- 1 -

EU accession and youth labour mobility from Slovakia: labour market

perspective 1

Lucia Mýtna Kureková *

Slovak Governance Institute, Bratislava and Central European University, Budapest

Prepared for the ECPR General Conference Glasgow, September 3-6, 2014

Abstract

Slovakia persistently faces one of the highest youth unemployment rates among the EU member

states and it also experienced large outflows of youth migrants after its accession to the EU. This

paper analyses the impact of structural factors on propensity of youth migrants to migrate. It

merges macro-level labour market and economic performance indicators at the regional level with

a unique micro-level dataset about migration intentions of graduating students collected at the

height of country’s emigration wave and before the world economic crisis. It evaluates relative

importance of structural, labour market and individual factors in affecting migration choice of

highly educated youth. The analysis shows that labour market conditions and indicators of

structural change at the level of regions are significant predictors of propensity to migrate among

the graduating students, net of regional earnings levels, individual characteristic and personal

perceptions about the ability to find work in the country. The paper links these findings to massive

and harsh adjustments that affected different Slovak regions disproportionately during the process

of transition and contributed to mismatches between the emerging job opportunities and

qualifications of labour force, including the graduating youth. It contributes to understanding the

role of sending country structural factors and broader non-migration policy framework, in addition

to economic conditions, on migration choices of well-educated youth in Europe.

Key words: structural change, youth, migration, development, return, labour market, unemployment,

mismatch

*Contact details: SGI, Gajova 4, 811 09 Bratislava, Email: kurekova_lucia@ceu-budapest.edu

1 Research for this paper was initiated during my doctoral studies at Central European University (CEU), and was

updated and finalized under the FP7 project STYLE, Grant agreement number: Proposal ID: 613256. Project details: http://www.style-research.eu/ I am grateful for comments on earlier versions of this work to: Anil Duman, Bela Greskovits, Kristin Makzsin, Mats Ohlen, and former colleagues at the Political Economy Research Group at CEU. I would like to thank to the Institute for the Study of Family and Labour, Bratislava for providing micro-data for the analysis. All errors remain my own.

- 2 -

Introduction

In the transition from the socialist to market economies, the countries of Central and Eastern Europe

(CEE)2 experienced rampant structural change characterized by the shifts in product and export markets,

sectoral transition and vast labour re-allocation from state to private enterprises and from old to new

sectors. These changes were unprecedented in their speed and scope and necessitated major

restructuring in the process of which skill endowments and the ability of labour to adjust to the new

modern economies became the key elements of a successful transition in economic and political terms

(Bohle and Greskovits 2012; Myant and Drahokoupil 2011). Rich literature emerged to study and to

explain labour market transitions, unemployment, job creation and job destruction in the post-

communist countries (Boeri 2000; Boeri and Terrell 2002; Jurajda and Terrell 2009; Boeri and Scarpetta

1995). In addition to varied effect of restructuring on utilization of labour of different skills, the

transition also had diverse regional effects resulting in growing regional inequalities (Boeri and Scarpetta

1995; Jurajda and Terrell 2009; Heidenreich and Wunder 2008; Heidenreich 2003). An important factor

behind these developments has been the spatial allocation of foreign direct investment (FDI) which

exacerbated unequal regional distribution of wealth within most CEE countries (Medve-Bálint 2014;

Brown, Greskovits, and Kulcsár 2007).

A vast majority of the studies have analysed these labour market adjustment processes as occurring

within countries, while little attention has been given to understanding the role of restructuring in

shaping international labour adjustment and stimulating international labour migration. In light of the

accession of the CEE countries to the European Union in 2004 and an intense outmigration wave that –

though unevenly across the countries - ensued (Kureková 2011a; Kahanec and Zimmermann 2010), the

question of the link between structural change, labour market deficiencies and international migration

remains understudied. Importantly, the post-accession East-West migration has been composed of a

high share of youth migrants, often with tertiary education, who migrated to the English speaking

destination countries which liberalized their labour markets, the UK and Ireland, in particular

(Kaczmarczyk and Okólski 2008; Kureková 2011a; Favell 2008). A majority of young migrants in the UK

and Ireland were mostly employed in low-skilled jobs with a limited degree of upward job mobility and

low earnings (Clark and Drinkwater 2008; Drinkwater, Eade, and Garapich 2009; Voitchovsky 2014;

Johnston, Khattab, and Manley 2014). This has been described as ‘brain overflow’ suggesting that a push

2 I refer to the eight countries that joined the EU in May 2004: Czech Republic, Estonia, Hungary, Latvia, Lithuania,

Poland, Slovakia and Slovenia.

- 3 -

for increased educational attainment among the youth cohorts has not matched well the structure of

the labour demand in the CEE (Kaczmarczyk and Okólski 2008; Galgóczi, Leschke, and Watt 2012; OECD

2012). A labour market slack in the sending countries has been argued to be an important determinant

of the choice of many young university graduates without any working experience to migrate (Kureková

2011a; Kahanec and Mýtna Kureková 2014; OECD 2012). Precise push factors beyond wage differentials,

however, remain to be understudied.

With the onset of the 2008-2009 economic crisis return of the youth migrants started to some CEE

countries (Zaiceva and Zimmermann 2014). The ultimate effects of post-accession migration on the

sending countries will depend on whether the returnees will have acquired (or lost) human but also

financial and social capital during their migration experience. However, evidence about return patterns

of youth migrants to date suggests that due to social and family ties and economic basis provided

through these ties, migrants tend to return to the localities of origin, and not to the more prosperous

urban centres. Returns to depressed labour markets result in the difficulties in integration of returnees

and make positive benefits of labour mobility for sending countries and regions within them rather

limited (Radu and Martin 2010; Barcevičius et al. 2012). 3

This paper thus seeks to investigate the link between structural change, labour market characteristics

and youth labour mobility on the regional level in Slovakia. It evaluates relative importance of structural,

labour market and individual factors in affecting migration choice of highly educated youth. Slovakia is a

useful case for the analysis of the spatial impact of structural change on migration patterns for several

reasons. First, unemployment and youth unemployment have become protracted symptoms of

country’s developmental trajectory. Youth unemployment rate stood at 33% in 2004, has never fallen

below 19% (2008) and by 2013 rose to 34%. Secondly, stark regional inequalities exist between the

