## Jacqueline O'Reilly, Janine Leschke, Renate Ortlieb, Martin Seeleib-Kaiser, and Paola Villa

Print publication date: 2018

Print ISBN-13: 9780190864798

Published to Oxford Scholarship Online: January 2019

DOI: 10.1093/oso/9780190864798.001.0001

Show Summary Details
Page of

PRINTED FROM OXFORD SCHOLARSHIP ONLINE (www.oxfordscholarship.com). (c) Copyright Oxford University Press, 2019. All Rights Reserved. Under the terms of the licence agreement, an individual user may print out a PDF of a single chapter of a monograph in OSO for personal use (for details see www.oxfordscholarship.com/page/privacy-policy).date: 18 March 2019

# What happens to young people who move to another country to find work?

Chapter:
(p.389) 13 What happens to young people who move to another country to find work?
Source:
Youth Labor in Transition
Publisher:
Oxford University Press
DOI:10.1093/oso/9780190864798.003.0013

# Abstract and Keywords

This chapter analyzes the labor market integration of South–North and East–West migrants, together with intra-European and non-European Union migrants, vis-à-vis native peers in main European destinations. The analysis considers individual characteristics and labor market outcomes by migrant origins. Labor market outcomes are estimated, controlling for sociodemographic characteristics and for country-fixed and year effects. Using interaction effects, the chapter estimates whether the work-related outcomes of young migrants differ vis-à-vis native peers. The econometric analysis using pooled European Social Surveys (2002–2015) suggests that individual characteristics explain part of the migrant–native peer differences. Particularly, migrants from Eastern and Southern Europe exhibit important gaps vis-à-vis native peers regarding unemployment, contract type, and overqualification. Overall, migrant youth and women seem to be in vulnerable situations in destination labor markets. In addition to nondiscriminatory treatment, transparent competence screening and smooth skills transferability could alleviate such youth and gender vulnerabilities.

# 13.1. Introduction

The freedom of movement of citizens across all of Europe has been one of the most important achievements of the European Union (EU).1 The size, composition, and direction of migration flows in Europe have evolved in a continuously changing pattern, reflecting various social, economic, and political conjunctures and circumstances resulting from both diverse and dynamic pull and push factors (Castles 1986, 2006; Constant and Massey 2003). Recent evidence suggests, however, that the mobility patterns of the past decade in Europe are mostly dominated by youth flows (Eurostat 2011). In particular, educated youth from Eastern and Southern Europe have been migrating to regions to the west and north that offer relatively more favorable labor market opportunities (Kahanec and Zimmermann 2010). However, the recent economic downturn, which has contributed to rising youth unemployment, and the challenges faced by young people transitioning from education to labor markets have put a strain on the labor market transitions of youth. Added to these difficulties are the challenges migrants normally face in integrating into destination-country labor markets.

Against the background of human capital and neoclassical models explaining migration patterns and motivations (Sjaastad 1962; Bowles 1970; Greenwood, Hunt, and McDowell 1986; Borjas, Bronars, and Trejo 1992), and given the evidence that migrants are ever more frequently young, female, and relatively well educated, these population movements raise questions concerning the ability of destination-country labor markets to integrate migrants in accordance with (p.390) their human capital endowments. Economic theory predicts a strong correlation between the circumstances of the labor market at origin and in the destination countries (Martin 2009). Based on this theory, if young individuals move mainly to escape stressful economic circumstances in their countries of origin, then one wonders what happens to them once they arrive in the destination country’s labor markets. Previous results from the migration literature generally find relatively worse labor market outcomes for foreign-born individuals vis-à-vis native peers. In this vein, if international transferability of skills or qualification recognition is an issue, then it is possible to observe education–occupation mismatches among migrant individuals (Chiswick 2009). In addition to sociodemographic differences such as education and age, the role of ethnic background in the labor market has also been highlighted in explaining some of the observed differences compared to native peers (Akgüç and Ferrer 2015). Furthermore, young migrants sometimes face a double disadvantage: the first for their youthfulness, which usually means that they lack work experience and therefore have difficulty in making the transition from education to the labor market (Brzinsky-Fay 2007), and a second one in the form of the differential and discriminatory treatment that is commonly meted out to migrants. All in all, analyzing the labor market integration of young migrants has important policy relevance because it evidences the (in)effectiveness of labor market institutions (e.g., in terms of recognition of foreign qualifications) in tackling possible labor market mismatches faced by foreign-born residents in destination countries.

To this end, this chapter addresses the following research questions: Do recently arrived young migrants in Europe differ from native peers with respect to socioeconomic and labor market indicators? How do recently arrived young migrants from different regions of origin differ among themselves? To what extent do the observable differences in sociodemographic characteristics explain the gaps in the labor market outcomes of young migrants from various regions relative to native peers? Do we observe gender gaps in labor market outcomes among young migrants?

To address these questions, this chapter conducts a comparative econometric analysis of the labor market integration of young migrants of different origins. In a departure from the main literature on labor market integration (one exception is Spreckelsen, Leschke, and Seeleib-Kaiser, this volume), the chapter focuses on youth aged 35 years or younger because this age group accounts for a large share of the migrants in Europe in the past two decades. In particular, the analysis considers recent migrants who arrived within the past 10 years. Regarding labor market integration, the chapter examines a wide range of outcomes, such as (un)employment, type of job contract (temporary or permanent), self-employment, hours worked, and various indicators of occupational mismatch.2 Unlike the general approach in much of the previous research, migrants are not treated as a homogeneous group, and attention is paid to differences in ethnic origins. In line with the recent mobility patterns in Europe, the focus is on young migrants from (p.391) Eastern and Southern Europe, but other migrant groups are also considered so as to give a broader picture. Moreover, the novelty of the chapter is that it analyzes the labor market integration of young migrants in a cross-country framework. Last, because the gender gap is highlighted as an important factor in migrants’ experience, the chapter also contributes to the literature by embedding gender aspects in the analysis of the labor market integration of young migrants.

The descriptive findings point to differences in socioeconomic characteristics (e.g., age and education) as well as in labor market indicators (e.g., employment and occupational mismatch) across different migrant groups and between migrants and native peers. Econometric analysis suggests that observable characteristics explain part, although not all, of the differential labor market outcomes of migrants. Young Eastern European migrants are found to be overqualified for their occupations compared to native peers of destination countries. Young Southern Europeans are more likely to be self-employed and to be on a temporary employment contract. Regarding broader age groups, the younger cohorts seem to be performing worse than the older cohorts in terms of unemployment, self-employment, contract type, and overqualification, but these differences are not always statistically significant and they vary by the origin of individuals. Furthermore, important gender gaps are observed among youth in favor of men with regard to employment and hours worked per week, and this pattern holds for all migrant groups considered.

The remainder of the chapter is organized as follows. We first provide a brief literature review with a short background on recent migration trends in Europe. We next provide a description of the data, variables of interest, and the econometric methodology used for the micro-level cross-country analysis, followed by a presentation of the descriptive analysis and the estimation results. Finally, we discuss the results along the youth and gender dimensions and provide concluding remarks, suggesting areas for future research and discussing issues related to policymaking aimed at alleviating migrant and youth vulnerabilities in destination labor markets.

# 13.2. Literature review

The majority of the literature has focused on migrant integration into English-speaking countries, examining single-country cases (Chiswick 1978, 1979; Borjas 1987; Ferrer and Riddell 2008; Constant, Nottmeyer, and Zimmermann 2012). Most of these papers examine a limited number of labor market outcomes, such as wages (Chiswick 1978; Borjas 1987; Ferrer and Riddell 2008). There are a few studies comparing several countries, but even these do not always use comparable data sources (Constant and Zimmermann 2005; Antecol, Kuhn, and Trejo 2006; Algan et al. 2010). One novelty of this chapter is that it takes a comparative approach and conducts an analysis using harmonized cross-country data on (p.392) labor market integration covering various outcomes. Notwithstanding a number of caveats—discussed in Section 13.3—pooled cross-country data add to our understanding of differences in the integration of migrant populations across countries (Adsera and Chiswick 2007).

