Out-group homogeneity bias is the well-replicated tendency to perceive members of groups we do not belong to as more alike one another than members of our own group — “they are all the same,” while “we” are richly varied. Quattrone and Jones (1980) demonstrated the effect experimentally, showing that people judged out-group members as more uniform in their traits and behavior than in-group members, even with equivalent information. The bias arises partly from unequal exposure — we encounter our own group as a crowd of distinct individuals but the out-group more thinly — and partly from how we categorize: the out-group is stored as a type, the in-group as a set of people.
As a manipulation, this bias is a foundational building block of propaganda and dehumanization. By presenting a broad, diverse population as a single-minded mass with one shared motive or threat, an operator makes it possible to argue from one member’s action to a claim about the whole — and to attach fear or contempt to the entire category at once. The technique is efficient precisely because it exploits a shortcut the mind already takes; the messaging need only supply the stereotype and let the bias do the flattening. A telltale signature is asymmetry: the in-group is rendered with full human variety and reasonable disagreement, while the out-group appears monolithic and menacing.
The harm rating sits in the high band because, while a single instance may look like mere rhetoric, out-group flattening is the mechanism that makes stereotyping, scapegoating, and ultimately dehumanization possible — it is where hostility toward a category begins. It pairs naturally with the fundamental attribution error (blaming the out-group’s character rather than circumstance) and with explicit us-versus-them framing. The defense is individuation: deliberately break the category back into specific people, demand named evidence rather than sweeping claims, and seek varied direct contact or first-hand accounts, since exposure to the real internal diversity of any group is the most reliable solvent for the illusion that “they are all alike.”