(Martin Kahanec, Klaus F. Zimmermann) EU Labor PDF

EU Labor Markets After
Post-Enlargement Migration
Martin Kahanec • Klaus F. Zimmermann
Editors
EU Labor Markets After

Post-Enlargement Migration
Editors
Dr. Martin Kahanec
Prof. Dr. Klaus F. Zimmermann
Institute for the Study of Labor (IZA)
Schaumburg-Lippe-Str. 5–9
53113 Bonn
Germany
kahanec@iza.org
zimmermann@iza.org
ISBN 978-3-642-02241-8 e-ISBN 978-3-642-02242-5

DOI 10.1007/978-3-642-02242-5
Springer Heidelberg Dordrecht London New York
Library of Congress Control Number: 2009937939
© Springer-Verlag Berlin Heidelberg 2010

This work is subject to copyright. All rights are reserved, whether the whole or part of the material is
concerned, specifically the rights of translation, reprinting, reuse of illustrations, recitation, broadcasting,
reproduction on microfilm or in any other way, and storage in data banks. Duplication of this publication
or parts thereof is permitted only under the provisions of the German Copyright Law of September 9, 1965,
in its current version, and permission for use must always be obtained from Springer. Violations are
liable to prosecution under the German Copyright Law.
The use of general descriptive names, registered names, trademarks, etc. in this publication does not
imply, even in the absence of a specific statement, that such names are exempt from the relevant protective
laws and regulations and therefore free for general use.
Typesetting: Thomas Wasilewski, IZA, Bonn
Cover design: WMXDesign GmbH, Heidelberg, Germany
Printed on acid-free paper
Springer is part of Springer Science+Business Media (www.springer.com)

Preface
This volume results from a continuing dedication of IZA’s Migration Program

Area at the forefront of research on the broad issues of international and internal
migration. Among the main themes covered from both the scientific and policy
perspectives are the impact of migration on the host and source labor markets, the
adjustment of migrants and their descendants in the destination countries, and the
role of ethnic identity in these processes.
The 2004 enlargement of the European Union, involving Cyprus and Malta and
eight Central and Eastern European countries: the Czech Republic, Estonia, Lat-
via, Lithuania, Hungary, Poland, Slovakia and Slovenia; and the 2007 enlarge-
ment, when the European Union opened up to Bulgaria and Romania, have con-
siderably changed the migration landscape in Europe. Within the migration
research program, the sub-area “EU Enlargement and the Labor Markets” ad-
dresses the current challenges and opportunities which accompany free mobility in
an enlarged European Union.
The migration research group at IZA produces numerous discussion papers,

policy papers and journal articles, research reports, monographs, edited volumes
and other publications each year. Cooperating with a network of more than 800
IZA Research and Policy Fellows and Affiliates and holding regular forums for
scientific exchange, including the Annual Migration Meeting, the biennial Migra-
tion Topic Week, and workshops on EU Enlargement and the Labor Markets, the
group has also become an international hub for migration research.
We wish to express our most sincere thanks to the contributors of this volume
for their persevering commitment and thought-provoking chapters. We are very
grateful to Barry R. Chiswick and Amelie F. Constant for their encouragement
and profound inspiration throughout this project, as well as their dedicated leader-
ship within the Migration Program Area. We also thank Anzelika Zaiceva, Holger
Hinte, Ross McRae, Andreas Kruepe, Thomas Wasilewski, and all the others who
have provided invaluable support at various stages of producing this volume. We
also acknowledge the continuous support and cooperation from our publishers at
Springer.
Bonn, May 2009
Martin Kahanec Klaus F. Zimmermann
v
Contents
Part I:
EU Enlargement and the Labor Markets: What Do We Know?
1 Lessons from Migration after EU Enlargement ...................................3

Martin Kahanec, Anzelika Zaiceva, Klaus F. Zimmermann
2 Labor Mobility in the Enlarged EU:

Who Wins, Who Loses? ........................................................................47
Timo Baas, Herbert Brücker, Andreas Hauptmann
3 Post-Enlargement Migration and Public Perception

in the European Union ..........................................................................71
Marcel Canoy, Anna Horvath, Agnès Hubert, Frédéric Lerais,
Myriam Sochacki
Part II:
The Impact of Migration on Destination Labor Markets
4 EU Enlargement under Continued Mobility Restrictions:

Consequences for the German Labor Market ..................................111
Karl Brenke, Mutlu Yuksel, Klaus F. Zimmermann
5 The Experience of Spain with the Inflows

of New Labor Migrants.......................................................................131
Sara de la Rica
6 EU Enlargement and Ireland’s Labor Market .................................145

Alan Barrett
7 Post-Enlargement Migration
and Labor Market Impact in Sweden................................................163
Christer Gerdes, Eskil Wadensjö
8 The Impact of the Recent Expansion of the EU

on the UK Labor Market ....................................................................181
David G. Blanchflower, Helen Lawton
vii
viii Contents
Part III:
The Impact of Migration on Source Labor Markets
9 Impact of the Post-Accession Migration

on the Polish Labor Market................................................................219
Paweł Kaczmarczyk, Marta Mioduszewska, Anna Żylicz
10 The Post-Enlargement Migration Experience

in the Baltic Labor Markets ...............................................................255
Mihails Hazans, Kaia Philips
11 The Case of Albania ............................................................................305

Zvezda Dermendzhieva, Randall K. Filer
Editors and Authors ..........................................................................................335
Index ...................................................................................................................343
1. Lessons from Migration after EU Enlargement
Martin Kahanec
Anzelika Zaiceva
Klaus F. Zimmermann
The Eastern enlargement of the EU was an institutional impetus to the migration

potential in Europe. While the overall numbers of migrants from the new member
states in the EU15 increased between 2003 and 2007, this increase was distrib-
uted unevenly among countries. The proportion of these migrants in the EU15 re-
mains smaller than that of non-EU27 migrants. The transitory arrangements may
have diverted some migrants from the EU8 mainly to Ireland and the UK. Mi-
grants from the EU2 continued to go predominantly to Italy and Spain. To date,
there is no evidence that these primarily economic migrants would displace native
workers or lower their wages (and even if crowding out happened in certain sec-
tors or occupation, aggregate data suggest that such natives found well-paid jobs
elsewhere), or that they would be more dependent on welfare than the natives. The
drain of mainly young and skilled people could pose some additional demo-
graphic challenges on the source countries. However, the anticipated brain circu-
lation may in fact help to solve their demographic and economic problems. While
the ongoing economic crisis may change the momentum of several migration tra-
jectories, free migration should in fact alleviate many consequences of the crisis
and generally improve the allocative efficiency of EU labor markets.
M. Kahanec and K.F. Zimmermann (eds.), EU Labor Markets After Post-Enlargement 3

Migration, DOI 10.1007/978-3-642-02242-5_1,
4 Martin Kahanec, Anzelika Zaiceva and Klaus F. Zimmermann
1 Introduction
What once was unimaginable, then a vision, turned into reality, when eight Cen-
tral and Eastern European countries1 (EU8), together with Cyprus and Malta, and
Bulgaria and Romania (EU2), joined the European Union in May 2004 and Janu-
ary 2007, respectively. While cheered by many, EU enlargement brought about a
number of concerns as well. The change it impinged on the European migration
landscape was unprecedented in many aspects: the population size of the acceding
countries was large; the differences in income between the old EU member states2
and the EU8 and the EU2 were substantial; essentially no migration between the
former Soviet bloc and the West was allowed during the decades of separation by
the “Iron Curtain”; and the new member states underwent a complex process of
societal transformation to a free society and a market-based economy prior to their
EU accession.
These specific circumstances partly explain the sensitivity of the migration is-
sue among the general public as well as policy makers across Europe, who feared
economic, social, cultural and political consequences of free east-west migration
in an enlarged EU. Competition in the labor markets and for welfare benefits has
prominently driven these fears. In contradiction of the European Union’s funda-
mental principle of free movement,3 these fears have materialized as transitional
periods of up to seven years, which restrict access of citizens from the EU8 and
EU2 to the labor markets in the old member states. These transitional arrange-
ments are based on the so-called “2+3+2” formula: for the first two years follow-
ing accession, access to the labor markets of the incumbent member states de-
pends on their national laws and policies. National measures may be extended for
a further period of three years and could continue for a further two years, but only
if there are serious disruptions in the respective receiving labor market.
Initially, only a few old member states opened their labor markets with no or
mild transitional measures. Following the 2004 EU enlargement, Ireland, the UK
and Sweden opened access to their labor markets immediately.4 Gradually, more
1 Including the Czech Republic, Estonia, Latvia, Lithuania, Hungary, Poland, Slovakia and
Slovenia. We use EU10 whenever we refer to Cyprus and Malta as well.
2 The old member states (EU15) in the context of these enlargements include Austria, Bel-
gium, Denmark, Finland, France, Germany, Greece, Ireland, Italy, Luxembourg, the Neth-
erlands, Portugal, Spain, Sweden and the United Kingdom. EU25 denotes EU15 and
EU10; EU27 includes EU2 as well.
3 Article 39 of the Treaty establishing the European Community entitles nationals of one EU
member state to work in another EU member state under the same conditions as that
member state’s own citizens.
4 Immigrants from the EU8 who wish to work in the UK have to register with the Home Of-
fice administered Worker Registration Scheme if they are employed in the UK for a month
or more. This requirement allows the authorities to monitor immigration and its impact on
1. Lessons from Migration after EU Enlargement 5
member states decided to lift restrictions, including Greece, Spain, Portugal and
Finland (all May 1, 2006); Italy (July 27, 2006); the Netherlands (May 1, 2007);
Luxembourg (November 1, 2007); France (July 1, 2008); and Belgium and Den-
mark (both May 1, 2009); whereas Germany and Austria have announced they
will continue to impose restrictions on labor mobility until 2011, although Ger-
many had passed legislation in late 2008 facilitating access for high-skilled mi-
grants.5 As for the 2007 enlargement, ten EU25 member states (the Czech Repub-
lic, Estonia, Cyprus, Latvia, Lithuania, Poland, Slovenia, Slovakia, Finland and
Sweden) liberalized the access of Bulgarian and Romanian workers to their labor
markets during the first phase.6 The second phase for the EU2 countries started on
January 1, 2009, and following the Commission’s report, Spain, Greece, Hungary
and Portugal have also opened their labor markets to the EU2 nationals. Denmark
granted free access on May 1, 2009. Most of the remaining EU25 Member States
have simplified their procedures or have reduced restrictions in some sectors or
professions.7
Between 2003 and 2007, there was an increase in the number of immigrants
from the new member states in the EU15. For EU8 migrants, the increase was par-
ticularly high in Ireland and the United Kingdom, while EU2 migrants continued
to go to Spain and Italy. Such a concentration of migrant flows into some old
member states may be an artifact of the variation in transitional arrangements, but
also other important factors, such as geographic, linguistic or cultural distances
between receiving and sending countries, as well as existing migrant networks.
Indeed, the early evidence that we discuss in section 3 shows that there is no sim-
ple link between the scale of migration and the transitional arrangements. Overall,
however, the proportion of non-EU27 nationals in the EU15 remains larger than
that of migrants from the new member states.
the British labor market. Individual identification in the form of a Personal Public Service
Number is required in order to gain employment or access to state benefits and public
services in Ireland. Access to the welfare systems and entitlement to social benefits in Ire-
land and the UK depends on the duration of residence and employment. Sweden decided
to apply European Community rules.
5 In April 2009, the German Federal Government notified the European Commission that
Germany will prolong the mobility restrictions in place since 2004. However, federal legis-
lators only recently passed the "Arbeitsmigrationssteuerungsgesetz" (law on the regulation
of labor migration, in effect since 1 January 2009), including rules for free mobility and la-
bor market access of high qualified EU10 citizens and their family members. This legisla-
tion adds to an inconsistent German migration policy. Nonetheless, its potential strategic
relevance should not be underestimated.
http://www.bgblportal.de/BGBL/bgbl1f/bgbl108s2210.pdf.
6 In Finland, Cyprus and Slovenia, employment must subsequently be registered for moni-
toring purposes.
7 As of May 2009, see http://ec.europa.eu/social/main.jsp?catId=466&langId=en
The impact of these migration flows on the labor markets of destination and
source countries depends primarily on their magnitude, composition, temporal na-
ture, as well as the overall functioning of the countries’ economies. The main ob-
jective of this chapter is to broadly map what we know about the determinants and
effects of the post-enlargement migration flows. Namely, we (i) discuss the theo-
retical underpinnings of migration causes and effects; (ii) describe the post-
enlargement migration flows; and (iii) investigate their impact on the receiving
and sending countries including the impact on the European growth potential.
Rather then discussing the potential repercussions of the ongoing economic and
financial sector crisis on migration flows and their effects in Europe throughout
the chapter, we devote a special section to this topic. Finally, we identify the main
scientific and policy challenges concerning migration in the enlarged EU, and re-
flect on how the chapters of this book contribute to our understanding of these
challenges.
2 Theoretical Framework:
Causes and Impacts of Migration
Early theories of migration stress the significance of (expected) regional dispari-

ties in prosperity for migration decisions (Harris and Todaro, 1970). In general,
migration theories imply the significance of international differentials, net of mi-
gration costs, in earnings and income levels; costs of living; unemployment rates;
the quality of public goods; and the generosity of the welfare systems. Other fac-
tors, such as age or skills, affect individual potential to benefit from migration or
ability to cope with the adjustment into the host society, as evident from the hu-
man capital theory (Becker, 1957; Sjaastad, 1962). The cultural, linguistic and
geographical distances between pairs of source and destination countries directly
affect not only the respective pecuniary, but also the psychological and social
costs of migration or adjustment. Having a child or spouse, or broader social rela-
tionships, such as ethnic networks, may also play a significant role (Mincer, 1978;
Massey, 1990). Migration can also serve as a risk-sharing strategy of a household
as is emphasized by the new economics of migration theories that view house-
holds rather than individuals as decision-making units (Stark, 1991). Other factors
mentioned in the literature include the earnings distribution (Borjas, 1985) and
welfare regimes (Borjas, 1999; De Giorgi and Pellizzari, 2006) in the receiving
and sending countries. Overall, migrants may be positively or negatively self-
selected with respect to their observable and unobservable characteristics, both
upon entry and exit (Borjas, 1987; Chiswick, 1999). Finally, the option of waiting
and not migrating may have a positive value in the presence of uncertainty and ir-
reversible costs of moving (Burda, 1995).
The impact of immigration on the destination labor market has been modeled
by a number of studies, including Chiswick et al. (1992) and Chiswick (1980,
1998). In these models, the effects of migration on receiving labor markets depend
on the substitutability or complementarity of migrant and native labor. We illus-
trate these effects in a simple theoretical setting in Figure 1. Consider the impact
of high-skilled immigration on an economy with labor markets for two types of
native workers: high-skilled and low-skilled.8 The rise in the stock of high-skilled
workers (see the shift in the supply curve from S 0 to S 1 in Figure 1a) moves the
equilibrium point down from C0 to C1 at a lower wage w1h . Due to the complemen-
tarity of high- and low-skilled labor, the demand for low-skilled workers increases
(see the shift of the demand curve in Figure 1b). Under a competitive market,
wages rise from wl0 to wl1 at the full employment level L . Under the regime with
rigid union wage ( wU ) that is binding for the wages of low-skilled workers, as is
the case in many European countries, the demand increase generates a higher level
of employment of low-skilled workers instead ( L0U to L1U ). This increase in low-
skilled employment feeds back into the market for high-skilled labor and causes
an upward shift in the demand for high-skilled workers, counteracting the original
wage decline from w1h to wh2 .9
Figure 1a, 1b
Immigration in an economy with heterogeneous labor
8 Whether particular migrants can be regarded as skilled or unskilled relative to the natives
also depends on the transferability of their skills acquired in the country of origin as well as
their ability and willingness to adjust to the skill requirements of the host labor market. See
the discussion in Kahanec and Zimmermann (2009).
9 The degree of complementarity determines whether the resulting wage is higher or lower
than w0 . In the Figure, partial counteraction is shown.
h
By a similar argument, low-skilled immigration decreases the wages to wl2

(under competitive markets) or increases unemployment from L0 − L0U to L1 − L0U
(under a rigid union wage) in the market for low-skilled labor. If the union wage is
lowered in response to the increased unemployment, employment of low-skilled
workers goes up. For example, if it is lowered to the pre-immigration competitive
level wl0 , employment will go from L0U to L0 , leaving L1 − L0 unemployed. This in-
creases demand in the market for high-skilled labor due to the complementarity of
high- and low-skilled workers in production, which increases wages of high-
skilled workers to wh3 .
This simple analysis has a straightforward message: skilled immigration bene-

fits unskilled and may hurt skilled native workers; and unskilled immigration
hurts unskilled and may benefit skilled native labor.10 Broader positive effects of
immigration may, however, overwhelm the potential negative effects that may
concern some of the native workers. Highly mobile economic migrants typically
improve the allocation of production factors, most notably human capital. The mi-
grants often act as agents of knowledge transfer and international trade, and pools
of skilled immigrants may attract high-tech investments (Bonin et al., 2008). In ef-
fect, immigration may expand a host country’s production possibility frontier and
thus increase the demand for native labor. On the other hand, some argue that im-
migrants may have unobserved characteristics that make them more likely to
choose to live in countries with more generous welfare benefits and put pressure
on their welfare systems. Furthermore, language problems, psychological trauma,
employers’ discriminatory practices, or legislative and institutional factors that in-
hibit adjustment and economic outcomes of immigrants may make them more de-
pendent on welfare (Borjas, 1999; Brücker et al., 2002). If so, immigration could
put pressure on the tax burden of native workers.
The effects in the source countries can be easily traced in the framework set up
in Figure 1. It is straightforward to see that high-skilled emigration increases
wages in the high-skilled market and via complementarities and the ensuing de-
mand decrease reduces the wage (in a competitive market) or increases unem-
ployment (under a union wage) in the low-skilled market. In the latter case, the
decreased employment in the low-skilled market feeds back into the high-skilled
market reducing demand and thus wages there. If it is the low-skilled who leave,
wages in the low-skilled market increase (under a competitive market) or unem-
ployment goes down (under a union wage). In the latter case, complementarities
imply negative effects on wages in the high-skilled sector. Again, these direct ef-
fects may be mitigated or even reversed if indirect effects are taken into account.
For example, if workers start returning to their home countries and circular migra-
10 Kahanec and Zimmermann (2008, 2009) analyze the ensuing consequences for inequality
in a theoretical and empirical framework.
tion emerges, the human capital they acquired while abroad would constitute brain
gain for the source countries.
3 Post-Enlargement Migration
a. Actual Migration Trends and Trajectories
To provide an updated picture of post-enlargement migration from the new mem-

ber states is a formidable task, as statistical data is continually being updated and
is often missing. Several countries do not report migrants stocks or flows in their
statistics. Migrants are often underrepresented in the official data. Moreover, the
definition of a migrant differs across countries, making cross-country comparisons
difficult. The distinction that is often ignored is between the foreign-born and for-
eign citizens, and between temporary and permanent migrants. Finally, illegal mi-
gration is not accounted for in the official data. In fact, much of the observed mi-
gration flows after enlargement may have been the legalization of people
originating from the new member states who already lived in the old member
states. Despite these difficulties, this section presents the statistics on the post-
enlargement migration trends in Europe making use of the latest and best available
data and reviews evidence from selected recent studies.11
Looking first at the stocks of people from the new member states shows that
the number of foreign residents from the EU8 in the old EU15 increased from al-
most 893,000 in 2003 to more than 1.91 million, or 0.5% of the population of the
EU15 by the end of 2007 (Brücker and Damelang, 2009; Brücker et al., 2009).
That suggests an average inflow (net immigration) of around 250,000 people from
the EU8 per year since 2004, compared to 62,000 per year between 2000 and
2003. The proportion of foreign residents from Bulgaria and Romania have been
continually increasing since 2000, and in 2007 amounted to around 1.86 million
people, or 0.5% of the EU15 population. Similarly, the European Commission
suggests that the proportion of nationals from the new member states have in-
creased, and in 2007 nationals from EU10 and EU2 amounted to 0.5 and 0.4-0.5%
of the EU15 population, respectively (European Commission, 2008a, b). The cor-
responding figure for 2003 was 0.2% for each group. However, these reports also
suggest that the overall increase in immigration after 2004 was rather limited
11 A separate strand of the literature investigated the migration potential, analyzing either mi-
gration intentions in the accession countries or using extrapolations or forecasts based on
historical data for other countries. For the reviews of these studies see, for instance, Alva-
rez-Plata et al. (2003), Boeri and Brücker (2005), Zaiceva and Zimmermann (2008),
Brücker et al. (2009).
when compared with the population sizes of both sending and receiving countries,
and in the case of EU10 migrants was mainly attributable to the increases in Ire-
land, the United Kingdom and also Austria (European Commission, 2006a, b,
2008a, b). For sake of comparison, nationals from other EU15 countries consti-
tuted 1.7% of the EU15 population in 2007 (1.6% in 2003); while non-EU27 na-
tionals amounted to 4.5% (3.7% in 2003) (European Commission 2008a, b). Fig-
ures 2 and 3 document these trends for the available origins and destinations.
Although the statistics on stocks of residents usually rely on registers of foreign

population, which tend to understate return migration, since no incentives to de-
register exist, they provide several interesting observations. First, they show that
the largest increase in the share of EU8 residents since the 2004 enlargement was
in Ireland and the UK. Indeed, almost 70% of the immigrants from the EU8 have
been absorbed by the UK and Ireland since 2003, and their share in the total num-
ber of foreign residents from the EU8 in the EU15 increased between 2002 and
2007; while the share of the traditional recipients, Austria and Germany, declined
(Brücker et al., 2009; Brücker and Damelang, 2009; European Commission,
2008a, b). Thus, it is likely that a “migration diversion” has occurred into the
countries that opened up their labor markets early (Brücker et al., 2009, Boeri and
Brücker, 2005). Otherwise, network effects and geographical proximity would
probably have attracted many more migrants to Austria and Germany.
However, other factors may also have contributed to the unexpectedly strong
rise in the numbers of Eastern European immigrants in the English speaking EU15
countries: many immigrants have studied English at home and valued the oppor-
tunity to use, improve or study it in these countries (Anderson et al., 2006);
enlargement may have contributed to the legalization of previously illegal immi-
grants already in the country, which would imply a pure statistical effect; and de-
mand-driven immigration, especially in the case of Ireland, was likely to occur in
any case to fill in the available vacancies.
Second, the picture is different for the EU2 countries because the main destina-
tion countries for these immigrants are Spain and Italy (Figures 2 and 3). The pro-
portion of these immigrants has also increased during the 2000s facilitated by bi-
lateral agreements between Spain and Italy and the sending countries and
legalizations of immigrants there (Brücker et al., 2009).. Moreover, given the rela-
tive geographic, and in the case of Romanians also linguistic proximity to Italy
and Spain, these migration trends suggest that geographical and linguistic distance
as well as networks are indeed very important pull-factors. In general, however,
the importance of geographical distance as a migration deterring factor has de-
clined over time with the increased availability of low-cost airlines (see Box 1,
Pollard et al., 2008; Brücker and Damelang, 2009; Brücker et al., 2009).
Figure 2
Share of foreign nationals from the new member states resident in the EU15
EU8
4.5
3.5
3
2000
2.5
2003
2
2007
1.5
0.5
0
k
m
15
nd
d
ly
y
K
e
ria
ce
en
ar
an
an
nd
ai
ur
ec
Ita
U
iu
EU
la
an
m
st
Sp
ed
bo
lg
rl a
nl
re
Ire
Au
en
Fr
Be
Fi
er
Sw
m
G
he
D
xe
et
Lu
EU2
1.8
1.6
1.4
1.2 2000
1 2003
0.8 2007
0.6
0.4
0.2
0
k
m
15
nd
d
ly
g
y
K
e
ce
ria
en
ar
an
nd
an
ai
ur
ec
U
Ita
iu
EU
la
an
m
st
Sp
ed
bo
lg
rl a
nl
re
Ire
Au
en
Fr
Be
Fi
er
Sw
m
G
he
D
xe
et
Lu
Source: Brücker and Damelang (2009), Tables 3b and 4b.

Notes: In % of the host population. Data is from National Statistics, apart for France and Greece (LFS, annual), Ireland (Irish
LFS, 4th quarter), Italy (2000: Eurostat), UK (UK LFS, 2nd quarter). In 2007, estimates are provided for EU8 residents in
Luxembourg and Spain. Partially no data available: Austria (2000), Ireland (EU8: 2000, 2003 (2004 instead), EU2: 2000,
2003), Luxembourg (2000). No data at all for Portugal.
Figure 3
Share of foreign nationals resident in the EU15
14
12
10
0
2)
6)
1)
re 4)
nl 1)
Ire 3 )
ed 2)
)
rl a 2)
Fr n 3)
1)
EU )
Sp 1)
3)
rtu )
Au s 3
1
Po a 2
G a rk
m
15
nd
Sw d
y
y
K
e
ce
ga
en
an
an
nd
l
ai
ec
U
et Ita
iu
ri
la
an
m
st
lg
m
en
Be
Fi
er
he
D
EU15 EU10 EU2 Non-EU27

Source: European Commission (2008a), Table A1.
Notes: 2007. In % of the total population. Luxembourg is excluded due to the scale comparability; proportion of EU15 na-
tionals in Luxembourg is 37.3, of EU10 nationals 0.8, of non-EU 3.4 and is missing for EU2 nationals. Choice of data sources
according to data availability for individual countries. For population statistics end of year data, for LFS 4th quarter data. Fig-
ures are not fully comparable between member states due to the use of different sources. Data from the LFS should be
treated with some caution due to limitations of the survey with regard to foreign populations, in particular concerning cover-
age of very recent migrants and collective households, relative levels of non-response and small sample sizes. 1) EU LFS
quarterly data, 4th quarter. 2) National and Eurostat Population Statistics. 3) Eurostat population statistics, 2007 DG Em-
ployment estimate. 4) CSO Ireland, Quarterly National Household Survey, 4th quarter; for 2003 3rd quarter 2004; Nationals
from BG and RO included under non-EU27 nationals until 2006 and under EU10 nationals in 2007. 5) EU LFS quarterly
data, 4th quarter. EU10 and EU2 nationals included under non-EU27. 6) EU totals and sub-totals are only of an indicative
nature, as they are the sum of country values that stem from different sources; EU totals and sub-totals include country data
which are not shown individually due to small sample sizes. Data for Ireland on EU2 migrants and for Portugal on EU10 mi-
grants are not available. See also notes in European Commission (2008a).
Third, based on the available data, Figure 3 also indicates that while the over-
whelming proportion of migrants from the EU10 is in Ireland, and the largest pro-
portion of EU2 migrants is in Spain and Italy, non-EU27 migrants constitute the
majority among all foreign nationals in all the old member states, except for Ire-
land (where it is EU10 migrants) and Luxembourg (not shown) as well as Bel-
gium (where it is EU15 migrants). It is also worth noting that between 2003 and
2007 the number of foreign residents from the EU8 also significantly increased in
countries outside the EU15, such as Iceland and Norway (Brücker and Damelang,
2009; Brücker et al., 2009).
Turning now to the available data on actual migration flows, it is important to

distinguish gross and net immigration. Since the Eurostat data on emigration are
missing for most countries and years, we present trends only for gross immigra-
tion rates for the four selected EU15 countries for which the data was available:
Sweden, the UK, Spain and Germany. Immigration rates are calculated as the pro-
portion of total immigrant inflows to the receiving country’s population. As can
be seen from Figure 4, gross immigration rates of EU12 citizens have been stead-
ily increasing in all four countries, including the period of EU enlargement.12
Figure 4
Gross immigration rates in selected EU15 countries
Germany
1.60
1.40
1.20
1.00
0.80
0.60
0.40
0.20
0.00
1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006
Total foreigners EU12
Spain
2.00
1.50
1.00
0.50
0.00
1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006
Sweden
1.00
0.80
0.60
0.40
0.20
0.00
1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006
12 Note, however, that when calculating total immigration from the EU12, the figures were
missing in several years and countries and in several cases zeros were reported.
Figure 4 (continued)
Gross immigration rates in selected EU15 countries
UK
1.00
0.80
0.60
0.40
0.20
0.00
1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006
Source: Own calculations based on data from the Eurostat online database for Population and Social Con-
ditions / International Migration and Asylum / International Migration Flows available at:
http://epp.eurostat.ec.europa.eu/portal/page?_pageid=1090,30070682,1090_30298591&_dad=portal&_sc
hema=PORTAL
Notes: Immigration rates are calculated in % of the receiving country’s population. For Germany, data for the Czech Repub-
lic, Slovenia and Slovakia are missing in 1991, for the Czech Republic, Cyprus, Malta and Slovakia in 1992, for Cyprus and
Malta in 1993-1996. For Sweden, the data for the Czech Republic and Slovakia are missing in 1992. For Spain, the data for
the Czech Republic, Slovenia and Slovakia are missing in 1992, for Slovenia in 1993, for Cyprus, Latvia, Malta and Slovakia
in 1995, for Estonia, Latvia and Lithuania in 1996. Zeros are reported in Germany for Cyprus and Malta in 1998, 1999 and
2001, for Estonia, Latvia and Lithuania in 2001; in Sweden for the Czech Republic, Lithuania, Slovenia and Slovakia in 1991,
for Malta in 1994 and 1996; in the UK for the Czech Republic and Estonia in 2001 and 2002, for Malta in 2001, for Slovenia
in 2001-2005, for Slovakia in 2002.
Figure 5 presents data on gross inflows to the UK and Ireland: the Worker Reg-
istration Scheme (WRS) applications and National Insurance Number (NINO) al-
locations in the UK and Personal Public Service Number (PPSN) allocations in
Ireland.13 Overall, Figure 5 shows that the total number of nationals from the EU8
have been increasing since 2004, and they reached a maximum in late 2006
(WRS) or early 2007 (NINOs) in the UK and in 2006 in Ireland; however, they
have been declining since then. The bulk of these applications are from Polish na-
tionals, followed by Slovaks and Lithuanians. Note also that compared with the
same period of 2007, there was an increase in the number of Hungarian nationals
applying to WRS and for NINOs in 2008. This most probably reflected the wors-
ening economic situation in Hungary. The figure also suggest some increase in the
numbers of Bulgarian and Romanian nationals after the 2007 enlargement in spite
of the transitional arrangements imposed on them.
13 Note that NINO and PPSN figures are directly comparable as social security number;
whereas WRS constitutes a different measure. These figures are likely to overstate the ac-
tual migration numbers due to the likely temporary and seasonal nature of immigration
from the new member states and a lack of incentives to deregister.
Figure 5
Registered immigration in Ireland and the UK
UK Worker Registration Scheme applications, May 2004-December 2008
65000
60000 Slovenia
55000
Slovakia
50000
45000 Poland
40000
Lithuania
35000
30000
Latvia
25000
20000 Hungary
15000
Estonia
10000
5000 Czech
Republic
0
q3 04
q4 4
q1 04
q2 5
q3 05
q4 5
q1 05
q2 6
q3 06
q4 06
q1 6
q2 7
q3 7
q4 07
q1 7
q2 08
q3 8
q4 08
8
00
00
00
00
00
00
00
00
00
00
0
0
/2
/2
/2
/2
/2
/2
/2
/2
/2
/2
/2
/2
/2
/2
/2
/2
/2
/2
/2
q2
UK National Insurance numbers (in thousands), January 2002-September 2008
120 Bulgaria
Romania
100
Malta
Cyprus
80
Slovenia
Slovakia
60
Czech R.
Poland
40
Hungary
Estonia
20
Lithuania
0 Latvia
2
04
06
07
08
8
/0
/0
/0
/0
/0
/0
/0
/0
/0
/0
/
/
03
09
03
09
03
09
03
09
03
09
03
09
03
09
2-
2-
3-
3-
4-
4-
5-
5-
6-
6-
7-
7-
8-
8-
/0
/0
/0
/0
/0
/0
/0
/0
/0
/0
/0
/0
/0
/0
01
07
01
07
01
07
01
07
01
07
01
07
01
07
Registered immigration in Ireland and the UK
Irish personal public service numbers, 2001-2008
150000
Romania
Bulgaria
120000
Slovenia
Slovakia
90000 Poland
Malta
Lithuania
60000
Latvia
Hungary
30000 Estonia
Czech R.
Cyprus
0
2001 2002 2003 2004 2005 2006 2007 2008
Source: UK WRS data are from Accession Monitoring Reports May 2004-June 2006, May 2004-December
2006, May 2004-December 2007, and May 2004-December 2008; UK NINO data are from Department for
Work and Pensions online statistics:
www.dwp.gov.uk/asd/tabtool.asp#ni_alloc;
Irish PPSN data are from the Department of Social and Family Affairs online statistics:
www.welfare.ie/EN/Topics/PPSN/Pages/ppsn_all_2008.aspx
Notes: WRS show applicants approved for initial applications only.
Finally, valuable evidence is based on country-level studies. Gilpin et al. (2006)

report that the number of nationals from the new member states increased substan-
tially in the UK. Blanchflower and Lawton (chapter 8 in this volume) document that
850,000 workers from the new member states have registered to work in the UK
since May 2004 until June 2008 under the Worker Registration Scheme, and there
have been an additional 19,525 and 31,080 worker registrations from Bulgaria and
Romania, respectively, between April and June 2008. The authors also argue that
these numbers do not reflect actual immigration, since these people are temporary
workers. However, the number of net migration from the eight new member states
was much lower and was equal to 71,000 in 2006. This highlights the importance of
the definition of migration. Doyle et al. (2006) document a small increase in immi-
gration from the new member states in Sweden in 2005, while Gerdes and Wadensjö
(2008) provide updated figures and show that the increase was larger in later peri-
ods. However, these immigrants still constitute a small part of the overall immigra-
tion to Sweden. Suggested reasons why immigration to Sweden did not increase by
a larger extent include fewer job vacancies, less flexible labor markets and language
issues. Doyle et al. (2006) document a remarkable increase in the number of for-
eigners in Ireland between 2003 and 2005, and the majority of this inflow in 2005
consisted of nationals from the new member states. Barrett et al. (2008) report that
nationals from the new member states in Ireland in 2006 constituted 3% of the coun-
try’s population, an remarkable increase from fewer than 25,000 people in 2002. Fi-
nally, Brenke and Zimmermann (2007) document an increase in net immigration
flows from the new member states into Germany despite the “closed door” policy.
b. The Source Country Perspective
To complete the picture of post-enlargement migration flows, we present available

evidence from the source countries’ perspective in this section. As shown in Figure
6, emigration to the EU15 has increased in all EU8 and EU2 countries since 2000.
In 2007, the largest shares of emigrants in proportion to the home population were in
Romania (7.2%), Bulgaria (4.1%), Lithuania (3.8%) and Poland (3.4%). According
to the European Commission (2008b), between 2003 and 2007, 3.1% of working-
age Lithuanians moved to another EU country, followed by Cypriots (3.0%), Roma-
nians (2.5%), Poles (2.0%), Slovaks (2.0%) and Bulgarians (1.7%). Including those
who emigrated more than four years ago, the countries with the highest share of
emigrants were Romania, Cyprus, Bulgaria, Lithuania and Poland. For sake of com-
parison, the share of emigrants to the EU15 in 2007 was 7.1% in Croatia and
amounted to an extraordinary 25.5% in Albania, the country with the highest emi-
gration rate in the region (Brücker and Damelang, 2009; Brücker et al., 2009).
Figure 6
Share of emigrants to the EU15 in the new member states
5
2000
4 2003
2007
3
0
a
ia
nd
a
2
ia
ia
.
ia
ia
ry
R
EU
EU
ni
ni
an
ak
ar
en
tv
ga
la
to
ua
ch
lg
La
om
Po
ov
ov
un
Es
th
ze
Bu
Sl
Sl
H
R
Li
C
Source: Brücker and Damelang (2009), Table 6b.

Notes: In % of the home population. Data are from National Population Statistics, Eurostat, LFS. 2000: without Austria and
Luxembourg, 2000-2003: without Ireland, 2004-2007: Ireland included with structure of PPSN. See also notes in Brücker and
Damelang (2009).
Figure 7
Emigration rates to Germany correlated with GDP per capita/unemployment rates in the EU12
Source: Migration data are from the Eurostat online database available at
hema=PORTAL. Data on GDP per capita in PPS are from Eurostat Yearbooks 2006-2007 and 2008; data
on unemployment and population are from Eurostat Yearbook 2008.
Notes: Migration rate is calculated as immigration from the respective sending country to Germany in % of the sending coun-
try’s population. Time period 1998-2006.
Another interesting fact is that in 2004, the destination countries for emigrants
from the new EU members changed, and Ireland and the UK became the preferred
destinations after enlargement. For example, the main destination countries for
emigrants from Lithuania in 2002 were Germany, Estonia, Russia, Ireland and the
US; whereas they largely moved to the UK and Germany in 2004 (World Bank,
2006). Kadziauskas (2007) confirms that emigration from Lithuania has increased
after enlargement and warns that the official emigration data might be severely
understated.
A similar upward trend in emigration is apparent from Polish Labour Force

Survey data, with 20% more Poles staying abroad in 2004 than in 2003, which
amounted to 250,000 people (World Bank, 2006). Kaczmarczyk and Okólski
(2008) report that based on Polish LFS data, the number of Polish residents who
stayed abroad for at least two months tripled from early 2004 to early 2007 from
around 180,000 to around 540,000. Germany remains the most important destina-
tion country for immigrants from Poland (especially regarding seasonal migra-
tion), although its share is decreasing; while the importance of the UK and Ireland
is increasing (World Bank, 2006; Frelak and Kazmierkiewicz, 2007;
Kaczmarczyk and Okólski, 2008). The migrants are predominantly male, work-
oriented, young, relatively well-educated and tend to stay abroad for less than one
year. The proportion of individuals with tertiary education is even larger among
those migrating after enlargement, leading to an emergence of two distinct emi-
grants groups – low-skilled individuals from the periphery and high-skilled ones
(Kaczmarczyk and Okólski, 2008).
Regarding potential macro-economic determinants of migration, Figure 7 is

based on Eurostat data on gross immigration flows and examines whether a cor-
relation exists between migration flows into Germany and the GDP per capita
and unemployment rates of the sending countries. It shows that both GDP per
capita and the unemployment rate are significant push-factors. Moreover, the
concave pattern of GDP per capita is in line with the migration “hump” hy-
pothesis (see Hatton and Williamson, 2005).14
c. Who Migrated, and How Do They Fare?
According to the European Commission (2008a, b), the post-enlargement mi-

grants demonstrate greater labor market participation and have higher employ-
ment rates than the populations in either the sending or receiving countries.
14 Some caution, however, is needed when interpreting these results. Although there were
large income and unemployment differences between old EU15 and new member states
upon accession in 2004, unemployment has until recently been declining in the new mem-
ber states.
However, there are notable exceptions: recent EU8 immigrants in Germany ex-
hibit lower employment and participation rates, but above average self-
employment rates (Brenke et al., chapter 4 in this volume). The share of em-
ployed among EU8 migrants in the EU15 is larger for cohorts arriving after
enlargement than for pre-enlargement migrants, and the shares of unemployed
and inactive are correspondingly smaller for the former than for the latter
(Brücker and Damelang, 2009). Interestingly, the employment rate is almost the
same for pre- and post-enlargement EU8 immigrants in Germany, while in the
UK the share of post-enlargement migrants in employment is much higher than
that for pre-enlargement migrants. Note also that self-employment seems to
have been used as a means to circumvent transitory arrangements in Germany
(Brenke and Zimmermann, 2007; Brenke et al., chapter 4 in this volume). In-
deed, while the share of self-employed among recent EU10 immigrants (those
residing for four years or less) is low in the UK, around half of employed recent
migrants from the EU2 in the UK and around 40% of recent EU10 migrants in
Germany are self-employed (European Commission, 2008b).
Migrants from the new member states are also more likely to be overrepre-
sented in low and medium-skilled sectors (such as manufacturing, construction,
hotels and restaurants, but also agriculture and private households, especially for
workers from EU2) and occupations in spite of the majority having a medium
educational attainment and almost a quarter of EU8 immigrants possessing high
education (European Commission, 2008b; Brücker and Damelang, 2009;
Brücker et al., 2009). According to Brücker and Damelang (2009), in 2006 about
29% of the immigrant population (of working age) from Bulgaria and Romania
had low educational attainment, compared to 17% from the EU8 and 27% of na-
tives from the EU15; the corresponding percentages with high education were
18, 22 and 27, respectively. The proportion of highly educated migrants from
the EU8 in the EU15 is lower for cohorts that arrived after enlargement than for
those who arrived before 2004, but also the share of the less-skilled has de-
clined, resulting in a slight improvement of the average education level of the
EU8 immigrants. Interestingly, the proportion of EU8 post-enlargement mi-
grants in Germany with low education is substantially larger than the proportion
of pre-enlargement migrants with low skills suggesting a negative selection of
migrants; while in the UK, the proportion of those with low education is smaller
and of those with higher education is larger after enlargement. In addition, mi-
grants from the new member states are disproportionately young and there is
also a relatively large share of females.
It is also important to note that a significant proportion of these migrants is

likely to be of a temporary or seasonal nature. For example, 62% of workers regis-
tering in 2008 under the Worker Registration Scheme (WRS) in the UK intended
to stay for less than three months, compared to 59% in 2007 and 55% in 2006
(Home Office, 2009).15 Germany also has a large number of migrants who are sea-
sonal workers (see also European Commission, 2008b; Pollard et al., 2008).
Regarding country-level evidence from the UK, Ireland and Sweden, in most
cases, the majority of EU8 migrants were found to be male, young, originating
from Poland, the Baltic states and Slovakia, with relatively high or medium skill
levels and concentrated in relatively low-skilled sectors, indicating such issues as
downgrading and imperfect transferability of human capital. Despite their rela-
tively high educational levels, in the UK these immigrants are concentrated in
low-skilled sectors, such as hotels and restaurants, manufacturing and agriculture
and in the low-skilled occupations (Gilpin et al., 2006; Drinkwater et al., 2006;
Blanchflower et al., 2007; Blanchflower and Lawton, chapter 8 in this volume).
They are also more likely to be self-employed, have a higher propensity to work
than natives and hold a qualification (Blanchflower et al., 2007; Blanchflower and
Lawton, chapter 8 in this volume). However, they were found to be relatively low
paid. It is also argued that the higher unemployment rates for the most recent EU8
migrants arise largely because they are relatively young.
Doyle et al. (2006) study immigrants from the new member states in Sweden in
2005. They document that they have a higher labor force participation rate than na-
tionals, and the majority work in hotels and restaurants, and in the construction and
manufacturing sectors. Gerdes and Wadensjö (2008) report that these immigrants have
relatively high education levels but lower employment rates than natives, and if they
work, they self-select into relatively low-skilled jobs and sectors. Barrett and Duffy
(2007) find that immigrants from the new member states in Ireland have the largest
occupational disadvantage compared to other immigrants and natives. Furthermore,
they show that the effect is largest for the post-enlargement arrivals and confirm the
existence of “downgrading”. According to Hughes (2007), there was a large increase
of the EU10 employees in the construction sector in Ireland in the beginning of 2007.
Barrett et al. (2008) provide the most recent analysis of EU8 immigrant performance
in Ireland and show that the difference in earnings is either non-existent or low for
people with low skill levels and for people at the lower end of the earnings distribu-
tion; while the difference is higher for those at the upper ends of the skills and earnings
distributions. The authors suggest that the transferability of human capital is a crucial
determinant of the immigrant-native earnings gap for these immigrants.
Building on Brenke and Zimmermann (2007), Brenke et al. (chapter 4 in this

volume) provide evidence on the impact of immigrant flow from EU8 countries
on the German labor market following EU enlargement, using more recent mi-
cro census data. As previously stated, Germany did not immediately open up its
labor market to immigrants from the new member states. Nevertheless, their
15 The temporal and seasonal nature of post-enlargement migration and the fact that there
are no incentives to deregister may also explain the observed discrepancies between reg-
ister and survey data.
analysis confirms a substantial inflow and suggests that the composition of EU8
immigrants has changed since EU enlargement. The majority of the new EU8
immigrants are male and young, and they are less educated compared to previ-
ous immigrant groups. They also find that recent EU8 immigrants are more
likely to be self-employed than employed as a wage earner. Furthermore, these
recent EU8 immigrants earn less conditional on being employed or self-
employed. The findings suggest that these recent EU8 immigrants are more
likely to compete for low-skilled jobs with immigrants from outside of Europe
and not with German natives. While Germany needs high-skilled immigrants,
the analysis suggests that the new EU8 immigrants only replace non-EU immi-
grants in low-skilled jobs. These results underline the importance of more open
immigration policies targeting high-skilled immigrants. The current German
migration policy not only fails to attract the required high-skilled workforce, but
also cannot avoid attracting the low-skilled immigrants.
The reasons for migration from the new to the old member states and the bar-
riers migrants face are important factors that shape the nature of east-west mi-
gration in Europe. Consistently with the above evidence, employment-related
factors were found to be the most important mobility motives for these emi-
grants (Fouarge and Ester 2007a, b; Bonin et al. 2008). The intentions to learn
English may also constitute a reason for emigrating as shown for migrants to the
UK by Anderson et al. (2006), and Pollard et al. (2008). East Europeans were
generally found to be unhappy with their lives, dissatisfied with their salaries
and working conditions, concerned about the availability of good jobs and inse-
cure about their current jobs. All these factors further contributed to reasons to
move abroad (Blanchflower and Lawton, chapter 8 in this volume; Blanchflower
et al., 2007). A similar picture emerges from studies in the new member states.
Kadziauskas (2007) reports that in Lithuania, a country with one of the largest
emigration rates, 90% of the respondents name low salaries as the main motive
to seek employment abroad. On the other hand, language and cultural barriers
play key roles as factors deterring intra-European migration. Furthermore, wor-
ries about finding a (suitable) job and the expected loss of social contacts with
family or friends also matter (Fouarge and Ester, 2007a; Bonin et al., 2008).
Is it likely that migration intentions in the new member states and thus actual
migration will decline in the future? The latest available evidence shows that im-
migration from the new member states to the EU15 has risen after enlargement, in
particular in Ireland and the UK, and these new networks may in turn trigger fu-
ture migration. However, the inflows of nationals from the new member states to
the UK and Ireland also appear to have slowed down in 2007 and 2008 (see Fig-
ure 5 and European Commission, 2008b). Moreover, there does not seem to be a
simple relation between the transitional arrangements and migration flows. On the
one hand, the proportion of individuals intending to emigrate after the 2004
enlargement was found to be larger in the new member states than in the EU15;
whereas it was smaller before the enlargement (see Fouarge and Ester, 2007a;
Zaiceva and Zimmermann, 2008).16 This may suggest that with open borders an
increasing number of individuals in the EU8 are willing to try their luck abroad,
since the option to return or migrate again is now always available. However, it
may also be due to the transitory enlargement-related shock and euphoria, which
may fade away over time. A number of additional factors could influence future
east-west migration: growth in the new member states; income convergence; fal-
ling unemployment rates; expected appreciation in their exchange rates in the long
term; new vacancies and skills shortages in these countries, combined with an
ageing population and remaining cultural barriers. A recent European Commission
report points out that “a further surge of labour mobility from the new Member
States seems unlikely” (European Commission, 2008a, p. 8). Indeed, Figure 8
shows that in 2006 although around 15% of respondents in the new member states
think about emigrating to another EU country in order to work “but haven’t de-
cided yet”, 13% have already “given up the idea”. The data from the 2007 Euro-
barometer on intentions to migrate abroad, although not directly comparable to
earlier waves, seem to indicate that moving intentions were on average lower in
2007 than they were in 2005, and the largest decline occurred in the Baltic States
(Bonin et al., 2008). Interestingly, cross-border moving intentions in 2007 were
highest in Bulgaria and Romania.
Figure 8
Intentions to migrate for work
“Have you, yourself, ever considered living in another Member State in order
to work?” Feb-Mar 2006
100
don't know
80
no, you have never
thought of it
60
yes, you have

already thought of
40 it, but gave up the
idea
yes, you think of it
, but you haven't
20 decided yet
yes, you have

already done it
0
10
15
EE
PL
LT
Y
SI
A
Z
SK
LV
EU
EU
C
L
TA
TA
TO
TO
Source: Own tabulations from the Eurobarometer EB 65.1.
16 Note, however, that the questions were asked in a slightly different manner before and af-
ter enlargement, which may partly account for the documented differences.
Box: Air passengers traveling to and from new EU member states via London
Since mid-1990s there has been an increase in the number of air passengers at London
airports traveling to and from the new member states. Furthermore, the most dramatic
increase occurred after 2004 (Figure B1). This trend is clear if London airports are
studied with the focus on passengers to and from Poland (Figures B2 and B3). In
1996, Heathrow was the most relevant airport and the number of passengers was
around 297,000; however by 2000 the number of passengers had increased to around
474,000, and by 2007 Gatwick, Heathrow, Stansted and Luton had reported the total
number of passengers to and from Poland was close to 3 million. There was also an
increase in the number of connections between Poland and London and by 2007 ten
Polish airports had established connections with four London airports. While the ad-
vent of low-cost airlines contributed to this development, the demand factors are unde-
niable. One should also note that these numbers reflect not only the increased migra-
tion to and from the UK, but also an increase in tourism. From another perspective,
the importance of distance in monetary terms is declining (see also Brücker and Dam-
elang, 2009 and Brücker et al., 2009 for documenting “the eroding role of distance”).
In early 2009 it cost almost the same amount of money for a person to travel from a
peripheral town of Rzeszów in south-east Poland to London on a low-cost air carrier
(around 30 euros, 1 hour and 35 minutes, one way) as to go by train to Warsaw
(around 26 euros and 5 hours and 20 minutes, one way) and somewhat less if traveling
to Warsaw by bus (around 16 euros and 4 hours and 45 minutes, one way).17
Figure B1
Number of air passengers between London airports and new member states
8000000
7000000
6000000
5000000
Charter
4000000 Sheduled
Total
3000000
2000000
1000000
0
1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007
Source: Own calculations from UK Airport Statistics, Table 12.1 Intl Air Pax Route Analysis
(various years):
www.caa.co.uk/default.aspx?catid=80&pagetype=88&pageid=3&sglid=3
Notes: Number of passengers traveling to or from London Airports to or from the following countries: Bulgaria, the
Czech Republic, Estonia, Hungary, Latvia, Lithuania, Poland, Romania, Slovakia, Slovenia.
Figure B2
Total air passenger traffic at London airports to and from Poland
1400000
1996 1997 1998 1999

1200000
2000 2001 2002 2003
2004 2005 2006 2007
1000000
800000
600000
400000
200000
0
Gatw ick Heathrow Stansted Luton
(various years):
Figure B3
Trends in the numbers of air passengers at London airports to and from Poland
1000000
900000
800000 Warsaw
Krakow
700000
Gdansk
600000
Katow ice
500000 Poznan
400000 Wroclaw
Bydgoszcz /
300000 Szw eredow o
Lodz
200000
Rzeszow
100000 Szczecin
(Golenow )
0
1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008
(various years):
17 This average one-way price is calculated for a week between February 2, 2009, and Feb-
ruary 8, 2009. These requests were made in November 2008. The time reported is the
fastest available.
4 The Impact of Post-Enlargement Migration Flows:

First Evidence
a. The Destination Countries
The possible negative impact of immigration on labor markets and welfare in des-
tination countries was one of the main reasons for introducing the transitional ar-
rangements for workers from the new member states. Were these concerns justi-
fied or exaggerated? While it is still too early to provide a complete answer, some
preliminary evidence already exists. In order to assess the impact on destination
countries, we first examine the aggregate labor market statistics to identify the
main trends before and after the recent enlargements.
Figures 9–14 present labor market statistics for the EU15 as well as for Ireland,
Sweden and the UK. It can be seen from Figure 9 that there is no evidence of a sig-
nificant slow-down of employment growth after the 2004 enlargement until at least
2007. The exception is Ireland, where employment growth started to decline during
the middle of 2005 from very high levels. Regarding the aggregate unemployment
rate, Figure 10 shows that until recently there had been a general decline in Ger-
many and the EU15 after the 2004 enlargement. The situation differs in the three
countries that opened up their labor markets in 2004: the unemployment rate in
Sweden increased in 2005 but afterwards declined to pre-enlargement levels, and in-
creased once again in 2008. The declining trend in the UK reversed in the middle of
2005 and then stabilized at a higher level. It had been slightly declining since mid-
2006 until the recent increase. The trend in Ireland was a slightly declining unem-
ployment rate until 2005, fluctuating between 4 and 5% between 2005 and 2007,
which started to pick up more recently. Therefore, one cannot identify a consistent
(and adverse) reaction pattern of unemployment with immigration.
Figures 11 and 12 show that there was demand for new labor in the destination
countries because the 2007 job vacancy rates in the UK and Sweden were the same
if not larger than before the 2004 enlargement. Even in manufacturing, a sector with
a high concentration of accession country citizens, the available data for the UK
show that after an initial decline immediately after enlargement, the job vacancy rate
was higher at the end of 2007 than before enlargement. Declining job vacancy rates
can be observed in all the time series after around 2007. Similarly, earnings data re-
veal that in the period following the 2004 enlargement up until recently nominal
wages and salaries (as part of the labor cost index) continued to grow even in manu-
facturing in the UK, Ireland and Germany (see Figures 13 and 14).18
18 Note, however, that these are nominal values.

