There can be no such thing, the classical economists held, as a general glut. A particular good might be overproduced — too many hats, too few coats, the hatter ruined while the tailor thrives — but not everything at once, for a universal surplus is a contradiction: to produce is to earn, and to earn is to be able to buy, so that every addition to supply is, in the same motion, an addition to demand. Goods, in James Mill's phrase, are bought with goods; money is only the veil over the barter beneath. Say put it most confidently of all, and the proposition took his name. For the better part of a century it stood as the rock on which the optimism of the science was built — until a decade arrived in which the warehouses filled, the workers were turned away, and the demand that supply was supposed to create was nowhere to be found.
The proposition has a name now — Say's Law, the law of markets — but it was less a discovery than the working-out of something that seemed, once seen, too plain to doubt. Begin with a single act of production. To make a coat is to hire labor, rent premises, use up materials, and leave a profit: every penny of the coat's value is paid out, in the making, as somebody's income. And income is the wherewithal of demand. The very act that adds a coat to the supply of goods has, in the adding, put into someone's hands the means to buy a coat's worth of something else. Carry the reasoning across the whole economy and the conclusion follows with the force of arithmetic: the total value supplied and the total income with which to demand it are one sum, seen from two ends. There can no more be a general deficiency of demand than a number can be less than itself.
What, then, of the gluts that plainly do occur — the unsold stock, the ruined trade, the warehouse no one will empty? The classics did not deny these; they denied only that they could be general. A glut, on their account, is always a thing of proportion. If hatters cannot sell their hats, it is because too much of the society's labor and capital has gone into hats and too little into the coats or the corn that the disappointed hat-buyers would rather have. The surplus of the one is the shortage of the other — a single error of allocation seen from opposite sides. Nor need anyone correct it by design, for the market corrects it unbidden: the price of the glutted good falls and the scarce good's rises, profit draws capital out of the first trade and into the second, and the proportions right themselves. The trouble is never that too much has been made in sum, only that the wrong things have been made.
Beneath the doctrine lay a particular picture of what an exchange economy really is, and its clearest draughtsman was James Mill. Strip away the money, Mill urged, and see the thing plain: a man brings his product to market not because he wants money but because he wants the products of others, and money is merely the contrivance that lets him sell to one person and buy from another instead of bartering with a single partner who happens to want just what he has. The demand for goods, at bottom, is goods. Every seller is a buyer in waiting, every supply an unspent demand. If this is the truth of the matter, the conclusion is irresistible — for a market so understood is a vast act of barter, and in barter a universal glut is unthinkable: one cannot offer without, in the same breath, asking. The economy cannot drown in goods, because goods are the very thing everyone is fishing for.
Everything turns, though the classics were slow to say so, on that little word merely — on money being merely the veil, merely the go-between. The picture of the economy as barter-with-a-token holds only if no one ever wishes to keep the token. The argument requires that money be wanted never for its own sake but solely as a thing in transit — taken in by a sale, paid out by a purchase, with no pause between — so that to receive money is already to be on the way to spending it. But suppose it false. Suppose money, being the one good that can be held against an uncertain future, is sometimes wanted precisely to be held — hoarded in fear, kept idle rather than risked. Then the proceeds of a sale may simply rest, the purchase that was to follow may never come, and the chain binding supply to demand is broken at its weakest link. The impossibility of a general glut, it emerges, was never a law of production. It was a hidden assumption about money.
Not everyone bowed. While Ricardo and the Mills built the orthodoxy, two economists looked at the wreckage of the post-war years — the gluts, the failures, the idle hands of the 1820s — and concluded that the thing the law forbade was happening before their eyes. Sismondi in Geneva and Malthus in England argued that demand could indeed fall short across a whole economy, and for a reason the law had overlooked: a society straining to accumulate might pour its income into new capacity faster than it found buyers for what that capacity produced. The drive to save and invest, a virtue in any single man, might in the aggregate outrun the willingness to consume — leaving goods unsold not because the wrong ones had been made but because, in sum, too many had, relative to a demand that thrift itself had thinned. Malthus, casting about for what might fill the gap, fixed on the idle consumption of landlords, the very class his rival most despised; and the oddity of the remedy did his case no good. He was judged confused, and lost.
