Abstract
The eastern enlargements of the EU and the extension of the free movement of labour to the new Member States have provided for increased mobility in the EU. This article reviews the impact of the newly emerged east-west mobility patterns on the labour market options of the labour force in host Member States and stayers in the source countries. The literature generally finds a positive aggregate effect on the host labour force, although micro-level adjustment may create winners and losers and the integration of immigrants has been limited. In spite of the fears of brain drain from the sending countries, empirical evidence also indicates positive effects on wages and employment of the stayers. A study on the effects of husbands’ employment abroad on the labour market status of Slovak women provides additional evidence of positive effects of husbands’ out-migration on their wives’ labour market outcomes. This study concludes that free mobility in an enlarged EU led to an increase in the efficiency of its labour markets.
Introduction
The eastern enlargements of the European Union (EU) in 2004, 2007, and, most recently, 2013, have, through the progressive extension of the right of free movement of workers, unleashed Europe’s mobility potential in an unprecedented way. The number of EU-10 and EU-2 migrants combined in the EU-15 increased from fewer than 2 million in 2004 to almost 5 million in 2009, an increase of about 150 per cent over just five years. 1 As work opportunities in the receiving countries dried up with the onset of the Great Recession, east-west mobility in Europe slowed down in 2009. Nevertheless, migration flows intensified in the early 2010s, as some of the sending countries continued to suffer from economic difficulties and in particular a lack of job opportunities. This large-scale reallocation of labour in the EU has had a profound impact on EU labour markets as well as on the broader migration debate in the EU.
A growing literature has looked at the causes and impacts of mobility in an enlarging EU. Kahanec and Zimmermann (2010) review Europe’s experience during the first years after enlargement, providing evidence of generally neutral-to-positive aggregate effects of post-enlargement mobility. Kahanec et al. (2013) find positive effects on GDP per capita in the receiving economies, Blanchflower and Shadforth (2009) and Kahanec et al. (2013) provide evidence of inflation-curbing effects, and Dustmann et al. (2003) find that there have been benefits for the sustainability of welfare systems. There is no evidence of negative effects on aggregate wages or employment (Kahanec, 2013).
Kahanec et al. (2016), and Fertig and Kahanec (2015) look at the determinants of mobility after the EU’s eastern enlargement. They find that EU accession and labour market opening, as well as stronger economic performance and the presence of diasporas in receiving labour markets, had independent positive effects on east-west mobility in Europe. Giulietti et al. (2013), De Giorgi and Pellizzari (2009), and Giulietti and Wahba (2013) argue that mobility in Europe is not driven by welfare benefits.
On the other hand, post-enlargement migrants are not yet fully integrated into receiving labour markets, as their jobs typically are below immigrants’ formal level of qualification (Dustmann et al., 2003; Kahanec et al., 2016). This finding may be an artifact of various impediments in the process of adjustment and skill transferability, but also of unequal treatment, or migrants’ choices (Guzi and Kahanec, 2015). Meardi (2012) warns that the post-enlargement migrants may be turning into a new underclass in the EU, and Anderson et al. (2006) find a degree of dissatisfaction among post-enlargement migrants in the UK.
From the sending countries’ perspective the risk of brain drain and adverse demographic consequences of out-migration, and signs of emerging patterns of circular migration that can potentially benefit the sending as well as receiving countries, have been studied by Kaczmarczyk et al. (2010), Kahanec and Zimmermann (2010) and Zaiceva and Zimmermann (2016). Few studies have looked at the labour market effects of post-enlargement migration on the stayers. Elsner (2013a, 2013b) shows that out-migration from Lithuania increased the wage level of the stayers. A similar study by Dustmann et al. (2012) finds that out-migration from Poland led to slightly increased aggregate wages, but the effect was the opposite for the low-skilled stayers. Pryymachenko et al. (2011) document the role of out-migration from the EU-8 (EU-10 minus Cyprus and Malta) to EU-15 countries in limiting unemployment rates in the sending countries.
