When Austria opened its labour market to workers from the new EU Member States, it was followed by a substantial inflow of workers from Central and Eastern Europe, including Slovakia. Did this mean fewer job opportunities for native Austrians? This is the question explored by Katarína Valková from the Faculty of Economics and Finance in her article Employment Effects of Immigration in Austria, published in the journal Empirica. Using detailed Austrian administrative data, she shows that immigration did not reduce the employment of native workers overall, although its effects varied considerably between men and women, across education groups, and between natives and earlier immigrants.
We spoke with Katarína about who benefited from the new wave of immigration and who faced greater competition, how the impact of immigration on jobs can be measured in the first place, and why this research also has a very personal dimension for her.
Let us start with the basics. What did you want to find out with this research, and why is Austria such an interesting setting for studying the impact of immigration on the labour market?
The aim was to find out whether Austrians should be concerned about their jobs as a result of a new wave of immigration. Austria is a textbook example of a destination country. I myself have made it my home. The research was conducted for the Austrian Public Employment Service (AMS) while I was still working at the Institute for Advanced Studies (IHS) in Vienna. AMS wanted to know how active labour market policies could be better targeted and whether some groups were more vulnerable than others. It is a great example of a public institution being genuinely interested in evidence-based policy.
One of the most common concerns surrounding immigration is that newcomers “take jobs” from people already living in the country. What do your results show? Did this happen in Austria following the substantial inflow of workers from the new EU Member States?
This is an entirely legitimate concern, one that is also related to higher public expenditure on unemployment and social benefits, and it needs to be taken seriously. For individuals, it can represent a real threat. My study examined the inflow resulting from Austria opening its labour market to the new EU Member States. However, I found no substantial negative effects. On the contrary, immigration even had mildly positive effects: a larger number of employees from the new EU Member States was associated with a slight increase in native employment.
Here it is important to understand what policymakers and economists actually mean by a displacement effect. Displacement does not occur if Austrians respond to immigration by moving to another, more favourable location and, for example, taking better-paid jobs because less attractive positions have been taken by immigrants. Migration often acts as a kind of buffer in the labour market, absorbing such shocks. In most cases, this is beneficial for native workers, although the results depend on a country’s institutions and policies.
My study did not examine mobility across sectors or occupations. It focused only on whether, overall, more or fewer people who had already been employed in the country remained in employment, and whether this was related to immigration. The answer is: more. International research generally reaches similar conclusions. David Card, a recent Nobel Prize laureate, has conducted important research on precisely this issue.
The average result, however, hides substantial differences. Men with intermediate vocational education experienced somewhat negative effects in the short run, while women’s employment increased. People who had immigrated to Austria earlier also faced stronger competition. How do you explain these differences?
Yes, and this is precisely what makes the paper important for policymakers. Whether immigration has positive or negative effects on the labour market depends on whether native and foreign workers are substitutes or complements — in other words, whether they compete directly because they have similar education and skills, or whether they complement one another because their skills differ. That is why my analysis explicitly takes education into account.
Women are one example. Research shows that women perform more unpaid work, including household work, caring for family members, and childcare. If some of this work is taken over by immigrant workers — for example, carers from Slovakia — and they help fill existing shortages in the care sector, native women can move into other occupations and, as a result, enter paid employment instead of remaining at home performing activities that are not counted as employment. This is one possible explanation consistent with what we observe in the data: positive effects, particularly among highly educated women.
By contrast, men with intermediate vocational education are more similar in terms of education and skills to incoming foreign workers, which makes the competition channel more important for them. We should also remember that although migrants from the newer EU Member States are relatively well educated on average, language barriers and requirements for recognising foreign qualifications often push them into jobs below their formal qualification level.
Despite these modest short-term fluctuations, the effects tend to even out after a few years as the labour market adjusts to the new situation.
One problem with this type of research is that migrants do not move randomly across regions — naturally, they tend to go where there are more jobs. How can you distinguish the genuine effect of immigration from the fact that migrants simply move to economically more successful regions?
That is the main challenge in virtually all economic estimates that try to get closer to identifying causal effects. Growing regions experience rising labour demand and are therefore better able to absorb an inflow of workers; in fact, they may even face labour shortages.
