Free Movement of Workers after Enlargement: The ignored Substitution Effects!

As a general rule, immigration is a good thing for receiving countries regarding their GDP (not GDP per capita!) and overall employment level. But some vulnerable groups are negatively affected in terms of wages and unemployment. Let’s have a closer look at the latter, although this is not an easy task at least for two reasons:  Firstly, political correctness doesn’t like us to talk about substitution effects of immigration. And, secondly, if you start talking about substitution effects, you easily find yourself in a situation of doing the job for right wing political parties – and who would like to be in that position?

In General

In a theoretical (textbook) framework immigration shifts the supply curve to the right, increasing overall employment levels but reducing the employment of nationals; substitution effects depend on the elasticities of demand and supply curves. Beyond that, immigration has a dampening effect on wages.  Of crucial importance in this reasoning is the assumption of foreign and native workers being perfect substitutes. This might be the case for unskilled workers and persons with same qualifications, motivation and reservation wages. If their skills and competences are complementary, then immigration will even increase employment of nationals. These positive effects are particularly large when migrants fit to vacancies which couldn’t be filled with natives. All in all, theory cannot predict the relation of immigration and employment of natives a priori; only empirical evidence can!

Anyway, employers normally like immigration because more labour supply has dampening effects on wages and more competition on the labour market makes it easier for employers to pick and choose. And normally, migration has positive effects for migrants themselves. But some labour market groups in the receiving countries – usually the ones with the same skills and competences as the immigrants – have to face negative effects regarding their incomes and employment probabilities. There are substitution effects.

Empirical research [1] normally concludes that an increase of the share of foreigners in total employment by 1 percentage point reduces wages of national employees in the short run by some 0.1%. In many studies, the correlation between immigration and wages or employment probabilities of nationals is insignificant. But these results might be due to “measurement errors” – this is my main point I would like to raise:  You will not be able to detect substitution effects if you average out on a too large scale, i. e. for several years or for sectors on a national level. If you allow your data to reveal substitution effects, you have to look on a regional and sectoral level, capturing occupations, skills and competences of migrants and nationals.

The long run empirical evidence of the labour market impact of immigration in Austria since 1989 is inconclusive in so far as it is measured at national level. But there are signs of displacement: Since the breakdown of the Iron Curtain employment of Austrian nationals increased by 8.8%, employment of non-nationals increased by 192.1%. During the same time, the unemployment rate of early school leavers increased by 94.6% (from 9.2% to 17.9% according to national definitions). Wage increases in Austria were quite moderate in this period. The yearly changes in employment show in many cases a parallel increase of both, nationals and foreigners. But in 2010, for example, 93% of overall employment increases was non-national, in 2011 the same figure was 61.2%. Beyond the overall situation at the national level, there is one important feature of immigration to the Austrian labour market, in particular from eastern European countries: Many of the migrants have good qualifications when they come to Austria, but find jobs only in occupations where their skills and competences are underutilised – so “over-qualification” is a big issue [2]. And this is a particular dangerous aspect for low qualified employees in Austria!

More specifically: Full Liberalization of Worker´s Mobility for Migrants and Commuters from EU8 – one Year after

In the first year of free labour mobility some 34.000 additional workers i. e. some 1% of total employment from EU8 (eastern European accession countries of 2004) have found jobs in Austria. This is at any rate a rather low inflow! But the influx was strongly concentrated at particular regions in the eastern part of Austria and in specific sectors like catering, construction, market services and agriculture. For example, employment from EU8 countries increased in the first year of free movement in Burgenland (bordering to Hungary) by some 22% in agriculture and forestry and by some 12% in lodging and catering. In Vienna, the same figures increase by some 11% in private household services. For 22 branches and 9 regions, the cross sectional and cross regional correlation of the year-on-year changes of EU8 employment and the unemployment of early school leavers was remarkable close with R2=0.46.

A more sophisticated empirical framework of a panel regression with fixed time effects [3] and changes of employment (of EU8, EU14, other non-nationals, nationals) und unemployment (overall and erarly school leavers) on a monthly basis revealed the following results: In certain sectors in some regions there were statistically significant substitution affects at work in the first year after full liberalisation (see table below). Particularly strong negative effects of inward migration could be found for the unemployment of unskilled workers. To give an example of the magnitude of the effect for Burgenland: An increase of 10 employees from EU8 in the construction sector reduced domestic employment by some 3.3 people in the same sector and employment of foreigners from third countries by 0.9 persons. The unemployment of early school leavers was increased by 1.7 persons.

To conclude

Like it or not: Immigration leads to substitution effects for domestic workers, at least to some extent. To ignore and deny the problems of the negatively affected vulnerable labour market groups would be a politically risky strategy. 


Bock-Schappelwein, J. Bremberger, Ch, Hierländer, R., Huber, P. Knittler, K., Berger, J, Hofer, H., Strohner, L. (2008), Die ökonomischen Wirkungen der Immigration in Österreich. See http://  www. Studien_

Huber, P., Bös, G. (2012), Monitoring der Arbeitsmarktauswirkungen der Zuwanderung aus den neuen EU-Mitgliedsländern im Regime der Freizügigkeit. See http://www.forschungsnetzwerk. at/downloadpub/WIFO_bmask_Monitoring_Endbericht_2012.pdf

Longhi, S., Nijkamp, P., Poot, J. (2004), Meta-Analysis of Empirical Evidence on the Labour Market Impacts of Immigration, Tinbergen Institute Discussion Paper, TI2004-134/3

Schweighofer, J. (2012), Gab es auf regional sektoraler Ebene Verdrängungseffekte im Gefolge der Arbeitsmarktöffnung vom Mai 2011, in: Wirtschaft und Gesellschaft 3/2012. See

[1] See for example Longhi (2004), Huber/Bös (2012) and Bock-Schappelwein (2008).

[2] See for example SEJ 11/10/2012, EU Labour Migration during the Crisis by Bela Galgoczi and Jeanine Leschke.

[3] For details see Schweighofer (2012), p. 609ff.


  1. Kathleen McMullen says

    The ‘free’ movement of labour includes many who don’t find work. I met a group the other day: 2 Irish men, I Polish woman, I French male, etc, homeless, without work, sleeping rough on the street.

    There is nothing ‘free’ about people migrating to remain homeless in a different country. In fact it’s outrageous. Labour needs to be properly housed and serviced. With our massive housing shortage expect social tensions to erupt next year, expect more unemployed migrants joining the charity soup queues.