View from
The Center

Ethnic Diversity and Social Trust: a Discussion with Merlin Schaeffer

Policies that try to tackle diversity problems make it always salient in people’s minds. And that can backfire, even if well-meant.”

Ethnic Diversity. The words will garner glee for some and dread for others. Whether ethnic diversity is already high within a country or in flux due to immigration, a top concern is how a nation addresses not the benefits but a known problems that ethnic diversity can bring, as well as whether less diversity is better. 

The logic of debate has balanced uneasily across the weighted shoulders of one prominent researcher. American political scientist Robert D. Putnam first published Bowling Alone: America’s Declining Social Capital in 1995, in which he concluded that more diversity in a community is associated with less trust. As Putnam said, in areas with high ethnic diversity, people tend to engage less in civic life, preferring to “hunker down.” Putnam’s research did not bode well for those aspiring to see ethnic diversity increase, especially through immigration. 

The social science fields exploded with further research after Putnam’s formative work. Yet, over the following 25 years, as numerous as the studies were, they seemed inconclusive on whether Putnam was right or not. 

This year, all that changed. Three researchers, Peter Thisted Dinesen, Merlin Schaeffer, and Kim Mannemar Sønderskov published a study that analyzed the vast pool of existing data and found the evidence was pointing toward an undeniable conclusion: Ethnic diversity does erode social trust. 

Putnam was right.

Was it the smoking gun that the anti-diversity crowd had been looking for? 

It was time to find out. Late in the afternoon from Denmark—where he teaches at the University of Copenhagen—one of the three researchers met me online for a virtual chat while I sipped a morning coffee on the west coast of Canada. We quickly realized a shared connection to his city of birth in Germany. This, of course, was irrelevant to the topic of social trust and ethnic diversity so the calm mannered professor, Dr. Merlin Schaeffer, and I, soon moved on to the task at hand.

Ethnic diversity really does erode social trust. That was the finding of your paper, “Ethnic Diversity and Social Trust: A Narrative and Meta-Analytical Review” published this year and co-authored with Peter Thisted Dinesen and Kim Mannemar Sønderskov. To date, it is arguably the most conclusive research paper on this topic. Were you surprised by the findings?

I wasn’t surprised at all. I’ve been working on the topic now for ten years. If you read the single studies, many are inconclusive. But the literature has a pattern. Single studies tend to report statistically inconclusive, but slightly negative results. If truly inconclusive, then [the findings] should be all over the place, not negative over and over again. So, we had this feeling that the meta-analysis—which takes all the evidence together—would show a systematic overall negative association. So no, I wasn’t surprised.

I guess one finding that mattered to me was Putnam’s most original finding. Not only is generalized trust lower in ethnically diverse settings, but also trust in the in-group. In the U.S., white people who live in diverse places trust white people less. Black people who live in diverse places trust black people less. In Europe, people with native born parents who live in diverse places trust other natives less. It’s not only a group competition thing. We re-established that finding in our meta study.

What got you interested in this line of research? Was there a trigger moment where you thought, “I really like this research topic”?

Um, no. It was first—you could say—opportunistic, and there was even a bit of refusal on my part. I wanted to do a PhD because I liked being an academic. I applied for all kinds of positions, and Ruud Koopmans had this PhD position. I got the offer. It was definitely the best offer I got because he’s a great scholar and WZB [Berlin Social Science Center] is an extremely good institution. Then I started to work on it, and, actually, I didn’t like the topic so much because I didn’t want to be the white, blue-eyed, blonde guy who studies why diversity is bad for society. 

But sociology is empirical research. The findings about ethnic diversity and lower levels of social cohesion were re-occurring. So eventually I learned to see this topic as a puzzle that needs to be solved and not as the agenda of a xenophobe. We need to understand what’s going on to know whether it’s a truly serious or a minor social problem our societies are facing. And if it is a problem, we need to understand it in order to find a solution. How do you say it in English—brushing it under the carpet won’t be the solution

So, at the beginning, I didn’t actually like the topic too much. But the results were there, and, as an empirical researcher, I take them seriously. I’ve met great scholars who work on this, and it’s been inspiring to work with them. 

Erosion of social trust happens mostly at neighborhood boundaries between otherwise relatively homogenous neighborhoods. Was that essentially the finding from your 2016 paper “Contested Boundaries”?

Let me contextualize a little bit. 

Overall, in this whole debate, you have two types of explanations. 

These declines in social trust could be driven by animosity: by conflicts between groups. That would be problematic because if that’s the main driver, then you could have deepening societal cleavages. You could have polarization and segregation. These can turn into really serious problems for society.