Figure 9
Employment growth in selected old member states
6
5
4
3
2
1
0
-1
-2
20 03
20 03
20 04
20 01
20 03
20 04
20 04
20 01
3
1
20 04
20 01
20 02
3
q0
q0
q0
q0
q0
q0
q0
q0
q0
q0
q
q
q
q
03
03
04
04
04
04
05
05
05
05
06
06
08
08
08
06
06
07
07
07
07
20
20
20
20
20
20
20
20
20
20
EU15 Germany Ireland Sw eden UK
Source: Own calculations based on data from the Eurostat online database for Population and Social Condi-
tions / Labour Market / Employment and Unemployment (Labour Force Survey):
http://epp.eurostat.ec.europa.eu/portal/page?_pageid=1090,30070682,1090_30298591&_dad=portal&_sche
ma=PORTAL
Notes: The indicator employment growth gives the change in percentage from one year to another of the total number of
employed persons on the economic territory of the country or the geographical area. The indicator is based on the European
System of Accounts.
Figure 10
Unemployment rates in selected old member states
12
10
0
2
4
20 03
20 04
20 01
20 03
20 04
20 01
20 03
20 04
20 01
20 02
20 03
20 04
20 01
20 02
20 03
20 04
20 01
20 02
q0
q0
q0
q0
q
q
q
08
08
06
07
07
07
07
08
08
03
03
04
04
04
04
05
05
05
05
06
06
06
20
20
20
20
EU15 Germany Ireland Sw eden UK
Source: Own calculations based on data from the Eurostat online database for Population and Social Condi-
tions / Labour Market / Employment and Unemployment (Labour Force Survey):
http://epp.eurostat.ec.europa.eu/portal/page?_pageid=1090,30070682,1090_30298591&_dad=portal&_sche
ma=PORTAL
Notes: Seasonally adjusted. Unemployment rates represent unemployed persons as a percentage of the labor force. The la-
bor force is the total number of people employed and unemployed. Unemployed persons comprise of people aged 15 to 74
who were: a) without work during the reference week; b) currently available for work, i.e. were available for paid employment
or self-employment before the end of the two weeks following the reference week; c) actively seeking work, i.e. had taken
specific steps in the four week period ending with the reference week to seek paid employment or self-employment or who
found a job to start later, i.e. within a period of at most three months.
Figure 11
Job vacancy rate in selected old member states
4.5
4
3.5
3
2.5
2
1.5
1
0.5
0
20 02
3
20 03
20 01
20 04
3
20 01
20 03
20 04
20 01
20 04
20 02
20 04
20 02
20 01
20 02
20 02
3
20 01
20 03
q0
q0
q0
q0
q0
q
q
q
q
q
q
q
q
08
08
07
07
08
06
07
07
05
05
06
06
06
04
05
05
03
03
03
04
04
04
20
20
20
20
20
EU15 Germany Sw eden UK
ditions / Labour Market / Job Vacancy Statistics available at:
hema=PORTAL
Notes: In %. A job vacancy is defined as a post (newly created, unoccupied or about to become vacant), for which the em-
ployer is taking active steps to find a suitable candidate from outside the enterprise concerned and is prepared to take more
steps; and which the employer intends to fill either immediately or in the near future. The data for Ireland are not available.
For EU15 and Germany, provisional values are reported.
Figure 12
Job vacancy rate in manufacturing in selected old member states
2.5
1.5
0.5
0
20 02
20 03
20 01
20 02
20 01
20 02
20 03
20 03
20 04
20 01
20 02
20 03
20 04
20 01
20 04
20 01
3
q0
q0
q0
q0
q0
q0
q
q
03
03
03
04
04
04
04
05
05
05
05
07
06
06
06
06
07
07
07
08
08
08
20
20
20
20
20
20
Germany Sw eden UK
hema=PORTAL
Notes: In %. For definitions, see notes to Figure 11. The data for Ireland and EU15 are not available. For Germany, provi-
sional values are reported.
Figure 13
Labor cost index: wages and salaries in selected old member states
160
140
120
100
80
60
40
20
0
20 02
20 01
20 02
3
20 04
20 03
20 01
20 02
20 01
20 02
20 03
20 02
20 03
20 04
20 02
3
q0
q0
q0
q0
q0
q0
q0
q0
q
q
03
03
03
04
04
04
04
05
05
05
05
06
06
06
06
07
07
07
07
08
08
08
20
20
20
20
20
20
20
20
Germany Ireland UK
ditions / Labour Market / Labour Costs available at:
hema=PORTAL
Notes: Index 2000=100. Wages and salaries in industries and services excluding public administration, seasonally adjusted
and adjusted by working days, nominal value. Data for EU15 are not available.
Figure 14
Labor cost index: wages and salaries in manufacturing in selected old member states
160
140
120
100
80
60
40
20
0
20 02
20 03
20 04
20 01
20 04
20 01
20 02
20 01
20 02
20 03
20 02
20 03
20 04
20 02
3
q0
q0
q0
q0
q0
q0
q0
q0
q
q
03
03
03
04
04
04
04
05
05
05
05
06
06
06
06
07
07
07
07
08
08
08
20
20
20
20
20
20
20
20
Germany Ireland UK
hema=PORTAL
Notes: Index 2000=100. Wages and salaries in manufacturing seasonally adjusted and adjusted by working days, nominal
value. Data for EU15 are not available.
While causal interpretations of these findings would be misleading, the stabil-

ity of the observed aggregate variables around the time of the 2004 EU enlarge-
ment is consistent with the hypothesis that there were not any significantly nega-
tive impacts on the labor markets in the receiving countries. This is true even for
the UK and Ireland, where the inflows of EU8 immigrants were the largest. A
general caveat of this analysis is that that aggregate data may suffer from aggrega-
tion bias and the impact of immigration from the new member states may be
higher for the local labor markets, i.e. certain skill groups, sectors or occupations.
The available literature also suggests that immigration from the new member
states has hitherto had little, if any, impact of on wages and employment and has
had no negative impact on the welfare systems in the receiving countries. Several
studies find very small or no impact of the recent immigration from the new
member states on the UK unemployment rate or wages, even after controlling for
potential observable and unobservable confounding factors in an econometric
analysis (see Gilpin et al., 2006; Blanchflower et al., 2007; Lemos and Portes,
2008). Blanchflower and Lawton (chapter 8 in this volume) suggest that recent
immigration from the EU8 may have had some (small) impact on the relative
wages of the least skilled (see also Blanchflower and Shadforth, 2009). It is rather
the “fear of unemployment” that has risen in response to immigration from the
new member states (Blanchflower and Shadforth, 2009). The number of EU8 na-
tionals applying for UK social benefits is low, and child benefits constitute the
largest category. On the macroeconomic side, Blanchflower et al. (2007) find no
evidence that the rise in the UK unemployment rate is related to structural factors
and argue that recent immigration helps to suppress inflationary pressures.
The post-2004 immigrants from the EU10 in Sweden are slightly overrepre-
sented in social assistance, but underrepresented in other income transfer pro-
grams, and the number of recipients of child allowance is also low among these
immigrants (Gerdes and Wadensjö, 2008).19 Regarding Ireland, no evidence of
“welfare tourism” is found, and it is argued that the immigration was primarily
demand-driven (Doyle et al. 2006; Hughes, 2007; Doyle, 2007). Some decrease in
earnings growth rates of some employees in Ireland is documented (Doyle et al.
2006), which has increased thereafter (Hughes, 2007). Although several cases of
displacement of Irish workers by EU8 citizens earning lower wages are docu-
mented, on aggregate, displacement does not appear to be a source of disturbance,
since the aggregate unemployment rate has remained constant after enlargement,
and even if Irish workers have been displaced in some sectors, they have most
probably moved to other better-paying jobs (Doyle et al., 2006; Hughes, 2007).
Moreover, vacancies data show that the demand for labor has remained strong af-
ter enlargement.
19 See also Doyle et al. (2006) and Wadensjö (2007).

Brücker et al. (2009) find that the immigration from the new member states
during 2004-2007 has decreased EU15 average wages by only 0.09% in the short
run (compared to a hypothetical situation of what they would have been without
the additional mobility from the EU8), and have no impact in the long-run. Aver-
age EU15 unemployment rates were found to increase in the short-run by 0.06
percentage points and by 0.02 percentage points in the long-run (see also Euro-
pean Commission 2008a, b). Analogously, Baas et al. (2009) calibrate the decline
in wages to be 0.09% and an increase in unemployment by 0.06 percentage points
in the EU15 in the short run with no significant effects in the long run.
The recent east-west migration has several additional impacts that are worth
mentioning. First, post-enlargement mobility has substantial positive effects for
the EU as a whole in terms of GDP, GDP per capita and productivity; and a
somewhat smaller effect for employment in the long run (European Commission,
2008a, b, 2009; Baas et al., 2009). For example, Brücker et al. (2009) find that
post-enlargement immigration from the EU8 increases the GDP (and GDP per
capita) of the enlarged EU by about 0.1 % in the short run and by about 0.2% in
the long run. The old member states also experience an increase in GDP, but their
GDP per capita decreases in the short run due to decreased productivity and wages
(D’Auria et al., 2008; Brücker et al., 2009). The effects are the opposite in the new
member states: GDP declines, but there are gains for real wages, productivity and
GDP per capita.
It is also likely that skilled immigration from the new member states may im-
prove the inequality situation in the old EU15 in the long run. Using data on im-
migration to OECD countries, Kahanec and Zimmermann (2008) provide empiri-
cal evidence on the positive relationship between high-skilled immigration and
economic equality in the destination countries. In addition, it is known that immi-
gration may alleviate the demographic burden and, especially skilled immigration,
may relieve the pressure placed due to an ageing population in the old EU15 (see
Bonin et al., 2000 for an analysis of the fiscal impact of earlier immigration in
Germany). Since immigrants from the new member states are predominantly
young and relatively skilled (although they seem to self-select into relatively low-
skilled occupations), they may decrease the fiscal burden of future resident gen-
erations; however, this effect depends on immigrants’ remaining lifetime net
payments to the public sector in the host country.
Overall, based on the early evidence, the World Bank (2006, p.1) concludes
that: “Evidence from the receiving countries indicates that fears of massive in-
flows of workers and devastating impacts on receiving labor markets were un-
founded, since inflows of foreign workers have generally supplemented rather
than replaced domestic labor and helped sustain solid economic growth, while at
the same time keeping local wages stable.” The European Commission reports
(see European Commission, 2006a, b, 2008a, b, 2009) also conclude that the im-
pacts of the recent east-west mobility on labor markets in receiving countries have
been non-negative on average, or even positive, and have not led to any serious
disturbances in the labor markets. Moreover, these immigrants seem to have
helped improve labor market efficiency by alleviating labor shortages in sectors
and occupations with the excess demand for labor that could not be satisfied by
the native labor force. They also contributed to easing inflationary pressures and
to increase economic growth in the receiving countries.
b. The Source Countries
An important part of the picture of the effects of post-enlargement migration is the

situation in the sending countries. Evidence reviewed above suggests that several
new member states have experienced significant outflows of relatively young and
skilled people. Is brain drain from the new member states indeed a worrisome is-
sue? What are the demographic and economic impacts of the post-enlargement
emigration in the EU8 and EU2, which are still troubled by post-transition struc-
tural and institutional weaknesses?
Figure 15
Unemployment rate in the new member states
20
18
16
14
12
10
8
6
4
2
0
ta
ia
15
nd
a
lic
ia
ia
ia
ia
s
ry
ni
ni
ru
an
ak
ar
al
en
tv
ub
ga
la
EU
to
ua
yp
M
lg
La
om
Po
ov
ov
ep
un
Es
th
Bu
Sl
Sl
R
R
H
Li
ch
ze
2004 2007
C
ditions / Labour Market / Employment and Unemployment (Labour Force Survey) available at:
hema=PORTAL
Notes: Annual data, not seasonally adjusted. Unemployment rates represent unemployed persons as a percentage of the
labor force.
As discussed above, labor migration from the new member states has increased
since enlargement. The significant outflows of predominantly young and skilled
individuals may have negative long-run impacts on these countries’ economic
growth, demographic situation and public budgets. For example, it has been esti-
mated that the elderly dependency ratio will more than double by 2050 in Lithua-
nia, and without fundamental changes (such as adjusting benefits, the age of eligi-
bility or promoting immigration) the social security system may fail entirely
(Kadziauskas, 2007). A significant decline in population and labor force has also
been forecasted for Poland, which would place a serious burden on its labor mar-
ket and public budgets over the next 50 years (Kaczmarczyk and Okólski, 2008).
This shrinking share of working age population is also likely to drive down the fu-
ture emigration potential from the new member states.
Figure 16
Employment growth rates in the new member states
-1
-2
15
ta
ia
nd
a
li c
ia
ia
ia
ia
s
y
ni
ni
ru
an
ak
ar
al
en
ub
tv
ga
EU
la
to
ua
yp
M
lg
La
om
Po
ov
ov
ep
un
Es
th
Bu
Sl
Sl
R
R
Li
ch
ze
C
2004 2007
ditions / Labour Market / Employment and Unemployment (Labour Force Survey) available at:
hema=PORTAL
Notes: The indicator employment growth gives the change in percentage from one year to another of the total number of
employed persons on the economic territory of the country or the geographical area. The indicator is based on the European
System of Accounts.
Emigration from the new EU members goes hand in hand with a decreasing
unemployment rate and an increasing employment growth rate, the number of va-
cancies rose until 2007 (see Figure 17), all indicating potential labor shortages.
Indeed, emigration may exacerbate structural weaknesses in national labor mar-
kets, as there are shortages of highly skilled specialists in a number of sectors
(World Bank, 2006). There were around 12,000 unfilled vacancies in Lithuania at
the end of 2005, and labor shortages were particularly large in manufacturing and
trade sectors (Kadziauskas, 2007). There were particularly significant shortages of
workers in Poland in manufacturing, trade, hotels and restaurants, and construc-
tion (Kaczmarczyk and Okólski, 2008). The mismatch between vacancies and
workers is also often present. One of the consequences is increasing immigration
in the new member states due to the increased demand for immigrant labor. In Po-
land, for example, the number of immigrants in 2004 was the highest since 1960,
and the main countries of origin in 2003 were Ukraine, Belarus and Russia (Frelak
and Kazmierkiewicz, 2007; Iglicka, 2005). This may be a sign of new migration
patterns when citizens of new member states are substituted by non-EU immi-
grants.
Figure 17
Job vacancy rate in selected new member states
Total
2.5
2
1.5
1
0.5
0
2
3
q0
q0
q0
q0
q0
q0
q0
q0
q0
q0
q0
q0
q0
q0
q0
q0
q0
q0
04
04
04
05
05
05
05
06
06
06
06
07
07
07
07
08
08
08
20
20
20
20
20
20
20
20
20
20
20
20
20
20
20
20
20
20
Latvia Lithuania Poland Slovakia
Manufacturing
3.5
3
2.5
2
1.5
1
0.5
0
2
3
q0
q0
q0
q0
q0
q0
q0
q0
q0
q0
q0
q0
q0
q0
q0
q0
q0
q0
04
04
04
05
05
05
05
06
06
06
06
07
07
07
07
08
08
08
20
20
20
20
20
20
20
20
20
20
20
20
20
20
20
20
20
20
Construction
8
7
6
5
4
3
2
1
0
2
3
q0
q0
q0
q0
q0
q0
q0
q0
q0
q0
q0
q0
q0
q0
q0
q0
q0
q0
04
04
04
05
05
05
05
06
06
06
06
07
07
07
07
08
08
08
20
20
20
20
20
20
20
20
20
20
20
20
20
20
20
20
20
20
hema=PORTAL
Notes: In %.
Figure 18
Average monthly earnings in industry and services (in euros)
3500
3000
2500
2000
1500
1000
500
0
BG CZ CY EE HU LT LV MT PL RO SI SK AT DE ES IE NL SE UK
2002 2006
Source: Own calculations from the the Eurostat online database for Labour Market / Earnings / Structure of
Earnings Survey 2002 (and 2006) / Monthly Earnings available at:
http://epp.eurostat.ec.europa.eu/portal/page?_pageid=1996,45323734&_dad=portal&_schema=PORTAL&
screen=welcomeref&open=/labour/earn/earn_ses/earn_ses_mo&language=en&product=EU_MASTER_lab
our_market&root=EU_MASTER_labour_market&scrollto=171
Figure 19
Labor cost index: wages and salaries in the new member states
500
450
400
350
300
250
200
150
100
50
0
a
ta
nd
ia
a
li c
ia
ia
ia
ia
s
ry
ni
ni
ru
an
ak
ar
al
en
tv
ub
ga
la
to
ua
yp
M
lg
La
om
Po
ov
ov
ep
un
Es
th
Bu
Sl
Sl
R
R
Li
ch
ze
2004 2007
C
hema=PORTAL
Notes: Index 2000=100.
Another development has been increasing wages, and thus unit labor costs in
the new member states, as suggested by Figures 18 and 19. However, wages in
these countries are still several times lower than, for example, in the UK or Ireland
(see Figure 18). The World Bank (2006) reports that the largest increases in the
first quarter of 2006 were in the health sector in Lithuania and in agriculture in
Poland and warns of potential inflationary pressures.
In general, one has to be careful, however, when correlating emigration with

these new developments in the labor markets. According to Rutkowski (2007), the
latest developments in the labor markets of new member countries are mainly
driven by economic restructuring and only to a small extent by emigration. This is
in line with conclusions in Kaczmarczyk and Okólski (2008) that fast growth rates
and growing demand play an important role.
The relatively high level of education of these migrants together with the ob-
served “downgrading” into less skilled occupations (due to incomplete transfer-
ability of migrants’ skills) reported above may suggest a potential danger of “brain
waste”, when human capital depreciates during periods of low-skilled work in the
host country. On the other hand, investing in language skills, expanding profes-
sional networks and learning other new skills abroad implies transferring these
skills back home on returning. In addition, migrants may bring back home a mi-
grant partner who is also likely to be well-educated, suggesting brain circulation
in the long run (European Commission, 2008b). The World Bank (2006) argues
that there was no incidence of massive “brain drain” from the new member states,
although in some sectors (such as health care) the emigration of highly skilled
specialists was relatively large. Indeed, around 4.3% of Polish physicians in 2006
were interested in working abroad and had acquired the required certifications,
which suggests a potential non-negligible brain drain from this sector (Frelak and
Kazmierkiewicz, 2007). Overall, although the proportion of the highly educated
emigrants from the new member states is above that of the stayers, especially for
the EU8 emigrants (European Commission, 2008b; Brücker and Damelang, 2009;
Brücker et al., 2009), the significant drain of skilled people is not observed
(Brücker et al., 2009). At the same time, there is a tendency of increasing enrol-
ment rates for tertiary education in the new EU member countries, which may
compensate for the outflow of skilled labor (European Commission, 2008b;
Brücker et al., 2009).
Finally, migrant remittances is another important issue when analyzing the

consequences of emigration. As shown in Figure 20, remittances constitute a sig-
nificant part of the GDP in Bulgaria and Romania, and they are also non-
negligible in the Baltic states. Their share increased in many countries between
2003 and 2006. The remittances are mostly seasonal in nature in Poland and the
Baltic states, which suggests a temporary seasonal pattern of migration from these
countries (Kaczmarczyk and Okólski, 2008). Remittances are largely used for
household consumption and the purchase of durable goods, but there has been a
recent tendency to invest in human capital, particularly tertiary education (World
Bank, 2006; Kaczmarczyk and Okólski, 2008).
Figure 20
Workers’ remittances and compensation of employees, received
10
9
8
7
6
5
4
3
2
1
0
a
ia
ta
nd
a
li c
ia
ia
ia
ia
s
ry
ni
ni
ru
an
ak
ar
al
en
tv
ub
ga
la
to
ua
yp
M
lg
La
om
Po
ov
ov
ep
un
Es
th
Bu
Sl
Sl
R
R
Li
ch
ze
C
2003 2004 2005 2006
Source: World Bank World Development Indicators CD-Rom 2007 (for 2003-2005) and World Bank (2008).
Notes: In % of GDP.
The impact of these remittances on the economic development so far is diffi-

cult to document, but they may play a positive role in the economic development
of the new member states inasmuch as they support aggregate demand and in-
vestments in education and business activities. Another issue that deserves further
investigation is how these remittances affect the decisions of “those who stay be-
hind,” including their labor supply and educational decisions. In addition, the con-
sequences of all these effects on public budgets in the source countries are still to
be fully studied. Overall, the World Bank (2006, p.1) concludes that although
emigration from the new member states “… has led to skills shortages and bottle-
necks in several sectors in the main sending countries and increased wage pres-
sures, sending countries are benefiting from increased remittances from workers
abroad who may also return with additional human capital.”
5 The Current Economic Crisis
With the downswing of European economies in early 2008, some deterioration of

labor market variables have recently been observed in the destination countries
(Figures 9-14). However, with the numbers of immigrants from the EU8 and EU2
declining and the rise of the global crisis, this can be hardly attributed to immigra-
tion. Similarly, the crisis is also affecting the new member states and potentially
will cause a slowdown there too in vacancies and employment growth (see also
Figure 17) as well as an increase in unemployment.
Obviously with the crisis, the “natural experiment” of politically determined

enlargement and the opening up of some labor markets is coming to an end, or
better said, moving to a different stage in the analysis. This is unfortunate, since
the study of the outflow of EU8 and EU2 emigrants caused by enlargement, their
size, performance and effects on the natives cannot be completed. Nevertheless,
more information about this regime will be available with the release of micro
data until 2008, since we currently often do not even have the 2007 data waves.
With the release of the 2009 micro data at the latest, we will have to focus on a
new research question, namely, the issue how under free mobility and an eco-
nomic downturn migration will perform.
It is known that labor migrants are very responsive to economic cycles, espe-
cially if there are no mobility restrictions for returning home and re-returning to
the host country. EU8 and EU2 immigrants in the EU15 countries do not face any
long-term mobility constraints. To the contrary, mobility barriers are known to
decline in the EU over the next years because some currently existing restrictions
have to be removed. Also many migrants, especially among Poles were evaluated
to be temporary movers. At times of high and rising unemployment, the general
public, policy makers and companies become more concerned about the potential
adverse effects of immigration on the job opportunities of natives. Consequently,
rightly or wrongly, the social pressure on migrants to leave rises. Since social
networks are very important for migrant job searches, an increase in unemploy-
ment will reduce the availability of valuable information, and hence employment
chances of ethnic networks. Migrants are further over-educated while working in
lower-paid jobs in the host country; if they become unemployed, they may prefer
to return to more skilled occupations back home. This all suggests that many for-
eign workers will now, or with a further worsening of the crisis, move home.
This view has to be qualified. Unlike in normal business cycles, the current
downturn is not asymmetric across economies but occurs parallel. This implies
that migrants face a deepening of the crisis in both the host and home countries
suggesting that it is less reasonable to return home. Even more, the relative situa-
tion in the old EU15 countries may still be better than the one in the new EU
members. The crisis also affects the host economies differently. For instance,
Germany is hit in the export-oriented investment goods industries with their many
skilled workers; whereas Ireland and the United Kingdom suffer from a more
global economic downturn. High-skilled immigrants, who are also typically
young, single and with no other strong family ties in the host country, are more
likely to return home than low-skilled immigrants who cannot expect to find a job
back home. Moreover, such factors like established networks, language or the
worsening economic situation at home may further serve as push or pull factors
for east-west European migration. Indeed, Blanchflower and Lawton (chapter 8 in
this volume) argue that inflows to the UK, for example, could remain strong well
into the future; although the weight of the source countries may change, for exam-
ple, with less people coming from Poland and more from Hungary, the Baltic
states, Romania or Bulgaria.
6 Conclusions
This chapter summarizes what we hitherto know about migration in the EU after
the 2004 and 2007 enlargements. The Eastern enlargement of the EU removed
some of the institutional migration barriers and was thus an impetus to the migra-
tion potential in Europe. While the overall numbers of migrants from the new
member states in the EU15 increased between 2003 and 2007, this increase was
unevenly distributed across countries. The proportion of migrants from the new
member states in the EU15 remains noticeably lower (and their inflows exhibit
slower dynamics) than that of non-EU migrants (apart from Ireland). In general,
there does not seem to be any simple link between the transitional arrangements
and the post-enlargement migration. The increase in immigration from the EU8
was largest in Ireland and the UK, while the proportion of EU2 citizens in Spain
and Italy continued to increase in 2007. The main source countries are Poland and
Romania among the EU2 and EU8 countries and the largest shares of emigrants
relative to the country’s population are in Romania, Bulgaria, Lithuania and Po-
land.
Have these primarily economic migrants affected the host labor markets? They
certainly brought skilled hands and new knowledge with them, and thus contrib-
uted to the economic development in the host countries. There has been no suffi-
cient evidence so far to conclude that they would crowd out native labor from em-
ployment or lower their wages. Even if crowding out happened in certain sectors
or occupations, aggregate data suggest that natives affected found new jobs else-
where. Nor is there any evidence that this new labor is more dependent on welfare
than the natives. Note, however, that the impacts may be different when disaggre-
gating by local labor markets, sectors or skill groups.
Concerning the new member states, the drain of mainly young and skilled peo-
ple could pose some additional demographic challenges on countries experiencing
significant outflows of people. It appears, however, that the migration trajectory is
of a temporary nature. To the extent that these people return with not only accu-
mulated human capital and new skills, but also financial resources, the anticipated
brain circulation may in fact help solve demographic and economic problems in
their countries of origin. Even before returning, they channel some resources back
home in the form of remittances; these finances are most significant in Romania
and Bulgaria, but also in the Baltic states. Those effects seem to be limited so far,
but it is too early and too little researched to reach a valid conclusion at this stage.
Several caveats on these findings need to be mentioned, however. First, the

time span may be still too short for any impact to be measurable. Second, the im-
pacts may be greater when disaggregating into local labor markets, skill and im-
migrant groups. Third, these effects are inherently difficult to measure, mainly be-
cause it is hard to track international flows of migrants, account for the
endogeneity of the migration decisions, and control for the confounding variables,
including the reactions of natives and stayers to migrant flows. Yet, the period
covered characterized by steady economic development lends itself to the analysis
of the impacts of migration on EU labor markets. The current economic crisis will
severely limit our ability to trace these impacts, since many migrants will return
home (and not be documented) and the labor market conditions will have changed
too substantially.
While the ongoing economic crisis may change the momentum of several mi-
gration trajectories, free migration should in fact alleviate many consequences of
the crisis. Labor migrants are typically very responsive to economic cycles, espe-
cially if there are no mobility restrictions when returning home and re-returning to
the host country. EU migrants do not face any such mobility constraints in the
long term. Furthermore, many immigrants from the new member states considered
themselves temporary movers. This, among other arguments, suggests that many
foreign workers will move home now or soon. However, if the economic crisis
has hit the home country stronger than the host country, or if labor market condi-
tions at home are especially poor for low-skilled workers, return migration might
be delayed.
Based on this account of the post-enlargement migration flows, it appears that

the key scientific and policy challenges from this process have to do with the fol-
lowing themes. First, we would like to have a better notion about the size of future
migration flows. Will many more people decide to leave new member states for
the old ones, or will the numbers stabilize at much lower figures than current? Our
preliminary account suggests that after the initial increase, which possibly in-
volved legalizations, the net numbers of migrants have started to decline. Second,
the composition of migration flows is a key factor driving their effects.
Third, the decisions of the post-enlargement migrants about the timing, dura-
tion and frequency of their stays abroad affect the characteristics of the effects
they may impinge upon the host and source labor markets. For example, these de-
cisions determine not only the size but also the composition of the stock of immi-
grants in the host labor markets at each time. Furthermore, immigrant intentions
regarding their planned length of stay influence the behavior of economic actors,
such as their own decisions to invest in human capital specific to the host econ-
omy, or the behavior of investors to invest in different technologies that use im-
migrant labor. Circular migration is also a vehicle of knowledge transfer and thus
may affect technological advancement in both source and destination countries.
Fourth, it is yet to be evaluated how the transitory arrangements have affected

the post-enlargement migration flows and their repercussions. It appears that these
arrangements have diverted some migrants to those old member states which
opened their labor markets early on. There is evidence that they have affected the
composition of migration flows and their consequences. For example, in contrast
to those countries which liberalized access to their labor markets, it seems that the
barriers applied in Germany possibly discouraged the most skilled migrants and
thus lowered the average quality of the migrant inflows. Those who came, dispro-
portionately work as entrepreneurs – probably as a means to penetrate the German
labor market despite the restrictions.
Finally, free mobility of people is a cornerstone of the European Union. How

to ensure that it is upheld in all aspects, that it contributes to economic prosperity
as well as the well-being of the individual and the society, and that it helps to alle-
viate economic and financial disturbances such as the ongoing economic crisis, is
a major policy task for the present and future.
The following chapters provide an in-depth analysis of these issues from pan-
European as well as country-specific perspectives. Chapter 2 evaluates and meas-
ures the effects of post-enlargement migration on the distribution of income and
labor market outcomes within host and source labor markets. Chapter 3 turns to
another momentous phenomenon accompanying post enlargement migration
flows: public perceptions. This chapter evaluates the public attitudes towards mi-
grants and their effects in the context of the general need for migrants in an aging
Europe, the political economy of migration and integration policy measures, as
well as the options for effective policy approaches to public attitudes in this con-
text. In a series of country case studies, Chapters 4 to 8 address the scope and ef-
fects of post-enlargement migration flows in particular destination countries:
Germany, Spain, Ireland, Sweden and the UK. Chapters 9 and 10 address the
same issue in selected source countries: Poland and the Baltic states. Chapter 11
looks at the migration potential in Albania, a potential candidate for EU member-
ship, thus benchmarking our results for the new member states and providing a
window to the future of European migration.
References
Alvarez-Plata, P./H. Brücker/B. Siliverstovs (2003), Potential Migration from Central and
Eastern Europe into the EU15 – An Update. Report for the European Commission, DG
Employment and Social Affairs, German Institute for Economic Research (DIW Ber-
lin).
Anderson, B./M. Ruhs/B. Rogaly/S. Spencer (2006), Fair Enough? Central and East Euro-
pean Migrants in Low-Wage Employment in the UK. Report published by the Joseph
Rowntree Foundation, London.
Baas, T./H. Brücker/A. Hauptmann (2009), Labor Mobility in the Enlarged EU: Who Wins,
Who Loses?, in: M. Kahanec/K. F. Zimmermann (eds.), EU Labor Markets After Post-
Enlargement Migration. Berlin et al., 47-70.
Barrett, A./D. Duffy (2007), Are Ireland’s Immigrants Integrating into its Labour Market?
IZA Discussion Paper No. 2838, Bonn.
Barrett, A./S. McGuinness/M. O’Brien (2008), The Immigrant Earnings Disadvantage
across the Earnings and Skills Distributions: The Case of Immigrants from the EU’s
New Member States in Ireland. IZA Discussion Paper No. 3479, Bonn.
Becker, G. S. (1957), The Economics of Discrimination. Chicago.
Blanchflower, D. G./H. Lawton (2009), The Impact of the Recent Expansion of the EU on the
UK Labour Market, in: M. Kahanec/K. F. Zimmermann (eds.), EU Labor Markets After
Post-Enlargement Migration. Berlin et al., 181-215.
Blanchflower, D. G./C. Shadforth (2009), Fear, Unemployment and Migration, in: The
Economic Journal 119 (535), F136-F182.
Blanchflower, D. G./J. Saleheen/C. Shadforth. (2007), The Impact of the Recent Migration
from Eastern Europe on the UK Economy. IZA Discussion Paper No. 2615, Bonn.
Boeri, T./H. Brücker (2005), Migration, Co-ordination Failures and EU Enlargement, in:
Economic Policy, 20 (44), 629-703.
Bonin, H./W. Eichhorst/C. Florman/M. O. Hansen/L. Skiöld/J. Stuhler/K. Tatsiramos/H.
Thomasen/K. F. Zimmermann (2008), Geographic Mobility in the European Union: Op-
timising its Economic and Social Benefits. IZA Research Report No. 19, Bonn.
Bonin, H./B. Raffelhüschen/J. Walliser (2000), Can Immigration Alleviate the Demo-
graphic Burden?, in: FinanzArchiv 57 (1), 1-21.
Borjas, G. J. (1985), Assimilation, Changes in Cohort Quality, and the Earnings of Immi-
grants, in: Journal of Labour Economics 3 (4), 463-489.
Borjas, G. J. (1987), Self-Selection and the Earnings of Immigrants, in: American Eco-
nomic Review 77 (4), 531-553.
Borjas, G. J. (1999), Immigration and Welfare Magnets, in: Journal of Labor Economics 17
(4), 607-637.
Brenke, K./K. F. Zimmermann (2007), Zuwanderung aus Mittel- und Osteuropa trotz Ar-
beitsmarktbarrieren deutlich gestiegen, in: DIW-Wochenbericht 44, 645-653.
Brenke, K./M. Yuksel/K. F. Zimmermann (2009), EU Enlargement under Continued Mo-
bility Restrictions: Consequences for the German Labor Market, in: M. Kahanec/K. F.
Zimmermann (eds.), EU Labor Markets After Post-Enlargement Migration. Berlin et al.,
111-129.
Brücker, H./A. Damelang (2009), Labour Mobility within the EU in the Context of
Enlargement and the Functioning of the Transitional Arrangements. Analysis of the
Scale, Direction and Structure of Labour Mobility. Deliverable 2. IAB, Nürnberg.
Brücker, H. et al. (2009), Labour Mobility within the EU in the Context of Enlargement
and the Functioning of the Transitional Arrangements. Final Report (IAB, CMR, fRDB,
GEP, WIFO, wiiw), Nürnberg.
Brücker, H./G. S. Epstein/B. McCormick/G. Saint-Paul/A. Venturini/K. F. Zimmermann
(2002), Managing Migration in the European Welfare State, in: T. Boeri/G. Hanson/B.
McCormick (eds.), Immigration Policy and the Welfare System. Oxford, 1-168.
Burda, M. C. (1995), Migration and the Option Value of Waiting, in: Economic and Social
Review 27, 1-19.
Chiswick, B. R. (1980), An Analysis of the Economic Progress and Impact of Immigrants,
Employment and Training Administration. U.S. Department of Labor. National Techni-
cal Information Service, Washington DC, PB80-200454.
Chiswick, B. R. (1998), The Economic Consequences of Immigration: Application to the
United States and Japan, in: M. Weiner/T. Hanami (eds.), Temporary Workers or Future
Citizens? Japanese and U.S. Migration Policies. New York, 177-208.
Chiswick, B. R. (1999), Are Immigrants Favorably Self-Selected?, in: American Economic
Review 89 (2), 181-185.
Chiswick, C. U./B. R. Chiswick/G. Karras (1992), The Impact of Immigrants on the Mac-
roeconomy. Carnegie-Rochester Conference Series on Public Policy, 37 (1), 279-316.
De Giorgi, G./M. Pellizzari (2006), Welfare Migration in Europe and the Cost of a Harmo-
nised Social Assistance. IZA Discussion Paper No. 2094, Bonn.
D’Auria, F./K. Mc Morrow/K. Pichelmann (2008), Economic Impact of Migration Flows fol-
lowing the 2004 Enlargement Process: A Model Based Analysis, European Economy.
Economic Papers No. 349, Brussels (European Commission).
Doyle, N. (2007), The Effects of Central European Labor Migration on Ireland, in: J. Smith-
Bozek (ed.), Labor Mobility in the European Union: New Members, New Challenges.
Washington DC (Center for European Policy Analysis), 35-59.
Doyle, N./G. Hughes/E. Wadensjö (2006), Freedom of Movement for Workers from Cen-
tral and Eastern Europe – Experiences in Ireland and Sweden. Swedish Institute for
European Policy Studies (SIEPS), Report No. 5, Stockholm.
Drinkwater, S./J. Eade/M. Garapich (2006), Poles Apart? EU Enlargement and the Labour
Market Outcomes of Immigrants in the UK. IZA Discussion Paper No. 2410, Bonn.
European Commission (2006a), Report on the Functioning of Transitional Arrangements
set out in the 2003 Accession Treaty (period 1 May 2004 – 30 April 2006). Communi-
cation from the Commission to the Council, the European Parliament, the European
Economic and Social Committee and the Committee of the Regions. Brussels, 8 Febru-
ary 2006.
European Commission (2006b), Enlargement Two Years After: An Economic Evaluation,
European Economy. Directorate-General for Economic and Financial Affairs, Occa-
sional Paper No. 24. Brussels, 3 May 2006.
European Commission (2008a), The Impact of Free Movement of Workers in the Context
of EU Enlargement, Report on the first phase (1 January 2007 – 31 December 2008) of
the Transitional Arrangements set out in the 2005 Accession Treaty and as requested
according to the Transitional Arrangements set out in the 2003 Accession Treaty.
Communication from the Commission to the European Parliament, the Council, the
European Economic and Social Committee and the Committee of the Regions. Brussels,
18 November 2008.
European Commission (2008b), Employment in Europe, Chapter 3: Geographical labour
mobility in the context of EU enlargement. Directorate-General for Employment, Social
Affairs and Equal Opportunities. Office for Official Publications of the European
Communities. Luxembourg, October 2008.
European Commission (2009), Five Years of an Enlarged EU - Economic Achievements

and Challenges. European Economy, No. 1/2009.
Fouarge, D./P. Ester (2007a), Determinants of Migration Intentions in Europe. Exodus or
Bounded Mobility? Institute for Labour Studies, Tilburg University, mimeo.
Fouarge, D./P. Ester (2007b), Factors Determining International and Regional Migration in
Europe. European Foundation for the Improvement of Living and Working Conditions,
Dublin.
Frelak, J./P. Kazmierkiewicz (2007), Labor Mobility: The Case of Poland, in: J. Smith-
Bozek (ed.), Labor Mobility in the European Union: New Members, New Challenges.
Washington DC (Center for European Policy Analysis), 60-79.
Gerdes, C./E. Wadensjö (2008), Immigrants from the New EU Member States and the
Swedish Welfare State. Swedish Institute for European Policy Studies (SIEPS), Report
No. 9, Stockholm.
Gilpin, N./M. Henty/S. Lemos/J. Portes/C. Bullen (2006), The Impact of Free Movement of
Workers from Central and Eastern Europe on the UK Labour Market. Department of
Work and Pensions, Working Paper No. 29, London.
Harris, J. R./M. P. Todaro (1970), Migration, Unemployment and Development: A Two-
Sector Analysis, in: American Economic Review 60 (1), 126-142.
Hatton, T. J./J. G. Williamson (2005), What Fundamentals Drive World Migration?, in: G.
J. Borjas/J. Crisp (eds.), Poverty, International Migration and Asylum. Palgrave Mac-
millan, 15-38, Basingstoke.
Home Office (2009), Accession Monitoring Report May 2004 – December 2008, UK Bor-
der Agency, London.
Hughes, G. (2007), EU Enlargement and Labour Market Effects of Migration to Ireland
from Southern, Central and Eastern Europe (Paper presented at Second IZA Migration
Workshop: EU Enlargement and the Labour Markets, Bonn, 7-8 September 2007).
Iglicka, K. (2005), The Impact of the EU Enlargement on Migratory Movements in Poland.
Center for International Relations, Report No. 12/05, Warsaw.
Kaczmarczyk, P./M. Okólski (2008), Economic Impacts of Migration on Poland and the
Baltic States. Fafo-Paper 2008: 1, Oslo.
Kadziauskas, G. (2007), Lithuanian Migration: Causes, Impacts and Policy Guidelines, in:
J. Smith-Bozek (ed.), Labor Mobility in the European Union: New Members, New
Challenges. Washington DC (Center for European Policy Analysis), 80-100.
Kahanec, M./K. F. Zimmermann (2008), Migration, the Quality of the Labour Force and
Economic Inequality. IZA Discussion Paper No. 3560, Bonn.
Kahanec, M./K. F. Zimmermann (2009), International Migration, Ethnicity and Economic
Inequality, in: W. Salverda/B. Nolan/T. M. Smeeding (eds.), The Oxford Handbook of
Economic Inequality. Oxford, 455-490.
Lemos, S./J. Portes (2008), The Impact of Migration from the New European Union Mem-
ber States on Native Workers. Department for Work and Pensions, Working Paper No.
52, Leeds.
Massey, D. S. (1990), Social Structure, Household Strategies, and the Cumulative Causa-
tion of Migration, in: Population Index 56 (1), 3-26.
Mincer, J. (1978), Family Migration Decisions, in: Journal of Political Economy 86 (5),
749-773.
Pollard, N./M. Latorre/D. Sriskandarajah (2008), Floodgates or Turnstiles? Post-EU
Enlargement Migration Flows to (and from) the UK. Institute for Public Policy Re-
search, London.
Rutkowski, J. (2007), Labor Markets in EU8+2: From the Shortage of Jobs to the Shortage
of Skilled Workers. World Bank EU8+2 Regular Economic Report Special Topic,
Washington DC.
Sjaastad, L. A. (1962), The Costs and Returns of Human Migration, in: Journal of Political
Economy 70(5), 80-93.
Stark, O. (1991), The Migration of Labor. Cambridge.
Wadensjö, E. (2007), Migration to Sweden from the New EU Member States. IZA Discus-
sion Paper No. 3190, Bonn.
World Bank (2006), International Migration, Remittances and the Brain Drain. Washington
DC.
World Bank (2008), World Development Indicators 2008. Washington DC.
Zaiceva, A./K. F. Zimmermann (2008), Scale, Diversity, and Determinants of Labour Mi-
gration in Europe, in: Oxford Review of Economic Policy 24 (3), 427-451.
2. Labor Mobility in the Enlarged EU:
Who Wins, Who Loses?
Timo Baas
Herbert Brücker
Andreas Hauptmann
The EU’s Eastern enlargement has triggered a substantial labor migration from
the new into the old member states of the EU. In this chapter, we analyze the im-
pact of this change in labor supply on macroeconomic aggregates and the distri-
bution of earnings and employment opportunities. For this purpose, we employ
two general equilibrium models which analyze the migration impact in a setting
with imperfect labor markets. The first model is based on a nested production
function, which enables us to examine the migration effects on wages and unem-
ployment in the different cells of the labor market. The second model is based on
CGE framework, which allows considering the restructuring of sectors and the in-
teraction of migration with trade and capital movements. Both models assume that
capital stocks adjust to labor supply shocks. Based on scenarios of labor migra-
tion from the EU8 into the EU15 from 2004 to 2007, we find substantial gains of
migration for the enlarged EU: the aggregate GDP of the integrated area in-
creases by about 0.2%, which corresponds to a sum of 24 billion euros. The total
factor income of the native population increases slightly in the receiving coun-
tries. The labor market effects are surprisingly small: in the short-run, wages de-
cline by about to 0.1% in the EU15, and the unemployment rate increases by
about 0.1 percentage points. Conversely, wages may increase by 0.3% in the new
member states, and the unemployment rate may decline by 0.4 percentage points.
In the long-run, after the adjustment of capital stocks, migration is by and large
neutral for wages and unemployment. Since the skill structure of the migrant
workforce does not differ largely from the native workforce in the sending and the
receiving countries, we observe only moderate distributional effects across the
different groups in the labor markets. Less-skilled workers in the receiving coun-
tries lose more proportionally, while high-skilled workers tend to benefit. The
converse holds for the sending countries.

Migration, DOI 10.1007/978-3-642-02242-5_2,
48 Timo Baas, Herbert Brücker and Andreas Hauptmann
1 Introduction
In this chapter, we analyze the impact of labor migration which has been triggered
by the EU’s Eastern enlargement on the distribution of factor income and em-
ployment opportunities in the sending and receiving countries of the enlarged EU.
The EU Eastern enlargement has been accompanied by substantial migration
movements. The migrant population from the eight new member states (EU8) who
joined the European Union (EU) in 20041 has increased from about 0.9 million by
the end of 2003 to 1.9 million by the end of 2007. At the same time, the number of
migrants from Bulgaria and Romania has increased from 0.7 million to 1.9 mil-
lion, although they acceded after 2007 (see Brücker et al., 2009 and Chapter 1 of
this book). Moreover, the selective application of transitional arrangements for the
free movement of workers by the incumbent EU member states is associated with
a substantial diversion of migration flows away from Austria and Germany to-
wards Ireland and the United Kingdom. As a consequence, the EU Eastern
enlargement has seen significant labor supply changes at least in the most affected
sending and receiving countries.
The economic rationale behind the free movement of workers is to increase the
productive use of human resources, and hence, output in the integrated area. Many
simulation models support this view and suggest that the benefits from opening
labor markets to international migration can easily dwarf potential gains from a
further liberalization of international goods and capital markets (Hamilton and
Whalley, 1984). This has been also demonstrated for labor migration within the
enlarged EU (Boeri and Brücker, 2005).
While it is not disputed that an enlarged EU benefits from the free movement
of workers in terms of aggregate output and earnings, the distributional conse-
quences are not. Concerns that immigration from the new member states might
depress wages and increase native unemployment are particularly widespread in
the receiving countries. Hence, many potential destinations for migrants from the
new member states, such as Austria and Germany, are reluctant to open their labor
markets before the end of the transitional periods.
International migration does not only create winners. The standard textbook
model of migration predicts that international labor mobility generates aggregate
gains for natives in the receiving countries, while natives left behind in the send-
ing countries tend to lose (e.g. Wong, 1995). Moreover, production factors in re-
ceiving countries which are net complements to migrant labor tend to win, while
those which are net substitutes tend to lose.
1 Czech Republic, Estonia, Hungary, Latvia, Lithuania, Poland, Slovakia, Slovenia.