Western versus the Central and Eastern regions in the country (Fidrmuc 2004; Pavlinek and Smith 1998;

Sokol 2001) (see also Table 2). Third, Slovakia belongs to the new accession states with a high share of

post-accession migration (Kureková 2011a; Kahanec and Mýtna Kureková 2014; Bahna 2012). The

number of migrants working abroad rose from around 100,000 in 2004 to a peak of 177,000 in 2007,

before sharply falling to around 129,000 by 2009 due to the crisis in the key destination countries, and

3 Studies in lower-income countries tend to find positive impact of migrants on development of region from which

they left. For example, (de Haas 2009) reviews evidence for Morocco and finds that migration and remittances have considerably improved living conditions in Moroccan regions with many migrants. He also documents that the effect is rather heterogenous and many factors, such as gender, socio-ethnic background or structural aspects, mediate the impact of migration and remittances on regional development. It might be that the impact of CEE returnees might be positive in the long run, but immediate positive impact appears partial.

- 4 -

stabilized thereafter (Labour Force Survey, Slovak Statistical Office). Fourth, a further motivation to

study this case is the availability of a unique dataset about migration intentions of graduating students

in 2006 which enables to study structural determinants of youth labour migration at the height of

emigration wave from the country.

The paper contributes to understanding the role of sending country structural factors and a broader

non-migration policy framework, in addition to economic conditions, on migration choices of well-

educated youth within Europe (Czaika and de Haas 2011; de Haas 2011a; de Haas 2011b; Castles 2010).

From a theoretical perspective, it goes beyond prevailing neoclassical approaches to migration and

provides a more nuanced understanding of migration determinants of the youth in high unemployment

country and in a free-mobility context. While the data analysis is carried out at a particular point in time,

conceptually this paper proposes a dynamic perspective on migration, which takes into account past

developmental trajectories on a regional level and studies an independent role of structural factors. The

quantitative approach base on merged macro- and micro-level data models migration as a decision

carried out by individuals embedded in specific environments which shape their opportunities and

constraints.

The paper is structured as follows. Section II reviews literature about migration, structural change and

the role of regional disparities on labour mobility. Section III describes data and analytical approach.

Models and expectations are presented in section IV and results are discussed in Section V. The last

section concludes and evaluates theoretical and policy implications of the study.

Literature review

Theoretical expectations about the effect of economic restructuring on labour flows are not

unidirectional and are closely connected to the global capital flows. The studies of migration based on

the world system theory (Sassen 1990; Silver 2003) and dual labour market theory (Piore 1979) are

based on the debates about interrelation of restructuring, structural change and migration. They

propose that globalization of world markets and the expansion of export manufacturing and export

agriculture lead to a disruption of traditional work structures and mobilize new segments of the

population into regional as well as long-distance migration. The changes in the structure of economy and

structural adjustments generate a pool of migrants while at the same time create new opportunities for

immigrant workers coming into the country. While these dynamics have been tested in other world

- 5 -

regions, developed and developing alike (Piore 1979; Athukorala and Manning 1999; Ruhs and Anderson

2010; de Haas and Vezzoli 2010), they have not been studied in the context of transition economies.4

The studies which investigate migration from the CEE region to the West are typically positioned

in the neo-classical theory of migration and base their estimations on wage differentials and probability

of employment at destination as the main determinants of migration (Bauer and Zimmermann 1999;

Dustmann et al. 2003). A key setback from the perspective of the focus of this paper is that this approach

masks large intra-country variations in labour market performance within the sending countries, which

are likely to produce uneven internal propensity to search for work abroad, and are important to

consider especially from the perspective of the sending countries and the underdeveloped regions within

them. The neoclassical approach has faced critique for ahistorical and simplistic approach to

understanding migration generally, and also in the context of East-West migration (de Haas 2011a;

Kureková 2011a; Kureková 2011b; Massey et al. 1993). A study by Kureková (2011b) attempts to

overcome these limitations and investigates the relationship between structural change, labour market

imbalances and labour migration from the CEE economies during the transition and after their accession

to the EU. It argues that structural change that characterized their transition created different labour

market imbalances across the CEE economies, hence creating different structures of employment and

unemployment and varied risks and opportunities for workers of different demographic and skill profiles.

These imbalances and labour market mismatches in turn induced some workers to seek migration as an

exit option more than others, and led to differences in migration rates and in the composition of

migrants.

In the instances of varied returns to labour and varied tightness of labour markets, the standard

economic theory of migration would anticipate a trend towards equalization of factors of production

4 For example, Athukorala and Manning (1999) and Manning (2002) offer an empirical illustration for the countries

in East Asia on their developmental paths between 1960s and late 1990s. The initial transition in East Asian economies from agriculture to manufacturing generated labour outflows in the earlier stages of industrialization. Facing labor shortages that development and restructuring brought in the early industrializers – Japan, Korea and Taiwan, the countries would initially export capital but essentially had to turn to importing unskilled labour from less developed economies in the region. This demand for mainly unskilled foreign labour was generated by up-skilling, technological upgrading and a change in the attitudes of domestic population to low-skilled and low-paid jobs. The authors note of two additional parallel migration processes. First, some of the countries continued to see outmigration while experiencing migrant inflows, such as Korea or Japan. Outflows from these countries were induced by the export of FDI that required corporate transfers to the countries of investment. A second type of outflows, typical for the Philippines (or New Zealand), were again highly skilled but were not related to FDI flows. The outmigration of professionals and highly skilled was rather driven by imbalances in the supply and demand for skilled workers at home created by well-developed higher education sector but a relatively underdeveloped economy and large wage differentials between the countries in the region (Manning 2002, 377).

