1. Seven Places Past the Decimal

There was once a rate of exchange between the pound sterling and the American dollar, and it read: one pound equaled 4.8665635 dollars.

Seven places after the decimal point.

That number was not estimated. It was defined. Britain had fixed the pound as a certain weight and fineness of gold; the United States had fixed the dollar as a certain weight and fineness of gold; divide one definition by the other and this is the figure that comes out. It required no transaction to hold it up, and no single day's trading could move it. It was the product of two statutes.

The product of two statutes is a phrase worth sitting with. It means the rate never had to consult a market, never had to ask a single merchant's opinion, never had to wait for a trade to actually occur. It became true the moment it was written down, and it stayed true every day afterward, whether or not anyone that day actually carried pounds across the water to become dollars. This is the thing the construct has always been reaching for: a number that holds without needing any particular event to hold it up.

For an exchange rate to arrive at that condition was, in the whole history of the ledger, a first. The scales examined across the preceding thirteen essays were all still dealing with something concrete — barley recorded on a clay tablet, silver coin that passed current inside a city wall and shattered into scrap the moment it crossed one, paper that needed a magistrate's office to stand behind it, a common field that could only be kept honest if every neighbor could see it. Here, for the first time, a scale no longer needed any particular person to vouch for it. It needed only two statute books and a metal.

But a fixed exchange rate is not a line. It is a band.

Shipping gold from London to New York cost money: freight, insurance, packing and handling, the interest lost on the days it sat in transit, a premium for risk, and a margin of profit for whoever was doing the shipping. Add all of this together and it opens a gap on either side of parity. Push the exchange rate above the upper edge of that gap and a private trader has an incentive — buy gold, load it, ship it across, convert it into the other currency, and pocket the spread. Push the rate below the lower edge and the same trade runs in reverse.

It is worth pausing on what actually set the width of that band. It was not decided by which country was stronger, and it was not decided by whose credit stood higher. It was decided by freight charges, insurance premiums, and the interest accrued over however many days a shipment spent at sea. The most flexible part of the whole arrangement, in other words, took its calibration from a shipping company's rate card. The tolerance built into the most abstract scale the world had yet produced was set, in practice, by cargo vessels and insurance policies.

So the exchange rate returned to its band without anyone issuing an order. The arrangement looked close to perfect, and its most elegant feature was this: the hand that pulled it back into line belonged not to any official, but to a class of merchants who wanted nothing more than a profit.

There is one thing, though, that has to be said plainly here, because it is the first nail driven into everything that follows.

A merchant was only willing to run that arbitrage because he believed something: that the authority on this end would truly sell him gold at the legal price, that the authority on the far end would truly take gold off his hands at the legal price, and that no government would suddenly clamp on exchange controls before his ship made port.

Strip out that belief, and the gold points are nothing but two numbers on a page.

Push the thought one step further. The first link in this whole self-acting mechanism was never the metal. It was expectation. The metal only does its work once people believe it will be handed over at the stated price; the instant that belief fails, the metal is still sitting exactly where it was, and the mechanism has already stopped. A Song-dynasty line, quoted already in Essay 6, applies here in a new setting: with the government holding no capital, how could the people be expected to trust it? The gold standard's answer was to stockpile the capital. What 1931 would show is that even with the capital stockpiled, the trust could still run short.

A system that advertised itself as anchoring currency to a metal, and thereby freeing itself from any need to depend on anyone's word, turns out, at its most finely engineered point, to still be running on a belief. That sentence would be re-tested in every crisis that followed, and every re-test came at a steep price.

The intellectual prehistory of the idea reaches back further still. In 1752, David Hume wrote a short essay on the balance of trade — Of the Balance of Trade — arguing that money stood in a certain proportion to goods, labor, and industry, and that a surplus of gold and silver could never permanently pile up in any one place. He left behind a line that later generations would quote for more than two centuries: that the money always finds its way back again, by a hundred canals, of which we have no notion or suspicion.

Finds its way back again, by itself — that phrase, on its own, is the whole source of the system's charm.

2. An Accident Dressed as Destiny

It is tempting, looking back, to tell the story of the gold standard as a rising line that had to happen: the world grew more developed, trade grew more complicated, and so, rationally, everyone settled on the best available scale.