This article reviews Europe’s experience with workers’ mobility following the EU’s eastern enlargements. I adopt a broad perspective and review the contexts, scale and effects of EU’s eastern enlargements. The article proceeds as follows. The next section discusses the institutional underpinnings of free movement in the EU. I then discuss theoretical arguments enabling us to discern key variables and hypotheses about labour market effects of labour mobility. The scale and composition of post-enlargement mobility in Europe are reviewed in the following section. In subsequent sections I discuss the economic effects on the receiving and sending countries. The following section shifts the focus to the household level, looking at the labour market status of Slovak wives left behind their migrating spouses following Kahanec (2015). Finally, I interpret the main findings and conclude.
Institutional underpinnings of free mobility in the EU
The EU’s eastern enlargements of 2004 and 2007 unleashed Europe’s potential for east-west mobility, which had been blocked by the Iron Curtain and various other barriers for several decades. The right of free movement of workers in the EU is guaranteed by Article 45 of the Treaty on the Functioning of the EU; and it has been extended by the Treaty on the European Union, Directive 2004/38/EC, and the case-law of the European Court of Justice to cover all EU citizens and their close family members, provided they have health insurance and means to support themselves. These free movement rights extend to Iceland, Liechtenstein and Norway (as members of the European Economic Area) and, based on a separate treaty, to Switzerland.
Given the decades-long period of economic, social, political as well as mental separation of eastern and western Europe, the creation of a single European labour market in which workers can freely move across the Member States was not free from problems, however. Economic and social gaps as well as the population size of the new entrants added to the controversy and resulted in the so-called transitional arrangements that provided old Member States with the option to apply national legislation to restrict the access of workers from the new Member States (excepting Cyprus and Malta) to their labour markets for up to seven years, with reviews after two and five years following enlargement. Transitional arrangements, however, did not restrict the mobility of entrepreneurs and self-employed workers, who could freely move across the EU immediately following their country’s EU accession under the provisions on freedom of establishment and freedom to provide services guaranteed by EU Law.
Although Ireland, Sweden and the UK opened up their labour markets immediately following the 2004 enlargement, and a further handful of countries opened up soon afterwards in 2007, most old Member States made use of transitional arrangements and delayed labour market opening to citizens from new Member States for up to seven years following the respective enlargement (Table 1). Eventually, however, citizens of the new Member States gained the full right to enter any other Member State and reside as well as seek and accept employment there. They also gained the right to access any social benefits associated with their employment.
Overview of transitional arrangements (date of termination)*.
* Ireland, Sweden and the UK chose not to apply transitional arrangements; hence we report EU enlargement as the date of labour market opening.
Theoretical arguments
EU enlargement and the opening up of EU labour markets lowered the costs of migration for the citizens from the new Member States. There were (and indeed still are) large socio-economic gaps between the old and new Member States. In such circumstances, theory dating back to Harris and Todaro (1970) predicts that a decrease in migration costs will lead to increased migration flows. Among the key factors of migration proposed by the literature are disparities in terms of expected earnings and income (adjusted for cost of living migration costs), career and employment opportunities, or the availability and quality of public goods and amenities (Massey, 1990; Borjas, 1999; Bonin et al., 2008). Risk sharing within households may be another factor for a household member to seek employment abroad (Stark, 1991). On the other hand, the existing statistical evidence shows the much feared ‘welfare tourism’ to be unfounded (De Giorgi and Pellizzari, 2009; Giulietti et al., 2013). However, Kureková (2011) shows that welfare systems in the sending countries may affect out-migration.
Other factors that may influence migration costs and hence migration decisions include geographical, linguistic and cultural distances between source and host populations (Chiswick and Miller, 2011). From the perspective of human capital theory, age and skills of potential migrants may affect the costs and benefits of migration as well as the migrants’ adjustment process (Becker, 1957; Sjaastad, 1962). These selection mechanisms may affect the composition of migration flows, that is, selection of migrants, which may vary across sending and receiving countries as well as over time (Borjas, 1987; Chiswick, 1999). It then follows that EU enlargement and opening up of EU labour markets can be expected to trigger substantial east-west mobility in the EU and that the scale and composition of these migration flows can be expected to vary across sending and receiving countries.
What does theory tell us about the expected effects of post-enlargement east-west mobility in EU labour markets? A simple economic textbook model would predict increasing wages in the sending countries, as out-migration should increase the scarcity of labour there. In contrast, inflows of workers from new Member States should, according to the basic model, decrease wages in the receiving EU Member States. This simple model seems to underpin much of the public debate and to fuel fears of labour substitution in receiving countries in Europe and elsewhere. Although based on a salient argument at the aggregate level, it ignores important heterogeneities and complementarities in the labour market.