- 6 -

and eventual levelling out of wages. The mechanism of intra-country labour market equalization when

labour moves from regions with a lack of work to those with abundant employment opportunities has

been in the Central and Eastern Europe very weak. High transportation costs and rigidity of housing

markets have been the main reasons of generally very low intra-country mobility in all CEE economies

(Paci 2007). Jurajda and Terrell (2009) investigate the intra-country migration in selected CEE

economies more systematically and argue that migration behaviour has differed between high skilled

and low skilled workers because of the fact that opportunity costs of not working are higher for the

highly skilled. According to their argument, social security nets provide disincentives for low skilled to

move in the countries where they have been extensive.

Only a few studies have investigated international migration from (or to) different regions within

the CEE economies. Among the exceptions are Fihel and Okólski (2009) who in their study of Poland

establish that the most strongly affected by the post-accession outflows were the underdeveloped

regions of eastern and southern Poland, especially the smaller towns and villages. Kaczmarczyk and

Okólski (2008) find in a Polish cross-regional comparison that migration loss was the lowest in

Mazowieckie region with the capital city Warsaw and the greatest in Podkarpackie region in the remote

south-east corner of Poland. The propensity to move was higher from less urbanized and more backward

regions, especially before the accession, and was strongly related to the level of economic development

in the regions. The composition of places of origin of Polish migrants altered after 2004: while before the

accession there was a clear prevalence of migrants from rural areas (deprived regions), immediately

after the enlargement there was a rise in migrants from urban centres and the impact of the place of

origin on migration decisions dropped.

Kahanec and Mýtna Kureková (2014) study determinants of migration from Slovakia after the

accession and during the Great Recession. Using LFS data on a pooled cross-section for 2003-2011, they

run a separate model to evaluate migration determinants for the youth aged 15-24. The study confirms

that the peak of youth work migration was in 2006 and 2007, with migration propensities having

subsequently declined. Among the youth, females are 3% more likely to work abroad than males, while

being a young Roma increases the likelihood by a staggering 36%. Having studied health or social work

increases the propensity to migrate by 9%, whereas other fields of study do not have a statistically

significant effect. Young unemployed are pushed out by almost 5% more than their employed

counterparts. Importantly, the region of origin has a very large impact, significantly stronger than among

the general population. Specifically, young people from Prešovský, Košický and Banskobystrický regions

- 7 -

are 43%, 26% and 28% more likely to work abroad than the youth from the Bratislava region in the

analysed period. The study measures the effect of region by regional dummy variables, which do not

reveal underlying labour market aspects contributing to higher mobility of the youth to seek work

abroad from these regions. This paper is more specific in measuring specific structural and labour

market characteristics of the Slovak regions, and their interaction with demographic characteristics and

perceptions about labour market prospects of the graduating students in 2006.

Data and approach

With the available data about intended migration of graduating students, I seek to test how significant is

the net effect of regional-level structural change variables on propensity to migrate from different

regions in Slovakia. I use logistic regression with clustered standard errors on a dataset that merges

macro (regional level) and micro (individual level) data. For two-level data (macro and micro) and in the

context of non-independent errors (heteroscedasticity), regression with clustered standard errors is the

appropriate modelling approach as non-corrected error estimation would yield biased estimates

(Cameron and Miller 2013).

Conceptually, this quantitative approach helps to model migration as a decision that is carried out by

individuals who are embedded in specific socio-economic contexts. The dataset enables testing the

effect of structural level variables on a very specific group of Slovak population – tertiary educated young

people – and by design controls for the variation in age and level of education. The analysis assumes that

the regions, measured at the NUTS3 level, represent micro-economies which have the capacity to affect

migration decisions. This is justified in the context of the administrative changes to the organization of

regional governance structures induced by the EU accession process at the end of 1990s.5

Micro-level data

The survey was collected in May-June 2006 among the graduating students in Slovakia with the

intention of identifying brain drain potential, push factors and barriers of potential mobility (Hanzelová,

5 NUTS regions became the statistical units on the basis of which a number of regional policies are conducted, i.e.

regional development, primary and secondary education, infrastructure projects, cohesion policies, investment promotion, etc.

- 8 -

Kostolná, and Kešelová 2007).6 It covered 16 higher education institutions in Slovakia with the master

level students (graduate level) representing 72.1% and the bachelor level (undergraduate) students

24.2% of the respondents. Different fields of study were widely covered. The total sample consisted of

802 graduates but only 769 cases contain information on all individual level variables used in the

analysis.7 Descriptive statistics is presented in Table 1. Distribution of individuals in the dataset was

uneven across different regions and the region of origin was not used as a sampling criterion - data was

collected on the level of the higher education institutions. 67% of the respondents studied in the region

other than the region of permanent residence. In addition to two demographic controls – gender and

marital status, I use the variable in which the respondents were asked whether they believe there is

work available for them in Slovakia after finishing their studies. This question allows me to proxy the

perceived prospects in domestic labour market, which has been suggested as a potential factor affecting

the decision to migrate (Cielinska 2008).

Table 1: Descriptive statistics

Gender % Region of origin %

Male 59.6 Bratislavský kraj 15.0

Female 40.4 Trnavský kraj 7.3

Marital status Trenciansky kraj 12.4

Single 95.3 Nitriansky kraj 11.3

Married (with children) 4.7 Žilinský kraj 10.3

Do you think there is work available in Slovakia for you after finishing your studies?

Banskobytrický kraj 10.8

No 19.2 Prešovský kraj 19.8

Yes 80.8 Košický kraj 13.3

Macro-level data

Table 2 gives an overview of the macro-level regional performance figures, showing averages between

2001 and 2005 in order to capture average performance in the years before the EU accession and before

the survey was conducted among the graduating students in 2006. The selected regional indicators

operationalize different theoretical propositions that stem from the neoclassical theory of migration

(earnings) on the one hand and the less orthodox migration indicators on the other hand. The two

variables approximating structural change are a) cumulative FDI stock and b) a change in the share of

6 I would like to thank to the Institute for the Research of Family and Labor for making the dataset available to me

for the purposes of this paper. 7 The remaining cases were deleted.

