Children play in a makeshift refugee camp at the northern Greek border point of Idomeni, Greece. (Petros Giannakouris/AP)

“What types of asylum seekers are European voters willing to accept?” This is the question that Jens Hainmueller and two other political scientists, Kirk Bansak and Dominik Hangartner, set out to answer in their latest study, a survey of 18,000 voters in 15 European countries. As part of the study, conducted in February and March, they showed respondents profiles of asylum seekers and asked them which ones they would prefer their countries to accept; all told, voters evaluated 180,000 profiles. Their answers revealed a preference for refugees who are young, highly skilled and non-Muslim. (The paper is currently under review.)

In a recent phone interview, Hainmueller discussed the findings, their policy implications and why, given the scope of the crisis — Europe received approximately 1.3 million new asylum claims in 2015 —researchers must collect much, much more data.

The interview has been edited and condensed.

Nikita Lalwani and Sam Winter-Levy: In your previous research, you’ve found that people tend to care less about an immigrant’s country of origin than they do about how educated he or she is. Does your most recent study back this up?

Jens Hainmueller — We saw that voters cared about three main factors. First, they evaluated refugees for the potential economic benefit or burden they would bring to the country as a whole. Refugees who worked in highly skilled professions — such as doctors or teachers — were much preferred over refugees who worked in low-skilled occupations or who were unemployed before coming over. For similar reasons, we also saw a preference for younger refugees over older ones, who might impose more of a burden on social welfare programs in Europe. And you see a preference for people who speak the host country’s language.

The second thing you see is strong Islamophobia, which is consistent across all the countries. We gave voters profiles of refugees who were Christian, agnostic and Muslim. We saw a big gap in the responses: The agnostics were accepted at similar rates to the Christians, but there was a big anti-Muslim bias.

The third factor is humanitarian concern. We showed respondents profiles of refugees who had fled their home countries, having been tortured and traumatized, and compared them to those who came in search of economic opportunity. We varied the consistency of the refugees’ testimony in their asylum hearings to give a sense of the credibility of their case. We found that Europeans were much more likely to grant asylum to those with a more credible case, who really were fleeing persecution rather than simply looking for better economic opportunities.

NL and SWL:  What’s the relative strength of these different considerations? Does the anti-Muslim bias outweigh the other factors?

JH — These are pretty equally sized factors. In the survey, a doctor is about 14 percentage points more likely to be let in than someone described as unemployed. Someone fleeing persecution is about 15 percentage points more likely to be let in than someone seeking economic opportunity. And Muslims are 11 percent less likely to be let in than Christians. All three factors are important drivers. The anti-Muslim bias is about four to five times larger among those who would place themselves on the far right of the political spectrum than on the far left.

Interestingly, these preferences were consistent across all 15 countries we looked at despite the fact that the crisis has affected each of them differently.

NL and SWL:  Were you surprised by that?

JH — Yes, we hadn’t anticipated that. We had chosen a heterogeneous sample of countries, some of which have been more exposed to the crisis and more willing to take in refugees, such as Germany and Sweden. We were surprised to see that despite major differences in exposure, people’s preferences were similar. But it does make sense if you think about people evaluating what is best for their country as a whole and not just for them personally.

NL and SWL: What are the policy implications of that?

JH — It’s a mixed message. On the one hand, the strong anti-Muslim bias will make it very difficult to lower public backlash and help successfully integrate these refugees. But on the other hand, given that people have expressed a preference for refugees who will contribute to their country, there is a clear narrative that policymakers could use to convince voters to be more accepting. For a lot of Europe’s aging societies, welcoming young, risk-taking refugees could be a big opportunity. But convincing the public will be tough.

NL and SWL: Another strand of your research has shown that naturalization can help immigrants integrate better.

JH — That finding was based on work that we did in Switzerland [published in 2015]. There is a heated policy debate about naturalization in Europe. On one side, people argue that naturalization is a catalyst for integration. When you give people citizenship, the argument goes, you equip them with the resources and incentives necessary to invest in a future in the host country. On the other side, people argue that naturalization is not a catalyst but a crown. In this view, it is the prospect of becoming naturalized that makes people work harder to integrate.

