Uncovering the roots of discrimination toward immigrants

All over the world, immigration has become a source of social and political conflict. But what are the roots of antipathy toward immigrants, and how might conflict between immigrant and native populations be dampened?

Political scientist Nicholas Sambanis has spent his career considering various forms of inter-group conflict, starting with civil wars and international interventions to help countries transition from war to peace. More recently, he has studied non-violent forms of conflict, including discrimination between individuals of different ethnic or religious background. Understanding why groups come into conflict and how to resolve those conflicts and avoid violent escalations is the mission of the Penn Identity and Conflict (PIC) Lab, which he founded when he arrived at Penn in 2016.

His newest research on identity politics, an experimental approach that explores the causes of discrimination against Muslim immigrants in Germany, was just published in the Proceedings of the National Academy of Sciences.

"Opposition toward immigration can be due to economic reasons because of competition for jobs or due to the perceived cultural threat that immigrants pose to their host country by challenging dominant norms and changing the national identity," he says.

He finds arguments centered on cultural threat more convincing than economic explanations of opposition to immigration, especially in Europe.

"Most previous research is limited to presenting survey-based attitudinal measures of antipathy toward immigrants or refugees and correlating them with socio-economic characteristics of the survey respondents or their political beliefs," Sambanis says. "We wanted to go beyond that and measure actual behavior in the field. We wanted to figure out what particular aspects of refugees or immigrants generate more hostility. Is it racial differences? Ethnic differences? Is it linguistic or religious differences? Is there merit to the idea that discrimination toward immigrants is due to the perception that they do not follow the rules and threaten dominant social norms?"

There's very little experimental research, Sambanis says, on the causes of anti-immigrant bias and even less research on how to reduce it.

Working with University of Pittsburgh assistant professor Donghyun Danny Choi, a former PIC Lab postdoc, and Mathias Poertner, a PIC Lab fellow and postdoc at the University of California, Berkeley, Sambanis designed the experimental study. They targeted Germany because of the high influx of immigrants and refugees and the political salience of immigration issues in recent elections there and because Germans are strongly inclined toward conforming with social norms, especially around keeping order.

Their hypothesis: If it is true that opposition to immigration is driven by the perception that immigrants threaten valued social norms and pose a cultural threat, then in a country that values norm adherence they would see a reduction in discrimination toward immigrants if immigrants show that they respect local social norms and care about their new society.

They staged an intervention against a native male German who littered in a public space, since not littering is a social norm there. A female researcher would approach the person littering, asking him to pick up his trash and dispose of it properly. Bystanders, unaware that they were being studied, observed the interaction. Shortly thereafter, the woman would take a call and while speaking on the phone would drop a bag of groceries, causing oranges to spill out on the floor. The observing researchers recorded whether the bystanders who had witnessed this entire interaction helped the woman pick up her oranges.

In some versions, the woman dropping the oranges would have sanctioned the norm violator, signaling her integration with the German culture. In others, she did not intervene so as to seem indifferent to the littering.

Researchers also used the woman's identity as a variable: In some versions, she was a native German, in others a Muslim immigrant wearing a hijab. Her degree of religiosity, her ethnic background, and her linguistic assimilation to German society were all manipulated as part of the experiment.

This allowed the researchers to measure whether immigrants who are more socially distant than the average German receive less assistance and whether following social norms offsets any bias toward them.

They ran this experiment more than 1,600 times in train stations in 30 cities in both western and eastern Germany using multiple teams of research assistants, with more than 7,000 bystanders unwittingly participating. Then, the researchers measured whether women who wore a hijab received less assistance than native Germans, whether ethno-racial differences between immigrants matters less than religious differences in generating bias, whether immigrants who wore a cross received more help than those who did not wear any outward symbols of religiosity, and whether good citizenship—enforcing anti-littering norms—generated more help from bystanders, eliminating any bias against immigrants.

"We found that bias toward Muslims is too pronounced and is not overcome by good citizenship; immigrant women who wore a hijab always received less assistance relative to German women, even when they followed the rules," Sambanis says.

