Economics & Money

What money actually is

Workers in shirtsleeves operating coining presses on the floor of the San Francisco Mint, with trays and sacks of blank coins beside the machines.
The San Francisco Mint striking coins under contract for the Chinese government — a reminder that a coin's value has always depended on whose stamp is on it rather than what it is made of. Photograph by Unknown author, U.S. National Archives and Records Administration · Public domain

Take a note out of your wallet and look at it properly. It's a rectangle of cotton and linen, or in a growing number of countries a sheet of polymer, printed with somebody's face and a number. Tear it in half and you've destroyed nothing of any use. Nobody can eat it, wear it or burn it for much heat.

Hand it over a counter, though, and someone gives you dinner.

That's the whole puzzle, and it's a much older one than economics. Money isn't a substance. It's a set of jobs, and any object that does those jobs well enough gets treated as money until something better turns up.

The three jobs

Money has to be a medium of exchange, a unit of account and a store of value. Most explanations stop at listing them. The interesting part is that they're separate jobs, and they come apart under stress.

A medium of exchange is the obvious one — you accept a thing you don't want because you're confident somebody else will accept it from you later. That confidence is the entire mechanism. It doesn't rely on the object being useful, only on everyone else believing everyone else will take it.

A unit of account is quieter and arguably more important. It's the ruler. Once prices are quoted in a common unit, you can compare a haircut to a bag of rice without swapping either. Contracts, wages, debts, tax bills, company accounts — all of it needs a shared measuring stick, and the stick doesn't have to be the same thing you actually hand over.

A store of value means it holds its worth across time well enough that you're willing to be paid on Friday and spend on Sunday. This is the job money does worst, and the one it's most often blamed for failing.

In countries with severe inflation you can watch the three come apart in real time. Shops quote prices in dollars, take payment in the local currency, and the moment anyone has spare money they convert it into something that isn't money at all. Three jobs, three different objects, one very tired population.

The barter story is probably wrong

Every textbook opens the same way. Once upon a time people bartered. The butcher wanted bread, the baker wanted shoes, the cobbler was a vegetarian, and the whole thing seized up on what economists call the double coincidence of wants. Then somebody invented money and everyone relaxed.

Adam Smith told a version of this in 1776 and it has been repeated ever since. The trouble is that anthropologists went looking for the barter society and couldn't find one.

Caroline Humphrey put it bluntly in 1985: no example of a barter economy, pure and simple, has ever been described, let alone the emergence of money from one. David Graeber built his 2011 book Debt: The First 5,000 Years on the same observation. Where barter does show up, it's usually between strangers or enemies — people with no ongoing relationship — or in the wreckage of a currency that's collapsed. Inside a village, people didn't swap chickens for sandals. They kept track of who owed whom, informally, for years.

What the written record actually shows first is accounting. The earliest clay tablets from Mesopotamia, well before coins existed, are records of debts and rations, denominated in the shekel — originally a weight of barley. Temples and palaces kept the books. Silver was the unit of account long before anybody was carrying it around in a purse.

England ran a version of the same idea for centuries with tally sticks: a hazel stick notched to record a debt, then split lengthways so each party held a half that matched only the other. The Exchequer used them until 1826. In 1834 somebody decided to dispose of the accumulated stock by burning it in the furnaces under the House of Lords, the flues overheated, and the resulting fire destroyed most of the Palace of Westminster. Britain's medieval accounting system took the parliament building with it on the way out.

So the tidy story runs backwards. Credit came first. Coins came later.

Why metal kept winning

Once you do want a physical token, the requirements are demanding. It has to survive being carried about. It has to divide without losing value, which rules out cattle — half a cow isn't worth half a cow. It has to be portable enough that a meaningful sum doesn't need a cart. It has to be hard to fake and hard to obtain, or everyone makes their own and the whole thing deflates into nothing.

Plenty of things have passed the test locally. Cowrie shells circulated across huge stretches of Asia and Africa. Cattle were wealth in enough societies that the Latin for money, pecunia, comes from pecus, livestock. Bronze spades and knives were cast as money in Zhou China, keeping the shape of the tools they'd replaced long after nobody would have dug with them.