2. Labor Mobility in the Enlarged EU: Who Wins, Who Loses? 49
These predictions of the standard model are based on a set of strong assump-
tions. In our view it is important to relax three assumptions of the standard model
for assessing the effects of migration in an empirically meaningful way. First, the
standard model relies on the assumption that labor markets clear, which is particu-
larly unrealistic in the European context. If we consider, however, wage rigidities
and unemployment, we find ambiguous effects of migration on the receiving and
sending countries. The aggregate income of natives in the receiving countries can
decline if migration involves an increase in unemployment there, whereas the ag-
gregate income of natives in the sending countries can increase (Boeri and
Brücker, 2005). The impact of migration on unemployment depends however on
the flexibility of wages in the different cells of the labor market: immigration can
even reduce unemployment in the receiving countries if migrants move into labor
market segments which are characterized by a high wage flexibility (Brücker and
Jahn, 2008).
Second, the standard model relies on the assumption that capital stock is fixed.
However, it is likely that at least in the long-run domestic and international in-
vestment adjusts to changes in labor supply. It is indeed one of the few robust em-
pirical findings in economics that the capital-output ratio, and hence, the produc-
tivity weighted labor intensity of production, remains constant over time (Kaldor,
1961). This implies that aggregate wages are not affected by labor supply shocks
in the long-run (Ottaviano and Peri, 2006).
Third, the standard one-sector framework ignores that economies adjust to migra-
tion by restructuring industries and changing the commodity composition of trade.
labor mobility is crucial to the optimal allocation of resources among sectors. If we
allow for geographical as well as sectoral labor mobility, the additional labor supply
is transferred to the most productive use. In receiving countries, migration reduces
wages in the most affected industries, and therefore, tends to push natives to other
industries; while in the sending countries, emigration leads to a restructuring of in-
dustries due to higher wages in certain sectors. The standard models of trade theory
predict that changes in factor endowments involve changes in the trade vector and
the sectoral structure of employment in small open economies, but leaves factor
prices unchanged if labor and other factors move swiftly across sectors.
Against this background, we address the impact of labor mobility on the distri-
bution of earnings and employment opportunities in a framework which considers
wage rigidities and unemployment, the adjustment of capital stocks, and the re-
structuring of industries and trade. To this end, we employ two models. The first
model analyses the changes in the different cells of the labor market on the basis
of a wage curve approach, which considers that wages adjust only imperfectly to
labor supply changes. Adjustments in other markets are only considered in this
model as long as they affect the capital-output-ratio, which may be the most im-
portant channel of adjustment.
The second model goes beyond this in considering the adjustment of the sec-
toral structure of the economy via international trade and shifts in the structure of
demand and production. Since the adjustment of the trade vector and of the sec-
toral structure of the economy mitigates the migration impact, we expect to find
smaller wage and employment effects of migration in simulations based on this
second model. Whilst we apply the first model for all EU15 and EU8 countries,
the second model will be used only for the two most affected sending and receiv-
ing countries: Poland and the United Kingdom.
The remainder of this chapter is organized as follows. The next section de-
scribes the structure of the labor supply shock and the migration scenarios (Sec-
tion 2). We then outline the two models which will be used for the empirical
analysis (Section 3). Section 4 presents the simulation results. Section 5 concludes
and discusses the possible implications of the financial crisis on the future effects
of the free movement of workers in the enlarged EU.
2 The Scale and Structure of Migration

and Migration Scenarios
2.1 The Scale of Migration
The income gap between the incumbent and the new member states is in the case
of the EU Eastern enlargement larger than in previous accession rounds. The GDP
per capita and wage levels in the new member states have converged rapidly since
EU Eastern enlargement, but the gap in output and earnings is still substantial.
Measured in purchasing power parities, the GDP per capita in the EU8 in 2007
amounted to 55% of that in the EU15. At current exchange rates, the GDP per
capita in the EU8 was about one-third of the EU15, and hourly labor costs were
about 24%. Note that nominal earnings differences affect migration incentives,
since a considerable part of the income earned in the receiving countries is con-
sumed in the home countries. Finally, the unemployment rates have converged in
the EU8 to the average level in the EU15 (see Brücker et al. 2009, for details).
The income gap between old and new EU member states has triggered substan-
tial migration flows since Eastern enlargement. As has been outlined in the previ-
ous chapter, the stock of foreign residents from the EU8 increased by about one
million in the first four years following EU Eastern enlargement. This corresponds
to 250,000 migrants per year, compared to 50,000 before enlargement. Recent
figures from the UK and Ireland suggest that net inflows started to decline in 2008
and are likely to decrease even further in main receiving countries in 2009 follow-
ing the course of the financial crisis.
Furthermore, the number of foreign residents from Bulgaria and Romania in-
creased from about 700,000 to almost 1.9 million from the end of 2003 to the end
of 2007. However, we cannot directly attribute this increase to enlargement, since
Bulgaria and Romania were not members of the EU before 2007. Moreover, all
EU countries, with the exception of Finland and Sweden, have maintained their
immigration restrictions vis-à-vis Bulgaria and Romania since 2007.
The labor supply shock affected EU member states in different ways. We observe
a distinct shift in the geographical structure of migration after enlargement. In 2003
about two-thirds of the EU8 migrants resided in Germany and Austria. Since East-
ern enlargement, about 70% of the net inflows from the EU8 have been absorbed by
Ireland and the United Kingdom. Similarly, Spain and Italy have received the over-
whelming share of migrants from Bulgaria and Romania since the beginning of this
decade; whereas Austria and Germany were the main destinations of migrants from
these countries in the beginning of the 1990s. As a result, we find the highest share
of migrants from the new member states in the following receiving countries: Ire-
land (4.7%), Spain (2.1%), Austria (1.5%), Luxembourg (1.3%), and the UK (1.1%)
in 2007.
From the perspective of the sending countries, 2.6% of the population from the
EU8 and 6.4% of the population from the EU2 have already moved to the EU15 in
2007. Most affected are Romania (7.1%), Lithuania (3.8%), Bulgaria (4.1%) and
Poland (3.4%). In contrast, the Czech Republic, Hungary and Slovenia have been
proportionately less affected by emigration to the EU15.
The migration potential from the new member states is not yet exhausted, al-
though we observe that net flows from the new member states into the EU15 have
started to decline. Based on a partial adjustment model, Brücker et al. (2009) es-
timate the long-run migration stock for the EU8 at about 6% of the population in
the sending countries, and for the EU2 at about 10 to 12% of the population in the
sending countries. Thus, the present migration stocks from the new member states
amount to about 50% of their long-run potential.
We cannot assess the implications of the financial crisis and the global recession
on potential migration flows from the new member states yet. According to EU
forecasts, the Baltic countries and Hungary will be more than proportionally af-
fected by the economic downturn in 2009, while important sending countries, such
as Poland and Romania, are forecasted to be proportionately less affected. Further-
more, migration depends heavily on employment opportunities in the receiving
countries. Hence, the economic downturn is likely to reduce migration flow even if
unemployment rates increase in the sending countries by the same amount as in the
receiving countries. Therefore, it is very likely that net migration rates from the new
member states into the EU15 will contract over the course of the economic down-
turn, but it is an open question when and to what extent they will recover over the
following years.
2.2 The Structure of Migration
The migrant population from the new member states is younger than the native
population in the receiving and sending countries, participates to a higher extent in
the labor force and is favorably self-selected on observable skills. Consequently,
the share of migrants from the new member states in the labor force is higher rela-
tive to the population.
The skill level of migrants from the new member states is heavily concentrated
in the group with medium educational attainment, according to the Eurostat La-
bour Force Survey: 22% of the migrant population from the EU8 have a university
degree (EU15: 27%), 61% a completed vocational training qualification or a simi-
lar level of schooling (EU15: 46%), and only 17% are without a completed voca-
tional training qualification (EU15: 27%) (Eurostat, 2008). Thus, migrants from
the new member states are proportionately less represented at the higher and the
lower ends of the skill spectrum compared to natives in the EU15. Relative to the
population in the sending countries, the skill level of migrants from the new mem-
ber states is higher than that of natives who stay behind, even when controlling for
cohort effects (Brücker and Damelang, 2009).
Overall, the skill structure of the migrant population is relatively similar to that
of the native population in the receiving and sending countries, so the skill com-
position of the workforce in the sending and the receiving countries remains by
and large unaffected by migration from the new member states into the EU15.
However, the occupational structure of employment suggests that migrants

from the new member states are employed below their educational levels: a large
share of migrants is employed in occupations which need only elementary skills.
As a consequence, the wage level of migrants from the new member states in Ire-
land and the UK is well below natives with similar education and work experience
(Barret and Duffy, 2008; Upward, 2009). Moreover, the returns to education do
not increase significantly with the time spend in the receiving countries, although
it is too early for a final assessment on the labor market assimilation of migrants
from the new member states (Upward, 2009). Overall, migrants from the new
member states compete to a large extent with natives and other foreigners in the
less-skilled segments of the labor market in the EU15, although migrant education
levels are relatively high.
2.3 Migration Scenarios
Our simulations focus on the impact of the labor supply shock triggered by EU East-
ern enlargement in the first four years of accession, i.e. on migration from the EU8
into the EU15 during the period from 2004 to 2007.2 This labor supply shock is de-
rived from the difference of two scenarios: the first scenario assumes counterfactually
that the pre-enlargement conditions for migration between the EU8 and the EU15 pre-
vail. Note that this counterfactual scenario does not assume that no migration takes
place, but that labor mobility continues at the pre-enlargement levels. This implies that
the stock of foreign residents from the EU8 increases moderately from 870,000 by the
end of 2003 to 1.07 million by the end of 2007.
Table 1
Migration scenarios
Foreign residents from EU8 Foreign residents from EU8

in persons in per cent of population
Counterfactual Enlargement Enlargement Counterfactual Enlargement Enlargement

Benchmark scenario scenario effect Benchmark scenario scenario effect
2003 2007 2007 2003 2007 2007
AT 60255 64596 89940 25344 0.75 0.81 1.12 0.32

BE 16151 23242 42918 19676 0.16 0.22 0.41 0.19
DE 427958 492123 554372 62249 0.52 0.60 0.68 0.08
DK 9807 11220 22146 10926 0.18 0.21 0.41 0.20
ES 46710 82863 131118 48255 0.11 0.20 0.31 0.12
FI 15825 19154 23957 4803 0.30 0.37 0.46 0.09
FR 33858 29690 36971 7281 0.06 0.05 0.06 0.01
GR 16413 21582 20257 -1325 0.16 0.20 0.19 -0.01
IE 34246 60657 178504 117847 0.86 1.52 4.47 2.95
IT 54665 74909 117042 42133 0.10 0.13 0.20 0.07
LU 1574 2568 5101 2533 0.36 0.58 1.15 0.57
NL 13048 16861 36317 19456 0.08 0.11 0.23 0.12
SE 21147 19301 42312 23011 0.24 0.22 0.47 0.26
UK 122465 154198 609415 455217 0.21 0.27 1.05 0.78
CZ 71019 95954 104442 8488 0.70 0.94 1.03 0.08

EE 26070 33922 36735 2813 1.93 2.51 2.72 0.21
HU 87680 88285 132582 44297 0.88 0.88 1.33 0.44
LT 53557 88922 128361 39439 1.55 2.58 3.73 1.14
LV 23863 32559 42547 9987 1.02 1.40 1.83 0.43
PL 532942 632111 1297647 665536 1.42 1.68 3.45 1.77
SI 35051 40958 35848 -5110 1.76 2.05 1.80 -0.26
SK 43938 60252 132207 71955 0.82 1.12 2.45 1.34
EU151) 874122 1072964 1910370 837406 0.24 0.29 0.52 0.23

EU8 874122 1072964 1910370 837406 1.21 1.48 2.64 1.16
Source: Own estimates and simulation, see text.

Notes: 1) Without Portugal.
The stock of foreign residents in 2003 is used as a benchmark. The counterfactual scenario assumes that immigration flows
continue at their pre-enlargement levels, while the enlargment scenario refers to the actual figures observed in 2007. There-
fore, the difference of the enlargement- and the counterfactual scenario is treated as the "enlargement effect".
2 For a simulation of the impact of the labour supply shock from the EU2 during the same
period of time and of the potential labour supply shocks during the 2008-2014 period see
Brücker et al. (2009).
The second scenario captures the current institutional conditions for migration
under the transitional arrangements. Therefore, it is based on the actual migration
movements. This involves the stock of migrants from the EU8 increasing from
870,000 by the end of 2003 to 1.91 million by the end of 2007. Thus, the EU
Eastern enlargement triggered an additional migration of 840,000 people accord-
ing to our scenarios during the 2004-2007 period (Table 1). Note that these two
scenarios consider, however, the diversion of migration flows, since the counter-
factual scenario captures the pre-enlargement regional structure of migration,
while the Eastern enlargement scenario is based the post-enlargement structure.
We also apply a number of assumptions regarding the structure of the labor

supply shock in our simulations. First, we use the actual participation rates of mi-
grants and natives in the labor force during the 2004-2007 period, which implies
that the migration share in the labor force is higher than the migration share in the
population. Second, we calculate the work experience based on the age informa-
tion in the LFS by using the average time of schooling for the different levels of
educational attainment. Third, and most importantly, we classify the skill structure
of the immigrants from the EU8 which arrived from 2004 onwards by their occu-
pational status and not by the information on their education in the LFS. This pro-
cedure addresses the problem of ‘brain waste’, i.e. the employment of migrants
from the new member states below their education levels. Using the LFS informa-
tion on education levels would clearly overstate the skill level of the employed la-
bor force from the new member states in the receiving countries.3
3 Outline of the Models
Our analysis of the impact of migration from the EU8 into the EU15 is based on two
models. The first model analyses the impact of migration on wages and employment
opportunities in different segments of the labor market. This model is based on a
structural general equilibrium framework which considers imperfect labor markets.
Whereas this model uses an aggregate production function with one sector, the sec-
ond model is based on a computable general equilibrium framework, which also en-
ables us to address the impact of migration on sectoral change. This model relies on
an open-economy framework which also allows us to consider the interaction be-
tween labor migration and the trade in goods and capital mobility.
3 See Baas et al. (2009) for details of the assumptions on the labour supply shock.
3.1 A Structural Model with Wage Rigidities
The first model we employ here builds on Brücker and Jahn (2008) and analyses
the wage and employment effects of labor migration simultaneously. Following
Boeri and Brücker (2005) and Levine (1999), it is assumed that wages respond to
changes in the unemployment rate, albeit imperfectly. As a consequence, addi-
tional labor supply through immigration can both reduce wages and increase un-
employment in the native labor force depending on the elasticities of substitutabil-
ity and complementarity in the different segments of the labor market.
The supply side of the labor market is modeled on an aggregate wage-setting

curve, which can be based on different theoretical foundations (see Blanchflower
and Oswald 1994, Layard et al. 2005, for a discussion). In our context, two model-
ing traditions are particularly important. First, the wage curve can be derived from
bargaining models (see e.g. Lindbeck 1993, Layard and Nickell 1986), which as-
sume that trade unions are concerned about both their employed and unemployed
members. Consider the case where wages are fixed in a bilateral bargaining monop-
oly between trade unions and employer federations. Once wages are fixed, firms
hire workers until the marginal product of labor equals the wage. Both parties that
participate in the wage bargain are aware of this. Higher unemployment means that
more union members are without work, and employed members who are dismissed
will have a lower probability of finding new employment. Consequently, the negoti-
ated wage is lower when unemployment is higher, and vice versa.
Second, in a completely non-unionized environment, a wage curve can be ex-

plained by efficiency wage or shirking considerations (Shapiro and Stiglitz, 1984),
where the productivity of workers is linked to the wage level. Unemployment
works here as a disciplining device since it determines the difficulties of finding a
new job. As a result, firms will reduce the remuneration of workers if the unem-
ployment rate is increasing, since they can achieve the same level of productivity
at a lower wage if unemployment is higher.
What both approaches have in common is that they replace the conventional
labor supply curve with a wage fixing function and they rely on standard assump-
tions about labor demand. However, different conclusions regarding the shape of
the wage curve can emerge from these different theoretical foundations. While the
bargaining models predict a flatter wage curve in labor segments with a high share
of unionized workers (e.g. workers with vocational training), the efficiency wage
models expect a flatter wage curve in labor market segments with a high level of
human capital (e.g. workers with a university degree). Nevertheless, in the empiri-
cal application of the model, we estimate the elasticities of the wage curves with-
out applying a priori restrictions on their curvature.
In contrast to labor supply, our framework relies on conventional assumptions on

labor demand. Once wages are fixed, firms hire workers until the marginal product of
labor equals the wage rate. This enables us to derive labor demand from a standard
production function framework. More specifically, we follow Borjas (2003), Ottavi-
ano and Peri (2006) and others in applying a nested CES production function, which
groups the labor force by education, work experience and national origin.
This approach allows us to derive the employment and wage response to an ex-
ogenous labor supply shock in a general equilibrium framework. The production
function and the wage curves determine a system of simultaneous equations,
which can be solved analytically and enable us to derive the employment and
wage response to migration at the same time. The model considers also the (par-
tial) adjustment of physical capital to labor supply shocks (for a detailed descrip-
tion of the model see Brücker and Jahn, 2008, Baas et al. 2009).
The simulation of the effects of migration requires the estimation of (i) the elastic-
ities of the wage-setting curves; (ii) the relevant parameters of the production function;
and (iii) the adjustment of the capital stock to exogenous labor supply shocks.
The wage-setting curves have been estimated at the national level depending on
data availability for different education and experience groups or the economy as
a whole. We found an overall elasticity of the wage-setting curve of -0.13 at the
level of the EU15, which is slightly higher than the average elasticities of the
wage-curves found in regional level studies (see Nijkamp and Poot, 2005).
The elasticities of the production function have been estimated in the EU15 coun-
tries based on data from the European Community Household Panel (ECHP). The
ECHP has a limited number of observations, which may affect a precise estimation
of the relevant parameters. As a robustness check, we carried out a sensitivity analy-
sis based on parameter values taken from various studies in the literature. Our find-
ings, however, turned out to be robust irrespective of the parameter values we use.
In the case of Germany and the UK, we also used detailed data from an administra-
tive source (IABS in Germany) and the national LFS (UK), which yielded once
more very similar simulation results.4 For the sending countries in the new member
states, we do not have as yet comparable data at hand. We therefore used the aver-
age of our parameter estimates for the EU15 in our simulations for the new member
states, assuming that the economies there will behave similarly to the EU15 coun-
tries.5
4 The results of the sensitivity analysis are available from the authors upon request.
5 Note that in the sending countries, we treated the native and the foreign labour force as
perfect substitutes.
Similar to Ottaviano and Peri (2006), we estimated the adjustment of the capi-
tal stock to labor supply shocks at the national level. We find that changes in labor
supply do not have a long-run impact on the capital-output ratio and that labor
productivity adjusts to an exogenous labor supply shock by more than 10% in one
year in most countries.
The estimation results for the elasticities of the wage-setting curves, the elastic-
ities of the production function and the adjustment of the capital-labor ratio are
presented in Annex Table A1. For detailed discussion, see Baas et al. (2009).
3.2 Sketch of the CGE Model
The model outlined in the previous section enable us to analyze the labor market
impacts of migration in detail, but it considers only a part of the potential adjust-
ment processes of economies to labor supply shocks. In particular, the adjustments
via trade in goods and services and the sectoral structure of the economy cannot
be considered, since the model relies on a one-sector framework. Therefore, we
complement this model by an analysis in a computable general equilibrium (CGE)
framework, which allows these adjustment mechanisms to be considered as well.
The CGE model employed here can be classified as a comparative static model
and follows the neoclassic-structuralist modeling tradition which has inspired
most of the CGE literature (Dervis and Robinson, 1982). More specifically, the
equations of the model are derived from standard microeconomic assumptions
about the behavior of price taking agents. Consumers maximize utility subject to
their budget constraints. Producers choose inputs so as to minimize production
costs. Production technologies are characterized by a CES or Leontief function
whereby resources are limited and distributed by market forces.
The model consists of sixteen commodities, sixteen domestic industries and

two types of households: migrants and natives. In total, there are two agricultural
industries, four manufacturing industries and ten service industries. Each com-
modity corresponds to an industry. Considering two types of households allows us
to capture differences in the consumption behavior of natives and migrants. For
the empirical application of the model, we use the current input-output matrices
from Eurostat. This enables us to cover the recent developments in production,
trade, factor movements and the linkages between them.
In order to capture the effects of the European integration process, we applied a

three country model, which reflects one country and two regions: the EU and the
rest of the world (see Baas et al., 2007). The national economies considered are
linked to the EU and to the rest of the world via trade in goods and services, capi-
tal flows and the migration of labor. Transaction costs within the EU are lower.
We thus consider different trade patterns which emerge from European integration
and distinguish between intra- and extra EU trade.
Similar to the first model, we allow for labor market imperfections by applying
a wage-setting framework, which is new in the CGE literature, on the effects of
the EU Eastern enlargement (see Baas and Brücker, 2008). We apply the same
elasticities of the wage-setting curves as in the first model outlined in the previous
section. Moreover, we consider the adjustment of capital stocks based on the same
parameter values used in the first model.
The technical features of the model and the data sources used for the calibra-
tion of the model are described in detail in Baas et al. (2009).
4 Presentation of the Simulation Results
The presentation of our simulation results starts with the first model, which en-
ables us to analyze the impact of the 2004-2007 labor supply shock triggered by
emigration from the EU8 into the EU15 on macroeconomic aggregates, and also
on the distribution of earnings and employment opportunities for the different
groups of the labor force. These simulations cover the EU15 countries and the
EU8. We then continue by presenting the simulation results from the CGE model,
which also enable us to analyze beyond the macroeconomic aggregates the sec-
toral shifts in output and employment. This analysis covers a selection of impor-
tant sending and receiving countries, i.e. Poland and the UK.
In each scenario, we calculate the labor supply shock for the sending and re-
ceiving countries as outlined in Section 2.3. Throughout the analysis, we distin-
guish between the short-run impact of migration, i.e. the impact under considera-
tion of the incomplete adjustment of the physical capital stock, and the long-run
impact, when capital stocks have completely adjusted.
4.1 The Structural Impact on Labor Markets

and Income Distribution
Table 2 presents the simulation results of the migration impact on GDP, wages and
unemployment. The first column presents the size of the labor supply shock trig-
gered by EU enlargement: migration from the EU8 increased the labor force in the
EU15 by 0.4%, and reduced it in the EU8 by 1.16%. Particularly affected are Ireland
(+4.9%), the UK (+1.3%), Luxembourg (+1%) and Austria (+0.4%) in the EU15,
and Poland (-1.8%), Slovakia (-1.3%) and Lithuania (-1.1%) in the EU8). Note that
migration is associated with an increase in labor market participation; consequently,

the total labor supply in the enlarged EU has increased by 0.1%.
Table 2
The impact of EU8 migration on GDP, wages and unemployment
Factor income
GDP GDP per capita per native Unemployment Wages
Change of
labour Short-run Long-run Short-run Long-run Short-run Long-run Short-run Long-run Short-run Long-run
Changes in % (unemployment rate: changes in percentage points)
AT 0.42 0.31 0.34 0.00 0.02 0.12 0.15 0.02 0.02 -0.02 0.00
BE 0.22 0.11 0.17 -0.08 -0.02 0.01 0.07 0.07 0.05 -0.04 0.00
DE 0.10 0.04 0.10 -0.03 0.02 -0.01 0.04 0.03 0.01 -0.03 0.00
DK 0.23 0.13 0.20 -0.08 -0.01 0.00 0.07 0.02 0.00 -0.05 0.00
ES 0.19 0.03 0.11 -0.08 -0.01 -0.04 0.04 0.05 0.02 -0.04 0.00
FI 0.09 0.03 0.08 -0.06 -0.01 -0.02 0.04 0.03 0.01 -0.03 0.00
FR 0.01 0.01 0.01 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00
GR -0.01 0.00 -0.01 0.01 0.00 0.00 -0.01 0.00 0.00 0.00 0.00
IE 4.87 0.80 2.93 -2.07 -0.02 -0.77 1.31 0.87 0.37 -1.61 0.00
IT 0.11 0.04 0.08 -0.03 0.01 0.00 0.04 0.02 0.01 -0.03 0.00
LU 1.00 0.81 1.13 0.23 0.55 0.34 0.65 0.12 0.05 -0.25 0.00
NL 0.14 0.09 0.12 -0.03 -0.01 0.02 0.04 0.02 0.01 -0.02 0.00
SE 0.38 0.25 0.33 -0.01 0.07 0.05 0.12 0.05 0.03 -0.06 0.00
UK 1.28 0.50 0.89 -0.28 0.10 -0.05 0.34 0.21 0.11 -0.29 0.00
CZ -0.08 -0.07 -0.11 0.01 -0.03 0.01 -0.03 -0.02 0.00 0.03 0.00
EE -0.21 -0.09 -0.19 0.12 0.02 0.12 0.02 -0.04 0.00 0.06 0.00
HU -0.44 -0.34 -0.49 0.10 -0.04 0.10 -0.04 -0.04 0.00 0.11 0.00
LT -1.14 -0.55 -1.15 0.61 -0.01 0.61 -0.01 -0.32 -0.01 0.31 0.00
LV -0.43 -0.26 -0.46 0.17 -0.03 0.17 -0.03 -0.09 0.00 0.12 0.00
PL -1.77 -0.88 -1.94 0.90 -0.18 0.90 -0.18 -0.59 0.03 0.43 0.00
SI 0.26 0.15 0.21 -0.10 -0.05 -0.10 -0.05 0.02 0.00 -0.04 0.00
SK -1.34 -0.53 -1.51 0.82 -0.18 0.82 -0.18 -0.55 0.00 0.43 0.00
EU151) 0.36 0.13 0.26 -0.09 0.03 -0.02 0.10 0.06 0.02 -0.09 0.00
EU8 -1.16 -0.52 -1.10 0.65 0.05 0.65 0.05 -0.42 -0.02 0.25 0.00
Total 0.11 0.11 0.20 0.11 0.20 0.16 0.25 -0.03 0.00 -0.07 0.00

Notes: 1) Without Portugal
This labor supply shock results in an increase in the GDP in the EU15 of
0.13% in the short-run and by 0.26% in the long-run, while the GDP in the
enlarged EU increases by 0.11% in the short-run and by 0.20% in the long-run.
This is a substantial income gain, which increases the overall GDP in the enlarged
EU by about 24 billion euros. In the sending countries, however, the loss of pro-
duction factors results in a decline in the GDP of 0.52% in the short-run and by
1.1% in the long-run.
The GDP per capita increases in the sending countries due to a higher capital
endowment per worker in the short-run; while the long-run effects are by and
large neutral. In the receiving countries of the EU15, the effects are ambiguous:
on the one hand, migrants bring no capital by assumption, so the GDP per capita
declines; on the other hand, the higher labor market participation of migrants from
the new member states exerts a positive impact on the GDP per capita. Overall,
the GDP per capita in the EU15 declines slightly in the short-run by 0.09%;
whereas it increases slightly in the long-run by 0.03%.
For welfare considerations, the impact of migration on factor income is more

relevant than the GDP per capita. We find that the migration from the new member
states is neutral on the total factor income of natives in the EU15 in the short-run;
while it increases in the long-run by 0.1%. In the sending-countries, the total factor
income per native increases by 0.65% in the short-run, a effect which almost com-
pletely disappears in the long-run, when the capital stock has adjusted. In the entire
region, the total factor income, however, increases substantially (+0.25% in the
long-run) since the income gain of migrants is considered here.
The aggregate labor market effects of migration are surprisingly small. The un-
employment tends to increase in the receiving countries of the EU15 by a mere
0.06 percentage points in the short-run and by almost zero in the long-run; while it
falls in the sending countries by 0.42 percentage points in the short-run. This ef-
fect disappears in the sending countries when capital stocks have adjusted. Over-
all, migration reduces aggregate unemployment in the enlarged EU slightly in the
short-run. Similarly, aggregate wages tend to decline by a mere 0.09% in the
EU15 and to increase by 0.25% in the EU8 in the short-run, but remain constant in
the long-run once the capital stock has adjusted.
The labor supply shocks can affect different groups in the labor market in dif-
ferent ways. Table 3 presents the impact of migration from the new member states
into the EU15 by qualification groups. Note again that in this context, we have
grouped the immigrants from the EU8 which arrived from 2004 onwards by occu-
pation and linked these occupation groups to education levels in order to capture
its actual impact on the different segments in the labor market.
We find that the effects are very balanced across the different groups of the la-
bor force in the EU15. Less-skilled workers are slightly more affected in the short-
term (-0.1) compared to medium-skilled workers (-0.09) and high-skilled workers
(-0.07). In the long-run, high-skilled workers benefit slightly; while medium and
low skilled workers lose slightly (Table 3). Similarly, the unemployment rate of
the less-skilled tends to increase slightly more than that of the medium and high-
skilled in the short-term. In the long-run, the unemployment rate remains by and
large unchanged for all groups in the labor market of the EU15.
In the sending countries, we find again that the wage impact of migration is
relatively balanced across the different groups in the labor market: wages tend to
increase by 0.23% for the less and medium-skilled, and by 0.3% for the high-
skilled. The long-run effects are negligible. However, we observe a different pat-
tern for the development of the unemployment risks: the unemployment rate of the
less-skilled declines by 0.8 percentage points, by 0.4 percentage points for the
medium-skilled and by 0.2 percentage points for the high-skilled in the short-run.
In the long-run, only the unemployment rate of the less-skilled declines signifi-
cantly by 0.2 percentage points.
Table 3
The impact of EU8 migration on wages by skill group
All Low-skilled Medium-skilled High-skilled
Short-run Long-run Short-run Long-run Short-run Long-run Short-run Long-run
Changes in %
AT -0.02 0.00 -0.02 0.00 -0.02 0.00 -0.02 0.00

BE -0.04 0.00 -0.03 0.01 -0.03 0.00 -0.05 -0.01
DE -0.03 0.00 -0.03 0.00 -0.03 0.00 -0.03 0.00
DK -0.05 0.00 -0.05 0.00 -0.05 0.00 -0.05 0.00
ES -0.04 0.00 -0.03 0.01 -0.14 -0.09 -0.01 0.04
FI -0.03 0.00 -0.03 0.00 -0.03 0.00 -0.03 0.00
FR 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00
GR 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00
IE -1.61 0.00 -1.72 -0.19 -1.84 -0.23 -1.34 0.30
IT -0.03 0.00 -0.03 0.00 -0.03 0.00 -0.03 0.00
LU -0.25 0.00 -0.13 0.12 -0.14 0.11 -0.63 -0.38
NL -0.02 0.00 -0.02 0.00 -0.02 0.00 -0.03 0.00
SE -0.06 0.00 -0.05 0.01 -0.05 0.01 -0.08 -0.02
UK -0.29 0.00 -0.35 -0.07 -0.35 -0.06 -0.19 0.11
CZ 0.03 0.00 0.03 0.00 0.02 0.00 0.03 0.01

EE 0.06 0.00 0.07 0.01 0.06 0.00 0.06 0.00
HU 0.11 0.00 0.09 -0.01 0.10 -0.01 0.12 0.01
LT 0.31 0.00 0.32 0.02 0.30 -0.01 0.33 0.01
LV 0.12 0.00 0.11 0.00 0.11 -0.01 0.13 0.01
PL 0.43 0.00 0.41 0.01 0.39 -0.03 0.51 0.06
SI -0.04 0.00 -0.06 -0.02 -0.04 0.00 -0.03 0.01
SK 0.43 0.00 0.36 -0.02 0.41 -0.02 0.49 0.05
EU151) -0.09 0.00 -0.10 -0.01 -0.09 -0.01 -0.07 0.02

EU8 0.25 0.00 0.23 0.00 0.23 -0.02 0.30 0.03
Total -0.07 0.00 -0.09 -0.01 -0.08 -0.01 -0.06 0.03

Altogether, migration from the new member states has only a moderate impact
on the distribution of wages and unemployment risks among different groups in
the labor markets of an enlarged EU, which can be inter alia traced back to the
fact that the skill structure of migrants from the new member states is relatively
balanced.
Finally, Table 4 presents the impact on natives and foreigners in the labor mar-
ket. Here we do indeed observe a distinct difference in the wage effects of migra-
tion: while native wages tend to decline in the EU15 only by 0.07% in the short-
run and to increase by 0.02% in the long-run, our simulation results suggest that
the wages of the foreign workforce tend to decline by 0.41% in the short-run and
by 0.34% in the long-run. Analogously, the unemployment risks are unequally
distributed between foreigners and natives: the unemployment rate of the foreign
workforce tends to increase by 0.19 percentage points and that of the native work-
force by 0.03 percentage points in the short-run. In the long-run, the unemploy-
ment rate of the foreign workforce increases by 0.16 percentage points, while that
of natives remains stable.
Table 4
The impact of EU8 migration on native and non-native wages and unemployment
Non-native
Native wages Non-native wages Native unemployment unemployment
Short-run Long-run Short-run Long-run Short-run Long-run Short-run Long-run
Changes in % (unemployment rate: changes in percentage points)
AT 0.01 0.03 -0.30 -0.28 0.00 -0.01 0.11 0.10

BE -0.03 0.01 -0.21 -0.17 0.01 -0.01 0.49 0.45
DE -0.02 0.01 -0.17 -0.14 0.01 -0.01 0.12 0.09
DK -0.05 0.00 -0.17 -0.13 0.01 -0.01 0.16 0.13
ES -0.04 0.00 -0.06 -0.01 0.04 0.00 0.20 0.15
FI -0.02 0.01 -0.50 -0.47 0.02 0.00 0.27 0.23
FR 0.00 0.00 0.00 0.00 0.00 0.00 0.01 0.01
GR 0.00 0.00 0.03 0.03 0.00 0.00 -0.03 -0.02
IE -1.58 0.03 -2.36 -0.74 0.59 0.07 2.10 1.87
IT -0.02 0.01 -0.17 -0.14 0.01 0.00 0.17 0.15
LU -0.20 0.05 -0.32 -0.07 0.03 -0.02 0.19 0.10
NL -0.02 0.01 -0.18 -0.16 0.00 0.00 0.14 0.13
SE -0.04 0.01 -0.43 -0.38 0.02 -0.01 0.37 0.33
UK -0.23 0.06 -2.89 -2.60 0.08 -0.02 3.28 3.21
EU151) -0.07 0.02 -0.41 -0.34 0.03 0.00 0.19 0.16

While we find no significant effects of migration on the distribution of earnings

and unemployment risks across the different education groups of the labor market,
immigration from the new member states has a significant impact on the distribution
of wages and unemployment risks between natives and foreigners. The different im-
plications of immigration from the new member states for the native and the foreign
labor force in the EU15 can be traced back to two facts: first, migrants from the new
member states are more similar to the foreign workforce with respect to observable
human capital characteristics such as education and work experience, and hence,
compete more in the same segments of the labor market. Second, according to our
estimation results, migrants and natives are imperfect substitutes in the labor market,
so that migrants tend to compete more with other foreigners than natives within the
same skill and experience cells of the labor market.
4.2 The Sectoral Impact of Migration from the EU8

into the EU15
Labor migration from the EU8 into the EU15 does not only affect wages and un-
employment risks at the aggregate levels of the economies involved, but also their
sectors in different ways. A complete understanding of the distributional effects of
migration also requires, therefore, an analysis of structural change triggered by the

EU’s Eastern enlargement, since this also involves benefits and costs for the af-
fected workers. To analyze the sectoral effects, we here apply the multi-sectoral
CGE model outlined in Section 3.2 to Poland and the UK.
The UK is the most important receiving country in the EU15 in absolute terms;
and Poland is the most important sending country. The UK had already opened its
labor market to migrants from the EU8 in 2004. Furthermore, the economic condi-
tions have been favorable there throughout the 2004-2007 period. In 2004, the un-
employment rate was at 4.7%, below the EU15 average, and GDP growth, at
2.8%, was slightly above the EU average. These favorable conditions changed
very little during the following years. According to the UK LFS, about 610,000
migrants from EU8 countries lived in the UK by the end of 2007. In our counter-
factual scenario, we estimate that migration would have been about 155,000 with-
out EU enlargement, and as a consequence, enlargement triggered an additional
stock of migrants of 455,000. The share of migrants in the labor force is larger
than their share in the population, since migrants from the new member states
have an extremely high participation rate (75%).
In contrast, the economic situation looked gloomy in Poland at the outset of ac-
cession. Low GDP growth of 1.5% in 2001 and 2002 resulted in high unemploy-
ment rates which peeked at 20% in 2002 and, at 19%, were still high in 2004. This
contributed to the relatively high emigration rates from Poland. According to our
estimates, more than 1.3 million migrants from Poland resided in the EU15 in
2007; while only 630,000 Polish migrants would have lived there in our counter-
factual scenario had enlargement not taken place. The difference accounts for
about 2% of the Polish workforce.
According to our simulations, the strong labor market shocks in Poland (-1.8%)
and the UK (+1.3%) decreased the GDP in Poland by 0.92% and increased it in
the UK by 0.86%. In the UK, our simulations show an improvement of the bal-
ance of payments: the increase in exports is stronger than the increase in imports
for both intra- or extra-EU trade. Nevertheless, intra-EU exports increase more
strongly than extra-EU exports (1.24% intra-EU to 1.09% extra-EU), while the
opposite holds true for intra- and extra-EU imports (0.81% intra-EU to 0.89% ex-
tra-EU).
In Poland, we can observe a worsening of the balance of payments: exports are

shrinking faster (1.25% intra-EU and 1.24% extra-EU) than imports (-0.8% intra-
EU and 0.81% extra-EU). In contrast to the UK, where we see only a small differ-
ence between intra-EU and extra-EU trade.
The trade effects are caused by a change in production structures. According to

our production function, intermediary goods are additional input factors. In an
open economy not all intermediaries are available within a country. Therefore,
imports have to react to a change in production, which is in turn triggered by
changes in factor endowments. Beside the channel of intermediaries, exports react
to a change in the real exchange rate. To meet demand in the goods market, a
change in production has to be accompanied by a change in prices. A general
price change affects the exchange rate and thus trade.
Table 5
Simulation results, key macroeconomic figures
UK Poland
Changes in %
GDP 0.86 -0.92
GDP per capita -0.03 0.81
Exports Intra-EU 1.24 -1.25
Exports Extra-EU 1.09 -1.24
Imports Intra-EU 0.81 -0.80
Imports Extra-EU 0.89 -0.81
Wages -0.34 0.32
Changes in
percentage points
Unemployment rate 0.13 -0.48
Source: Own estimates.
In our model, trade and sectoral effects strongly interact. On the one hand, a
change in labor supply of a tradable sector instantly affects imports and exports.
On the other hand, a change in the exchange rate affects the demand of goods of
both an industry producing tradable goods and an industry demanding tradable in-
termediaries.
We see a strong reaction by the manufacturing industry in both the UK and Po-
land caused by the respective labor supply shocks. Manufacturing goods are
mainly tradable goods, and the intermediaries needed for production in this sector
are tradable, too. Consequently, a strong reaction in trade can be seen in a change
in production in this sector. In the UK, manufacturing production increases by
proportionately 1.1% more than the overall change in production of 0.8%. How-
ever, not only the tradable sectors benefit from the additional labor supply in the
UK. Some service industries like the health and social work sector and education,
also enhance their production at 0.9% above the average. The sectors which bene-
fit only modestly from the additional labor supply are agriculture, and mining and
quarrying.
In Poland, the production of the manufacturing sector is reduced by 1% (see

Table 6), which is above the average decrease in production of 0.9%. In contrast,
some industries producing domestic goods like hotels and restaurants and con-
struction reduce their production at 0.8% below the average. However, the sec-
toral effects are in general smaller in Poland than the shift in the sectoral structure
of the economy in the UK.
Table 6
Simulation results, sectoral impact
UK Poland
Changes in %
Agriculture, hunting and forestry 0.7 -0.8
Fishing 0.5 -0.9
Mining and quarrying 0.4 -1.0
Manufacturing 1.1 -1.0
Electricity, gas and water supply 0.7 -0.8
Construction 0.7 -0.8
1)
Wholesale and retail trade 0.9 -0.9
Hotels and restaurants 0.8 -0.8
Transport, storage and

0.8 -0.9
communication
Financial intermediation 0.6 -0.8
Real estate, renting and business

0.6 -0.8
activities
Public administration and defence;
0.9 -0.9
compulsory social security
Education 0.9 -0.9
Health and social work 0.9 -0.9
Other community, social and

0.7 -0.9
personal service activities
Activities of households 0.9 -0.7
Total 0.8 -0.9
Source: Own estimates.

Notes: 1) Includes also repair of motor vehicles, motorcycles and personal and household goods.
Again, our results predict moderate effects of migration on wages and unem-
ployment in both countries. The migration effect is mitigated by the partial ad-
justment of the capital-output ratio, by international trade and by the redistribution
of factors across sectors. Therefore, we observe that the increase in labor supply is
accompanied by an increase in the physical capital stock in the UK. Furthermore,
the balance of payments has improved there. In Poland, the capital stock corre-
spondingly declines while the balance of payment worsens. As a third effect, the
sectoral factor mobility assures that the new factor endowments are distributed to
their most productive use, which yields additional gains in factor productivity and
output.
6 Conclusions
In this chapter, we analyze the effects of migration on the distribution of factor in-
comes, employment opportunities and structural change. We find that the migra-
tion flows triggered by the EU’s Eastern enlargement has increased the GDP in
the integrated area by about 0.2%, or by about 24 billions euros.
The impact on aggregate wages and unemployment is surprisingly small. Our

simulation results indicate that wages in the receiving countries of the EU15 de-
cline by less than 0.1% in the short-run, while they remain stable in the long-run.
Similarly, the unemployment rate increases only slightly in the EU15 in the short-
run, but remains by and large unaffected by immigration in the long-run. Con-
versely, we observe increasing wage and declining unemployment rates in the
sending countries in the short-term; while the aggregate labor market effects of
migration are more or less neutral there in the long-term.
This moderate impact of migration can be traced back to two main factors:
first, capital stocks tend to adjust to labor supply shocks even in the short-term. In
the long-run, our empirical findings suggest that capital stocks adjust completely
to changes in labor supply, both in the receiving and sending countries. As a con-
sequence, aggregate wages und unemployment risks remain largely unaffected by
the emigration or immigration of labor. Second, the economies also adjust with
trade and sectoral change. The simulations in our second model, which consider
these adjustment mechanisms, find therefore particularly small effects of migra-
tion on labor markets.
Moreover, we find only a moderate impact of migration on the distribution of

wages and unemployment risks. This is caused largely by the relatively balanced
skill structure of the migrant population from the new member states. In contrast
to other groups of migrants, migrants from the new member states are character-
ized by a similar skill structure compared to the population of the receiving coun-
tries and are only moderately better qualified than the population in the sending
countries. However, migrants from the new member states are employed well be-
low their skill levels in the EU15. Nevertheless, this has been considered in our
simulations of the labor supply shocks. Overall, we find that the less-skilled are
slightly more affected by falling wages and increasing unemployment risks than
medium and high-skilled workers in the EU15. In the long-run however, the ef-
fects are pretty small, even for less-skilled workers. High and medium-skilled
workers are either unaffected by immigration from the new member states in the
long-run or tend to benefit.
There is, however, one important distributional impact of immigration from the
new member states: while the native workforce tends to benefit, foreign workers
which already reside in the EU15 tend to lose substantially. The wages of foreign
workers decline by about 0.4% in the short-term, while the unemployment rate in-
creases by about 0.2 percentage points. We thus find that natives in the receiving
countries tend to benefit from migration triggered by the EU’s Eastern enlarge-
ment, while foreigners tend to lose out. This can be traced back to our finding that
foreigners and natives are imperfect substitutes in the labor market, and that the
skill and work experience structure of the employed workforce from the new
member states is more similar to that of other foreigners than to that of natives.
Finally, our simulation results demonstrate that the sectoral structure of the
economy is affected in different ways by migration from the new member states.
In the receiving countries, particularly the manufacturing sectors benefit from
immigration; while these sectors contract in the sending countries.
Altogether, our findings suggest that the additional migration created by EU

Eastern enlargement has triggered substantial aggregate gains for the enlarged EU,
but it has affected the distribution of earnings and unemployment risks only very
moderately. Opening the labor markets in all EU15 countries, including Austria
and Germany, would therefore generate additional benefits for the enlarged EU
according to our simulations, without adversely affecting the receiving countries.
In contrast, the native population and the native workforce tend to gain from fur-
ther immigration from the new member states, while the foreign workforce tends
to lose there.
The financial crisis will, however, reduce the benefits from EU Eastern
enlargement. It is very likely that net migration from the new member states into
the EU will contract during the following years since migration is largely driven
by employment opportunities in the receiving countries. As a consequence, immi-
gration from the new member states will decline in the EU15 and return migration
will increase, although the unemployment rates are likely to increase in the new
member states as well.
It is speculative to assess whether the EU15 or the new member states will be
more than proportionately affected by the financial crisis. According to the latest
forecasts by the EU, GDP will decline less in important sending countries such as
Poland and Romania than in important destination countries such as the UK and
Spain. This would further contribute to the contraction of net migration flows.
These developments will generate further benefits for natives in the receiving
countries: a part of the employment decline in the most affected receiving coun-
tries will be absorbed by the adjustment of migration flows. However, net return
migration would generate additional costs to the sending countries: the unem-
ployment rate will increase there compared to a case without return migration.
Nevertheless, from the perspective of the enlarged EU, open labor markets can
contribute to reduce unemployment if a part of the macroeconomic shock is ab-
sorbed by increasing labor mobility.
References
Baas, T./H. Brücker/E. Hönekopp (2007), EU-Osterweiterung: Beachtliche Gewinne für

die deutsche Volkswirtschaft, in: IAB-Kurzbericht 6/2007.
Baas, T./H. Brücker (2008), Macroeconomic Impact of Eastern Enlargement on Germany
and UK: Evidence from a CGE Model, in: Applied Economics Letters, forthcoming.
Baas, T./H. Brücker/A. Hauptmann/E. J. Jahn (2009), Labour Mobility within the EU in the
Context of Enlargement and the Functioning of the Transitional Arrangements: The
Macroeconomic Consequences of Labour Mobility. Background Report, IAB, Nürn-
berg.
Barrett, A./D. Duffy (2008), Are Ireland’s Immigrants Integrating into its Labor Market?,
in: International Migration Review 42(3), 597-619.
Blanchflower, D. G./A. J. Oswald (1994), The wage curve. Cambridge et al.
Boeri, T./H. Brücker (2005), Why are Europeans so tough on migrants?, in: Economic Pol-
icy 20(44), 629-703.
Borjas, G. J. (2003), The Labor Demand Curve is Downward Sloping: Reexamining the
Impact of Immigration on the Labor Market, in: The Quarterly Journal of Economics
118(4), 1335-1374.
Brücker, H./E. J. Jahn (2008), Migration and the Wage Curve: A Structural Approach to
Measure the Wage and Employment Effects of Migration. IZA Discussion Paper No.
3423, Bonn.
Brücker, H./A. Damelang (2009), Labour Mobility within the EU in the Context of
Enlargement and the Functioning of the Transitional Arrangements: Analysis of the
Scale, Direction and structure of Labour Mobility. Background Report, IAB, Nürnberg.
Dervis, K./S. Robinson (1982), A General Equilibrium Analysis of the Causes of a Foreign
Exchange Crisis: The Case of Turkey, in: Review of World Economics 118 (2), 259-
280.
Hamilton, B./J. Whalley (1984), Efficiency and Distributional Implications of Global Re-
strictions on Labour Mobility: Calculations and Policy Implications, in: Journal of De-
velopment Economics 14(1), 61-75.
Kaldor, N. (1961), Capital Accumulation and Economic Growth, in: F. A. Lutz/D. C.
Hague (eds.), The Theory of Capital, New York.
Layard, R./S. Nickell (1986), Unemployment in Britain, in: Economica 53, 121-169.
Layard, R./S. Nickell/R. Jackman (2005), Unemployment: Macroeconomic performance
and the labour market (2nd Edition). Oxford et al.
Levine, P. (1999), The welfare economics of immigration control, in: Journal of Population
Economics 12(1), 23-43.
Lindbeck, A. (1993), Unemployment and Macroeconomics, Cambridge, MA.
Longhi, S./P. Nijkamp/J. Poot (2006), The Impact of Immigration on Employment of Na-
tives in Regional Labour Markets: A Meta-Analysis. IZA Discussion Paper No. 2044,
Bonn.
Longhi, S./P. Nijkamp/J. Poot (2005), A Meta-Analytic Assessment of the Effect of Immi-
gration on Wages, in: Journal of Economic Surveys 19(3), 451-477.
Nijkamp, P./J. Poot (2005), The last word on the wage curve?, in: Journal of Economic
Surveys 19(3), 421-450.
Ottaviano, G./G. Peri (2006), Rethinking the Effects of Immigration on Wages. NBER
Working Paper No. 12497.
Shapiro, C./J. E. Stiglitz (1984), Equilibrium Unemployment as a Worker Discipline De-

vice, in: American Economic Review 74 (3), 433-444.
Untiedt, G./B. Alecke/T. Baas/G. Biffl/H. Brücker/O. Fritz/B. Gardiner/E. Hönekopp/P.
Huber/A. Lamour et al. (2007), Auswirkungen der EU-Erweiterung auf Wachstum und
Beschäftigung in Deutschland und ausgewählten EU-Mitgliedsstaaten. Bundesministe-
rium für Wirtschaft und Technologie, Berlin (Project 33/05).
Upward, R. (2009), Labour Mobility within the EU in the Context of Enlargement and the
Functioning of the Transitional Arrangements: Country Study: United Kingdom. Final
draft, Leverhulme Centre for Research on Globalisation and Economic Policy, Notting-
ham.
Wong, K. (1995), International trade in goods and factor mobility, Cambridge et al.
.
3. Post-Enlargement Migration and Public
Perception in the European Union
Marcel Canoy
Anna Horvath
Agnès Hubert
Frédéric Lerais
Myriam Sochacki
Marcel Canoy et al.
This chapter analyzes the role of public perception in intra EU migration. Differ-
ent member states have different histories, cultures and political realities vis-a-vis
migration in general and intra EU migration in particular. This leads to a role of
public perception that varies between topics and between countries. The paper
analyzes the veil, Poles in the UK and Roma in Italy.The study reaches four con-
clusions. First, there is a clear link between public perception and migration poli-
cies. It is clear that Roma migrants are treated differently in Italy than in other
similar member states with relatively small native Roma population irrespective of
their numbers or profile. The same picture emerged in the discussion on the veil.
Second, the dual causality of migration and public perception was confirmed. In
particular, the case of Poles in the UK shows that public perception prior to arri-
val of Poles was more negative than afterwards, partially because initial fears
were based on imaginary things that might have happened after migration. Third,
there is also a link between performance of migrants on the labor market and per-
ception. With Poles in the UK case, the causality mainly runs from performance to
perception. In the case of Roma in Italy, the causality was mainly from perception
to performance, as it is in the case of the veil. Fourth, there has been a silent but
noticeable recent positive shift in public perception on migration. It is, however,
too early to conclude that there is a clear trend for three reasons. First, the trend
is very recent. Only data from the last two years show the trend. Second, there
could be a backlash resulting from the financial crisis. Third, there are still large
differences across member states, with some member states seeing a worsening of
the tone in the debate. The discussion on Roma in Italy underlines this point.