- 9 -

industrial employment. I use cumulative foreign capital stock in 2005 (rather than average flows in 2001-

2005) to capture overall development of investments throughout the transition.8 The change in the

share of industrial employment on overall employment from 2001 to 2005 approximates the degree of

change in employment structure in the economy.

The remaining regional level indicators are related to labour market performance

(unemployment rate), general economic performance (average net nominal earnings, balance of primary

income, social benefits as the share of disposable income) and relative position of the region within the

country (regional GDP as share of country GDP). More specifically, unemployment rate measures the

degree of labour market slack in the regional economy. Regional GDP as a share of the national GDP is a

measure of the distribution of country’s wealth across regions and a measure of inequality. Balance of

primary income is a proxy for the ability of a region to provide employment and generate economic

activity (market income)9 and should be inversely correlated with social benefits as a share of net

disposable income that measures the degree of dependency of a region on the state transfers. The data

reveals great diversity across the eight regions (Table 2).

A correlation matrix of regional level variables provides important insights for deciding which

regional level variables to include in the models (Table 3), as strongly correlated variable would cause

multicollinearity problem. There are very strong and significant relationships between several measures

of general economic performance, relative positioning of regions and some of the structural change

variables: the average net earnings are very strongly and significantly correlated with the balance of

primary income, FDI stock, regional GDP share and social benefits as the share of disposable income. This

suggests that factors like regional earnings levels are strongly affected by the FDI stock and that lower

ability to generate market income (balance of primary income) or higher dependence on state transfers

(social benefits as the share of disposable income) are projected in the average net earnings and vice-

versa. High correlation between these variables means that replacing average earnings with FDI stock

8 Correlation between FDI cumulative stock in 2005 and 2004-2005 average FDI flows (longer time series not

available) is strong and significant (0.961, p<0.01) suggesting that there is not too much of a difference between these variables. 9 Primary incomes are considered to be a very important indicator of regional accounts, as they indicate the ability

of residential households to generate incomes either as entrepreneurs, employees or receivers of property

incomes. Low primary incomes of resident households show dependence of a given region on the support from

state or international bodies. They are a marker of less developed regions, and point to the necessity to introduce

measures oriented towards improvement of working conditions, creation of new working possibilities and support

of entrepreneurial activities.

- 10 -

should generate similar statistical effect on the dependent variable, although these variables measure

different factors. Strong and significant relationship between average earnings and cumulative foreign

direct investment stock in 2005 at the level of regions further implies that transnational capital has not

only been the driver of marketization and structural change but that it is a good predictor of earnings at

the regional level.10 A different measure of structural change – change in the share of industrial

employment - is not correlated with the average earnings. This allows entering this variable into the

analysis together with average earnings.

10

For a convincing empirical evidence on the impact of FDI flows on regional level output in mid-2000s see (Medve-Bálint 2010).

11

Table 2: Regional performance indicators: 2001-2005 average

Average net nominal monthly

earnings (Euro)

2001-2005

Unemployment (%)

2001-2005

Change in industrial

employment 2005/2001

FDI stock (cumulative SKK

mil. 2005)

Balance of primary income

(net, mil SKK) 2001-2005

Social benefits as share of disposable

income 2001-2005

Regional GDP as share on country

GDP (%)

Bratislavský 519.25 4.19 0.97 279,802 155,321 0.16 231.32

Trnavský 399.84 11.11 1.12 24,461 81,875 0.24 102.86

Trenčiansky 374.26 9.67 1.07 20,294 83,873 0.26 90.85

Nitriansky 363.69 17.98 1.02 13,255 93,678 0.28 85.95

Žilinský 374.80 12.96 1.00 25,862 91,617 0.28 81.08

Banskobystrický 363.23 21.59 0.85 10,754 82,164 0.31 82.15

Prešovský 337.66 19.96 0.98 7,086 86,168 0.32 60.19

Košický 402.78 21.67 1.03 35,506 95,159 0.31 89.03

Source: Slovak Statistical Office. FDI cumulative: SARIO (reports about FDI dynamics in each Slovak region).

Table 3: Correlations Matrix

Average earnings

Unemployment Rate

Regional GDP on country

GDP

Balance of net primary

income

Social benefits as share of

disposable income

Change in industrial

employment

FDI stock cumulative

Average Earnings 1 -0.704* 0.975

*** 0.918

** -0.901

** 0.018 0.956

***

Unemployment rate -0.704* 1 -0.744

** -0.609 0.919

** -0.359 -0.690

*

Regional GDP/country GDP 0.975***

-0.744**

1 0.946***

-0.928**

-0.065 0.983***

Balance of net primary income 0.918**

-0.609 0.946***

1 -0.816**

-0.169 0.981***

Social benefits/disposable income -0.901**

0.919**

-0.928**

-0.816**

1 -0.211 -0.872**

Change in industrial employment 0.018 -0.359 -0.065 -0.169 -0.211 1 -0.141

FDI stock cumulative 0.956***

-0.690* 0.983

*** 0.981

*** -0.872

** -0.141 1

Notes: N=8. All data: averages for 2001-2005 period, FDI stock cumulative in 2005. *** - Significant at the 0.01 level, ** - Significant at the 0.05 level, * - Significant at the

0.1 level.

12

Model and expectations

Different model specifications are tested performing logistic regression with clustered standard

errors on a dataset that merged the presented macro-level regional data and micro-level data. The

question on whether the respondent is thinking about looking for work abroad after finishing his/her

studies is used as the dependent variable to proxy the propensity to migrate. Out of the whole

sample, 43.8% of the respondents were not thinking about looking for work abroad, 23.9% were

thinking about looking for work abroad in their field of study and at the university level position,

14.2% in a different field of study but for a university level position and 18.1% in a different field of

study and not for a university level position. I dichotomize this variable as ‘0’ for those not thinking to

look for work abroad (43.7%) and ‘1’ as those thinking to look for work abroad (56.3%).