The record does not support that story.

The real international turning point fell between 1871 and 1873. After unification, the German Empire moved away from the silver standards its constituent states had used and onto gold. One historian of the period has summarized it this way: that reform set off a chain of steps in other countries that eventually brought the worldwide system of bimetallism to an end. France followed almost immediately, halting the free coinage of silver in September 1873; the other members of the Latin Monetary Union fell in behind France one by one, and bimetallism lost the last thing that had kept it steady.

But the same historian adds, without missing a beat, that Germany did not have to take this road.

In the 1860s, Germany had seriously debated joining the bimetallic system led by France. What actually happened later looks much more like two uncoordinated decisions, made separately by Germany and France in the early 1870s, that simply happened to stack on top of each other. One historian who has studied the episode calls it, flatly, a historical accident.

Accident is a heavy word to use in economic history. It does not mean the thing might not have happened at all; it means the thing might have turned out differently, and that every piece of reasoning built afterward simply takes the shape it did happen to take as its starting point. How many countries in the world ended up using gold was never deduced from any economic law. It grew out of a handful of decisions that a handful of countries happened to make in the years after 1871.

The spread that followed has a timetable. Germany in 1871. Denmark in 1872. Sweden in 1873. The Netherlands in 1875. France in 1878. The United States completed the move in law in 1873 and 1874, though effective restoration did not arrive until 1879. Japan followed in 1897. India moved in 1898 to a gold-exchange standard pegged to sterling.

That timetable can be read a second way. Behind every year on it sits a domestic political fight: debates on the floor of a parliament, silver-mine owners lobbying against it, creditors and debtors each running their own arithmetic, and the question of whether a given country wanted London to regard it as a borrower worth trusting. The dates look like a technical matter. The processes behind them never were. Between the United States' legal completion of the move and its effective restoration lay something like six years — years eventful enough to fill a book of their own.

By 1879, Britain, France, Germany, and the United States, the four core economies, were all on gold at the same time. What later generations would call the international gold standard is dated from that year to 1914: thirty-six years in total.

Thirty-six years is not a short stretch of a human life, but it is not a long stretch of institutional history either. And yet the entire inherited image of the gold standard, across two generations, rests on those thirty-six years. When people call it stable, they mean those thirty-six years. When they call it automatic, they mean those thirty-six years. When they call it a natural order, they still mean those same thirty-six years. An experience of under forty years got mistaken for the nature of the thing itself.

On why this diffusion happened at all, the historical literature offers two accounts, and neither one is weak.

One emphasizes network effects: the more trade a country did with countries already on gold, the stronger its incentive to join, and the maturity of its financial system shaped exactly when it did so. On this account, the spread was a logical adaptation.

The other emphasizes contingency: Germany and France could have chosen differently, the collapse of bimetallism was not calculated in advance, it was stumbled into. On this account, the spread was the sum of a string of policy moves that could easily have gone somewhere else entirely.

No verdict will be rendered here. But setting the two accounts side by side makes something visible that neither one shows on its own.

A scale that countless people afterward would treat as natural, objective, beyond anyone's will, has a birth certificate that can be dated to a specific piece of legislation, in a specific country, in a specific year — legislation that, at the time, could easily have been written otherwise. The construct is most successful exactly when it makes people forget it was ever made. Thirty-six years is not long, but it was long enough for a whole generation to assume the thing had simply always been there.

This is a standard maneuver of the construct, and it has appeared before. In Essay 10, a process spanning some twenty years, disconnected and pulling in different directions, got compressed into a single year and three names set side by side. Here, a sequence of separate policy choices made by several different countries gets compressed into a natural condition called the gold standard. The construct does not only press down on things outside itself. It also presses down on its own origins, flattening its own history into a backdrop with no starting point at all.

3. A Machine With Hands Inside It

On paper, the workings of the gold standard can be described with total cleanliness.

A country's central bank issues notes and deposit liabilities, and must back them with a certain proportion of gold reserves. A deficit in the balance of payments gets settled by an outflow of gold; a surplus gets settled by an inflow. If the monetary authority does nothing to offset the effect of these flows, an outflow of gold compresses the country's money supply and credit, prices fall in turn, exports grow cheaper, and the deficit narrows on its own; an inflow works the same way in reverse. This came to be known, later, as the price-specie-flow mechanism.