Migration, DOI 10.1007/978-3-642-02242-5_3,
72 Marcel Canoy et al.
1 Introduction
This chapter builds on Beutin et al. (2007), who show in general terms the impor-
tance of public perception in the success or failure of migration policies in Europe.
The value added by this contribution was that it pointed to an issue often over-
looked in policy discussions within an EU context, particularly where 27 member
states have to share common views on migration with the Schengen agreements.
Public perception of migration is not consistent across the member states. Polls
on attitudes towards migrants show large differences between member states, in-
come groups and social classes, and between the types of fear and apprehension
that migration arouses. However, despite the differences between member states
and some notable exceptions, the overall conclusion is that although the picture is
mixed, public perception of migration tends to be increasingly negative through-
out Europe.
This article links public perception more concretely with intra-EU migration.
By comparing different member states and different situations, we show that out-
comes (e.g. in terms of migrant’s levels of integration) differ wildly across
Europe. These differences cannot be solely explained by traditional methods such
as comparing employment levels, but can also be attributed to the role of public
perception.
The paper builds on Beutin et al. (2007) in four ways. First, it provides more
detail by looking at concrete issues (e.g. Roma in Italy, Polish workers in the UK,
or discussions on the wearing of a veil). Second, it updates data from surveys and
opinion polls. Third, it focuses on intra-EU migration. Fourth, it provides a cross-
country comparison. This enables us to show how public perception matters.
The chapter is organized as follows. Section 2 summarizes the aforementioned

paper by Beutin et al. (2007) pointing at the broader context and potential links
between public perception and policy in different areas. Sections 3 and 4 delve
deeper into the role of public perception in policy, theory and surveys. Sections 5,
6 and 7 are case studies. Section 5 discusses the policy and public debate around
the veil in various member states. While perhaps deviating from the overall topic
of the book, the section neatly reveals the broader importance of perception to-
wards migration policies in a concrete way. Section 6 analyzes the migration of
Poles to the UK. Section 7 does the same for Roma in Italy. Section 8 concludes.
3. Post-Enlargement Migration and Public Perception in the European Union 73
2 Migration and Public Perception –

Insights from Beutin et al. (2007)
Although there are many different motivations to immigrate to Europe or within

Europe, a human being’s natural search for a job that offers the prospect of a bet-
ter life is uppermost. Even the pressure to reunite families has roots in labor mi-
gration, albeit indirectly. The public perception is that migration is by and large a
supply driven phenomenon. Evidence shows that this perception is not entirely
correct. Whatever the precise reason, the demand for labor which migrants can
meet is often the source of people’s decision to migrate and of their choice of des-
tination. As long as there are jobs for migrants, they will try to come, one way or
the other. This phenomenon explains a substantial portion of migration to Europe
in the past. It explains the presence of Mexicans in the US today; and will explain
migration pressures on Europe in the future.
As Beutin et al. (2007) argue, European labor markets will need migrants in the
coming decades. Europe needs to be an attractive place to work for the highly
skilled, and to a certain extent, migrants with a large variety of skills are needed to
partially offset demographic trends and to be employed in areas where there is an
insufficient supply of local labor (such as for skilled and unskilled jobs in the car-
ing sector). While migration alone is not sufficient to meet the future demand for
labor, increasing labor participation by the native population - often erroneously
mentioned as an alternative to ‘low-skilled’ migration - does not suffice either, or
in the words of the former European Commission Vice-President Frattini: “The
new key message is: Europe will need more migration, since labor and skills
shortages are already noticeable in a number of sectors and they will tend to in-
crease” (Frattini 2005).
Migration is potentially beneficial for receiving countries and societies, as well

as for sending countries and migrants themselves. Immigrants can make valuable
contributions by solving labor shortages, increasing labor market efficiency, and
acting as catalysts for job creation and economic growth. Immigrants can enhance
cultural diversity and stimulate interest in other cultures and regions of the world
among native populations. Immigrants have played such a role for Europe
throughout history. In addition, they can make important contributions to the eco-
nomic development in their countries of origin by means of remittances, by pro-
moting the investment in education or by building networks of knowledge.
Reaping the benefits from migration is, however, far from automatic. The
speed with which migration takes place requires careful management. In the ab-
sence of such careful management, migration can accentuate existing problems
with social housing, schools, labor markets and feelings of insecurity. Some of
these problems are already felt and visible today, but none of these problems is
linked to migration per se. However, failing to act today could result in these
problems worsening with serious consequences for European societies.
2.1 Analyzed Areas
In their analysis, Beutin et al. (2007) pay special attention to the following areas.
Integration is the key to beneficial immigration. Labor market integration is es-

pecially important as it encompasses other aspects of integration (e.g. language,
culture). The link between public perception and integration policies is recipro-
cated: a lack of sufficient integration on the part of migrants and their descendants
in the receiving country has repercussions on public attitudes; and conversely,
public perception and resistance to change in society often makes integration more
difficult. Thus, integration is a two-way responsibility. Migrants need to make
more a serious effort to integrate, while receiving countries need to promote inte-
gration. The latter involves labor market reform which yields equal opportunities
to all; while public perceptions of migrants as ‘welfare scroungers’ or ‘job
thieves’ need to be tackled.
Illegal migration is dangerous for the people involved, creates incentives for il-
legal activities and impacts negatively upon public opinion, damaging perceptions
of legal migrants in almost equal measures. Breaking this cycle requires Europe to
offer some opportunity for legal migration to be possible beyond the strict condi-
tions attached to family reunion. This provides a legitimate route of entry for some
‘bogus’ asylum seekers or irregular migrants who would prefer to enter as legal or
regular migrants. For such an expansion of migration opportunities to be benefi-
cial and politically feasible, careful selection of new and existing migrants is nec-
essary on the basis of employability, and the size of the informal economy needs
to be reduced as well.
The asylum channel was used increasingly as a response to global migratory

pressure, but it was also used by many who did not require protection. This caused
serious handling problems resulting in backlogs, long decision procedures, rises in
legal costs and the overloading of domestic court systems. The administrative fail-
ings made asylum a favorite subject of media critics and the result was a general
and growing malaise in public opinion towards migrants in general. Given that
asylum application numbers are now at their lowest level for some years, the EU
has a good opportunity to make positive changes to public perceptions and posi-
tively influence the current debate on migration and asylum.
Security is an additional aspect which often frames public perception of migra-

tion. Terrorism and crime are serious security threats that need a robust and ap-
propriate response. Both may in certain instances have links to migratory phe-
nomena, but they are not intrinsic problems of migration itself. Were the public
discourse to ‘securitize’1 migration and perceive immigrants mainly as security
risks, it would become much more difficult for migrants to integrate. Effective
migration management has to strike a balance between the human security of mi-
grants and their situation in both receiving and sending countries and the impor-
tant duty of the EU and its member states to protect the security of its citizens.
Gender issues need to receive more attention as women play a pivotal role in
successful integration. The role and vulnerabilities of first and second generation
female migrants often remains overlooked. Improving their status by empowering
them with equal opportunities and rights should help integration in European so-
cieties. It may result in cultural differences being debated with the full participa-
tion of everybody concerned. This would have a positive effect both on percep-
tions and unlocking potential areas of conflicts linked to gender relations.
2.2 Results
Beutin et al. (2007) convey five main messages. First, European labor markets
need migrants in the future, which puts pressure on the need to effectively manage
migration. Neither the current profile of migrants, nor the level of integration or
public perception is in a state that allows Europe to reap the benefits of migration.
Second, the most immediate concern is breaking the negative cycle of hostility
towards migration leading to restrictive policies dealing with legal migration
which in turn lead to a diversion of migratory flows to asylum seekers or illegal
migrants. Breaking the cycle requires a sufficient quantity of legal migration be-
yond family reunion. Third, a number of critical conditions must be met for such
an expansion to be beneficial and politically feasible: these should include selec-
tivity, improved integration efforts and a reduction of the informal economy.
Fourth, none of these policies will be effective unless issues relating to public per-
ception of migration are explicitly addressed. Justified concerns have to be tackled
and misperceptions have to be cleared up without presenting an overly optimistic
view of the migration challenge. Finally, the need to put migration in a broader
context implies that non-migration policies are as important as migration policies.
On a global basis, policy areas such as development, trade, foreign relations, agri-
culture and fisheries should be seen through migration standpoint. On a national
basis, the same applies for labor market policies, education, gender and housing
policies.
1 ‘Securitization’ refers to the process of turning a policy issue into a security issue, thus
presenting its substance as a potential threat (See e.g. Buzan et al. 1998).
3 The Role of Public Perception in Policy-Making
There are various ways in which public perception influences politics and policy-
making. In the most obvious way, public perception matters for democratic rea-
sons. If politicians defy public opinion, it could lead to them being voted out of
office. However, this does not mean there is a one to one relationship between
popular opinions and policies. Often policies and policy-makers have a longer
time horizon than opinion polls. Politicians have to make a trade-off between
short term gains in the polls and longer term gains that result from solid policies.
This central driving force could have several different implications.
Reactive: Public perception drives policy. Some politicians treat public per-
ception in a reactive way. They consider opinion polls and the general current atti-
tudes and then act accordingly. Such behavior is most prominent among populist
parties, but it is by no means limited to them. Reactive policy-making can have
both positive and negative consequences. On the negative side, suggestive stories
in tabloids can lead to panic policies, e.g. on asylum seekers. For example, The
Daily Express once came up with the title ‘Asylum seekers ate my donkey’, a
phony story about some lost donkey, which in the end did not have anything to do
with asylum seekers at all. If tabloids shape perception and politicians act accord-
ingly, one can see stricter asylum rules as a result, which are in the end based on
tabloid headings rather than facts. Nevertheless, reactive political behavior can be
positive as well if it follows valid information. For example, the public outcry af-
ter the suffocation of Chinese illegal migrants in trucks created a momentum for
policy to tackle trafficking. Certainly one also has to consider the balance between
the weight of such information and the policies that follow.
Pro-active: Policy drives public perception. The causality does not always
run from perception to policy making: policies themselves can also have an im-
pact on public perceptions. Again, positive and negative examples can be equally
cited. Let us assume that policy makers created a successful policy. For example,
consider that in member state X integration policies increased the labor market
participation of migrant women by 30%. This will shape the general perception
towards migrants (as contributors to society), which will then have ramifications
for migrants and other migration policies even if these are de facto unrelated to the
initial successful policy. In contrast, a negative variant would be an unsuccessful
policy to stop illegal immigration, creating negative feelings not only towards il-
legal but also legal migrants.
It is important to keep this distinction in mind when analyzing migration. For

instance, if we analyze the migration of Polish workers to the UK, one can say
that it is policy which drives perception. The policy of opening up was the trigger
that shaped perceptions. Subsequently, the perception that emerged fed into fol-
low-up policies and policies elsewhere in Europe. When analyzing Roma in Italy,
it is the other way around: it is public perception that drives policy.
3.1 Public Perception and Social Disorganization Theory
If we want to single out public perception as a driving force of the success of pub-
lic policy, we first have to put it into a perspective. If we want to analyze the per-
formance of migrants (e.g. in the labor market) and also try to understand why in
some countries, regions or periods, it seems to work, while in others it does not
seem to work, we need to analyze the underlying causes of social success and
failure in communities. This requires us to look beyond economic theory.
Social disorganization theory deals with the inability of a community to inte-

grate (i.e. realize some minimum threshold of common values and effective social
controls; see Kornhauser 1978, Sampson and Raudenbush 1999). Determinants
which drive such inability mentioned by Sampson and Groves (1989) are sparse
friendship networks, unsupervised teenage groups and low organizational partici-
pation.
But these factors by themselves, although illuminating, do not answer the ques-
tion ‘why?’. Further research digs deeper into the underlying causes of these de-
terminants (Veysey et al. 1999) to answer why we witness some social groups
with these characteristics and not others? Here public perception comes into play.
It is interesting to see why groups with similar socio-economic, ethnic and cultural
backgrounds ‘perform’ differently in different member states. We claim that pub-
lic perception is one of the categories which explain these differences.
The following hypothetical example highlights the underlying assumptions of

such theories. In this example, both Member State A and B have minority com-
munities X with similar backgrounds and size. There is some tension in both
member states. Person-X performs worse in the labor market than other minority
groups and has higher crime rates as well.
Member State A has a strong populist movement, which hits hard at both per-
son-X as well the government, which fails to act. Member State A has two routes
now available. The first one involves a weak government that follows the popu-
lists. It implements negative policies and expels the criminal element, person-X.
This in turn will lead to stigmas being enforced and the worsening of underlying
factors such as unsupervised teenage groups and low organizational participation.
Integration records of X-man will not improve, but some of its consequences for
the original inhabitants of Member State A may be dampened by the strict poli-
cies. In the second route, Member State A has a strong government that imple-
ments a successful integration policy with the aim to silence the populists. As a
consequence of the success, the populists will look elsewhere and more support of
constructive integration policies emerges.
Member State B has a reasonably favorable public opinion on migration. There

are two potential routes here, too. With the first route, the government of Member
State B is strong and applies a successful policy, so that the situation of the X-man
community gradually improves. But a favorable public perception is not a guaran-
tee for success. Even under favorable circumstances, bad policies can shift public
perception downwards and erode support for migration policies.
4 Emerging Indications of a More Nuanced

Public Perception on Migration
Migration is neither a new nor a temporary phenomenon. It is here to stay, and it

cannot be constantly viewed as a long-lasting crisis. Over the last two years, the
issue of immigration has been increasingly debated and new policies have been
developed, both at the national and European level. Some most recent surveys on
public perceptions of migration and public opinion on migration policies show the
positive impact of talking about immigration-related issues. Public perception on
migration appears to be more nuanced than it used to be.
After European countries emerged as major immigration countries in the

1950s, the nature of migration went through deep and quick changes. The global-
ization of population flows has been accompanied by an increase in immigration
from North Africa, Sub-Sahara Africa and Asia, while the temporary guest work-
ers often turned out to become permanent residents and illegal migration grew to
be a more prominent trend. Integration, cultural diversity and security became key
issues.
These major shifts have occurred in an evolving global socio-political and eco-
nomic environment increasingly characterized by uncertainties and the highest
level of unpredictability. Fear of the unknown and the feeling of a growing vul-
nerability became key political challenges. These attitudes can be, and indeed,
have been used instrumentally by populist political parties.
Obviously these different developments have had a significant impact on the

public perception of migration. Public opinion shifted from openness towards a
mainly European labor force in time of a post-war economic revival to a more re-
luctant attitude towards a more diverse immigrant population. This pattern has re-
peated itself since 2004 in the attitude towards labor migration from new EU
member states - in time of economic difficulties and growing human security con-
cerns.
Migration became an important political issue playing a significant role in na-

tional elections. A recent survey for Transatlantic Trends: Immigration (Transat-
lantic Trends 2008) exploring public opinion on migration in the United States
and six EU member states – France, Germany, Italy, the Netherlands, Poland and
the United Kingdom – finds that when asked whether political parties' agenda on
immigration would influence their vote in the next elections, 50% of Europeans
answer "yes". The highest percentage is found in the UK, with 61% of the respon-
dents saying that it would influence their vote "a little" or even "a lot" (compared
to only 28% of Polish respondents).
However, when it comes to European elections, immigration does not seem to

play such an important role. A special Eurobarometer survey on the 2009 Euro-
pean elections (European Commission 2008b) shows that European citizens want
the campaign to focus first and foremost on economic matters (unemployment
47%; economic growth 45%; inflation and purchasing power 41%) and global and
security-related issues (crime 37%; terrorism 35%; combating climate change
33%). 32% of respondents want the campaign to focus on immigration. That being
said, a majority of Europeans continue to consider that decisions regarding immi-
gration should be made jointly within the EU (60%; European Commission
2008d).2
Over the recent years, surveys on public opinion in the European Union have
regularly found migration to be among the most important concerns for European
citizens immediately after economic and security-related issues. However, the Au-
tumn 2008 Standard Eurobarometer survey (European Commission 2008d) finds a
significant decline in concerns about immigration at the national level (immigra-
tion is now mentioned by only 9% of the respondents, down 12 percentage points
compared to 2006).
Concerns about immigration have also declined in the UK, although they re-
main high – one of the highest in Europe after Malta – and come as the third most
important issue (23%) after "crime" and the "economic situation" (34% each) .The
latter – "economic situation " – now tops the list of concerns at the European level
(37%, up 14 percentage points compared to 2006) together with "rising
price/inflation" (37%, up 21 percentage points compared to 2006) , overtaking
"unemployment" (26%, down 14 percentage points compared to 2006) and fol-
lowed by "crime" (17%, a fall of 6).
2 This means plus 2 percentage points compared to Eurobarometer 69 (European Commis-

sion 2008c).
For the first time, the Autumn 2008 Eurobarometer also asked respondents to
indicate the two most important issues they are personally facing. The impact of
the economic results is clearly reflected in European citizens' personal concerns,
with inflation being ranked first (50%) followed by the economic situation (23%).
Immigration is even mentioned less often as a personal concern (4% - position
twelfth out of the fourteen issues presented to the respondents) than as a national
concern (9% - seventh out of the fourteenth issues).
The persistence of global economic turbulences and the related rising concern
about purchasing power is certainly reflected in the decline of concerns on other
issues such as immigration. It may be also true that European public attitudes to-
wards migration have become more nuanced.
Over the recent years, migration has increasingly been at the top of the national
and European political agenda. It has been much talked and written about. A wide
range of policies to address this complex issue has been adopted at national and
European levels. This may well have led to a very nuanced and complex mixture
of skepticism and optimism about migration's effects as clearly shown by the 2008
Transatlantic Trends: Immigration survey. One of the key finding of this study is
that the public in the countries surveyed do not have clear conceptions of migra-
tion as either "a problem" or an "opportunity".
47% of European respondents in this transatlantic survey perceive migration to

be "more of a problem" than an "opportunity" (compared to 50% in the United
States). The UK is the most skeptic country, with 61% of respondents considering
immigration as being "more of a problem". The French and the Dutch hold the
opposite view with 46% and 42% of respondents, respectively, answering that it is
"more of an opportunity."
The main concern is very clearly related to illegal – not legal – migration. 45%
of Europeans express concern only about illegal immigration and just 3% are only
worried about legal migration; while as many as 29% of respondents are not wor-
ried about either illegal or legal migration. When asked whether they believe that
most of the immigrants in their respective countries are there illegally, 68% of
Italians and 53% of Poles answered yes, while 77% of Germans, 68% of Dutch,
60% of French and 46% of Britons considered that they were there legally.
Among policies to reduce illegal immigration, European respondents express
strong support for dissuasive policies such as the reinforcement of border controls
(80%), and the imposition of tougher penalties on employers who hire illegal im-
migrants (84%); however, they are also in favor of pro-active policies such as
making it easier for immigrants to enter a host country to work or study (70%),
and to increase development aid to poorer countries whose citizens immigrate il-
legally to Europe (67%). When asked how to deal with illegal immigrants, many
European respondents favor "deportation" (48%) over "legalization" (38%). With
regard to legal immigrants, a majority of Europeans (64%) are in favor of giving

them the opportunity to stay permanently rather than being admitted temporarily
and then be required to return to their country of origin.
As the issue of immigration is often linked to security and terrorism concerns,

it is interesting to stress that the Transatlantic Trends: Immigration survey finds a
close link between immigration and crime but not between immigration and ter-
rorism. Majorities in all European countries (66% in Italy, 61% in the Nether-
lands, 57% in Poland and 53% in UK) besides France (28%) believe that immigra-
tion will increase crime in their society in general. However, majorities in all
European countries do not believe that there is an increased likelihood of a future
terrorist attack in their country because of immigration. Only 35% of Europeans
agree that this is a possibility.
Migration-related economic concerns do not appear to be so strong in European

countries except in the UK. A strong majority (65%) of Britons believe that immi-
gration will lead to higher taxes as a result of increased demand for social services
by immigrants compared to the European average of 53% (56% of Dutch, 52% of
Poles, 51% of Germans, 51% of Italians and 45% of French). Even more striking
is the difference between respondents in the UK and in the other European coun-
tries surveyed: when asked whether they believed that immigrants take jobs away
from their respective countries, 52% of Britons agreed that immigrants were cur-
rently taking jobs away from native-born workers, compared to a European aver-
age of 34% (47% of Poles, 28% of Germans, 27% in the Netherlands and Italy,
and 24% in France). Unsurprisingly a majority of Britons (51%) are against "en-
couraging immigration for employment purposes" in the face of an aging society;
while all other Europeans surveyed are in favor of immigration to solve potential
shortages in the work force (67% of Dutch, 55% in France and Germany, 54% of
Italians, but with a more mixed picture in Poland with 43% of Poles in favor and
47% against).
Europeans also appear to be optimistic about immigrants' cultural influence.

65% of Europeans agree that immigration will improve their culture with new
ideas and customs. As integration is obviously a two-way street, Europeans expect
immigrants to know the national language of the host country. This criterion for
admittance is considered as important as having a job offer (87% of respondents
in both cases); while having a relative in the host country or coming from a coun-
try with a Christian heritage are not considered to be important criteria for admit-
tance. The great importance given to the knowledge of the national language is re-
flected in terms of integration policies in the strong support (85%) for free
language courses for immigrants. With regard to perceptions of national identity,
this is clearly reflected in the view that "being able to speak the language" is con-
sidered by 94% of European respondents as an important attribute for national
identity right after "respect for political institutions and law" (96%) and before
"having citizenship" (81%), "having feelings of nationality" (78%) or "having

been born in the country" (50%).
Europe exhibits openness to immigrants' cultural influence, but there is the ex-
pectation that immigrants will show a similar openness towards the host society.
This is illustrated in the strong support (94%) given to the promotion of the teach-
ing of mutual respect in schools.
More specifically concerning Muslim culture, 53% of Europeans do not agree

that Western European and Muslim ways of life are irreconcilable. However, they
are evenly divided on Muslims' contribution to their countries' culture (47% agree
that Muslims have a lot to offer to their countries' culture and 46% disagree) and
on the willingness of Muslims coming to their countries to integrate (47% agree
that Muslims want to integrate and 43% disagree). In fact, a July 2006 Pew Global
Attitudes survey (The Pew Global Attitudes Project 2006) found that large per-
centages of Muslims in Europe think of themselves first as a Muslim rather than a
citizen of their country. This is the case for 81% of British Muslims, 69% of Span-
ish Muslims, 66% of German Muslims, and only 46% of French Muslims.
This 2006 Pew Global Attitudes survey came in the wake of the rioting by
Muslim youth in France, the subway bombings in London and the controversy
over the Danish cartoons. Interestingly, it found few signs of backlash from West-
ern Europeans, who still mostly welcome immigrants from the Middle East and
North African countries. Indeed in the four EU member states surveyed – France,
Germany, Spain and the United Kingdom – immigrants from predominantly Mus-
lim countries were no less welcome than immigrants from Eastern Europe, with
the exception of Germany. Immigration from Middle East and North African
countries was considered "a good thing" by 62% of Spanish (compared to 60%
considering Eastern Europe immigration as "a good thing"), 58% of French (58%
also in favor of immigration from Eastern Europe), and 57% of Britons (58% in
the case of Eastern Europe) in sharp contrast with only 34% of Germans consider-
ing Middle East and North African immigration as "a good thing" (59% as "a bad
thing") and just 36% of Germans saying that Eastern Europe immigration was a
good thing (59% "a bad thing"). These findings show that: (1) as confirmed by the
2008 Transatlantic Trends: Immigration, Europeans do not associate immigration
with an increased risk of terrorism; (2) Europeans are sympathetic to the frustra-
tions of Muslim youth with their place in society and their economic situation3;
and (3) although most Europeans believe that Muslim intolerance was to be
3 49% of European respondents surveyed in the 2008 Transatlantic Trends: Immigration

agree that Muslim youth is frustrated by their economic situation, and the July Pew Global
Attitudes survey show that 64% of Germans, 62% of Britons, 46% of French – but only
37% of Spanish – were sympathetic to youth who felt frustrated by their place in French
society.
blamed for the controversy surrounding the Danish cartoons, they can neverthe-
less understand that Muslims felt offended by those cartoons.
This more nuanced public perception on migration is also reflected in the re-
sults of the Spring 2008 Eurobarometer (European Commission 2008c). When
asked about the contribution of immigrants to society, Europeans (EU27) are al-
most evenly divided: 44% of respondents agree with the statement that "immi-
grants contribute a lot to [their] country" (up 4 percentage point compared to
2006) and 47% disagree. However, there are very important differences between
countries: there are ten member states where (strong) majorities agree with this
statement (83% in Sweden, 67% in Portugal, 64% in Spain, 61% in Finland, 60%
in Luxembourg, 58% in Ireland, 57% in Denmark, 56% in the Netherlands, 54%
in Romania and 49% in France); while significant percentage of respondents in
Hungary (82%), Slovakia (81%), the Czech Republic (77%), Latvia (76%) and
Malta (75%) take the opposite view. The difference between the most positive at-
titude towards migration contribution to host society (Sweden) and the most nega-
tive (Hungary) is 73 percentage points.
Education, age, contact with immigrants and political preferences consistently

correlate with attitudes towards migration. The younger and the more educated
hold more positive views on migration. The 2008 Transatlantic Trends: Immigra-
tion survey finds that immigration is "more of a problem" for 40.7% of respon-
dents aged 18-24 compared to 49.3% of respondents over 65 and for 46.4% of re-
spondents with only a high school education compared to 27.8% holding a post-
graduate degree. The results of this survey also show that 54% of Europeans with-
out immigrant friends or colleagues say that immigration is "more of a problem"
compared to 42% of Europeans with at least a few personal or professional immi-
grant contacts. Similarly, the 2008 Eurobarometer survey finds that significant dif-
ferences depend on the place of birth of the respondents' parents: 64% of respon-
dents whose parents were both born in another EU country and 71% of those with
at least one parent born outside the EU and the other born in an EU member state
consider that "immigrants contribute a lot to their country". Only 42% of respon-
dents whose parents were both born in the native country and 45% of those with
one parent born in the native country and the other in another EU country hold the
same view. Finally, the political views play a significant role. A majority of those
on the left of the political spectrum hold more favorable views on migration (55%
agree that "immigrants contribute a lot to their country" and only 33% say that
migration is "more of a problem") compared to those on the right (53% disagree
that "immigrants contribute a lot to their country" and 58% consider that migra-
tion is "more of a problem").
The emergence over the last two years of a more nuanced public perception on
migration demonstrates the value of debating this issue. In this regard, the 2008
Transatlantic Trends: Immigration survey contributes some interesting evidence.
Some respondents were asked to define immigration as "more of a problem" or

"more of an opportunity" at the beginning of the survey, some were asked the
same question at the end of the survey, and a third group was asked the question
twice – at the beginning of the survey and at the end. On average, the percentage
of respondents saying that immigration was "more of an opportunity" at the end of
the survey was 7 percentage points higher than when they were asked at the be-
ginning. This is clearly a very significant message to policymakers.
5 Public Perception and the Veil
Women now make up the majority of migrants but their visibility in public policy
is still limited except for Muslim women who have made the front pages with is-
sues surrounding the veil. Debates around the veil or Islamic dress have been par-
ticularly fertile in interactions of a diverse nature between public perceptions and
policies. This section analyses how the hijab4 worn by some Muslim women and
girls has come to symbolize polarizing debates and fears which influences the in-
tegration of migrants and reveals characteristics of the relationship between Euro-
peans and "others".
While diverting from the overall object of the study (intra-European migra-
tion), the case is still highly relevant for our study since there are many different
countries where Muslim women wear veils, so that differences in perceptions that
color outcomes can be discerned.
Before getting into issues of a religious and political nature which are crucial to
understand the volatility and permanence of public perceptions, we will recall
some key facts about the changes in various member states as they were triggered
by events which shaped public perceptions.
5.1 When the Veil Makes Headlines
The public discourse concerning Islamic dress became a major public issue in
most member states following the tensions linked to the fear of Islamist radicals
post-2001 and after the London and Madrid bombings. Earlier on, the imposition
of the burqa by the Taliban in Afghanistan in the late nineties had been instrumen-
tal in creating a debate about the head and body covering of Muslim women in the
4 The word hijab comes from the Arabic word hajaba, meaning to hide from view or conceal,
containing also the meaning of modesty. Hijab has come to represent the variety of Mus-
lim women's outer garments, in particular the Muslim headscarf. For more details on defi-
nitions.
public sphere. The veil became a symbol of the backwardness of some Muslims in
their attitudes towards women. The Islamic headscarf was not an issue when Al-
gerian, Moroccan, Pakistani or Indonesian wives joined their husbands in the sev-
enties and eighties, but it recently became the subject of heated debate and legal
activity in France, Denmark, Holland, Germany and the United Kingdom. In other
countries such as Greece, Spain or even Austria, debates took place but the issue
did not give rise to public policy moves.
The reactions of governments, courts and the public to Islamic dress have dif-
fered across the EU. In France, headscarves were banned by law for students in
state schools in 2004. In Germany, regional laws banning the display of religious
symbols in schools have been implemented to prevent teachers in many Länder
from wearing headscarves. In the Netherlands, the controversial ex-immigration
minister Rita Verdonk proposed banning the all-covering burqa in public spaces
on the grounds that it disturbs public order, citizens and safety. In Sweden, the
right to freedom of religion was invoked by the national School Board rejecting
bans on headscarves. In the UK, Cabinet Minister Jack Straw's comment that the
full face veil is “a visible statement of separation and of difference” sparked fur-
ther heated debate.5 His position against the burqa was supported by the Italian
Prime Minister, Romano Prodi and by Wolfgang Schäuble, the German home sec-
retary, who declared himself against the burqa on the grounds that it prevents
communication. This question developed across Europe into a polemic out of pro-
portions to the issues involved.
5.2 Facts and Figures about Islam and Public Perceptions
History, size and distribution of Muslim populations as well as variations in eco-

nomic environments and legal systems make strongly heterogeneous contexts
within which the issue of the “wearing or not, banning or not” of the veil is raised.
It is nevertheless useful to assess public perceptions against the background of ex-
isting facts and figures.
The data on religious affiliation are not systematically collected in most mem-
ber states, so figures on the percentage of practicing or declared Muslims only
give a rough indication of the proportion of the population whom the issue might
concern. The Muslim population in member states widely differ: France (4-7% of
the population), the Netherlands (5.8%), Germany (3.2%) and the UK (2.8%) are
in the top league; whereas in the majority of member states the Muslim population
is around 1% or below: with 1.5% in Italy, less than 2% in Spain, 1% in Poland
and 0.08% in Lithuania (European Parliament 2007). Amongst these, the propor-
5 As far as public opinion is concerned, a 2006 poll found that 61% of people believed "Mus-
lim women are segregating themselves" by wearing a facial veil, but 77% believed that
they had the right to wear it.
tion of women wearing the veil is very low even if it has increased in recent years
as assessed by a group of academics working on Values, Equality and Differences
in Liberal Democracies (VEIL)6. A survey conducted in 20067 in France, Ger-
many, Great Britain and Spain finds that a majority of Muslim women in Great
Britain and Spain and under half of Muslim women in Germany cover their heads
every day or almost every day. About a third of Muslim women in Spain and 41%
in Germany never cover their heads. In France, a large majority (73%) declare
they never wear the veil.
What do we know about public perceptions about the wearing of a veil? A

2005 Pew Global Attitudes project survey in the Netherlands found that 51% of
the Dutch public favored banning head scarves in public; in Germany, 54% also
supported a ban; while in France 78% did so. In Britain by contrast, 62% thought
the banning of scarves a bad idea. Opinion was more evenly divided in Spain,
where 43% said the ban was a good idea and 48% opposed it.
5.3 The Situation in Some Member States
Numerous factors come into play in determining the importance of the issue of Is-
lamic dress across Europe. When considering the situation in different member
states, it is interesting to note for instance that in Poland or Italy, tolerance to-
wards the veil is similar for Muslim women and catholic nuns8; in other member
states such as Romania, a ban would be ruled out by the strong anti-discrimination
council; nor could it be envisaged in Hungary, a former part of the Ottoman Em-
pire. Why has the topic made headlines in recent years and given rise to heated
debates and court action in some member states but has gone almost unnoticed in
others? The cases mentioned hereafter underline the prominent events surrounding
the veil in countries where public perceptions and government action have inter-
played.
A ban in France on Muslim headscarves and other 'conspicuous' religious

symbols (Jewish skullcaps, large Christian crosses, Sikh turbans) for both teachers
and students at state schools was introduced in February 2004 after a wide-ranging
enquiry conducted by the 'Stasi Commission' (Commission de Réflexion 2003). In
6 The VEIL Project is an EU 6th framework program project which focuses on the debates,
conflicts and regulations concerning head and body coverings of Muslim women in the
public sphere in 8 European countries. See: www.veil-project.eu/
7 See the Pew Global Attitudes Project, Europeans Debate the Scarf and the Veil: “Except
in France, most Muslim women choose to cover their heads - but many among the general
public disapprove” The survey was conducted in April-May involving 400 Muslims in
France, 413 in Germany, 412 in Great Britain and 402 in Spain by Richard Morin and
Juliana Menasce Horowitz.
8 They are allowed to wear a veil even on passport pictures.
a country where the separation of state and religion is enshrined in the law, about
70% of French people, of those 40% were Muslim women, backed the controver-
sial law. The ban does not apply to religious schools or universities. According to
an evaluation in June 2005 (Bronner 2005), the majority of pupils agreed to with-
draw their religious symbols, but several associations pointed to 'quiet exclusions'
(i.e. abandoning of schooling, schooling in private schools or abroad; see EUMC
2006, 82).
Germany is the only other member state in which a ban is specifically targeted
by law. It is the right of every state (Land) to pass a law prohibiting the display of
religious references in schools within the limits of constitutional requirements.
This means that the legislature must take into reasonable account the freedom of
faith of the teachers and pupils affected, the right to education and the state's duty
to uphold ideological and religious neutrality.9 At least four German states have
banned teachers from wearing headscarves, and in the state of Hesse, the ban ap-
plies to all civil servants. Nevertheless, in September 2003, the Federal Constitu-
tional Court ruled in favor of a teacher who wanted to wear an Islamic scarf to
school.10 Students everywhere are allowed to wear headscarves, and a ban would
be widely considered unconstitutional. However, in April 2006, two 18-year-old
students were suspended from a school near Cologne for wearing all-covering
burqas/niqabs. The ban was not contested in court.11 These events are not only re-
stricted to the education system, in 2006 one District Court (Landgericht) judge
who wanted to wear a headscarf during a hearing was relieved of her duties as the
Chamber was of the opinion that the religious symbol endangered the neutrality of
the court, in particular with regards to the equal treatment of men and women.
UK: there are no general bans on Islamic dress. Public servants such as police
officers, parking attendants and medics are permitted to wear headscarves, so long
as their dress does not impede their work. However, institutions such as schools
(including state schools) are allowed to forge their own dress codes for students.
In March 2006, the House of Lords overturned a Court of Appeals ruling in favor
of Shabina Begum, who complained that her school sent her home for wearing a
jilbab. The case gained much publicity, in part due to Shabina being represented
by Cherie Booth (Blair) QC. The court said the school made sufficient conces-
sions by allowing the Islamic trousers and tunic, and that Shabina could go to a
school with a dress code permitting the jilbab should she so choose. After the
9 See the Judgment of the Second Senate of 24 September 2003: BVerfG, 2 BvR 1436/02
of 09/24/2003, paragraphs No. (1 - 138),
www.bverfg.de/entscheidungen/rs20030924_2bvr143602en.html
10 See the website of Federal Constitutional Court: Decision, September 24 2003, 2 BvR
1436/02, www.bundesverfassungsgericht.de
11 For more information, see:
www.bezreg−koeln.nrw.de/brk_internet/presse/pressemeldungen/archiv_2006/
presse_032_2006/index.html
Court of Appeals ruling, it became known that Shabina's brother Sherwas Rah-
man, who often spoke on her behalf, was a supporter of the extremist Islamist
group, Hizb-ut-Tahrir (HT), which took credit for having 'advised' Shabina.
The Netherlands: in October 2005, the Equal Treatment Commission ruled in

favor of an Arabic language teacher who was refused a post after refusing to wear
a headscarf whilst teaching.12 In 2006, the Dutch cabinet backed a proposal by the
country's immigration minister at the time, Rita Verdonk, to ban Muslim women
from wearing the burqa in public places. The cabinet said burqas disturb public
order, citizens and safety. An estimated 5% of people living in the Netherlands are
Muslim, but only a few dozen women choose to wear the burqa. This proposal for
a blanket style ban was the most radical.
Belgium: schools can ban certain religious symbols at their discretion. A 2004
survey found that out of 111 schools surveyed in Brussels, only 8 allowed teachers
or students to wear headscarves.13 In January 2007, the city council of the town of
Antwerp banned visible religious symbols for counter clerks dealing directly with
the public.
Sweden: on January 23rd 2007, the Swedish School Board ruled that the head-
scarf cannot be banned in schools due to considerations of rights to freedom of re-
ligion. Since 2003, however, the burqa can be banned on pedagogical or security
grounds. In supermarkets, cashiers and shelf-stockers are allowed to wear a head-
scarf.
Denmark: in the early 1990s, the government upheld the right of a hospital to
ban a doctor from wearing a headscarf for 'practical and hygienic' reasons and also
because 'they diverge from the hospital's normal standards for attire' (quoted in
Klausen 2005, 183). In 1999, a court sided with a supermarket chain which fired a
cashier for wearing a headscarf on the basis that it might get stuck in the register,
rejecting the defense's argument that this held also for the permitted baseball hats
and Christmas headgear. In 2005, a proposal by the Danish Peoples Party for a
parliamentary resolution forbidding public employees from wearing 'cultural-
related headgear' was withdrawn after indications that it might constitute indirect
discrimination.
5.4 The Stakes of the “Veil Controversy”
The discourse surrounding the wearing of Islamic dress has a strong political di-
mension, raising questions relating to freedom of expression, freedom of religion,
12 More information at www.cgb.nl (25/01/2006)

13 See: www.changement-egalite.be/spip.php?article765
the role of the state and gender equality and overall integration. The debate in the
context of the fight against terrorism also raises concerns of security. We will start
here to unfold some of these issues, including the following point that is often
overlooked. This debate represents not simply a discussion or a 'clash' between
cultures, but also divergences internal both to Muslim societies and to Europeans.
European Islam is in transition along with European societies. It is governed by
European laws, by European democratic values stressing individual freedom and
choice, but its members, and in particular women, are struggling with contradic-
tory pressures to find their space and identity. On the other hand, as underlined by
a group of academics working on Values, Equality and Differences in Liberal
Democracies (the VEIL project), the debates about headscarves are deeply em-
bedded in the current struggles to define European nations and relations to “other-
ness” in increasingly multicultural societies.
Security
Concerns of the niqab (facial veil) and the burqa (veil covering also the eyes) were
raised in individual stories reported in the press. In December 2006, Mustafa
Jama, 26, wanted for robbery and the murder of police constable Sharon Besh-
enivsky, fled the UK from Heathrow airport and returned to his native Somalia.
The picture of Britain's most wanted man, with 21 criminal convictions, had been
circulated to every police force, port and airport in the country. He got through by
disguising himself as a Muslim woman, adopting his sister's identity, using her
passport and wearing a niqab. Security personnel made no attempt to match his
face with her passport picture, even though the airport was on a heightened state
of alert after the 7/7 attacks. When challenged by the press, government sources
conceded that immigration officer surveillance of people leaving Britain had been
reduced to random checks to save money. And although both immigration control
officers and airline staff have the right to demand that a Muslim woman show her
face to help prove her identity, it seems that concerns not to offend religious cus-
tom often take precedence. Despite the argument of antiterrorist agencies that the
security of planes is unaffected, as any concealment of weapons or explosives in a
person shrouded in a head-to-toe chador would be detected by electronic sensors,
the story raised a new dilemma between the need for security and respect of the
individual's privacy and faith to coexist. Moreover, the security concerns of Is-
lamic dress are not confined to easily remedied lapses in high security environ-
ments such as border control points. In October 2006, a male suspect in a major
anti-terrorist investigation evaded capture in Britain for several days by dressing
in a burqa. If the niqab and the burqa were found to pose a real security problem,
then open discussion with the Muslim community could be engaged in. Let us re-
member that very few European Muslims wear these all-covering garments, and a
ban on such garments in high security spaces is not likely to cause much contro-
versy among Muslims or the wider public. Moreover, according to researcher
Jytte Klausen, most European Muslim leaders regard the burqa as an ethnic cus-
tom that is debilitating for routine social intercourse and detrimental to pluralist
coexistence (Klausen 2005, 182).
Islamic Dress and Women's Rights
Apart from the limited (but growing) proportion of the population that would be
affected by a ban on facial veils, such coverings also represent to many the most
offensive symbol of the oppression of women under religious pretext. Secular
Muslims warn of the dangers of cultural relativism, and of Western countries'
complicity in the rise of fundamentalism when governments tolerate the veil as a
religious or cultural symbol. They state that the veil represents the oppression of
women and is contrary to democratic values.
This point was underlined by the findings of the Stasi Commission in France,
which recommended banning 'ostentatious' religious symbols in state schools. The
Stasi Report identified the following positions with regard to wearing the Muslim
headscarf:
“For those wearing it, the headscarf can have different meanings. The
wearers may have exercised a free personal choice to wear it; or external
pressure may have forced them to do so. Most French people find this idea
of constraint or pressure particularly intolerable as it fundamentally con-
tradicts the principle of equality between men and women. Also when they
have to wear it against their wishes the headscarf can be seen as a way to
keep women hidden and subservient, and as a way to justify violence to-
wards women who choose not to wear it” (Commission de Réflexion 2003).
At the other extremity, the role of the veil as an instrument of women's eman-
cipation in fundamentalist Islamic societies has also been stressed. The argument
is that the private sphere represented by Islamic dress may be a safer and more
comfortable place to be. As Gita Saghal and Nira Yuval-Davis point out:
“One of the paradoxes is the fact that women collude, seek comfort, and
even at times gain a sense of empowerment within the spaces allocated to
them by fundamentalist movements. Being active in a religious movement
allows women a legitimate place in a public sphere which otherwise might
be blocked to them, and which in certain circumstances they might be able
to subvert for their purposes” (Saghal and Yuval-Davis 2000, 15).
So it is argued that thanks to the veil, women who otherwise would be confined
to the private domain can enter into the public and professional sphere. The exam-
ple of Iran, where female university students outnumber males is often quoted. As
shown later, however, the argument is of limited value in a European context
when looking at integration. It is not only unlikely to solve issues of discrimina-

tion against Muslims and violence against women, but in many cases it only rein-
forces those problems. In the wider Muslim world, however, women who consider
themselves feminists and believers most often accept the veil but challenge what
they consider as more fundamental offences such as misogynistic readings of the
Koran, and discrimination and violence against women, be it domestic or honor
crimes. In the case of European Muslims, the argument that the veil can be seen as
an instrument of emancipation only stands in terms of a way to 'achieve respect
equally regardless of physical appearance' (as noted by the Stasi Commission) and
as protection from male harassment.
Identity: European and Muslim
The fact is that increasing numbers of European Muslim women may actually
choose to wear Islamic dress is often explained by the role of the veil as an ex-
pression of cultural identity and a political tool in the context of the wars in Af-
ghanistan, Iraq and Lebanon as well as regards to the Palestinian question. As
Yasmin Alibhai-Brown explains: “What critics of Islam fail to understand is that
when they see a young woman in a hijab she may have chosen the garment as a
mark of her defiant political identity and also as a way of regaining control over
her body” (Alibhai-Brown 2000, 246). Bhikhu Parekh makes the same point:
“In France and the Netherlands several Muslim girls freely wore the hi-
jab (headscarf), partly to reassure their conservative parents that they
would not be corrupted by the public culture of the school, and partly to re-
shape the latter by indicating to white boys how they wished to be treated.
The hijab in their case was a highly complex autonomous act intended to
use the resources of the tradition both to change and to preserve it” (Parekh
1997).
Rising religious tensions in Europe are creating a vicious circle of lowering

tolerance and demand for integration on the part of non-Muslims, and in return a
more conscious and insistent expression of Islamic identity among Muslim popu-
lations. This can be seen in the rise of fundamentalism and, stemming from this,
greater pressure on women to wear Islamic dress. However, it must also be under-
stood in terms of women themselves choosing to express their distinct socio-
cultural as well as religious identity, and a rejection of Western values seen as
demeaning to women. This transition is particularly visible among young, politi-
cally active Muslim women, many of whom have adopted the headscarf as a po-
litical statement. Jytte Klausen points to the generational politics in this, noting
that prominent women leaders rarely wear a headscarf, while their mothers in
most cases did, and the opposite is true for many of the young activists (Klausen
2005, 173).
Identity: European and … English, French, Danish, …
The other side of identity involved here is, according to some political scientists,
identities of European nations resisting the threat to traditional values represented
by Europe, gender equality and multiculturalism. Researchers working on the
VEIL Project assume within the debates, conflicts and regulations on the covering
of Muslim women in the public sphere that the fundamental values and norms of
liberal democracies are not only being re-negotiated but collective identities are
also being re-constructed within the European integration process. They also es-
tablish that “the issue reveals the gendered nature of values and norms that widely
prevail in liberal democracies in Europe”.14 Birte Siim, professor of social sci-
ences at Aalborg University, describes the Danish situation:
“Many of the actors in the debates about veiling in Denmark are white,
ethnically Danish men who have not previously argued in favor of gender
equality; and often representing political parties which historically have not
been involved in any kind of feminist struggles. These men seem to suddenly
be very involved in whether Islam and Muslim headscarves are female op-
pressive. Their discourses are interesting examples of how feminism is in-
strumentalized in attempts to negotiate the nation as a white and ethnically
monolistic. By drawing upon my analyses from parliament debates and the
news media's coverage of debates about headscarves from the previous
decade, I will show how the debates about headscarves are deeply embed-
ded in the current struggles to define the nation and Danishness.”15
Integration
The low labor market participation of migrant women in general and their over-
qualification regarding the jobs they do has been established by the OECD
(OECD/SOPEMI 2007) and more recently by the RAND Corporation in a re-
search for the European Commission (Rubin et al. 2008). A recent study by the
Young Foundation in the UK (The Young Foundation 2008) concentrates specifi-
cally on Muslim women. It establishes that “British Muslim women are the most
disadvantaged faith group in the UK labor market, with 68% defined as inactive in
the labor market and only 29% in employment. The assumption that they have low
levels of education is not verified by data: only 22% of British Muslims do not
have a qualification compared to 29% of the general population. 36% of the Brit-
14 VEIL compares policy and media debates in eight European countries and the European
Union. The aims of the project are first to map out and compare the fundamental values
and political principles of the main actors in the headscarf debates, and second to explain
the differences and similarities in the conflicting values as expressed in the policies on
Muslim headscarves.
15 See: www2.iisg.nl/esshc/programme.asp?selyear=9&pap=5012
ish Muslim women respondents in the survey undertaken for the study have
graduate or postgraduate qualifications. “Second generation British Muslim
women who are young, generally well educated, is a strikingly underused resource
in the UK labor market.” So what is going wrong? Some of the barriers they meet
affect all women: gender discrimination, inflexibility and a lack of childcare; but
“British Muslim women face additional challenges, including discrimination
based on clothing and faith”. More specifically, “fifty% of women respondents in
work wear the hijab, which clearly identifies them as a Muslim and in the course
of the survey declared that they had missed out on progression opportunities be-
cause of religious discrimination and that the wearing of the hijab could have been
a factor”. The recommendations from this study start with the need first and fore-
most to tackle misperceptions of employers and the general public regarding the
educational levels and career aspirations of Muslim women.
5.5 Results
Has tension over the veil diminished in the recent period? It is symptomatic that
the recent decision of the European Court of Human Rights of 4/12/2008 in cases
Dogru v. France and Kervanci v. France, in which the expulsion from school of
two young girls who refused to remove their headscarf during sport classes was
confirmed, went hardly noticed and certainly did not provoke a flow of critical or
enthusiastic comments by politicians and the press, as it would have done a few
years ago around the time of the Danish cartoon controversy.
In addition, the fact that Muslim women themselves are increasingly express-
ing themselves publicly, that they are interviewed directly rather than spoken for
is good news. Their levels of education prove their worth and hopefully they are
progressively becoming more proactive in denouncing discrimination where they
meet it.
However, the fate of Muslim women and the waste of their resources for the
economy and society is only the tip of an iceberg of prejudices and misconcep-
tions of the general public, employers and politicians which have emerged in the
often heated debates about the headscarf. Surprisingly, while there has been a
“politicization” (Klausen 2005) of the issue, the political discourse on the causes
and consequences of the focus on the Islamic dress of Muslim women has been
minimal. A political discourse is nevertheless needed if this issue is not to be cap-
tured by populist parties. If integration and religious tolerance are to be effectively
addressed beyond primary perceptions, it is important to continue the debate on
Islamic dress.
6 Poles in the UK
The UK experience with migration and in particular with internal mobility within
the EU is of particular importance for various reasons: (i) a large number of mi-
grants from new member states came to the UK; (ii) despite fairly positive experi-
ences this did not imply opening up to the two newcomers (Bulgaria and Roma-
nia); (iii) the current economic slowdown will probably change the situation
(some public speeches have already hinted along this line). There is probably a
kind of “turning point”. Some British MPs for instance argued when the economy
was enjoying a boom, there was a case for an “open immigration policy” – but
now that most countries will undergo a long recession, the immigration policy
suitable for a boom may seem unsuitable for a recession.
6.1 Labor Mobility Following the Recent Enlargements
On May 2004, UK gave free access (with a mandatory workers registration

scheme for monitoring) to ten new member states but not for workers from Bul-
garia and Romania. The debate to decide to extend or not to Bulgaria and Roma-
nia is of crucial importance. For a large part, growth in the UK has been the con-
sequence of post-enlargement mobility flows. Figures from the Office for
National Statistics show the population grew by nearly two million to almost 61m
people between 2001 and 2007. Various official projections predict this to rise to
77m in 2051 or 110m in 2081 (Office for National Statistics 2006).
Four years after the 2004 enlargement and over a year after the accession of
Bulgaria and Romania, practically all of the available evidence suggests that the
economic impact of recent intra-EU mobility has been on balance positive, and it
has not led to serious disturbances in the labor market, even in member states such
as the UK, which have seen a relatively large inflow of migrants from the new
member states.
In terms of recent mobility from the EU10, Ireland and the UK have been the
main ‘receiving countries’, and to a lesser extent Austria and Germany. Neverthe-
less, despite their significant size, intra-EU mobility flows following enlargement
never acquired the ‘tidal wave’ dimensions initially feared by some observers. Be-
tween 2003 and 2007, the average population share of EU10 foreigners resident in
the EU15 increased from around 0.2% to 0.5% (European Commission 2008e).
Moreover, there is no indication that recent intra-EU mobility flows have ex-
ceeded the absorption capacities of the labor markets. In both the main receiving and
sending countries, local workers’ wages have continued to rise and unemployment
has declined since enlargement. Even when analyzing the isolated effects of migra-
tion and mobility on wages and unemployment, empirical studies have consistently
found very small impacts on local workers’ wages and employment.
In fact, mobility flows to the UK and Ireland, which appear to have peaked in
2006, declined significantly in 2007 and the first quarter of 2008. Indeed, there are
indications of increasing return flows, particularly from the UK. However, hidden
behind these macro-effects could be non-trivial local effects that color public per-
ception. Macro-numbers do not make tabloid headlines, unemployed locals that
‘suffer’ from Polish influx do, even if their total number is negligible on a national
scale.
From the perspective of the new member states, in particular the ‘high-
mobility’ ones, substantial outflows of workers are often perceived as a mixed
blessing. On the one hand, outflows have helped to reduce unemployment. On the
other hand, outflows of predominately young and high-skilled people have raised
concerns about ‘brain drain’ and labor shortages in the sending countries. But
there are a number of factors that may help to alleviate these problems, such as the
significant recent rise in higher enrolment rates for university education in most
new member states, the temporary nature of much of the mobility, the fact that
many of those who return often do so with improved working skills and interna-
tional contacts, which can be of benefit to the home country, and finally partial
substitution (in the case of Poland with Ukrainians).
Table 1
Share of foreigners resident in the EU27 in % of total population
UK EU15
EU15
2003 1.6 1.5
2004 1.7 1.6
2005 1.6 1.6
2006 1.5 1.7
2007 1.6 1.7
EU10
2003 0.2 0.2
2004 0.2 0.2
2005 0.3 0.3
2006 0.6 0.4
2007 1.1 0.5
EU2
2003 0 0.2
2004 0 0.2
2005 0.1 0.3
2006 0.1 0.3
2007 0.1 0.4
Non-EU27
2003 3 3.6
2004 3.1 4
2005 3.2 4.1
2006 3.6 4.2
2007 3.8 4.4
Source: European Commission (2008e).
Figure 1
Worker Registration Scheme registrations approved by nationality
(UK, May 2004 - September 2007)
250,000
200,000
150,000
100,000
50,000
0
2004 2005 2006 2007
Czech Rep Estonia Hungary Latvia Lithuania Poland Slovakia Slovenia
Source: Home Office (2008), Table 3.
6.2 General State of Public Perception

towards Migration in the UK
What has happened in terms of public perception during this period? We will fo-
cus on the perception of Poles, as they represent nearly two thirds of the EU mi-
grants coming to the UK. “Next Stopski London” is a key study (Fomina and Fre-
lak 2008). The main aim of the study is to present the image of Polish immigrants
as it appeared in the British press before Poland joined the European Union and
three years after.
The general attitude in the UK towards migration is a mixed picture. On the

positive side, the majority of migrants, or second or further generation migrants,
have been present in the UK for a long time (allowing them time to integrate),
they often speak the language (since they come from former colonies), and they
often have a job (since labor markets are flexible). The employment rate among
nationals from the eight Central and Eastern European countries that joined the
EU on 1 May 2004 stands at 84%. The figure is 9% higher than the UK-born av-
erage and is the highest employment rate of all immigrant groups in the UK (Insti-
tute of Public Policy Research 2008).
On the negative side, the UK has relatively high sales of tabloids, which tend to
be fairly hostile towards migrants. This is simply because these stories sell better
than friendly stories, although as we show below, the picture is more nuanced than
that. Moreover, the UK has reservations about Europe (also through tabloids but
not exclusively so). There has been no significant improvement in the attitudes of
UK citizens towards the EU. According to a Eurobarometer survey (European
Commission 2007), only half of the British people think that EU membership has
brought benefits: from the free market of goods, a cleaner environment and that
now the UK has more say in trade negotiations. However, there is bad news: (i)
fewer citizens see the positive effects of the EU on working conditions and the
benefits of the single market compared to the numbers registered in 2006
information about new EU initiatives only reaches one in two UK citizens at best
awareness about the availability of the EU’s information sources remains low; see
European Commission 2007). This attitude may have a general impact on more
specific attitudes towards intra-European migration. The housing market in the
UK is also fairly segregated.16
6.3 Public Perception towards Poles
The reaction of the public in the UK towards Poles (or more the general accession
of new member states) typically illustrates the role of public perception. To help
persuade skeptics, the UK government issued an estimate of the number of ex-
pected migrants from new member states. They came to an assessment of 5,000 to
13,000, a startling underestimation of the realized 605,375 people who registered
between May 1, 2004 and March 31, 2007 (of which 65% were Polish). In our
view, this underestimation was very likely deliberate. It also contrasted with some
results of studies forecasting bigger waves of migration (see European Commis-
sion 2008e). On the one hand, it was meant to be reassuring; and on the other
hand, it was used to pave the way for migration, which the government assessed to
be favorable but the general public at the time did not. The idea was that after mi-
gration had taken place, the general public would see the benefits and public opin-
ion would change. This is exactly the scenario that we described in Section 3
when public perception is unfavorable but the government is strong.
Fomina and Frelak (2008) describe the debate in that period extensively. Their
main conclusions are:
• One argument in favor of free flow of labor is based on what happened after
the accession of Southern European countries, and estimated numbers were
extrapolations from that. Obviously, it was not taken into account that the
wage gap between new member states and old member states was much larger
16 See e.g. www.jrf.org.uk/knowledge/findings/housing/pdf/2064.pdf

now than in the period of accession of the Southern European countries. Also,
the fact that Germany closed its borders was mentioned as an effect that in-
creased the number of Poles coming to the UK.
• Posing limits to social security benefits was one response of the British gov-
ernment to public concerns. The slogan was “You can come to the UK to
work, if you register, but you can not claim benefits”.
• Despite the disparity between ex-ante public perception and beliefs and ex-
post realizations, similar discussions appeared when discussing opening the
labor market for Bulgarian and Romanian workers (we will elaborate further
below).
• Poles mainly filled vacancies in the UK, thereby challenging the public belief
that they would “‘steal jobs from locals”.
• The tone in the debate from 2004 onwards was mixed, with negative tones in
tabloids and more positive one in the quality press. The negative stance to-
wards Poles was in part explained by anti-European sentiments. Some quotes:
“Seen through the narrow lens with which Britain suspiciously gazes at all
things that emanate from Brussels, the enlargement of the European Union
may seem to present little cause for celebration and big cause for anxiety”
(The Guardian, April 30, 2004); and “the business-hungry newcomers may
find they have exchanged one totalitarian regime for another. Brussels won’t
be brutal, but it will be bureaucratic” (The Sun, April 30, 2004).
• In fact, the main reason for a positive stance by the UK on enlargement is that
the UK saw enlargement as a useful countervailing power against perceived
dominance in Europe of Germany and France.
• There were significant changes over time. Comparing 2004 with 2007, the
most striking was a shift towards a more neutral tone, also notably in tab-
loids (see tables below). This is also exemplified by quotes: “Poles are
working everywhere. They are cleaning City offices, but they are also run-
ning them. They are building homes, but also buying and selling them.
Great success has come to some, hardship to others, and the majority lie in
between” (Times, June 16, 2007). And even anti-immigration newspapers
often presented positive images of Polish migrants. The Sun, for example,
was trying to placate its readers’ possible fears about Poles’ assimilation
and impact by drawing attention to the experience of the archetypal immi-
grant country, the USA: “America also has a large number of people de-
scended from Polish immigrants who went on to achieve fame and fortune”
(The Sun, April 30, 2004). The article went on to present its readers the
achievement of Liberace, the world famous pianist, Ruth Handler, creator of
the Barbie doll, Arthur Miller, the playwright, or Zbigniew Brzeziński, na-
tional security advisor to Jimmy Carter.
Table 2
Tone of the analyzed newspapers in % (2004)
Tone
Positive Negative Neutral Ambivalent Total

Newspapers
Tabloids 9 17.5 8 13 47.5
Broadsheets 19 5 17 11,5 52.5
Total 28 22.5 25 24.5 100%
Source: Formina and Frelak (2008), 31.
Table 3
Tone of the analyzed newspapers in % (2007)
Tone
Positive Negative Neutral Ambivalent Total