The goal is to test the significance of regional level variables in addition to average earnings and

controlling for individual level variables. The selection of macro-level regional variables – earnings,

unemployment rate, change in industrial employment and cumulative FDI stock – represent key

variables of different migration theories. Average earnings are the main migration determinant

proposed by the neoclassical theory of migration. Unemployment rate measures labour market

performance of a given region and can be considered a key variable of labour market slack of a given

region. The change in the share of industrial employment on total employment proxies structural

change. Finding the latter two variables significant in addition to or instead of average earnings would

confirm an independent effect of structural change indicators and support the argument that it is not

only wages and wage differentials, but the character of economic opportunities, labour market

conditions and economic structure that are affecting migration choices.

I anticipate that the different socio-economic performance of regions in Slovakia, partly caused by the

process of structural change and partly inherited, should lead to different propensity to migrate. Also,

while young people with tertiary education are likely to compete in the national rather than regional

labour markets, it is reasonable to expect that students’ perceptions about their employment chances

and prospects are strongly shaped by the experience and situation of their parents who are

embedded in specific regional labour markets.

Due to the fact that regression with clustered standard errors is bounded in degrees of freedom by

the number of clusters which is in this case the eight regions in Slovakia, I do not test more than 7

parameters simultaneously. In order to avoid multicollinearity, the models were constructed in a

following way. The first model tests only the individual level variables (M1), while the following

specifications test a number of regional level variables in different combinations (M2-M8). Model 2

(M2) adds average earnings, M3 unemployment rate, M5 change in industrial employment and M7

13

replaces earnings variable with FDI stock. M4, M6 and M8 differ from the previous models in leaving

out the individual level variable about the perceptions about the availability of work in Slovakia in

order to see if the effect of macro-level variables changes if we leave out personal expectations and

perceptions about labour market performance and individual chances. Intercepts were estimated in

all models but are not reported. Table 4 presents the results reporting odds ratios. 11

Results and discussion

Individual level variables show very consistent results across different models. Gender is not

a significant predictor of propensity while being married decreases the odds of migrating

approximately 7 times. The ‘work in Slovakia’ variable, which measures perceived prospects of

finding a job in the Slovak labour market, is a significant predictor of odds of thinking about migrating

for work abroad at 90% level; it decreases the odds of migrating by about 56%. This indicator does

not attain significance in the first two models that do not test macro-indicators or only include

earnings. This seems to suggest that the significance of personal perceptions of labour market

chances are accentuated and brought out in the context of regional macro-performance.

All tested regional level macro-indicators are significant predictors of propensity to migrate

among the graduating students in Slovakia. This is in line with the expectation that the regional

performance significantly shapes the decisions of young people to migrate, although they might not

be necessarily competing in the regional labour market but rather in the national labour market.

Higher regional unemployment rate and a larger change in the share of industrial employment on

total employment increase the odds of thinking about migration after graduating significantly and

strongly. In fact, change in industrial employment is the strongest factor of all tested variables and

increases the odds of thinking about migrating by more than 4 times when controlling for individual

level determinants, unemployment rate and earnings (Model 5) and by nearly 11 times when

controlling for individual level determinants, unemployment rate and FDI stock (Model 7).12

The results show that higher average earnings increase rather than decrease the odds of

migration although the actual effect is small (a euro increase in average net nominal monthly

earnings increases the odds of thinking about migration by about 0.005%). Similar effect – significant

but relatively small – is true for cumulative FDI stock tested in Models 7 and 8. In line with

11

Standard logistic regressions without error correction and with robust standard errors were also carried out; where the individual level predictors gain greater significance, while regional level predictors have lower significance. Parameter estimates remained unchanged. The results are available upon request. 12

Let me emphasize that this is the case even though the students in the sample are from all fields of study, not only technical fields which would be more directly affected by the change in industrial employment.

14

expectations, FDI is a significant predictor of propensity to think about migration but, as was the case

with earnings, not in the expected direction. This seems to suggest that for the potential young

migrants, earnings have a positive and enabling effect on migration rather than a solely ‘pushing’

effect as argued by the neoclassical theory (de Haas 2007; de Haas 2010). It also suggests that

reasons for migration of young and educated people are complex. Their mobility is inspired by

school-to-work transition and the difficulties related to this process. The unexpected direction of the

effect of cumulative FDI stock could in turn imply that although more foreign direct investment

means more employment which should translate into less migration, it is not only the number of jobs

but equally the match between the emerging employment opportunities and skill endowments of

the graduating students that are important.

In sum, the analysis showed that both individual and regional level factors matter in affecting

the propensity to migrate. Marital status, perceptions about the ability to find a job in Slovakia and

the degree of change in industrial employment have the strongest effect on shaping decisions about

seeking work abroad after graduating. The role of one’s perceptions of finding employment in

Slovakia after graduating is strengthened when regional level indicators related to economic

performance are taken into account. In addition to this, there is strong evidence demonstrating that

regional performance in the early 2000s in Slovakia is an outcome of the process of transition and

the degree and the form of structural change, strongly shaped by the dynamics of transnational

capital inflows.

- 15 -

Table 4: Logistic Regression with clustered standard errors, odds ratios

Odds (Exp. B) of thinking about migrating after graduation

Odds ratio M1(B) M2(B) M3(B) M4(B) M5(B) M6(B) M7(B) M8(B)

Individual level

Gender (1= female)

0.967 (0.835)

0.957 (0.789)

0.954 (0.777)

0.967 (0.840)

0.952 (0.770)

0.966 (0.835)

0.958 (0.796)

0.972 (0.861)

Marital status (1= not single)

0.141*** (0.000)

0.141*** (0.000)

0.141*** (0.000)

0.145*** (0.000)

0.139*** (0.000)

0.143*** (0.000)

0.142*** (0.000)

0.147*** (0.000)

Work in Slovakia (1= yes)

0.659 (0.107)

0.652 (0.103)

0.639* (0.084)

0.634* (0.079)

0.634* (0.078)

Regional level

Average net earnings (average 2001-2005)

1.002 *** (0.000)

1.004*** (0.000)

1.004*** (0.000)

1.005*** (0.000)

1.005*** (0.000)

Unemployment rate (average 2001-2005)

1.038*** (0.000)

1.036*** (0.000)

1.050*** (0.000)

1.046*** (0.000)

1.057*** (0.000)

1.053*** (0.000)

Change in industrial employment (2005 over 2000)

4.543*** (0.000)

4.043*** (0.000)

10.84*** (0.000)

9.067*** (0.000)

FDI stock (Cumulative in 2005, SKK mil.)

