№ VIII
The twelve plates
Every price is supposed to measure a want — the satisfaction a good brings the one who buys it. But want cannot be seen: no instrument reads pleasure, no scale weighs desire, and the quantity on which the whole science of value rests is inferred only from the choices it supposedly drives. This volume anatomizes utility, the ghost in the machine: the subjective theory of value that placed it at the center, the unobservability that makes a price's claim to measure it run in a circle, and the welfare, the happiness, and the justice that economics tried, and failed, to read from it.
if want can never be seen, only inferred from the choices it supposedly drives, then what does a price measure — the satisfaction it claims to, or merely the willingness, and the power, to pay?
Economics is built on a quantity it cannot see. Beneath every demand curve, every market price, every calculation of gain and loss, lies the idea of utility — the satisfaction a good affords, the want it answers, the pleasure or the use it brings the one who has it. Utility is what the consumer is supposed to maximize, what the price is supposed to reflect, what the whole apparatus of value is built to track. It is the bedrock on which the modern theory of value rests: a thing is worth what it is worth to someone, and what it is worth to him is the utility it yields. Remove the concept and demand has no foundation, welfare no meaning, and price no anchor in anything human. And yet, as this volume will show, utility has a peculiarity that unsettles everything built upon it — it can be felt by the one who has it, and seen by no one at all.
The elevation of utility to this central place was the work of a revolution in thought, and a recent one. For most of the history of economics, value was held to come from the side of production: a thing was worth the labor that made it, as Smith and Ricardo broadly taught, or the cost of bringing it forth. The diamond's great price and water's small one were a standing embarrassment to this view, since water is useful beyond measure and diamonds barely at all. In the eighteen-seventies, working separately, William Stanley Jevons in England, Carl Menger in Austria, and Léon Walras in Switzerland overturned it: value, they argued, comes not from the cost of a thing but from the want it satisfies — from its utility to the person who wants one more of it. Jeremy Bentham had prepared the ground a century before, by making pleasure and pain the springs of all human action and the foundation of morals. With the marginalists, want became the source of worth, and value passed from the object into the mind that desires it.
But the want they had enthroned at the center of value has a property that no amount of theory can wish away: it cannot be observed. Satisfaction is felt, not seen; it registers on no instrument, leaves no residue, casts no shadow that another can measure. One may know one's own pleasure with perfect intimacy and never glimpse another's at all, save by guess and sympathy. The utility on which every price is said to rest is thus a ghost in the machine — the cause of every choice and the content of no observation, driving the visible world of demand from somewhere wholly hidden. This is not a difficulty at the edges of the science but a flaw at its foundation, for a theory that explains prices by utility owes some account of how the utility might be known; and the honest answer, that it cannot be known directly at all, threatens to make the whole explanation a tissue of the unobservable. The want that governs everything can be read nowhere.
Economics met this embarrassment not by finding the want but by ceasing to look for it. If utility cannot be measured, perhaps it need not be: one may simply watch what people choose and let the choice stand for the want. This was the path from the cardinal utility of the early marginalists, who imagined satisfaction came in countable units, to the ordinal utility of Vilfredo Pareto, who showed that most of the theory needs only that wants be ranked, not measured, and finally to the revealed preference of Paul Samuelson, who proposed to dispense with utility as an inner state altogether and define it by the choosing: if a man takes A when he might have had B, then A is, by definition, preferred. The maneuver bought rigor by retreating from the mind to the deed. But it carried a difficulty of its own, which this volume will not let pass: that to infer the want from the choice and then explain the choice by the want is to argue in a circle, and a circle, however elegant, explains nothing it did not assume.
And there is a further trouble, graver than circularity, which returns us to the argument of the whole work. Even if a price faithfully tracked the choices people make, it would not thereby measure their wants — for what governs a market choice is not how much a person wants a thing but how much he is willing, and able, to pay for it. Willingness to pay is want filtered through means: the rich man's idle fancy commands more dollars than the poor man's desperate need, and so speaks louder in the only language the market hears. A price is a verdict on a hidden contest, mistaken for a fact of nature; and the price that claims to measure utility measures, in truth, willingness to pay — want weighted by the wealth behind it — so that what looks like a reading of satisfaction is a reading of the distribution of money. The demand curve sums dollars, not desires; and a thing is "worth" what someone will give for it, which depends on what he has to give.
This is the quiet scandal beneath the doctrine that price measures utility: the doctrine launders a distribution of purchasing power into a measure of value, and calls the result the worth of things. A famine can rage while grain is carried out of the stricken land — as Amartya Sen showed, famines have struck where food was ample but the hungry had lost the means to buy it — because the starving, having no money, generate no demand, while the well-fed elsewhere, having money, do; and the market, faithfully measuring willingness to pay, records that the grain is worth more in the mouths that are already full. A price is a verdict on a hidden contest, mistaken for a fact of nature; and the contest, here, is over who holds the means to make his wants count. Utility was offered as the bedrock natural fact beneath every price, the subjective ground that would explain value once and for all. It turns out to encode, beneath its show of measuring desire, the wholly unnatural question of how the means of desiring are shared out.
None of this makes utility useless, and the volume will not pretend that it does. The concept organizes the whole theory of demand; it lets us speak, however carefully, of people being made better or worse off; and the retreat to ranking rather than measuring rescued much of its work from the charge of weighing the immeasurable. But it leaves utility able to bear far less than it is often asked to carry. It cannot compare one person's satisfaction with another's, as Lionel Robbins insisted, and so cannot of itself justify taking from the rich to give to the poor; it cannot be summed into a coherent will for a whole society, as Kenneth Arrow proved; and it may not, in the end, track the happiness it was first invented to name. The volume's task is to mark, with care, the exact edge of what this central and indispensable quantity can and cannot be made to tell us — to honor its uses without crediting its pretensions.
The twelve plates follow utility from the want at its root to the worth at its end. They set out the want itself — the subjective theory that made desire the source of value — and the ghost, the unobservability that makes a price's claim to measure it run in a circle; the calculus by which Bentham dreamed of counting pleasure, and the wager through which Bernoulli and his heirs measured it by the willingness to bet; the gulf between persons that no utility can bridge, and the welfare economics built to do without the crossing; the leveller, the case for equality that diminishing want seems to promise and cannot prove, and the sum that Arrow showed could never be coherently taken; the happiness utility may have lost in gaining its rigor, and the currency it claims to be when it renders all goods in one coin; the axioms beneath the modern theory, and the choices that break them; and, at the last, the worth of a thing, and what utility was ever supposed to mean. Other volumes meet this one at every turn: the volume on the margin, which traced how value came to be reckoned at the last unit; the choice, which studies the chooser utility is built to describe; the surplus, the gain a want fulfilled leaves over its price; the stock, the purchasing power whose distribution every price quietly records. Utility is the quantity at the center of the science, indispensable and unseen — the want every price is built to measure, and the ghost no instrument can read. The argument remains open.