The strangest surviving example is on Yap, in the western Pacific, where enormous carved limestone discs served as high-value money. The discs were quarried on another island and were far too heavy to shift, so ownership changed while the stone stayed put. One was reportedly lost at sea during transport, and by common agreement it still counted — everybody knew it existed and everybody knew whose it was. Milton Friedman wrote about it in 1991 partly to make the point that this is less absurd than it sounds, given that most of the world's monetary gold sits in vaults and never moves either.

Metal won out because it hit every requirement at once. The first true coins we know of came from Lydia, in what's now western Turkey, around 600 BC — lumps of electrum, a natural gold-silver alloy, stamped with a mark. The stamp was the innovation. It meant you didn't have to weigh and test the metal yourself; you trusted whoever's badge was on it.

Which is exactly where the trouble starts

The moment a coin is worth more as a token than as metal, whoever issues it has found free money. And they will take it.

Rome ran the experiment thoroughly. The denarius started as a high-purity silver coin. Nero trimmed both weight and fineness in AD 64. Successive emperors trimmed further, and by the middle of the third century the standard silver coin was a copper disc with a wash of silver on the surface that wore off in use. Prices in Egypt, where the papyrus record is unusually good, rose by orders of magnitude over that period.

Ordinary people played the same game from below. Clipping — shaving slivers off the edge of a coin and keeping the metal — was so common in seventeenth-century England that the currency was genuinely unreliable by weight. The fix was the milled edge, those fine ridges still on the rim of a modern coin, which make clipping obvious. Isaac Newton spent the last three decades of his life running the Royal Mint, oversaw the Great Recoinage of 1696, and personally pursued counterfeiters, including William Chaloner, who was hanged in 1699.

The clipping era also gave us Gresham's law, usually shortened to "bad money drives out good". The mechanism matters more than the slogan: if two coins are legally worth the same but one contains more silver, you'll spend the poor one and keep the good one. So the good coins vanish from circulation and only the rubbish stays in your pocket.

Paper, and the leap it required

China got there first. Merchants in the Song dynasty were already issuing deposit receipts in the eleventh century, and the state took the idea over. By the Yuan dynasty, paper notes were the official currency, which is why Marco Polo devoted an incredulous passage to the emperor who made money out of the bark of trees. China also discovered, several times, what happens when you print more of it than the economy can absorb.

Europe took longer. Stockholms Banco issued Europe's first proper banknotes in 1661, under Johan Palmstruch, and collapsed within a few years for exactly the reason you'd expect — it issued more notes than it could redeem. London's goldsmiths, who stored other people's gold and issued receipts for it, noticed that the receipts circulated perfectly well on their own and that the gold mostly sat there. Lending out receipts against gold you were only holding is the beginning of modern banking. The Bank of England was founded in 1694, and the immediate reason was to fund a war.

Most money is created by banks typing numbers

Here's the part that surprises people, and it isn't a fringe view — the Bank of England set it out plainly in its 2014 quarterly bulletin.

Banks don't lend out deposits that savers have placed with them. When a bank makes a loan, it credits the borrower's account with new money it has created by making an entry. The loan and the deposit appear at the same instant. Repay the loan and that money is extinguished again. Commercial banks, not the central bank, create the overwhelming majority of the money in circulation, and in the UK something like 97% of the money supply is bank deposits rather than notes and coin.

Which means the cash in your pocket is the unusual case, not the normal one.

Central banks influence all this indirectly. They set a short-term interest rate, which changes how much borrowing people and firms want to do, which changes how much money gets created. It's a steering wheel with a long delay built into it, and the delay is why monetary policy is so often described as acting with long and variable lags.

The gold standard and the trade it forced

Under the classical gold standard, roughly the 1870s to 1914, major currencies were fixed to a set weight of gold and therefore to each other. It made exchange rates predictable and made it very hard for a government to inflate away its debts.

It also handed control of the money supply to geology. If mining output didn't keep pace with a growing economy, the result was falling prices — and falling prices are brutal for anyone carrying debt, because the debt stays the same size while everything else shrinks. That's the backdrop to American farm politics in the 1890s and William Jennings Bryan's "Cross of Gold" speech in 1896.

Britain suspended gold in the First World War, went back in 1925 at the pre-war rate, and found that the rate was far too high for post-war costs. Keynes wrote a pamphlet about it with the acid title The Economic Consequences of Mr Churchill. Britain came off gold again in 1931. The Bretton Woods system after 1944 fixed the dollar to gold at $35 an ounce and other currencies to the dollar, which worked while the United States held most of the world's gold and stopped working when it didn't. Nixon suspended convertibility in August 1971, and by 1973 the major currencies were floating.

