Fundamental Attribution Error: Blaming the Person, Not the Situation

When you watch someone act, the person is in the foreground and the situation is the background, and explanations follow attention. The driver who cut in is aggressive; the fact that their lane was ending is not visible from your seat. The colleague who snapped in the meeting is difficult; the call they took ten minutes earlier is not part of what you saw. Lee Ross named this pattern the fundamental attribution error in 1977 and argued that it was the most basic distortion in how people explain one another: a bias toward dispositions over situations.

The essay task shows the pattern at its sharpest, because the situation is not merely visible but stated in advance. Readers are told that the writer was assigned the side, which fully explains the essay. Readers still move their estimate of the writer's own view toward the side argued, as if the words had leaked something about the person who wrote them. Daniel Gilbert and Patrick Malone, reviewing this literature in 1995 under the name correspondence bias, described it as a failure to fully correct: the reader sees the behavior, infers the matching attitude almost automatically, and then adjusts for the situation only partly, especially when tired, distracted or in a hurry.

The error is often paired with a claim about the self: Jones and Nisbett proposed in 1971 that people explain their own behavior by circumstances more readily than other people's, the actor-observer asymmetry, which has its own page. That claim is weaker than textbooks suggest. Bertram Malle's 2006 meta-analysis of 173 studies found average effect sizes close to zero, with the asymmetry appearing mainly for negative events, for intimates and for hypothetical or idiosyncratic behavior, and reversing for positive events. The correspondence bias does not depend on it. Where a self-other difference does appear, the reasons are partly perceptual, since you can see your own situation and not your own face, and partly informational, since you know your own history and have no such record for a stranger.

The size of the effect depends on culture and on how much the situation is made salient. Joan Miller found in 1984 that adults in the United States explained behavior by traits far more than adults in India, who leaned on context and relationships, and Choi, Nisbett and Norenzayan's 1999 review found that East Asian participants made the correspondence bias too but corrected for strong situational information more fully. The bias is widespread rather than universal, and it is reduced when people are given both the motive and the time to think about the circumstances.

Edward Jones and Victor Harris published "The attribution of attitudes" in the Journal of Experimental Social Psychology in 1967, ten years before Ross coined the modern name. Their participants, students at Duke University, read short essays about Fidel Castro's Cuba and were asked to estimate the true attitude of the person who had written each one. Ten seven-point items were summed into a score from 10, strongly against, to 70, strongly in favor.

The essays came in two versions and under two conditions. Some essays were pro-Castro and some anti-Castro, and all were presented as written by students for a political science course examination. In the choice condition, readers were told that the instructor had let the writer pick which side to argue. In the no-choice condition, they were told that the instructor had assigned the side. The question was how far readers would discount the essay's content when they knew it had been ordered.

They discounted it, but not nearly enough. In the choice condition, writers of pro-Castro essays were rated at a mean of 59.6 and writers of anti-Castro essays at 17.4, which is what you would expect when a person chooses their side. In the no-choice condition the gap shrank but stayed large: 44.1 for pro-Castro writers and 22.9 for anti-Castro writers. Readers who had been told plainly that the writer had no choice still placed the pro-Castro writer well above the anti-Castro writer.

Jones and Harris had predicted that assignment would shrink the difference, with the size of the discount depending on whether the position argued was the popular or the unpopular one; the discount they found was much smaller than the instructions warranted. Later work made it the standard demonstration. Ross, Amabile and Steinmetz showed the same thing with a quiz game in 1977: questioners who were randomly assigned to invent hard questions were rated as more knowledgeable than the contestants who failed to answer them, by the contestants and by observers alike, even though everyone knew the roles had been assigned by chance.

The error appears whenever one person judges another from the outside.

The bias runs on attention, so the counters are ways of bringing the situation into view before the trait judgment settles.

• Ask the situational question first. Before deciding what kind of person did something, ask what circumstances would lead an ordinary person to do it, and whether those circumstances could have been present.
• Look for the record. One act is evidence about a situation; a pattern across situations is evidence about a person. Judge dispositions only from patterns.
• Notice roles. Lawyers, spokespeople, debaters, customer service staff and quiz hosts all behave as their role requires. Jones and Harris's readers knew the role and still inferred the attitude; knowing is not enough, so make the correction a deliberate step.
• Swap seats. Ask how you would describe the same act if you had done it, and borrow that explanation. Your own account would probably give the circumstances more room, because you can see them.
• Slow down the judgment when it matters. Gilbert and Malone's account places the failure in the correction step, which needs attention. A hiring decision or a performance review deserves the extra minute.

Several biases in the attribution family are easy to confuse, and three of them are near neighbors of this one.

• It is not the actor-observer bias, which is the difference between how you explain your own behavior and how you explain other people's. The fundamental attribution error describes the observer's side; the self-other difference itself is smaller and more conditional than textbooks suggest.
• It is not the self-serving bias, which credits your own successes to yourself and your failures to circumstances. That pattern depends on the outcome being good or bad; the fundamental attribution error applies to other people regardless of outcome.
• It is not the ultimate attribution error, which is the same move applied to groups: the negative acts of an out-group member are explained by what the group is like, and positive acts by luck or exception.
• It is not the halo effect, where one known trait colors the judgment of others. The halo effect spreads a trait across traits; the attribution error produces a trait from an act.