In general, the evidence is limited on this question. Studies have tried to compare naturalized and non-naturalized immigrants, but it’s hard to learn much from that comparison because the people who select to become naturalized are probably different from the immigrants who select not to become naturalized. They might be better informed, they might have better resources, they might be more hooked into social networks, they might have been in the country for a longer time. Any of these might explain why they integrate better.

But Switzerland provided a unique opportunity. Because Swiss municipalities vote on individual naturalization applications, we were able to compare people who received 51 percent of the vote, and who therefore received Swiss citizenship, to those who got 49 percent of the vote and therefore did not. This allowed us to compare two very similar groups of immigrants and see the difference naturalization made. These referenda were held in the 1990s and early 2000s, and we recently went back to measure how well the immigrants had integrated. And what we found is that the results are more in line with the “naturalization as a catalyst” perspective. The people who barely got Swiss citizenship were much more politically and socially integrated than the ones who had just missed out. They participated in elections at the same rate as people who’d been born and raised in Switzerland, whose parents are Swiss.

NL and SWL: Is that because they feel they have a stake in the system?

JH — Yeah, exactly. They get the right to vote, so they can express their concerns and their grievances in the political process, and they can run for office. They can participate in local referenda to decide local issues. They become informed about politics and more engaged in their communities.

They also seemed much more socially integrated, in terms of their interaction with native Swiss, their membership in clubs, and the degree to which they now consume Swiss media rather than foreign media. Their long-term plans involve staying in Switzerland as opposed to returning to their original homes. Native Swiss people were also less likely to discriminate against them.

NL and SWL:  Does the Swiss case offer lessons for other countries?

JH — It’s too early to draw broad conclusions. What’s absurd is that we know so little about the effects of different integration policies. Different countries are trying different policies, especially in Europe, but we just don’t really know what the returns to these different policies are. We’re starting to evaluate some of them.

NL and SWL:  We know you’re working on a paper on how lengthy asylum processes can increase unemployment among refugees.

JH — Yes. It’s part of this bigger question about the extent to which countries are successfully integrating refugees or not. But there’s not much tracking of refugees over time, so it’s hard to know. In Switzerland, we were lucky enough to get the relevant data: We knew all the people who applied for asylum and what jobs they were doing every year afterwards.

If you go to Switzerland and submit an asylum claim after entering the country, you enter an asylum limbo, where you wait for two or three years in an asylum center that has been assigned to you, often in isolation from the native population, for a decision on your claim. We found that there are massive negative effects from waiting one additional year in limbo. It reduces your chance of finding a job subsequently by about 25 percent. And if you can reduce the wait time by 10 percent — by hiring one or two additional caseworkers, for example — you could save 12 million francs [$12.6 million] in a single year alone. There would also be less push-back from the native Swiss, and it would be better for the refugees.

That’s just one example of one policy where research is important. But there are many more. For example, in reception centers, is it better to keep people of the same ethnicity together, or to mix people from different countries? Should people be allocated to countries randomly, or should they be assigned to municipalities by some sort of matching system? How effective are language courses? These are questions all European countries are struggling with, and there is currently very little evidence. We’ve started finding answers in Switzerland, and we’re now collecting data in Germany, France, and Italy.

NL and SWL: Do the findings from Europe hold in the United States? Is there as little evidence here?

JH — There is not much data here either. If you want to evaluate the refugee resettlement process in the United States, all you have to go by are 90-day outcomes. Ninety days after they arrive, refugees must report whether or not they’ve secured a job. After that, they simply disappear.

The one thing the research has shown is that there is much less push-back if the immigrants integrate better, but we know much less about how to get better integration through policy. On many questions where we thought there would be answers, there just aren’t any. It’s amazing how little data there is compared to questions of education or welfare, which people have studied for a long time. When it comes to immigration, there are lots of ideological debates but very little evidence about what works and what doesn’t. That’s what we’re trying to change.

Nikita Lalwani is a staff editor at Foreign Affairs.

Sam Winter-Levy is an assistant editor at Foreign Affairs.