The number in your account
Look at a bank balance and ask what the number is. It buys groceries and settles rent, so it behaves like money, yet there is no coin behind it and no note in a vault with your name on it. The figure is an entry in the bank’s ledger recording what the bank owes you, which is to say it is a debt, denominated in a unit almost everyone around you also keeps their accounts in. That is the whole of it. Before asking, with the next chapter, where the bank found the money for a mortgage, it is worth asking the prior question the balance raises: what is this thing, and where did it come from.
The usual answer arrives as a story most readers were told young, and the story is tidy enough that its confidence is part of the problem. This chapter sets the story beside the record and finds the two do not match, then asks what the mismatch means for the number on the statement.
The story in the schoolbook
The textbook account begins with barter and treats money as its cure. Picture a world without it. A baker who wants shoes must find a shoemaker who happens to want bread, and wants it today, in the amount the shoes are worth. Economists call this the double coincidence of wants, and the phrase captures why the arrangement is said to strangle trade: every exchange waits on two people’s needs lining up at once. Adam Smith gave the story its enduring shape in The Wealth of Nations (1776), rooting exchange in “a certain propensity in human nature … to truck, barter, and exchange one thing for another,” and arguing that the awkwardness of direct swaps drove prudent people to keep a stock of some good few would refuse, until one such good hardened into money.
A century later Carl Menger turned the story into a mechanism. In “On the Origin of Money” (1892) he asked why any one commodity should win the role, and answered with saleableness: some goods can be sold again more easily than others, and a trader will accept a highly saleable good even without wanting it, because it can be passed on. Each person reaching for the most saleable good on offer nudges the whole market toward a single medium, with no assembly and no decree. Menger insisted the outcome was “the spontaneous outcome … of particular, individual efforts,” neither accident nor state compulsion. This is the strongest form of the barter-first case, and it is worth stating at full strength before weighing it: it explains money without anyone having to design it, and it fits how commodity monies have in fact emerged in unplanned settings, from prison cigarettes to wartime rations.
What the record shows instead
The trouble with the barter-first story is that no one has found the barter economy it starts from. Anthropologists have looked. Where money is absent or scarce, people do not fall into spot barter of the textbook kind; they keep running tallies of who owes whom, settle debts unevenly over months and years, and lean on gift, obligation, and credit. Caroline Humphrey, surveying the ethnographic record, put the point flatly.
What the record does show, in the earliest places writing lets us read, is credit. In the temple and palace economies of Mesopotamia, scribes kept accounts in silver and barley long before either changed hands as coin. Rations, rents, and taxes were reckoned in fixed weights of silver; debts were recorded on clay, carried interest, and were periodically forgiven by decree. The unit of account and the interest-bearing loan are both in evidence by around 3000 BCE, and the clay habit of writing numbers down begins earlier still, in the accounting tokens and tablets from which cuneiform itself descends. Money as a measure and a debt is thus attested more than two thousand years before money as a stamped coin.
The order matters because it inverts the schoolbook sequence. In the story, coin comes first and credit is a later convenience built on top of it. In the record, the reckoning in a unit of account and the recording of debts come first, and the coin is a portable token that arrives late and travels well. The numbers in the temple accounts and the number in a modern bank balance are the same kind of thing, an entry in a ledger denominated in a shared unit; the coin is the interruption, not the origin.
The matching problem, whatever the origin
None of this retires the double coincidence of wants. Whatever its history, money does the job the schoolbook credits it with: it dissolves the matching problem that direct swaps run into. The point of separating function from origin is that the function survives either story. A shared unit that everyone will accept lets each half of an exchange find a different partner at a different time, and that is true whether the unit began as a temple’s silver weight or as Menger’s most saleable commodity.
Once function and origin are held apart, the barter story reads as a thought experiment about what money is good for that got mistaken for a chronicle of how it began. It is a serviceable parable and a poor history. The medium the parable reaches for at its end, some universally accepted thing, had already existed for millennia as an entry in an account.
Why the state keeps turning up
If markets did not conjure the unit of account on their own, the other recurring author is the state, and its fingerprints are on both the measure and the coin. The numismatist Philip Grierson argued that standard values first appear not in trade but in law: the compensation tariffs of early legal codes, which set exactly how much a life, a wound, or a theft must be paid for, required a common scale of value before markets were dense enough to grow one. The unit of account, on this reading, is a creature of custom and command, and it long predates the marketplace it later served.
