Extrinsic Incentive Error: Why Other People Seem to Work for the Money

People have direct access to their own reasons and only a theory about everyone else's. When you ask yourself why you stayed late, the answer comes from the inside: the problem was interesting, the team needed it, you wanted to see it finished. When you ask why a colleague stayed late, you have no such access, so you reach for the most visible, most general explanation available, and in working life that is money. The lay theory of motivation that most people carry says that others respond to carrots and sticks; the theory they apply to themselves says that they care about the work.

Heath's account separates this from simple flattery. A self-serving bias would make people claim every attractive motive for themselves and none for others. The extrinsic incentive error has a specific shape: people grant others the motives that are easy to see from outside, pay, security, perks, and keep for themselves the motives that are only visible from inside, interest, learning, meaning. Heath noted that this runs against the classic actor-observer pattern, in which actors explain their own behavior by the situation and observers explain it by the person's traits; here the observers reach for a situational force, the paycheck, and the actors reach for something internal. The explanation he offered is nonetheless the same: observers see the environment, actors feel their own states.

The error has consequences because people act on their lay theories. A manager who believes the team works for money designs bonuses and monitoring; a team that actually works for interest and autonomy reads those as a lack of trust, and effort falls. Heath framed this as a problem for organizing itself: principals who hold the wrong theory about agents offer the wrong deals. Kovach's 1987 survey of industrial employees is the practical illustration most often cited alongside it. Employees ranked interesting work first, full appreciation of work done second and a feeling of being in on things third; good wages did not make their top three.

Two cautions. The evidence is about predictions and rankings, and a ranking is not a measurement of what moves anyone's hand at work. And pay does matter: nobody in these studies rated it as irrelevant, and when pay is low or unfair it dominates everything else. The error is in the gap between what people say about themselves and what they assume about others, not in caring about money.

Chip Heath, then at Duke University, published "On the social psychology of agency relationships: Lay theories of motivation overemphasize extrinsic incentives" in Organizational Behavior and Human Decision Processes in 1999. The paper's starting point was economic: an agency relationship is one in which a principal, an employer or a client, has to design incentives for an agent whose motives the principal cannot see. If principals hold a mistaken theory of what motivates agents, they will design the wrong incentives, and Heath set out to test whether the ordinary theory of motivation is mistaken in a predictable direction.

He reported four studies, three in the laboratory and one in the field. In three of them people rated or ranked a set of motivators, outside rewards such as pay, benefits and job security alongside inside rewards such as learning new things, developing skills and doing something worthwhile, once for themselves and once for other people, such as peers, co-workers or a group of employees they were asked to think about. The third study asked a different question, about how incentives are communicated: people rated alternative framings of a fixed incentive program, once for how motivating they would find each framing themselves and once for how motivating the people the program was aimed at would find it. The comparison between the two sets of answers is the finding in every study: people put the extrinsic items, and the framings that stress them, higher for others than for themselves and the intrinsic items lower for others than for themselves. Heath named the pattern the extrinsic incentives bias.

Two further results shaped how the finding is read. First, the bias could be separated from a self-serving bias: it was not that people claimed all the good motives, but that they assigned to others the motives visible from outside. Second, Heath presented it as an empirical counterexample to the traditional actor-observer effect in social psychology, while arguing that the theoretical explanation is similar: observers notice the situation around a person, and in working life the most salient feature of the situation is the paycheck.

The finding has been reproduced in spirit many times since in work on what people believe about motivation, now often called metamotivation, where people tend to overestimate the motivating power of rewards and underestimate how much they themselves will be moved by interest. Kovach's 1987 survey, which found that employees ranked interesting work above wages, predates Heath and is the field counterpart most often placed beside his laboratory results.

The error shows up wherever one person has to guess what will move another.

The counter is to grant other people the same inner life you grant yourself, and to check rather than assume.

• Use your own answer as the estimate. When you catch yourself assuming a colleague is in it for the money, ask what moves you in the same role, and start from the assumption that the answer is similar.
• Ask, and ask in private. People will tell you what matters to them if the question is not attached to a negotiation. Exit conversations and one-to-ones are where the real ranking appears.
• Design incentives for the person you would be. A plan built on interest, autonomy, recognition and fair pay fits most people better than a plan built on money alone, because most people describe themselves that way.
• Keep pay fair and then stop talking about it. Low or unfair pay dominates every other motive; adequate pay recedes into the background and leaves room for the rest.
• Notice the asymmetry in your explanations. If your own late nights are about the work and theirs are about the bonus, the difference is more likely to be in your vantage point than in the two people.

The extrinsic incentive error belongs to a family of mistakes about other minds, and its neighbors point in different directions.

• It is not the false consensus effect, in which people assume others share their own views and choices. The extrinsic incentive error runs the other way: people assume others differ from them, specifically in caring more about money and less about the work.
• It is not the fundamental attribution error, the habit of explaining other people's behavior by their character rather than their situation. Here observers reach for a situational force, the paycheck, which is why Heath presented the finding as a counterexample to the usual actor-observer pattern.
• It is not naive cynicism, the expectation that others will claim credit and shift blame self-servingly. Naive cynicism is about how others judge; the extrinsic incentive error is about what others want.