One historian, summarizing why the classical economists prized it, gives a reason simple enough to state plainly: they believed this mechanism would deliver long-run price stability across the whole world, and keep national price levels roughly in line with one another.

That vision has a genuine pull to it, and it should not be dismissed lightly. It promised something the history of the ledger had never delivered before: without depending on any monarch's self-restraint, and without depending on any parliament's discipline, a country's money supply would carry an automatic ceiling, set by whatever happened to be dug out of the ground. The debts of Essay 2 that never stopped compounding on their own, and the paper of Essay 6 that never stopped multiplying — this vision, in principle, sealed off both.

The trouble showed up in the actual operation.

One historian reviewing the classical gold standard wrote, in the introduction to an academic conference on the subject, that the conference had failed to settle a basic question: under fixed exchange rates, what really drove international adjustment — specie flows, purchasing-power parity and interest-rate parity, or short-term capital movements? She noted as well that scholars since the late nineteenth century have increasingly come to see short-term capital flows as playing a role in adjustment no smaller than the metal itself.

She then listed what the Bank of England actually did, and the list looks nothing like an automatic machine.

The Bank of England did not simply raise and lower its discount rate in mechanical response to gold flows. It also adjusted the buying and selling prices of gold bars and foreign coin, used a set of techniques later termed gold devices, charged different rates on different classes of loan, expanded and contracted its own position in short-term international lending, and used open-market operations to try to keep its defensive moves abroad as separate as possible from the domestic economy.

Nothing on that list was automatic. Every item required a judgment call: when to move, by how much, which lever first, and whether to let the market see it happening. And the point of these moves was, more often than not, not to keep the mechanism running at all, but precisely to cushion the domestic economy from the shock of the mechanism running. The machine advertised as automatic always had someone standing beside it, quietly keeping it from grinding up the ground it stood on.

The rules of the game, whatever else the phrase means, never described a machine that could be left unattended.

This tells us something sharp about the nature of the construct. The more insistently a system proclaims itself automatic, the more hands it requires working behind the scenes, and those hands have to work as quietly as possible. The instant anyone is caught turning the dial, the claim of automaticity collapses, and automaticity was the entire source of the system's persuasive power.

The remainder shows its face for the first time in this essay here, and in an odd location: not among the things being compressed, but on the side doing the compressing. Keeping the machine looking automatic required someone to keep making non-automatic judgment calls, and those judgment calls had no column of their own to be entered into. In the phrase the rules of the game, the rules are the part that got written down. The game is the part that never was.

There is a harder fact underneath this one: those who could turn the dial and those who could not were never standing in the same place.

Looking back on this period, Barry Eichengreen has pointed out that the cooperation sustaining the system was largely confined to a handful of core European countries, with the United States joining in only intermittently; the smaller countries on the periphery received no comparable support, and were correspondingly quicker to buckle under pressure.

One scale, unevenly distributed capacity to bear its weight. The construct's universality is always stratified: it applies a single standard to everyone, but not everyone can go on living under that same standard.

This sentence has already appeared, in variant form, in earlier essays. In Essay 7, one act of commensuration spread responsibility so thin that no one bore it. In Essay 12, one wage-scale priced the labor of whichever side worked harder at a lower rate. Here is a third variation: one discipline, with core countries propped up by unseen hands and peripheral countries left to stand on their own. The rule itself can be written down identically for everyone; the ring of support surrounding the rule never gets written down at all.

4. The Invisible Column

To sustain a gold standard, a country needs, at minimum, some rough sense of what its external accounts look like. At the time, that was simply not available.

Prewar Britain did not seem to need much in the way of statistics, because its accounts were comfortable enough to spare it the trouble: a deficit in visible trade was covered, over the long run, by a surplus on invisible items, and once that was covered there was still capacity left over for a great deal of investment abroad. In the years before the war, Britain's overseas investment ran to roughly four percent of national income annually, or about thirty percent of annual savings.

Invisible items meant shipping, financial services, interest income, commissions, and the returns on overseas investments. All of it was settleable, but none of it could be counted on a dock the way a shipload of cotton or a consignment of steel rails could be.