Perhaps most importantly, immigrants and natives are assumed to be perfect substitutes in such a model; an assumption that is clearly incongruent with the reality. The literature indeed documents that migrants concentrate in different sectors and occupations than natives, and very often downskill into jobs that are below their level of formal education in terms of required skills (Dustmann et al., 2003; Kahanec, 2013). In addition, neither immigrants nor natives are perfectly substitutable among themselves. Kahanec (2013) provides a stylized model which treats high- and low-skilled labour markets in sending and receiving countries separately. The implications of this model depart from those of the simple model treating labour as homogenous in that in terms of wages and/or employment immigration benefits complementary labour and may impair substitutable labour in receiving countries. The opposite effects arise in sending countries: wages of workers complementary (substitutable) to the out-migrating workers are negatively (positively) affected. 2 It then follows that the effects of mobility is an empirical question and the composition of migration flows is key to understanding their effects in receiving and sending labour markets.
The scale and composition of post-enlargement mobility
EU enlargement and the opening up of EU labour markets were followed by a large increase in east-west migration flows in the EU. According to Holland et al. (2011), from 2004 to 2009 the number of citizens of the new Member States (EU-8+EU-2) residing in the EU-15 increased from somewhat below 2 million to almost 5 million, an increase of about 150 per cent. Indeed, Kahanec et al. (2016) show that EU enlargement as well as the opening up of labour markets in the EU-15 countries have had independent, statistically significant positive effects on east-west migration. In effect, in 2009 the combined migrant population from the new Member States (EU-8+EU-2) in the EU-15 constituted 1.22 per cent of the combined host population of the EU-15, and 4.75 per cent of the sending countries’ population combined. After an initial surge during the early to mid-2000s, migration dynamics slowed down during the Great Recession, only to pick up again in the early 2010s, probably in response to protracted hardship in some of the sending countries following the Great Recession (Kahanec, 2013).
While the scale of post-enlargement migration is an important variable determining its effects, from the theoretical argument advanced above the composition is equally important, for it determines not only the scale, but also the sign of the effects on various segments of the labour market. Based on the existing literature we can summarize the main qualitative characteristics of post-enlargement migration flows as follows. Although the largest fraction of post-enlargement migrants has a medium level of education, migrants with a high and low level of education are over-represented with respect to most source populations (Holland et al., 2011). Compared to the receiving countries’ populations, relatively high levels of educational attainment of citizens from the new Member States were documented by Elsner and Zimmermann (2016) for Germany, Barrett (2010) for Ireland, Gerdes and Wadensjö (2010) for Sweden, and Blanchflower and Lawton (2010), and Clark and Drinkwater (2016) for the UK.
Kahanec (2013) reconstructs the educational attainment of cohorts of immigrants from the EU-10 and EU-2 in the EU-15 based on the 2009 wave of the EU Labour Force Survey and the information about the year of arrival. The study finds that EU-10 nationals in the EU-15 were more educated than the EU-15 population, with 26.1 per cent in the high-educated category and 22.5 per cent in the low-educated category. Among EU-2 nationals in the EU-15 the corresponding shares were 12.2 per cent and 37.5 per cent, which compares unfavourably to 18.9 per cent high-educated EU-15 nationals but favorably with respect to the 45.7 per cent low-educated EU-15 nationals. Kahanec (2013) confirms that both EU-10 and EU-2 migrants were positively selected from their respective source populations, exhibiting 14.4 per cent high-educated and 27.4 per cent low-educated in the EU-10 and 10.3 per cent high-educated and 40.9 per cent low-educated in the EU-2.
Positive selection on educational attainment of out-migration to the EU-15 has been confirmed by Kaczmarczyk (2016) for Poland and Andren and Roman (2016) for Romania. Hárs (2016) corroborates this finding for out-migration from Hungary, although she finds the positive effect of education to be declining. Hazans and Philips (2010) and Hazans (2012) show that there was variation across the source countries, as post-enlargement out-migration was less educated than the source populations in the Baltic states. Kahanec and Mýtna Kureková (2016) find a small negative effect of having a higher educational attainment on out-migration from Slovakia. Holland et al. (2011) corroborate a relatively high educational attainment of post-enlargement east-west migrants, but find that vis-à-vis the source populations low-educated were somewhat over-represented and medium- and high-educated under-represented among migrants.