Most contributions find relatively worse outcomes for migrants compared to native peers in the labor markets for various reasons (Chiswick 1978; Adsera and Chiswick 2007; Jean et al. 2007). Although part of the nativity gap is related to socioeconomic background, such as education—where the latter has been obtained (Akgüç and Ferrer 2015)—and previous labor market experience, another part could be caused by skills recognition or transferability issues in destination countries (Chiswick 2009). Earlier studies also emphasize the assimilation process, whereby migrants catch up—if ever—with native outcomes only after a certain amount of time has been spent in the country and after obtaining country-specific skills (Chiswick 1978). Country of origin and cultural background are another set of related factors that determine labor market outcomes (Fernández and Fogli 2009; Blau, Kahn, and Papps 2011). Migration motivations, such as economic goals, education, political beliefs, or family reunification, might also be associated with integration patterns (Akgüç 2014), whereby the experience of economic and student migrants seems to more closely approximate that of native peers. Last, differential treatment in the form of discrimination might also lie behind native–immigrant gaps. Considering these dimensions, this chapter contributes to the literature by providing further insights into the labor market integration of recent young migrants in Europe by controlling for socioeconomic and ethnic backgrounds.

In the migration literature, the main focus is usually on working-age individuals rather than on migrating youth, except in some contributions, such as Seeleib-Kaiser and Spreckelsen (2016) and Spreckelsen et al. (this volume). Examining recent young European migrants in the United Kingdom, Seeleib-Kaiser and Spreckelsen find that although these migrants are highly integrated in terms of employment, they end up in poor-quality jobs. Similarly, Clark and Drinkwater (2008) find that recent Eastern European migrants to the United Kingdom experience relatively low returns on their education and work in unskilled occupations. This chapter likewise focuses on young migrants, but in a cross-country framework; the findings are nevertheless similar to those of previous papers. Although most of the aforementioned reasons for poor integration outcomes can be valid for young migrants as well, this group might also face the additional challenge of being young and the related risks to labor market transitions posed by lack of previous market experience and particularly of skills that are specific to the destination country. Finally, to our knowledge, none of the earlier studies addresses gender gaps while examining the labor market integration of youth migrant groups, as is done in this chapter.

As mentioned in Section 13.1, the chapter mainly focuses on Southern and Eastern European young migrants, even though other origins are included in (p.393) order to have a complete picture. The main reason for the focus on these groups are the recent mobility patterns in Europe. Regarding Southern Europe, Spain has turned from a migration destination during the boom years of 1995–2000 into an emigration country during the recent recession, whereby many young native peers and foreign residents have left to find employment elsewhere as jobs have become scarce (González Gago and Kirzner 2013; Izquierdo, Jimeno, and Lacuesta 2016). In the Italian case, despite the stable emigration in the pre- and postcrisis periods, the recent composition of migrants has changed to include more highly educated youth older than age 25 years, which suggests that the usual out-migration for study abroad has been replaced by work motives with lower return rates (Constant and D’Agosto 2008; Ciccarone 2013), thus raising the issue of brain drain (Beine, Docquier, and Özden 2011; Docquier and Rapoport 2012). Regarding migrants from Eastern Europe, the major policy change influencing their mobility has been the Eastern enlargement of the EU. However, EU accession did not immediately give the right of free movement and work to the citizens of the new member states,3 with transitional measures of up to 7 years restricting free movement for work purposes (Galgóczi, Leschke, and Watt 2011; Galgóczi and Leschke 2012).4 Regardless of the transition measures, a striking feature of recent migrant flows from Eastern Europe is that they are mainly dominated by young and well-educated individuals, as will be shown in the empirical analysis.

# 13.3. Data and methodology

To conduct the econometric analysis of labor market integration of migrants within a cross-country framework, we have at least two options regarding data sources: the European Union Labour Force Survey (EU-LFS) and the European Social Survey (ESS). Given the focus on Southern and Eastern European origins, we opted for the ESS, mainly because it provides detailed country-of-origin information. For example, we are not able to distinguish Southern European migrants in the EU-LFS, which gives only a broader country-of-origin categorization, such as EU15.

The ESS is a biennial—partly repetitive—cross-section survey including conventional demographic and socioeconomic variables as well as labor market indicators relating to diverse populations in more than 30 countries. The survey covers all persons aged 15 years or older who are residents within private households—regardless of their nationality, citizenship, language, or legal status—in the 36 participating countries (mainly in Europe). The survey is accessible via the Norwegian Social Science Data Services.

Using the ESS, we focus on 15 destinations, namely Austria, Belgium, Denmark, Finland, France, Germany, Ireland, Italy, Luxembourg, the Netherlands, Norway, Spain, Sweden, Switzerland, and the United Kingdom. These are countries that (p.394) have received important migrant flows during the past few decades not only from within but also from outside Europe (Brücker, Capuano, and Marfouk 2013). The migration flows to these destinations have been influenced by and have evolved through various economic, social, and political developments during this period—for example, the Eastern enlargement of the EU, occasional amnesties offered to illegal migrants (e.g., in Spain), rising youth unemployment, and widening socioeconomic inequalities. Not all of the 15 countries participated in all rounds of the survey, but quite a few of them participated in almost all rounds (see Table A13.1 in the Appendix). The total sample has 145,564 observations, composed of native- and foreign-born individuals from diverse origins; in fact, the sample includes 198 different countries of origin.

In order to have a large enough sample for the econometric analysis, we use all available ESS rounds (1–7) during the period 2002–2015. We pool the countries together and over time and include individuals aged 15–65 years at the time of the survey. Given that young people from various origins have been more mobile in Europe in recent years, we pay particular attention to the youth dimension, searching for possible heterogeneities and patterns across various countries of origin. To this end, we create two age bands using 35 years as the cut-off age, whereby individuals are defined as being young if they are aged 35 years or younger. In addition to providing standard summary statistics including everyone, we report additional descriptive information on the youth dimension so as to inspect the differences in outcomes by age group.

Regarding the definition of migrants, an individual is defined as a migrant if his or her country of birth is different from his or her country of residence at the time of the survey. However, this definition of migration, although standard in the literature, can be rather broad because it can also include migrants who arrived as small children and hence would be considered second-generation migrants, which is not the focus of this chapter. Because the focus is mainly on first-generation migrants who move for work, we address this potential issue by limiting the sample to “recent” migrants who migrated within the previous 10 years. In this way, we capture—to a large extent—individuals who recently migrated as adults or youth. Moreover, because there is no particular information on seasonal, circular, or cross-border migration in the data, we are not able to capture such temporary migration here.

Given the focus on Southern and Eastern European migrants, because they have been among the most mobile groups in Europe recently, we create aggregate categories of origins for migrants, in addition to the native peers:5 (1) Southern Europe, which includes individuals from Greece, Italy, Portugal, and Spain; (2) Eastern Europe, which includes individuals from EU10 countries—that is, Bulgaria, Czech Republic, Estonia, Hungary, Latvia, Lithuania, Poland, Romania, Slovakia, and Slovenia; (3) intra-EU, which consists of individuals from other EU countries, excluding Southern and Eastern Europeans; and (4) non-EU, which consists of individuals from countries other than the 28 member states (p.395) of the European Union.6 The main focus is on Southern and Eastern European individuals, but to give a complete picture, residents from non-EU origins as well as the other intra-EU countries are also included. In total, approximately 10.3% of the population in the sample is foreign born of diverse origins.

While carrying out the descriptive analysis, we also run several t-tests (not reported here but available upon request) of mean differences in characteristics across various groups in order to check whether the observed unconditional differences are statistically significant, in which case analysis across groups is justified. Results from these tests point to statistically significant heterogeneities in almost all observed characteristics across diverse origins. Therefore, we distinguish these various subgroups, taking native peers as the reference in the remainder of the econometric analysis.