Newspapers
Tabloids 4 17 23 10 54
Broadsheets 11 4 25 6 46
Total 15 21 48 16 100%
Source: Formina and Frelak (2008), 31.
6.4 Interaction between Policy and Public Perception
One of the more interesting features is the combination of two insights. On the
one hand, Fomina and Frelak (2008) observe a positive shift in the debate. On the
other hand, the government (may have) deliberately underestimated the number of
Polish workers. If one analyzes this in absence of the public perception dynamics
discussed in the previous section, an apparent paradox unfolds. If people were
worried about the numbers and the numbers turned out to be ten times the number
estimated, why did the public perception shifted upwards?
There are two explanations here. On the one hand – as often the case is– the
fears were based on imaginary events of what might have happened after migra-
tion, such as mass crime, drinking or the unemployment of Poles (or Brits), none
of which was likely to occur. Since it did not occur despite the much higher num-
bers, it explains one part of the upward spiral. A second explanation is that once
the Polish migration became a fact, the perception was based more on realities,
which in turn led to a disassociation between Polish migration and anti-European
sentiments (which still persists).
So what we see here is negative perception to start with, positive policy and a
subsequent spiral of positive perception. One could be tempted to say that the
same will occur in the case of Romanians and Bulgarians, where the starting point
on perception is negative, as it was for the Poles back in 2004. This seems not to
be the case, however (Fomina and Frelak 2008). There could be four reasons for
this.
Firstly, it may be the case that the British government has reasonably assessed
that society has reached its saturation point for moment in terms of acceptance of
migrants from Eastern Europe. Secondly, nobody will ‘fall’ for a second underes-
timation ‘trick’. Thirdly, the British government might assess the added value of
these new Eastern Europeans as being low. Fourthly, this British government is
politically weaker than the previous one, therefore not willing to challenge public
perception.
7 The Securitization of Roma Migration in Italy
Italy is a relatively new destination for international migration. It became a net

immigration country in the eighties, and its annual growth rate to date of migrant
presence is one of the highest in the EU. Regular migrants amount to 6.2% of the
total population.17 Immigrants who arrive in Italy in the largest numbers are from
Romania, Albania and Morocco (OECD/SOPEMI 2008).
Italy lifted all labor market restrictions in 2006 towards all EU member states
which joined the EU in 2004. In the case of Romania and Bulgaria, Italy opened
its labor market in certain sectors – e.g. agriculture, tourism and domestic work –
immediately after these countries joined the EU in 2007. As a result, the legally
residing Romanian population rose by about 50% in 2007, which also meant that
Romania replaced Albania as the most significant country of origin for migrants in
the country (OECD/SOPEMI 2008).
The attitudes of Italians towards immigrants are, however, somewhat more

negative than the European average. According to Standard Eurobarometer 69,
36% of Italian respondents agreed with the statement that ‘immigrants contribute
a lot’ to Italy, while 57% disagreed with it (European Commission 2008c, 53).
However, the picture becomes more negative when it comes to public attitudes
towards ethnic groups: Italy is among the three European countries where respon-
dents would be the least comfortable with having a person from a different ethnic
origin as a neighbor (European Commission 2008a, 41). The public perception of
Roma immigrants is a good illustration for this discomfort.
17 See the website of the International Organization of Migration (IOM):

www.iom.int/jahia/Jahia/pid/835 (accessed: 09.11.2008).
7.1 Public Perception and the Roma in Italy
There are approximately 150,000-160,000 Roma living in Italy, half of whom are
Italian citizens. About half of non-Italian Roma are EU citizens, mainly Romani-
ans. In the nineties, most Roma migrants who arrived in Italy came from the for-
mer Yugoslavia. However, since the lifting of visa restrictions in 2002 and the ac-
cession of Romania to the EU in 2007, there has been a large increase in the
number of Roma from Romania migrating to Italy. Romanian Roma, if employed,
usually work in informal sectors of the economy (ERRC 2008).
The majority of the Roma population – Romanian citizens among them – lives
in ‘authorized’ and ‘unauthorized’ camps, mainly in Northern Italy. This policy of
establishing and offering camps reflects the perception of Italian authorities that
the Roma are ‘nomads’ and travelers, who do not need permanent accommodation
(ECRI 2002, 20). Thus, these camps are seen to serve the ‘protection of nomadic
cultures’ (International Helsinki Federation for Human Rights 2005, 18). This
perception prevails despite the fact that such camps are contrary to the ‘needs and
preferences’ of the Roma themselves (FRA 2008). Living in camps which lack the
‘basic infrastructure and facilities’ separate the Roma from the ‘mainstream Italian
society’ (ECRI 2002, 9). This separation contributes to the deterioration of the al-
ready problematic education and employment performance of the Roma. Thus,
while integration problems concerning Roma communities exist in many member
states, such separation or segregation on the basis of a misperception (i.e. that the
Roma are nomads) certainly negatively influences their potential for integration in
comparison to the situation of the Roma in other countries.
As summarized in Section 2, there is a reciprocal link between integration

problems and negative public perception. Public attitudes towards the Roma have
been largely negative in Italy, which has manifested from time to time in violent
attacks against Roma or Roma camps (FRA 2008, 8-9). According to a recent
Eurobarometer survey, Italians have the second least favorable opinion about the
Roma in the EU: almost half of the respondents in Italy would feel uncomfortable
having a Roma neighbor (European Commission 2008a, 43). Furthermore, this
survey shows that self-reported discrimination is also very high in Italy (European
Commission 2008a, 13).
As far as the media is concerned, a study on Italian media in the nineties argues
that the press usually ‘supported mobilizations against Roma settlements’ and de-
scribed the Roma as a ‘social pathology’ (ter Wal 2002, 244-245). Furthermore,
Italian media tended to ‘criminalize’ the Roma – and immigrants in general – rep-
resenting criminal behavior and deviance ‘as the inherent features of an entire
group’ (ter Wal 2002, 247). The second (2002) and third (2006) reports of the
European Commission against Racism and Intolerance (ECRI) drew special atten-
tion to the situation of Roma communities in Italy and to the problem that ‘wide-
spread societal prejudice against Roma/Gypsies is exploited and corroborated by

inflammatory statements made by politicians’ (ECRI 2002, 23).
7.2 The Securitization of Roma Migration since November

2007: Interaction between Perceptions and Policies
Such attitudes and related policies gained a new dimension after Romania’s acces-
sion to the EU in January 2007, and especially after November 2007. According
to an NGO report, ‘since the arrival of Romanian Roma in recent years, the Italian
people’s perception of Roma, which was already quite negative, has increasingly
changed’ in a negative way (ERRC 2008, 5). This period is marked by two main
events which triggered large media attention, increasingly negative attitudes and
even hostility towards both Romanians and the Roma – including violent attacks –
and resulted in the adoption of stricter migration policies that tended to ‘securi-
tize’ the discourse around Romanian, and especially Roma migrants (see Ange-
lescu 2008). Thus, these events illustrate the negative consequences of reactive
political behavior outlined in Section 3 (see the hypothetical example of the first
route taken by member state A).
The first event was the rape and murder of Giovanna Reggiani by a Romanian
Roma in Rome in October 2007. Following a media uproar, the Italian govern-
ment issued a decree the next day which granted powers to local authorities to ex-
pel EU citizens if they presented a threat to public order and security. As The
Economist argues, the decree explicitly targeted Romanians, with the preamble
stating that ‘the proportion of crime committed by foreigners has increased, and
those who commit most crime are the Romanians’ (The Economist 2007). This
statement might be true in absolute terms – given that Romanians are the largest
immigrant group – but not necessarily in relative terms. Nevertheless, the decree
led to the expulsion of 20 people within 24 hours after adoption (The Economist
2007).
Public opinion was largely favorable towards these measures. A poll published
in Corriere della Sera showed that only 30% of Italians think that it is possible to
integrate the Roma. Accordingly, Romanian authorities tried to shift attention to
the Roma instead of Romanian migrants, talking about the events as ‘the most
problematic phenomena, like the migration flows involving the Roma ethnic
group’ (The Economist 2007). Italian national politicians, especially the opposi-
tion and particularly the Lega Nord (Northern League) – already preparing for the
elections in April 2008 – also took part in the securitization of the ‘Roma prob-
lem’ with making public statements suggesting that the Roma are to a large extent
criminals (Angelescu 2008, 2).
The second incident occurred after the general elections when in May 2008 a
young Roma girl was caught in a private apartment in Naples holding the six-
month old baby of the family and was accused of attempted kidnapping. Follow-
ing this event, national newspapers widely brought up the traditional stereotype of
the ‘Roma stealing children’ (FRA 2008, 21). This event again triggered a series
of violent crimes against the Roma, including an assault against a big Roma camp
around Naples. A subsequent poll in La Repubblica showed that 68% of Italians
would like to have all the Roma expelled from the country. In another poll, 81%
of respondents said that they found the Roma ‘barely likeable or not likeable at
all’, while 64% had the same attitude towards non-Roma Romanians (see Kington
2008).
The new government also responded with stricter policies and proposed a secu-
rity package. The package – targeting mainly the Romanian Roma population in
the country – again included measures facilitating the expulsion of EU citizens
who committed criminal offences, as well as criminalized unauthorized entry into
Italy (FRA 2008, 11). Silvio Berlusconi, Italy’s prime minister – who previously
described the Roma as an ‘army of evil’ (see Fraser 2008) – introduced the pack-
age claiming that the government wanted to ensure ‘the right of Italians not to be
afraid – a fundamental right that the state must guarantee’ (see Hooper 2008). The
government also declared a ‘state of emergency with regard to nomad community
settlements’ (ERRC 2008, 7). Furthermore, it started to carry out a census and fin-
gerprint all Roma who lived in camps, including children, in order ‘to avoid phe-
nomena like begging’, as Roberto Maroni, Italy’s interior minister put it (BBC
News 2008). This plan brought about widespread international disapproval (see
FRA 2008, 26).
7.3 Results
Italian events in the year between October 2007 and September 2008, labeled as
the ‘Roma emergency’ in the Italian press, illustrate well the vicious circle of se-
curitization with regard to negative attitudes and perceptions, media stereotyping
and strict migration policies. As a result of securitizing and criminalizing public
and political discourses, a group that is already separated from the Italian society –
being treated as ‘nomads’ – becomes even more marginalized and stigmatized.
One way to break this vicious circle and to reduce existing prejudices is to change
such political discourse and reactive policies (see discussion in Section 3). As it
was formulated by the ECRI in a statement issued in June 2008: there is ‘the ur-
gent need for the Italian authorities to take a firm stand against all forms of racism
and xenophobia, including hate speech, so as to curb and prevent the development
of these phenomena in Italian society’ (ECRI 2008).
8 Conclusions
On the basis of the material in this chapter we come to four conclusions.
First, there is a clear link between public perception and migration policies. It
is clear that Roma migrants are treated differently in Italy than in other similar
member states with relatively small native Roma population irrespective of their
numbers or profile. The same picture emerged in the discussion on the veil.
Second, the dual causality that was established in Beutin et al (2007) was con-
firmed with empirical evidence. In particular, the case of Poles in the UK shows
the point. Public perception prior to arrival of Poles was more negative than af-
terwards, partially because initial fears were based on imaginary things that might
have happened after migration. So a negative perception, combined with positive
policy led to a subsequent positive spiral of perception.
Third, there is also a link between performance of migrants on the labor market
and perception. Although we should tread carefully here, not being able to draw
too bold conclusion on the limited empirical information collected in this chapter,
there are nevertheless clear signals. In the case of Poles in the UK, the causality
mainly runs from performance to perception. In the case of Roma in Italy, the cau-
sality was mainly from perception to performance, as it is in the case of the veil.
Fourth, there has been a silent but noticeable recent positive shift in public per-
ception on migration. This is not only confirmed by Eurobarometer surveys but
also other sources such as the 2008 Transatlantic Trends: Immigration survey, as
well as the discussion on the veil and the shift in perception towards Poles in the
UK. It is, however, too early to conclude that there is a clear trend for three rea-
sons. First, the trend is very recent. Only data from the last two years show the
trend. Second, there could be a backlash resulting from the financial crisis. Third,
there are still large differences across member states, with some member states
seeing a worsening of the tone in the debate. The discussion on Roma in Italy un-
derlines this point.
References
Alibhai-Brown, Y. (2000), Who do We Think We Are? Imagining the New Britain. London.
Angelescu, I. (2008), All New Migration Debates Commence in Rome: New Developments
in the Securitization of Migration in the EU, in: Across Fading Borders: The Challenges of
East-West Migration in the EU.
www.eumap.org, (14 April)
BBC News (2008), Italy Fingerprint Plan Criticised, June 26.
Beutin, R./M. Canoy/A. Horvath/A. Hubert/F. Lerais/M. Sochacki (2007), Reassessing the
Link between Public Perception and Migration Policy, in: European Journal of Migration
and Law 9(4), 389-418.
Bronner, L. (2005), Un rapport dresse un bilan positif de la loi sur le voile à l’école. Le
Monde, 27 August.
Buzan, B./O. Wæver/J. de Wilde (1998), Security: A New Framework for Analysis. Boulder.
Commission de Réflexion sur l'Application du Principe de Laïcité dans la République. 2003.
Rapport au Président de la République.
http://lesrapports.ladocumentationfrancaise.fr/BRP/034000725/0000.pdf
The Economist (2007), Disharmony and Tension, 8 November.
European Monitoring Centre on Racism and Xenophobia (EUMC) (2006), The Annual Re-
port on the Situation Regarding Racism and Xenophobia in the Member States of the EU.
http://fra.europa.eu
European Commission (2007), Attitudes towards the EU in the United Kingdom. Analytical
Report, Flash Eurobarometer 203, Field Work: January 2007, Publication: May 2007.
http://europa.eu.int/comm/public_opinion/index_en.htm
European Commission (2008a), Discrimination in the European Union: Perceptions, Experi-
ences and Attitudes. Report on the Special Eurobarometer 296, Field Work: February-
March 2008, Publication: July 2008.
European Commission (2008b), The 2009 European Elections. Report on the Special Euro-
barometer 299. Field work: March-May 2008. Publication: September 2008.
European Commission (2008c), Eurobarometer 69: Values of Europeans. Report on the Stan-
dard Eurobarometer 69, Field Work: March-May 2008, Publication: November 2008.
European Commission (2008d), Eurobarometer 70: First Results. Field work: October-
November 2008, Publication: December 2008.
European Commission (2008e), Employment in Europe 2008. Brussels: European Commis-
sion, Directorate-General for Employment, Social Affairs and Equal Opportunities.
European Commission against Racism and Intolerance (ECRI) (2002), Second Report on Italy
(Adopted on 22 June 2001), Strasbourg.
European Commission against Racism and Intolerance (ECRI) (2006), Third Report on Italy
(Adopted on 16 December 2005), Strasbourg.
European Commission against Racism and Intolerance (ECRI) (2008), Statement of the Euro-
pean Commission against Racism and Intolerance on Recent Events Affecting Roma and
Immigrants in Italy.
www.coe.int/t/e/human_rights/ecri/ (20 June).
European Parliament (2007), L’islam dans l’Union européenne: quel enjeu pour l’avenir.
European Parliament, Direction générale Politiques internes de l'Union, PE 369.031.
European Roma Rights Centre (ERRC; in cooperation with the Open Society Institute, Ro-
mani Criss, Roma Civic Alliance in Romania, and Centre on Housing Rights and Evic-
tions) (2008), Security a la Italiana: Fingerprinting, Extreme Violence and Harassment of

Roma in Italy, Budapest.
European Union Agency for Fundamental Rights (FRA) (2008), Incident Report: Violent At-
tacks against Roma in the Ponticelli District of Naples, Italy. http://fra.europa.eu
Fomina, J./J. Frelak (2008), Next Stopski London. Public Perceptions of Labour Migration
within the EU. The Case of Polish Labour Migrants in the British Press. Warsaw (Institute
of Public Affairs).
Fraser, C. (2008), Italy Police to Protect Gypsies, BBC News, 14 May.
Frattini, F. (2005), Speech at Harvard University, 7 November 2005.
Home Office (2008), Accession Monitoring Report, May 2004 - September 2008 (A8 Coun-
tries). A joint online report between the UK Border Agency, Department for Work and
Pensions, HM Revenue and Customs and Communities and Local Government.
Hooper, J. (2008), Berlusconi Clamps Down on Gypsies, The Guardian, 22 May.
Institute of Public Policy Research (2008), Half of Migrants from New EU Countries Have
Now Left UK.
www.ippr.org/pressreleases/
International Helsinki Federation for Human Rights (2005), The Situation of Roma in Se-
lected Western European Countries. Report to the OSCE Conference on Anti-Semitism
and on other Forms of Intolerance, Cordoba, Spain, 8-9 June.
Kington, T. (2008), 68% of Italians Want Roma Expelled – Poll, The Guardian, 17 May.
Klausen, J. (2005), The Islamic Challenge: Politics and Religion in Western Europe. Oxford.
Kornhauser R. R (1978), Social Sources of Delinquency: An Appraisal of Analytic Models,
Chicago.
OECD/SOPEMI (2007), International Migration Outlook, Paris.
OECD/SOPEMI (2008), International Migration Outlook, Paris.
Office for National Statistics (2006), National Population Projections 2006–based. Series PP2
No 26. London.
www.statistics.gov.uk/downloads/theme_population/pp2no26.pdf
Parekh, B. (1997), A Varied Moral World: A Response to Susan Okin’s ‘Is Multiculturalism
Bad for Women?’, Boston Review (October/November 1997).
http://bostonreview.net/BR22.5/parekh.html
The Pew Global Attitudes Project (2006), Muslims in Europe: Economic Worries Top Con-
cerns about Religious and Cultural Identity, July. www.pewglobal.org
Rubin, J./M. S. Rendall/L. Rabinovich/F. Tsang/C. van Oranje-Nassau/B. Janta (2008), Mi-
grant Women in the EU Labour Force: Current Situation and Future Prospects. Technical
Report of the RAND Corporation.
www.rand.org
Saghal, G./N. Yuval-Davis (2000), Refusing Holy Orders: Women and Fundamentalism in
Britain.
www.wluml.org
Sampson R. J./S. W. Raudenbush (1999), Systemic Social Observation of Public Spaces: A
New Look at Disorder in Urban Neighborhoods, in: American Journal of Sociology 105
(3), 603-651.
Sampson R. J./W. B. Groves (1989), Community Structure and Crime: Testing Social-
Disorganization Theory, in: American Journal of Sociology 94 (4), 774-802.
ter Wal, J. (2002), Italy, in: J. ter Wal (ed.), Racism and Cultural Diversity in the Mass Media:
An Overview of Research and Examples of Good Practice in the EU Member States 1995-
2000. On Behalf of the European Monitoring Centre on Racism and Xenophobia. Vienna
(EUMC).
Transatlantic Trends (2008), Transatlantic Trends: Immigration. Key findings 2008.
www.transatlantictrends.org
Veysey, B. M./S. F. Messner (1999), Further Testing of Social Disorganization Theory: An

Elaboration of Sampson and Groves’s ‘Community Structure and Crim’, in: Journal of
Research in Crime and Delinquency 36 (2), 156-174.
Young Foundation (2008), Valuing Family, Valuing Work: British Muslim Women and the
Labour Market.
www.youngfoundation.org.uk/research/news/valuing-family-valuing-work-british-
muslim-women-and-labour-market
4. EU Enlargement under Mobility Restrictions: Consequences for Germany
4. EU Enlargement under Continued Mobility

Restrictions: Consequences
for the German Labor Market
Karl Brenke
Mutlu Yuksel
Klaus F. Zimmermann
The numbers of migrants from the accessions countries have clearly increased
since the enlargement of the EU in 2004. Following enlargement, the net inflow of
EU8 immigrants has become 2.5 times larger than the four-year period before
enlargement. Poles constitute the largest immigrant group among the EU8 immi-
grants: since enlargement, 65% of all net immigrants and 71% of EU8 immi-
grants are from Poland. This chapter presents new evidence on the impact of im-
migrant flow from EU8 countries on the German labor market since EU
enlargement. Unlike other EU countries, Germany has not immediately opened up
its labor market for immigrants from the new member states. Nevertheless, our
analysis documents a substantial inflow and suggests that the composition of EU8
immigrants has changed since EU enlargement. The majority of the new EU8 im-
migrants are male and young, and they are less educated compared to previous
immigrant groups. We also find that recent EU8 immigrants are more likely to be
self-employed than employed as a wage earner. Furthermore, these recent EU8
immigrants earn less conditional on being employed or self-employed. Our find-
ings suggest that these recent EU8 immigrants are more likely to compete with
immigrants from outside of Europe for low-skilled jobs instead of competing with
German natives. While Germany needs high-skilled immigrants, our analysis sug-
gests that the new EU8 immigrants only replace non-EU immigrants in low-
skilled jobs. These results underline the importance of more open immigration
policies targeting high-skilled immigrants. The current policy not only cannot at-
tract the required high-skilled workforce, but also cannot avoid the attraction of
low-skilled immigrants, and is a complete failure.
Migration, DOI 10.1007/978-3-642-02242-5_4,
112 Karl Brenke, Mutlu Yuksel and Klaus F. Zimmermann
1 Introduction
Germany has been experiencing a large inflow of immigrants for at least the last
four decades. Post Second World War labor immigration into Germany began dur-
ing the 1960s and 1970s, when Germany invited approximately 3 million guest-
workers especially from Southern Europe, the former Yugoslavia and Turkey.
Since reunification some 15 years ago, 2.5 million people from Eastern Europe
have moved to Germany. Immigration has kept Germany’s population constant
since the enlargement of the European Union in the spring of 2004.
According to the Central Register of Foreigners, the total number of immi-

grants between the age of 15 and 65 in Germany at the end of 2007 was close to
5.5 million, or approximately 10% of the total population in that age group. Immi-
grants from EU8 countries constituted only 9.3% of all immigrants at the end of
2007.1 However, net inflows have been increasing from the new member states
into Germany despite a “closed door”2 policy aimed at restricting immigration.
A number of recent studies extensively analyze migration intentions, determi-

nants and scale of migration, effects of the post-enlargement migration flows on
the receiving as well as the sending countries' labor markets, welfare systems, and
growth and competitiveness for all European countries after EU enlargements
(Kahanec and Zimmermann, 2008; Fouarge and Ester, 2007a, 2007b; Bonin et al.,
2008; Zaiceva and Zimmermann, 2008; De Giorgi and Pellizzari, 2006; Blanch-
flower and Lawton, chapter 8 in this volume). This chapter adds to this strand of
literature. Here, we investigate the influx of recent EU8 immigrants, their charac-
teristics and how they impact the German labor market. In the next section, we
provide the legal framework governing Germany’s restrictive policy, and any ex-
ceptions made for new immigrants. In section 3, recent migration and the back-
ground of EU8 immigrants are studied. In section 4, we offer a broad picture of
the main characteristics of this group of immigrants, their observed skills and their
labor market outcomes. Section 5 compares the labor market situation of these re-
cent EU8 immigrants to that of other immigrant groups and natives. Finally, sec-
tion 6 discusses the main findings and provides some immigration policy recom-
mendations in view of the results.
1 The percentages of immigrants from the more traditional source countries are as follows:
24.6% of the 5.5 million are from EU15 countries; 24.1% from Turkey; 13.2% from the
former Yugoslavia; 5% from the ex-Soviet Union; and 20.1% from outside of Europe.
2 For more detailed information see Brenke and Zimmermann (2007).
4. EU Enlargement under Mobility Restrictions: Consequences for Germany 113
2 The Legal Framework
During the negotiations regarding the enlargement of the EU which took place on
May 1st 2004, the individual countries of “old” Europe (EU15) were able to de-
cide for themselves how far they would relax their border restrictions for new mi-
grants from the new member states. Germany has since had a restrictive policy
and denied both workers and companies from the accession countries complete
freedom, with the exceptions of Cyprus and Malta. Therefore, Germany, together
with Austria, are unique because they are the only member states which have not
extensively relaxed restrictions on migrants from the accession countries of 2004.
However, holes have appeared in the fence which has been raised to block immi-
grants from the new member states. The following summarizes the exceptions for
workers and businesses.
Self-employed workers from the EU8 are allowed to settle in Germany and run
a business. However, they may not employ workers from their home country.
Temporary work permits are given to employees from EU8 countries (including
Bulgaria and Romania), but only if German workers or established immigrants are
not put at a disadvantage.3 Further training can also count as employment. In
2007, government job agencies issued 61,000 work permits (without seasonal
workers, domestic servants) for EU8 citizens and 19,000 for workers from Bul-
garia and Romania. It should also be said, however, that 12,000 applications from
the new member states were rejected.4 In addition to these permits, 286,000 sea-
sonal work permits were issued, where 79% were used by Poles.
Even though politicians evidently expect German companies to face strong

competition - especially due to lower wages - and have restricted the free move-
ment of most services, businesses of certain sectors can operate with different
permits. For instance, the IT industry or consultation companies can operate freely
in Germany, and they may also employ workers from abroad. However, some for-
eign companies are allowed to be active in sectors such as the building industry,
the decorating business, and cleaning companies but only with so-called “key per-
sonnel”, who are “management and personnel with the qualifications and exper-
tise for the job or area of work, and skills which are necessary for the home (Ger-
man) company base.”5
The other exceptions are so-called “contingency contracts”, which are mainly
for farm helpers, and “special contracts” for high-skilled workers. Such contin-
gency contracts have been around for decades and are not just associated with the
3 See § 284, Abs. 1 SGB III. SGB means Sozialgesetzbuch (social law).
4 See Bundesagentur für Arbeit (2008).
5 See Christen (2004).
enlargement process. These types of contracts are mainly established by bilateral

agreements between Germany and certain EU8 countries, and they are referred to
as “contingent labor forces”, which are for workers who only stay in Germany for
a specific period of time.
Generally, the immigration of high-skilled workers is possible, especially in the

IT industry, but this is not related to EU8 expansion. The earnings are fixed for
this type of immigrants, who need to earn in Germany a minimum gross annual
salary of 86,400 euros. However, as few people immigrated to Germany as a re-
sult, the German government in the summer of 2007 decided to lower the limit to
63,600 euros.6 In addition, there are also special rules for individual professions
e.g. scientists working in research or people who work at fairgrounds. These high-
skilled permits are issued in very low numbers relative to the total number of work
permits issued in Germany. Ultimately, it is also possible for a foreigner outside
the EU15 to obtain a work permit by marrying a German national.
There has been little debate or disagreement between the political parties on
limiting access to Germany for immigrants from EU8 countries. Therefore, it is
expected that Germany’s current policy will remain until 2011, when the transi-
tional period of the free movement of workers and services expires. This is all the
more likely given the slowing down of the economy since 2008.
3 Migration Movement and Background

of People from EU8 Countries
The following findings are based on two official data sources on the population
from regional registry offices in the community (“örtliche Melderegister”) and the
Central Registry of Foreigners (“Ausländerzentralregister”). The data sets contain
personal information about immigrants (and emigrants) such as age, sex, national-
ity and length of stay.
Table 1 presents annual inflows and net migration numbers in Germany from
different countries, including EU8 countries, from 2000 to 2006. The number of
migrants from the accessions countries has clearly increased since the enlargement
in 2004. The net gain of EU8 immigrants is 2.5 times larger than in the four-year
period before enlargement. Had it not been for migration from the EU8, Germany
would have had a net loss of migration. A decrease of net-migration resulted from
a more than 50% drop in the number of immigrants from other important source
6 Numbers for the issued work permits for IT workers exist only for countries outside of the
EU. This number was 3,400 in 2007. See Bundesagentur für Arbeit (2008).
countries, including Turkey, Russia and Ukraine, along with countries further out-
side Europe. The outflows of German citizens and EU15 immigrants have been
higher than the inflows, resulting in a net loss over the last 2 years. The total net
gain is predominantly caused by Polish migration and is at around 65%.
Table 1
Total annual inflow and net migration numbers by citizenship in Germany
Citizenship 2000 2001 2002 2003 2004 2005 2006
A. Total annual inflows
Poland 70,431 75,012 77,405 84,693 119,551 141,314 145,761

Hungary 15,514 16,455 15,982 13,790 16,834 18,026 18,111
Slovakia 10,413 10,985 10,889 10,006 11,053 11,395 10,933
Czech. Rep. 10,392 10,207 9,601 7,873 8,275 7,762 7,125
Slovenia 1,770 2,504 2,197 1,975 2,260 1,397 1,080
Estonia . . . 738 667 635 529
Latvia . . . 1,642 2,085 2,248 1,857
Lithuania . . . 2,775 4,150 4,859 4,477
EU8 108,520 115,163 116,074 123,492 164,875 187,636 189,873
Romania 22,521 18,547 22,376 22,329 22,203 22,052 22,532
Bulgaria 10,008 12,516 12,371 12,613 10,891 8,492 7,260
Turkey 39,575 44,189 47,616 41,908 36,275 30,002 26,059
Russia 25,996 28,208 29,072 25,671 23,168 19,072 14,274
Other countries 153,116 175,217 174,215 150,830 136,713 125,127 121,849
Germany 132,773 134,616 128,238 118,019 131,752 92,953 77,082
EU15 115,359 105,896 97,127 86,745 81,643 78,175 78,535
Europe 400,957 411,951 402,009 375,376 396,597 392,656 380,117
Total 686,846 721,784 704,462 644,225 665,062 610,736 579,048
B. Total annual net migration
Poland 11,841 12,836 11,461 13,003 26,600 44,951 40,569

Hungary 1,500 2,010 685 -770 793 2,716 3,426
Slovakia 2,000 1,548 1,462 806 1,240 2,522 1,696
Czech. Rep. 2,070 2,154 1,113 58 433 1,868 1,034
Slovenia -32 255 26 -5,867 76 -37 1
Estonia . . . 268 -38 203 60
Latvia . . . 292 510 891 379
Lithuania . . . 831 1,888 2,465 1,437
EU8 17,379 18,803 14,747 8,621 31,502 55,579 48,602
Romania 6,900 1,319 5,840 3,634 2,896 2,334 1,627
Bulgaria 3,511 4,819 4,027 2,860 1,137 2,746 226
Turkey 7,992 14,871 17,526 11,650 5,666 2,766 881
Russia 15,816 17,770 17,202 14,094 11,360 8,366 4,078
Other countries 53,820 76,198 71,707 43,516 20,558 21,075 16,269
Germany 47,930 49,330 37,330 19,484 13,524 -17,796 -41,009
EU15 5,868 2,345 -8,486 -10,019 -25,507 -4,401 -2,524
Europe 29,867 78,832 62,272 45,497 33,231 70,124 56,461
Total 131,617 204,360 171,309 108,497 67,313 73,403 31,721
Source: Own calculations based on Statistisches Bundesamt (Federal Statistical Office) -
Population Register.
Notes: Outflow numbers are inflow minus net migration numbers. "Other countries" include remaining countries, stateless
individuals and individuals with unknown nationality.
As mentioned above, Poland, responsible for 71% of EU8 immigrants, is the

major source country for EU8 migration. Hungary is the second most significant
source country and Estonia the least.7 For each EU8 immigrant living in Germany,
there are 116 who live in the home country. Poles clearly dominate the recent im-
migrant flow. That is not surprising given that Poland is by far the most populated
of the EU8 countries. Similar vein, if the number of immigrants in Germany is
compared to their source country’s population, Poles rank clearly above the other
EU8 countries as well. In particular, for each Polish immigrant living in Germany,
there are 81.4 Poles residing in Poland. Relatively few migrants come from the
Czech Republic8, which is surprising given its proximity to Germany. This may
indicate that recent EU8 migration is intentionally temporary.
When we analyze the distribution of recent immigrants according to age co-

horts, we find evidence suggesting that EU8 immigrants’ main aim when migrat-
ing to Germany is employment. Immigrants from EU8 countries mainly comprise
of those in the age group 25 to 45. For every immigrant in Germany from an EU8
country aged between 25 and 35, there are 66 people of the same age living in the
EU8 countries.9 Furthermore, immigrants from the EU8 who have migrated after
2004 have migrated with fewer children than other immigrants: 14.7% of other
immigrants and only 5.4% of EU8 immigrants were younger than 15 years old in
2007. However, it is quite likely that their children have remained in the source
country.
When we explore the composition of recent immigrants by gender, there is a

slight decline in the share of working age females amongst EU8 immigrants,
which decreased from 55.3% in 2004 to 52.7% in 2007. Furthermore, the share of
women increased from the Baltic States and the Czech Republic during the same
period. However, these national groups have fewer immigrants in Germany,
which is in contrast to Poles, whose numbers have disproportionately increased,
with the share of Polish women decreasing by around 4 percentage points10 in
2007. There are also gender differences along with age differences between immi-
grant cohorts from these countries. For instance, there were twice as many men as
women aged between 25 and 34 who migrated in the period 2005 to 2007.
7 10% from Hungary, 6% from the Czech Republic, 4% from Slovakia, 3% from Slovenia
and Lithuania, 2% from Latvia and 1% from Estonia.
8 For each Czech immigrant in Germany, there are 356 Czechs living in the Czech Repub-
lic; 1:342 for Hungary; 1:265 for Estonia; 1:219 for Slovenia; 1:195 Latvia; 1:139 Lithuania;
1:64 Slovakia.
9 One immigrant who is in the same age cohort living in Germany to 310 people who are be-
tween 15 to 20 years old living in the EU8 countries; 113 aged 20 to 25; 88 aged 35 to 45;
121 aged 45 to 55; 212 aged 55 to 65.
10 In 2007, the share of females from Estonia was 74.2%; Latvia 69.8%; Lithuania 73.8%;
Poland 51.1%; Slovakia 59%; Slovenia 51.9%; the Czech Republic 67.7%; and Hungary
40.8%.
Not only do men have higher levels of immigration, but they have been more
likely to leave Germany since 2004. Both male inflow and outflow numbers are
larger than those of females, and hence, so are the net migration numbers. In 2007,
one quarter of EU8 immigrants stayed less than a year in Germany, and 70%
stayed less than three years. Compared to the earliest data, which are from 2005,
the average duration of residence has risen considerably for EU8 immigrants;
while it has remained the same for EU15 and non-EU immigrants.
To summarize, there was a definite increase in immigration to Germany after

EU enlargement. There would have been a net loss of migration as early as 2006
without immigrants from EU8 countries, mainly Poland. The migration flow after
EU enlargement has consisted mostly of Polish migrants, especially men, who
have moved to Germany, and migrants from the EU8, which have exhibited large
fluctuations. The most significant group is aged between 25 and 45; hence, few
children have migrated. This suggests that migration was not permanent by inten-
tion, but has been a planned temporary move to Germany.
4 Descriptive Statistics
Although the data in the registry contains information on age, sex, nationality and
length of stay, it does not provide much information about the socio-economic
characteristics of immigrants in Germany. Therefore, we use microcensus data,
which are based on an official survey of households obligated to give information.
The microcensus includes 1% of the resident population, and it is a large, repre-
sentative, random sample containing comprehensive information on socio-
demographic and economic characteristics of individuals in Germany. We use
data from the 2006 wave, with approximately 380,000 households and 820,000
people in total, which are generally based on face-to-face interviews. The survey
also contains information on all household members and has a very high response
rate of approximately 97%.
Table 2 reports detailed information from the 2006 microcensus on the labor
force participation, employment status, and occupational skill levels according to
the source country of the immigrant. EU8 immigrants are divided into two sub-
groups of those who migrated before 2004 and those after. The labor force partici-
pation rate of EU8 immigrants is lower compared to Germans and EU15 immi-
grants but higher than or equal to immigrants from other countries. The unem-
ployment rate of recent EU8 immigrants is much lower compared to other
immigrants. However, when we exclude students who are also employed, the un-
employment rate is about 20% for recent EU8 immigrants and nearly the same as
the rate for earlier ones from the EU8. Moreover, the percentage of students from
EU8 countries increased from 10% to 15% after enlargement. Although EU15
citizens have free access, recent years have seen a decrease in their migration to
Germany.
Table 2
Employment and residential characteristics in Germany by citizenship
Earlier Recent S.E. Other Middle North

Natives EU8 EU8 EU15 Russia Turkey Eur. Eur. East Asia Am. Others
Labor Force Participation
Students, apprentices 14 10 15 10 15 14 11 13 17 14 13 18
Active labour force 62 53 54 64 37 41 54 50 32 48 58 37
Unemployed* 7 14 9 8 22 13 11 12 20 8 7 13
Non-active population 17 23 22 18 26 31 25 25 32 30 23 32
Unemployment rate** 10 21 19 11 38 25 17 19 39 15 11 26
Job Quality
Low-skilled 29 48 52 44 59 65 56 44 53 51 27 58
Medium-skilled 58 45 37 45 33 33 40 47 39 43 52 37
High-skilled 13 8 12 12 9 2 4 9 8 6 21 5
Employment Status
Self-employed 12 11 40 16 11 9 7 18 12 21 21 15
Employed 88 89 60 84 89 91 93 82 88 79 79 85
Residence
Core cities 23 43 42 38 39 41 43 42 55 45 40 52
Other cities 16 15 19 22 13 22 20 15 16 14 15 16
Other types of regions 61 42 39 40 48 37 37 43 29 41 45 32
Source: Own calculations based on Microcensus data (2006).
Notes: All numbers are in %. "Recent EU8" shows immigrants migrated after 2004 from EU8 countries. "Earlier EU8" shows
immigrants arrive before 2004 from EU8 countries. "S.E. Eur." includes former Yugoslavia countries (without Slovenia), Bul-
garia, and Romania. " Other Eur" indicates remaining EU countries. "Middle East" includes Middle East and North African
countries. "Job Quality" is classified by skill intensity of occupations.
* includes employed students.
** excludes employed students.
Even though the employment rate is higher for EU8 immigrants compared to
other immigrants, half of the employed EU8 immigrants are self-employed or
work in part-time or mini-jobs11. This may be due to legal restrictions faced by
EU8 migrants who live in Germany which act as a barrier to the labor market. The
percentage of self-employed immigrants is generally higher than native Germans,
with the exception of people from Turkey and South-East Europe. The incidence
is especially high for people from the EU8. Obviously, many EU8 immigrants
have taken advantage of the opportunity made possible by EU enlargement to take
up residence in Germany and become self-employed. This finding probably has to
do with the fact that such employment is often associated with simple unskilled
jobs. Alternatively, it might be that the type of work an immigrant tends to do is in
a legal gray zone or is in fact illegal. It could explain why many immigrants, both
from the EU8 and non-EU countries, are marginally employed. Exceptions to this
are people from Turkey and the former Yugoslavia: national groups which be-
longed to an earlier guest worker program. Table 2 also shows that more than half
11 A mini-job is a form of employment in which up to 400 euros a month can be earned with-
out having to pay any tax or social security contributions. The employer has to make a
small social security payment. However, there may also be comparable employment ar-
rangements in an illegal form.
of all immigrant groups reside in the core cities of metropolitan areas. The figures
are similar for recent and earlier EU8 immigrants, where 42% and 43% of them
live in big metropolitan cities, respectively. This finding suggests that immigrants
are attracted by the economic opportunities to be found in big cities. It also seems
that enclave networks are important in the choice of residence.
The main interest group is EU8 immigrants, and five main subgroups of the total
sample are defined and presented in the following empirical analysis. The first dis-
tinction leads to two groups which are based on the criteria of having German na-
tionality. The first group is called “native” and includes those born in Germany and
with German nationality since birth. The second group is called “native-others” and
contains individuals who were born in Germany but naturalized later (such as sec-
ond generation immigrants), or who migrated to Germany and have only German
nationality (such as ethnic-Germans). Immigrants form the remaining population
and are divided into 3 sub-groups. The first sub-group is called “EU8” and consti-
tutes immigrants who migrated from and are a national of an EU8 country. The sec-
ond sub-group is called “EU15” and includes migrants from EU15 countries and the
US. The last sub-group is called “non-EU” and covers those from non-EU countries,
mainly Turkey, Africa, the Middle East and Asia. Furthermore, individuals who
immigrated to Germany before and after the 2004 enlargement were also separately
analyzed for each of the groups. The target group that we mainly focus on in our
analysis is “recent EU8 immigrants”, who migrated from enlargement countries af-
ter 2004. For a detailed analysis, we also separate all immigrants groups into two by
their year of migration, where “recent” refers to the immigrants who migrate after
2004, and “earlier” refers to the ones that migrated before 2004.
EU8, non-EU, EU15 and native-others constitute 1%, 10%, 3% and 5%, re-
spectively, of the weighted sample containing all individuals who are aged be-
tween 15 and 65. 7% of all immigrants are from EU8 countries, 22% from EU15,
and the remaining 71% are from other countries in our sample. Immigrants from
EU8 countries generally migrate at older ages compared to immigrants from all
other countries. However, they are also on average more educated than other im-
migrant groups. The mean age at migration of a EU8 immigrant is 26; whereas it
is around 22 for other immigrant groups. On the other hand, if we only focus on
immigrants who arrived just 3 years before and after enlargement, we have a dif-
ferent picture. The average age at the time of migration is 32 for an immigrant
from the EU15 and 29 for a non-EU country. Furthermore, this average age is the
same before and after enlargement. However, an EU8 immigrant’s average age at
migration was 27 before enlargement, and the mean age of recent immigrants is
31 years old. These figures hint that EU enlargement might have altered the com-
position of immigrants from EU8 countries, which we will analyze in detail in the
remainder of the chapter.
Figure 1 presents the stock of EU8 immigrants in 2006 by year of arrival and
gender, where the years of arrival are divided into 3-year intervals. In order to cal-
culate the stock, we take the number of male immigrants in 2006 from a year of
arrival interval and divide this by the total number of males in 2006 and then mul-
tiply this share by 100 to express it in percentage points. Female stock percent-
ages are calculated similarly. EU8 immigrants are divided into 12 categories by
year of migration providing information on immigrants who migrated between
1971 and 1973, and so forth until 2004 and 2006, which contains the most recent
after enlargement influxes of EU8 migrants to Germany.
The stock of EU8 immigrants who migrated between 1989 and 1991 constitutes
0.19% for all females and 0.17% for all males in 2006. After the collapse of the Ber-
lin Wall, there was a dramatic increase in migration from EU8 countries, and then it
fell steeply until 1995. After 1995, migration from these countries started to increase
once more with the percentage of females from the EU8 reaching 0.19% at the point
of EU enlargement. On the other hand, the share of EU8 male immigrants increased
to 0.13% after 2004, but it is still lower than the percentage of female migrants from
these countries. The percentage of recent immigrants (those who migrated between
2004 and 2006), other immigrant groups and the immigrants who migrated between
2001 and 2003 are the same or have decreased.
Figure 1
The stock of EU8 immigrants in 2006 by year of arrival and gender
0.20
0.17
0.18
0.19
percentage of population by gender
0.16 EU8 male

0.19 0.17
0.14 0.12 0.13 EU8 female
0.14
0.12
0.13
0.10 0.12
0.08 0.10 0.07

0.06
0.06 0.04
0.03
0.03 0.05
0.04
0.02
0.01 0.04
0.02 0.01
0.03
0.00 0.01 0.01 0.01
1971-1973
1974-1976
1977-1979
1980-1982
1983-1985
1986-1988
1989-1991
1992-1994
1995-1997
1998-2000
2001-2003
2004-2006
year of arrival
Source: Own calculations based on Microcensus data.

Notes: The X-axis shows the year of arrival in 3-year intervals and the Y-axis shows the percentage of total population by
gender. In order to calculate the stock, we take the number of male immigrants in 2006 from a year of arrival interval and
divide this by total number of males in 2006 and then multiply this share by 100 to express it in percentage points. Female
stock percentages are calculated similarly.
The microcensus data provides information on the highest secondary school

degree acquired and information related to the type of vocational training that the
individual has received. These two variables allow us to construct a continuous
measure of years of education. The mean year of schooling is 13.3 years for EU8
immigrants and 12.6 for non-EU immigrants. However, a more detailed analysis
reveals that differences in educational attainment have been decreasing over time
between those two groups, which we will incorporate in our analysis.
Figure 2a and 2b show the average years of education of immigrant males and
females according to the year of arrival in Germany. After 1990, the average years
of schooling for immigrant males increased for all immigrant groups. However, be-
fore enlargement the average education of EU8 immigrants was higher than of non-
EU immigrants and the EU15 group before 2000. The picture has changed since
2004: the average education of recent male EU8 immigrants as become the lowest
among all recent male immigrants from non-EU countries. This could suggest that a
cohort of lower educated immigrants from the EU8 have migrated to Germany since
enlargement. Albeit, there is a slight decrease in the average years of education of
male immigrants from EU8 countries; however, this may be true only for this pe-
riod. The questions regarding this issue will be better addressed once more recent
data become available. On the other hand, figures show that more educated immi-
grants from non-EU countries migrated to Germany in the last decade and less edu-
cated ones left, so the average education has increased for this group. However, the
situation is different for females. After 1990, with the collapse of the Berlin Wall,
the average education of female immigrants from the EU8 was higher than that of
males. Even after enlargement, there has still been an increasing trend in average
educational attainments of EU8 females, whose average education is higher than
female immigrants from non-EU countries.
In order to understand the characteristics of recent EU8 immigrants better, we

also broaden our analysis to the federal state level. After enlargement, as with all
other immigrant groups in Germany, there are no or very few recent EU8 immi-
grants in the states of Saarland, Brandenburg, Mecklenburg-Vorpommern, Saxony
and Thuringia. This may indicate that recent EU8 immigrants prefer to migrate for
economic reasons, since they prefer to migrate to states where job opportunities
are high. On the other hand, these figures also suggest that the existence of earlier
immigrants from EU8 countries in certain states or cities might lead to more re-
cent immigrants from those countries. To put some context to this issue, Hamburg
has the largest EU8 immigrant population, at approximately 2% of the total popu-
lation of the state and 12% of them are post-enlargement EU8 immigrants. Fur-
thermore, 1% of the population in the Rhineland is from an EU8 country, and
25.3% of those came after enlargement. Recent EU8 male immigrants who live in
Bavaria have the lowest average years of education, 12.2 years, and are relatively
older than those who live in other states, with an average age of 35.3 years. On the
other hand, Berlin and Hamburg attract relatively younger immigrants, who are 27
years old on average at the time of migration and have the highest average years
of education with 14.3 years.
Figure 2a
Average years of schooling of male immigrants by year of arrival and country of origin
15.5
15
14.5
14 13.88
years of schooling
13.45 13.39 13.56

13.38 13.33
13.5
12.80 12.80
13
12.42
12.32
12.5 EU8 male
11.80
12 Non-EU male
11.5 11.87
EU15 male
11
1971-1973
1974-1976
1977-1979
1980-1982
1983-1985
1986-1988
1989-1991
1992-1994
1995-1997
1998-2000
2001-2003
2004-2006
year of arrival

Notes: The X-axis shows the year of arrival in 3-year intervals and the Y-axis shows average years of schooling in 2006.
“EU8 male”, “EU15 male”, “Non-EU male” indicate average years of schooling of male immigrants from the EU8, EU15, and
remaining countries respectively.
Figure 2b
Average years of schooling of female immigrants by year of arrival and country of origin
15.5
15
14.5
14 13.71 13.77
years of schooling
13.60 13.52 13.44 13.51 13.69

13.41
13.28 13.35
13.5 13.10
13
12.5
EU8 female
12
Non-EU female
11.5
EU15 female
11
1971-1973
1974-1976
1977-1979
1980-1982
1983-1985
1986-1988
1989-1991
1992-1994
1995-1997
1998-2000
2001-2003
2004-2006
year of arrival

Notes: The X-axis shows the year of arrival in 3-year intervals and the Y-axis shows average years of schooling in 2006.
“EU8 female”, “EU15 female”, “Non-EU female” indicate average years of schooling of female immigrants from the EU8,
EU15, and remaining countries respectively.
Unfortunately, the microcensus data set does not contain information regarding
the wages of individuals; instead, it has information concerning the “main source
of income”12. We only use the information on net monthly income if the individ-
ual reports that the “main source of income” is from “working in a job”. We also
disregard earnings obtained while at school, or during vocational training or ma-
ternity leave. The net monthly income is reported in intervals with 24 income
brackets in all. This variable is converted into a continuous variable by taking the
midpoint of each interval. Furthermore, we calculate hourly wages by dividing the
monthly net income by 4 times the weekly hours worked. Employed people are all
those who are 15 and older and have worked at least one hour in the reference
week.
It is observed from raw data that EU8 immigrants earn less on average. Aver-
age income is 1,652 euros for EU8 men and 1,051 euros for women, while native
men earn 2,097 euros and women earn 1,298 euros. Furthermore, when we com-
pare the average income of immigrants who migrated to Germany just 3 years be-
fore and after enlargement, it seems that the average income of male EU8 immi-
grants dropped from 1,648 euros to 1,407; while it increased for women from EU8
countries from 806 euros to 1,111.
However, all these numbers are raw averages. As we will explain in detail in
the next section, observable characteristics of all these groups are different, and
this may lead to the wrong conclusion from just using raw averages. To address
the differences in observable characteristics between different immigrant groups
as well as the immigrants of the same group with different years of migration to
Germany, we effectively control for the observable characteristics of immigrants
in our analysis.
5 Empirical Analysis
In the previous section, we presented descriptive characteristics for different im-

migrants groups. As explained, all these groups differ in terms of their observable
characteristics including educational attainment, age at migration, employment
choices, location of migration choices, and so on. In this section, we compare im-
migrants’ income, hourly wage, working hours, employment status, and self em-
ployment probabilities relative to natives and across immigrant groups. We pre-
sent conditional means on these labor market measures using regression analysis,
in which we compare immigrants to natives after controlling for observable char-
12 Possible answers for “main source of income” are employment income, retirement pay-
ments, rents, investment income, unemployment benefits, social welfare or financial
grants.
acteristics. We analyze the natives and immigrant groups which were defined in
the previous section.
We report a cross-sectional analysis in Table 3 using the 2006 microcensus and

follow the basic specifications for immigrant earnings equations widely applied in
the literature: the variable of interest is regressed on individual characteristics
such as sex, age, education, marital status; state dummies to control for fixed dif-
ferences across German states; and dummies for each group. The reported stan-
dard errors are clustered by immigrant groups, accounting for the correlations in
outcome of individuals within the same group. The dependent variable is the natu-
ral logarithm of monthly income (in column 1), natural logarithm of hourly wage
(in column 2), weekly hours worked (in column 3), probability of employment (in
column 4), and the likelihood of self-employment (in column 5). In all these re-
gressions, the omitted group is “natives”. In the first 3 regressions, reported coef-
ficients can be interpreted as the mean difference in outcome of interest with re-
spect to German natives. For the remaining regressions, it is the difference in the
probability of being in work for each particular group of immigrants with respect
to natives, once observable controls are included.
Table 3
Labor market outcomes by citizenship
Logincome Loghourly-wage Work Hours Employment Self-employment

EU8 - Recent
immigrants -0.278 (0.005) -0.246 (0.006) -0.740 (0.038) -0.233 (0.006) 0.158 (0.004)
EU8 - Earlier
immigrants -0.211 (0.001) -0.166 (0.000) -1.198 (0.007) -0.128 (0.005) -0.009 (0.001)
Non-EU
immigrants -0.125 (0.009) -0.100 (0.008) -0.776 (0.034) -0.145 -(0.001) -0.020 (0.000)
EU15
immigrants -0.027 (0.006) -0.037 (0.005) 0.612 (0.013) -0.032 (0.002) 0.027 (0.000)
Native-Other
-0.100 (0.004) -0.086 (0.004) -0.630 (0.021) -0.063 (0.001) -0.021 (0.000)
Gender
-0.469 (0.009) -0.180 (0.007) -8.597 (0.052) -0.129 (0.006) -0.041 (0.000)
Age
0.266 (0.015) 0.147 (0.027) 1.998 (0.403) 0.001 (0.022) 0.058 (0.003)
Years of
education 0.081 (0.006) 0.066 (0.005) 0.489 (0.035) 0.028 (0.003) 0.009 (0.000)

Notes: Number of observations are 240,413 for first three regressions and 421,840 for last two regressions. Standard errors
clustered by state are shown in parentheses. Each regression also controls for polynomial of age up to degree four, and 16
state indicators. "EU8 - Recent immigrants" shows immigrants who arrived after 2004 from EU8 countries. "EU8 - Earlier
immigrants " shows immigrants who arrived before 2004 from EU8 countries. "EU15 immigrants" shows all immigrants from
EU15 countries. "Non-EU immigrants" indicates immigrants outside the EU. "Native-Other" indicates German citizens who
obtained German nationality after birth. The omitted group is German natives in all regressions.
The monthly income of recent EU8 immigrants is on average 28% less than
that of natives who have similar observable characteristics. In addition, earlier and
recent EU8 immigrants earn less compared to other immigrant groups. In a similar
vein, recent EU8 immigrants’ hourly wage is the lowest among all immigrant
groups, and recent EU immigrants earn on average 24% less relative to German
natives. They also work on average 0.75 hours less a week than natives. There-
fore, we may conclude that recent EU8 immigrants work on average more hours,
and their hourly wage is less compared to non-natives: hence, their monthly in-
come is lower than other immigrant groups.
In column 4 of Table 3, we analyze the employment status of immigrant

groups. We construct an employment status indicator which takes the value of 1 if
an individual is employed and zero otherwise. In Table 3, reported coefficients are
the difference in probability of being employed for the relevant group relative to
natives with similar observable characteristics. From column 4 of Table 3, it ap-
pears that recent EU8 immigrants are 23.7% less likely to be employed, and they
have the lowest labor force participation among all immigrant groups. Having
shown that immigrants from the EU8 have lower labor force participation, we
quantify whether this immigrant group is substantially different from other immi-
grant groups with respect to other dimensions of employment, i.e. self-
employment. The estimation results are displayed in column 5 of Table 3. This
column suggests that recent EU8 immigrants are 15% more likely to be self-
employed relative to natives. This finding is not surprising as Germany has very
few restrictions on being self-employed for EU8 migrants. However, it is still an
open question why the recent immigrants from the EU8 outperform other immi-
grant groups in terms of self-employment.
There are additional checks that might help us to understand the nature of im-
migration to Germany and how it has altered since EU enlargement. To gain a bet-
ter understanding of the effect of enlargement and the distribution of immigrants
who migrated to Germany after 2004, we divide all immigrant sub-groups into
two categories, recent and earlier, and report the results in Table 4. This specifica-
tion can help us to account for the assimilation or the integration effects. First, ear-
lier immigrants from the same source country with similar observable skills may
earn more than the recent arrivals only because they have spent more time in the
host country. Second, there might be differences in earnings between recent and
earlier immigrants due to the fact that the earlier immigrants may have had better
information about the local labor market in the host country, acquired the lan-
guage skills required from the higher skilled jobs, and had better access to training
in order to obtain country specific skills which are all unobservable in this data
set.
Table 4
Labor market outcomes by citizenship for all immigrant groups by arrival status
Logincome Loghourly-wage Work Hours Employment Self-employment
EU8 - Recent
immigrants -0.277 (0.003) -0.245 (0.005) -0.745 (0.037) -0.231 (0.008) 0.157 (0.005)
EU8 - Earlier
immigrants -0.210 (0.001) -0.166 (0.001) -1.199 (0.014) -0.127 (0.007) -0.009 (0.001)
Non-EU Earlier
immigrants -0.126 (0.007) -0.104 (0.006) -0.741 (0.041) -0.125 (0.002) -0.021 (0.001)
Non-EU Recent
immigrants -0.371 (0.003) -0.256 (0.005) -2.595 (0.076) -0.442 (0.012) -0.037 (0.002)
EU15 - Earlier
immigrants -0.031 (0.004) -0.045 (0.002) 0.794 (0.033) -0.023 (0.001) 0.025 (0.001)
EU15 - Recent
immigrants 0.058 (0.009) 0.114 (0.004) -1.627 (0.138) -0.220 (0.008) -0.010 (0.002)
Native-Other -0.099 (0.003) -0.085 (0.003) -0.630 (0.024) -0.061 (0.003) -0.021 (0.000)
Gender -0.469 (0.009) -0.179 (0.007) -8.597 (0.050) -0.129 (0.006) -0.041 (0.000)
Age 0.263 (0.014) 0.144 (0.024) 2.019 (0.373) 0.003 (0.024) 0.059 (0.002)
Years of
education 0.081 (0.006) 0.066 (0.005) 0.492 (0.032) 0.029 (0.002) 0.009 (0.000)

Notes: Number of observations are 240,413 for first three regression and 421,840 for last two regressions. Standard errors
clustered by state are shown in parantheses. Each regression also controls for polynomial of age up to degree four, and 16
state indicators. "EU8 - Recent immigrants ", "EU15 - Recent immigrants", " Non-EU Recent immigrants" show immigrants
who arrived after 2004 from EU8 countries, from EU15 countries, and outside the EU respectively. "EU8 - Earlier immi-
grants", "EU15 - Earlier immigrants", "Non-EU Earlier immigrants" shows immigrants who arrived before 2004 from EU8
countries, from EU15 countries, and outside the EU, respectively. "Native-Other" indicates German citizens who obtained
German nationality after birth. The omitted group is German natives in all regressions.
Table 4 suggests that the net monthly wage of recent EU8 and non-EU immi-
grants is 28% and 37% less, respectively, relative to natives with similar observ-
ables. Although recent EU8 immigrants work 1.8 hours a week more than recent
non-EU immigrants, their mean levels of hourly wages are not statistically differ-
ent from each other; hence, both groups’ hourly wage is on average 24% less than
that of natives. However, recent EU15 immigrants earn 6% more a month than na-
tives with similar characteristics. Furthermore, the hourly wage of recent EU15
immigrants is not statistically different from that of the natives, and they only
work 0.1 hours more a week, which is statistically significant. These results on re-
cent EU15 immigrants indicate that those who are generally better educated and
are high-skilled have a higher income premium during the recent years of migrat-
ing to and working in Germany. On the other hand, the net-migration number of
EU15 immigrants also indicates that unsuccessful ones have left Germany in re-
cent years. Recent EU8 immigrants work in low-paid jobs and work longer hours,
and their monthly income is the lowest on average among all immigrant groups
except recent non-EU immigrants.
Table 4 also suggests that EU8 immigrants are 23% less likely to work relative
to natives, and this is also similar for recent EU15 immigrants. Earlier EU15 im-
migrants are 2% more likely to be self-employed than natives. Also, the probabil-
ity of being self-employed for recent EU8 immigrants is 15% higher than for na-
tives and higher than all other earlier immigrant groups. However, recent EU8
immigrants have a lower probability of being employed compared to earlier im-

migrant groups.
We also estimate alternative specifications to address different concerns in the

immigration literature. First, we investigate whether the results presented in Table
3 and Table 4 are driven simply by the fact that earlier/recent immigrant groups
are more likely to work in specific industries and occupations. In this case, we
might mistakenly attribute the differences in the earnings profile across indus-
tries/occupations to earnings differences between recent and earlier immigrants.
The results in Table 3 and Table 4 are very similar after controlling for industry
and occupation fixed characteristics in addition to other control variables, which
further bolster our confidence that our results are not driven by differences in im-
migrants’ occupations or industries.
Second, we control for individuals’ years of schooling in our estimations.

