There is a Latin tag the economists are fond of — de gustibus non est disputandum, of tastes there is no disputing — and it marks the exact spot where the science of choice declines to look. The model of the rational chooser begins with his preferences already in hand: complete, settled, his own, the fixed point from which all his choosing flows. Where they came from is held to be no business of the economist, who takes them as the physicist takes the charge on an electron, a given of the problem and not a thing to be explained. This is a convenience, and for many purposes a harmless one. But it conceals an assumption of great consequence — that a person's wants are his own — and that assumption, looked at squarely, is not obviously true.
The first plate left its maximizer waiting, blandly, to be told what he wants, and it is time to ask who tells him. The model itself does not; that is the point of beginning here. Preferences enter the theory of choice as a premise, never a conclusion: the economist is handed a ranking and asked what follows, as a navigator is handed the winds and asked to plot a course, with no duty to say why the winds blow. The wants are exogenous — they come from outside the model, set down at its edge, and the machinery takes them as it finds them. For a vast range of work this is exactly right: to explain why butter sells better when its price falls, one needs no theory of where the love of butter was born. But the silence has a cost, hidden precisely because it is so convenient. By placing the wants outside the frame, the model excuses itself from ever asking the most consequential thing about them — whether they are, in any deep sense, the chooser's own.
The refusal to look is not mere laziness; at its best it is a discipline, and its ablest defenders made a virtue of it. Stigler and Becker, in a paper whose title is the very tag we began with, argued that the economist should resist always the temptation to explain a change in behavior by a change in taste. For tastes so invoked explain anything: did consumption rise? then the taste grew; did it fall? then the taste shrank — and an explanation that fits every case illuminates none. Far better, they urged, to assume that tastes are stable over a life and much the same from one person to the next, and to insist that every shift in conduct be traced to something observable — a movement in prices or incomes. On this account the fixity of preferences is no naivety but a self-denying ordinance, forcing the economist to find real causes rather than retreat into the soft sands of changing whim. There is genuine rigor in it.
Behind the methodological case stands a moral one, weightier still. To take a person's wants as given is, in the end, to take them as authoritative — to hold that he is the rightful judge of what is good for him, and that his choices deserve a certain respect simply as his. This is the doctrine of consumer sovereignty, bound up with the deepest commitments of a liberal order: the refusal to rank one person's conception of the good life above another's, the suspicion of any authority claiming to know a man's true interests better than he knows them. The moment one divides wants into the authentic and the spurious, it must be asked who does the dividing — and the answer is always some person or board or party that thereby gains a power over others' lives which the respect for given preferences was meant to withhold. The economist's incuriosity about where wants come from is, in this light, a kind of restraint, even a kind of decency.
And yet the wants will not stay where the model puts them, for it is plain on the least reflection that they are made as much as had. Consider how a person comes by them. No one selects his first preferences; they are laid down before he could consent — by the parents who shaped his palate and his faith, the class whose habits he absorbed, the nation and age that decided, before his birth, what would seem to him normal and desirable and fine. The Frenchman's taste for wine and the Englishman's for beer are given in the most literal way: given to each, from outside, by an accident of origin he had no hand in. They are his now, and he will defend them as his own. But to call them his own creation, the free expression of some inner self that precedes them, is to forget that the self was furnished with them before it was a self that could have wanted otherwise.
If formation shapes wants without our leave, persuasion sets out to do so on purpose, and here the model's serenity grows hardest to keep. An entire industry exists whose business is the manufacture of desire — the advertiser's art, which does not merely tell us a thing is for sale but labors to make us want it, and would best of all make us want what would never have occurred to us unbidden. Galbraith gave the thing its name. In a society poor enough that its wants are the urgent ones of hunger and cold, he allowed, production answers to wants that plainly come first; but in an affluent society the sequence can run backward, the want called into being by the very process that stands ready to satisfy it, so that production and desire issue from one source. He called it the dependence effect, and its sting is sharp: a system that produces the wants it fills can claim little credit for filling them, and the sovereignty of a consumer whose wishes are supplied to him by the seller is a sovereignty in name only.