For a comparative analysis of the socioeconomic characteristics and the labor market integration of various populations, we initially examine the unconditional differences in individual characteristics such as age, gender, household size, marital status, number of children, residential area, and educational attainment, in addition to several labor market indicators, such as employment, unemployment, self-employment, weekly total hours worked in main job (overtime included), contract type (temporary/permanent), and education–occupation mismatch. With regard to mismatch, we mainly have in mind overqualification, referring to individuals who are capable of handling more complex tasks and whose skills are underused, as defined by the Organization for Economic Co-operation and Development (OECD 2012; see also McGuinness, Bergin, and Whelan, this volume).7 Technically, we construct the overqualification indicator based on the definition used by Chiswick and Miller (2010) and Aleksynska and Tritah (2013): Using information on the average years of educational attainment per occupation in each country, an individual is defined to be overqualified if his or her education is one standard deviation above the average within the occupation.8

The different access years for citizens from Eastern Europe to the labor markets of the old member states because of various transitional measures can potentially raise issues when one analyzes migration for work, but it is outside the scope of this chapter to analyze labor market integration incorporating all possible restrictive transitional periods. However, evidence from aggregate data by Akgüç and Beblavý (2015) suggests that there has already been a substantial and continuous migrant flow from Eastern European countries to Western and Northern European countries since the early 1990s. Moreover, taking into account country and time effects in the econometrics analysis partially captures the differential transition periods as well.9

We address the differences in labor market integration by controlling for socioeconomic characteristics and their interactions across different groups analyzed within a multivariate regression framework. For the baseline model, each binary dependent variable (employment, unemployment, self-employment, contract (p.396) type, and overqualification) $Yict$ of individual i in country c at time t is estimated by probit using the following model:10

$Display mathematics$
(13.1)

where X includes dummy variables (ORIic) for five broad origin groups for each individual i in country c (native peers, Southern Europeans, Eastern Europeans, intra-EU, and non-EU migrants); demographic/socioeconomic controls ($Xict$) such as age and age squared, gender, household size, marital status, children, educational attainment in years, and residential area; and country-fixed effects ($ηc$), year effects ($μt$), and a random error term ($εict$). To facilitate the interpretation of the coefficients, all the estimation results with binary variables report the estimated marginal effects of the respective control variable.

For the continuous dependent variable (weekly total hours worked) of individual i in country c at time t, we estimate an ordinary least squares version of Eq. (13.1):

$Display mathematics$
(13.2)

where the same notation as before follows. For self-employment, contract type, hours worked, and overqualification, we add the condition of “being employed.” In this way, we compare, for example, the number of hours worked among employed individuals only and not also among unemployed. In the models, the coefficients of interest are those in front of the origin dummies as well as the youth dummy—where relevant—and they are interpreted as the deviation in the outcomes from the reference population, consisting of native-born and older individuals.

Next, with the aim of exploring heterogeneities in these initial results for different age cohorts by origin, we estimate the previous models by interacting the origin dummies with the youth dummy. This implies adding the term$β1ORIic*YOUTHict$ into the previous equations, where YOUTHict is an indicator of youth (1 if aged 35 years or younger). Furthermore, we explore the gender dimension in the analysis by running similar interaction models as with the youth dimension but replacing the youth dummy by the gender dummy FEMALEict. Finally, we estimate gender gaps across native-born and migrant groups for selected labor market outcomes among young individuals only. This last exercise allows us to explore the potential heterogeneities and vulnerabilities experienced by young migrant women.

With respect to the pooling of data across different countries and over time, as we have elected to do in this chapter, there are both advantages and disadvantages to this exercise. We acknowledge that pooling different destination countries with different economic and welfare-state configurations combined with changes (p.397) over time makes it difficult to interpret the results—especially in a causal way for a particular country. For this reason, we note that outcomes would be likely to differ from one destination to another if countries were analyzed separately (see Spreckelsen et al., this volume). At the same time, pooling helps smooth out heterogeneities between countries and years and provides a comprehensive overview of the general situation that is complementary to the single-country analysis at a point in time or over time. Pooling also boosts the sample size, particularly for migrants. Furthermore, inclusion of country and time effects in models with pooled data—as done in this chapter—takes into account part of the cross-country and period-related heterogeneities. Finally, in order to have representative results both nationally and across countries, we include country and design weights provided by ESS when pooling all countries throughout the empirical analysis.

# 13.4. Descriptive statistics

## 13.4.1. Summary Statistics of Main Variables

Table 13.1 displays the main summary statistics for native peers and recent migrant groups of all age groups in the sample. The female ratio is mainly approximately 50% across various population groups, reaching between 55% and 60% for Eastern European and intra-EU migrants. This finding is consistent with the feminization of migration during recent decades. Migrants tend to live in more urban areas than do native peers. The latter finding might be related to the prediction by Harris and Todaro (1970) that individuals from less developed rural regions are more likely to move to developed urban areas.11 Regarding educational attainment, the numbers suggest that recent migrants from Eastern Europe, followed by those from intra-EU countries, have acquired more years of education compared to native-born individuals. The educational profiles of migrants overall seem to be in line with the human capital theory of migration, which postulates that migrants tend to be relatively well educated notwithstanding differences across different origins.

Table 13.1 Summary statistics of main variables (all age groups)

Native peers

Southern European migrants

Eastern European migrants

Intra-EU migrants

Non-EU migrants

Female

0.515

0.496

0.549

0.597

0.512

(0.500)

(0.500)

(0.497)

(0.492)

(0.500)

Household size

3.050

3.115

3.062

2.800

3.430

(1.377)

(1.394)

(1.423)

(1.330)

(1.619)

Married

0.533

0.688

0.587

0.559

0.623

(0.499)

(0.463)

(0.492)

(0.497)

(0.485)

No. of children

0.803

0.993

0.781

0.815

1.124

(1.067)

(1.094)

(1.028)

(1.096)

(1.284)

Residence in urban area

0.274

0.352

0.362

0.314

0.459

(0.446)

(0.478)

(0.481)

(0.464)

(0.498)

Education (years)

13.21

11.52

13.41

14.36

12.96

(3.837)

(4.984)

(3.538)

(4.213)

(4.500)

Employed

0.643

0.682

0.653

0.639

0.596

(0.479)

(0.466)

(0.476)

(0.480)

(0.491)

Unemployed

0.053

0.062

0.083

0.062

0.099

(0.224)

(0.240)

(0.275)

(0.241)

(0.294)

Self-employment

0.135

0.143

0.114

0.148

0.120

(0.341)

(0.351)

(0.318)

(0.355)

(0.325)

Total hours of work (week)

38.98

38.80

39.46

38.48

39.25

(13.46)

(12.80)

(16.02)

(13.54)

(13.80)

Contract type (temporary)

0.107

0.109

0.170

0.093

0.165

(0.309)

(0.311)

(0.376)

(0.290)

(0.371)

Education–occupation mismatch

0.147

0.152

0.201

0.199

0.223

Overqualified

(0.354)

(0.359)

(0.401)

(0.399)

(0.416)

No. of observations

129,395

1,389

2,011

3,832

8,711

Notes: Means are reported, standard deviations are in parentheses. Only migrants who arrived within the previous 10 years are included. Intra-EU refers to EU countries other than Southern and Eastern Europe.

Source: ESS (2002–2015).

Regarding the labor market variables, the employment rate is approximately two-thirds for all groups, whereas unemployment is approximately 5% or 6%, on average, for native-born individuals, Southern migrants, and intra-EU migrants, and it is higher for Eastern European and non-EU migrants (8%–10%). Self-employment is more common among intra-EU migrants and Southern European migrants. The average number of weekly hours worked is approximately 39 hours for everyone. Regarding contract type, migrants from Eastern European and non-EU countries are more likely to be on temporary contracts compared to the rest of the sample. This could be due to the fact that these groups are younger than the others. At the same time, there has been a general increase in the share of temporary (p.398) contracts since the early 2000s. Therefore, an econometric estimation that controls for sociodemographic characteristics together with time trends can shed light on this finding. Finally, the constructed overqualification indicator suggests that native peers are the least likely to be overqualified in their jobs, whereas non-EU migrants are the most likely to be overqualified. Southern Europeans are relatively similar to native peers in this regard, whereas Eastern Europeans and intra-EU migrants are more likely to be overqualified compared to native peers.