However, years of schooling may not be as a good measure of education, since
source countries may have different time spans to obtain similar educational lev-
els. Therefore, we estimate the same specifications but include 3 educational cate-
gories instead of years of schooling in our analysis. We find that coefficients in
Table 3 and 4 are higher in absolute terms in this specification; nevertheless, the
results are in accordance with previous specifications for all groups.
Finally, another confounding factor for the main analysis is potential differ-
ences in earnings profiles across different age groups. In particular, individuals’
labor market earnings increase with age and experience. Therefore, if we do not
account for this factor, a differential age composition of immigrant groups can
lead to misleading conclusions, even though we exclude students and trainees
from the analysis. We might conclude that recent immigrants earn less relative to
earlier immigrants, while in reality the composition of EU8 immigrants has been
changing, i.e. some immigrant groups comprise of younger individuals than oth-
ers: hence, they have lower mean labor market earnings. To address this concern,
we exclude from the analysis individuals younger than 25 who work in mini-jobs.
The main results virtually remain unchanged after excluding these individuals, al-
though some of the coefficients change slightly. This additional analysis further
confirms that our results are not caused by the change in the age composition of
EU8 immigrants or other immigrant groups.
6 Conclusions and Policy Recommendations
The number of migrants from the accessions countries has clearly increased since
the enlargement of the EU in 2004. With enlargement, the net gain of EU8 immi-
grants has become 2.5 times larger than in the four-year period before enlarge-
ment. Among EU8 immigrants, Poles constitute the largest net immigrant group:
since enlargement, 71% of EU8 immigrants and 65% of all immigrants are from
Poland. On the other hand, without the immigration from the EU8, Germany
would have had a net loss of migration. Therefore, it is essential to understand
how this new immigrant influx from the EU8 countries has affected the labor
market conditions in Germany, and whether the composition of these immigrants
has altered since enlargement. This chapter presents unique evidence on the im-
pact of immigrant flows from EU8 countries on the German labor market after EU
enlargement.
Our analysis suggests that the composition of EU8 immigrants has indeed
changed since EU enlargement. The recent immigrants from EU8 countries are
comparably older and have lower levels of education relative to previous immi-
grants from EU8 countries. On the other hand, the majority of EU8 migrants are
male and young, and come from either Poland or the Baltic states. The change in
the distribution of recent immigrants is also reflected by their labor market out-
comes in Germany. We find that recent EU8 immigrants work in low-paid jobs
and work longer hours, and their net monthly income is the lowest on average
among all immigrant groups except recent non-EU immigrants.
In a similar vein, recent EU8 immigrants are 23.7% less likely to be employed,
and they have the lowest labor force participation among all immigrant groups ex-
cept recent non-EU immigrants. On the other hand, recent EU8 immigrants are
15% more likely to be self-employed relative to natives. EU8 immigrants work
1.8 hours a week more than recent non-EU immigrants, but their mean levels of
hourly wages are not statistically different from each other, whereas both groups’
hourly wage is 24% less than natives on average.
Taking these findings together suggests that recent EU8 immigrants in Ger-
many are more likely to be self-employed than employed as a wage earner in
Germany. Furthermore, these recent EU8 immigrants also earn less conditional on
being employed or self-employed. Given the large numbers of inflows and out-
flows of recent EU8 immigrants, they may serve as seasonal workers in low-paid
jobs or in agriculture. Our results suggest that recent EU8 immigrants are more
likely to compete with immigrants from outside Europe for low-skilled jobs than
compete with natives in Germany. Although Germany needs high-skilled immi-
grants, our analysis suggests that these immigrants only replace non-EU immi-
grants in low-skilled jobs. These results underline the importance of more open
immigration policies targeting high-skilled immigrants.
References
Blanchflower, D. G./H. Lawton (2009), The Impact of the Recent Expansion of the EU on
the UK Labour Market, in: M. Kahanec/K. F. Zimmermann (eds.), EU Labor Markets
After Post-Enlargement Migration. Berlin et al., 181-215.
Bonin, H./W. Eichhorst/C. Florman/M. O. Hansen/L. Skiöld/J. Stuhler/K. Tatsiramos/H.
Thomasen/K. F. Zimmermann (2008), Geographic Mobility in the European Union: Op-
timising its Economic and Social Benefits. IZA Research Report No. 19, Bonn.
Brenke, K./K. F. Zimmermann (2007), Zuwanderung aus Mittel- und Osteuropa trotz Ar-
beitsmarktbarrieren deutlich gestiegen, in: DIW-Wochenbericht, 44, 645-653, Berlin.
Bundesagentur für Arbeit (2008), Arbeitsmarkt in Zahlen, Arbeitsgenehmigungen/ Zu-
stimmungen 2007. Nürnberg.
Christen, T. G. (2004), Der Zugang zum deutschen Arbeitsmarkt nach der EU- Erweite-
rung, in: Bundesarbeitsblatt Nr. 3/2004, 10.
De Giorgi, G./M. Pellizzari (2006), Welfare Migration in Europe and the Cost of a Harmo-
nised Social Assistance. IZA Discussion Paper No. 2094, Bonn.
Fouarge, D./P. Ester (2007a), Determinants of Migration Intentions in Europe. Exodus or
Bounded Mobility? Institute for Labour Studies, Tilburg University, mimeo.
Fouarge, D./P. Ester (2007b), Factors Determining International and Regional Migration in
Europe. European Foundation for the Improvement of Living and Working Conditions,
Dublin.
Kahanec, M./K. F. Zimmermann (2008), Migration in an Enlarged EU: A Challenging So-
lution. IZA Discussion Paper No. 3913, Bonn.
Zaiceva, A./K. F. Zimmermann (2008), Scale, Diversity, and Determinants of Labour Mi-
gration in Europe, in: Oxford Review of Economic Policy 24 (3), 428-452.
5. The Experience of Spain with the Inflows
of New Labor Migrants
Sara de la Rica
The 2004 and 2007 EU enlargement has led to a significant increase in the immi-
gration flow to Spain. Individuals from the new EU12 countries accounted for no
more than 10% of the whole Spanish immigrant population in 2004, but by 2008
they accounted for almost 20% of the total flow of immigrants. Immigrants from
Bulgaria, Poland and Romania have accounted for 97% of new EU12 immigrants
since 2008. These immigrants are younger, and the vast majority of them are edu-
cated to the secondary level. Their employment rate is higher than that of natives,
but they are hit harder by unemployment. Our results point to two conclusions
from a policy prospective: first, EU enlargement has significantly improved legal
immigration from new EU12 countries. Second, the lack of employment assimila-
tion in terms of job quality for workers from the new EU12 countries may dis-
courage the entrance of highly qualified workers. The Spanish authorities should
provide on-the-job training for these qualified workers, so that they can find ade-
quate job prospects in Spain and decide to stay.
Migration, DOI 10.1007/978-3-642-02242-5_5,
132 Sara de la Rica
1 Introduction
Since the mid-1990s, the Spanish economy has been characterized by a continu-
ous growth in immigrant flows from African, Latin American and Non-EU15
countries. By January 2008, a total of 5,220,000 foreigners – 11% of the popula-
tion – resided in Spain (Padrón Municipal - 2007, Spanish Institute of Statistics).
The graph below charts the growth of immigration over the last 12 years.
Figure 1
Share of foreigners as a percentage of the total population of Spain (1996-2007)
12
10
8
% of foreigners
0
1996 1998 2000 2002 2003 2002 2005 2006 2007
Source: Spanish Padrón Municipal (Spanish Institute of Statistics).
As Figure 1 shows, the number of foreign-born people living in Spain grew

over the 12-year period from less than 1% of the population to approximately
10%. Various elements drove this trend: the rise of democracy; rapid economic
growth fuelled in part by Spain’s incorporation into the European Common Mar-
ket in 1986; the unrestricted entry of foreigners as tourists combined with lax im-
plementation of immigration laws; and close linguistic and cultural ties main-
tained with Latin Americans due to colonial history together with preferential
treatment (Escrivá 2000, Ribas-Mateos 2000).
In spite of stricter immigration restrictions comprising of limiting work and resi-

dency permit renewals, and immigration quotas implemented during the 1990s1,
Spain is currently considered the most popular destination country for Latino immi-
1 Since 1993 the Spanish government has maintained a quota system for farm laborers and
domestic servants. See Escrivá (2000) for more details.
5. The Experience of Spain with the Inflows of New Labor Migrants 133
grants (Millman and Vitzthum 2003). Spain also receives significant immigrant flow
from Africa, particularly Morocco, which is close to the Spanish mainland.
Furthermore, the enlargement of the European Union from fifteen to twenty-five

countries in 2004 and to twenty seven in 2007 has also contributed to a significant
increase in immigrants from the new EU12 countries. Immigrants from these new
member countries accounted for only 10% of the total immigration flow into Spain
in 2004, but now they account for 18% of the country’s total immigrant population.
This chapter seeks to address the impact of the 2004 and 2007 post-enlargement
migration flows into the Spanish labor market. In section 2 we offer a broad picture
of the main characteristics of this group of immigrants, their observed skills and
their labor market situation. Section 3 compares the labor market situation of these
new EU12 individuals to that of natives. Section 4 is devoted to the study of the la-
bor market assimilation of these immigrants into the Spanish labor market. By “la-
bor market assimilation” we mean how their labor market situation over the course
of their stay. We compare their assimilation process with those experienced by other
immigrants. Finally, section 5 presents a summary of the main findings and dis-
cusses some aspects of immigration policy in view of the results.
2 Main Characteristics of New EU12 Immigrants

in Spain
The table below presents the number of individuals born in each of the 12 new EU
countries who have lived in Spain in the last four years from 2004 to 2008. Taking
all individuals together, the first point to note is the significant increase in the in-
cidence of this group within the total number of immigrants. In 2004, about 10%
of the whole immigrant population were from the EU12. However, this figure has
currently increased to 18%. The second interesting issue is that Bulgaria, Poland
and particularly Romania have by far the highest incidence of the new EU12
countries. Even though Bulgaria and Romania joined the EU in 2007, the flow of
immigrants to Spain from these two countries has been very significant since
2004. These three countries carry great importance within the new EU12 group:
97% of the new EU12 immigrants who come to Spain are from one of the three
countries. Therefore, we pay particular attention to them in our descriptions of the
location and situation of immigrants.
134 Sara de la Rica
Table1
Number of individuals born in the new EU12 countries living in Spain
New EU12
countries 2004 2005 2006 2007 2008
Bulgaria 68,795 91,265 99,083 118,478 148,822

Cyprus 146 154 192 211 237
Slovenia 344 461 572 729 942
Estonia 461 626 750 939 1,108
Hungary 2,039 2,676 3,447 4,634 6,522
Latvia 1,113 1,377 1,680 2,021 2,380
Lithuania 9,040 12,128 15,116 17,615 19,963
Malta 137 154 157 182 217
Poland 26,526 34,636 43,291 58,296 74,436
Czech
Republic 3,408 4,310 5,305 6,558 8,018
Slovakia 2,387 3,405 4,342 5,673 7,078
Romania 203,173 308,856 394,078 507,736 701,065
All new EU12 317,569 460,048 568,013 723072 952,788

All foreign-
born 3,034,326 3,730,610 4,144,166 4,519,554 5,220,000
Source: Spanish Padrón Municipal (Spanish Institute of Statistics).
Before studying the individual and labor market characteristics of this immi-
grant flow, it is interesting to see whether they decide to locate in similar areas, or
whether individuals from different countries make different decisions regarding
their geographical situation. We do this by presenting the geographical situation of
each of the main three groups – Bulgarians, Poles and Romanians – with the loca-
tions of the remaining new EU12 immigrants aggregated into a single category.
Table 2 presents the regional distribution for the three most recent years, 2006-
2008, which is the period we analyze when evaluating the labor market situation
of this group of immigrants. Interestingly, the decision on where to locate varies
greatly depending on the country of origin.
The table shows that 31% of Polish immigrants reside in Madrid; while the pre-
ferred location for Bulgarians is clearly Castille-Leon (over 34%); and over 22%
of Romanians choose to live in Valencia. For the rest of the new EU12 immi-
grants taken as a single group, almost 25% choose to live in the Balearic and Ca-
nary Islands. The decision on where to locate has to do with two main issues: first,
given their lack of fluency in Spanish, these groups of immigrants need to reside,
at least initially, in areas where other immigrants from their country are already
located (networks). Second, it is very likely that networks are created in locations
where the probability of finding a job is highest for each nationality, given their
skills.
Table 2
Regional distribution of new EU12 individuals living in Spain (%)
Rest of new
Poland Bulgaria Romania
EU12
Andalusia 6.7 4.88 11.5 10.24
Aragón 3.8 2.09 10.9 3.6
Balearic and Canary
Islands 10.6 4.89 2.59 23.4
Castilla-León 7.25 34.6 6.22 4.14
Castilla-La Mancha 10.63 8.37 14.5 2.27
Catalonia 7.25 4.9 9.5 11.02
Valencia 9.66 18.1 22.13 18.53
Madrid 31.4 8.4 10.9 6.1
Other regions 30.6 16.7 11.76 20.7
Source: Spanish Current Labour Force Survey (2006-2008) - Spanish Institute of Statistics.
Now let us examine the main individual and labor market characteristics of
New-EU12 immigrants compared with those of natives. Table 3 summarizes the
main human capital characteristics as well as their work status. Information is
taken from the 2004 and 2008 Spanish Labour Force Surveys. We present data for
the most recent pre-enlargement period (2004) and for the most recent post-
enlargement period (2008) in order to assess possible differences.
Starting with demographic characteristics, it can be seen that the incidence of

immigrant women increased by 4 percentage points over the 4-year period. The
proportion of females among these immigrants is currently similar to that of na-
tives. In addition, the immigrants in this group are younger than the natives, as
expected, and the proportion of those married is similar to natives.
When education is analyzed, we find the majority of new EU12 immigrants are
educated to the secondary level (60% in 2004 and 65% in 2008). This increase in
the percentage of individuals with secondary education has coincided with a de-
crease of university-educated individuals (from 24% - similar to natives - in 2004
to 20% - significantly lower than natives - in 2008). Therefore, there seems to be a
change in the structure of the new EU12 immigrants who decide to come to Spain.
Highly educated individuals from the new EU12 countries seem to be less eager to
ome to Spain now than they were some years ago. It would be very interesting to
investigate further the potential reasons for this observed change, but unfortu-
nately we do not have any information about the underlying reasons why these in-
dividuals decide not to come to Spain. One potential reason is suggested in our
following examination of the labor market situation and labor market assimilation
of this group of immigrants as their stay in Spain lengthens.
136 Sara de la Rica
Table 3
Main characteristics of individuals born in the new EU12 countries compared to natives
Natives New EU12 countries
2004 2008 2004 2008

0.5 0.51 0.48 0.52
Female
-0.49 -0.49 -0.5 -0.49
39.7 41.09 32.54 33.16
Age
-13.8 -13.8 -9.37 -10.1
0.48 0.56 0.55 0.57
Married
-0.5 -0.49 -0.49 -0.49
Education
0.26 0.21 0.15 0.15
Primary or less
-0.43 -0.41 -0.36 -0.36
0.5 0.52 0.6 0.65
Secondary
-0.49 -0.49 -0.49 -0.48
0.24 0.27 0.24 0.2
University
-0.42 -0.44 -0.43 -0.39
Work status
0.54 0.58 0.7 0.66
Employed
-0.49 -0.49 -0.45 -0.47
0.08 0.067 0.13 0.14
Unemployed
-0.28 -0.25 -0.33 -0.35
Type of employment
for the employed
0.045 0.19 0.04 0.38

Fixed-term
-0.21 -0.39 -0.2 -0.48
0.56 0.59 0.3 0.43
Indefinite
-0.49 -0.49 -0.46 -0.49
0.11 0.01 0.4 0.001
No written contract
-0.31 -0.1 -0.49 -0.09
No. of observations 110.839 101.798 577 1567
Source: Spanish Labour Force Survey. Individuals between 16 and 65 years of age.
The third group of variables depicted in Table 3 describes the labor market
situation of individuals from the new EU12 countries as compared to natives. The
percentage of workers is much higher among immigrants than among natives, re-
flecting at least partially their lower reservation wage. On the other hand, unem-
ployment hits new EU12 immigrants harder than it hits natives. Among workers,
type of employment is another interesting feature that reveals information about
employment quality. Spain is a country where the incidence of fixed-term con-
tracts is particularly high among young workers. Natives tend to associate such
contracts with job insecurity, because on-the-job training is normally lower in jobs
with fixed-term contracts than in those under indefinite contracts. However, given
that the long-term perspectives of immigrants tend to differ from those of natives,
having a fixed-term contract does not necessarily mean holding a low-quality job.
The percentage of new EU12 immigrant workers with fixed-term contracts has in-
creased significantly over this 4-year period, coinciding with the practical disap-
pearance of non-written contracts. This change represents a great improvement in
job quality for workers, as non-written contracts do not provide workers with any
employment protection.2 In summary, workers from the new EU12 countries have
greatly improved their working status. Not only has the proportion of workers
with an indefinite contract increased, but also (and mainly) there are hardly any
workers with non-written contracts.
3 Labor Market Situation of New EU12 Immigrants

Compared to Similarly Skilled Natives
In Section 2 we saw that natives and the group of new EU12 immigrants differ in
some observable individual, demographic and human capital characteristics. If the
labor market situation of these immigrants is to be compared with that of natives,
it is important to control for these observed differences and study the extent to
which they persist when immigrants are compared to similarly skilled natives. In
this section we analyze two specific issues regarding the comparison of new EU12
immigrants and natives: first, we estimate the adjusted propensity to work for all
individuals (natives, new EU12 immigrants and other immigrants) once demo-
graphic, individual, regional and time indicators are taken into account. In addi-
tion, new EU12 immigrants and other immigrants are disaggregated into recent (5
years and less since arrival) and non-recent immigrants (more than 5 years since
arrival). The results of the estimates are presented in the first column of Table 4.
Second, we estimate the propensity of having a blue-collar job relative to a

white-collar job.3 The rationale behind this distinction is that, by definition, blue-
collar jobs involve more manual tasks than white collar jobs. They also tend to be
generally low-qualified and less desirable jobs.4 We want to measure to which ex-
tent new EU12 immigrants are found in low-qualified jobs compared to other im-
migrants and natives. The results are displayed in column (2) of Table 4.
2 In principle, non-written contracts are not legal in Spain. However, compliance with the
lawhas been enforced more successfully in recent years than before.
3 Specifically, we classify blue collar jobs as the occupations with the following 2-digit
ISCO88 codes: 51 (personal and protection service workers), 61 (Skilled agricultural and
fishery workers), 71-74 (Craft and related trade workers), 81-84 (Plant/machine operators
and assemblers) and 91-94 (Elementary Occupations). The remaining occupations are
grouped into the so-called white collar job group.
4 In our sample, the incidence of native workers compared to EU12 workers in blue-collar
jobs is 39.5% to 72%. If we take education into regard, we find the incidence of blue-collar
jobs is 71% to 79% among workers with primary education; 48% to 74% among workers
with secondary education; and among highly educated workers, the incidence of blue-
collar jobs is 12% for natives to 63% for new EU12 immigrants.
138 Sara de la Rica
Table 4
Labor market situation of new EU12 individuals compared to natives
Probability of Being Probability of Working in
Independent Employed a Manual Job
Variables (1) (2)
0.03 0.91
Recent New EU12
(0.028) (0.04)
-0.12 0.39
Non-recent New EU12
(0.04) (0.05)
Recent Other than 0.06 0.66
New EU12 (0.01) (0.01)
Non-recent Other than -0.07 0.22
New EU12 (0.01) (0.018)
-0.68 -0.81
Female
(0.004) (0.006)
0.23 0.022
Age
(0.001) (0.001)
2 -0.002 -0.0003
Age
(0.0001) (0.00002)
0.24 -0.7
Secondary Education
(0.006) (0.009)
0.58 -1.8
University Education
(0.007) (0.011)
Region Dummies (17) Yes Yes
Year Dummies (2) Yes Yes
No. of observations 358.492 192.391
In column (1) of Table 4, the dependent variable is set to one if a respondent in

the Spanish Labour Force Survey reports that he/she is employed and zero if ei-
ther unemployed or out of the labor force. The reference group is “Natives”, so the
coefficient reported in the table are the difference in the probability of being em-
ployed for each particular group of immigrants with respect to natives, once ob-
servable individual, human capital and other (regional and time) controls are in-
cluded. Looking at these coefficients, it can be seen that there is no difference in
the probability of being in work for recent EU12 immigrants and natives. If we re-
fer to non-recent EU12 immigrants, we see that their probability of being in work
is smaller than that of natives, i.e. non-recent EU12 immigrants are doing worse in
terms of employment than their recent counterparts. When we compare this group
of immigrants with the rest, we can see that the other recent immigrants are doing
better than natives in terms of the probability of being in work. Thus, they are also
doing better than recent EU12 immigrants. However, when comparing the prob-
ability of being employed for two groups of workers we must consider both sup-
ply and demand factors. It is not possible to infer from this result whether this ob-
served better performance of other recent immigrants when compared to recent
EU12 immigrants is due to their lower reservation wage or to a higher demand for
other immigrants from the side of employers. The analysis in the next section en-
ables us to investigate this further.
What about differences in types of job? Column (2) of Table 4 displays differ-
ences in the probability of workers in a blue-collar job compared to a white-collar
job. The reference group is employed natives. When comparing new EU12 immi-
grants with natives, we can see that, once differences in observed individual and
human capital characteristics are controlled for: (i) the probability of new EU12
immigrants with a blue-collar job is much higher than that of comparable natives;
(ii) that probability diminishes if their stay in Spain is long enough (at least 5
years); (iii) when we compare the new EU12 group with other immigrants, we see
that the probability of doing manual jobs is much higher for the former than the
latter. This result tells us that compared with their counterpart immigrants from
elsewhere, the new EU12 group seem to be more concentrated in manual and
hence, low-qualified jobs.
In the following section, we look at this result in greater depth in search of a

neater idea of the assimilation process that different immigrant groups experience
when entering the Spanish Labor Market.
4 Labor Market Assimilation of New EU12

Immigrants Compared to Similarly Skilled Natives
and Other Immigrants
So far we have looked at the main features of the overall labor market situation of
new EU12 immigrants compared to natives and other immigrants. However, given
the observed increase of immigrant flow into Spain, we must look not only at the
overall situation but also at their adaptability to the Spanish labor market during
their stay in Spain. A good understanding of the assimilation process of these
workers into the Spanish labor market is crucial in the development of social poli-
cies facilitating the integration of these newcomers into Spanish society. In the
following, we examine changes in the likelihood of being employed for new
EU12 immigrants, as well as changes in the probability of being in a blue-collar
job, the longer they stay.
To do this, we estimate (i) the probability of having a job, and (ii) the probabil-
ity of being employed in a blue-collar job relative to a white-collar one for natives
and immigrants by allowing for differences in the labor market situation of these
new EU12 individuals during their stay. We do this by including dummy variables
that indicate whether a migrant has been in the country for 0, 1, 2, 3, 4, 5 or more
than 5 years. Thus, the immigrant dummy variable simply reveals the difference
between the labor market situations of an immigrant who has been in the country
for a particular number of years and that of a native. Assimilation takes place
140 Sara de la Rica
when the difference between natives and immigrants diminishes the longer they
stay in Spain.
Table 5
Labor market assimilation of new EU12 individuals compared to natives
as their stay in Spain lengthens
Probability of being in Probability of working in a
Independent variables work manual job
(1) (2)
-0.220 0.816
New EU12 0 years since entry
(0.086) (0.132)
-0.137 0.949
(0.070) (0.1069
-0.038 0.800
(0.070) (0.100)
0.168 0.962
(0.069) (0.090)
0.110 0.933
(0.070) (0.091)
0.253 1.017
(0.071) (0.090)
-0.051 0.469
New EU12 >5 years since entry
(0.037) (0.049)
-0.168 0.808
Other immig 0 years since entry
(0.046) (0.071)
-0.019 0.700
(0.038) (0.055)
0.079 0.629
(0.038) (0.051)
0.159 0.678
(0.036) (0.0471)
0.099 0.668
(0.034) (0.045)
0.116 0.660
(0.032) (0.042)
-0.034 0.313
Other immig >5 years since entry
(0.013) (0.017)
No. of observations 358,492 192,391
Notes: Both estimates include indicators for gender, age, age square, education, region dummies (17) and year dummies
(2). The reference group is a native individual.
The first column of Table 5 reports differences between natives (the reference
group) and immigrants in the probability of working. The latter are disaggregated
by place of origin (new EU12 and others) and by the length of the stay in Spain.
When we compare new EU12 immigrants with natives, we observe that the prob-
ability of working is much lower for newcomers from the new EU12 countries
than for their native counterparts, once controls for gender, age, education, region
and time have been included. However, this differential diminishes greatly as their
stay in Spain increases, and disappears by their second year in the country. Indeed,
new EU12 immigrants are more likely to work than natives once they have been
in the country for at least three years. We can also compare the process of assimi-
lation into the labor market of this group of immigrants with that of other immi-
grants by comparing the differences in the probability of being in work of each

group of immigrants with that of natives. With regards to assimilation, we can see
that new EU12 immigrants do worse than the others at entry and until they have
been in Spain for three years, as their differential in the probability of being in
work in comparison with natives is bigger. However, differences between new
EU12 immigrants and the rest disappear once they have been in Spain for at least
3 years. This difference in the assimilation process between new EU12 immi-
grants and others is very likely to be related to language fluency. Most immigrants
in the “other immigrants” group are Latinos, who speak Spanish, so they do not
face the language barrier. After three years of living in Spain, it is very likely that
the language barrier disappears. Therefore, we see that the difference among dif-
ferent groups of immigrants in the probability of being employed also disappears.
However, the likelihood of having a job only partially indicates assimilation:

another important factor is whether or not there is an increase in the probability of
finding a “better” job as time passes. Most studies that focus on assimilation into
the labor market analyze changes in the wages of immigrants as their stay in the
destination country increases. Examples can be seen in Chiswick (1978) and Bor-
jas (1985) for the US; Hunt (2004) for Canada; Longva and Raaum (2003) for
Norway; Hartog and Winkelmann (2003) for the Netherlands; Bevelander and
Nielsen (2001) for Sweden; Constant and Massey (2003) for Germany; and
Wheatly Price (2001), Dustmann (2000) and, more recently, Clark and Lindley
(2005) for the UK. However, we do not have any information on wages in the re-
cent period we analyzed. Instead, we can try to address the issue of whether im-
migrants obtain better jobs as time passes by comparing the type of jobs they hold
compared to their native counterparts. Generally speaking, blue-collar jobs are
less qualified, and we could regard them as “worse” jobs than white-collar jobs.
We saw in the descriptive section that the vast majority of new EU12 immigrants
are educated to secondary level. The incidence of blue-collar jobs among natives
with secondary education is 48%; whereas the incidence of blue-collar jobs
among similarly educated new EU12 immigrants is 74%.
We pose the question whether new EU12 immigrants tend to change to white-
collar jobs the longer they stay. This is answered in the second column of Table 5.
As before, the coefficient must be understood as the differential in the probability
of holding a blue-collar job (as opposed to a white-collar one) with respect to a
native. The first thing to note is that the probability of newcomers from new EU12
countries having a blue-collar job as opposed to a white-collar job is much higher
than for their native counterparts. In addition, there does not seem to be any sig-
nificant assimilation in terms of a change to white-collar jobs the longer they stay
(except for those who have been in Spain for more than 5 years, but that is a very
heterogeneous group). If we indirectly compare the assimilation process with
other immigrants in terms of job qualification, we see that although newcomers
from different origins seem to behave very similarly, the differential in job quali-
142 Sara de la Rica
fication between immigrants not from the new EU12 and natives diminishes to a
great extent over time. There is some evidence of “job quality” assimilation for
other immigrants, but not for new EU12 immigrants. This lack of job quality as-
similation might go at least some way towards explaining the fact that the educa-
tion level of new EU12 immigrants is decreasing over time. We saw in Table 3
that the incidence of individuals from new EU12 countries with university studies
had decreased by 4 percentage points over the last 5 years. If highly educated
workers do not see prospects for job quality improvement, they may tend either to
stay in their country or choose another EU country where the job prospects are
better.
5 What Have We Learnt?

Summary and Policy Conclusions
The 2004 and 2007 EU enlargements have led to a significant increase in the im-
migration flow to Spain. In 2004 individuals from the new EU12 countries ac-
counted for no more than 10% of the whole Spanish immigrant population, but by
2008 they accounted for almost 20% of the whole immigrant flow. In 2008 immi-
grants from Bulgaria, Poland and Romania accounted for 97% of the new EU12
immigrants. They do not all decide to settle in the same areas of Spain: Bulgarians
tend to locate in Castille-Leon; Poles in Madrid; and Romanians in Valencia. The
geographical concentration of immigrants by nationality reveals the importance of
networks, which can help newcomers when they first arrive.
These new EU12 immigrants are younger, and the vast majority of them are
educated to secondary level. A decrease has been observed in the incidence of
highly educated immigrants since 2004. Their employment rate is higher than that
of natives – partly because their reservation wage is probably lower; however, un-
employment hits them harder than natives. Another interesting feature of the labor
market situation of these workers is that the percentage of workers with non-
written contracts has fallen to practically zero. The fact that workers from the new
EU12 countries have enjoyed job mobility throughout the EU since enlargement
seems to be a very important determinant for having a written contract – and thus
contributing to the social security system and obtaining job security.
They are less likely to be employed than their native counterparts and other
immigrants (most of whom are Latinos, who speak Spanish), but after a two-year
stay in Spain that differential disappears. However, there is no evidence of an im-
provement in job qualification of new EU12 immigrants compared to other immi-
grants. This lack of prospects may discourage highly educated workers from these
countries coming to Spain, and encourage them to look for other European coun-
tries which could provide them with better future prospects in terms of job quality.
Our results point to two conclusions from a policy perspective: first, EU

enlargement has significantly improved legal immigration from the new EU12
countries. This is very positive from the workers’ viewpoint – they can enjoy the
same job security as any native worker – and also from the Spanish perspective –
these workers contribute to social security. Second, the lack of employment as-
similation in terms of job quality for workers from the new EU12 countries may
discourage highly qualified workers from entering. Spanish employers repeatedly
voice the need for a highly qualified workforce from abroad, but if potential im-
migrant workers do not see good job prospects they may decide to try other Euro-
pean countries which offer better qualified jobs. The Spanish authorities should
provide on-the-job training for these skilled workers so that they can find adequate
job prospects in Spain and decide to stay.
144 Sara de la Rica
References
Bevelander, P./ H. S. Nielsen (2001), Declining employment success of immigrant males in

Sweden: Observed or Unobserved characteristics?, in: Journal of Population Economics
14 (3), 455-471.
Borjas, G. (1985), Assimilation, Changes in Cohort Quality, and the Earnings of Immi-
grants, in: Journal of Labor Economics 3, 463-489.
Chiswick, B. (1978), The Effect of Americanization on the Earnings of Foreign-Born
Men, in: Journal of Political Economy 86 (5), 897-921.
Clark, K./J. Lindley (2005), Immigrant Labour Market Assimilation and Arrival Effects:
Evidence from the Labour Force Survey. Sheffield Economic Research Paper Series
No. 2005004, University of Sheffield.
Constant, A./D. S. Massey (2003), Self-selection, earnings, and out-migration: A longitudi-
nal study of immigrants to Germany, in: Journal of Population Economics 16 (4), 631-
653.
Dustmann, C. (2000), Temporary Migration and Economic Assimilation, in: Swedish Eco-
nomic Policy Review 7(2), 213-244.
Escrivá, A. (2000), The Position and Status of Migrant Women in Spain, in: F. Anthias/G.
Lazaridis (eds.), Gender and Migration in Southern Europe: Women on the Move. Ox-
ford et al., 199-225.
Hartog, J./R. Winkelmann (2003), Comparing migrants to non-migrants: The case of Dutch
migration to New Zealand, in: Journal of Population Economics 16(4), 683-705.
Hunt, J. (2004), Are Migrants More Skilled than Non-migrants? Repeat, Return, and Same-
Employer Migrants, in: Canadian Journal of Economics 37 (4), 830-849.
Longva, P./O. Raaum (2003), Earnings assimilation of immigrants in Norway – A reap-
praisal, in: Journal of Population Economics 16 (1), 177-193.
Millman, J./C.Vitzthum (2003), Changing Tide: Europe Becomes New Destination for La-
tino Workers. The Wall Street Journal, 12 September.
Ribas-Mateos. N. (2000), Female Birds of Passage: Leaving and Settling in Spain in: F.
Anthias/G. Lazaridis (eds.), Gender and Migration in Southern Europe: Women on the
Move. Oxford et al., 173-197.
Wheatly Price, S. (2001), The employment adjustment of male immigrants in England, in:
Journal of Population Economics 14 (1), 193-220.
6. EU Enlargement and Ireland’s Labor Market
Alan Barrett
Ireland, along with Sweden and the UK, allowed full access to its labor market to
the citizens of the accession countries when the EU enlarged in May 2004. Given
the limited number of countries that opened up and the rapid pace of economic
growth in Ireland around 2004, a significant inflow was expected. However, the
rate of inflow exceeded all expectations. Based on census information, the number
of EU10 nationals living in Ireland grew from around 10,000 in 2002 to 120,000
in 2006. Data on inflows suggests that this number could have reached 200,000
by 2008 or 4.5% of the population. The EU10 immigrants have very high em-
ployment rates and also have levels of education that are comparable to the native
labor force in Ireland. However, they appear to earn considerably less than the
native labor force and also to be in lower grade occupations. They have impacted
positively on the Irish economy in terms of GNP growth. This is because wages
grew more slowly in Ireland than would otherwise have been the case as a result
of the labor supply increase brought about by this immigration flow.
Migration, DOI 10.1007/978-3-642-02242-5_6,
146 Alan Barrett
1 Introduction
The ten years up to 2008 saw extraordinary growth in the Irish economy. The
unprecedented period of growth lead to many changes in Ireland’s economy and
society. Among the most significant of these changes was the reversal of years
of migratory outflows, and the inflow of large numbers of immigrants. These in-
flows began around the mid-1990s but accelerated dramatically after EU
enlargement. Given the speed with which the economy was growing around
2004, the Government decided to allow full access to the Irish labor market to
all citizens of the European Union’s New Member States (EU10) from the date
of accession, i.e. 1 May 2004.1 Only the governments of Sweden and the UK
joined them in making this decision. This gave rise to a situation in which po-
tential migrants from the EU10 had a limited number of destinations from which
to choose if they wanted to work legally. With Ireland growing so strongly
around 2004, it was perhaps inevitable that a large inflow from the EU10 would
result.
In this chapter, we will explore the size, nature and impacts of this migratory
inflow from the EU10 to Ireland. In Section 2, we present figures on the size of
the inflow, relative to the population and to the labor force. In Section 3, we look
at the inflow in greater detail and review what recent research tells us about the
characteristics and experiences of EU10 immigrants in Ireland, with a particular
emphasis on the labor market. In Section 4, we move to the issue of impact, and
we present estimates of the impact of EU10 immigration. In Section 5, we draw
the various strands together and provide an overall assessment of how the migra-
tion of people from the EU10/12 impacted upon Ireland’s labor market.
2 The Size of the Inflow
It is important to set this discussion of migration into Ireland in the broader con-
text of macro-economic developments. Let us begin this section by looking at Ire-
land’s growth performance since 1990, its levels of employment and rates of un-
employment.
1 The same freedoms were not extended to Romanian and Bulgarian citizens when their
countries joined the EU. As a result, most of the discussion in this chapter refers to immi-
grants from the EU10 only. On occasions, we do make reference to immigrants from the
EU12 because official data is sometimes published on that basis. The precise group to
which we are referring is specified where relevant throughout the text, although the differ-
ence in numbers between the EU10 and the EU12 is slight.
6. EU Enlargement and Ireland’s Labor Market 147
Figure 1
Real GNP growth rates (1990-2007)
12.0
10.0
8.0
6.0
4.0
2.0
0.0
1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007
Source: Central Statistics Office.
Figure 1 shows growth rates of real GNP. The exceptional performance of the
Irish economy is clear: growth rates averaged 5.7% over this period, earning Ire-
land what is now the familiar label of the “Celtic Tiger”. What is of more rele-
vance for current purposes is the growth performance in the period around EU
Accession. Like many countries, Ireland experienced a dip around 2001/2. How-
ever, the economy soon bounced back and recorded average growth rates of just
over 5% between 2003 and 2007.
We can get another view of the extraordinary growth performance by looking

at changes in the level of employment over the same time period. In 1990, there
were 1.16 million people working in Ireland. By 2007, this number had risen to
2.1 million, an increase of 80%. Within the period 2003 to 2007 alone, employ-
ment grew from 1.79 million to 2.1 million, an increase of almost 17%. As a final
element in this macro-overview, we consider the rate of unemployment. The rate
peaked at almost 16% in 1993 before beginning a downward slide through to
2001. In that year, the rate fell to below 3.7%. The rate did rise again after 2001
but remained low by contemporaneous European standards and only reached 4.5%
in 2007.
Let us now consider migration in the context of a longer historic picture by

way for the later discussion of migration in the period after 2004.
148 Alan Barrett
Figure 2
Net migratory flows 1987 to 2008, with forecast for 2009
80
60
40
20
-20
-40
-60
87
88
89
90
91
92
93
94
95
96
97
98
99
00
01
01
03
04
05
06
07
20 8
(f)
0
19
19
19
19
19
19
19
19
19
19
19
19
19
20
20
20
20
20
20
20
20
20
09
Source: Economic and Social Research Institute (ESRI) databank.
Figure 2 shows net inflows into Ireland over the period 1987 to 2008, and a
forecast for 2009. The years from the late 1980s capture that fact that Ireland was
experiencing large population outflows as recently as then. Between 1987 and
1990, the net outflow totaled 131,000 – as the population in 1991 was just 3.5 mil-
lion, that net outflow was around 4% of the population and came on top of an ad-
ditional net outflow of some 70,000 in the earlier part of the 1980s.
Even though the economy began to grow rapidly in 1994, it was not until 1997
that significant net inflows emerged. In that year, the net inflow was 19,200. In
general, the net inflow had been increasing up to 2001 when it reached 41,300.
However, even this figure was to be somewhat eclipsed following EU Accession.
The net inflow averaged almost 65,000 in the three years between 2005 and 2007.
Having looked at the economy and migration in broad terms, let us look in
greater detail at the period 2004 to 2007 and, in particular, at the inflow of EU10
nationals. Our first task is to get a sense of the size of the EU10 inflow, both rela-
tive to the population in general and also relative to other immigrant groups. We
begin by looking at measures of stocks in the censuses of 2002 and 2006, and in-
formation on flows.
Table 1
Stocks of Irish and non-Irish nationals in Census 2002 and Census 2006
2002 2006
Population totals 3858495 4172013
Total Irish 3584975 3706683
Total non-Irish 224261 419733

Not stated 48412 44279
% Non-Irish (including not stated) 7 11
Non-EU Europe (including 9,101 people from Russia, 23105
Romania and Ukraine)
EU10 120534
Table 1 shows how the population grew by over 300,000 or 8%. It also shows
that the population of non-Irish people living in Ireland grew by almost 200,000:
almost two thirds of Ireland’s population increase between 2002 and 2006 was
driven by the inward migration of non-nationals. If those who failed to provide a
nationality in the census questionnaires in the group of non-nationals are included,
the non-national proportion of Ireland’s population grew from 7% in 2002 to 11%
in 2006. We can put this rate of growth in context by noting that the immigrant
share of the UK population grew by 2 percentage points over the thirty-year pe-
riod, 1960-1990 (Bell, 1997). Over the same period, the immigrant share in the
US grew by 2.5 percentage points (Bell, 1997). In this context, the four percentage
point increase in the share of immigrants in Ireland’s population over a four year
period is quite remarkable.
It is not possible to determine an exact figure for the number of immigrants

from the EU10 in 2002 based on published data, but it is possible to infer an upper
bound figure. Table 1 shows there were 23,105 immigrants from non-EU Euro-
pean countries living in Ireland in 2002. We know that 9,101 were from Russia,
Romania and the Ukraine. Hence, a maximum of 14,000 were from what we now
refer to as the EU10. We do have exact figures on the number of immigrants from
the EU10 for 2006. The total was 120,534, including 63,276 Poles, 24,628
Lithuanians and 13,319 Latvians. These numbers imply that between 2002 and
2006 the stock of immigrants from the EU10 in Ireland grew by over 100,000.
In order to get a sense of the size of the continuing flow after the census in
April 2006, we can look at figures from the Central Statistics Office. We present
CSO information on flows into Ireland between 2003 and 2008 in Table 2. In the
years before 2004, immigrants from the EU12 were included in the “rest of world”
category. However, for 2005 onwards, they are separately identified and a con-
tinuing strong inflow after 2006 is clear. In the two years ending April 2008,
150 Alan Barrett
86,400 immigrants from the EU12 had settled in Ireland. While there was also an
outflow to these countries of 16,000, the net inflow of 70,000 between 2006 and
2008 suggests that the population of EU12 immigrants would have been around
200,000 by April 2008. As the total population was estimated to be 4.42 million in
2008, this implies that immigrants from the EU12 made up 4.5% of the total popu-
lation. If we recall that there proportion in 2002 was less than 0.5%, the size of the
inflow relative to the existing population is clear.
Table 2
Inflows by nationality, 2003-2008 (and outflow to EU12)
2003 2004 2005 2006 2007 2008
Irish 17.6 16.7 18.5 18.9 20 16.2
UK 9.1 7.4 8.9 9.9 5.9 7
Rest of EU 15 8.8 13.3 9.3 12.7 10.4 8.6
EU 12 34.1 49.9 52.7 33.7
USA 2.1 2.3 2.1 1.7 2.8 2
Rest of world 22.4 18.8 11.6 14.7 17.8 16.3
Total 60 58.5 84.5 107.8 109.6 83.8
Total outflows 29.3 26.5 29.4 36 42.2 45.3

Of which, number to EU12 0.8 2.3 7 9
As a final view on the size of the inflow of EU12 immigrants into Ireland after
accession, we can look at administrative records and, in particular, the issuing of
social security numbers (known as PPS numbers in Ireland). We do this because a
very different impression of the size of the inflow is generated if we take this per-
spective, and so it is useful to explore the discrepancy between the data just pre-
sented and the administrative data.
In Table 3 we present data on the numbers of PPS numbers issued to EU 10 na-

tionals between 2002 and 2006. It should be recalled from above that our com-
parison of the stocks of EU 10 nationals in the censuses of 2002 and 2006 sug-
gested that the net inflow was in the order to 100,000. However, we can see from
Table 3 that well over 300,000 PPS numbers were issued to EU10 nationals over
this period. Even allowing for the fact that the 2006 Census was in April, and so
many of the PPS numbers issued in 2006 would have been to people arriving after
April, there is a huge discrepancy between the census/labor force survey data and
the administrative data. However, a clue to the reason underlying the discrepancy
is included in the administrative data and is shown in Table 3.
Table 3
Number of PPS numbers issued to EU10 nationals 2002-2006, along with % where
no employment activity is registered and % with employment activity in each year
% with no
Allocation employment % with employment activity by year
activity
2002-2006 2002 2003 2004 2005 2006
2002 8975 14 75 69 59 55 51
2003 8967 12 6 73 73 62 57
2004 58079 1 0 1 79 75 65
2005 107451 18 0 0 1 71 69
2006 127693 34 0 0 0 0 66
The first column in Table 3 after the “allocation” column shows the proportion
of PPS numbers issued for which no employment activity was recorded. Taking
2005 as an example, we see that of the 107,000 numbers issued almost 20% were
issued to people who subsequently did not show up as working. There are a num-
ber of possible reasons for this. First, the individuals could have worked in situa-
tions where they evaded tax. Second, the individuals could have looked for work
in Ireland but may have moved on to the UK or back home without having
worked here. A third possibility is that people applied for PPS numbers but with-
out any intention of working, although this is probably the least likely. As the Irish
Government imposed a two-year residency requirement in order to be eligible for
welfare payments, there would be little point in EU 10 immigrants applying for
PPS numbers unless they did intend working.
The last four columns of Table 3 show the proportion of PPS numbers issued in
each year that were “active” across years, where “active” refers to some level of
work being recorded. Taking 2005 again as an example, we can see that 69% of
PPS numbers were still active in 2006, implying that 31% of those receiving num-
bers in 2005 were not working (legally at least) in 2006, and so may have left Ire-
land. A similar picture emerges for 2004.
While the administrative data suggest a much bigger inflow into Ireland rela-
tive to the census/labor force survey data, they also point to a short-term dimen-
sion in some of the migratory moves. This is an important point because it is
likely to have implications which may not be captured in the way we model the
impacts of immigration. As discussed in the section on the impacts of EU flow,
the effects of immigration in Ireland has been estimated by using a statistically
generated model of the labor market and by conducting simulations. If this large
and temporary inflow of labor has altered the way in which the labor market func-
tions, then the modeling approach might miss this.
152 Alan Barrett
3 The Characteristics and Experiences of the Inflow
We have looked at the size of the EU12 immigrant inflow in broad terms, but let
us now focus on the labor market dimensions: firstly, by discussing the labor-
related characteristics of EU12 immigrants; then by examining the labor market
experiences of these immigrants in terms of their earnings and occupational at-
tainment. In this section, we draw on existing research and begin with the most
up-to-date data published by the CSO.
In Table 4, we show the numbers employed, unemployed and inactive in Ire-

land by national groups and also the unemployment and employment rates for
early 2008. The first point to be taken from the table is the size of the EU12 group
as a proportion of Ireland’s labor force. There are 178,800 EU12 immigrants in
Ireland’s labor force, which amounts to 8% of the total of 2,223,900. We esti-
mated above that the EU made up about 4.5% of the population, so the 8% points
is a significant over-representation in the labor force. The two main reasons for
this would be the under-representation of children in this immigrant group and
also the high employment rate. While the Irish have an employment rate of 59%,
the rate for the EU12 is 80%. This employment rate places the EU12 group above
all other immigrant groups and by a significant margin. Their unemployment rate,
at 6.5%, is above that of the Irish but could well be explained by the presence of
recent arrivals and thus jobseekers.
Table 4
Labor force statistics on Ireland’s immigrants aged 15 and over (2008)
Employed Unemployed UE rate (%) Inactive Total Employment rate (%)
Irish 1770.7 90.7 0.049 1164.7 3026.1 0.59
UK 54.5 4.2 0.072 41 99.7 0.55
EU15 31.2 1.8 0.055 10.2 43.2 0.72
EU12 167.2 11.6 0.065 29.3 208.1 0.8
Other 84.8 7.2 0.078 41 133 0.64
Total 2108.4 115.5 0.052 1286.2 3510.1 0.6
In order to get a sense of the educational profile of EU10 immigrants, we need

to draw on earlier research. Barrett and Duffy (2008) show the education levels of
EU10 immigrants, which we present in Table 5, along with those of other immi-
grants based on data from 2005. The first point to be taken from the table is that,
in general, Ireland’s immigrants are relatively highly educated. We know from
Barrett et al. (2006) that about 30% of the Irish labor force have third level quali-
fications. Hence, the proportion of immigrants with third level qualifications, at
over 40%, points to a high-skilled inflow. With regard to immigrants from the
EU10, although they have the lowest proportion of highly educated across the
immigrant groups, they still compare favorably with the domestic labor force in
terms of skill levels.
Table 5
Educational distributions of immigrants by national group (2005)
UK EU13 EU10 Other USA All immigrants
% % % % % %
No formal/ primary education 2.4 1.1 6.4 4.7 0 4
Lower secondary 18.3 2.2 9.3 3.5 7.1 8.4
Upper secondary 19.6 22.4 37.8 25.2 17.9 26.8
Post Leaving 11.4 8.2 14.6 7.1 10.7 10.4
Third level 15.1 14.2 12.6 13.6 3.6 13.6
Third level - degree or above 33.3 51.9 19.2 45.9 60.7 36.8
100 100 100 100 100 100
N 378 268 452 508 28 1634

