Plate ix in print

The Surfeit

choice overload and the paradox of choice; how more options, though any one of them may be ignored, reduce both welfare and the willingness to choose at all.

The theory promises that one more door can never harm, since a door unwanted may simply be left shut; but every door must first be weighed before it is shut, and the weighing tires, and the doors not taken accuse — so that past some small number the abundance meant to free the chooser instead confounds him, and offered everything, he chooses nothing at all.

— the argument of this plate, after Sheena Iyengar, 2000

There is a proposition buried in the theory of choice so obvious that it is almost never stated, and it is this: that more is better, or at any rate never worse. Add an option to the set a person chooses from and you cannot make him worse off, the reasoning runs, for if the new option is inferior he will pass it by untouched, and if it is superior he will take it and gain; either way the larger menu weakly dominates the smaller, and a chooser can only welcome the widening of his field. On this small theorem a great deal rests — the celebration of variety, the faith that competition serves us by multiplying our options, the intuition that freedom is the having of alternatives and that to have more of them is to be more free. The proposition is, within the frictionless world of the theory, perfectly correct. And it is, in the world where people actually choose, frequently and demonstrably false. The chooser offered more does not reliably do better; past a surprisingly low threshold he does worse — buys less, enjoys it less, regrets it more, and sometimes, confronted with too much, declines to choose at all. The plainest demonstration of this came from the psychologist Sheena Iyengar, and it involved nothing more exotic than a table of jam.

The theorem deserves a fair hearing before it is dismantled, because its logic really is compelling. Take any option you please and drop it into the menu. A rational chooser surveys the enlarged menu, identifies the alternative he most prefers, and selects it; the presence of the new option changes his selection only if the new option happens to be the one he most prefers, in which case he is better off by his own lights, and otherwise leaves his selection exactly as it was, in which case he is no worse off. There is no fourth possibility. The extra option is, at worst, inert — a dish on the menu he does not order, a road on the map he does not take. To be harmed by it he would have to be harmed by the mere availability of something he declines, and how could the simple existence of an unchosen alternative cost him anything? The argument is so clean that its conclusion feels less like an empirical claim than a truth of logic. And that is precisely the tell, for it has smuggled in an assumption it never examines: that surveying the menu and identifying the best is free.1

1. The theorem's logic is compelling. Drop any option into the menu: a rational chooser surveys it, finds the alternative he most prefers, and selects it. The new option changes his choice only if it is the one he most prefers — making him better off — and otherwise leaves his choice unchanged, making him no worse off. There is no fourth possibility; the extra option is at worst inert. To be harmed by it, he would have to be harmed by the mere availability of something he declines. So clean it feels like logic, not fact — which is the tell: it has smuggled in the assumption that surveying the menu and finding the best is free.

Iyengar, with Mark Lepper, set up a tasting booth in an upscale grocery and let it offer, on some hours, a modest array of six jams, and on others a lavish array of twenty-four. The lavish display did what abundance is supposed to do at the first stage: it drew a larger crowd, more passersby stopping to look, the greater variety the greater lure. But at the second stage, the stage that matters, it did the opposite of what the theorem predicts. Of those who paused at the small display, a substantial share went on to buy a jar; of those who paused at the large one, almost none did — the purchase rate collapsing roughly tenfold when the options quadrupled. The very profusion that attracted people repelled them from deciding. Faced with six, they could compare, prefer, and act; faced with twenty-four, they browsed, hesitated, and walked away with nothing, the surplus of choice having converted a pleasant small decision into an overwhelming large one. The finding was not a fluke of jam. Iyengar found the same shape wherever she looked: employees offered more mutual funds in their retirement plans grew less likely to enroll in any, forfeiting the employer's matching money rather than face the thicket; students given a longer list of essay topics wrote worse essays and fewer of them; tasters of an extravagant range of chocolates enjoyed their chosen piece less than tasters offered a few. More options, less action; more options, less satisfaction; more options, more paralysis. The theorem's inert extra dish was not inert at all.2

2. Iyengar and Lepper set a tasting booth in a grocery offering six jams some hours, twenty-four others. The large display drew a bigger crowd — abundance as lure. But of those who paused at the small display a substantial share bought; of those at the large one, almost none, the purchase rate collapsing roughly tenfold when options quadrupled. The profusion that attracted repelled them from deciding. Not a fluke of jam: more retirement funds, lower enrollment; more essay topics, worse and fewer essays; more chocolates, less enjoyment of the one chosen. More options, less action, less satisfaction, more paralysis.
⁂