After EU enlargement the share of highly educated among EU-10 nationals in the EU-15 increased significantly, whereas the proportion of low-educated initially decreased but later increased (Kahanec, 2013). On the other hand, Brenke et al. (2010) find that EU-8 migrants in Germany were less educated and older after EU enlargement than previously. The educational attainment of EU-2 nationals in the EU-15 somewhat deteriorated during the first two years following Bulgaria’s and Romania’s accession to the EU in 2007, primarily due to an increase in the share of low-educated among these migrants. In 2009, however, the share of highly educated increased whereas low-educated decreased among EU-2 migrants (Kahanec, 2013).
It would however be erroneous to equate migrants’ educational attainment with the skills these migrants actually use in destination countries. Kahanec and Zimmermann (2010), Dustmann et al. (2003) or Kahanec et al. (2016) document significant downskilling of post-enlargement migrants into jobs below their level of formal qualification. Tijdens and van Klaveren (2012) show that among EU-10 nationals in the EU-15 only 65 per cent report a correct match between their job and education, almost 10 percentage points below the corresponding figure for the host population and 7 percentage points below the figure for all migrants. In sum, whereas EU-10 and EU-2 nationals in the EU-15 are relatively well skilled, the jobs they get do not always correspond to their level of qualification. This discrepancy may decrease with years since migration and vary across immigrant cohorts; however, de la Rica (2010) shows little adjustment of post-enlargement migrants over time. Kahanec and Mýtna Kureková (2016) argue that the prevalent profile of Slovak out-migration has changed from young and educated before the Great Recession to older and medium-educated during the recession; with the degree of downskilling being lesser for migrants with the latter profile. Hence, changing patterns of downskilling may interfere with simple implications about changing patterns of migrants’ skills as measured by their formal educational attainment. Another aspect is that a large proportion of post-enlargement migrants are temporary and report that they have only temporary migration intentions (Kahanec and Zimmermann, 2010). Such migrants, however, may prefer to invest less in skills specific for destination countries (Kahanec and Shields, 2013).
Generally, migrants from the new Member States in the EU-15 are relatively young, although their population is ageing and distinct groups of migrants in their forties and fifties have emerged (Kahanec and Zimmermann, 2010; Kahanec and Zimmermann, 2016). Post-enlargement migration is also gendered, with males and females over represented in some sectors and destinations (Kahanec and Zimmermann, 2010; Kahanec and Zimmermann, 2016). Finally, EU-10 and EU-2 nationals concentrate in certain sectors and occupations, including construction, manufacturing, services and agriculture (Kahanec et al., 2010).
The effects on the receiving labour markets
The picture that emerges from the previous section indicates that post-enlargement migrants specialize in certain sectors and occupations in the receiving labour markets. Moreover, their labour force participation rates are higher than the participation rates of the natives in most EU Member States (Kahanec et al., 2016). Although they also exhibit higher unemployment rates, their employment rates still often exceed those of the natives (Kahanec et al., 2016). Such observations may indicate that migrants from the new Member States respond to sectoral labour demand in receiving labour markets and fill existing labour and skill shortages. It could also be, however, that they are crowding out native workers using a low-pay strategy as their competitive advantage.
At the aggregate level, workers from the new Member States may in theory complement but also substitute the native labour force. As discussed above, this is a key factor determining the effects of immigration. The aggregate effects of post-enlargement migration on natives’ labour market status thus depend on the extent of the two types of interaction, complementarity vs. substitution, which is an empirical question.
A number of studies have evaluated the effects of east-west migration in the EU-15. Kahanec and Zimmerman (2010) review early evidence to conclude that the aggregate effects on the native workers were generally neutral or positive. Even in Ireland, the country with the largest inflows of EU-10 workers relative to the country’s size among EU-15 countries, local labour markets adjusted and native workers were not negatively affected on aggregate (Doyle et al., 2006). Similar results are found for the UK (Gilpin et al., 2006; Blanchflower et al., 2007; Lemos and Portes, 2008). Brenke et al. (2010) show that in Germany immigrants from the EU-8 countries specialize in low-skilled jobs, competing with other immigrants rather than with the natives.