1.000004*** (0.000)

1.000004*** (0.000)

N 769 769 769 769 769 769 769 769

Log likelihood -512.539 -511.758 -509.565 -512.280 -508.83 -511.649 -508.909 -511.717

Pseudo R square 0.0282 0.0296 0.0338 0.0287 0.0352 0.0299 0.0350 0.0297

Note: p values in parentheses. ***- significant at the 0.01 level; ** - significant at the 0.05 level, *-significant at the 0.1 level.

16

Conclusion

Central and Eastern European post-communist countries have experienced a successful

economic and political transition which culminated in their accession to the European Union in 2004.

The transition process was marked by vast restructuring with negative social and economic impact

resulting in labour market imbalances and structural problems. In several countries, youth labour

markets were particularly affected. The EU accession in effect widened employment opportunities to

the youth and labour mobility was perceived as a mechanism for remedying youth unemployment in

the respective countries. A significant decline in unemployment rates, including youth

unemployment, took place which was partly caused by significant outflows of workers (OECD 2012;

Rutkowski 2007; Galgóczi, Leschke, and Watt 2009).

To date the link between economic restructuring, structural change and international

mobility has not been fully understood. This paper addresses this gap and investigates the link

between structural change, labour market characteristics and youth labour mobility on the regional

level in Slovakia. It evaluates relative importance of structural, labour market and individual factors in

affecting migration choice of highly educated youth in a free mobility context. The analysis shows

that labour market conditions and indicators of the degree of structural change at the level of

regions are significant predictors of propensity to migrate of graduating students, in addition to net

regional earnings, individual level characteristic and personal perceptions about the ability to find

work in the country. This confirms that massive and harsh adjustments that the Eastern and Central

part of Slovakia experienced during the process of transition carried over to the post-accession

migration as the employment potential remained relatively limited or was characterized by the

mismatches between emerging jobs and labour force qualifications. This suggests that structural

change has been affecting migration patterns also of the young graduates that face difficulties in

school-to-work transition.

The findings of this paper have important theoretical and policy-related implications.

Studying specific conditions of localities and their change over time and taking into account wider

range of migration determinants related to labour market conditions and factors such as the match

between employment opportunities and human capital endowments can help us to understand (and

to predict) future migration flows and their composition much better than the oversimplified neo-

classical framework. With respect to earnings, I find that for the graduating students in Slovakia

higher regional earnings seem to enable rather than inhibit migration. This suggests that the crucial

factor is not (higher) wages but job opportunities that match the migrants’ profiles in terms of skill

requirements, preferences or location. The analysis shows that sending countries need to be given

17

more research attention about determinants of migration: in addition to the structures, policies and

institutions in receiving countries, structural conditions in home states are equally important in

helping to understand migration structures, patterns, and changes.13

A finding of the independent impact of structural factors has further policy implications.

Regional inequalities in Slovakia have not evened out and continue to persist, which has implications

on return migrants and their integration trajectories. Due to social and family ties and economic basis

provided through these ties, migrants tend to return to the localities, where they left from. These

often continue to be characterized by depressed labour markets leading to difficulties in integration

of returnees and making positive benefits of labour mobility for sending countries and regions within

them limited (Radu and Martin 2010; Barcevičius et al. 2012). Contrary to the macro-economic

benefits in sending countries shown in a marked decline of youth unemployment rates (OECD 2012;

Rutkowski 2007; Galgóczi, Leschke, and Watt 2009; Pryymachenko and Fregert 2011), micro-level

impact of youth mobility therefore remains blurred. On the one hand migration experience might

generate important individual-level benefits and give signals to employers who value a set of skills

and characteristics that living and working abroad helps to develop. On the other hand, persisting

structural problems in regional labour markets might deter migrants from return or make their

integration into local labour market a difficult and discouraging process. While migrants are often

perceived as bearers of economic development and growth, in the context of persisting structural

problems, they might become discouraged and enter unemployment registries of home states rather

than contribute to development (Schroth 2013; White 2014). Policy focus should therefore consider

return migrants, especially educated youth, as a specific target group with a large potential, which

might however need specific guidance upon return.

13

For an account of the role of sending countries’ non-migration policies and institutions see (Kureková 2013).

18

References

Athukorala, Prema-Chandra, and Chris Manning. 1999. Structural Change and International Migration in East Asia: Adjusting to Labour Scarcity. Oxford University Press.

Bahna, Miloslav. 2012. ‘Intra-EU Migration from Slovakia’. European Societies, July, 1–20. doi:10.1080/14616696.2012.707669.

Barcevičius, Egidijus, Krystyna Iglicka, Daiva Repečkaitė, and Dovilė Žvalionytė. 2012. Labour Mobility within the EU: The Impact of Return Migration. EUROFOUND. http://www.eurofound.europa.eu/publications/htmlfiles/ef1243.htm.

Bauer, Thomas, and Klaus F. Zimmermann. 1999. Assessment of Possible Migration Pressure and Its Labour Market Impact Following EU Enlargement to Central and Eastern Europe. IZA.

Boeri, Tito. 2000. Structural Change, Welfare Systems, and Labour Reallocation: Lessons from the Transition of Formerly Planned Economies. Oxford University Press, USA.

Boeri, Tito, and Stefano Scarpetta. 1995. Regional Dimensions of Unemployment in Central and Eastern Europe and Social Barriers to Restructuring. 17. EUI Working Papers in Economics. Florence.