Coinage carries a maker’s mark more plainly, because a coin is stamped by whoever issues it. One influential reading is that a state coins metal to pay soldiers, who are mobile, numerous, and disinclined to wait on the double coincidence of wants, and then demands the same coin back as tax, which obliges everyone within reach to obtain it and so gives it currency. David Graeber, drawing the threads together in Debt: The First 5,000 Years (2011), tied ancient coinage to the financing of armies in exactly this way. The interpretation is contested, and it fits some episodes better than others; even who struck the earliest electrum coins is debated, whether royal mints or private issuers such as merchants and local magnates. What is clear is that the first coins were stamped by an issuer rather than bubbling up from private swaps, and that taxation is a durable reason a particular money stays in demand. Money spread, in this telling, because power spent it and then required it back.
From coin to paper to the ledger again
The later history rhymes with the early one, moving away from the coin as fast as it once moved toward it. China issued the first government paper money under the Song, printing notes called jiaozi from 1023, redeemable claims that stood in for heavy strings of coin. In 1609 the Bank of Amsterdam opened as a public deposit bank, and merchants learned to settle with one another by moving figures between accounts on its books, a payment completed by debiting one balance and crediting another with no coin leaving a drawer. That is giro money, and it is the direct ancestor of the balance on a phone screen. Each step returned money to what the Mesopotamian scribes would have recognized: a number in a ledger, standing for a debt, accepted because others accept it.
Which is why the number in the account resists a simple answer to “what is it.” It is not a coin and not a claim on one. It is the newest layer of the oldest form money took, credit written down and passed around, now kept by banks rather than temples. How those banks come to write new figures into being, and why most of the money in the economy is figures they wrote rather than notes a mint struck, is the next chapter’s question.
Money was invented to replace barter.
Oversimplified Moderate confidence
The functional half is sound: a common money does dissolve the double coincidence of wants that direct swaps run into, and that is much of what money is for. The historical half does not hold. No barter economy of the textbook kind has been documented for money to have replaced, and the earliest evidence, from Mesopotamia, shows a unit of account and written, interest-bearing debt roughly two thousand years before the first coin. Money as measure and record predates money as circulating token. The claim goes wrong by reading a parable about what money does as a chronicle of how it began; the strongest opposing case, Menger’s account of money emerging spontaneously from the most saleable commodity, is a theory of money’s origin that offers a coherent path from barter to a single medium, but the barter starting point it assumes is the very thing the record does not show.
Sources
- C. Humphrey, “Barter and Economic Disintegration,” Man 20 (1985) — the ethnographic survey finding no pure-barter economy on record; supports the historical leg.
- M. Hudson, “Origins of Money and Interest: Palatial Credit, Not Barter” (2004; repr. 2018), and W. Goetzmann, Money Changes Everything (2016) — Mesopotamian silver-and-barley units of account and interest-bearing debt by c. 3000 BCE, predating coinage.
- C. Menger, “On the Origin of Money,” Economic Journal 2 (1892) — the strongest barter-first case, named per the observer rule; a theory of how a medium of exchange can emerge without design, not evidence of a barter stage.
- W. S. Jevons, Money and the Mechanism of Exchange (1875) — the double-coincidence argument the functional leg rests on. A. Smith, Wealth of Nations (1776), Bk I — the barter-origin narrative the claim restates.
- Confidence is moderate under the rubric: the pip scores the weaker of evidence directness and construct match, and directness binds here. The claim is a counterfactual about a stage of history; the case against it rests on the absence of any documented pure-barter economy (an argument from ethnographic silence) plus the positive Mesopotamian record, which is strong but partial. The direction is well supported; the evidence is inference from absence and from one region’s early archives.
Where the argument goes next
Reading money as debt rather than as coin is what makes the modern system legible. If money is a record of who owes what, in a unit everyone keeps their accounts in, then the interesting question is who gets to write the records, and the surprising answer is that ordinary banks do, every time they lend. Value creation through exchange, the surplus a trade leaves behind, has its own home in Frameworks, in the chapter on whether the economy is zero-sum; this volume stays with the institution. The next chapter follows the ledger into the bank and asks where new money actually comes from.