Working with migrants: does it increase or decrease support for anti-immigration political parties?

Download this episode

In Episode 6 of Series 3 of the DIAL Podcast, Sirus Dehdari from the Swedish Institute for Social Research at Stockholm University and a member of DIAL’s Populism, Inequality and Institutions (PII) project, talks about his research looking at whether support for anti-immigration political parties increases or decreases when native-born voters work alongside migrants.

Workplace Contact and Support for Anti-Immigration Parties is a DIAL Working Paper by Henrik Andersson and Sirus H. Dehdari 


Christine Garrington  0:00 

Welcome to DIAL, a podcast where we tune into evidence on inequality over the life course. In this series we discus emerging findings from DIAL research. My guest today is Sirus Dehdari from the Swedish Institute for Social Research at Stockholm University, who was part of the DIAL’s Populism, Inequality and Institutions project, has been looking at whether support for anti-immigration, political parties, increases or decreases when native born voters work alongside migrants. I started by asking him about the social and political backdrop to the research.

Sirus Dehdari  0:31  

I’m sure both you and your listeners know that we’ve seen a rather large increase in immigration to most European countries in the last couple of decades. In some EU countries the share of foreign born is around the close to 10% and is in a few cases is approaching 50%. At the same time, we’ve seen that several countries, several of these countries, are also seeing how anti-immigration parties are increasing their support. So, one might be tempted to conclude that the former is actually causing the latter, meaning that increasing immigration is causing opposition to immigration. And there are several, several studies in political science, sociology and some in economics that consider within countries, regional variation in immigration and voting for these types of policies, and they find a positive correlation or in some cases, a positive causal relationship between these two phenomena. But there are also studies that find either no correlation, or even in some cases, a negative relationship between the presence of immigrants or the visibility of immigrants and support for anti-immigration policies. It seems that there are some local and perhaps regional conditions that influences how immigration is affecting opposition to immigration. For instance, the skill level of the newly arrived immigrants compared to the skill level of the natives, or the socio-economic status of the natives in the neighbourhoods where immigrants are placed, if they are placed, or where they choose to settle. And another important factor that been seen in the literature is the type of contact that occurs when navies and immigrants interact, which is simplified a bit by separating these types of contact into two categories: basic superficial and meaningful contact, where the former is basically just seeing immigrants at the local supermarket or on your daily commute to work without really interacting with them or getting to know them, while meaningful contact would involve getting to know each other or cooperating towards a common goal. Superficial contact might actually increase opposition to immigration, but we believe that meaningful and cooperative interactions or meaningful and cooperative integral contact between native and immigrants, we believe that they’re expected to reduce opposition. So this is basically the political and social setting that got us to start thinking about our project.

Christine Garrington  3:12  

Yeah, that’s great. And so what then specifically was it that you wanted to look at and why? 

Sirus Dehdari  3:17  

What we set out to do is to look and investigate how contact between natives and immigrants in the workplace influences opposition to immigration. And why is this important? Well, first of all the workplace is a very important social setting for most working adults, where we spend a large amount of our time. And many workplaces are characterised by teamwork, where we would expect this type of meaningful and cooperative contact that I mentioned earlier. And second, immigrant co-workers might actually signal to natives that immigrants are indeed here to lower their wages to take away their high wages. So, we have two potential effects going in opposite directions. And this is why we have found that this particular setting, the workplace, to be very interesting, interesting in terms of its effect on the opposition to immigration. 

Christine Garrington  4:10  

So what do you actually do with this particular piece of research and where did you get your information from?

Sirus Dehdari  4:16  

So we use Swedish registered data, where we are able to connect almost all of working adults to a specific workplace, so we know which firm they work at, but we also know which actual workplace that work at, so their physical address. For each native worker, we compute the share of non-European co-workers and this is basically a measure of the type of, or the amount of meaningful and cooperative contact that each native has. The government that people voted for, but we are able to aggregate these individual level shares, workplace contact shares, to the smallest political geographical unit, which is the election precinct. So, we aggregate this to the precincts and we create a share of non-European co-workers among the average native born worker living in that precinct. At the precinct level, we also have election results, we know the share of votes for Social Democrats or the Green Party and so on. And which means that we also know the share of votes for the Sweden Democrats, which is the Swedish version of an anti-immigration and xenophobic type of party. A typical party that exists in many other European countries. We match this to the precinct level and then we conduct the analysis on this geographical unit. 

Christine Garrington  5:37 

I see, and what did what did you find when you did that matching when you match those things together?

Sirus Dehdari  5:41 

We find that an increase in the share of non-European co-workers decreases support for the Sweden Democrats. What does this mean? Well, it means that non-European co-workers might signal or might highlight the threat of increased labour market competition amongst the natives working at that particular workplace. This is offset by the potential benefits stemming from, from meaningful and cooperative workplace contact. So at the end, the net effect of additional immigrant co-worker is lower support for anti-immigration parties. 

