Washington Hotel — twenty-two days of negotiation, a system of fixed rates and gold convertibility, the institutional architecture of postwar monetary order. It turned into an invoice. The invoice shows $13.59 billion in gold at the fixed price, over $56 billion in implicit subsidy measured against prices within three years of the system's collapse, and an unquantifiable cost in monetary policy autonomy surrendered to a peg that Triffin proved structurally unsustainable a full eleven years before it finally broke. Whether a generation of trade stability was worth that invoice — or whether the reconstruction would have proceeded anyway, because the underlying economic conditions of postwar demand and American industrial surplus made it inevitable with or without a gold anchor — is a question the archival record raises but does not settle. If you have a framework that resolves it, the literature has been waiting since 1971.