Greater social identity complexity predicts more tolerance toward out-groups

Evidence

In a number of preliminary studies in the United States and Israel, Roccas and Brewer (2002) asked samples of American and Israeli university students to rate the perceived overlap and perceived similarity between a number of predefined groups (for U.S. students, the categories used were white, American, college students, and religious denomination; for Israeli students, the categories used were secular, Israeli, college students, and Jewish). In the U.S. sample, the authors found that individuals who were higher on overlap complexity also scored significantly higher on measures of tolerance toward racial or religious out-groups. The nature of the relationship between similarity complexity and outgroup tolerance was similar, yet did not reach statistical significance. Similarly, in the Israeli sample, the authors found that both overlap and similarity complexity were significantly positively associated with willingness to engage in and acceptability of contact with immigrants from the former Soviet Union. In a further study involving 210 Ohio citizens, Brewer and Pierce (2005) found social identity complexity (measured by means of an overlap complexity measure) to be significantly associated with a range of social attitudes and tolerance measures, including, for example, greater support for multiculturalism and affirmative action and feeling thermometer ratings of specific out-groups. — Roderick M. Kramer, Geoffrey J. Leonardelli, Robert W. Livingston, Social Cognition, Social Identity, and…, 533

Brewer and colleagues have argued that social identity complexity should positively covary with out-group attitudes (Brewer & Pierce, 2005; Roccas & Brewer, 2002), for reasons similar to those underlying the notion of crossed-categorization. Doise’s (1978) category differentiation model predicts that once social identity becomes salient, this gives rise to intergroup discrimination whereby intragroup differences are attenuated and intergroup differences accentuated. This process provides the cognitive basis for in-group bias to occur (see also Vanbeselaere, 1991). … In the context of multiple crossed categories, this general tendency to minimize differences within but maximize differences between categories should be reduced or even eliminated, given that individuals rely on more complex category combinations. Moreover, when multiple categories are salient, individuals are thought to rely less on single in-groups for bolstering self-evaluation and self-esteem 82 • Katharina Schmid and Miles Hewstone (see Vanbeselaere, 1991) or satisfying opposing needs for assimilation and differentiation (Brewer, 1991; see also Leonardelli, Pickett, Joseph, & Hess, Chapter 6, this book; we return to a more detailed description of optimal distinctiveness theory at a later point in this chapter). — Roderick M. Kramer, Geoffrey J. Leonardelli, Robert W. Livingston, Social Cognition, Social Identity, and…, 529

Categorization

• QUESTION: Notably, similarity complexity (perceiving groups as similar to each other) did NOT reach statistical significance in the US sample — only overlap complexity did. Worth chasing why similarity and overlap complexity diverge.
• SEED: Angle: the ‘cross-cutting identities as anti-prejudice’ intervention — could tests of overlap complexity double as a measurable, trainable cognitive skill for reducing bias, better than implicit-bias training?
• XREF: Connects to the reader’s social capital material from Coleman (intergenerational closure) and to Milgrom’s institutional analysis — both about how group boundaries and overlap affect trust and cooperation. — connects to: Accepting relationships with out-group members model new identity structures for exiting extremists. — Amy Fisher-Smith, Charles R. Sullivan, John D. Macready, and Geoffrey Manzi, Identity Reconfiguration and the Core…, 475; the MOC “The False Compass: How Racial Categories Reorder Bodies, Minds, and the Self”
• ADDENDUM: Relates directly to the reader’s prejudice-psychology thread; the crossed-categorization basis for reduced out-group bias is what prior Allport argument predicted.
• corroborated: 4 independent passages support this claim — Roderick M. Kramer, Geoffrey J. Leonardelli, Robert W. Livingston, Social Cognition, Social Identity, and…, 533; Roderick M. Kramer, Geoffrey J. Leonardelli, Robert W. Livingston, Social Cognition, Social Identity, and…, 529