The second types of mechanisms that could drive this are processes we see with modernization. When communities urbanize, when residential instability increases—you know, you don’t stay your whole life where you grew up—then we see a decline in community, decline in social control, decline in communication, more anonymity among the residents. Declining community could be an explanation of why ethnic diversity is associated with lower trust. [This] would be less problematic because our societies are able to deal with these things. Although urbanization has meant a decline in the classic tight-knit community, society hasn’t fallen apart, and there are also benefits. It’s in cities where you see the most creativity, where you see innovation and patents being developed, where a lot of economic growth comes from. And that parallels other research: Ethnically diverse workgroups produce more patents, research papers, and so on. So, if those types of mechanisms erode social trust, it would not be as problematic. We are able to deal with it. 

As it turns out, it can be either. 

The “Contested Boundary” paper was one attempt to show—under conditions where you have strong residential segregation in a city between groups, and where these groups meet at neighborhood boundaries—that is where animosity and conflict drive down trust. People cannot solve their local neighborhood problems together anymore. They call the authorities because they don’t talk to each other. If you have such stark segregation where groups build place-based identities—This is our neighbourhood. This is a white neighbourhood. This is a black neighbourhood—at the boundaries where they meet, you have conflict. Is this still our neighborhood or is it already yours?

If diversity goes with residential segregation, it will become problematic.

In the paper, “Relational diversity and neighborhood cohesion” you proposed that it is not the size of the minority groups that mattered, but variety. More variety equates to having less in common, which could lead to dysfunctional governance and breakdown in social trust. Is that the idea?


In a typical European or German setting, we don’t have the kind of segregation you see in the United States. In the European setting, people are more spatially integrated.

So, the question is, “Okay, what are the mechanisms here that lead to erosion of trust?”

In our setting, communication problems could be the link. But it could also be cultural misunderstandings like, “Hey, when will we solve this problem with the speeding cars on our street?”

“Yes, soon.” 

But you have different ideas of what “soon” means. Then, you get the impression the other person is unreliable. Things like that. Or you could have different preferences. We want to tackle the speeding cars. The others want more playgrounds for children. 

Is it these things that drive down trust? 

Overall, these things are associated with diversity, not polarization. 

Think about language. If you have two groups speaking a different language, it is easy to bridge the language boundaries. You basically need a couple translators that understand both sides. But if you have a hundred groups that all have a different language, then you need way more translators. There is way more potential for misunderstanding. So, if it is these processes—rather than animosity—we should see that many, many groups should drive down trust, and that’s exactly what we find.  

These problems of diversity are good to understand, but, as a sociologist, I wouldn’t say that’s super worrying. What’s more worrying is the animosity of segregation. 

With regard to segregation, is it better to have only immigrants from one origin instead of many?

I would say, no. I don’t think that is necessarily better. On the one hand, the benefits that come with diversity–creativity, patents—you would not have to the same extent. And, at the same time, you have more potential for the kind of segregation and animosity that you see in the U.S.

Instead, communication problems and the like will be the bigger challenge with more diverse groups. 

Also, people in cities trust less. People who are not homeowners trust less. These things have increased over the past, but, nevertheless, social trust is actually pretty stable. Well, in some countries. Not in the U.S.  

But Denmark, for instance, has seen an increase in all of these things: ethnic diversity, immigration, residential instability, urbanization, but trust has gone up because we have other means to tackle these things. 

What we are not good with is tackling polarization. That’s what we see in the U.S. Such a polarized context now. Many people study it, and yet there’s so little insight on how to overcome polarization. 

Like Northern Ireland where you have two polarized groups?

That’s exactly the danger. 

That doesn’t necessarily mean their communities have low levels of cohesion and trust. They can actually have quite high trust and also high in-group trust, particularly in Northern Ireland. The trust in in-group is probably pretty high, in comparison to super diverse places. But polarization is the more problematic situation to have, rather than complex diversity.

As immigrants become better integrated over generations, one would assume there should be fewer grievances, but you found it’s actually the opposite—a.k.a. The Integration Paradox. Your research found, for example, that perceived discrimination is more common among second generation rather than first generation immigrants, especially when the first-generation parents are educated but the second-generation children have less education. Are these second-generation individuals using the crutch of discrimination instead of admitting to their own personal shortcomings? Or is that too harsh?

The problem with this debate about the integration paradox and about findings like mine is—so far—it is impossible to know. 