Boeri, Tito, and Katherine Terrell. 2002. ‘Institutional Determinants of Labor Reallocation in Transition’. The Journal of Economic Perspectives 16 (1): 51–76.

Bohle, Dorothee, and Béla Greskovits. 2012. Capitalist Diversity on Europe’s Periphery. Cornell University Press. http://books.google.com/books?hl=sk&lr=&id=WNESVI4kbHQC&oi=fnd&pg=PR9&dq=bohle+and+greskovits+2012&ots=o9K1hHJ23a&sig=qRzvmkIXy-j79VUHW7YrJetj52g.

Brown, David L., Béla Greskovits, and Lazslo J. Kulcsár. 2007. ‘Leading Sectors and Leading Regions: Economic Restructuring and Regional Inequality in Hungary since 1990’. International Journal of Urban and Regional Research 31 (3): 522–42.

Cameron, A. Colin, and Douglas L. Miller. 2013. ‘A Practitioner’s Guide to Cluster-Robust Inference’. Forthcoming in Journal of Human Resources. http://cameron.econ.ucdavis.edu/research/Cameron_Miller_Cluster_Robust_October152013.pdf.

Castles, Stephen. 2010. ‘Understanding Global Migration: A Social Transformation Perspective’. Journal of Ethnic and Migration Studies 36 (10): 1565–86.

Cielinska, Barbara. 2008. ‘Labour Emigration in Social Debate and Public Opinion (the Case of Bialystok)’. Warsaw: CMR Working Paper, no. 39/97.

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

Czaika, Mathias, and Hein de Haas. 2011. ‘The Effectiveness of Immigration Policies: A Conceptual Review of Empirical Evidence’. IMI/DEMIG Working Paper, International Migration Institute, University of Oxford.

De Haas, Hein. 2007. ‘Turning the Tide? Why Development Will Not Stop Migration’. Development and Change 38 (5): 819–41.

———. 2009. ‘International Migration and Regional Development in Morocco: A Review’. Journal of Ethnic and Migration Studies 35 (10): 1571–93. doi:10.1080/13691830903165808.

———. 2010. ‘Migration and Development: A Theoretical Perspective’. International Migration Review 44 (1): 227–64.

———. 2011a. ‘The Determinants of International Migration’. IMI/DEMIG Working Paper, International Migration Institute, University of Oxford. http://www.devstud.org.uk/aqadmin/media/uploads/4dc3c9cac333a_WP32%20The%20Determinants%20of%20International%20Migration.pdf.

———. 2011b. ‘The Determinants of International Migration: Conceptualizing Policy, Origin and Destination Effects’. International Migration Institute, University of Oxford, Working Paper 32.

19

De Haas, Hein, and Simona Vezzoli. 2010. ‘Migration and Development Lessons from the Mexico-US and Morocco-EU Experiences’. International Migration Institute, University of Oxford, Oxford. http://www.imi.ox.ac.uk/pdfs/migration-et-developpement-lecons-tirees-des-experiences-mexique-etats-unis-et-le-maroc-ue/at_download/file.

Drinkwater, S., J. Eade, and M. Garapich. 2009. ‘Poles Apart? EU Enlargement and the Labour Market Outcomes of Immigrants in the United Kingdom’. International Migration 47 (1): 161–90.

Dustmann, Christian, Maria Casanova, Michael Fertig, Ian Preston, and Christoph M. Schmidt. 2003. The Impact of EU Enlargement on Migration Flows. 25/03. Research Development and Statistics Directorate, Home Office. http://eprints.ucl.ac.uk/14332/.

Favell, A. 2008. ‘The New Face of East–West Migration in Europe’. Journal of Ethnic and Migration Studies 34 (5): 701–16.

Fidrmuc, Jan. 2004. ‘Migration and Regional Adjustment to Asymmetric Shocks in Transition Economies’. Journal of Comparative Economics 32 (2): 230–47.

Fihel, Agnieszka, and Marek Okólski. 2009. ‘Dimensions and Effects of Labor Migration to EU Countries: The Case of Poland.’ In EU Labour Migration since Enlargement: Trends, Impacts and Policies, edited by B. Galgóczi, Janine Leschke, and Andrew Watt. Ashgate Publishing. http://books.google.com/books?hl=en&lr=&id=JubPDZSyS4oC&oi=fnd&pg=PR7&dq=galgoczi,+leschke&ots=-eHfU3Cj6O&sig=wdhMfaSKnG-WwRTBpk_ZMrg-eL0.

Galgóczi, B., J. Leschke, and A. Watt. 2012. Eu Labour Migration in Troubled Times: Skills Mismatch, Return and Policy Responses. Ashgate Publishing.

Galgóczi, B., Janine Leschke, and Andrew Watt. 2009. EU Labour Migration since Enlargement: Trends, Impacts and Policies. Ashgate Publishing.

Hanzelová, Eneke, Zuzana Kostolná, and Daniela Kešelová. 2007. Migračný Potenciál Obyvateľstva SR. Sprístupnenie Trhov Práce vo Vybraných Krajinách EÚ a Vývojové Trendy Na Trhu Práce v SR. (Migration Potential of the Population of the Slovak Republic). Bratislava: Inštitút pre výskum práce a rodiny.

Heidenreich, Martin. 2003. ‘Regional Inequalities in the Enlarged Europe’. Journal of European Social Policy 13 (4): 313–33.

Heidenreich, Martin, and Christoph Wunder. 2008. ‘Patterns of Regional Inequality in the Enlarged Europe’. European Sociological Review 24 (1): 19–36.

Johnston, Ron, Nabil Khattab, and David Manley. 2014. ‘East versus West? Over-Qualification and Earnings among the UK&#039;s European Migrants’. Journal of Ethnic and Migration Studies 0 (0): 1–23. doi:10.1080/1369183X.2014.935308.

Jurajda, Štěpán, and Katherine Terrell. 2009. ‘Regional Unemployment and Human Capital in Transition Economies’. Economics of Transition 17 (2): 241–74. doi:10.1111/j.1468-0351.2009.00351.x.

