Mixed Signals Summary and key ideas

Mixed Signals, by Uri Gneezy.

An action can do more than produce a practical result. It can also tell other people what someone values and help that person understand who they are. Incentives, then, have more than a direct economic effect: they can change the meaning people attach to an action. A payment might make a task more worthwhile, but it may also change what the task appears to say about the person doing it.

The Disney example makes that second message visible. Uri Gneezy recounts telling staff that his son was almost three to avoid a $117 ticket, although the boy had turned three a couple of months earlier. Gneezy had taught his son that honesty distinguishes good people from bad people. The boy noticed the contradiction. To him, his father’s action seemed to say that saving money could matter more than telling the truth. The ticket price was the immediate saving; the child’s interpretation was another consequence of the choice.

This is why a claim supported by an action can be more convincing than words alone. Words are often easy to say without accepting any cost. A choice that carries a real cost can make the message harder for someone with different priorities to imitate. The cost need not be money. It can be effort, lost opportunity, or a risk to how others see you. What matters is whether the choice is especially burdensome for the person who does not hold the value being signaled.

Gneezy illustrates the idea with Jim and his friends, who ride Harleys and wear biker clothes during annual breaks. The clothes can come off when the break ends, so someone can adopt the appearance temporarily without changing much about their life. A neck tattoo is a more lasting declaration. For someone planning to return to conventional office life, visible tattoos could carry a career cost. For a person who identifies with biker culture, that same choice may fit their desired life and recognition. The tattoo is therefore a stronger signal in this example: the person who values the identity has more reason to accept its cost, while the person who does not may prefer not to take it on.

The price of a signal depends on who sees it and what they make of it. A tattoo that is covered, or in a place customers never see, may communicate little to an employer or guest-facing audience. A visible one can affect what an employer thinks the business should present to its guests, regardless of the wearer’s job performance. Meanings also change across settings and over time. Clothes, hairstyles, tattoos, and other choices do not carry one permanent interpretation for every audience.

The early Prius offers a product example. Early hybrid cars cost more and had drawbacks in areas such as speed, acceleration, comfort, and safety, even as they used less fuel. For buyers concerned about the environment, accepting those disadvantages could signal a willingness to make a sacrifice. Toyota later gave the Prius a distinctive shape, making hybrid ownership easy for coworkers and others to notice. A subtle marker might tell a careful observer what kind of car it was; a conspicuous design made the choice legible at a glance.

That visibility mattered because a signal has little social value if its intended audience does not notice it. In a cited 2007 survey, 57 percent of Prius buyers said they bought it because it made a statement about them; 36 percent cited fuel economy, and 25 percent cited low emissions. Those responses suggest that identity and practical benefits could both matter. They do not establish that the styling alone caused the Prius’s sales growth.

A signal can also lose force when the sacrifice behind it shrinks. As the Prius improved in fuel savings, comfort, and reliability, buyers could choose it for practical advantages without having strong environmental concern. The car could still express an identity, but the purchase no longer carried the same evidence of willingness to give something up. The signal weakened as it became less costly for people with different priorities to make the same choice.

The same distinction between a choice’s direct value and its message appears in a gift example from Seinfeld. Jerry gives Elaine $182 in cash, although she wanted something thoughtful. The money has an obvious face value, but it can suggest that little effort went into choosing it. Kramer gives her a less costly bench she had been looking for. Its value lies partly in what it communicates: he remembered what she wanted. The examples show why judging a choice only by its price or usefulness can miss the meaning received by its audience.

Signals can also be directed inward. Social signaling concerns what other people infer; self-signaling concerns what we infer about ourselves from our own actions. The two can reinforce each other, but they can also pull apart. A person may feel generous because they gave, while also caring about whether someone else noticed. Whether the act is public or private can change which of those motives matters most.

Gneezy and his colleagues explored this at Der Wiener Deewan, a restaurant in Vienna where customers chose how much to pay. In their field experiment, some customers handed a questionnaire and their payment to a waiter. Others placed both in a sealed envelope and dropped it in a box. The second arrangement made the payment anonymous to the restaurant staff, reducing the chance of gaining a social image from the amount paid.

If customers paid only to impress others or avoid looking cheap, anonymity should have made payments fall sharply. Self-image could still matter: even without an audience, a diner might not want to think of themselves as taking advantage of the restaurant. Observed customers paid an average of €4.66. Anonymous customers paid significantly more, by €0.71 on average. That result was not what a simple account of social and self-signaling as added benefits would predict.

The authors suggest that being watched may sometimes weaken the private meaning of generosity. An observed customer may feel they are paying to impress or avoid disapproval, making the act seem less like evidence of personal generosity. Alone, the customer can take the payment as a clearer sign to themselves. This is an interpretation of the result, not proof that every diner had the same motive or that anonymity always increases giving. It comes from one restaurant and one experiment.

When interpreting an action or designing an incentive, ask four questions. Who can see the choice, and whose judgment matters? What does it cost the person making it? How easily could someone who does not hold the stated value imitate it? And could a different audience, setting, or convention give the same action another meaning? These questions help distinguish a choice’s direct payoff from the message it sends to others and to the person who makes it.