The word invisible was well chosen, and it was the word people at the time used themselves. It did not mean nonexistent, and it did not mean unimportant — quite the opposite: it named the single most profitable line of business Britain had. That a country's strongest sector should also be its hardest to count is, on its own, reason enough to be wary of how complete the books really were. But in years when the accounts came out comfortable regardless, no one had much incentive to count it precisely, because things worked out fine without the counting.

The Bank of England later produced its own retrospective account of this period, and that account reads almost like an ironic footnote to the whole myth of the common measure.

It records that before Britain's return to gold in 1925, whatever balance-of-payments statistics could be had barely entered into policy discussions at all; people at the time spoke of Britain's external position only in the vaguest terms. By 1926, the relevant statistics committee understood the accounting principles of the balance of payments well enough, but the data actually usable in practice remained extremely limited. Keynes complained in 1927 that those invisible items remained, in the most literal sense, invisible.

That date, 1927, is worth holding onto. By then Britain had already been back on gold for two years, had already paid two years' worth of the cost of that decision, and the external accounts meant to justify the decision were, statistically speaking, still a blur. The decision had been made against a number precise to several decimal places; the reality that number was supposed to capture, no one at the time could actually measure.

It got worse from there. Between 1930 and 1931, the Macmillan Committee produced an estimate of Britain's net short-term external liabilities, an estimate later shown to have been badly too low.

And by September 1931, an internal Bank of England memorandum admitted outright that capital movements had grown so large that not even the roughest numerical hint could be offered.

Set that sentence beside the seven decimal digits from the opening of this essay.

Parity could be carried out to the seventh place after the decimal point, and yet how much short-term external debt the country itself was carrying, and how much more of it might flee the following morning, could not be given so much as the loosest approximation.

This is the shape the remainder takes at this point in the story, and it is different from the shapes it has taken in earlier essays.

Some earlier remainders went unmeasured. Some were measured accurately and then left unexamined. Some were classified out of bounds for the discipline in question. Some simply lacked evidence. Here the remainder takes a different form again: the scale itself was extremely precise, and the object it was measuring was unknowable.

These two things are routinely confused with each other, and they are not the same thing at all. Precision belongs to the scale's side of the transaction — how fine its gradations are, how many digits its reading can report. Knowability belongs to the object's side — how large the thing is right now, and how large it might become tomorrow. A scale can be graduated as fine as a hair, while the thing it is meant to measure may simply refuse to hold still. This is exactly where the construct makes its easiest mistake: it mistakes the precision on its own side for clarity on the other.

The trouble in the 1931 crisis was never that gold lacked a definition. Gold was defined with total clarity. The trouble was that nobody could say what the real position actually was, or whether tomorrow would bring another wave of flight — and neither of those questions comes with decimal places, or with units.

The gold standard did not make the world measurable. What it did was something else: it kept forcing country after country to turn things that resisted a common dimension into statistics, into balance sheets, into exchange rates. That project kept advancing the whole time, and at the one moment it mattered most, the numbers still could not be produced.

And the project itself was genuinely useful — this cannot be said in reverse. Balance-of-payments accounting principles, statistics committees, the entire apparatus of external accounts that came afterward: all of it was forced into being by this system, and all of it is still in use today. What the construct forces into existence often outlives the construct itself. It simply could not keep pace with that particular emergency: by the time the reckoning could finally be made, the thing it needed to reckon with had already passed.

5. Four Hundred Ounces, Out of Reach

On April 28, 1925, Winston Churchill announced in his budget speech that Britain was returning to the gold standard.

But he immediately attached a qualification: this did not mean reissuing gold coin for circulation. That, he said, would be entirely unnecessary, and given the state of the nation's finances at the time, it would be, in his own words, "an unwarrantable extravagance which our present financial stringency by no means allows us to indulge in." He appealed, he said, "to all classes in the public interest to continue to use notes and to make no change in the habits and practices they have become used to for the last ten years."

There is a detail in this passage that rarely gets noticed. Churchill was not explaining a technical arrangement; he was offering reassurance. He wanted the public to feel comfortable continuing to use paper, rather than rushing to the Bank to demand gold the moment they heard the standard had been restored. Which is to say: on the very day of the announcement, the system had already conceded something — that if people really did all come to redeem their notes, the arrangement could not withstand it.