## (p.399) 13.4.2. Further Inspection of Age Structures, Migrant Backgrounds, and Gender Gaps

Regarding the age structure, a comparative report by Eurostat (2011) on the migrant population in Europe suggests that compared to native peers, the foreign-born population is younger and more concentrated in the lower working-age group. The figures from the ESS sample, as displayed in Table 13.2, suggest parallel results. Although the share of native peers aged 35 years or younger is approximately one-third, the numbers jump almost twofold among migrants who arrived within the past 10 years; for example, approximately two out of three migrants from Eastern Europe and non-EU countries are young, whereas slightly more than half of Southern Europeans are young. In line with the youth shares, recent migrants are, on average, much younger than the native-born population (aged 41 vs. aged in their early 30s, respectively).

Table 13.2 Youth shares and average age by country of origin

Native peers

Southern Europe

Eastern Europe

Intra-EU

Non-EU

Youth population share (%; recent migrants only)

33.6

53.5

65.5

45.4

65.1

Average age (years)

41.1

33.6

32.1

36.4

32.3

Source: Authors’ calculations based on ESS (2002–2015).

The youth dimension among migrants is given further inspection in Figure 13.1, which shows the evolution of youth shares among migrants from the main sending regions per survey year. Each column gives the composition of migrants aged 35 years or younger by region of origin. For example, in 2002, the majority of young migrants (almost 70%) were from non-EU countries, whereas less than 10% were from Southern and Eastern Europe combined. In 2009, the total share of young European migrants increased to more than 40%. Moreover, the relative share of young Eastern Europeans has increased significantly since 2008, which is likely due both to the changing economic circumstances brought on by the global recession and to the Eastern enlargement of the EU. Overall, an increasing number of young people of diverse origins seem to be on the move in Europe during the past decade.

Figure 13.1 Distribution of youth share among migrant groups by survey year.

Source: Authors’ calculations based on ESS (2002–2015).

Table 13.3 examines the gender gaps in different age cohorts in general, without distinguishing between migratory origins. To do this, we first estimate the mean gaps in outcome between men—the reference group—and women for a selected set of variables that are closely associated with labor market performance (e.g., educational attainment, employment status, hours worked, contract type, and mismatch indicators). In order to investigate whether gender gaps differ by age structure, we repeat the first step for young individuals younger than age 35 years and for individuals aged 35 years or older, respectively. In this (p.400) way, we get a hint as to how gender gaps in selected outcomes evolve across the life cycle.12 From the results shown in Table 13.3, we observe that young women have significantly more years of education (0.25) compared to young men, whereas the difference goes in the opposite direction among older individuals. Young women are also 8% less likely to be employed compared to young men, and this gap widens to 12% among the older cohorts. In terms of unemployment, women in general (regardless of their age cohort) are 1% less likely than men to be unemployed, which could be explained by the higher inactivity shares among women. The gender gap in self-employment is also in favor of men and widens with age, whereas the gender gap in weekly working hours widens by almost half in favor of men aged 35 years or older. For the remaining outcomes (e.g., contract type and overqualification), the gender gaps remain significant but do not differ across age groups. Without claiming causal relations, the econometric analysis in Section 13.4.3 acknowledges these differences by taking into account sociodemographic and ethnic background as well as variation across countries and time.

Table 13.3 Mean gender gaps in labor market outcomes by age groups

(1)

(2)

35 years or younger

35+ years

Education (years)

0.249***

–0.088***

(8.02)

(–3.39)

Employed

–0.080***

–0.121***

(–18.02)

(–41.59)

Unemployed

–0.010***

–0.009***

(–4.67)

(–6.73)

Self-employment

–0.038***

–0.096***

(–12.83)

(–35.51)

Hours worked (week)

–7.076***

–9.702***

(–46.17)

(–100.47)

Temporary contract

0.03***

0.024***

(6.15)

(12.64)

Overqualified

–0.013***

–0.016***

(–2.87)

(–5.79)

No. of observations

49,068

96,459

Notes: t statistics in parentheses. Reference group is men.

* p < .10.

** p < .05.

(***) p < .01.

Source: ESS (2002–2015).

## 13.4.3. Baseline Estimation Results for Recent Migrants

Table 13.4 reports the baseline results of estimating Eqs. (13.1) and (13.2). By default, we always include the broad origin variables (first column of each outcome variable) and then add the common set of explanatory variables, comprising age, age squared, female dummy, household size, children, education, marital status, and urban dummy (second column of each outcome variable) in order to (p.401) determine whether holding observed characteristics constant modifies the initial effects of origins on the labor market outcomes of interest among native-born individuals and recent migrants. The improvement of the (pseudo/adjusted) R2 when additional explanatory variables are added implies a better fit of the models when the positive influence on this coefficient due to the increase in the number of covariates is taken into account.

Table 13.4 Baseline estimations of labor market performance with full set of control variables

Employment

Unemployment

Self-employment

Temporary contract

Hours of work (weekly)

Overqualified

(1)

(2)

(3)

(4)

(5)

(6)

(7)

(8)

(9)

(10)

(11)

(12)

South

0.030

–0.021

0.014

0.012

0.023

0.022

0.033*

0.046**

–0.436

–0.529

0.020

0.022*

(0.028)

(0.029)

(0.011)

(0.010)

(0.023)

(0.022)

(0.019)

(0.018)

(0.781)

(0.726)

(0.023)

(0.013)

East

–0.026

–0.072***

0.027***

0.021***

–0.030*

–0.000

0.060***

0.042***

1.615

0.679

0.048**

0.030***

(0.020)

(0.021)

(0.007)

(0.007)

(0.018)

(0.018)

(0.014)

(0.013)

(1.385)

(0.762)

(0.019)

(0.009)

Intra-EU

–0.015

–0.078***

0.016

0.017*

0.017

0.022

–0.001

0.004

–0.229

–0.347

0.071***

–0.000

(0.017)

(0.019)

(0.010)

(0.009)

(0.014)

(0.014)

(0.012)

(0.012)

(0.612)

(0.553)

(0.014)

(0.007)

Non-EU

–0.073***

–0.120***

0.039***

0.033***

–0.023***

–0.015*

0.073***

0.064***

0.605*

0.580*

0.075***

0.037***

(0.009)

(0.010)

(0.004)

(0.003)

(0.009)

(0.009)

(0.007)

(0.006)

(0.356)

(0.338)

(0.008)

(0.004)

Age

0.098***

0.007***

0.010***

–0.017***

0.965***

–0.001*

(0.001)

(0.000)

(0.001)

(0.001)

(0.055)

(0.001)

Age squared

–0.001***

–0.000***

–0.000***

0.000***

–0.011***

0.000

(0.000)

(0.000)

(0.000)

(0.000)

(0.001)

(0.000)

Female

–0.153***

–0.004***

–0.072***

0.017***

–9.524***

–0.019***

(0.004)

(0.002)

(0.003)

(0.003)

(0.127)

(0.002)

Household size

–0.006**

0.001

0.006**

0.007***

–0.214

0.003**

(0.003)

(0.001)

(0.003)

(0.002)

(0.139)

(0.001)

Education (years)

0.017***

–0.004***

0.002***

–0.000

0.242***

0.031***

(0.001)

(0.000)

(0.000)

(0.000)

(0.018)

(0.001)

Married

0.055***

–0.034***

–0.006

–0.032***

–0.462***

–0.008***

(0.005)

(0.002)

(0.004)

(0.003)

(0.158)

(0.002)

No. of children

–0.024***

–0.005***

0.001

–0.007***

–0.737***

–0.003*

(0.004)

(0.001)

(0.003)

(0.002)

(0.162)

(0.002)

Living in urban area

–0.019***

0.003

–0.004

0.001

–0.643***

–0.009***

(0.005)

(0.002)

(0.004)

(0.003)

(0.138)

(0.002)

Pseudo R2

0.008

0.198

0.023

0.062

0.024

0.063

0.026

0.113

0.017

0.159

0.008

0.409

No. of observations

140,813

139,641

140,813

139,641

92,543

91,960

92,543

91,960

89,902

89,445

92,226

91,670

Notes: Reference group is native-born individuals. Robust standard errors are in parentheses. Individual controls include age, age squared, gender, household size, education, marital status, children, and urban residence. Only recent migrants who arrived in the destination countries within the previous 10 years are included. Intra-EU refers to EU countries other than Southern and Eastern Europe.

