The Contact Hypothesis Offers Hope for the World

Demonstrators during the Selma to Montgomery March in 1965. Photo: Joseph Aaron Campbell/Getty Images

If you’re the sort of person who thinks about human nature, about who we are as a species, it’s easy to experience some whiplash these days. Here’s a story about Germans welcoming migrants at a train station in Munich. There’s a story about Swedes burning down housing for migrants. Here are Americans electing Donald Trump, who proposed a ban on all Muslims entering the country, as president. But there are Americans turning out by the thousands to protest at airports after he signed an executive order they viewed as anti-Muslim.

There’s no right “answer” to the question of whether human beings are fundamentally good or bad, of course. Humans are a lot of things, and social psychology has known for a while that our behavior is profoundly affected by context and culture. It’s safe to say the average human is capable on the one hand of profound goodness and on the other of profound evil — or at the very least of looking away when profound evil occurs.

But in the last few months I’ve found a bit of solace and much-needed solidity in a social-psychological idea that has been developed for the better part of the last century: the contact hypothesis. It’s the simple, inspiring idea that when members of different groups — even groups that historically dislike one another — interact in meaningful ways, trust and compassion bloom naturally as a result, and prejudice falls by the wayside.

The contact hypothesis, or contact theory as it’s sometimes known, is a really powerful, promising idea for a country like the United States — one that is big and diverse and whose national conversation on a host of subjects ranging from poverty to crime is veined through with implicit and explicit racism. At the moment, for example, a solid minority of Americans look upon Muslim people unfavorably — a fact that’s readily apparent in the sorts of crazy rumors about Muslim terrorists that often bounce around the internet. The contact hypothesis makes a simple claim: If you could get more non-Muslims to interact with Muslims, whether as neighbors or business partners or in a host of other contexts, that percentage would likely drop. And while this idea sounds idealistic, there’s solid evidence behind it — significantly more than there is behind other ideas, like corporate diversity trainings, for reducing prejudice that focus more on information and awareness than personal relationships.

The kernels of the contact hypothesis started bouncing around social psychology not long after the subfield first fully established itself, in the 1930s. Even early on, researchers took a fairly nuanced and cautious stance on the idea: Rarely did they argue you could simply mash together two different people from two different groups and expect a kumbaya moment as a result. Rather, certain conditions need to be met. A key work laying out some of the potential limitations of contact theory was The Reduction of Intergroup Tensions, a monograph by the Cornell University sociologist Robin Williams Jr. Williams laid out 102 propositions — empirically testable ones — concerning under what conditions contact might or might not work. For example, he hypothesized that “[p]ersonal contacts between members of different groups are generally most effective in producing friendly relations when the individuals are of the same, or nearly the same, economic and social status and share similar interests and tastes.”

Building off what was by then a couple decades worth of serious work on the subject, including Williams’s research, in 1954 Gordon Allport, one of the titans of 20th-century social psychology, offered up the clearest and simplest explanation of contact theory yet, published in book The Nature of Prejudice. In it, Allport trimmed down and adapted Williams’s 102 ideas to four “optimal conditions” under which contact might lead to improvements in intergroup relations: as summed up in a meta-analysis we’ll get to shortly, prejudice is most likely to be reduced when contact between groups entails “equal status between the groups in the situation; common goals; intergroup cooperation; and the support of authorities, law, or custom.”

It’s easy to understand, intuitively, why these conditions might matter. For example, if two members of different groups are interacting, but one has a huge amount of power over the other, it stands to reason that from a prejudice-reduction point of view the interaction is unlikely to be productive. The most obvious example here is American slavery — many American slaveowners interacted with their slaves constantly, and — not surprisingly — this didn’t tend to reduce their prejudice toward black people. To take a more contemporarily relevant example, think about a wealthy white Christian New Yorker who regularly patronizes bodegas run by Arab recent immigrants to the U.S. In the Allportian view, these interactions are unlikely to lead to reductions in prejudice, since the white customer and the Arab bodega clerks are approaching the interaction from significantly different places on the social hierarchy.

In addition to providing a useful common-sense framework for contact theory, Allport’s criteria provided useful guideposts for researchers hoping to test practical interventions geared at reducing prejudice — an area of great concern to researchers in the postwar years, which coincided with the start of the civil-rights movement. And in the decades since Allport laid out his conditions, researchers have done so in a myriad number of ways. The basic template of a contact-hypothesis experiment is something like this: First, experiment participants, all members of a majority group (white people, say), are split into two groups. Then the members of one of the two groups are exposed to some sort of pro-contact intervention — maybe they’re assigned to a cooperative task with black partners — while the members of the other complete no task or a placebo task. Finally, the participants are asked how they feel about the other group to see if completing the task had any measurable effect.

While there have been hundreds of studies dealing with all different sorts of different groups and flavors of interventions, according to Linda Tropp, a social psychologist and contact-theory expert at the University of Massachusetts Amherst, by the 2000s there was significant controversy within social psychology about whether contact “worked.” Tropp explained that, “A lot of the time, from a practical standpoint, people would say, ‘Oh, there are too many conditions — we could never construct contact situations to meet all of them, so why should we bother trying?’”