It is possible that immigrants who are less successful than they had aspired for use discrimination as an excuse for their personal failure. That way, you can keep your personal self-esteem. It is not me that failed; I was discriminated against. But it’s inconclusive because it could also be that they actually were discriminated against and—for that reason—are less successful.

That’s the reason why I am starting a new research project. The problem with current research is that we send out fake applications to jobs to find out about true discrimination. But, of course, fake applications have no personal feelings and perceptions of having experienced discrimination. What is missing is research data on actual discrimination and perceived discrimination, together. No one has that so far. It’s a bit complex, but I’m starting experiments that will hopefully give us that data.

So, it could be what you say, but that is one big question mark.

It’s a timely topic.

You also see it in other regards. Women complain more about gender inequality (and perceive more gender inequality) in countries where there is actually more gender equality. 

Some people call it Tocqueville’s Paradox. It is not only related to immigration.

He said that if you have improvement in societal conditions, then remaining inequalities that people beforehand might not have cared about become important. For instance, in countries with more equality, women report more sexual partner violence. The question is: Do women in more egalitarian countries experience more partner violence or do women in less egalitarian countries under-report?

Going back, are there definitive actions that governments could take to reduce the harmful effects of ethnic diversity on social trust? 

Social trust is not only an outcome of diversity. 

There are other things that affect it too. One of the most important is, for instance, corruption, transparent government, things like that. That’s probably what drives up the incredible Danish increase in social trust—Denmark now being one of the least corrupt countries in the world. 

Because these things are even stronger predictors of trust, they will reap more benefits. 

But, also, we found in our research that the constant emphasis on diversity also lowers trust. Policies that try to tackle diversity problems make it always salient in people’s minds. And that can backfire, even if well-meant. That’s also true for multiculturalism. As well-intended as it might be, it emphasizes difference all the time. That can be problematic. 

Instead of diversity politics, definitely consider segregation. That is really important. 

Associated to that, home ownership goes along with ethnic segregation because natives tend to have more money and will move out into more homogenous places. The immigrants will stay in the social housing and rental housing market, and so you have a stronger divide. A smart housing policy would, on the one hand, help immigrants become homeowners. Homeownership goes along with more engagement in the place where you live. It ties you to the community. It makes you have an interest in your neighbors. So homeownership is a good thing, but we need to make sure the minorities can also buy homes—not in separate places but in a spatially integrated way. 

Danish politics is struggling with that a lot. If I had to say where is the strongest angle, I would say, “Think about more even dispersal.”

Here in Canada, Koreans immigrants are very good at geographic dispersal whereas Sikhs tend to live among themselves. With a government policy of dispersal, you might have resistance from some ethnic groups.

Preference for immigrants to live among co-ethnics, relatives, and kin is stronger for some groups than it is for others. There’s no question. Of course, you don’t have to follow the Singaporean example where every single house needs to reflect the overall population shares of different groups, so that there is no segregation whatsoever. That’s not necessary. 

But if you reach levels of segregation like you often have in the U.S.—think about a city like Chicago—it really becomes long-term problematic. The German word for real estate is “Immobilie” [literally immobile or fixed in place]. Once it has materialized, it is very difficult to change. It’s cement. It’s immobile. That’s why housing policy is [extremely] important.

Segregation can happen very quickly, as you’ve noted. Are there certain actions that can prevent rapid segregation? 

Well, there are things that make it happen more [quickly]. You see it here in Denmark but also in the U.S. If you have a public housing policy that builds big public housing blocks, then you have it more [quickly]. 

In Germany, we even had an explicit housing policy that tried to buy single flats within houses on the private market, whereas in the U.S. you sometimes have these whole housing blocks that are public housing. Then you have stronger segregation. 

Social housing that is dispersed is definitely number one.

Now that you, as a German, are living in Denmark, have you learned to speak Danish?

I need to step up my game, particularly as a scholar of immigrant integration. I can order my coffee and things like that and [have] small chats. 

Do you find it ironic that you are now an immigrant yourself?

I know as a scholar that language is extremely important. I am now—in a way—an immigrant, and now I know why it’s not so easy to learn a language. You can always have private excuses why you will start tomorrow. 

And there are things that you don’t so easily see if you always live in your own country; [there are] difficulties that even I, a very privileged immigrant and EU citizen, face. There are things that are more difficult for me than for the Danes. That’s a good experience to have. It puts things into perspective. What would be good in terms of integrating: learn Danish fast.

I look forward to reading your future research. It has been a pleasure talking with you today. Vi ses.

Yes, vi ses.

Leave a Reply

Your email address will not be published. Required fields are marked *

This site uses Akismet to reduce spam. Learn how your comment data is processed.