(*) p < .10.

(**) p < .05.

(***) p < .01.

Source: ESS (2002–2015).

The results of the baseline employment regressions before introducing additional controls suggest that there is no significant difference in employment across groups, except for migrants from non-European countries. Once we take into account differences in personal characteristics, however, significant gaps emerge: For example, migrants from Eastern Europe and intra-EU have lower employment levels compared to the native-born population. The explained employment gap between native-born individuals and non-EU migrants rises to 12 percentage points once individual controls are held constant. The change from column 1 to column 2 in Table 13.4 suggests that migrants have characteristics that lead to lower employment compared to native peers. The remaining coefficients in column 2 have expected signs: Age increases employment at a (p.402) (p.403) (p.404) decreasing rate, being female is negatively related to employment, and an additional year of education increases employment. Regarding unemployment, migrants from Eastern Europe and of non-EU origins have higher chances of being unemployed, and adding individual controls does not modify the results to any great extent. In terms of self-employment, Eastern Europeans and non-EU migrants are less likely (although the significance of the coefficient is barely 10% for the former group) to be self-employed compared to native peers; however, this difference almost disappears once individual controls are introduced. Regarding contract duration, most migrants—except for intra-EU migrants—are more likely (to a varying extent by origin) than native peers to hold a temporary job. The estimated gaps in temporary contracts between native-born workers and migrants remain significant even after introducing individual controls.

As seen in the unconditional means from the descriptive statistics, weekly hours of work do not differ across groups in general for the main groups of interest, except for the non-EU migrants, who work slightly more hours than the others. Concerning occupational mismatch, all migrants except Southern Europeans have a higher chance of being overqualified compared to native-born individuals, but this picture changes somewhat once sociodemographic controls are introduced. For example, whereas migrants from intra-EU no longer differ from native peers, Southern Europeans now appear to be overqualified in terms of their educational attainments (although only at 10% significance), together with individuals from Eastern Europe and non-EU countries, even though the extent of mismatch is reduced for the latter origins once control variables are added.

## 13.4.4. Results with Youth Interactions and Gender Gaps Among Youth

Following the baseline estimations, we investigate the labor market outcomes of various migrant groups by distinguishing between different age cohorts in order to obtain insights into the possible vulnerabilities that young people might experience in destination labor markets. To this end, we conduct several additional exercises.13 First, we rerun similar models by adding an interaction term for the youth indicator and the origin dummies (Table 13.5). Next, based on these estimation results with youth interactions, we choose the migrant origins in which we are interested—Eastern and Southern Europe—and conduct a post-estimation mean-differences test (i.e., a t-test) to compare the labor market outcomes of young migrants to those of native-born young people (Table 13.6). We illustrate the results with youth interactions graphically for selected labor market outcomes by origin and by age group, broken down by the 35-year cut-off (see Figure 13.2). Finally, we augment the econometric analysis thus far with the gender dimension by estimating labor market performance models across different migrant origins by gender and by age groups only among individuals (p.405) (p.406) (p.407) (p.408) younger than age 35 years. We display the predicted gender gaps among young people and by origin in selected labor market outcomes (see Figure 13.3)

In Table 13.5, the single coefficients of the origin dummies for Southern and Eastern Europe give the average effect for these groups without distinguishing the age group, whereas the interacted terms with the youth dummy give the effects for young individuals from these regions. Therefore, to obtain the overall effect of being young and being of a particular origin on the outcome variable, we need to add these coefficients together. Before assessing the overall effects, a quick glance at the estimated coefficients suggests that compared to older individuals, young individuals are less likely to be employed or self-employed, are more likely to be unemployed or have temporary job contracts, and are more likely to be overqualified for their occupations.

Table 13.5 Estimations of labor market performance with youth interactions (with full set of controls)

Employment

Unemployment

Self-employment

Temporary contract

Hours (weekly)

Overqualified

(1)

(2)

(3)

(4)

(5)

(6)

South

0.049

0.018

0.003

0.044*

–0.607

0.012

(0.034)

(0.014)

(0.025)

(0.024)

(0.838)

(0.018)

East

–0.076***

0.044***

–0.015

0.052***

–0.115

0.030**

(0.028)

(0.010)

(0.023)

(0.019)

(0.917)

(0.014)

Intra-EU

–0.108***

0.019*

0.015

0.031**

–0.666

–0.006

(0.022)

(0.012)

(0.016)

(0.014)

(0.662)

(0.008)

Non-EU

–0.064***

0.047***

–0.018*

0.069***

0.960**

0.048***

(0.013)

(0.005)

(0.010)

(0.008)

(0.418)

(0.005)

Young (age < 35)

–0.071***

0.016***

–0.082***

0.083***

–1.091***

0.017***

(0.005)

(0.002)

(0.005)

(0.003)

(0.184)

(0.002)

South*Young

–0.038

–0.002

0.081

0.001

1.243

0.030

(0.062)

(0.021)

(0.052)

(0.036)

(1.684)

(0.025)

East*Young

0.119***

–0.036***

0.025

–0.024

2.038

0.002

(0.041)

(0.014)

(0.035)

(0.026)

(1.505)

(0.018)

Intra-EU*Young

0.145***

0.001

0.020

–0.069***

1.389

0.016

(0.039)

(0.019)

(0.032)

(0.023)

(1.187)

(0.014)

Non-EU*Young

–0.006

–0.018***

–0.004

–0.013

–0.683

–0.026***

(0.020)

(0.007)

(0.019)

(0.012)

(0.715)

(0.009)

Female

–0.138***

–0.006***

–0.073***

0.018***

–9.501***

–0.019***

(0.004)

(0.002)

(0.003)

(0.003)

(0.128)

(0.002)

Household size

–0.055***

–0.003***

–0.001

0.018***

–0.778***

0.004***

(0.003)

(0.001)

(0.003)

(0.002)

(0.137)

(0.001)

Education (years)

0.028***

–0.003***

0.002***

–0.001**

0.284***

0.031***

(0.001)

(0.000)

(0.000)

(0.000)

(0.018)

(0.001)

Married

0.056***

–0.035***

0.009**

–0.049***

–0.008

–0.011***

(0.005)

(0.002)

(0.004)

(0.003)

(0.151)

(0.002)

No. of children

0.103***

0.005***

0.004

–0.020***

0.229

–0.003**

(0.003)

(0.001)

(0.003)

(0.002)

(0.152)

(0.002)

Living in urban area

–0.020***

0.003

–0.004

0.000

–0.650***

–0.010***

(0.004)

(0.002)

(0.004)

(0.003)

(0.139)

(0.002)

Individual controls

Yes

Yes

Yes

Yes

Yes

Yes

Year effects

Yes

Yes

Yes

Yes

Yes

Yes

Country-fixed effects

Yes

Yes

Yes

Yes

Yes

Yes

Pseudo R2

0.099

0.048

0.057

0.099

0.149

0.408

No. of observations

139,641

139,641

91,960

91,960

89,445

91,670

Notes: See notes to Table 13.4.

Source: ESS (2002–2015).

In order to determine whether the joint effect of being young and from a particular migratory origin on labor market outcome is statistically significant, we run post-estimation significance tests of linear combinations of the coefficients of the youth and respective origin dummies from Table 13.5. Table 13.6 summarizes these post-estimation test results for young Southern and Eastern European migrants by taking native-born young people as the reference group.14

Table 13.6 Labor market performance differences between native-born youth and young Southern/Eastern migrants

Young Southern European migrants vs. young native-born

Young Eastern European migrants vs. young native-born

Employment

+

+

Unemployment

+

+

Self-employment

+**

+

Temporary contract

+*

+

Hours of work (weekly)

+

+**

Overqualified

+

+***

Notes: The table displays post-estimation t-test results of linear combinations of origin interacted with youth dummies. A plus sign indicates that the respective migrant group has a higher value of the outcome variable compared to native-born. Asterisks indicate the significance level of the t-tests based on conventional notation. No asterisk means nonsignificance of the tested coefficients. Only recent migrants who arrived in the destination countries within the previous 10 years are included in the analysis.

