Social Comparison Effect: How Neighbors Change What Enough Feels Like

Most of the things people want to know about themselves have no absolute scale. There is no instrument that says how good a cook you are or whether your income is enough; there are only other people. Leon Festinger's 1954 theory of social comparison processes starts from this: people have a drive to evaluate their opinions and abilities, and when objective standards are not available they do so by comparison with others, preferring similar others because a comparison with someone very different tells you little. The theory has since been applied to almost everything a person can have more or less of.

Income is the clearest case. Erzo Luttmer's 2005 study matched the self-reported happiness of American adults to the average earnings of the people living around them and found that, holding a person's own income fixed, having richer neighbors went with lower reported happiness. The effect was large enough that a rise in neighbors' earnings and a similar-sized fall in one's own income reduced happiness by about the same order of magnitude. Boyce, Brown and Moore reported in 2010, from a British panel, that the rank of a person's income among comparable others predicted life satisfaction better than the income itself did.

Comparison runs in two directions. Thomas Wills set out in 1981 how downward comparison, against people worse off, can protect well-being under threat, and upward comparison, against people better off, can motivate or deflate depending on whether the gap looks closable. Under stress, downward comparisons become more attractive; in pursuit of a goal, upward ones do.

People will pay to be on the right side of the comparison. Solnick and Hemenway asked respondents in 1998 to choose between a world in which they earned more than others but less in absolute terms and one in which everyone earned more and they earned less than others. For income, half preferred 50 percent less real income with a higher relative position. The preference was strongest for attractiveness and praise from a supervisor and weakest for vacation time, which suggests that what gets compared is what is visible and signals standing.

Leon Festinger's "A theory of social comparison processes" appeared in Human Relations in 1954. It is a theory paper, built from numbered hypotheses rather than a single experiment: people have a drive to evaluate their opinions and abilities, they do so by comparison with others when objective means are unavailable, the tendency to compare with a particular person falls as the difference between them grows, and a group's members act to reduce discrepancies. Festinger also proposed a unidirectional drive upward for abilities, so that people want to be slightly better than those they compare with, which has no counterpart for opinions.

Its most direct descendant for the question on this page is the economics of relative income, where the comparison standard is measured rather than assumed. Erzo Luttmer's 2005 paper, "Neighbors as negatives: Relative earnings and well-being," used the 1987 to 1988 and 1992 to 1994 waves of the National Survey of Families and Households, a panel of American adults that asked about happiness and other measures of well-being. He matched each respondent to the average earnings in their Public Use Microdata Area, a Census geography of about 150,000 inhabitants, estimated from national earnings patterns applied to the area's population.

Controlling for a respondent's own income and a long list of personal characteristics, higher earnings among the neighbors went with lower self-reported happiness, and the association was robust to changes in specification. The panel structure let Luttmer check whether the pattern came from people sorting into neighborhoods or from changes in how they defined happiness, and he found no support for either. His preferred explanation was interpersonal preferences: utility that depends on consumption relative to others as well as on consumption itself.

The study is observational, and Luttmer describes the finding as an association with evidence against the main alternative explanations. The demo on this page is the experimental version in miniature: random assignment to a richer or poorer comparison group and a single rating, which can show the direction of the effect once many visitors have answered but says nothing about any one visitor.

The rows below are illustrations of where a reference group does the work of a measurement, not findings from the studies cited on this page.

Comparison cannot be switched off, and Festinger's point was that it does a job nothing else does. The counters are about choosing the comparison on purpose instead of taking the one in view.

• Choose the reference group before you judge. A comparison with the richest neighborhood you can see, or the best-paid person you know, is the one that happens by default; a comparison with your own position five years ago is one you have to choose.
• Convert relative judgments to absolute ones where you can. Ask what the money buys, what the house does, what the job allows, and write those down before asking how they rank.
• Notice who is in view. Luttmer's finding was about the neighbors because the neighbors are visible; a feed of strangers' highlights is a neighborhood too, and one assembled to make the comparison unfavorable.
• Use downward comparison with care. Wills's review found it protective under threat, but a habit of measuring yourself against people worse off trades accuracy for comfort.
• In organizations, be deliberate about disclosure. Card, Mas, Moretti and Saez found in 2012 that employees who learned they were paid below the median for their unit reported lower satisfaction and more intention to look for another job, while those above the median were unchanged.
• Expect the raise to fade. If the satisfaction from a gain depends on rank, and everyone gains, the satisfaction is short-lived.

The social comparison effect is about the standard used for a self-evaluation. Several neighbors concern standards too, and the differences matter.

• It is not the contrast effect, the general finding that a stimulus is judged differently depending on what was seen just before it. Social comparison is the special case in which the standard is another person's situation and the thing judged is your own.
• It is not illusory superiority, the tendency to rate yourself above average on desirable traits. That is a direction the comparison tends to take; the social comparison effect is the dependence of the judgment on who is being compared with, in either direction.
• It is not anchoring, where a number seen first pulls an estimate toward it. The neighbors' earnings do not pull your estimate of your own income; they change how that income feels.
• It is not the focusing effect, where one feature of a situation is given too much weight in a judgment of the whole. Comparison can operate on a single feature or on everything at once.