Source: Own calculations from the Quarterly National Household Survey.
Notes: “EU13” denotes the group of EU15 countries without the UK and Ireland, which are shown separately.
Although the figures on education levels presented in Table 5 point to a highly

skilled inflow, research on immigration in Ireland has identified a tendency for
immigrants to be employed in situations which do not fully reflect their skills. Of
course, this finding is typical of migration research in many countries and has
given rise to a number of theories, such as the lack of location-specific human
capital and discrimination. Such factors may be at play in Ireland, but before dis-
cussing them, we will look in greater detail at the research on labor market out-
comes.
The labor market experiences of immigrants in Ireland have been explored

along two dimensions: wages and occupational attainment. Looking firstly at the
work on wages, Barrett and McCarthy (2007a) used the Irish component of the
European Union’s Survey on Income and Living Conditions (EU-SILC) from
2004 to explore the earnings of immigrants in Ireland in general, and also those of
EU10 immigrants in particular. They found that immigrants earned 18% less than
natives on average, controlling for characteristics such as gender, experience and
education.
However, this aggregate figure of 18% hid large differences across immigrant
groups. For example, when broken down by immigrants from English-speaking
and non-English-speaking countries, they found that the immigrant/native wage
difference was essentially zero for the former group and 31% for the latter group.
They then went on to look within the non-English-speaking group and found a
wage disadvantage of 45% for EU10 immigrants relative to natives. This was the
largest disadvantage across the different national groups.
As the data used by Barrett and McCarthy in that paper were collected during
2004, it may have been the case that some of the EU10 immigrants in the sample
were interviewed prior to May 2004. Hence, they would have been working ille-
154 Alan Barrett
gally. Even those who arrived after May would have been very recent arrivals, and
so the high degree of wage disadvantage may not be surprising.
In order to get a second look at this issue, Barrett and McCarthy (2007b) re-
peated the analysis using EU-SILC data for 2005. Although the later estimate of
the EU10 immigrant earnings disadvantage was lower at 32%, this was still a
sizeable wage gap and bigger than those of other immigrant groups. Given that the
sample was drawn in 2005, the wage disadvantage would not have been related to
illegality. However, the possible lack of location-specific human capital is clear.
The EU-SILC data did not include information on the year of arrival, but we know
from the material above that most of the EU10 immigrant would have been recent
arrivals.
Another view of immigrant earnings, including those of the EU10 immigrants,

can be found in Barrett et al. (2008b). This paper contains significant advances on
the two Barrett and McCarthy papers, partly because of the data used and partly
because of the techniques employed. The data used by Barrett et al. (2008b) is
from a survey of 50,000 employees taken in 2005, of which 10% were immi-
grants. Whereas Barrett and McCarthy worked with samples of around 200 immi-
grants, Barrett et al. have a vastly bigger dataset set to work with. And whereas
Barrett and McCarthy relied on OLS regression, Barrett et al. use quantile regres-
sion to get a richer sense of the immigrant earnings disadvantage across the earn-
ings distribution.
Similar to Barrett and McCarthy, Barrett et al. (2008b) find that immigrants
from the EU10 have the highest earnings disadvantage across all immigrants
groups. The figure they report is of an EU10 gap of 18% relative to natives. This
is smaller than the estimates in the two Barrett and McCarthy papers. The more
interesting results in the paper arise in the context of the quantile regression. The
wage gap is essentially zero for the first earnings decile but rises steadily across
the deciles reaching 16% in the highest decile. Barrett et al. also run OLS wage
regressions within educational categories to see if the immigrant earnings disad-
vantage varies across educational level. They find that immigrants with low levels
of education (i.e. primary or secondary only) earn similar amounts relative to na-
tives with similarly low levels of education. They also find that the wage gap is
evident for those with post-secondary and tertiary education. They interpret these
results as providing evidence that the wage disadvantage for EU10 immigrants is
not about discrimination of low-skilled and vulnerable immigrants but about the
failure of more skilled immigrants to fully capture returns to human capital.
As noted above, the second approach to looking at labor market outcomes was
by looking at occupational attainment. Barrett and Duffy (2008) use data from Ire-
land’s official labor force survey (known as the Quarterly National Household
Survey) from 2005 to explore whether immigrants are employed in high-level oc-
cupations at rates comparable to native employees, controlling for characteristics

such as age, education and gender. They do this by running ordered probit regres-
sions in which the dependent variable is a four-way categorization of occupations,
with a ranking from high-level to low level.
The first regression presented by Barrett and Duffy show that immigrants (gen-
erally) are 2% less likely to be in the highest level occupations and 4% less likely
to be in the next category down, relative to natives. They are also 2% more likely
to be in the bottom category and 4% more likely to be in the category second from
bottom.
As was the case with the earnings results, the results for immigrants generally
hide important differences across immigrant groups. And again, as was the case
with the earnings results, the biggest difference between natives and immigrants
arose in the case of EU10 immigrants. They were 9% more likely to be in the
lowest occupational category relative to natives and also 9% more likely to be in
the second lowest category. The results also showed they were 5% less likely to
be in the highest category and 13% less likely to be in the second highest group.
We noted under the discussion of earnings that a part of (indeed much of) the
disadvantage experienced by EU10 immigrants relative to natives may be the re-
sult of the EU10 immigrants being recent arrivals and so lacking location-specific
human capital. If this were the case, the disadvantage would not be a concern, and
it would be expected to disappear the longer immigrants were in Ireland and ac-
cumulated the required location-specific human capital. The data used by Barrett
and Duffy had information on the year in which immigrants arrived in Ireland.
Hence, they were able to explore whether the occupational disadvantage disap-
peared over time.
The results from the EU10 immigrants and for most of the other immigrant
groups were disappointing. Little evidence was found in support of a decline in
the “occupational gap” between immigrants and natives, including those from the
EU10. As Barrett and Duffy write: this could be because the time period being ex-
amined was too short. It could also have been the case that the cell sizes being
used in the analysis were too small to generate statistically significant differences
in the estimates. But either way, no evidence of labor market integration was
found in the sense in which they defined it.
Before leaving this section on the characteristics of the immigrants from the
EU10, other sources of information should be mentioned. The Central Statistics
Office (2008) used the Census 2006 to provide profiles of the immigrant commu-
nities from ten countries, including Poland, Latvia and Lithuania. The profiles are
purely descriptive with no regression analyses. However, they still provide inter-
esting insights at the nationality level, something which is missing from the stud-
156 Alan Barrett
ies just referred to due to aggregation of nationalities in the micro-data provided to

researchers by the CSO. For example, over half of the Polish males were working
in either construction or manufacturing; while over half of the Polish females were
working in shops, hotels and restaurants. Although among Lithuanians, 23% have
a third level qualification, only 2% were working as professionals. This fits with
the results in Barrett et al. (2006) and Barrett and Duffy (2008).
Yet another interesting finding from this CSO report relates to the extent to
which married immigrants were not living with their partners. Among Poles, 59%
of married men were not living with their spouses when the census was taken;
among Latvian married men, the corresponding figure was 48%. Our interest in
this figure arises because it is a possible indicator of a mindset among these immi-
grants: they may view migration as being temporary. By way of contrast, the pro-
portion of married Nigerian immigrants who were not living with their spouses
was just 8%.
4 The Impacts of the EU10 Inflow
Research that has been conducted into the labor market impacts of immigration in
Ireland has not focused on the EU Accession States in particular but has instead
taken all immigrants as a block. We can, however, make some inferences about
the possible impacts by drawing on that work.
The impacts of immigration into Ireland have been considered in two papers,
Barrett et al (2002) and Barrett et al (2006). In the case of both papers, the ap-
proach taken was along the lines of Borjas, Freeman and Katz (1997) by using a
model of the Irish labor market to simulate the impact of migration and to trace
through the impact on variables such as GNP, GNP per worker, employment, un-
employment and wages. A key feature of the model is that it incorporates a crucial
determinant of output growth in a small open economy such as Ireland, namely,
competitiveness. In essence, output in Ireland is driven by its ability to attract in-
ternationally mobile capital and this in turn is achieved by having wages that are
lower relative to trading partners. Migration impacts in the model by constraining
wage growth through a labor supply increase. This wage dampening effect im-
pacts positively on competitiveness and thereby on labor demand.
One problem with using the model to simulate the impact of immigration is
that the classification of immigrants into the high-skilled and low-skilled catego-
ries is not clear-cut. In Barrett et al. (2002), immigrants who described themselves
as having particular levels of education were treated as being the same as natives
with the corresponding levels of education. However, Barrett et al. (2006) were
dealing with a later flow, for which this approach would tend to overstate the true
inflow of skills given that there was evidence of skilled immigrants working in
occupations below their skill levels. They approached the problem in one of their
simulations by using occupations as the indicator of skills as opposed to reported
education levels.
Table 6
Inferred impact of EU10 immigration of 180,000 labor force participants
% change
GNP per worker 1.7
GNP 5.9
Total employment 4.4
Average wage -7.8
Source: Own calculations.
In Table 6, we present what can be described as an “inferred” measure of the

impact of EU12 immigration. We use the term “inferred” because we have taken
results from Barrett et al (2006) and scaled them to fit with the higher level of
immigration. The figures in Table 6 show the inferred impacts of an inflow of
180,000 immigrants, with 180,000 being the number of EU12 immigrants in the
labor force as of 2008.
Barrett et al (2006) present simulation results under two scenarios. In one, ad-
justment within the labor market occurs through the rate of unemployment; in the
other adjustment is through wages. As the rate of unemployment was generally
constant around 2004 (at 4.5%), it seems reasonable to assume that most of the ad-
justment to immigration was through wages. For this reason, we only present here
results based on that scenario.
The first point to be taken from Table 6 relates to the impact on employment.
Although an inflow of 180,000 represents 8% of the labor force, the simulation
suggests that the impact on employment is actually much lower: employment only
increases by 4.4% because wages fall (on average) by 7.8% reducing the labor
supply. Given the net increase in employment and a positive impact on output per
worker, GNP increases by close to 6%.
Even with this concise presentation of results from the impact simulation, it is
possible to get a sense of the mechanism through which EU10/12 immigration
impacted positively upon the Irish economy. The Irish economy experienced rapid
growth around 2004, and so demand for labor was strong. In the absence of a
large inflow, wages would have risen. This would have choked off the increased
labor demand and so employment growth, and thus GNP growth would have been
constrained. In a sense, inward migration into Ireland helped to prevent further
losses in competitiveness as the deterioration in Ireland’s competitiveness became
a key concern in the middle part of this decade.
158 Alan Barrett
5 Conclusions
The ultimate objective of this chapter is to explore the labor market impacts of EU
Accession on the Irish labor market. In order to arrive at an assessment of such
impacts, it was necessary to move through a number of steps. These steps in-
volved determining the number who immigrated and their labor-related character-
istics; analyzing their labor market outcomes; and then simulating the impacts us-
ing a model of the Irish labor market.
EU Accession occurred at a time when the Irish economy was growing rapidly
and after a period of almost 10 years of sustained growth. It was against this back-
ground that the Irish Government chose to allow full access to the Irish labor mar-
ket for all citizens of the EU10. Given the rapidly growing economy and the lim-
ited numbers of countries who granted such free access as of May 2004, it was
perhaps not surprising that a large inflow arrived into Ireland from the EU10. The
number of EU10 nationals resident in Ireland grew from under 14,000 in 2002 to
over 120,000 in 2006. The continued inflows in 2007 and 2008 suggest that the
population of EU12 immigrants in Ireland in 2008 was around 200,000, or around
4.5% of the total population.
With regard to the inflow of EU10 nationals into the Irish labor market, the
proportionate inflow was much higher. According to the official labor force sur-
vey, 8% of the labor force was from the EU12 as of 2008. However, administra-
tive data collected through the issuing of social security numbers suggest a much
bigger inflow. Between 2002 and 2006, over 300,000 PPS numbers were issued to
EU10 nationals. The discrepancy between the two sources is likely to be the result
of very short-term stays in Ireland for many of the immigrants from the EU10.
While the immigrants from the EU10 appear to have levels of education com-
parable to the existing labor force, two pieces of research suggest that the EU10
immigrants are employed in occupations below what might be expected given
their education. The wage gap between immigrants and natives is largest for the
EU10 nationals: estimates of the gap range between 18 and 45%. It is also the case
that the EU10 immigrants experience the largest “occupational gap” relative to na-
tives, defined as occupational attainment given education levels. These results
combined suggest that the EU10 inflow may have been low-skilled in its impact.
Drawing on previous work that has estimated the impact of immigration into
Ireland in general, we have attempted to infer the impacts of the EU10 migration.
The inferred results suggest that the inflow allowed for employment to increase,
although by an amount substantially below the labor force increase implied by the
inflow. The modeling exercise suggests that wages would have fallen, thereby di-
luting the employment impact as some natives withdrew from participation. As
the rate of unemployment remained largely unchanged during the period under
question, we would tend to believe that the impacts of immigration were felt
through lower wages and hence, these withdrawals from the labor market. How-
ever, it should be stressed that the model used to establish these results was esti-
mated using time series data from 1966 to 2002. If it were the case that the inward
migration altered the workings of the Irish labor market, then this could have been
missed in the simulations.
The process within the model whereby inward migration dampened wage in-
creases is likely to have reflected reality quite well, even if the precise estimation
of these impacts is difficult. As Ireland is a small and open economy, it is gener-
ally understood that competitiveness is a critical factor in economic growth. From
around the beginning of this decade, Ireland had been losing competitiveness as
the prolonged period of economic growth resulted in wage increases above those
of Ireland’s trading partners. In the absence of large labor force inflows, such
wage increases would have been even greater, thus choking off labor demand.
Given the downturn in the Irish economy in 2008, the context has now changed
dramatically. From a growth rate of over 5% in 2007, Ireland’s economy is
thought to have contracted in 2008. This new situation gives rise to a set of ques-
tions including, how the recent inflow will react. To date, the answer to this is un-
known, but the prospect of outward migration on the part of the EU10 nationals
brings to mind aspects of Ireland’s migration experience of the 1980s and 1990s.
Looking firstly at the 1980s, and indeed earlier, Honohan (1984 and 1993)
showed how migration to and from Ireland tended to react to differentials between
the rates of unemployment in Ireland and the UK, whereby a stable gap between
the rates tended to hold. In essence, population movements out of Ireland at times
of high unemployment tended to keep the rate of unemployment lower than it
would otherwise have been. With unemployment on the increase in Ireland, fore-
casts for 2009 see a return to outward migration2 and hence a rate of unemploy-
ment below that which would hold in the absence of that outflow (Barrett et al,
2008a).
Turning to Ireland’s experience of migration in the 1990s, many of those who

left Ireland in the 1980s returned during the upturn of the 1990s. One interesting
by-product of that out-and-return migration experience was an apparent increase
in the human capital of those who left and returned relative to those who never left
(Barrett and O’Connell, 2001). Such an effect could now be part of the labor mar-
ket impacts on the EU10 as return migration occurs. However, the poor labor
market situations in which many of these immigrants found themselves in Ireland
and in the UK may mean that little human capital will have been accumulated
while working away and hence may lessen any such effect.
2 See Figure 2.
160 Alan Barrett
As a final note, migration from Ireland was often characterized as a “safety

valve” through which increases in unemployment were modified, with unem-
ployed people leaving. As noted above, with unemployment increasing it remains
to be seen if the “safety valve” will continue to operate although this time with
EU12 nationals leaving as opposed to Irish natives.
References
Barrett, A/A. Bergin/D. Duffy (2006), The Labour Market Characteristics and Labour Mar-
ket Impacts of Immigrants in Ireland, in: The Economic and Social Review 37 (1), 1-26.
Barrett, A./D. Duffy (2008), Are Ireland’s Immigrants Integrating into its Labour Market?,
in: International Migration Review 42 (3), 597-619.
Barrett, A./J. FitzGerald/B. Nolan (2002), Earning Inequality, Returns to Education and
Immigration into Ireland, in: Labour Economics 9 (5), 665-680.
Barrett, A./I. Kearney/J. Goggin/M. O’Brien (2008a), Quarterly Economic Commentary,
Autumn. Economic and Social Research Institute, Dublin.
Barrett, A./Y. McCarthy (2007a), Immigrants in a Booming Economy: Analysing their
Earnings and Welfare Dependence, in: Labour: Review of Labour Economics and In-
dustrial Relations 21 (4-5), 789-808.
Barrett, A./Y. McCarthy (2007b), The Earnings of Immigrants in Ireland: Results from the
2005 EU Survey of Income and Living Conditions, in: Quarterly Economic Commen-
tary, Winter, 42-62.
Barrett, A./S. McGuinness/M. O’Brien (2008b), The Immigrant Earnings Disadvantage
across the Earnings and Skills Distributions: The Case of Immigrants from the EU’s
New Member States in Ireland. IZA Discussion Paper No. 3479, Bonn.
Barrett, A./P. O’Connell (2001), Is There a Wage Premium for Returning Irish Migrants?,
in: Economic and Social Review 32 (1), 1-21.
Bell, B. D. (1997), The Performance of Immigrants in the United Kingdom: Evidence from
the GHS, in: The Economic Journal 107 (441), 333-344.
Borjas, G. J./R. B. Freeman/L. F. Katz (1997), How Much Do Immigration and Trade Af-
fect Labor Market Outcomes, in: Brooking Papers on Economic Activity 1, 1-90.
CSO Ireland (2008), Census 2006 – Non-Irish Nationals Living in Ireland. Dublin: The Sta-
tionery Office.
Honohan, P. (1984), The Evolution of Unemployment in Ireland 1962-83, in: Quarterly
Economic Commentary, May, 41-58.
Honohan, P. (1992), The Link between Irish and UK Unemployment, in: Quarterly Eco-
nomic Commentary, Spring, 33-44.
7. Post-Enlargement Migration and
Labor Market Impact in Sweden
Christer Gerdes
Eskil Wadensjö
Sweden was the only country that did not apply any transitional rules for migrants
coming from the ten new European Union member states in May 2004. The num-
ber of immigrants arriving in Sweden from these countries also increased, espe-
cially from Poland and the Baltic states, although not to the same extent as was
found in Ireland and the UK (two countries with transitory rules of only minor
importance). The composition of the migrants changed compared with that of mi-
grants who had arrived earlier from the same countries. Immigrants are repre-
sented in all sectors of the economy but overrepresented in some sectors, and
while earlier many more women than men arrived, now the gender composition is
much more even. The wages of the new immigrants from EU10 countries, control-
ling for education and age, are somewhat lower than for those born in Sweden.
This contrasts with the wages for new immigrants from EU14 countries being
higher than for those for those born in Sweden.
Migration, DOI 10.1007/978-3-642-02242-5_7,
164 Christer Gerdes and Eskil Wadensjö
1 Introduction1
Sweden was the only European Union country that did not introduce any transi-
tional rules for immigrants coming from the ten countries who became members
on 1 May 2004.2 Sweden maintained its policy of no transitional rules for immi-
grants when Bulgaria and Romania became members of the Union on 1 January
2007. This means that Sweden is an interesting test case for the likely effects of a
free labor market.
Immigration has economic effects on the country of origin. An increase of im-

migration due to an expansion of the European common labor market may have
various effects, both positive and negative. The magnitude of these effects, of
course, depends on the size of the increase in immigration and on the composition
of the new immigrants and the structure of the destination economy. We focus on
the dimension and composition of the actual migration to Sweden in the first years
after the 2004 enlargement of the European Union and analyze the labor market
situation of the immigrants from the new member states.
2 The Development of Immigration to Sweden

from the New EU Member Countries3
A prediction of the size of the new immigration is crucial to better understand the
economic effects of the enlargement of the European Union labor market.4 This
section presents the development of the number and composition of immigrants
1 This is a further development of two earlier studies; Doyle, Hughes and Wadensjö (2006)
and Gerdes and Wadensjö (2008). Note that throughout this chapter, “EU14” denotes the
group of EU15 countries without Sweden.
2 Ireland and the UK introduced some, but not very important, restrictions.
3 There are a few follow-up studies of immigration from the new member states after the
enlargement of the European Union. See Dølvik and Eldring (2005) for the migration to the
Nordic countries, and Commission of the European Communities (2006) comparing migra-
tion from the EU10 and the EU15 to all the EU15 countries. Note that the numbers based
on residence permits in the latter study are strongly misleading regarding the immigration
from EU15 countries to Sweden. Danish and Finnish citizens, who constitute the majority
of immigrants from other EU15 countries to Sweden, do not need a residence permit.
4 The predictions of the size and effects of migration made before the enlargement of the
European Union vary greatly. See for example Boeri, Hanson and MacCormick (2002),
Dustmann et al. (2003), Eriksson (2004), Pedersen, Pytlikova and Smith (2008) and Sinn
and Ochel (2003). The study most critical to free migration for citizens of the new member
states is by Sinn and Ochel (2003). They argue that migration without restriction would
lead to a dismantling of the welfare state. For a survey and discussion of the different pre-
dictions, see Zaiceva (2006).
7. Post-Enlargement Migration and Labor Market Impact in Sweden 165
from the new member states to Sweden before and after accession. However, it is
an incomplete picture for different reasons. One reason is that not all immigrants
in Sweden are registered and, more importantly, that the share registered may
have changed. It is likely that some immigrants who have lived and worked in
Sweden for a period but have not been registered before may register as a result of
the legal change. It is also important to study the emigration to the new member
states. The emigration to those states will mainly be return migration of earlier
immigrants. The return migration will increase, given the propensity to re-
emigrate, with the growth of the immigrant population living in Sweden. How-
ever, it may also increase if deregulation of immigration makes it more attractive
to return, resulting in the possibility of immigrating to Sweden again after a period
in the home country. Hence, the decision to go back to the home country is easy to
change if it is regretted.
We first look at the flows of immigrants and emigrants from the ten new
member states in the period 2001–2007 (see Table 1). We have three alternative
statistical series: the inflow and outflow of citizens of the ten new EU members;
the inflow and outflow of people born in the ten new EU members; and, finally,
people coming from and leaving to one of the new EU member states. We have
in Table 1 chosen the first series as the regulations refer to citizenship.
The total numbers show an increase in immigration during the first years of the
decade but a much higher increase in 2004, 2005 and especially 2006 and 2007.
That the increase continues in 2005 to 2007 is an indication that it is not only a re-
sult of higher registration, i.e. a registration effect, but also a real increase. The
immigration from the new member states, however, is still only a small part of the
total immigration to Sweden.
Emigration to the new member states is also increasing, but it is much smaller
than immigration. Net immigration is still a large part of gross immigration.
The majority of the immigrants from the new member states in all the years
of the period studied up to 2006 were women. The immigration of men in-
creased more than the immigration of women in each of the years from 2004 to
2007, so that more men than women came to Sweden from the ten new member
states in 2006 and 2007. This shift was a result of a large increase in male im-
migration from Poland and the immigrants were mainly of a young active age
(20–34 years old).
Table 1
Immigrants and emigrants according to citizenship (2001–2007)
2001 2002 2003 2004 2005 2006 2007
Country
Im Em Im Em Im Em Im Em Im Em Im Em Im Em
Czech R.
Men 33 9 54 17 41 18 54 34 54 28 88 23 119 34
Women 47 16 45 21 42 17 46 24 59 28 82 19 110 30
Cyprus
Men 2 3 2 7 6 10 5 2 7 7 24 3 47 7
Women 1 1 1 1 2 2 5 0 3 1 10 1 43 5
Estonia
Men 71 23 86 25 66 24 147 45 136 43 165 40 206 56
Women 206 28 204 43 211 38 256 53 247 64 257 68 265 73
Hungary
Men 66 30 96 38 61 44 89 40 115 36 243 50 465 70
Women 101 34 126 53 98 53 139 49 154 44 219 48 387 79
Latvia
Men 55 23 47 13 44 21 63 15 89 18 142 27 156 36
Women 99 11 114 13 108 19 143 22 143 18 210 27 221 63
Lithuania
Men 72 31 75 12 67 18 181 26 352 21 436 42 494 68
Women 141 14 184 3 163 25 257 26 343 14 459 45 412 56
Malta
Men 6 1 4 1 1 3 2 1 2 1 7 3 5 3
Women 3 3 2 1 2 0 3 0 4 2 4 4 3 0
Poland
Men 310 65 393 61 386 72 1120 94 1750 105 3416 159 4250 354
Women 499 75 672 85 630 100 1338 119 1670 136 2931 203 3290 282
Slovakia
Men 25 7 32 6 16 7 37 9 36 7 60 13 92 9
Women 23 9 36 4 31 7 68 11 61 13 90 13 70 23
Slovenia
Men 5 2 4 1 8 5 15 2 20 6 26 9 35 7
Women 16 5 8 4 8 3 19 3 16 6 24 8 38 7
EU10 1781 390 2185 409 1991 486 3987 575 5261 598 8893 805 10708 1262
Men 645 194 793 181 696 222 1713 268 2561 272 4607 369 5869 644
Women 1136 196 1392 228 1295 264 2274 307 2700 326 4286 436 4839 618
Source: Own calculations based on data from Statistics Sweden and Swedish Migration Board.
The immigrants to Sweden from the ten new member states were mainly from
Poland, which accounted for more than half of all immigrants. The relative impor-
tance of this immigration greatly increased in 2004 to 2007.5 Besides Poland, the
source of immigration is largest from the Baltic States, especially Lithuania, and
from Hungary. Furthermore, immigration of Bulgarian citizens increased from
123 in 2006 to 1159 in 2007 and Romanian citizens from 337 in 2006 to 2587 in
2007 (figures are not shown in Table 1).
5 Poland is the largest of the new member states and a neighbour to Sweden on the other
side of the Baltic Sea, and with a fast-growing and changing economy it is not only a coun-
try of origin for international migration but also a country of destination, especially for peo-
ple coming from some of the successor states to the Soviet Union. See Iglicka, 2005, and
Iglicka, Kazmierkiewicz and Weinar, 2005). A comparison with the migration statistics of
other countries for the same migration flows indicates that international migration is proba-
bly much under-estimated in Polish statistics.
Immigration has increased considerably from the new member states, espe-
cially from Poland, but it is still small compared with the total immigration and
size of the Swedish labor market. One explanation may be that there have been
few job vacancies for newly arrived immigrants. The Swedish unemployment rate
is low compared to several other European countries, but the job vacancies are
also few. Another explanation for the low immigration may be that the propensity
to emigrate from the new member states has been lower than expected by many
taking part in the public debate. An interview survey of migration intentions in the
Baltic states carried out a few years before the accession date shows that the will-
ingness to move abroad was not very high, and only a few stated the Nordic coun-
tries (including Sweden) as the preferred destination.6 A third explanation is that
those migrating chose Ireland and the UK instead of Sweden due to easier access
to the labor market and the fact that English is the language in those two countries.
We cannot exclude that immigration from other countries, registered and un-
registered, may have declined as a result of the accession agreement. Employers
who earlier employed immigrants from other countries may have turned to em-
ploying immigrants from the new member states.
3 The New Immigrants and the Labor Market
We now turn to the immigrants from the new member states and their position in
the Swedish labor market and analyze data containing information on all immi-
grants from the new member states who now live in Sweden. The latest year for
which data are available is 2006. The population included are those who are regis-
tered as living in Sweden at the end of the year (for those from the new member
states the criterion is to have a residence permit and be registered as living in
Sweden on 31 December). The employment information is for September or No-
vember, depending on the sector. This means that the data set does not include any
information on whether people are employed or not for those who arrived late in
2006; only that they have arrived and are living in Sweden at the end of the year.
We compare the employment rates for those who were born in one of the ten new
member states with someone born in Sweden (see Table 2). Anybody who was
born in one of the countries is included, irrespective of whether they arrived in
Sweden from May 2004 onwards or earlier.
6 See Brunovskis, Djuve and Haualand (2003).

Table 2
November employment rate of those born in one of the new member states
and in Sweden aged 16–64 living in Sweden at the end of 2006
Country Men Women All

Cyprus 47.6 50.5 48.8
Czech R. 54 46.9 49.8
Czechoslovakia 60.7 66.8 63.9
Estonia 50.9 53.8 52.8
Hungary 58.6 57 57.7
Latvia 43.3 48.6 46.9
Lithuania 59 51.4 54.5
Malta 62.7 70.4 66.1
Poland 61.9 58.4 59.7
Slovakia 44.6 40.7 42
Slovenia 66.7 58 62.2
EU10 60 57.6 58.6
EU10, May 2004– 57.6 33.6 45.4
EU14 59.9 62.8 29.9
EU14, May 2004– 31.5 27.6 29.9
Sweden 76.4 73.9 75.2
The employment rates for those from the ten new member states are considera-
bly lower than for people born in Sweden, but larger than for refugee immigrants
according to information from the labor force surveys.7 A closer look behind the
figures shows that many of the unemployed from the new member states do not
have any form of income, not even from income transfer programs (unemploy-
ment benefits, sickness benefits, pensions etc.). This category may hide several
different groups. The first group is not very large and consists of those who have
immigrated to Sweden in December. Hence, they did not work in September of
the same year. The second consists of immigrants who have (re)emigrated without
notifying the tax authorities, and the registration of emigration has been delayed
due to that. Some studies of this group show that quite a few immigrants have re-
turned without registering.8 Statistics Sweden constructed a variable for us indicat-
ing whether a person is likely to have left the country (there is no information on
7 The Commission of the European Communities (2006) gives information on employment

rates based on labour force surveys for 2004 and 2005. For citizens from the EU10, the
employment rate was 62 per cent in Sweden in 2005. This is a slightly higher value than
shown in Table 3. There are two explanations for this difference: one is that citizens of
EU10 countries and people born in the EU10 are two different populations; another is that
data collecting methods are different. In the labour force, only those participating in the
sur-vey are included. Those who have left the country without registering it are therefore
not in-cluded. One problem with labour force surveys is those employed may be overrep-
resented among those answering, leading to a selection problem.
8 See, for example, Edner and Johansson (2006). Statistics Sweden has also made esti-
mates of the share of different groups who have emigrated but are still registered as living
in Sweden. The shares vary greatly between different groups and are over 10% for some
groups.
any event which indicates that the person has been in Sweden during the year in
any of a number of registers at Statistics Sweden). The variable probably underes-
timates the number of those who have left Sweden. According to this study, 2.5%
of people who were born in an EU10 country are no longer in Sweden. The corre-
sponding share of those who were born in an EU14 country is 3.7% and for those
born in Sweden 0.2%.
Figure 1 shows that the share is especially high for those who have arrived in
recent years, both for those who arrived from EU10 and EU14 countries. Thus, the
immigrant population is overestimated and the employment rate is underesti-
mated, especially among recently arrived immigrants.
Figure 1
Share of persons registered as living in Sweden but who most likely do not do it
according to year of stay in Sweden (2005)
0.2
0.18
0.16
0.14
Not in Sweden
0.12
0.1
0.08
0.06
0.04
0.02
0
0 20 40 60
Number of years in Sweden
EU10 countries EU14 countries
A third group consists of students (in secondary or higher education) who do

not combine studies with work. A fourth group consists of immigrants who are
registered as immigrants in Sweden but only work for part of the year, for exam-
ple seasonal work in the summer. A fifth group contains people who do not work
and are supported by other family members, for example housewives. A sixth
group consists of people who work in the unregistered part of the economy (the
shadow economy). We do not have any estimates of the size of this group.
Even if the employment rate estimations are to be interpreted with care, infor-
mation on working hours and wages for those employed does not suffer from such
limitations. Information on working hours in 2006 are shown in Table 3. If we in-

clude everyone, irrespective of when they arrived, we find only small differences
between those born in Sweden and those born in the new member states or in the
old member states. If we study only those who arrived in May 2004 or later, the
working hours are longer for men who were born in the EU10 or the EU14. The
variation in working hours among the ten countries and between the two years
may be explained by the small number of observations.
Table 3
Working hours among those born in one of the new member states
and in Sweden aged 16–64 in September or November 2006
Cyprus 136.7 125.7 131.3
Czech R. 142.5 106 119.7
Estonia 135.7 116.7 121.6
Hungary 132.7 120.1 125.1
Latvia 134.7 115.4 119.4
Lithuania 141.7 110.7 119.1
Malta 158.3 114.6 138
Poland 137.3 119.7 124.5
Slovakia 130.2 104.3 112.6
Slovenia 143.2 107.9 123.2
EU10 136.3 118.9 124.1
EU10, May 2004– 146.3 117.2 129.2
EU14 138 121.9 128.5
EU14, May 2004– 142.4 119.8 132
Sweden 138.7 117.9 126.8
Table 4 shows the average monthly wage (recalculated to full-time monthly

wage for those not working full-time) for people who were born in the EU10
member states, the EU14 member states and Sweden. The average monthly wage
is slightly higher for men and women born in the other member states than that for
those born in Sweden, both for those from the EU10 and the EU14. There are also
differences between those from different EU10 countries, and it should be stressed
that some groups contain only a small number of individuals. There are differ-
ences in the composition according to age and education which may contribute to
explaining the differences. The table shows that new arrivals born in EU14 coun-
tries have higher labor incomes than those born in Sweden.
Table 4
Monthly wage (for those working less than full-time the wage is recalculated to
full-time wage) among those born in one of the new member states and in
Sweden aged 16–64 in September or November 2006; thousands SEK
Cyprus 30.1 21.4 25.8
Czech R. 27.2 24.6 25.6
Estonia 29.8 22.1 24.1
Hungary 28.5 23.5 25.5
Latvia 31.3 22.1 24
Lithuania 30.1 22.2 24.4
Malta 25.5 19.2 22.6
Poland 28.8 22.5 24.3
Slovakia 32.6 25 27.4
Slovenia 25.3 20.7 22.7
EU10 29.3 22.9 24.8
EU10, May 2004– 28.3 23.8 25.6
EU14 28.7 22.9 25.3
EU14, May 2004– 35.2 25.4 30.8
Sweden 28.3 22.2 24.8
Table 5 shows that the educational level is higher on average for those coming
from the new member states than those born in Sweden. There are large variations
between the different states. Note also that we lack information for a much larger
share of the immigrants than for those born in Sweden. It is mainly for the newly
arrived immigrants that information on education is missing. This leads to diffi-
culties in estimating the effects of education on the labor outcome for new immi-
grants in Sweden. Figure 2 plots the amount of information against the number of
years since arrival. The graph shows that we have least information on those who
have recently arrived.
Table 5
Distribution of people born in one of the new member states and in Sweden
according to education, in % (2006)
Education
Country
1 2 3 4 5 6 9 all
Cyprus 9 14 41 6 18 2 9 100
Czech R. 2 3 27 6 36 5 22 100
Czechoslovakia 5 8 41 6 34 4 2 100
Estonia 2 8 28 6 37 3 15 100
Hungary 5 7 46 6 28 2 5 100
Latvia 1 8 20 6 38 4 23 100
Lithuania 1 5 16 5 35 5 33 100
Malta 7 16 36 5 26 1 10 100
Poland 3 7 40 5 29 2 13 100
Slovakia 5 9 20 4 28 5 29 100
Slovenia 10 11 49 4 18 1 7 100
EU10 4 8 39 6 30 2 12 100
EU10, May 2004– 1 2 23 3 27 1 43 100
EU14 8 11 43 5 29 2 3 100
EU14, May 2004– 2 3 14 3 43 5 30 100
Sweden 4 16 47 6 25 1 1 100
Figure 2
Share of persons without information on education
according to years of immigration (2005)
0.6
0.5
0.4
Unknown education
0.3
0.2
0.1
0
0 20 40 60
Number of years in Sw eden
EU10 countries EU14 countries
When estimating probit equations including besides years of stay in Sweden (in
linear and quadratic form) also age, gender and country of origin, some significant
results are obtained. Young people report their education less often than older
people; and women more often than men. The differences between countries of
origin are small and in most cases not significant. We compared the newly arrived
with and without information in different respects and found the differences to be
quite small: the salary of immigrants from the EU10 is 21,638 SEK with informa-
tion on education and 21,467 SEK without such information; the figures for the
EU14 are 34,142 and 34,313, respectively.
Table 6 compares the industry distribution for those who were born in the EU
to those who were born in Sweden. The distribution is very much the same for
both. The main difference is that people born in EU10 countries are overrepre-
sented in the health care sector and if from the EU14 in manufacturing. If we
study only those who arrived in May 2004 or later, the differences are larger. Im-
migrants from the EU10 are overrepresented the greatest in agriculture and con-
struction and underrepresented the most in education; those born in the EU14 are
overrepresented, especially in financial and business services.
Table 6
Distribution of people born in one of the new member states and in Sweden
according to industry, in % (2006)
Industry
Country
0 1 2 3 4 5 6 7 8 9 10 All
Cyprus 2 0 13 0 3 18 10 14 17 19 4 100
Czech R. 1 2 17 0 1 15 12 12 20 17 3 100
Czechoslovakia 2 0 17 0 2 14 14 13 24 8 5 100
Estonia 2 1 10 0 6 18 15 12 20 9 5 100
Hungary 2 0 17 1 3 17 15 13 20 8 5 100
Latvia 3 4 11 1 3 13 15 13 22 11 5 100
Lithuania 2 13 13 0 8 12 15 11 16 7 3 100
Malta 1 0 18 0 1 15 16 16 11 21 0 100
Poland 1 2 15 0 7 16 15 10 22 8 4 100
Slovakia 2 0 17 0 2 13 16 11 29 11 0 100
Slovenia 1 0 28 0 3 18 17 7 16 6 4 100
EU10 2 2 15 0 6 16 15 11 22 8 4 100
EU10, May 2004– 2 7 14 0 19 15 19 3 11 9 1 100
EU14 1 1 20 1 5 15 14 12 19 8 4 100
EU14, May 2004– 2 1 14 0 6 17 21 12 14 11 2 100
Sweden 1 2 16 1 6 19 14 11 16 7 6 100
We estimated Mincer equations with the logarithm of the monthly full-time

wage as the dependent variable. The independent variables were age, age squared,
female, educational levels, and either dummy variables for being an immigrant
who was born in an EU10 country or a dummy variable for coming from any of
them (see Table 7). We did not have complete observations for many recently ar-
rived individuals, due to the lack of information on education. The result for the
coefficient of the EU10 dummy is a negative value indicating a negative wage
premium of about 3% for women and 4% for men.9 When dummies for the differ-
ent countries are included in the estimations, the coefficients vary: they are nega-
tive in most but not all cases (people born in Sweden are the reference group). The
coefficients are negative for some groups with many recent arrivals – Poland,
Hungary and the Baltic states, and the largest negative wage effect is found for
immigrants from Lithuania.
9 We used the logarithm of wage as the outcome variable, so the effect of a change in one
of the explanatory variables is estimated by the antilog given by exp(estimated coeffi-
cient)-1.
Table 7
Wage equation estimates with log monthly wage for full-time work in 2006
as the dependent variable
Men Women All Men Women All
8.995 9.269 9.247 8.990 9.267 9.247
Constant
(0.0042) (0.0028) (0.0028) (0.0046) (0.0028) (0.0025)
-0.185 -0.185
Female
(00005) (0.0005)
0.0392 0.0227 0.0299 0.0392 0.0227 0.0300
Age
(0.0002) (0.0001) (0.0001) (0.0002) (0.0001) (0.0001)
-0.00037 -0.00021 -0.00028 -0.00037 -0.00021 -0.00028
(0.000003) (0.000002) (0.000001) (0.000003) (0.000002) (0.000001)
Age squared
Primary school 0.073 0.062 0.067 0.073 0.062 0.067

9 or 10 years (0.0018) (0.0014) (0.0011) (0.0018) (0.0014) (0.0011)
Secondary 0.152 0.102 0.127 0.152 0.103 0.117
school (0.0017) (0.0012) (0.0010) (0.0016) (0.0012) (0.0010)
Higher 0.337 0.247 0.299 0.337 0.247 0.299
education less (0.0021) (0.0018) (0.0014) (0.0021) (0.0018) (0.0014)
than two years
Higher 0.391 0.308 0.346 0.391 0.308 0.346
education two (0.0017) (0.0012) (0.0011) (0.0017) (0.0012) (0.0011)
years or more
Post-graduate 0.632 0.628 0.626 0.532 0.628 0.626
education (0.0030) (0.0035) (0.0023) (0.0030) (0.0035) (0.0023)
-0.125 -0.020 -0.058
Czech Republic
(0.041) (0.028) (0.024)
0.020 0.043 0.032
Czechoslovakia
(0.014) (0.009) (0.008)
-0.038 -0.052 -0.040
Cyprus
(0.035) (0.027) (0.022)
-0.034 -0.069 -0.059
Estonia
(0.019) (0.009) (0.008)
-0.075 -0.017 -0.041
Hungary
(0.011) (0.006) (0.006)
-0.001 -0.070 -0.056
Latvia
(0.032) (0.014) (0.013)
-0.065 -0.117 -0.102
Lithuania
(0.030) (0.013) (0.012)
-0.066 -0.110 -0.084
Malta
(0.075) (0.027) (0.041)
-0.047 -0.042 -0.046
Poland
(0.005) (0.003) (0.002)
0.080 0.008 0.033
Slovakia
(0.053) (0.032) (0.028)
-0.037 -0.057 -0.047
Slovenia
(0.023) (0.016) (0.013)
-0.044 -0.036 -0.040
EU10
(0.004) (0.002) (0.002)
R squared 0.25 0.266 0.305 0.25 0.266 0.305
Number of 675238 876168 1551406 675238 876168 1551406
observations
We re-estimated the equations with dummies for the period of arrival and for
women, and we find, as expected, that the earlier the EU10 immigrants arrived,
the smaller the wage disadvantage (see Table 8). Surprisingly, men born in the
EU10 who arrived in 2000–06 have a positive wage premium compared with
those born in Sweden. For someone who arrived in the period 2000–06 the differ-
ence is minus 4% for women and plus 3% for men. For someone who arrived be-
fore 1970, there is no difference compared with those born in Sweden for men and
a positive difference for women (4%).
Table 8
as the dependent variable for natives and immigrants from EU10 countries
Men Women All
8.994 9.268
Constant
(0.0042) (0.0028) 9.246 (0.025)
-0.185
Female
(0.0005)
0.0392 0.0227 0.0300
Age
(0.0002) (0.0001) (0.0001)
Age squared
-0.00037 -0.00021 -0.00028
(0.000003) (0.000002) (0.000001)
Primary school 9 or 10 0.073 0.062 0.067
years (0.0018) (0.0014) (0.0011)
0.152 0.102 0.127
Secondary School
(0.0016) (0.0012) (0.0010)
Higher education less 0.337 0.246 0.299
than two years (0.0021) (0.0018) (0.0014)
Higher education two 0.391 0.307 0.346
years or more (0.0017) (0.0012) (0.0011)
Post graduate 0.632 0.627 0.625
education (0.0030) (0.0034) (0.0023)
Arrived before 1970
born in EU10 0.011 (0.012) 0.042 (0.009) 0.028 (0.007)
Arrived 1970–74 born -0.067 -0.017
in EU10 (0.012) 0.012 (0.007) (0.006)
Arrived 1975–79 born -0.071 -0.021 -0.039
in EU10 (0.013) (0.006) (0.005)
Arrived 1980–84 born -0.081 -0.017 -0.041
in EU10 (0.010) (0.005) (0.005)
Arrived 1985–89 born -0.077 -0.054 -0.066
in EU10 (0.009) (0.005) (0.004)
Arrived 1990–94 born -0.065 -0.081 -0.081
in EU10 (0.015) (0.005) (0.005)
Arrived 1995–99 born -0.057 -0.085 -0.080
in EU10 (0.020) (0.006) (0.007)
Arrived 2000–06 born -0.043 -0.015
in EU10 0.030 (0.012) (0.007) (0.006)
R squared 0.25 0.266 0.305
Number of
observations 675238 876168 1551406
The following table compares migrants born in the EU10 who arrived in May
2004 or later to immigrants born in the EU14 who arrived in the same period and
with those born in Sweden.
Table 9
as the dependent variable for natives and immigrants from EU14 and EU10 countries
who have arrived in 2004-2006
Men Women All All
8.999 9.276 9.252 8.912
Constant
(0.0041) (0.0027) (0.024) (0.014)
-0.185 -0.184
Female
(0.0004) (0.0004)
0.0393 0.02255 0.0300 0.0300
Age
(0.0002) (0.0001) (0.0001) (0.0001)
Age squared
-0.00037 -0.00021 -0.00028 -0.00028
(0.000003) (0.000002) (0.000001) (0.000001)
Primary school 9 or 0.066 0.057 0.061 0.061
10 years (0.0016) (0.0012) (0.0012) (0.0011)
0.144 0.086 0.120 0.120
Secondary school
(0.0015) (0.0012) (0.0010) (0.0009)
Higher education
0.329 0.240 0.291 0.292
less than two years
(0.0019) (0.0017) (0.0013) (0.0013)
Higher education 0.387 0.303 0.342 0.341
two years or more (0.0016) (0.0011) (0.0009) (0.0009)
Post-graduate 0.623 0.619 0.617 0.617
education (0.0028) (0.0032) (0.0021) (0.0021)
Arrived in 2004 -0.008 -0.069 -0.044 0.000
born in EU10 (0.030) (0.016) (0.015) (0.015)
Arrived in 2005 -0.037 -0.069 -0.053 -0.004
born in EU10 (0.028) (0.018) (0.016) (0.016)
Arrived in 2006 -0.038 0.024 0.001 0.028
born in EU10 (0.026) (0.024) (0.017) (0.018)
Arrived in 2004 0.032 -0.006 0.016 0.019
born in EU14 (0.017) (0.014) (0.012) (0.011)
Arrived in 2005 0.044 -0.000 0.024 0.028
born in EU14 (0.019) (0.011) (0.011) (0.011)
Arrived in 2006 0.097 0.017 0.065
born in EU14 (0.017) (0.011) 0.062 (0.11) (0.011)
ln (GDP per capita
in EURO in home 0.033
country) (0.004)
R squared 0.248 0.266 0.303 0.303
Number of
observations 718957 939079 1656036 1656138
The estimations in Table 9 show a negative sign on wages for immigrant

women who were born in the EU10 and who arrived in 2004 or 2005. The differ-
ence may be due to the immigrants not working in the occupations they are trained
for, so that, at least in the short run, they are over-educated. We find no significant
wage differences compared to being born in Sweden for women who immigrated
in 2006 and for men independent of whether they immigrated in 2004, 2005 or
2006. More surprising is that the male immigrants who were born in one of the
EU14 countries have higher wages than those born in Sweden, controlling for age
and education. One explanation may be that their type of education leads to occu-
pations with high wages, for example in medicine, technology or business admini-
stration.
An explanation to the pattern described here may be that reservation wages for
the prospective migrants differ depending on the wage level in the country of ori-
gin. Those coming from countries with low wages accept lower wages than those
from countries with high wages. We included GDP per capita in euros as a rough
measure of the wage level in the different measures. The variable is highly signifi-
cant and has the expected positive sign. The variables which indicate that the indi-
viduals have emigrated from an EU10 country are no longer significantly different
from zero, but the EU14 variables are still positive and significant.
4 Conclusions
Four years have passed since a further ten countries became members of the
European Union, which is a very short period to follow up on what has happened.
Some of our data covers periods of only thirty-six to forty-four months; other data
covers an even narrower time period. It takes time to react to a new legal frame-
work, and we cannot expect the full effect of the changes to have taken place yet.
However, we believe that the information on this short period may contribute to a
better understanding of the likely effects of deregulation of migration from the
new member states.
• Migration increased from the new member states after the enlargement of the
European Union. It was more than five times larger in 2007 than in 2003. The
immigration of men has increased more than that of women. The immigrants
are mainly Polish citizens. Among the other countries, Hungary and the three
Baltic states are the most important countries of origin.
• According to our data, the immigrants from the new member states have an
employment rate much lower than people born in Sweden. Part of this is ex-
plained by lags in the registration of returning migrants. A number of other
factors may explain other parts of the difference,10 but further studies are
needed.
10 Those not registered as employed are only to low extent on income transfer programs.
See Gerdes and Wadensjö (2008).
• Among employed immigrants from the new member states the working hours
are about the same as for those born in Sweden; however, for men from EU
countries who arrived in May 2004 and later, the working hours are longer
than for those born in Sweden.
• The monthly wage income in 2006 (recalculated to monthly wage for full
time work for those working part-time) is slightly higher for the recently ar-
rived who were born in EU10 countries than for those born in Sweden. The
new arrivals from EU14 countries have considerably higher monthly wage
incomes than those born in Sweden.
• Information on education is lacking for many of the new immigrants. How-
ever, for those on whom we do have information, the educational level is
higher than for those born in Sweden. The difference is explained by most of
the new immigrants being young and belonging to cohorts with a longer edu-
cation period on average.
• The new EU10 immigrants are overrepresented in agriculture and construc-
tion; and underrepresented in education (few are teachers) and public admini-
stration. The new EU14 immigrants, on the other hand, are overrepresented
in financial and business services.
• The wage rate is slightly lower for new immigrants from EU10 countries than
for those born in Sweden, controlling for education and age. This is in con-
trast to new immigrants from EU14 countries who earn higher wages than
those born in Sweden, controlling for age and education.
• It is important to improve the official statistics in some respects and collect
information on education for a greater part of the immigrants. It is also very
important to study the activities of those who are not employed in the month
when employment is measured. Are those who are not employed re-
emigrants, seasonal workers not in Sweden in the month in which the em-
ployment is measured, in the grey economy, students or housewives?
References
Boeri, T./G. Hanson/B. MacCormick (eds.) (2002), Immigration Policy and the Welfare
System.Oxford.
Brunovskis, A./A. B. Djuve/H. M. Haualand (2003), Facing a Baltic Invasion? Mobility of
Baltic Labour towards the Nordic Countries. Fafo-Report 432, Oslo.
Dølvik, J. E./L. Eldring (2005), Arbeids- og tjenstemobilitet etter EU-utvidelsen. Nordiske
forskjeller og fellestrekk, TemaNord 2005: 566. Nordic Council of Ministers, Copenha-
gen.
Doyle, N./G. Hughes/E. Wadensjö (2006), Freedom of Movement for Workers from Cen-
tral and Eastern Europe. Experiences in Ireland and Sweden. Swedish Institute for
European Policy Studies (SIEPS), Report No. 5, Stockholm.
Dustmann, C./M. Casanova/M. Fertig/I. Preston/C. M. Schmidt (2003), The Impact of EU
Enlargement on Migration Flows. Home Office Online Report 25/03.
Edner, B./N. Johansson (2006), Kartläggning av “folkbokföringsfelet”, Skatteverket (the
Swedish National Tax Board). Stockholm.
Eriksson, J. (2004), Nettomigrationen från de nya EU-länderna – utvärdering och prog-
nostisering, Swedish Institute for European Policy Studies (SIEPS), Report No. 2u,
Stockholm.
European Commission (2006), Report on the Functioning of the Transitional Arrangements
Set Out in the 2003 Accession Treaty (Period 1 May 2004–30 April 2006). Communica-
tion from the Commission to the Council, the European Parliament, the European Eco-
nomic and Social Committee and the Committee of the Regions. Brussels 8 February
2006.
Gerdes, C./E. Wadensjö (2008), Immigrants from the New EU Member States and the
Swedish Welfare State, Swedish Institute for European Policy Studies (SIEPS), Report
No. 9, Stockholm.
Iglicka, K. (2005), The Impact of the EU Enlargement on Migratory Movement in Poland.
Center for International Relations, Warsaw, Report No. 12/05.
Iglicka, K./P. Kaźmierkiewicz/A. Weinar (2005), Poland in: J. Niessen/Y. Schibel/C.
Thompson (eds.), Current Immigration Debates in Europe: A Publication of the Euro-
pean Migration Dialogue. Migration Policy Group (MPG), Brussels.
Pedersen, P./M. Pytlikova/N. Smith (2008), Selection or network effects? Migration flows
into OECD countries 1990-2000, in: European Economic Review 52 (7), 1160-1186.
Sinn, H.-W./W. Ochel (2003), Social Union, Convergence and Migration, in: Journal of
Common Market Studies 41 (5), 869–896.
Zaiceva, A. (2006), Reconciling the Estimates of Potential Migration into the Enlarged
European Union. IZA Discussion Paper No. 2519, Bonn
8. The Impact of the Recent Expansion
of the EU on the UK Labor Market
David G. Blanchflower
Helen Lawton
We examine the impact of the influx of workers from Eastern Europe on the UK
and look at the characteristics of the workers who have come to the UK since
2004. We also use data from a number of Eurobarometers (2004-2007) as well as
the 2005 Work Orientation module International Social Survey Programme to
look at the attitudes of residents of these countries. Eastern Europeans report that
they are unhappy with their lives and the country they live in, they are dissatisfied
with their jobs, and they would experience difficulties finding a good new job or
keeping their existing job. Relatively high proportions express a desire to move
abroad. Eastern Europeans’ expectations for the future of their domestic econo-
mies and their personal situations remain low, but have improved since 2004.
There has been some deterioration in the availability of jobs in the UK as the
economy moves into recession. However, the UK is an attractive place for Eastern
Europeans to live and work. We argue that rather than dissipate, flows of Eastern
European workers to the UK could remain strong well into the future.
Migration, DOI 10.1007/978-3-642-02242-5_8,
182 David G. Blanchflower and Helen Lawton
1 Introduction
The European Union began its process of enlargement in spring 2004, and since
then its membership has grown from fifteen to twenty-five countries. This process
has greatly benefited the UK: nearly 900,000 workers from the accession coun-
tries have registered to work in the UK since 2004. In contrast to other EU coun-
tries, the UK government allowed full access to its labor market to the nationals of
eight former Soviet-bloc countries (the Czech Republic, Estonia, Hungary, Latvia,
Lithuania, Poland, Slovakia and Slovenia – the EU8 countries).1 However, the UK
did restrict access to state benefits. On 1st January 2007, Bulgaria and Romania
(the EU2 henceforth) joined the European Union; workers from these two coun-
tries were given much less open access to the UK labor market than those from
the EU8.2 We refer to the group of ten Eastern European countries as the EU10.
EU10 nationals are allowed access to the UK labor market under the Worker
Registration Scheme (WRS). Those who wish to work as employees in the UK for
a period of at least a month are required to register with the WRS. Workers who
are self-employed do not need to register. Applicants must register more than once
if they are employed by more than one employer. They must also re-register if
they change employer. An individual who has registered to work and who leaves
employment is not required to deregister, so some of those counted will have left
the employment for which they registered. Indeed, some are likely to have left the
UK. Employed and (in contrast to the WRS) self-employed workers from EU10
countries are legally required to register for a National Insurance number (NINO)
in order to work in the UK.3
A number of questions arise regarding the arrival of EU10 workers in the UK:
1) How many have arrived and how many have returned?