"But we also found that good citizenship has some benefit, as the degree of discrimination toward Muslims goes down if they signal that they care about the host society. And ethnic or racial differences alone do not cause discrimination in our setup. Nor is religious assimilation—wearing a cross rather than a hijab—necessary to be treated with civility."

On average, women wearing a hijab who did not enforce the norm got help in about 60% of cases, whereas "German" women who did scold the litterer got help in 84% of the cases. The rates of assistance offered to a Muslim who enforced social norms by scolding the litterer were equivalent to those for a German who did not enforce the norm.

"The reason to run such an experiment focusing on everyday interactions is that it gives you a sense of the accumulated impact of discrimination in shaping perceptions of identity and belonging," Sambanis says. "Getting help to pick up something you drop on the floor seems like a small thing. But these small things—and small slights—add up to form lasting impressions of how others perceive you and, in turn, can inform the immigrants' own attitudes and behavior toward the host society."

Now, Sambanis, Choi, and Poertner are extending their research to new questions trying to understand the mechanisms underlying the effects they picked up with their experiments in Germany.

They found gender was a key factor, as it was German women who discriminated against Muslim women. Sambanis says he didn't expect this result since existing research implies that men are more likely to discriminate, and certainly media portrayals of anti-immigrant backlash tend to center on men.

"We puzzled over the fact that German women withheld assistance from Muslim women who needed help. Based on survey data we collected after our experiment, it seemed that this effect was particularly due to secular women, women who do not register a religious preference," he says. "This led us to hypothesize that part of the reason we observed this behavior is that German women who might otherwise be open to immigration have developed hostile attitudes toward Muslims because they perceive their cultural practices as threatening to hard-won advances in women's rights. It's basically a feminist opposition to political Islam."

The team has now designed a new experiment that explicitly tests this hypothesis. Two new experiments test whether signaling one's political ideology regarding key issues related to women's rights can offset discrimination toward Muslim women.

This collaborative effort between Sambanis, Choi, and Poertner will become a book on how conflict between immigrants and native populations can be managed and whether norms can form the basis for the reduction in discrimination. The German experiments will be expanded next year and applied to a different social context in Greece, which also faces an intense political crisis due to unsustainably high levels of immigration and which differs from Germany with respect to the degree of public adherence to laws and rules.

Individuals there are less likely to follow rules and contribute less to the public good. So Sambanis and his co-authors think they may observe even lower effects of the ability of social norms to offset discrimination due to ethno-religious differences. That research will provide a useful comparison to better understand the existing experimental results.

"A key idea in socio-biological theories of inter-group conflict is that there is an almost innate antipathy or suspicion toward members of "out groups" [immigrant], however those groups are defined. But clearly societies can manage sources of tension and avoid conflict escalation since there is very little observed conflict relative to how many different types of inter-group differences exist out there," Sambanis says. "A lot of the literature on immigration has suggested that assimilation is the key to reducing conflict between natives and immigrants: Immigrants must shed their names, change their religion, or hide their customs so they can be more accepted.

"Is this really necessary? Or is it enough for immigrants to just signal credibly that they care about being good citizens as much as everybody else?"

Understanding these types of questions is at the heart of the PIC Lab's mission. A unifying theme of Sambanis' work has been reducing inter-group conflict, particularly inter-ethnic conflict.

His interests were shaped by the wars in Bosnia and Rwanda, which were going on when he was in graduate school and pushed him away from international economics and toward studying peacekeeping. At the PIC Lab, researchers tackle questions both at the larger country level and at the smaller individual and group level, integrating ideas from political science, social psychology, and behavioral economics to understand human behavior and explore the outcomes of different policy interventions to reduce conflict. The lab conducts data-based, mostly quantitative research that can inform policy design but also theory-building in political science, Sambanis says.

"Ethnic differences, religious differences, racial differences—they all matter for politics, but they do not need to produce conflict," he says. "When people are faced with the hard realities of ethnic wars, separatist conflicts, genocides, or hate crimes, they usually assume that these are inevitable outcomes of innate human prejudices or fears and that people just can't get along because of deep differences in their preferences or their customs.

"A lot of the work that I do shows that ethnic conflict is not inevitable. The key is to understand the conditions that make ethnic differences salient and then find ways to defuse or manage conflict."