So what, exactly, had been restored?

The actual form convertibility took was this: the Bank of England, at its head office, would sell gold bars at a fixed price, and no single sale could come to less than four hundred fine ounces.

A few days later, the Gold Standard Act 1925 wrote the arrangement into law. Its meaning was unambiguous: the Bank of England was no longer obliged to redeem its notes in legal gold coin; what it was obliged to do was sell, to anyone who asked and paid in legal tender, a bar of approximately four hundred fine ounces of gold.

Four hundred fine ounces comes to a little over twelve kilograms.

That figure was not chosen carelessly. It was large enough to keep the overwhelming majority of banknote holders out of the redemption business, while staying small enough not to interfere with large-scale settlement between banks. Set at that threshold, convertibility stopped being a right anyone could exercise and became a right only institutions could afford to.

An ordinary person walking into the Bank of England with a five-pound note would come away with nothing. Convertibility remained, in law, perfectly intact — and remained, in practice, open to almost no one.

A very clean split appears here. As a matter of law, the pound could still be converted into gold; that statement is entirely true, and it is precisely that statement that held up the credibility of the whole system. As a matter of fact, the only parties who could exercise the right were banks, large merchant houses, and the central banks of other countries. Credibility rested on a right that, for the vast majority of people holding the currency, did not actually apply to them. There was no deception in this — the statute said so plainly, and the threshold was public knowledge. It only demonstrates, once again, that the same rule is a different thing depending on where a person stands in relation to it.

Churchill also announced that Britain had accumulated a gold reserve of one hundred fifty-three million pounds, and declared that this reserve would be used, in his words, "without hesitation, if necessary with the Bank Rate, in order to defend and sustain" the newly restored system.

The point of this section is not what Churchill said. It is the direction of the arrangement itself.

Essay 3 told the story of what the Lydians did: they gave the scale a body it could carry — the coin. For more than two thousand years afterward, the scale and the people using it kept some kind of hands-on relationship: money could be touched, weight could be hefted, purity could be doubted with a person's own two hands.

The step taken in 1925 runs in the opposite direction.

Gold was pulled further out of everyday circulation, becoming an institutional backstop that only ever surfaced in large settlements. The very act of restoring the gold standard was, in effect, the act of taking gold out of ordinary hands. The system did not become more concrete as a result. It became more abstract, more concentrated, more dependent on institutions.

Anonymity was pushed to its limit. And the result of pushing it that far was that it moved further away from ordinary people.

Following this thread back, one can look again at the road already traveled. The earliest accounts had to rely on a temple hierarchy to back them up; later, coined money relied on the stamp verifying its face; later still, jiaozi relied on whether the government's office recognized it. Every step moved further in an impersonal direction, and every step pushed the ultimate backstop up one more level. By the time of the gold standard, the backstop had moved to a metal and a handful of central banks, somewhere an ordinary person could not even touch it. The further this process went, the fewer people remained who were in any position to check it.

This, too, is one destination available to the remainder. It was not eliminated; it was relocated somewhere fewer people could see. So long as anyone at all can still check a thing, that thing stays within the range of what can be disputed. Once checking it requires four hundred fine ounces as the price of admission, it exits the field of vision of almost everyone. And exiting the field of vision, in day-to-day life, is very hard to tell apart from not existing at all.

6. Ten Percent, Made Flesh

Keynes's opposition arrived quickly. On July 22 through 24, 1925, he ran three consecutive articles in an evening newspaper, and shortly afterward expanded them into a pamphlet titled The Economic Consequences of Mr. Churchill.

His arithmetic was about as simple as arithmetic gets.

Sterling's external value after the war stood roughly ten percent below its prewar gold parity. Pulling it back up to that prewar parity meant that, from then on, whenever Britain sold anything abroad, either the foreign buyer would have to pay ten percent more, or Britain would have to accept ten percent less. There was only one way to avoid losing competitiveness: cut the sterling prices of things like coal, steel, and shipping by roughly ten percent.

