Group Attribution Error
🇳🇴GruppeattribusjonsfeilDefinition
The group attribution error is the tendency to (a) assume that an individual group member's behavior reflects the whole group's preferences, or (b) assume that group decisions reflect each individual's views. We hop between individual and collective without accounting for the fact that groups and individuals aren't identical.
Real-world example
A politician from party X says something controversial. The conclusion 'party X believes this' appears in the comment section immediately – even if the party is deeply split on the issue. Conversely: A board decision is reached by a narrow majority, and afterwards people assume 'everyone on the board agreed' – even though the vote was 5–4.
Allison and Messick (1985) found the effect strongly present in judgments of jury and committee decisions. The effect is related to out-group homogeneity but more specifically concerns how we conflate individual and group levels. It leads people to misread election results ('the people have spoken' when 51% won), to overinterpret committee decisions as consensus, and to brand individual members' escapades as organizational culture.
Supplementary perspective
The bias is dangerous in political analysis because it masks internal disagreement and creates illusions of monolithic groups. The difference between a 'group preference' (aggregated outcome of majority rule) and 'member preferences' (individuals' views) is central to voting theory and Arrow's impossibility theorem – majority decisions cannot be read directly as consensus.
Practical advice
Recognize
- —Notice if you generalize one representative's statement to the whole group.
- —Check whether a 'group decision' hides a narrow margin.
- —Notice if you use 'they think' where you should say 'a majority in that group thinks.'
Counteract
- —Ask for distributions, not just outcomes: How large was the majority?
- —Distinguish official party line from individual members' views.
- —Test generalizations against at least three independent voices from the same group.
Ethical use
- —In journalism: Report vote distributions, not just winning outcomes.
- —In leadership: Don't use individuals' behavior as evidence of organizational culture without data.
- —In political communication: Acknowledge internal variation in your own and others' groups.