There is a subtler shaping yet, which needs no advertiser, for the wants bend of their own accord toward whatever the world allows. Aesop's fox, finding the grapes beyond his leap, decides they are sour and wants them no more; and in that small fable lies a large discomfort about desire — that it adapts to its circumstances, shrinking to fit the possible so as to spare its owner the ache of wanting what he cannot have. The poor learn not to crave what they will never afford; the captive makes peace with his confinement; the girl raised to expect little learns, in self-protection, to ask for little. These adaptive preferences, as they have been called, are a defense the mind mounts against perpetual frustration, and a mercy to those who hold them. But they spring a trap on any science that would read a person's good from his wants, for a contentment made by the shrinking of hope is not the same as a life that is genuinely good — and the two can look, from outside and in the record of choice, exactly alike.
The conclusions press in hard — and would press harder still if the critique did not carry a danger of its own. If wants are formed, manufactured, and adapted, then the comfortable claim that the market gives people what they want grows hollow, for the market supplies the wants along with the goods, and a welfare read off such wants may measure only the success of the conditioning. So far the critic. But Hayek returned a powerful answer that must be granted its force. Nearly every want worth having above brute necessity, he observed, is produced in us by others: no one is born wanting symphonies or philosophy, and such desires are instilled by a culture as surely as any jingle instils a thirst — yet we do not count them false, or less our own, for having been taught. If being produced made a want spurious, the noblest human wants would be the most spurious of all. The critique, pushed to its end, would dissolve the very self it set out to defend; and the line between a taste cultivated and a taste manufactured, on which all of it turns, is one no one has drawn to general satisfaction.
Strip the matter to its core, and the want stands exposed as the deepest instance of this work's recurring discovery. Through every volume the pattern has held: a thing wearing the face of a natural fact proves, on inspection, to be a verdict — the price that seemed a property of the goods, the wage that seemed the worth of the work, the idleness that seemed an act of nature. The model of choice rests its whole weight on the one thing it declines to examine, the preferences, treating them as the bedrock beneath all the verdicts, the single fact that is simply and innocently given. But the preferences are the least innocent thing of all. They too are verdicts — laid down by an upbringing the chooser did not pick, sharpened by a persuasion aimed at him on purpose, bent by the quiet adaptation of hope to circumstance. A price is a verdict on a hidden contest, mistaken for a fact of nature; and the want, the most private-seeming possession a person has, is the verdict on the longest and most hidden contest of all — the one between the person and everything that made him.
What the plate cannot do is resolve the matter cleanly, and the inability is itself the lesson. We cannot say the wants are simply the chooser's own, for we have watched them installed, manufactured, and trimmed by forces he neither chose nor saw. But neither can we say they are simply not his, and reach past them to some truer interest he would hold if only he were undeceived — for that road runs straight to the planner who knows our good better than we do, and to every tyranny that ever called itself liberation. The taste is the seam where the spare machinery of choice meets the whole untidy world that forms a person, and the seam will not be sewn shut. To respect a person is to take his wants as his own; to understand him is to see that they are not only his. Both are true, and they will not lie down together.
The want is the most private thing a person owns, and among the least purely his — the bedrock the whole model rests on, and the one stone that proves to be sediment.
This is the question the maximizer could not ask, and it completes the volume's opening: the one plate gave us the engine of choice, this one the doubtful fuel it burns. The plates that follow work outward from the seam. If the wants are shaped, then how a choice is framed may shape it too, and the nudge becomes a power; if the wants are adapted, then the welfare we read from choice is unreliable, and the measure of a person's good begins to totter; if the wants are not wholly his, then the self that holds them may be divided, or social, or other than the lone calculator the model drew. A price is a verdict on a hidden contest, mistaken for a fact of nature; and the deepest such verdict, it turns out, is the one we are most sure is ours alone — what we want. The argument remains open.