In 2006, the University of Cape Town in South Africa randomly allocated incoming students into dormitories. Due to its history of race relations, many of these students initially held biases against other races. One study evaluated the effects of being assigned to a mixed-race room on stereotypes and academic performance. The results were overwhelmingly positive: White students who were paired with a Black roommate exhibited lower levels of prejudice, and there were more cross-racial friendships. On the academic side, grades of historically disadvantaged Black students improved when paired with a White roommate, and grades of their White roommates were similar to those of students in all-White rooms.
This study is a successful example of what scientists call the contact hypothesis. The idea is not new. In his 1954 book, The Nature of Prejudice, G. Allport coined the hypothesis by stating that personal contact between members of different groups can reduce prejudice. The hypothesis has been tested in hundreds of studies in the fields of psychology, sociology, and economics.
Until the early 2000s, however, the bulk of the literature used observational studies to examine the effects of contact. For instance, one would compare the levels of prejudice between people who have minority friends and people who do not. It is, however, clear that these studies would suffer from selection bias—that is, one chooses to make friends with minorities because one is less prejudiced. The clear causal link between meeting people from different groups and reduced prejudice was therefore not established.
A recent trend in the academic literature on the contact hypothesis has emerged, in which researchers use randomized experiments to induce contact between group members. For instance, one study randomly allocated Christians and Muslims to play in football leagues in Iraq. Another study randomly allocated new recruits in the US Naval Academy into squadrons with or without Black peers.
In a meta-analysis, I identified 37 research papers that experimentally test the effects of intergroup contact on prejudice and discrimination. The contexts of these papers range from inter-religious conflicts in Nigeria to prejudice against immigrants in the United States.
The results from the meta-analysis are clear: contact interventions work. While there exists a large variation in the effect sizes, on average, following a contact intervention, participants exhibit less prejudiced views and more inclusionary attitudes.
If intergroup contacts are so beneficial, should we implement them in any contexts? The answer to this question is more ambiguous for two main reasons. The first is that there exists large variation in the magnitude of the effects of contact. However, my analysis shows that no condition of the intervention successfully explains this variation, meaning that there is room to understand the causes of these differences to maximize the potential effects of contact.
The second and most important limitation of the evidence on intergroup contact interventions is that these interventions cannot be directly translated into policies. It’s hard to imagine that we would put every member of society into a shared dormitory with a migrant or enroll every individual in the military. The evidence on the effectiveness of intergroup contacts in everyday situations is much scarcer, and the results from existing studies are not as promising. For instance, sitting next to minority Roma peers in school does not seem to reduce anti-minority discrimination among native Hungarian children.
Turning the findings of research on the contact hypothesis into actual policy that improves social cohesion is certainly a difficult task, and researchers and practitioners must put a lot of effort in to transforming these experiments into policy. Few endeavors are as simple as college dorm assignments, after all. Still, the findings are an encouraging suggestion that such efforts have a great potential to bear fruits.
The famous sociologist Robert Putnam said that “the central challenge for modern, diversifying societies is to create a new, broader sense of ‘we.’” In essence, my findings reflect on this statement, putting forward the idea that creating links between communities in ever-more diverse societies can help overcome the upfront costs and reap all the benefits of diversity. I hope my research can help make societies more inclusive and, because of diversity, more prosperous.