Kahanec and Pytliková (2015) and Kahanec et al. (2013) use a unique dataset about bilateral migration flows between new and old Member States to evaluate the effects of east-west mobility on receiving countries’ economies. They find positive effects of migration from the EU-8 on GDP, GDP per capita and employment rate and negative effect on output per worker in the EU-15.
Blanchflower and Shadforth (2009) and Blanchflower et al. (2007) argue that expectations of increased immigration from the new Member States reduced inflationary pressures in the UK. Barrett (2010) credits post-enlargement migrants to Ireland with helping the country to moderate wage growth and hence enhancing Ireland’s competitiveness during the boom preceding the Great Recession.
Several studies show that the much feared effect of post-enlargement migration on welfare systems in the receiving countries are unjustified and that immigrants may in fact benefit receiving countries’ welfare systems (Gerdes and Wadensjö, 2010; Doyle, 2007). Dustmann et al. (2003) and Dustmann and Frattini (2014) find that post-enlargement migrants in the UK are net contributors to UK’s public finances. Giulietti et al. (2013) and De Giorgi and Pellizzari (2009) show that welfare generosity is an essentially insignificant factor when it comes to mobility in the EU. 3
The effects on the sending labour markets
The implementation of free movement of workers was generally welcomed by the new Member States, although fears of brain drain surfaced, too. Some authors warned early on that out-migration may aggravate already adverse demographic trends in new Member States and further undermine their public finances (Kadziauskas, 2007). Others reported growing shortages in some sectors and occupations in the sending countries following out-migration to the EU-15 (Kaczmarczyk and Okólski, 2008; Kadziauskas, 2007; Kureková, 2011). A new trend emerged whereby these shortages were filled by immigrants from the EU’s eastern and south-eastern neighbours (Frelak and Kazmierkiewicz, 2007; Iglicka, 2005; Kureková, 2011).
In spite of emerging shortages in some sectors, such as health care, several studies on post-enlargement mobility found no sign of significant brain drain from the new Member States (Frelak and Kazmierkiewicz, 2007; Brücker et al., 2009; European Commission, 2008; Hazans, 2012). More recent evidence, however, warns that return migration may be negatively selected (Hazans, 2012) and the scale of out-migration during the Great Recession might have lasting adverse consequences for the Baltic states (Hazans, 2016).
The economic loss due to outflow of workers may be compensated by remittances. Kahanec et al. (2010) report growing remittances to EU-2 countries as well as the Baltic states. Remittances to the EU-8 and EU-2 declined during the Great Recession, however (Comini and Faes-Cannito, 2010). Kaczmarczyk and Okólski (2008) find that the use of remittances changed over time: whereas they were used to finance consumption and durable goods during the early post-enlargement period, subsequently human capital investments were financed.
As for labour market effects, Elsner (2013a, 2013b) finds that out-migration from Lithuania increased wages and decreased unemployment in the country. A similar study by Dustmann et al. (2012) finds that out-migration from Poland led to slightly increased aggregate wages, but the effect was the opposite for the low-skilled stayers. Pryymachenko et al. (2011) document an attenuating role of out-migration from the EU-8 (EU-10 minus Cyprus and Malta) to EU-15 countries on unemployment rates in the sending countries. These findings are in line with theoretical arguments outlined above. Kaminska and Kahancová (2011) show that institutional contexts mattered as well, documenting that trade unions in Slovakia were able to use the argument of out-migration to obtain wage increases and Polish workers were able to achieve a similar outcome through public protests co-organized by trade unions.