(*) p < .10.

(**) p < .05.

(***) p < .01.

Source: ESS (2002–2015).

The results show that young Southern and Eastern Europeans are not significantly more likely than young native peers to be employed or unemployed. However, young Southern Europeans are more likely than young native peers to be self-employed (which is not a general result for Southern migrants of all ages, as seen in Table 13.4). Finally, the results suggest that young migrants from (p.409) both Eastern and Southern Europe are more likely than young native peers to be overqualified for their occupations, where the gap compared to native peers is statistically significant for Eastern Europeans, in particular.

To further illustrate these results visually, Figure 13.2 shows the predicted probabilities of selected labor market indicators—such as unemployment, self-employment, contract type, and overqualification—across origins for two age bands (cut-off age is 35 years). The graphs in Figure 13.2 are based on the estimated probit models with the full set of controls, and each point in the figure gives the marginal effect of a particular age group on the predicted outcome for a given origin. The top left panel shows that unemployment is generally higher for all young groups of different origins except for Eastern Europe and that the overall predicted unemployment is highest among non-EU migrants. Regarding self-employment, young individuals of all origins have lower predicted self-employment compared to older individuals. Among young people, Southern Europeans have the highest level of self-employment. Similar to self-employment, young individuals of all origins are more likely than older individuals to have a temporary contract, but among the youth of different origins, there is quite a bit of heterogeneity in predicted outcomes. For example, young migrants from non-EU countries and Southern Europe have higher predicted values for having a temporary contract compared to young intra-EU migrants and native-born workers. Finally, younger individuals are generally more likely to be overqualified across all groups, except for non-EU migrants. However, as the post-estimation test from Table 13.4 suggested, the difference is significant mainly for Eastern Europeans.

Figure 13.2 Predicted labor market outcomes by origins and age groups.

Source: ESS (2002–2015).

Finally, we examine the gender gaps among individuals of different origins and aged 35 years or younger for selected labor market outcomes, such as employment, contract type, overqualification, and hours worked per week. We choose the labor market outcomes for which we observed significant (unconditional) gender differences, as reported previously (see Table 13.3). Figure 13.3 is based on the estimation of predicted probabilities for these selected outcomes after including all control variables as before. We see that there is an important gender gap in favor of men in employment and hours worked per week and that this pattern holds for all migrant groups considered. As observed previously for other outcomes, there are also variations in the outcomes among the migrant origins. For example, young women of non-EU origins have the lowest employment and hours worked per week compared to young women of other origins and compared to young men in general. Concerning contract type, we observe that the previous gender-gap patterns are somewhat broken but that they still seem to exist. For example, young migrant men from Southern Europe have a higher probability of having a temporary job compared to their female counterparts of the same origin, whereas the predicted probability of being on a temporary job contract is almost the same for young native-born individuals and for Eastern European (p.410) male and female migrants. Regarding the gender gap in the overqualification outcome, it seems that there is again a slight gender pattern, although this time it is more in favor of women, whereby young men of most origins, including native-born men (except for Southern Europe), are more likely than young women to be overqualified for the jobs they hold.

Figure 13.3 Predicted gender gaps in labor market outcomes by origin of youth.

Source: ESS (2002–2015).

# 13.5. Discussion

Overall, regarding the main groups of interest in the destination countries analyzed, the results from baseline estimations show clearly that migrants from Eastern Europe and non-EU countries (as well as from Southern Europe, to a lesser extent) display important differences in certain labor outcomes, such as employment, unemployment, and overqualification for the occupation held, even after taking into account differences in their socioeconomic characteristics. This comes as a surprise given the strong educational and socioeconomic background of some migrants. At the same time, examining the fit of the models in different columns, we observe that the performance of the model estimation varies across outcomes of interest, whereby the fit of the models for employment, hours worked, and overqualification is better than for the rest.

The finding of a relatively worse labor market performance of migrants compared to native peers is not very new in the literature (Chiswick 1978; Adsera and Chiswick 2007; Jean et al. 2007; Akgüç and Ferrer 2015). Chiswick asserts that the earnings gap between native-born individuals and immigrants in the labor markets narrows the longer the migrant stays in the destination country and that this assimilation period can last for a relatively long time (10–15 years). The fact that we focus our analysis only on recent migrants could partially explain these nativity gaps because it might take a longer time for recent migrants to accumulate country-specific skills. Other reasons behind the persisting gaps between various populations in European destination labor markets could be related to factors not accounted for here, such as individual unobserved heterogeneity, language proficiency gaps, and so on. A further explanation for labor market outcome gaps between native-born individuals and migrants could be related to differential labor market treatment in the form of discrimination.

Regarding the main results with youth interactions, we find that youth generally have worse outcomes in employment, unemployment, contract type, and education–occupation match compared to older cohorts but that these differences are not always significant. This is in line with the findings from the literature pointing to various transitional challenges faced by youth in general (Brzinsky-Fay 2007). Moreover, this differential performance varies by the origin of the young individuals. Our results also suggest that Eastern and Southern (p.411) migrants are more likely than native-born people to be overqualified and that the overqualification of Eastern Europeans seems to be mainly found among young migrants. These findings, again, could be associated with the theses that there is imperfect international skills transferability across countries (Chiswick 2009) or that these young migrants need more time to fully assimilate and accumulate skills that are specific to the destination country so that they can catch up with the native-born individuals (Chiswick 1978). Moreover, we note that because the estimated models are based on pooled data from a number of relatively heterogeneous destination countries with different labor market institutions, welfare systems, and compositions of migrant populations, it is impossible to pin down the exact mechanism explaining why the migrant–native gaps persist in the labor markets.

Regarding the gender dimension in labor market integration among youth migrants, our findings highlight the fact that the gender gaps seem to generally exist among young individuals regarding certain labor outcomes such as employment and hours worked, although some differential patterns are also observed in contract type and occupational mismatch. Moreover, the predicted outcomes also vary by different migratory origins. In summary, various factors—such as different labor market institutions in terms of their flexibility for work–life balance, differences in childcare access, as well as different cultural attitudes toward labor market participation among various migrant groups—could be behind these gender gaps. A comprehensive understanding of the causal mechanisms behind these differences is beyond the scope of this chapter; however, we highlight these gender differences among youth migrants by controlling for various sociodemographic and ethnic backgrounds and by exploiting the variation across countries and time.

# 13.6. Conclusions

Using a microeconometric framework, this chapter examined the labor market integration of recent migrant populations vis-à-vis native-born individuals, with a focus on youth in major European countries that have received important inflows in recent decades. Although the quantitative analysis is carried out including all migratory origins, particular attention is paid to migrants from Southern and Eastern Europe, given that these two regions have been the largest source of young migrants within Europe especially during the past decade. In this vein, examining the recent migration flows from within Europe, Akgüç and Beblavý (2015) point to a shift from Southern Europe to Eastern Europe as an important region of origin. The stock figures suggest, however, that Southern European migrant stocks are still larger than those of Eastern Europeans across many destinations in Europe, such as France, Germany, and the United Kingdom.

(p.412) This chapter focused on youth migrant integration and investigated outcomes following migration because—based on several theories outlined previously—(1) migration is an essential part of a strategic transition in an individual’s life and (2) youth is a particular group with possibly different migration behavior and human capital endowment compared to the rest of the population. With this aim in mind, the microeconometric analysis using individual-level data from the ESS across 15 European countries specifically examined how young migrants differ from older migrants and from native peers, and especially whether young migrants from Southern and Eastern Europe have different labor market outcomes compared to young migrants from the rest of Europe and from outside the EU. The chapter treated migrants as a heterogeneous group and distinguished ethnic origins via broader country clusters. The descriptive analysis highlighted that recent migrants (who arrived within the past 10 years) are, on average, much younger than the native-born population. The findings from the micro-level analysis suggest that migrants from Eastern and Southern Europe show important differences compared to native-born people regarding certain outcomes, such as employment, unemployment, contract type, and overqualification, even after taking into account differences in socioeconomic characteristics such as education, gender, age, and country-fixed and year effects. Furthermore, young migrants from both Eastern and Southern Europe are more likely to be overqualified compared to young native-born workers. These findings imply that individual characteristics explain only part of the differential performance of migrants in the destination-country labor markets. Moreover, we also find important gender gaps in favor of men in employment and hours worked per week and that this pattern holds for all migrant groups considered (and very significantly so for non-EU migrants).

