Ultimate attribution error

The ultimate attribution error is a proposed group-serving pattern in causal explanation. Negative outgroup behavior is more readily blamed on disposition, positive outgroup behavior is discounted, and ingroup outcomes may receive the reverse treatment.

Thomas Pettigrew introduced the concept in 1979. Later evidence supports several ingroup-favoring attribution patterns, but not a universal four-part rule.[^pettigrew][^hewstone]

A fictional example

This scene is wholly fictional and is not study data. Two neighborhood lantern clubs share a workshop. When the visiting club misses a setup deadline, a home-club member calls them disorganized. When the home club missed its previous deadline, the explanation was a late key handoff.

The visiting club later finishes early. Instead of revising the first judgment, the same member credits an easy checklist. The causal standard shifts with group membership and outcome: character explains the outsider’s setback, circumstance explains the home group’s setback, and circumstance discounts the outsider’s success.

The example does not prove prejudice or reveal anyone’s true motive. It illustrates the comparison researchers ask people to inspect.

Pettigrew’s original proposal

Pettigrew extended attribution research into intergroup judgment. For a negative act by an outgroup member, the proposal predicts more dispositional explanation than for the same act by an ingroup member. A positive outgroup act may instead be treated as an exception, luck, special advantage, unusually high effort, or a product of the situation.[^pettigrew]

This is wider than the claim that people simply “blame outsiders.” It concerns how both positive and negative outcomes are explained, and whether the explanation changes when the actor’s group changes.

What the evidence supports, and what it does not

Miles Hewstone reviewed 19 studies of intergroup causal attribution in 1990. Most contained some tendency for explanations to favor the ingroup, across studies of positive and negative acts, success and failure, and perceived group differences.[^hewstone]

The same review concluded that support for Pettigrew’s complete formulation was limited. Results often appeared on a particular attribution dimension rather than as the full pattern. A result on one scale cannot be assumed to recur on every other: ratings of personal causation, ability, effort, chance, task difficulty, and exceptional status can diverge.

So the ultimate attribution error is best treated as a useful theoretical pattern with bounded empirical support. It is not a law applying to every person, group, or intergroup setting. The APA Dictionary retains the standard definition and credits Pettigrew with first describing it in 1979.[^apa]

What the pattern cannot establish

An attribution answer does not directly measure discriminatory behavior. It cannot by itself establish what caused an event, whether a respondent holds explicit or implicit prejudice, or whether a group stereotype produced a later decision.

The pattern may vary with group identification, prior beliefs, status, conflict, the typicality of the target, and the wording of the causal options. Social-identity protection and schema-consistent interpretation are plausible explanations, not a complete account of every result.