That ten percent was not an estimate. It was the difference calculated between two parities, produced by the same arithmetic that had produced the seven-digit decimal in the opening section of this essay. What the construct does here is remarkably clean: it converts a country's entire external trade into a single ratio, and then hands that ratio down the line. Exactly how it gets distributed among individual industries, individual firms, individual people, is not a question this equation is obliged to answer.

Cut prices by ten percent reads cleanly enough on paper.

In the real world it has exactly one way of being carried out. The largest cost in those industries is wages; cutting prices means cutting wages; and wages have to be accepted one person at a time.

Between the equation and the pay packet lies a long road: board meetings, foremen passing the word down the line, contracts renegotiated, strikes, a return to work, then strikes again. None of what happens on that road shows up anywhere in the equation, and yet the equation's conclusion cannot land without that road being walked to its end. The construct can calculate the result. It cannot calculate the process of arriving at the result, and people live their lives inside the process, not the result.

Keynes wrote the sharpest line of all: "The policy of deliberately intensifying unemployment with a view to forcing wage reductions is already partly in force."

He did not soften it: the point of the policy, he said outright, was to force wages down. And he named exactly who would be paying for it — the miners, he wrote, "are the victims of the economic Juggernaut. They represent in the flesh the 'fundamental adjustments' engineered by the Treasury and the Bank of England to satisfy the impatience of the City fathers to bridge the 'moderate gap' between $4.40 and $4.86. They (and others to follow) are the 'moderate sacrifice' still necessary to ensure the stability of the gold standard."

Represented in the flesh — that phrase needs no footnote. It is not a figure of speech. It is an exact description of how a policy actually gets carried out: the adjustment has nowhere else to land, it can only land on specific, particular people, and which people it lands on is decided by what industry they happen to be in and how much bargaining power they happen to hold.

So the remainder in this case has a specific face, and it was the people of the time themselves who pointed to it. Keynes did not say the miners had been treated unfairly. He said the miners were the adjustment itself. A thing that exists, in the equation, as a percentage, exists, in reality, as a human being, and there is no conversion table between those two modes of existing.

The Bank of England's own retrospective account, written decades later, renders the same episode in the flattest official language, but does not dodge it.

That account states that before the 1925 restoration, policymakers had barely scrutinized the balance-of-payments statistics with any seriousness; for most people at the time, not returning to the 1913 dollar parity was almost unthinkable; and if this meant sterling ended up overvalued, the classic remedy was forced deflation at home and inflation abroad, driving prices and costs back into alignment.

The trouble was that British unemployment was already high, and the government had no appetite for bearing further deflation on top of it; and the General Strike of 1926 demonstrated just how nearly impossible an actual cut in nominal wages was to achieve. The result was that from 1925 to 1931, keeping the external balance in order became an unending problem, and Britain had to hold interest rates above American levels for years on end simply to keep capital from fleeing.

The cost was measured.

Across the six years the restored gold standard lasted, the unemployment rate among Britain's insured workers averaged 12.9 percent, and never once fell below 8.5 percent. In the first nine months of 1931 alone, it averaged 21 percent.

Four hundred fine ounces, and twenty-one percent.

Both figures came out of the same system, and only one of them was ever written into law. The other was that law's byproduct: published every year, visible to anyone who cared to look, and answerable to no clause of the statute at all.

For as long as the Bank of England kept handing over gold bars at the legal price, the unemployed stayed exactly where they were, underneath deflation, and underneath high real interest rates.

7. Nine Billion Locked, Two Billion Free

The most influential account of the relationship between the gold standard and the Great Depression comes from Barry Eichengreen's Golden Fetters, together with a series of papers he co-wrote with another researcher.

The core of this account is that the gold standard placed a fundamental constraint on economic policy, and did a great deal to manufacture an unstable environment in its own right. Eichengreen and his co-author also proposed a harder concept to argue with, which they called the gold-standard mentality: the idea that this was not merely an institutional arrangement but a standard of judgment shared by political elites and central bankers alike, in which holding onto gold meant holding onto credibility, and abandoning convertibility meant forfeiting it.

This concept matters a great deal for the framework of this essay series. Across the preceding thirteen essays, the construct has always operated from the outside: compressing things, renaming things, refusing to ask certain questions, drawing boundaries, laying down tracks, demanding evidence. Here, for the first time, the construct has taken up residence inside the heads of the very people enforcing it. The system itself had, in fact, left a way out. The people running it simply refused to use it.