Why does the inert dish bite? Because the theorem's frictionless survey is a fiction, and once friction is admitted the whole tidy argument unwinds. An option cannot be passed by until it has been considered, and consideration is not free: each alternative must be examined, appraised, set against the others, and this labor mounts with the number of options, until the cost of deciding swamps the gain from deciding well. You cannot ignore the twenty-fourth jam; you can only decide to ignore it, and to decide that you must first take it in. This is the first cost, the sheer effort of evaluation, and it is why a large menu fatigues where a small one delights. But there are subtler costs beneath it. A larger field raises the chooser's expectation, for somewhere in twenty-four jams there surely sits a more perfect jam than any of six, and the choice he finally makes is now measured against that loftier possibility and found wanting, so that the same jar tastes worse for having been picked from plenty. And every option not taken becomes a road not traveled, a possible regret; with six alternatives there are five foregone, with twenty-four there are twenty-three, and the mind, choosing, cannot help glancing down all those untaken roads and wondering. Effort, inflated expectation, multiplied regret — these are the charges the extra option levies, and none of them appears in a theory that imagines choosing as the costless reading-off of a preference already formed.3

3. Why does the inert dish bite? An option cannot be passed by until considered, and consideration is not free: each must be examined and appraised, the labor mounting with number until the cost of deciding swamps the gain. You cannot ignore the twenty-fourth jam; you can only decide to, and must first take it in. Beneath this effort lie subtler costs: a larger field raises expectations, so the chosen jar tastes worse for having been picked from plenty; and every option not taken becomes a possible regret, twenty-three foregone where there were five. None of these appears in a theory that imagines choosing as the costless reading-off of a preference already formed.

The psychologist Barry Schwartz gathered these findings into a phrase that named a mood of the age — the paradox of choice — and pressed the case that the modern multiplication of options, exalted as the very substance of freedom, had curdled past some point into a source of anxiety, paralysis, and quiet dissatisfaction, a tyranny dressed as liberation. He may have pressed it too hard, and honesty requires the qualification. The overload effect is real but it is not iron: later studies found it fickle, strong in some settings and absent in others, sensitive to whether the chooser knows his own taste, whether the options are easy to tell apart, whether he has the time and expertise to sort them. A connoisseur is not overwhelmed by a long wine list; a novice is. The lesson is therefore not the crude inversion that less is always more, a slogan as false as the theorem it would replace. The lesson is that the theorem's monotonic promise — more can only help — holds solely in the frictionless limit, and that in the friction-filled world the relation between the number of options and the chooser's welfare is not a rising line but an arch: climbing while options remain few and each adds real reach, cresting at some modest number, and then descending as the costs of evaluation, the inflation of hope, and the accumulation of regret begin to outweigh whatever a marginally better alternative could provide.4

4. Barry Schwartz gathered these into a phrase — the paradox of choice — pressing that the modern multiplication of options, exalted as freedom, had curdled into anxiety and paralysis, a tyranny dressed as liberation. Perhaps too hard: the effect is real but fickle, strong in some settings, absent in others, sensitive to whether the chooser knows his taste and can tell the options apart. A connoisseur is not overwhelmed by a long wine list; a novice is. So not the crude inversion that less is always more, but that welfare against number is not a rising line but an arch — climbing while options are few, cresting, then descending.

So the door that could always be left shut turns out to exact a toll merely for standing there to be shut, and the would-ask of this plate has its answer: added options cannot in fact be costlessly ignored, because the act of ignoring is itself an act of appraisal, and appraisal, expectation, and regret are real burdens that a large enough menu piles past the point where any better choice could repay them. There is, for a given chooser facing a given kind of choice, a best size of menu, neither bare nor boundless, and the abundance heaped beyond it is not a gift but a weight.5 This is one more seam in the elegant cloth — kin to the reversals and the cycles, and closest of all to the inversion of the eighth plate, where the menu itself was shown to be part of what is chosen. For if the contents of the menu shape the chooser, so too does its sheer size; the set he selects from is not a neutral backdrop to his preference but an active force upon it, capable of freezing the very faculty it was meant to serve. The theory hands the chooser a world of open doors and calls it freedom. The chooser, standing before too many of them, discovers that an open door is also a question asked, and that there can be more questions than a person has the heart to answer.

5. The door that could be left shut exacts a toll merely for standing there to be shut: ignoring is itself an act of appraisal, and appraisal, expectation, and regret are real burdens a large menu piles past the point any better choice could repay. There is a best size of menu, neither bare nor boundless. Kin to the reversals and cycles, and closest to the eighth plate: if the menu's contents shape the chooser, so does its sheer size — the set selected from is an active force upon preference, capable of freezing the faculty it was meant to serve.