The Contact Hypothesis, Superordinate Identities and Interdependence

1A football league in northern Iraq

After the armed group ISIS was driven out of northern Iraq, many Christians who had been displaced distrusted their Muslim neighbours. In 2019 the political scientist Salma Mousa organised football leagues for Christian amateur players. By lottery, some Christian players were placed in teams with only Christian teammates, and others in teams that also included Muslim players.

By the end of the season, Christian players with Muslim teammates were more likely to vote for a Muslim player from another team to receive a sportsmanship award, to sign up for a mixed team the following season, and to be training with Muslims six months later. But their behaviour towards Muslims outside football, such as visiting a restaurant in the mainly Muslim city of Mosul, did not change (Mousa, 2020).

Contact on a team worked, and its effects were real but limited. This lesson asks why contact works, under what conditions, and how shared identities and shared goals can strengthen it.

21. Allport's four conditions

Allport argued that contact reduces prejudice best when:

1. Equal status: within the contact situation, members of both groups have equal standing (for example, all are players, not players and servants).
2. Common goals: they work towards the same goal.
3. Intergroup cooperation: they must cooperate, not compete, to reach it.
4. Support of authorities, law or custom: leaders and institutions clearly support the contact.

The evidence

• Thomas Pettigrew and Linda Tropp (2006) combined 515 studies with more than 250,000 participants in 38 countries. Contact was associated with lower prejudice in the great majority of them. Effects were larger when Allport's conditions were present, but contact helped even when they were not fully met, and the effects extended from the individuals met to the out-group as a whole.
• Their later analysis (2008) found that contact works mainly by reducing anxiety about the out-group and increasing empathy and perspective-taking; simply learning facts about the other group mattered less.
• Even extended contact, knowing that a friend from your own group has a friend in the other group, is linked with less prejudice (Wright and colleagues, 1997).

The limits

• Most studies are correlational: people with less prejudice may simply seek more contact. Elizabeth Levy Paluck, Seth Green and Donald Green (2019) found only 27 randomised studies that measured effects at least a day later; they confirmed positive effects but noted that few involved adults and that effects on ethnic and racial prejudice were weaker.
• Negative contact, such as an insult or threat, can increase prejudice, and a single negative experience may weigh more than a positive one (Barlow and colleagues, 2012).
• As Mousa found, changes in behaviour may stay within the setting where contact happened.

32. Superordinate goals and identities

In the final stage of the Robbers Cave study (Lesson 2), simply bringing the Eagles and Rattlers together for meals and films led to food fights. Hostility fell only when the researchers created superordinate goals: the camp's water supply "broke down" and both groups had to find the fault; they had to pool money to rent a film; a lorry carrying their food "got stuck" and all the boys pulled it together with a rope. By the end, most chose to travel home on the same bus.

Samuel Gaertner and John Dovidio's common ingroup identity model (2000) explains why: shared goals lead people to recategorise "us" and "them" as a single "we", and the in-group favouritism they feel for "us" now extends to former out-group members (Figure 1).

Must the old groups disappear? Not necessarily. When people value their subgroup, being told to give it up can feel threatening. Matthew Hornsey and Michael Hogg (2000) found that a dual identity, a shared superordinate identity that still respects subgroups (the right-hand side of Figure 1), can reduce bias more reliably than asking people to drop their subgroup.

43. Interdependence in practice

Morton Deutsch (1949) showed that groups whose members' goals were positively interdependent communicated better and were friendlier and more productive than groups set up so that members competed. Elliot Aronson's jigsaw classroom, first used in newly desegregated schools in Austin, Texas, in 1971, applies this: each pupil in a mixed group of five or six learns one part of the lesson and must teach it to the others, so everyone depends on everyone. Jigsaw classes showed more liking across groups and higher self-esteem than ordinary classes.