This is the construct's most durable mode of existing. A rule can be amended, suspended, voted down. A standard of judgment needs no one to enforce it at all: it has already lodged itself inside the mind of the person making the decision, as the measure by which he judges his own sense of responsibility. At that point, nobody has to force anyone to do anything. The people keeping the rule will weld the exit shut themselves.

Several cross-country comparisons lend this account real weight.

First, the interwar gold standard was remarkably short-lived. In the classical period, the four core countries were on gold together for thirty-six years; between the wars, that same configuration held for only four to six years.

Second, the system was technically tighter than it had been before the war. Two researchers produced a striking figure: in 1929, forty-one countries together held 9.378 billion dollars in gold reserves, but of that total, only 2.178 billion dollars was genuinely surplus, available to meet a temporary international imbalance. The rest sat locked onto balance sheets by statutory cover requirements.

Out of 9.3 billion, only 2.1 billion could actually move.

This is worth sitting with on its own. A reserve is supposed to exist so that it can be drawn on in hard times, and yet this system required it to sit in place permanently, precisely so that everyone could see it sitting there. The function of a reserve turned into being seen rather than being used. And on the one day it was actually needed, the portion that could move turned out to be the smallest slice of all.

This point sits alongside a thread already running through this essay. To make people believe it had a floor beneath it, the construct had to display that floor for everyone to see, and the very thing put on display, the moment it was actually drawn on, would amount to admitting the floor was not enough. So the more a reserve gets treated as a guarantee of credibility, the less it can ever actually be used. In the end, a thing's usefulness turned into its staying unused.

Third, the gold standard transmitted contraction across borders. Between 1928 and 1929, the United States Federal Reserve tightened credit to rein in stock-market speculation, and because the international gold standard bound every central bank to every other, the rest were forced to raise their own rates in step; the resulting contractionary monetary policy pushed a great many economies into recession. One researcher summed up the global deflation of the early 1930s in a single sentence: it was a monetary contraction transmitted through the international gold standard.

Fourth, whoever left earlier tended to recover earlier. Spain, which never restored the gold standard and let its exchange rate float, escaped the fall in prices and output that hit most of Europe. The Scandinavian countries, which left the gold standard alongside Britain in 1931, recovered markedly faster than the European countries that stayed on it. Across a sample of twenty-four countries, industrial production between 1932 and 1935 grew, on average, about seven percentage points faster in the countries off gold than in the countries still on it. The Bank of England's own historical writing concedes that Britain's real economic recovery did not arrive until early 1932, and that it depended entirely on the suspension of the gold standard late in 1931, which finally made interest-rate cuts and a devalued pound possible.

Fifth, two countries drew gold in and would not let it back out. One researcher has called France a gold sink: France's share of the world's gold reserves rose from seven percent in 1927 to twenty-seven percent in 1932, and it locked most of that new gold away under an even higher cover ratio than before. His counterfactual simulation suggests that had the historical relationship between gold reserves and prices from 1929 to 1933 simply continued to operate as usual, world prices would have risen slightly rather than collapsed catastrophically. He further estimates that the excess gold held by the United States and France in 1929 amounted to about six percent of the world's gold stock, rising to twelve percent by 1930 through 1932; by 1932, France alone had withdrawn something like thirteen percent of the world's gold stock from effective circulation.

Britain, in the middle of all this, was squeezed from both sides: an overvalued pound on one side, and, on the other, the pressure created by France's decision to hold its reserves in gold rather than in sterling or dollars. Between July and September 1931, British authorities spent roughly two hundred million pounds in gold and foreign credit trying to support the pound; the outflow did not stop, and in September Britain was forced to suspend convertibility. Sterling fell to $3.89 by October and to $3.37 by December.

8. The Good Housekeeping Seal of Approval

The other side of this argument has to be laid out in full, or the preceding seven sections will read as a verdict rather than an account.