Christine Garrington  6:15 

So Sirus, you went on and dug a little deeper into the level of skills of the non-natives. Can you tell us a bit about why you did that, and what you found there?

Sirus Dehdari  6:25  

If you consider the two types of contact that I mentioned earlier, and also to consider what I mentioned about the immigrants in the workplace highlighting the threat of labour market competition and natives’ access to employment and high wages, we expect immigrants of the same skill to highlight this treat even more. So consider a native working in a workplace and then his or her new manager is a foreign-born immigrant or non-European. If the native is for instance a carpenter or a plumber, she might not feel threatened, in the same extent, as if the new co-worker was actually a plumber, or a carpenter or performing a skill, performing tasks that is similar to a task the native is performing. For this reason, we might actually expect that the threat of labour market competition only highlighted or id more highlighted when immigrants of the same skill appear in the workplace. But at the same time, we will also expect natives and immigrants of the same skill to be involved in even more cooperative tasks. So, we might not expect the plumber and the manager to share tasks, even if they share the same goal, that would be the profit of the company, we might not expect them to work towards that goal in the same manner. So even in this case of the same skill contact, we might expect the threat to be highlighted even more, but also the contact to be even more  meaningful and more cooperative. The net effect is still ambiguous. So what we do is that we separate between having co-workers of the same skill  and having co-workers of the opposite skill and then we estimate the effect of these, these average shares, again on support for the Sweden Democrats and what we find is that the results, the negative results that we saw in the main results, or the baseline results are solely driven within skill contact. So the estimate is still negative for within skill contact but for opposite skill contact, we find basically nothing. And what does this mean? Well, it means that even in this case of the presence of within-skilled non-Europeans, where we expect labour market competition to be more highlighted, the benefits of meaningful workplace contact is still offsetting this potential negative impact of opposition to immigration. And at the same time, we don’t find anything for our cross-skilled contact, which means that we shouldn’t expect, for instance, low educated workers and high educated workers to actually interact, regardless of the type of interaction is superficial or if it’s meaningful. And at the same time, we shouldn’t expect natives of a particular skill to feel threatened when immigrants of the opposite skill are present at the workplace.

Christine Garrington  9:24  

Yeah, that’s really interesting. So I wonder if I can ask you whether it mattered if the job was say under threat from something like computerization or artificial intelligence or something where you know those sorts of developments might be taking place within a company setting?

Sirus Dehdari  9:41  

So what I’ve said so far is basically that labour market threat is highlighted when immigrants become visible in the workplace. However, and as I just mentioned, comparing within skill and across skill, we said that we might not expect natives of a particular skill to be threatened when immigrants of the opposite skill appear in the workplace because they don’t compete. Another way to think about this is that natives that do not feel that their employment is actually under threat, they will not feel even more threatened when immigrants become visible in the workplace. So basically, what we wanted to do is that we wanted to separate occupations into secure and non-secure, or vulnerable and non-vulnerable occupations. And to do so we use a routine task index score, which determines the probability or the threat of a particular occupations tasks to be replaced by computers. More or less in occupations with a high routine task index score the workers there are under the threat of being replaced and what we do is similar to what we do within the cross-skilled contact which is to separate, again we separate occupations into vulnerable and non-vulnerable occupations and then we re-estimate the model again. And what we find is that the negative effect that we found from the contact between natives and immigrants in the workplace only appears when natives are occupied in non-vulnerable occupations. So when they are basically less threatened by automation and computerization. On the other hand, if we consider natives employed in vulnerable sectors, we actually find a positive relationship, so when immigrants are visible in in these workplaces, natives are actually more likely to oppose immigration.

Christine Garrington  11:35

So Sirus, just finally then I wonder what you feel we learn from all of this? There’s quite a lot of important insights I think that are coming out of this work. But what would you say are   the key takeaways?

Sirus Dehdari  11:47  

The main takeaway is that similar to other studies that have found that cooperative and meaningful contact, integral contact between members of different groups, for instance ethnic groups or religious groups, is that the contact really matters, the type of contact matters. If we are interested in reducing opposition to immigration for instance, we have to facilitate these types of contact, for instance through the workplace but also through the different social settings, for instance different sporting events. So the results I believe are not specific to the workplace but can also be extended to other important social settings. 

Christine Garrington  12:25

Workplace Contact and Support for Anti-Immigration Parties is a DIAL Working Paper by Henrik Andersson and Sirus Dehdari. You can find the paper and more about the wider DIAL research programme at and don’t forget to subscribe to the DIAL podcast, to access earlier and forthcoming episodes.