There could be various reasons for the unexplained gaps between different young migrant groups and native peers, such as differential treatment of these groups in destination countries. Regarding the vulnerabilities faced by—especially female—migrants in the labor markets, there is also the issue of their selection into the labor force (and employment), which could lie behind the discrepancies compared to the performance of native-born workers. However, dealing with selection issues, in general, is outside the scope of this chapter and has been left for future research. Last, we note that given the pooled nature of the cross-country data, we can expect different outcomes and findings if the analysis is carried out on a single country; nevertheless, these findings on differential outcomes for migrants in destination-country labor markets call for further research on the underlying channels leading to native–migrant gaps. In this vein, panel data would prove very useful in controlling for unobserved individual heterogeneity.

To tackle issues of persisting native–migrant gaps in labor market performance, policies could be geared toward further integration and nondiscriminatory treatment of foreign-born residents in the destination labor markets. Employers (p.413) could adopt anonymous job applications to avoid discriminatory hiring based on ethnicity. On the education–occupation mismatch issue, better screening and more transparent evaluation schemes could be developed to compare and recognize the degrees, qualifications, and skills possessed by the migrants so that their skills and competences could be put to better use in destination countries. Similarly, mechanisms that facilitate international skill transferability and on-the-job training possibilities could be offered to (young) migrants so as to avoid skill mismatches in occupations. Regarding the persisting gender gap found among migrants, especially in outcomes such as employment and hours worked, policymakers could take a targeted approach, whereby they inform migrant women about existing facilities, such as family-friendly work schedules and access to childcare, depending on the destination-country context and labor market flexibilities.

# Notes

(1) We thank Silvana Weiss, Paweł Kaczmarczyk, and the editors of this volume—Jacqueline O’Reilly, Janine Leschke, Renate Ortlieb, Martin Seeleib-Kaiser, and Paola Villa—for valuable comments and feedback.

(2) We are not able to analyze wages because the data we used contain no information on this point.

(3) Except for countries such as Ireland, Sweden, and the United Kingdom, which opened their labor markets immediately to migrants from the new member states.

(4) See http://ec.europa.eu/social for more information regarding the year when free access to the receiving-country labor markets in the old member states was given to citizens of new member states.

(5) We acknowledge the existence of further heterogeneities among migrants within country-of-origin clusters; however, this compromise is offset by the possibility of getting an overall effect for these broader groups of origin, which still have certain sociodemographic characteristics in common. We leave the more detailed analysis of the peculiarities of migration experiences by specific origins to future research.

(6) In this construction, non-EU also includes Switzerland and Norway; however, given the relatively low emigration rates from these countries compared to the rest of the non-EU, the data are not significantly affected by this inclusion. Moreover, the results are also not sensitive to including these two countries in the intra-EU cluster.

(7) We also estimated models with indicators for underqualification and correct matches; the results are not reported here but are available from the authors upon request.

(p.414)

(8) There may be other ways to define overqualification that take into account migrant niches in certain occupations, where migrants might be overrepresented (see Kacmarczyk and Tyrowicz 2015).

(9) We also ran the analysis dropping the first round of the survey (hence, years 2002 and 2003) so as to account for the first year of the enlargement period, but the results remained substantially the same. Therefore, we decided to use all the survey rounds.

(10) As a robustness check, we estimated the mismatch variables using a multinomial logit specification; the results (available upon request) remain qualitatively unchanged compared to binary probit estimations.

(11) Of course, there could also be network effects, in which the existing migrant networks in urban areas attract further migrants.

(12) Note that we do not observe the same individuals over their exact life cycle in the ESS data set; rather, we observe different cohorts of representative individuals at various cycles in their lives.

(13) We also ran models without native-born individuals and included controls for years since migration, but the results did not change substantially; thus, we present the findings with the full set of population groups.

(14) We note that the comparison of young migrants to older native-born individuals would be a different exercise, which we also performed but have not reported here (available upon request). We also note that these results are based on pooled country estimations and hence might show different patterns if applied to and tested in separate country studies.

References

Bibliography references:

Adsera, Alicia, and Barry R. Chiswick. 2007. “Are There Gender and Country of Origin Differences in Immigrant Labor Market Outcomes Across European Destinations?” Journal of Population Economics 20: 495–526.

Akgüç, Mehtap. 2014. “Do Visas Matter? Labor Market Outcomes of Immigrants in France by Visa Classes at Entry.” TSE Working Paper. Toulouse: Toulouse School of Economics.

Akgüç, Mehtap, and Miroslav Beblavý. 2015. “From South or East? Re-emerging European Migration Patterns and Labor Market Outcomes.” STYLE Working Paper 6.3. Brighton, UK: CROME, University of Brighton. http://www.style-research.eu/publications/working-papers

Akgüç, Mehtap, and Ana Ferrer. 2015. “Educational Attainment and Labor Market Performance: An Analysis of Immigrants in France.” IZA Discussion Paper 8925. Bonn: Institute for the Study of Labor.

Aleksynska, Mariya, and Ahmed Tritah. 2013. “Occupation–Education Mismatch of Immigrant Workers in Europe: Context and Policies.” Economics of Education Review 36: 229–44.

(p.415) Algan, Yann, Christian Dustmann, Albrecht Glitz, and Alan Manning. 2010. “The Economic Situation of First- and Second-Generation Immigrants in France, Germany and the United Kingdom.” Economic Journal 120: 4–30.

Antecol, Heather, Peter Kuhn, and Stephen J. Trejo. 2006. “Assimilation via Prices or Quantities? Sources of Immigrant Earnings Growth in Australia, Canada, and the United States.” Journal of Human Resources 41 (4): 821–40.

Beine, Michel, Frédéric Docquier, and Çağlar Özden. 2011. “Diasporas.” Journal of Development Economics 95 (1): 30–41.

Blau, Francine D., Lawrence M. Kahn, and Kerry L. Papps. 2011. “Gender, Source Country Characteristics, and Labor Market Assimilation Among Immigrants: 1980–2000.” Review of Economics and Statistics 93 (1): 43–58.

Borjas, George J. 1987. “Self-Selection and the Earnings of Immigrants.” American Economic Review 77 (4): 531–53.

Borjas, George J., Stephen G. Bronars, and Stephen J. Trejo. 1992. “Self-Selection and Internal Migration in the United States.” Journal of Urban Economics 32 (2): 159–85.

Bowles, Samuel. 1970. “Migration as Investment: Empirical Tests of the Human Investment Approach to Geographical Mobility.” Review of Economic Statistics 52 (4): 356–62.

Brücker, Herbert, Stella Capuano, and Abdeslam Marfouk. 2013. “Education, Gender and International Migration: Insights from a Panel-Dataset 1980–2010.” Mimeo. Nuremberg: IAB Institute for Employment Research.

Brzinsky-Fay, Christian. 2007. “Lost in Transition? Labour Market Entry Sequences of School Leavers in Europe.” European Sociological Review 23 (4): 409–22.

Castles, Stephen. 1986. “The Guest-Worker in Western Europe: An Obituary.” International Migration Review 20 (4): 761–78.

Castles, Stephen. 2006. “Guestworkers in Europe: A Resurrection?” International Migration Review 40 (4): 741–66.

Chiswick, Barry R. 1978. “The Effect of Americanization on the Earnings of Foreign-Born Men.” Journal of Political Economy 86 (5): 897–922.

Chiswick, Barry R. 1979. “The Economic Progress of Immigrants: Some Apparently Universal Patterns.” In Contemporary Economic Problems, edited by William J. Fellner, 357–99. Washington, DC: American Enterprise Institute.

Chiswick, Barry R. 2009. “The International Transferability of Immigrants’ Human Capital.” Economics of Education Review 28 (2): 162–69.

