Although the outgroup homogeneity effect is a robust one in the intergroup relations literature, multiple studies have suggested that its magnitude is moderated by intergroup differences in social power and status. Low power and status groups may actually report greater variability for the outgroup than the ingroup. The present research sought to examine some of the mechanisms responsible for this moderation. Three possibilities exist: (1) one's relative power and status position affects one's perception of group variability; (2) the power or status of a target group affects how one judges its variability; and (3) high power and status groups are actually more variable than low power and status groups, due to the fact that the possession of social power may reduce the strength of normative controls on behavior.

Three studies were conducted to document this third mechanism and to demonstrate differences in the strength of normative social control as a function of social power. In the first, participants were randomly assigned to either high or low power groups. These subsequently interacted and all participants then made judgments about all group members. Participants in study 2 observed videotapes of the group interactions from study 1 and rated all group members, either having been informed or not about the power differential between the two groups. Finally, in the third study, power was manipulated again, but this time a confederate who made racist comments during the group interactions was included in each group. The results of all three studies suggest that individuals in high power groups may actually behave in more idiosyncratic and variable ways than those in low power groups. Additionally, results from the third study suggest that this is because members of high power groups are less likely to attempt the exercise of social control over deviant behavior in their group.