One school of research calls the classical gold standard a good housekeeping seal of approval. The reasoning runs like this: between 1870 and 1914, a peripheral country joining the gold standard sent a signal to the London market that its finances were in order and that it intended to honor its commitments; as a result, it could borrow more easily, and more cheaply. One quantitative study has put a number on the benefit: before 1914, adherence to the gold standard could lower a country's borrowing spread by something like forty to sixty basis points.

This same school has also advanced a more refined version of the argument, describing the classical gold standard as a rule with an escape clause. The idea is that it was never a dead letter: convertibility held in normal times, could be suspended during a war or a major crisis, and the market expected the country to return afterward to its original parity. On this account, the classical gold standard was not without a human dimension at all. It simply folded discretion into a more credible framework.

This school generally does not deny that the interwar years were a failure. It simply places the failure elsewhere: what failed was the gold-exchange standard rebuilt after the war, not the classical gold standard itself. Some of its defenders go further still, arguing that many of the era's disasters should be blamed on central banks that violated the rules of the gold standard, not on the underlying idea.

Both sides can produce evidence, and no verdict will be rendered here, but the reason needs to be stated clearly.

The two camps are not actually looking at the same thing. One is mainly asking why the classical period achieved relative stability. The other is mainly asking why the interwar years turned an ordinary recession into the Great Depression. A third question, pursued by others still, asks why countries kept holding on even once the costs were plain to see, and why some left earlier than others. Each of these three questions has its own time frame, its own set of countries, its own body of evidence. Mash them together into a single question, was the gold standard good or bad, and what comes out will not be an answer. It will be a flattening.

There is a point about method worth recording in passing here. A debate can only reach a close when every side is arguing about the same proposition, and a great many disputes in economic history that never seem to close are not stuck because the evidence is lacking. They are stuck because the participants are answering different questions while sharing a single title. Sharing a title is easily mistaken for arguing over the same thing.

One study, though, pushes the question further, into the realm of politics, and its conclusion rewards a slow reading.

Across a panel of twenty-four countries, higher per-capita income, a stronger position as a creditor nation, and a history of hyperinflation all made a country more likely to stay on the gold standard for the long haul; democratic institutions, unemployment, and banking crises all shortened how long it lasted.

Translate that into plain language: what determined how long a country could stay on the gold standard was never just how much gold it held, but how long it could go on compressing the pain. And how long the pain could be compressed depended on whether the people bearing it had any way to push back.

One thing remains unaccounted for, and it is the deepest irony in the whole arrangement.

In the name of conserving gold, what got rebuilt after the war was not a purer gold system but a more abstract gold-exchange standard. Many countries stopped gold coin from circulating after the war and switched to a gold bullion standard; still more simply held gold-convertible foreign exchange as their reserves, and that foreign exchange was, in practice, mostly sterling and dollars. The foreign-exchange reserves held by European central banks, which had fluctuated in the decade before the war somewhere between 250 and 400 million dollars, had risen to 844 million dollars by the end of 1924, and to 2.513 billion dollars by the end of 1928. Of the four core interwar economies, only the United States remained on a gold coin standard; Britain and France were on a gold bullion standard; most of the rest were on a gold-exchange standard.

Which is to say: this world, which advertised itself as resting on gold, was increasingly sitting on paper that other people had merely promised to redeem in gold.

Essay 6 told the story of jiaozi: strip the metal out of money, and precisely that much weight comes to rest on trust instead. What the gold standard tried to do was the reverse — move the weight off trust and back onto metal. What it produced, in the end, was a new layer of trust, one more distant, more concentrated, and harder for anyone to verify than the layers it had replaced. The effort to escape trust altogether ended by relocating trust somewhere even less visible.

That thread, the one running through national credit, market reputation, central-bank cooperation, and social compliance, was never actually cut. The institutional shift of the 1870s, the war finance of 1914, the overvalued restoration of 1925, the capital flight of 1931, the global banking panic of 1933: every real stress test lit that thread up again.

Return to the two figures this essay opened with.

The seventh decimal place, and the sentence about not being able to offer even the roughest hint, both came out of the same system, less than a decade apart.

They are not, in fact, contradictory. A scale can be made extremely precise, precise enough that people come to assume everything precise has already been measured. But what it measures is always only what it is capable of measuring. What it cannot measure keeps moving regardless, and moves most violently exactly at the moment it matters most.

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