Chiswick, Barry R., and Paul W. Miller. 2010. “Does the Choice of Reference Levels of Education Matter in the ORU Earnings Equation?” Economics of Education Review 29 (6): 1076–85.

Ciccarone, Giuseppe. 2013. “Geographical Labour Mobility in the Context of the Crisis: Italy.” Ad-hoc Request. Brussels: European Employment Observatory.

(p.416) Clark, Ken, and Stephen Drinkwater. 2008. “The Labour-Market Performance of Recent Migrants.” Oxford Review of Economic Policy 24 (3): 496–517.

Constant, Amelie F., and Elena D’Agosto. 2008. “Where Do the Brainy Italians Go?” IZA Discussion Paper 3325. Bonn: Institute for the Study of Labor.

Constant, Amelie F., and Douglas S. Massey. 2003. “Self-Selection, Earnings and Out-Migration: A Longitudinal Study of Immigrants.” Journal of Population Economics 16: 630–53.

Constant, Amelie F., Olga Nottmeyer, and Klaus F. Zimmermann. 2012. “Cultural Integration in Germany.” In Cultural Integration of Immigrants in Europe, edited by Yann Algan, Alberto Bisin, Alan Manning, and Thierry Verdier, 69–124. Oxford: Oxford University Press.

Constant, Amelie F., and Klaus F. Zimmermann. 2005. “Legal Status at Entry, Economic Performance, and Self-Employment Proclivity: A Bi-National Study of Immigrants. IZA Discussion Paper 1910. Bonn: Institute for the Study of Labor.

Docquier, Frédéric, and Hillel Rapoport. 2012. “Globalization, Brain Drain, and Development.” Journal of Economic Literature 50 (3): 681–730.

Eurostat. 2011. Migrants in Europe: A Statistical Portrait of the First and Second Generation. Luxembourg: Publications Office of the European Union.

Fernández, Raquel, and Alessandra Fogli. 2009. “Culture: An Empirical Investigation of Beliefs, Work, and Fertility.” American Economic Journal: Microeconomics 1 (1): 146–77.

Ferrer, Ana, and W. Craig Riddell. 2008. “Education, Credentials, and Immigrant Earnings.” Canadian Journal of Economics 41 (1): 186–216.

Galgóczi, Béla, and Janine Leschke. 2012. “Intra-EU Labor Migration After Eastern Enlargement and During the Crisis.” ETUI Working Paper 13. Brussels: European Trade Union Institute.

Galgóczi, Béla, Janine Leschke, and Andrew Watt. 2011. “Intra-EU Labor Migration: Flows, Effects and Policy Responses.” ETUI Working Paper 03. Brussels: European Trade Union Institute.

González Gago, Elvira, and Marcelo Segales Kirzner. 2013. “Geographical Labour Mobility in the Context of the Crisis: Spain.” Ad-hoc Request. Brussels: European Employment Observatory Publication.

Greenwood, Michel J., Gary L. Hunt, and John M. McDowell. 1986. “Migration and Employment Change: Empirical Evidence on the Spatial and Temporal Dimensions of the Linkage.” Journal of Regional Science 26 (2): 223–34.

Harris, John R., and Michael P. Todaro. 1970. “Migration, Unemployment and Development: A Two-Sector Analysis.” American Economic Review 60 (1): 125–42.

Izquierdo, Mario, Juan F. Jimeno, and Aitor Lacuesta. 2016. “Spain: From (Massive) Immigration to (Vast) Emigration?” IZA Journal of Development and Migration 5 (10).

(p.417) Jean, Sebastian, Orsetta Causa, Miguel Jimenez, and Isabelle Wanner. 2007. “Migration in OECD Countries: Labour Market Impact and Integration Issues.” OECD Economics Department Working Paper 562. Paris: Organization for Economic Co-operation and Development.

Kacmarczyk, Paweł, and Joanna Tyrowicz. 2015. “Winners and Losers Among Skilled Migrants: The Case of Post-Accession Polish Migrants to the UK.” IZA Discussion Paper 9057. Bonn: Institute for the Study of Labor.

Kahanec, Martin, and Klaus F. Zimmermann. 2010. “Migration in an Enlarged EU: A Challenging Solution?” In Five Years of an Enlarged EU: A Positive Sum Game, edited by Filip Keereman and Istvan Szekely, 63–94. Berlin: Springer.

Martin, Philip. 2009. “The Recession and Migration: Alternative Scenarios.” International Migration Institute Working Paper 13. Oxford: University of Oxford.

Organization for Economic Co-operation and Development (OECD). 2012. Better Skills, Better Lives: A Strategic Approach to Skills Policies. Paris: OECD.

Seeleib-Kaiser, Martin, and Thees Spreckelsen. 2016. “Dimensions of Labour Market Integration Among Young EU Migrant Citizens in the UK.” Barnett Working Paper 01. Oxford: Department of Social Policy and Intervention, University of Oxford.

Sjaastad, Larry A. 1962. “The Costs and Returns of Human Migration.” Journal of Political Economy 70 (5): 80–93.

(p.418)

Appendix

Table A13.1 European Social Survey (2002–2015)

Country

ESS Round

Total sample

1

2

3

4

5

6

7

Austria

o

o

o

7,322

Belgium

10,266

Denmark

8,729

Finland

11,314

France

10,209

Germany

16,294

Ireland

12,435

Italy

o

o

o

2,993

Luxembourg

o

o

o

o

o

2,712

Netherlands

10,678

Norway

9,868

Spain

10,823

Sweden

10,200

Switzerland

9,890

United Kingdom

11,831

Total sample

23,827

23,742

20,901

19,228

18,498

19,774

19,594

145,564

Note: A checkmark indicates that the country was included in the survey round.

Source: ESS (2002–2015; rounds 1–7).

## Notes:

(1) We thank Silvana Weiss, Paweł Kaczmarczyk, and the editors of this volume—Jacqueline O’Reilly, Janine Leschke, Renate Ortlieb, Martin Seeleib-Kaiser, and Paola Villa—for valuable comments and feedback.

(2) We are not able to analyze wages because the data we used contain no information on this point.

(3) Except for countries such as Ireland, Sweden, and the United Kingdom, which opened their labor markets immediately to migrants from the new member states.

(4) See http://ec.europa.eu/social for more information regarding the year when free access to the receiving-country labor markets in the old member states was given to citizens of new member states.

(5) We acknowledge the existence of further heterogeneities among migrants within country-of-origin clusters; however, this compromise is offset by the possibility of getting an overall effect for these broader groups of origin, which still have certain sociodemographic characteristics in common. We leave the more detailed analysis of the peculiarities of migration experiences by specific origins to future research.

(6) In this construction, non-EU also includes Switzerland and Norway; however, given the relatively low emigration rates from these countries compared to the rest of the non-EU, the data are not significantly affected by this inclusion. Moreover, the results are also not sensitive to including these two countries in the intra-EU cluster.

(7) We also estimated models with indicators for underqualification and correct matches; the results are not reported here but are available from the authors upon request.

(8) There may be other ways to define overqualification that take into account migrant niches in certain occupations, where migrants might be overrepresented (see Kacmarczyk and Tyrowicz 2015).

(9) We also ran the analysis dropping the first round of the survey (hence, years 2002 and 2003) so as to account for the first year of the enlargement period, but the results remained substantially the same. Therefore, we decided to use all the survey rounds.

(10) As a robustness check, we estimated the mismatch variables using a multinomial logit specification; the results (available upon request) remain qualitatively unchanged compared to binary probit estimations.

(11) Of course, there could also be network effects, in which the existing migrant networks in urban areas attract further migrants.

(12) Note that we do not observe the same individuals over their exact life cycle in the ESS data set; rather, we observe different cohorts of representative individuals at various cycles in their lives.

(13) We also ran models without native-born individuals and included controls for years since migration, but the results did not change substantially; thus, we present the findings with the full set of population groups.

(14) We note that the comparison of young migrants to older native-born individuals would be a different exercise, which we also performed but have not reported here (available upon request). We also note that these results are based on pooled country estimations and hence might show different patterns if applied to and tested in separate country studies.