To measure the effect of current immigration on the current employment of native workers, we need, technically speaking, a variable that is related to current immigration but not to favourable current economic conditions in the region. Card proposed a very clever instrument for this purpose. The idea is to use the historical distribution of foreign workers — in my case, data from the 1991 and 2001 censuses. Because we know that immigrants tend to settle in areas where people from the same country of origin have already settled, I divided them into groups according to origin and used this historical distribution to predict today’s distribution.
In practice, this instrument distributes the current total number of foreign workers across regions according to their historical settlement patterns — in other words, it tells us what the distribution would look like if historical settlement patterns had remained unchanged. The total number of migrants stays the same; only their distribution across regions comes from the past. This variable is no longer driven by today’s favourable economic conditions and therefore allows us to move much closer to a causal interpretation of the estimate.
When the article was published, you mentioned on social media that this research is also personally very close to you: when you were younger, your mother went to work in Austria. Did that experience influence the questions you asked in your research? And looking at the results today, do they also help you understand what you experienced at the time?
Thank you for the question. We have to separate several different aspects here. My study obviously does not explain how I felt when I was left at home without my mother — that is not its purpose. But it does help me understand why these questions interest me in the first place.
The original research question came from the Austrian AMS, so it was not even something I chose myself. But the fact that I enjoyed working on it, that I applied for a position at the institute and joined a research group focusing on labour market inequalities and social policy — that is already part of the story. So is the fact that I later developed the project further into this study.
Part of our lives is shaped by chance, and part of it we subsequently shape according to our own preferences. In any case, I have come to realise that the questions we ask — whether in research or in everyday life — are connected to our own histories and to the way we perceive the world. Experiencing difficult situations can make us more open to understanding issues that might never even occur to someone else.
That is also why I see enormous potential in migration. Migration is a difficult experience for anyone and, for people fleeing their homes, it can also be traumatic. Because of my own experience, I volunteered in Austria with children from migrant and refugee families. We did not just study together; we also spent ordinary enjoyable days together, visiting museums, going to the theatre and attending different events.
One of the children I used to mentor is now studying for his second university degree, has a family, and volunteers himself. The project is still running today, is highly successful, and demonstrates how much potential there is in people who have been through difficult periods in their lives. In Slovakia, I know of a similar project in Roma settlements, Omama, and I believe there are now others as well. I am convinced that everyone who works in such a project has their own personal reasons for doing so — along with a certain amount of chance in how they ended up there in the first place.
Your results therefore suggest that the question is not simply whether immigration is good or bad for the labour market, but rather who benefits from it and who may lose in the short run. What do you think follows from this for economic and social policy — and perhaps for Slovakia in particular?
Yes. As with automation and many other economic changes, there are so-called “winners” and “losers”. In Slovakia, we do not yet have sufficiently detailed data to estimate these effects in advance, but we can learn from international experience, and we do have current data on the educational and employment structure of foreign workers coming to Slovakia. ÚPSVaR regularly monitors these data using information from the Ministry of Interior. They should be actively used to estimate possible scenarios.
If we know how many workers have arrived in each education group, and we also know where labour shortages exist, we already have enough information to identify where losses due to greater competition are more likely and where native workers are instead more likely to switch occupations.
We also know that collective bargaining coverage is almost 98% in Austria, compared with roughly 25–28% in Slovakia. This means that in Slovakia a larger share of the adjustment would probably occur through wages rather than employment. Austrian results therefore cannot simply be transferred directly to Slovakia.
The fastest and most severe adjustment would be likely to occur where worker protection is weakest: in illegal employment and among bogus self-employed workers, a category in which Slovakia has one of the highest shares in the EU. These workers can lose their jobs relatively easily, and this is where I would expect the largest losses, particularly among low- and medium-skilled workers.
At the same time, Slovakia’s Labour Code is relatively inflexible, which serves as a form of social protection against the risk of sudden job loss. However, we now know that a certain degree of flexibility is necessary for labour market dynamism. In my view, this combination would also contribute to greater wage competition and downward pressure on wages.
Finally, active labour market policies should identify and target vulnerable groups early. The current system does not detect unemployment risks soon enough and often reaches people only once they have already become long-term unemployed. In many cases, employment office staff begin working intensively with a person only after they have been registered as unemployed for a year. As a result, even though overall unemployment is low, Slovakia continues to have a persistent core of long-term unemployed people who are extremely difficult to help back into employment.