The trade-off is real in both directions. A metal standard stops a government debasing the currency, and it also stops it doing anything at all when demand collapses. Countries that left gold earliest in the 1930s generally recovered earliest. That's about the least controversial finding in the whole literature.

So what actually backs it now

Nothing you can dig up, which is why the question feels uncomfortable. The honest answer has several parts.

  • Taxes. Governments demand payment in their own currency and won't take anything else, which creates a permanent floor of demand for it among everyone who owes tax — which is everyone.
  • Law. Legal tender rules and the requirement that domestic contracts settle in the domestic unit keep it embedded in everything.
  • Everyone else. A currency is a network, and the value of being on the network rises with the number of people already on it.
  • The central bank's willingness to restrict supply, and the institutional credibility that makes people believe it will.

That last one isn't decoration. Currencies that fail almost always fail because the public stops believing the issuer will exercise restraint, and once that belief goes, it goes fast.

Inflation is a mechanism, not a character flaw

Inflation gets talked about as though it were a moral event — greed, waste, decadence. It's more useful to treat it as plumbing.

Prices rise when the amount of spending in an economy grows faster than the amount of stuff available to buy. That can happen because the money supply expanded, or because supply was disrupted, or because everyone tried to spend savings at the same moment. The 1970s had a textbook supply shock when oil prices quadrupled after 1973 — no extra money required, just far less of something everything else depended on.

Expectations then do the rest. If you're confident prices will be 8% higher next year, you'll ask for a bigger wage rise, and firms expecting to pay bigger wages will set higher prices now. The expectation makes itself true. This is why central banks talk endlessly about anchoring expectations, and why they treat their own credibility as an economic asset.

The extreme cases are instructive because the mechanism is identical, just faster. Germany in 1923 ended with a dollar costing about 4.2 trillion marks, having cost 4.2 before the war. Hungary in 1946 holds the record — prices roughly doubling every 15 hours at the peak. Zimbabwe in 2008 printed a hundred-trillion-dollar note. Each followed a government funding itself by creating money because it couldn't raise enough by taxing or borrowing.

Inflation also redistributes, silently and without anyone voting for it. Debts are fixed in nominal terms, so inflation shrinks them in real terms — good for borrowers, bad for lenders, bad for anyone on a fixed income. Governments are usually the largest borrowers around.

Deflation isn't the safe opposite. Falling prices make every existing debt heavier and give people a reason to postpone buying, which reduces demand and pushes prices down further. That loop is a large part of why the 1930s were as bad as they were. It's also why most central banks aim for a small positive number rather than zero — New Zealand was first to adopt a formal inflation target in 1990, and 2% became the common choice not because it's magic but because it's far enough from zero to leave room for error.

The ruler keeps shrinking

Because money is the measuring stick, and the stick changes length, comparing money across time is harder than it looks. Irving Fisher called the confusion money illusion in 1928: people react to the number on the price tag rather than what it buys.

Any statement about what a Victorian labourer's wage would be "in today's money" is an estimate built on a basket of goods that has changed beyond recognition. Bread and rent still exist. Antibiotics, refrigeration and electric light didn't, at any price, and no amount of adjusting handles that.

The form of money keeps changing too, and much faster than the questions do. Notes, cheques, cards, instant bank transfers, contactless payment on a phone — each shift moved money further from anything you can hold, and each was met with the same worry about whether the thing was still real. Cash is now a minority of transactions across much of the world, and central banks are actively studying digital versions of it.

None of which changes the underlying arrangement. Money is a claim on other people's willingness to give you things, held together by the expectation that they'll keep doing it. The paper is just the paperwork.

Quizzes on this subject

All articles Play a quiz
Keep reading

More from the Blog

Interior of a rural American schoolroom in the 1930s, with pupils of several ages sitting at wooden desks in rows facing a blackboard.
Education & Learning

Why we teach the way we do

Rows of desks, children sorted by birth year, fifty-minute periods, six weeks off in summer…

A crowded street under a bright blue sky during a Philippine fiesta, with people spraying water from hoses over one another beside parked vehicles and a decorated arch reading Saint Peter.
Culture & Society

Why traditions survive

Many practices that present themselves as immemorial are younger than the railway. The interesting…