He lost because Ricardo had what looked like a complete answer, and pressed it without mercy through years of letters. The flaw in Malthus, said Ricardo, was to imagine that what is saved is not spent. But the saver is no miser stuffing a mattress; he lends his savings, or invests them, and they are spent as fully as any wage — on labor, on materials, on the building of the very capacity Malthus feared. Saving is not a subtraction from demand but a transfer of it, from consumer goods to capital goods; the baker who saves a shilling sets a bricklayer to work as surely as if he had bought a loaf. And since the new capacity will in its turn generate the income to buy its own output, the process can run on without limit and without glut. The reasoning was elegant, largely persuasive, and it carried the day — resting, as few then noticed, on the quiet assumption that every shilling saved is a shilling promptly and certainly invested.
The most penetrating of the classical minds came nearest the truth, and then declined it. In an essay written in his twenties, John Stuart Mill conceded a remarkable thing: there might come a moment of commercial crisis in which almost everyone, seized by alarm, wishes to sell goods at once and to defer the purchases he can. This, he saw, is precisely a general glut of commodities — the warehouses overflowing, no one buying. But he saw, too, what such a moment is when turned around: a sudden, universal craving for money, a flight into the one thing all are willing to hold. Goods superabundant and money scarce are the same crisis told twice. Here, conceded by the doctrine's ablest defender, was the general glut the doctrine denied — and conceded along with its cause, the wish to hold money rather than spend it. Yet Mill set the concession aside as a passing fever of confidence, real but transient, no threat to the law in its long-run majesty. He had opened the door a crack, glanced through, and shut it.
And so the law endured. For something like a hundred years Say's proposition was the unquestioned foundation of the orthodox view: the economy, left alone, tends of itself to employ all who wish to work, and the slumps that came and went were laid to particular causes — a banking panic, a harvest failure, a bout of speculation, a meddling government — never to any general insufficiency of demand, which the law had shown could not exist. Then came the decade this plate opened with, in which the impossible thing stood plainly in view: not a partial glut soon righted, not a passing crisis of nerve, but a deep and stubborn deficiency of demand that held a quarter of the workforce idle, year upon year, while the means to employ them sat unused. It was Keynes who reopened the case Malthus had lost, and who built upon the ruins of the law the theory the next plate sets out. What logic had ruled out, events had ruled back in.
It is worth seeing exactly what the long argument was about, for it is the master question of this table in its purest form. Say's Law presents the self-clearing of the market — its tendency to absorb whatever it produces and to set every willing hand to work — as a matter of logical necessity, a thing that simply cannot fail to happen. It is called a law, and the name carries the claim: not a tendency, but a necessity of the same order as the laws of nature. And yet the necessity was never in the production at all. It lived entirely in an assumption about money and motive — that to sell is always to intend to buy, that no one holds money for its own sake — an assumption nowhere proved and not always true. Lift it, and the iron law softens into a mere likelihood: holding when confidence is high and money flows, failing when fear sets in and money is hoarded. The impossibility of the general glut was a fact of nature only so long as one did not look at what held it up. A price is a verdict on a hidden contest, mistaken for a fact of nature; and here the very possibility of a slump, denied for a century as a contradiction in terms, proves to be a question about money wearing the mask of a theorem.
The glut that could not happen had merely been defined out of existence — and it was the definition, not the world, that gave way.
This is why the volume begins here, with a question that sounds like settled arithmetic and proves to be nothing of the kind. Whether an economy can suffer a general failure of demand is no technicality at the edge of the subject; it is the hinge on which all the rest turns. For if the answer is no, the slump is always someone's folly or nature's mischance, and the market needs only to be let alone — while if the answer is yes, the slump may be no one's fault and the market's own, and letting it alone may be the cruelest course there is. The classics answered no, and built a century of confidence upon it. The dissenters answered yes, and were waved away. The Depression answered for them. What the next plate takes up is not whether the general glut can occur — that question the breadlines closed — but what governs how much an economy produces, and why it can fall short of the work willing hands are waiting to do. The argument remains open.