Kaczmarczyk, P., and M. Okólski. 2008. ‘Demographic and Labour-Market Impacts of Migration on Poland’. Oxford Review of Economic Policy 24 (3): 599–624.

Kahanec, Martin, and Lucia Mýtna Kureková. 2014. ‘Did Post-Enlargement Labor Mobility Help the EU to Adjust during the Great Recession? The Case of Slovakia’. In EU Enlargement and the Labor Markets., edited by Martin Kahanec and K.F. Zimmermann. Springer.

Kahanec, Martin, and K. F. Zimmermann. 2010. EU Labor Markets after Post-Enlargement Migration. Springer.

Kahanec, Martin, Klaus F. Zimmermann, Lucia Kureková, and Costanza Biavaschi. 2013. Report No. 56: Labour Migration from EaP Countries to the EU – Assessment of Costs and Benefits and Proposals for Better Labour Market Matching. IZA Research Reports 56. Institute for the Study of Labor (IZA). http://ideas.repec.org/p/iza/izarrs/56.html.

Kureková, Lucia. 2011a. ‘From Job Search to Skill Search: Political Economy of Labor Migration in Central and Eastern Europe’. Central European University. http://www.etd.ceu.hu/2011/kurekova_lucia.pdf.

20

———. 2011b. ‘The Effects of Structural Factors in Origin Countries on Migration: The Case of Central and Eastern Europe’. IMI Working Paper no. 45. http://papers.ssrn.com/sol3/papers.cfm?abstract_id=1962349.

———. 2013. ‘Welfare Systems as Emigration Factor: Evidence from the New Accession States’. JCMS: Journal of Common Market Studies 55 (4): 721–39. doi:10.1111/jcms.12020.

Manning, Chris. 2002. ‘Structural Change, Economic Crisis and International Labour Migration in East Asia’. The World Economy 25 (3): 359–85.

Massey, Douglas S., Joaquin Arango, Graeme Hugo, Ali Kouaouci, Adela Pellegrino, and J. Edward Taylor. 1993. ‘Theories of International Migration: A Review and Appraisal’. Population and Development Review, 431–66.

Medve-Bálint, Gergő. 2010. ‘Explaining Uneven Regional Development in Post-Socialist Central Europe: The Link between Governmental Policies, Foreign Investments and EU Structural Funds’. In Regional Disparities in Central and Eastern Europe, edited by Marek Radvansky and M.W. Tiruneh, 79–97. Bratislava: Institute of Economic Research, Slovak Academy of Sciences.

———. 2014. ‘The Role of the EU in Shaping FDI Flows to East Central Europe’. JCMS: Journal of Common Market Studies 52 (1): 35–51. doi:10.1111/jcms.12077.

Myant, Martin R., and Jan Drahokoupil. 2011. Transition Economies: Political Economy in Russia, Eastern Europe, and Central Asia. Wiley Hoboken.

OECD. 2012. Free Movement of Workers and Labour Market Adjustment. Recent Experiences from OECD Countries and the European Union. Paris: OECD. http://www.oecd-ilibrary.org/social-issues-migration-health/free-movement-of-workers-and-labour-market-adjustment_9789264177185-en.

Paci, Pierella. 2007. Internal Labor Mobility in Central Europe and the Baltic Region. Vol. 105. World Bank Publications.

Pavlinek, Petr, and Adrian Smith. 1998. ‘Internationalization and Embeddedness in East-Central European Transition: The Contrasting Geographies of Inward Investment in the Czech and Slovak Republics’. Regional Studies 32 (7): 619–38.

Piore, Michael J. 1979. Birds of Passage: Migrant Labor and Industrial Societies. Cambridge: Cambridge University Press.

Pryymachenko, Y., and K. Fregert. 2011. ‘The Effect of Emigration on Unemployment in Source Countries: Evidence from the Central and Eastern European EU Member States’. http://www.nek.lu.se/NEKfng/yana.pdf.

Radu, D., and R. Martin. 2010. ‘Return Migration: The Experience of Eastern Europe’. International Migration. http://westminsterresearch.wmin.ac.uk/8230/.

Ruhs, Martin, and Bridget Anderson. 2010. Who Needs Migrant Workers? Labour Shortages, Immigration, and Public Policy. Oxford Univ Press.

Rutkowski, Jan. 2007. ‘Labor Markets in EU8+2: From the Shortage of Jobs to the Shortage of Skilled Workers’. World Bank EU8+2 Regular Economic Report, Part II. http://papers.ssrn.com/sol3/papers.cfm?abstract_id=1078895.

Sassen, Saskia. 1990. The Mobility of Labor and Capital: A Study in International Investment and Labor Flow. Cambridge University Press.

Schroth, Jan. 2013. International Comparative Working Paper ‘Barriers for Successful Capitalization on Returning Migrants’. Prague Office: IOM. http://www.re-migrants.eu/download/International%20Comparative%20Working%20Paper%20on%20Barriers.PDF.

Silver, Beverly J. 2003. Forces of Labor: Workers’ Movements and Globalization since 1870. Cambridge University Press.

Sokol, Martin. 2001. ‘Central and Eastern Europe a Decade after the Fall of State-Socialism: Regional Dimensions of Transition Processes’. Regional Studies 35 (7): 645–55.

Voitchovsky, Sarah. 2014. ‘Occupational Downgrading and Wages of New Member States Immigrants to Ireland’. International Migration Review, no. 48: 500–537. doi:10.1111/imre.12089.

21

White, Anne. 2014. ‘Polish Return and Double Return Migration’. Europe-Asia Studies 66 (1): 25–49. doi:10.1080/09668136.2013.855021.

Zaiceva, Anzelika, and Klaus F. Zimmermann. 2014. ‘Returning Home at Times of Trouble? Return Migration of EU Enlargement Migrants during the Crisis’. In Migration and the Great Recession: Adjustments in the Labour Market of an Enlarged European Community’, edited by Martin Kahanec and Klaus F. Zimmermann. Springer Verlag.