Around that time, Tropp teamed up with Thomas Pettigrew, a renowned social psychologist at the University of California Santa Cruz to conduct the largest ever meta-analysis — or study of studies — of the contact literature. The pair was able to collect 713 samples from 515 studies to work with. Their findings, published in the leading Journal of Personality and Social Psychology in 2006, were quite promising: A wide variety of contact interventions, conducted over the course of the decades, appeared to reduce prejudice, and to do so by respectable amounts. “There were a couple of ways in which we found really strong effects,” explained Tropp. For one thing, they found that the more rigorous studies yielded bigger effect sizes (meaning contact between groups appeared to be more effective) than less rigorous studies. This was a good sign for the solidity of the basic concept — all else being equal, it’s considered a red flag when less rigorous experiments deliver more impressive results. Second, while Tropp and Pettigrew found “stronger effects, on average” in studies where all four of Allport’s conditions were met, there were plenty of successful studies in which one or more of the conditions didn’t hold.

Tropp’s interpretation of this latter finding is that Allport’s four criteria aren’t fully independent, but rather “work in conjunction with one another.” So on paper, maybe a given experiment induces a sense of intergroup cooperation but not of common goals, but in reality it’s pretty hard to have the former without it promoting the latter. In Tropp’s view, social psychology may have been putting too much emphasis on the idea that all four conditions need to be met for contact to work, leading to unwarranted pessimism. “What our research is suggesting is every little bit helps, and the more we’re able to approach those optimal conditions … that takes us a good part of the way,” she explained. “So rather than worry about having everything on that checklist, let’s do what we can to move in this direction.” So in real-world settings, even just finding ways to put members of different groups on the same “team,” in support of the same goal, might make a difference — even if Allport’s other conditions are unattainable in that specific setting.

Before the meta-analysis, said Tropp, “There wasn’t real clarity about whether [contact] was a viable approach to trying to reduce prejudice or promote other positive outcomes in intergroup relations,” she said. But she said the paper elicited a big reaction from social psychologists. “I have been humbled by the number of people who have told me or Tom that we have helped to contribute to a renaissance of contact research,” said Tropp, adding that there’s been an uptick in work on the subject since her and Pettigrew’s work was published.

None of this means contact always works, or that one should adopt a cartoonishly rosy view toward it. For one thing, there’s a solid body of research suggesting that in certain instances, contact can have negative effects. Most common, perhaps, are instances in which members of a majority group come away from an interaction feeling better about the outgroup, but members of the minority group come away feeling worse. This jibes with intuition and anecdote: If you talk to any members of a minority group, they can tell you stories of majority-group members acting in condescending or just plain strange ways toward them, sometimes out of a well-intentioned but bumbling attempt to appear friendly or inclusive. These sorts of awkward interactions happen, and they can stymie the positive effects of contact.

There could also, in some cases, be unanticipated side effects even to mutually positive contact. Take an upcoming paper by John Dovidio, Angelika Love, Fabian M.H. Schellhaas, and Miles Hewstone, in which the researchers sum up the last two decades of contact research. They write that “an increasing number of studies indicate that positive intergroup contact may inadvertently reduce minority-group members’ willingness to participate in collective action to improve their position in society.” It might be the case that when minority group members have friendly day-to-day interactions with majority group members, “an emphasis on commonalities reduces identification with the disadvantaged group and diverts attention away from group differences and injustice.” That said, this isn’t inevitable — Tropp cited research by Julia Becker and Stephen Wright that suggests that in these interactions — this is Tropp’s paraphrasing — “If the majority group member expresses that they do not believe the inequalities between the group are legitimate, then you are less likely to see that effect.” In other words, “It might really depend on the nature of these members of the minority and majority groups.”

There are also some intriguing contact-theory findings which suggest that positive intergroup contact can cause majority-group members to be more likely to agitate on behalf of oppressed groups. Tropp is particularly excited about this line of research. “One of the reasons why I’m excited about this research is because it suggests that in addition to changing attitudes,” she said, is that “there’s also the potential for contact to change the future behavior of historically advantaged groups to benefit the disadvantaged.”

All this fits well with the underlying theoretical framework of both contact theory and many other areas of psychology, which suggest that personal relationships — and the strong emotions they can elicit — will almost always be a bigger drivers of belief and behavior than, say, factual information about how members of an outgroup are treated (perhaps presented in a diversity-training context). In summing up the original tenets of Allport’s theory, for example, Dovidio and his colleagues highlight the importance of “personal acquaintance between the members [of different groups], especially when personalization occurs with those who do not meet stereotypic expectations, and the development of intergroup friendships.” Maybe that’s the most important takeaway message from contact theory: When you have close, meaningful relationships with members of another group, it’s simply much harder to embrace negative stereotypes about members of that group, or to look the other way when their members are mistreated.

As I read about these experiments and spoke with Tropp, I kept thinking about the airport protests two weekends ago. Suffice it to say that many of the protesters were simply there because they thought it was the right thing to do, because they were motivated by politics or religion or their social networks or whatever else. But think about how much more potent that drive is when you know and value and worry about people who could be personally affected: Think about the difference between I am protesting this policy because it is wrong and I am protesting this policy because it is wrong and could hurt people I care about.

That’s the ultimate promise of the contact hypothesis: You don’t need fancy educating or lecturing or anything else to get people to treat one another better. To a certain extent, you just need to get them to interact on the same level, and progress will follow. That doesn’t mean it’s easy — but it does mean there’s good reason not to lose hope, even when the worst part of human nature rears its head.

The Contact Hypothesis Offers Hope for the World