The household level: labour market status of wives left behind
The literature reviewed above indicates that out-migration has increased wages and decreased unemployment in sending countries (Elsner 2013a, 2013b; Dustmann et al., 2012). These studies adopted a meso-level approach, whereby individual observations were grouped by educational attainment and experience (skill), and out-migration from such skill groups was studied as an explanatory factor of the observed variation of wages of workers across groups and over time. A recent study by Kahanec (2015) shifts the focus to the household level, studying the effect of the husband’s employment abroad on the wife’s labour force participation, employment, part-time or temporary employment. In other contexts existing studies generally find a negative effect on wives’ labour force participation and employment (Acosta, 2006; Amuedo-Dorantes and Pozo, 2006; Mendola and Carletto, 2009; Binzel and Assaad, 2011). Remittances are proposed as a key factor behind these findings, as they increase wives’ reservation wages. On the other hand, husbands’ employment abroad may result in positive peer effects on their wives. Under credit constraints they may also enable households to start a new business (Binzel and Assaad, 2011). Finally, employment opportunities for women may improve if they can substitute the out-migrated men in production.
Using the Slovak Labour Force Survey (SK LFS), Kahanec (2015) is the first to study the link between husbands’ foreign employment on their wives’ labour market status in the context of post-enlargement migration in the EU. A key methodological challenge is that households’ decision to send the husbands to work abroad is not exogenous, and may be affected by the labour market status of the wives. Kahanec addresses this issue by adopting an instrumental variable framework, constructing an instrumental variable that combines the information about labour market opening and the industrial structure of receiving labour markets with the husbands’ field of study. This approach is based on the idea that the opening of labour markets increases the probability of migration to these labour markets especially of those husbands whose education matches the sectoral composition of open labour markets; and that there is no direct effect of sectoral composition of receiving labour markets on the labour market status of wives in sending labour markets.
Using this instrumental variable in a Two-Stage Least Square (2SLS) model, Kahanec (2015) finds that husbands’ employment abroad affects female spouses’ participation rate positively, decreases unemployment and part-time employment, and has no effect on temporary employment. The effects on unemployment and part-time employment disappear, however, if past labour market status and field of study are controlled for. In a seemingly-unrelated Probit framework estimated using a maximum likelihood method, the author shows that the effect of husbands’ foreign employment on wives’ labour force participation is not statistically significantly different from zero; however, husbands’ employment abroad decreases the probability of wives’ unemployment or temporary employment. No significant effects are found for part-time employment.
The results of Kahanec (2015) hence correspond to those found by Elsner (2013a, 2013b), Dustmann et al. (2012) and Pryymachenko et al. (2011) in that in all these studies out-migration improves the labour market status of those left behind. They also indicate that the detrimental role of spouse’s out-migration found in other contexts such as Albania, Egypt, El Salvador or Mexico (Acosta, 2006; Amuedo-Dorantes and Pozo, 2006; Mendola and Carletto, 2009; or Binzel and Assaad, 2011) is not confirmed for Slovakia. Among the reasons for this may be a relatively low scale of remittances in Slovakia, strong peer effects from male to female spouses, or a broader general equilibrium effect.
Conclusions
This article reviewed the scale and effects of east-west migration in the EU following its eastern enlargements. The picture painted in the literature is rather positive: most studies find neutral or positive economic effects on receiving labour markets. In particular, the fears of negative effects on wages or employment are not supported by available evidence. Similarly, post-enlargement migration does not undermine European welfare systems: incoming workers seem to be net contributors and there is essentially no evidence supporting the welfare magnet hypothesis. In spite of the generally neutral-to-positive aggregate effects, local adjustments may create winners and losers.
Available studies show that out-migration increases wages and employment in the sending countries. Similarly, there is evidence supporting neutral-to-positive effects of out-migration of husbands on their wives’ labour market outcomes. On the other hand, the risk of brain drain, downskilling and dissatisfaction with own experience of migration may pose challenges for free movement in Europe.
Overall, this study contributes to the literature on east-west mobility in an enlarged EU. A review of the available literature indicates that at the aggregate level the effects of post-enlargement migration on EU labour markets have been generally positive. On the other hand, adjustment at the individual level may have been benefiting some but hurting others. The long-run effects will depend on the extent to which the EU succeeds in building a truly common EU labour market with fluid movement between its Member States, such that circular migration patterns emerge and provide for brain circulation and brain gain.
Footnotes
Acknowledgements
I am grateful to an anonymous referee for very constructive and helpful comments. I remain responsible for any remaining errors.
Funding
The author acknowledges the financial support of the Eduworks Marie Curie Initial Training Network Project (PITN-GA-2013-608311) of the European Commission’s 7th Framework Programme.