2) What are their characteristics?
3) Why did they come?
4) What impact have they had on the UK labor market?
5) Will the flow continue in the future?
1 Residents of Cyprus and Malta were also permitted to work in the UK from 2004, but the
size of the flows are small, and hence, we concentrate on the more important flows from
the Eastern European ten.
2 For details of the work rules for nationals of Bulgaria and Romania see the UK Border
Agency (2008), Bulgarian and Romanian Accession Statistics, #s1-5.
www.bia.homeoffice.gov.uk/sitecontent/documents/aboutus/reports/bulgarianromanian/
3 The National Insurance numbers reported here relate to the numbers allocated for em-
ployment purposes. A further 6,597 were allocated for benefit purposes and 13,998 for tax
credit purposes.
8. The Impact of the Recent Expansion of the EU on the UK Labor Market 183
In the following, we examine these questions using data from a number of

sources. In the UK, we make use of data from the WRS, NINOs, the Labour Force
Survey (LFS), a number of Eurobarometers (2004-2007), and the 2005 Work Orien-
tation module International Social Survey Programme. The latter looks at the atti-
tudes of EU10 residents to examine factors likely to impact on their propensity to
move to the UK.
In summary, EU10 residents report that they are unhappy with their lives and
the country they live in, they are dissatisfied with their jobs, and they would ex-
perience difficulties finding a good new job or keeping their existing job. Rela-
tively high proportions express a desire to move abroad. Eastern Europeans’ ex-
pectations for the future of their domestic economies and personal situations
remain low, but have improved since 2004. There has been some deterioration in
the availability of jobs in the UK as the economy moves into recession. A number
of commentators have taken this to imply that the flow of EU10 workers to the
UK will slow. We examine the macroeconomic factors that are likely to impact
the potential flow of workers in the future.
2 How Many EU10 Workers Have Arrived

and How Many Have Returned?
Table 1 provides the most recently available data from the Worker Registration
Scheme (WRS) based on the Accession Monitoring Report May 2004–June 2008. A
cumulative total of 850,000 applicants were approved on the WRS between 1st May
2004 and 30th June 2008 (column 1). In addition, evidence from the Bulgarian and
Romanian Accession Statistics April to June 2008 suggests there were 19,525
worker registrations from Bulgaria and 31,080 from Romania under various
schemes available. However, this is possibly an over-estimate: the numbers may in-
clude double-counting because it is possible to reapply and be issued a further cer-
tificate.4
4 Between January 2007 and June 2008, 2,290 Bulgarians and 2,815 Romanians were ap-
proved to receive accession worker cards. In addition, 5,640 Bulgarians and 2,420 Romani-
ans in 2007 received approval to work under the Seasonal Agricultural Workers Scheme.
Other Bulgarian and Romanian nationals exercising a Treaty right in the United Kingdom
may seek a registration certificate. This includes those exempt from the worker authorisation
restrictions; highly skilled migrants; those with restricted access to the labour market (stu-
dents, the self-employed and self-sufficient people); and Bulgarian and Romanian family
members of these main applicants. Registration cerificates have been approved for 15,605
Bulgarians and 25,845 Romanians since 007. The largest proportion of these applications
(43% in the second quarter of 2008) was for registration certificates confirming that the appli-
cant is exercising a Treaty right as a self-employed person. The next largest group of appli-
Table 1
Worker Registration Scheme approvals, re-registrations and
National Insurance number applications (May 2004 – June 2008)
Worker
Multiple re- National Insurance
Registration Re-registrations
registrations number
Scheme
2004 Total 125,880 10,130 3,855 62,539
2005 Total 204,970 51,360 13,945 218,521
2006 Total 227,875 81,845 23,105 260,909
2007 Total 210,575 115,640 19,870 246,923
2007 Q1 50,315 28,370 6,090 68,782
Q2 52,340 30,025 5,430 55,592
Q3 57,270 29,295 4,475 66,281
Q4 50,650 27,955 3,880 55,638
2008 Q1 42,790 32,700 4,205 52,858
2008 Q2 37,955 31,140 3,760 n/a
Total 850,045 407,320 64,980 841,120
Source: Accession Monitoring Report, May 2004 – June 2008, Home Office UK Border Agency.
www.ukba.homeoffice.gov.uk/sitecontent/documents/aboutus/reports/accession_monitoring_report/
Notes: National Insurance numbers reported here relate to the numbers allocated for employment purposes. A further 6,597
were allocated for benefit purposes and 13,998 for tax credit purposes.
Overall, there have been around 900,000 WRS approvals from the EU10 coun-
tries. In addition, there have been a large number of re-registrations and multiple
re-registrations (columns 2 and 3 of Table 1). Re-registrations occur when those
who have previously registered change employer. Multiple re-registrations are
made by individuals working for more than one employer simultaneously. Re-
registrations are widespread and account for approximately half of total registra-
tions, suggesting there is considerable movement between employers (a point we
return to below). Interestingly, despite a slowing in the number of new registra-
tions in 2008, the number of re-registrations in the second quarter of 2008 is still
higher than for any quarter in 2007.
Data on the number of individuals who registered to obtain National Insurance

numbers are also available. Between May 2004 and March 2008, 841,200 Na-
tional Insurance numbers were allocated to EU8 applicants. In addition, there
were 37,968 registrations for National Insurance numbers for employment pur-
poses from Bulgaria and Romania. Between January 2007 and March 2008, ap-
proximately two thirds of these applications were from Romanians.5 In total, be-
cants for registration certificates are those applying on the basis that they are family mem-
bers of the main applicant (23%), and those applying on the basis of study (17%).
5 For example, in the latest data available (January – March 2008), the total number of Na-
tional Insurance number applications from Bulgarian and Romanian nationals was 11,102.
98.0% of these were allocated for employment purposes, 1.3% for benefit purposes, and
0.6% for tax credit purposes. Between April and June 2008, 55% of applicants were male
and 73% of applicants were aged 18-34 years. Source: Bulgarian and Romanian Acces-
sion Statistics, April –June 2008, Home Office UK Border Agency.
www.bia.homeoffice.gov.uk/sitecontent/documents/aboutus/reports/bulgarianromanian/
tween mid-2004 and March 2008, approximately 880,000 individuals from the
EU10 countries received a NINO for employment purposes.
Table 2 shows that two-thirds of those who have registered on the WRS came
from Poland, followed by Slovakia (10%) and Lithuania (9%). Fewer than 1,000
who registered came from Slovenia. The country mix of registrations in the sec-
ond quarter of 2008 was broadly similar to that of 2004.
Table 2
Nationality of approved WRS applicants by quarter and year of application
(May 2004 – June 2008)
Czech
Estonia Hungary Latvia Lithuania Poland Slovakia Slovenia Total
Rep.
2004 Total 8,255 1,860 3,620 8,670 19,270 71,025 13,020 160 125,880
2005 Total 10,575 2,560 6,355 12,960 22,990 127,325 22,035 175 204,970
2006 Total 8,345 1,475 7,060 9,490 17,065 162,500 21,755 185 227,875
2007 Q1 1,825 275 1,965 1,835 3,740 35,800 4,835 45 50,315
Q2 1,800 210 2,085 1,630 3,690 37,280 5,600 40 52,340
Q3 1,985 275 2,305 1,545 3,715 41,170 6,230 50 57,270
Q4 1,895 210 2,515 1,270 3,090 35,865 5,760 55 50,650
2007 Total 7,510 965 8,875 6,285 14,260 150,245 22,425 190 210,775
2008 Q1 1,725 200 2,615 1,440 2,740 32,205 5,420 50 46,395
2008 Q2 1,600 205 2,470 1,475 2,645 24,895 4,610 55 37,955
Total 38,010 7,265 30,995 40,320 78,970 568,195 89,265 815 853,850
% 4% 1% 3% 5% 9% 67% 10% 0% 100%
Source: Accession Monitoring Report, May 2004 – June 2008, Home Office UK Border Agency.
www.ukba.homeoffice.gov.uk/sitecontent/documents/aboutus/reports/accession_monitoring_report
It appears from the data in Tables 1 and 2 that there has been a slowing in the
numbers of individuals registering on the WRS and also acquiring NINOs. Unfor-
tunately, this information does not tell us about the number of workers in the UK
at any one time. The vast majority of these workers are not migrants. The United
Nations’ definition of a long-term international migrant is someone who moves to
a country other than that of his or her usual residence for a period of at least a
year, so that the country of destination effectively becomes his or her new country
of usual residence.
It turns out that the number of migrants from the EU8 is quite small. According
to a recent ONS publication, “International Migration” (2008), net migration from
the EU8 increased from 61,000 in 2005 to 71,000 in 2006. The estimate of the
numbers of EU8 citizens migrating to the UK for a period of at least a year in-
creased from 53,000 in 2004 to 76,000 in 2005 and 92,000 in 2006. The increase
between 2004 and 2005 can be explained by the fact that 2005 was the first full
calendar year following accession in May 2004 in which EU8 migration could be
estimated. These statistics are much lower than WRS and NINO statistics because
the vast majority of workers who come to the UK from Eastern Europe are not
migrants. These workers may subsequently change their mind and become mi-
grants (thus not returning home in the future), but at the time they enter the UK,
migration is not their apparent intention. These people are temporary workers or
commuters who have the legal right to work in the UK. These workers move in
and out of the UK as jobs are available, providing a degree of flexibility not seen
in a group of migrants who arrive permanently in a country.
A receiving country benefits from temporary workers: if and when the coun-
try’s economy slows down, temporary workers return to their home countries thus
slowing any potential rise in unemployment. It is also possible that temporary
workers cause wage growth to be contained, because the incumbent workforce
may have a fear of being replaced by the temporary workers.
As part of the WRS application process, applicants are asked how long they in-
tend to stay in the UK. In the twelve months ending June 2008, only 11% said
they intended to stay a year or more; 61% said they intended to stay less than three
months. So intended lengths of stay are short (and consistent with the number of
migrants from the EU8 being small). Furthermore, some applicants from Bulgaria
and Romania are only allowed to stay for six months. So given that workers from
Eastern Europe generally intend to, and actually do, stay in the UK for relatively
short spells, it is in our view inappropriate to call these people migrants: they
should more appropriately be considered temporary workers.
Previous work has been done on return migration. For example, LaLonde and
Topel (1997) find that 4.8 million of the 15.7 million US immigrants who arrived
between 1907 and 1957 had departed by the latter year. Yang (2006) examines the
economics of return migration for temporary labor migration by Filipinos. Yang
finds that, on average, a 10% improvement in the exchange rate reduced the
twelve-month migrant return rate by 1.4 percentage points. This is a large effect,
amounting to nearly one-fifth of the mean twelve-month return rate in his sample.
Constant and Zimmermann (2007) examine return (what they call “circular migra-
tion”) and argue that it is potentially a way to minimize psychological costs due to
long separations from family members. Using evidence on the guest worker popu-
lation in the German Socio-Economic Panel, they find that more than sixty% were
repeat migrants.
The probability of a migrant returning to his or her place of origin within a year
is 13%; the probability of moving to another location is 15% (Devanzo, 1983 and
Dustmann, 2003). Zaiceva (2006) summarizes the empirical literature on potential
European migration flows. She shows that between 2 and 4% of the residents of
Central and Eastern European countries (CEECS) will move west in the long run,
and this constitutes about 1% of the EU15 population. Zaiceva also presents evi-
dence from simulations suggesting that the majority of migrants will be from Ro-
mania, Poland and Bulgaria. This is consistent with other estimates in the litera-
ture.
Christian Dustmann, in a number of papers, has undertaken the most complete

analysis in the UK of the economics of return migration. For example, Dustmann
(1994) suggests three potential motives for return migration: (i) the migrant pre-
fers consumption in the home country; (ii) if prices are lower in the home country
than in the host country, this allows the entrant to take advantage of high wages
abroad and low prices at home; and (iii) human capital acquired in the host coun-
try is more valuable in the home country. Dustmann (1996) find that return pro-
pensities in Europe increase with age and decrease with the number of years of
residence. Recently, Dustmann and Weiss (2007) have shown that return migra-
tion in the UK is not a new phenomenon. They explore this issue empirically be-
fore the influx of workers from Eastern Europe using data from the LFS between
1992 and 2004. The authors take the population of immigrants who were still in
the country one year after arrival as the base and find that about 40% of all males
and 55% of all females had left the UK five years later. As we will show below, it
appears that the return rate for workers from the EU10 is even more rapid than for
those who have arrived in the UK from other countries.
3 What are the Characteristics of EU10 Workers?
A considerable amount of information on the characteristics of EU10 workers is

available from the WRS and the LFS. The following information comes from
WRS data in the Accession Monitoring Report May 2004 – June 2008 (unless
otherwise stated).
1. The vast majority (57.4%) of those registered on the WRS are young males,
43% are in the 18-24 age group and 39% are in the 25-34 age group.
2. EU10 workers who have come to the UK are highly educated. While the LFS
contains data on education, it does not provide details of qualifications ob-
tained outside the UK. However, it does provide information on the age when
full-time education was completed. Based on weighted LFS data, the average
age at which EU10 workers complete full-time education was 20.1 years
(20.3 years for Poles). This compares with 17.6 years for natives. A survey of
more than 900 EU8 workers in Fife found that nearly 30% had a university
degree, and a further 22% had an under-graduate level qualification (Fife Re-
search Coordination Group, 2008). The finding that arrivals are predomi-
nantly young men is consistent with the findings in Blanchflower (2001). He
examines the attitudes of Eastern European countries towards market reforms
using data from Eastern European Eurobarometers 1-8 for 1991 to 1997. The
highest levels of support for market reforms are found among the young,
men, and the most educated. Survey questions asked were: (i) whether re-
spondents were satisfied with the way democracy was developing; (ii)
whether the “free market was right for the country’s future”; and (iii) whether
“things in the country are going in the right direction”. A positive response to
all three questions was especially high in Poland in 1997.
3. Only 8% of people registered under the WRS have dependants.
4. The most common occupation is “process operative (other factory worker)”.
This occupation accounted for 28% of workers, followed by “warehouse op-
erative” (8%).
5. 96% of WRS applicants in the twelve months to June 2008 worked more than
sixteen hours per week, and 86% worked more than thirty-five hours per
week.
6. 52% of WRS applicants in the twelve months to June 2008 worked for tem-
porary employment agencies.
7. 68% of WRS applicants in the twelve months to June 2008 stated that they
had a wage of between £4.50 and £5.99 per hour; 24% stated that it was be-
tween £6.00 and £7.99 per hour.
8. The highest proportion of EU10 workers worked in East Anglia (15%), fol-
lowed by the Midlands (13%), and London (12%).
9. London had the highest proportion National Insurance number applications
(26%), followed by the South East (13%), and Scotland (10%).
10. EU10 nationals living in the UK are much more likely to be in a household
with at least three adults (47.7% of EU10 nationals, compared to 24.3% of
natives and 27.6% of non-EU10 nationals). See Blanchflower and Shadforth
(2009).
11. Only relatively small numbers of EU10 nationals have obtained state benefits
such as Income Support or Jobseeker’s Allowance (only around 4,650 have
received a Jobseeker’s Allowance). They came to work and not to claim
benefits. The main exception is that over 111,000 EU10 nationals have re-
ceived Child Benefit (but entitlement to this benefit is near-universal for UK
residents with dependent children; entitlement is not dependent on income or
employment status).
The next part of this section examines the extent to which EU10 nationals in
the UK differ from natives and from non-EU10 migrants, and it is based on the
most recent LFS data available (December 2006 to March 2008). We examine (i)
the propensity of EU10 nationals to be in-work (the so-called employment-to-
population ratio, EPOP); (ii) the incidence of self-employment; (iii) usual working
hours; and (iv) wages and union membership. We find that EU10 nationals have a
high probability of being in-work and of being self-employed, and receive rela-
tively low wages.
3.1 Propensity to Work
We investigate the EU10 arrivals’ propensity to work in the UK (Table 3, column

1). This is known as the employment-to-population ratio (EPOP). The dependent
variable is set to one if the respondent to the Labour Force Survey said they were
working (as an employee or self-employed) and zero otherwise (this includes be-
ing unemployed or out of the labor force, OLF). The sample is restricted to those
aged 16 to 70, and excludes students.6 The equation includes controls for age and
its square, gender, race, region of residence and three year dummies. The equation
also includes a variable identifying whether an individual’s country of origin is an
EU10 or non-EU10 country, and whether the individual arrived in the UK before
or after 2003.
Table 3
Labor market characteristics (ages 16 to 70)
(2) Self- (4) Work limiting (5) Log hourly (6) Union
(1) Work (3) Usual hours
employment health problem earnings membership
New EU10 .0958 (16.76) -.0128 (2.19) 3.8890 (21.52) -.0563 (12.61) -.1342 (8.69) -.1160 (5.26)
Old EU10 .0409 (3.63) .1992 (16.70) 1.5946 (4.66) -.0402 (5.07) -.0412 (1.21) -.0468 (1.18)
New Non-EU10 -.0266 (4.90) -.0581 (11.45) 1.0992 (6.33) -.0464 (11.25) .0637 (4.41) -.1097 (5.85)
Old Non-EU10 -.0053 (1.93) .0213 (8.57) .7840 (9.40) -.0012 (0.65) .0337 (4.71) -.0294 (3.37)
Age .0209 (73.55) .0094 (31.85) 1.0930 (116.21) .0014 (6.60) .0610 (74.92) .0221 (19.99)
2
Age -.0002 (82.37) -.00007 (19.67) -.0132 (117.50) .0000 (1.45) -.0006 (65.41) -.0002 (17.19)
Male .0933 (71.99) .0917 (79.78) 7.7264 (180.27) .0039 (3.86) .1582 (44.16) .0454 (9.84)
Degree .2232 (136.07) -.0171 (8.37) 3.1057 (39.17) -.0365 (23.59) .6972 (101.73) .0832 (8.99)
Higher education .1825 (99.08) -.0280 (12.14) 1.6972 (19.07) -.0191 (10.96) .4597 (60.69) .0916 (8.80)
Apprenticeship
etc .1896 (113.91) .0016 (0.81) 1.3460 (17.76) -.0180 (11.89) .2874 (43.36) .0431 (4.87)
GCSE A-C .1552 (90.57) -.0250 (12.08) .6404 (8.35) -.0160 (10.37) .1789 (27.09) .0079 (0.90)
Other
qualifications .1265 (67.56) -.0299 (13.68) 1.3504 (16.04) -.0072 (4.23) .1144 (15.70) .0248 (2.55)
Don't know .1422 (26.55) .0018 (0.30) 4.1166 (17.13) -.0357 (7.78) .1593 (6.74) -.0077 (0.27)
Mixed race -.0810 (9.62) .0272 (3.40) -.8899 (3.57) .0261 (4.08) -.0368 (1.75) -.0099 (0.36)
Asian -.1172 (29.64) .0298 (8.30) -1.0916 (9.35) .0026 (0.90) -.1470 (14.25) -.0083 (0.64)
Black -.0698 (13.75) -.0421 (10.23) -1.2605 (8.39) -.0079 (2.23) -.2005 (15.37) .0404 (2.38)
Chinese -.0348 (3.25) .0248 (2.64) -.2477 (0.79) -.0215 (2.86) -.1323 (4.79) -.1068 (3.25)
Other race -.1110 (18.04) -.0090 (1.70) -.9581 (5.21) -.0012 (0.27) -.1717 (10.54) -.0204 (0.95)
Self-employed 2.5554 (40.85) .0122 (8.65)
Region dummies
(19) Yes Yes Yes Yes Yes Yes
Industry
dummies (61) No No No Yes Yes Yes
Organisn
dummies (8) No No Yes Yes Yes Yes
2 2
R /Pseudo R 0.1275 0.0688 0.2549 0.0334 0.3988 0.2398
N 445,462 329,037 321,192 326,746 83,193 48,232
Source: LFS, December 2006 to March 2008.

Notes: equations also include three year dummies. Excluded categories are no qualification; white. Columns 2 to 6: workers
only. Column 1: excludes students. Columns 1, 2, 4 and 6 are dprobits; columns 3 and 5 are OLS. “'New EU10” and “New
Non-EU10” are post-2003 arrivals. “Old EU10” and “Old Non-EU10” are pre-2003 arrivals. T-statistics in parentheses.
6 We exclude students on the sensible suggestion of Jonathan Wadsworth. In private com-

munication, he suggested that his work with John Schmitt had shown that results were
sensitive to the inclusion of students.
EU10 nationals have a higher propensity to work than both natives and non-
EU10 nationals. Both EU10 variables are significantly positive, with EU10 nationals
who arrived post-2003 having a near 13% higher probability of working than natives
(compared to only 5% for EU10 nationals who arrived before 2003). In contrast, the
coefficients on both non-EU10 variables are significantly negative. This suggests
that non-EU10 nationals have a significantly lower propensity to work than either
natives or EU10 nationals who arrived before or after 2003. Interestingly, in contrast
to the EU10 nationals who arrived post-2003, the post-2003 arrivals from non-EU10
countries have a significantly lower probability of working than non-EU10 nationals
who arrived prior to 2003.
3.2 Incidence of Self-Employment
In almost all countries for which data are available, the OECD (2006) find that self-
employment among immigrants has increased in both absolute numbers and as a
percentage of total self-employment over the past few years. The OECD find a par-
ticularly marked increase in some countries,. In 2004, foreign-born persons ac-
counted for about 11% of total self-employment in both the UK and France; it was
12% in Belgium and nearly 14% in Sweden. These figures are generally higher than
the share of immigrants in the total labor force. Dustmann and Kirchkamp (2002)
study Turkish immigrants in Germany who subsequently returned home. They
found that approximately half of the immigrants became entrepreneurs when they
returned to Turkey.
Based on Flash Eurobarometers (“Entrepreneurship” 2000 to 2004), Blanch-

flower and Shadforth (2009) find that a high proportion of workers in Lithuania
(62%) and Poland (57%) said they would like to be self-employed.7 They also find
that a relatively high proportion of workers in both countries (52% in Lithuania and
42% in Poland) report that it was hard to start-up a business because of a lack of fi-
nancial support (this compares to only 24% in the UK and 26% in the US). Hence,
part of the desire to come to the UK may be to raise capital in order to overcome li-
quidity constraints at home.
We next examine the incidence of self-reported self-employment among EU10

workers, given that the self-employed do not need to register under the WRS (al-
7 The question asked was: “Suppose you could choose between different kinds of jobs.
Which one would you prefer: being an employee or being self-employed?” For the remain-
ing EU8 countries, the percentage preferring self-employment was as follows: the Czech
Republic 37 per cent; Estonia 49 per cent; Hungary 47 per cent; Latvia 44 per cent; Slova-
kia 36 per cent; and Slovenia 35 per cent. This compares to: 43 per cent in France; 46 per
cent in Germany; 47 per cent in the UK; and 66 per cent in the US. For more on the desire
to be self-employed across countries, see Blanchflower, Oswald and Stutzer (2001).
though they do need to apply for a NINO). Column 2 of Table 3 reports the results.
The probability of a worker being self-employed is estimated using a dprobit with
the dependent variable set to one if self-employed and zero if an employee.8 The re-
sults are standard in that the probability of being self-employed is higher for men,
Asians and Chinese; it rises with age; and is especially high for those with a trade.9
It is low for blacks, the young, and for those with only an HND, teaching or nursing
(“other”) qualification. Foreign workers who arrived before 2004 have a higher
probability of being self-employed than natives. EU10 workers have a particularly
high probability of being self-employed; however, the more recent arrivals have a
lower probability (recent EU10 arrivals have a particularly low propensity to be self-
employed). This may be partly due to the difficulty of capturing self-employment
among EU10 workers in the Labour Force Survey.
3.3 Usual Working Hours
Usual working hours is estimated in Table 3, column 3. Workers from the EU10
work longer hours, and those who arrived post-2003 work on average 3.9 hours
longer per week than natives. In the LFS, workers report whether they have a con-
dition that limits their ability to work. The probability that a worker reports such a
condition is estimated in column 4. Recent arrivals from the EU10 are much less
likely than natives to report such a condition.
3.4 Wages and Union Membership
There is also evidence (including LFS data) to suggest that EU10 workers in gen-
eral, and the more recent arrivals in particular, are being paid relatively low wages,
ceteris paribus. This is shown in Table 3, column 5, where the results of estimating
log hourly wage equations using LFS data (2004 to 2007) are reported. Sample sizes
are smaller than for columns 2 to 4 because the sample is restricted to employees
only, and restricted further because wages are only asked in waves one and five of
the survey.10 There are approximately 83,000 observations in total. Column 5 shows
that recent EU10 workers receive 12.5% lower wages than natives (obtained by tak-
ing anti-logarithms and deducting one). Interestingly, non-EU10 migrants have sig-
nificantly higher wages (6.6%) than natives, holding characteristics (including race)
8 The dprobit command in STATA reports the marginal effect, that is, the change in the
probability for an infinitesimal change in each independent, continuous variable. By de-
fault, it reports the discrete change in the probability for dummy variables.
9 See Blanchflower (2004) and Blanchflower and Shadforth (2009).
10 In the LFS since 1997, a fifth of the sample is replaced each quarter. Individuals remain in
the sample for five consecutive waves or quarters.
constant (column 6).11 The OECD (2006, p. 222) reports that one of the main fea-
tures of labor immigration in the UK, prior to the influx of EU10 workers, was the
high proportion accounted for by corporate transfers. The OECD noted that in 2005,
a quarter of interviewees who were working in the UK at the time of the interview,
but were working abroad in the previous year, were working for the same employer
(both in the UK and abroad). This may partly explain the higher earnings of the non-
EU10 workers.
Column 6 suggests that lower wages partly arise because foreign-born workers are
much less likely to be union members than natives. It is well understood that union
membership brings significant wage benefits (Blanchflower and Bryson, 2003, 2008).
In summary, the new arrivals from Eastern Europe who have come to work in the
UK tend to be young, male, healthy, educated, non-unionized and unmarried. Ap-
proximately one third work for recruitment agencies. They disproportionately work
in East Anglia, and the West and East Midlands. Holding constant a variety of char-
acteristics (including age, qualifications, and location), EU10 workers who have ar-
rived since accession in 2003 have higher self-employment rates, longer working
weeks, are fitter, have lower wages, are less likely to be union members and have
higher employment-to-population ratios than natives. In contrast, non-EU10 mi-
grants who have arrived in the UK since 2003 are less likely to work, but when they
do, they obtain higher wages than natives.
4 Why Did EU10 Workers Come to the UK?
In this section, we examine the reasons for the influx of workers from the EU10
countries since accession. We make use of data from a number of Eurobarometers
(2004 onwards), with the most recent taken at the end of 2007. On a large number of
measures and from several different data files, respondents from former communist
countries consistently say they are unhappy with their lives, jobs, the country they
live in and its economic situation, as well as prospects for the future (Blanchflower
and Freeman, 1997; and Blanchflower, 2001, 2008). Reasons given are broadly con-
sistent across data sets and vary little across different questions. A significantly high
proportion of EU10 residents report that they have thought about living abroad. A
relatively high proportion of EU10 workers say they would be prepared to move
abroad if they became unemployed. Unemployment appears to be a particular con-
cern among the nationals of EU10 countries. In this section, we provide evidence
11 Drinkwater et al (2006) analyse wages in Labour Force Surveys between 2001 and 2006.
They find that Poles have lower rates of return to their human capital than other recent mi-
grants, even after controlling for other personal and job-related characteristics.
that, in comparison with other EU countries, EU10 respondents report that the fol-
lowing is true of their home countries:
1) It is difficult to find a good new job.

2) Levels of job satisfaction are low.
3) Workers feel insecure about their jobs and fear that they could be replaced
easily.
4) The work/life balance is poor.
5) Life satisfaction is low, and people report low levels of happiness with their
lives and the country they live in.
6) Expectations of the macro-economy are low.
7) Inflation and unemployment are of particular concern.
8) Workers would be prepared to move abroad if they were made unemployed.
Each of these factors shows a low level of satisfaction and contentment among
EU10 nationals and demonstrates that there are strong “push incentives” to leave.
There is also evidence that the desire to learn English has been a strong “pull incen-
tive” in coming to the UK for EU10 nationals.
In the remainder of this section, we closely examine the main reasons for the in-
flux of EU10 workers to the UK since accession.
4.1 Job Satisfaction
Table 4 measures the difficulty in finding a good job, job satisfaction and work/life
balance. It covers workers only (with the exception of column 1) and uses data from
Eurobarometer #64.1 for 2005, covering the EU25 (i.e. including the EU8). Col-
umns 1 and 2 report ordered logits where the dependent variable “difficulty in find-
ing a good job” is scored as 4 = “totally agree”; 3 = “'tend to agree”; 2 = “tend to
disagree”; and 1 = “totally disagree”. The results show that it is very difficult to find
a good job in an EU8 country, and the coefficients are large for each of the EU8
country dummies, especially for Hungary.
Columns 3 to 5 estimate job satisfaction ordered logits in relation to salary, num-

ber of working hours, and the work/life balance. Column 6 estimates an OLS where
the dependent variable is the sum of ten job satisfaction variables, each coded 1 to 4
(mean = 30.9; sd = 4.7). Job satisfaction on each of the four measures is U-shaped in
age (Clark, Oswald and Warr, 1996). The data shows that workers from the EU10,
and from Hungary in particular, were dissatisfied with their jobs. This is consistent
with findings by Blanchflower and Freeman (1997) who find that, using data from
the 1989 ISSP survey, job satisfaction levels are lower in Hungary than in Western
countries. The data also shows that the work/life balance (column 5) is lowest in the
EU10 countries (particularly Latvia) and is lower for the most educated.
Table 4
Difficulty in finding a job and job satisfaction (2005)
Difficult to Difficult to Job satisfaction
find a job find a job Salary # hours Work life All
Ordered logit Ordered logit Ordered logit Ordered logit Ordered logit OLS
Czech Republic 1.1605 (14.48) 1.1468 (10.23) -.6714 (5.70) -.3049 (2.64) -.1360 (1.17) -1.3419 (4.53)
East Germany 2.4117 (21.26) 2.4252 (14.11) -.6379 (3.96) -.0969 (0.60) .1661 (1.02) -1.1292 (2.75)
Estonia .9161 (11.37) .9057 (7.70) -.7641 (6.24) -.0883 (0.73) -.1815 (1.49) -1.8504 (5.97)
Hungary 2.4462 (26.93) 2.7023 (18.38) -1.5834 (12.11) -.5317 (4.03) -.2605 (1.95) -2.9140 (9.09)
Latvia 1.0884 (13.60) 1.0293 (8.98) -1.3954 (11.69) -.3385 (2.87) -.6546 (5.49) -2.9984 (10.24)
Lithuania 1.5810 (19.09) 1.3793 (11.08) -1.2300 (9.52) -.4994 (3.89) -.2647 (2.08) -2.5145 (7.82)
Poland 1.9012 (22.51) 1.7601 (13.01) -1.1206 (8.11) -.3936 (2.85) -.3796 (2.75) -2.3754 (6.38)
Slovakia 1.6820 (20.86) 1.6443 (14.48) -.8669 (7.26) -.5163 (4.43) -.3073 (2.63) -2.4399 (8.33)
Slovenia 1.3586 (17.24) 1.3459 (11.45) -.4575 (3.68) -.5144 (4.23) -.3774 (3.09) -1.8955 (6.40)
Austria .8507 (10.73) .8731 (7.81) .5020 (4.17) .0432 (0.37) .6176 (5.28) .9651 (3.34)
Belgium .2413 (3.10) .0392 (0.35) .1935 (1.59) .2949 (2.49) .1625 (1.36) .0294 (0.10)
Cyprus 1.3743 (13.44) 1.3987 (9.65) -.1031 (0.67) -.4133 (2.79) .0640 (0.44) -1.4760 (3.68)
Denmark -.8675 (10.60) -1.2644 (10.87) .7648 (6.16) .9793 (8.08) 1.0672 (8.73) 2.2686 (7.75)
Finland -.1691 (2.15) -.3461 (2.98) -.3650 (2.89) -.1431 (1.18) .0989 (0.80) -1.2530 (3.75)
France .6294 (8.00) .5513 (4.85) -.9238 (7.60) -.4335 (3.60) -.1346 (1.11) -.8758 (2.91)
Greece 1.9041 (22.45) 2.0568 (15.87) -.4027 (3.04) -.2363 (1.81) -.1254 (0.97) -1.8999 (6.24)
Ireland -.4873 (6.05) -.4563 (3.95) .3767 (3.03) .3253 (2.69) .4002 (3.31) .4441 (1.46)
Italy 1.1609 (14.59) 1.1547 (10.08) -.3870 (3.16) -.5361 (4.54) -.2331 (1.95) -2.7680 (9.28)
Luxembourg .6009 (6.06) .5002 (3.31) .9350 (5.62) .9203 (5.70) .4335 (2.68) 1.1058 (2.77)
Malta .6283 (6.29) .7514 (4.39) -.6968 (3.77) .3424 (1.87) .0003 (0.00) -1.5257 (3.22)
Netherlands .0531 (0.67) -.0647 (0.60) .4205 (3.58) .6465 (5.65) .3869 (3.37) .5856 (2.08)
Portugal 1.8364 (21.72) 1.9446 (14.92) -.7338 (5.54) -.3918 (3.06) -.0786 (0.61) -2.0299 (6.18)
Spain 1.3948 (16.99) 1.4493 (11.62) -.2069 (1.57) -.4322 (3.38) -.0051 (0.04) -1.6260 (5.18)
Sweden -.2404 (2.96) -.5252 (4.72) -.3896 (3.35) .1224 (1.07) .1439 (1.25) -.6336 (2.24)
West Germany 1.5858 (19.61) 1.6019 (13.83) -.0126 (0.10) .1045 (0.88) .1716 (1.43) -.5072 (1.72)
Age -.0028 (2.52) .0019 (1.20) -.0450 (4.36) -.0405 (4.02) -.0575 (5.62) -.1065 (4.07)
2
Age .0005 (4.19) .0005 (4.50) .0007 (6.65) .0014 (4.76)
Male -.1866 (6.98) -.2584 (6.75) .2843 (7.17) -.1177 (3.00) -.0060 (0.15) .2367 (2.40)
15-19 years schooling -.2471 (6.69) -.2383 (3.72) .0076 (0.12) .0006 (0.01) -.1201 (1.86) -.0500 (0.30)
≥20 years schooling -.5455 (12.85) -.4874 (6.93) .0879 (1.22) -.0099 (0.14) -.2245 (3.13) .0271 (0.15)
Minority .2285 (4.22) .3083 (4.03) -.3400 (4.30) -.0760 (0.97) -.0962 (1.23) -.6375 (3.21)
Student -.6518 (8.85)
Unemployed .6413 (9.25)
Retired .0327 (0.62)
Professional -.5767 (5.36) -.3737 (2.33) 1.3177 (7.86) 1.4279 (8.59) .6337 (3.79) 2.6202 (4.94)
Shop owner -.2270 (2.60) -.0284 (0.20) 1.0409 (6.98) .9235 (6.24) .4893 (3.27) 2.3728 (4.69)
Business proprietor -.6405 (6.14) -.4236 (2.70) 1.5541 (9.52) 1.0957 (6.81) .6874 (4.22) 3.1334 (5.96)
Employed professional -.7373 (8.53) -.5175 (3.52) 1.0092 (6.72) 1.3974 (9.27) .5381 (3.54) 1.6144 (3.33)
General management -.7789 (6.34) -.5510 (3.23) 1.8943 (10.60) 1.2970 (7.30) .4995 (2.80) 2.6514 (5.00)
Middle management -.5122 (7.78) -.2783 (2.05) 1.0878 (7.85) 1.5037 (10.82) .5782 (4.11) 1.5339 (3.32)
Employed - desk -.2242 (3.70) .0056 (0.04) .8406 (6.20) 1.6613 (12.19) .6362 (4.62) 1.0896 (2.38)
Travelling salesman -.1980 (2.33) .0647 (0.45) .6266 (4.30) 1.1060 (7.58) .1828 (1.24) .0680 (0.14)
Employed not at a desk -.2492 (3.88) .0033 (0.03) .5937 (4.35) 1.3769 (10.02) .5156 (3.71) .6522 (1.42)
Supervisor -.3656 (2.60) -.0960 (0.52) 1.0255 (5.33) 1.5588 (8.26) .5989 (3.13) 1.4063 (2.50)
Skilled manual -.0080 (0.13) .2604 (1.98) .5275 (3.98) 1.2623 (9.49) .3478 (2.58) -.1703 (0.38)
Unskilled manual .1334 (1.71) .3703 (2.62) .4115 (2.90) 1.2545 (8.81) .3369 (2.33) -.7355 (1.55)
cut1/constant -2.4404 -2.1081 -2.9094 -2.7471 -4.0367 32.6511
cut2 -0.71 -0.322 -1.2209 -1.1085 -2.2257
cut3 0.8844 1.2653 1.6827 1.4604 0.3257
Workers only No Yes Yes Yes Yes Yes
N 23954 11127 11134 11151 11115 9188
2
Pseudo/Adj R 0.1012 0.1101 0.0609 0.0336 0.0231 0.134
Source: Eurobarometer #64.1, Mobility, Food Risk, Smoking, AIDS Prevention and Medical Errors,
September-October 2005 (ICPSR 4641).
Notes: excluded categories are UK; responsible for shopping (column 1); farmer; and less than sixteen years schooling. T-
statistics in parentheses. Question 1: It is difficult to find a good job in (OUR COUNTRY)? 4 = totally agree; 3 = tend to
agree; 2 = tend to disagree; 1 = totally disagree. Question 2: Generally speaking, when you think about your professional
life, could you tell me whether you are 4 = very satisfied; 3 = fairly satisfied; 2 = fairly dissatisfied; or 1 = not at all satisfied
with each of the following: (i) your salary; (ii) your work contract; (iii) the number of hours you work; (iv) your commuting time;
(v) your career prospects; (vi) the content of your job; (vii) your colleagues; (viii) your training opportunities; (ix) the balance
between your private life and your working life; and (x) the health and safety conditions in your company.
Table 5
Job satisfaction, job security and the likelihood of being replaced (2005)
Job satisfaction Job security Easy to Replace

Bulgaria -.1965 (1.63) -.8219 (7.04) .9985 (8.30)
Czech Republic -.4521 (4.03) -.5342 (5.01) .7196 (6.46)
East Germany .1898 (1.37) -.4336 (3.17) .1058 (0.75)
Hungary -.3305 (2.65) -.3633 (3.08) .3348 (2.73)
Latvia -.1673 (1.32) -.4289 (3.46) -.2721 (2.13)
Russia -.4404 (4.04) .0072 (0.07) .4878 (4.65)
Slovenia -.4606 (3.85) .4020 (3.42) .1315 (1.12)
Australia -.3047 (3.00) -.2052 (2.09) .1789 (1.78)
Canada -.1259 (1.08) -.0997 (0.87) .2886 (2.49)
Cyprus .9777 (7.75) -.1224 (1.00) .7440 (5.67)
Denmark .2724 (2.63) .4700 (4.58) -.0716 (0.70)
Dominican Republic .1546 (1.47) .1512 (1.49) 1.2832 (11.90)
Finland -.0972 (0.89) -.2926 (2.70) .1166 (1.05)
Flanders -.4811 (4.46) -.1161 (1.12) .0892 (0.84)
France -.5886 (5.73) -.3827 (3.74) .6971 (6.85)
Ireland .5505 (4.77) .3035 (2.67) .4101 (3.49)
Israel .2115 (1.80) -.2189 (1.92) .2237 (1.92)
Japan -.7262 (6.18) .1857 (1.51) .2314 (1.92)
Mexico 1.0700 (9.36) .3568 (3.18) 1.0513 (8.74)
New Zealand -.1578 (1.49) .0240 (0.23) -.1365 (1.30)
Norway -.1857 (1.76) -.2855 (2.80) .2313 (2.23)
Philippines .4397 (3.74) .1179 (1.05) .8860 (7.49)
Portugal -.0374 (0.36) -.0889 (0.86) .5682 (5.46)
South Africa .2902 (2.64) -.0425 (0.40) .7716 (6.96)
South Korea -1.1639 (11.02) -.8819 (8.68) .4163 (3.90)
Spain -.0617 (0.53) .2159 (1.88) .3344 (2.90)
Sweden -.2896 (2.71) -.0963 (0.93) .0813 (0.77)
Switzerland .6848 (6.21) .0172 (0.16) .0522 (0.47)
Taiwan -.6054 (6.11) -.4757 (4.97) .5297 (5.28)
USA .2254 (2.16) .1226 (1.21) .3218 (3.08)
West Germany .2047 (1.74) .3242 (2.78) .0918 (0.78)
Age .0106 (10.25) -.0255 (4.24) -.0191 (3.04)
2
Age .0003 (4.94) .0001 (2.55)
Male -.0031 (0.13) -.0706 (2.83) -.2663 (10.47)
Lowest formal .3959 (5.18) .1044 (1.40) -.2731 (3.36)
Above lowest .4816 (6.64) .1029 (1.46) -.3509 (4.57)
Higher secondary .6206 (8.71) .2641 (3.81) -.4347 (5.78)
Above higher sec. .7277 (10.09) .3218 (4.58) -.5740 (7.54)
Degree completed .7972 (11.11) .5050 (7.20) -.6075 (8.01)
Other qualification 1.4434 (1.49) .3123 (0.34) .4312 (0.51)
Self-employed .4917 (13.91) -.1205 (3.44) -1.1464 (28.55)
Part time -.0289 (0.83) -.2211 (6.26) .2363 (6.56)
<part-time -.0467 (0.48) -.6727 (7.06) .3718 (3.81)
Helping family member -.4492 (3.45) .0081 (0.06) -.3200 (2.08)
cut1 -3.6753 -3.411 -3.1686
cut2 -2.591 -1.8103 -1.6748
cut3 -1.5238 - 0.9233 -0.5425
cut4 -0.5438 0.882 0.9037
cut5 1.3103
cut6 2.9072
N 23,221 23,066 21,707
2
Pseudo R 0.0259 0.0148 0.0301
Source: International Social Survey Programme, Work Orientation module, 2005

Notes: excluded categories are UK; full-time; and no formal qualification. T-statistics in parentheses. All columns are or-
dered logits. The age squared variable in column 1 was insignificant and hence is omitted. Question 1: How satisfied are
you in your main job? 7 = completely satisfied; 6 = very satisfied; 5 = fairly satisfied; 4 = neither satisfied nor dissatisfied; 3
= fairly dissatisfied; 2 = very dissatisfied; 1 = completely dissatisfied. Question 2: “My job is secure.” Do you: 4 = strongly
agree; 3 = agree; 2 = neither agree nor disagree; 1 = strongly disagree with this statement. Question 3: How difficult or easy
do you think it would be for your firm or organisation to replace you if you left? 5 = very easy; 4 = fairly easy; 3 = neither
easy nor difficult; 2 = fairly difficult; 1 = very difficult.
Table 5 provides evidence on job satisfaction and workers’ views of job secu-
rity and the likelihood of being replaced. It uses data from the 2005 International
Social Survey Programme (ISSP) and covers thirty-two countries, including five
from the EU10 (Bulgaria, the Czech Republic, Hungary, Latvia and Slovenia),
Russia, eight developed non-EU countries (Australia, Canada, Israel, Japan, New
Zealand, Norway, Switzerland and the US), and six less developed countries (the
Dominican Republic, Mexico, the Philippines, Taiwan, South Africa and South
Korea). The data shows that workers in the five EU10 countries were especially
likely to report low levels of job satisfaction and job security. This confirms the
Eurobarometer findings shown in Table 4 (above). The data also shows that work-
ers in Bulgaria, the Czech Republic and Hungary are especially likely to report
that it would be easy for their employers to replace them.
4.2 Life Satisfaction
It is well known that EU10 residents are particularly unhappy with their lots
(Blanchflower and Freeman, 1997; Blanchflower, 2001; and Blanchflower and
Shadforth, 2009). However, there has been some improvement over time in life
satisfaction levels. Table 6, column 1, presents life satisfaction levels in the EU10
countries (and the UK). It uses the mean scores from Eurobarometers #62.0 (for
2004), and #68.1 (for 2007). The scoring method used was 1 = “not at all satis-
fied”; 2 = “not very satisfied”; 3 = “fairly satisfied”; and 4 = “very satisfied”.
Eurobarometer #62.0 is the first in the series to survey individuals from the EU10.
Results show that levels of life satisfaction increased in eight of the ten accession
countries between 2003 and 2007 (satisfaction levels fell in Hungary and Slove-
nia). With the exception of Slovenia, satisfaction scores for the EU10 countries
were well below the UK score. Bulgaria’s satisfaction score was particularly low.
Table 6 also reports the EU10 and UK respondents’ expectations of the macro-
economy and respondents’ more personal situations for the twelve months ahead.
Overall, there was some improvement in the 2007 responses (compared to the
2004 responses), especially in Poland, Lithuania, Latvia and Slovakia; however,
expectations in Hungary worsened in all categories. Workers in Bulgaria, the
Czech Republic and Hungary became less optimistic their personal job prospects.
Table 7 also makes use of data from Eurobarometer #68.1 (for 2007). It in-
cludes data for Turkey, Croatia and Macedonia (these countries are in the process
of applying for EU membership). Column 1 reports the econometric results of es-
timating a life satisfaction ordered logit with standard controls. It shows that life
satisfaction is high in Denmark, the Netherlands, Luxembourg and Sweden.12 Life
12 The question used in the Eurobarometers is, “On the whole, are you: 4 = very satisfied; 3
= fairly satisfied; 2 = not very satisfied; and 1 = not at all satisfied with the life you lead?”
satisfaction is especially low in Bulgaria, Romania and Hungary. Compared with

life satisfaction in the UK, life satisfaction in all of the EU10 countries is signifi-
cantly lower, ceteris paribus.
Table 6
Life satisfaction and percentage of respondents
reporting that the twelve months ahead will be better (2004 and 2007)
Life Expectations 12 months ahead (% better)

satisfaction Economic Financial Employment Personal Job
Life Situation Situation Situation Situation
2004 2007 2004 2007 2004 2007 2004 2007 2004 2007 2004 2007
Bulgaria 2.06 2.15 25 26 24 26 19 24 23 30 35 20
Czech Republic 2.82 2.91 17 22 9 18 8 19 8 27 26 18
Estonia 2.74 2.8 39 40 39 30 32 35 35 32 14 35
Hungary 2.44 2.38 24 15 19 14 20 12 17 10 26 11
Latvia 2.52 2.68 31 34 26 21 27 31 27 32 26 33
Lithuania 2.55 2.63 39 35 37 31 33 31 44 48 28 38
Poland 2.81 2.85 35 30 26 33 21 25 23 40 17 20
Romania 2.32 2.39 50 44 44 36 44 41 35 33 34 38
Slovakia 2.59 2.74 24 29 18 24 18 22 24 33 18 22
Slovenia 3.17 3.1 32 26 28 24 22 20 19 21 16 16
UK 3.22 3.22 44 35 16 17 32 30 17 17 28 30
Source: 2004 data from Eurobarometer #62.0 and Standard European Trend Questions and Sport, Octo-
ber-November, 2004 (ICPSR 4289). 2007 data from Eurobarometer #68.1, “Public Opinion in the Euro-
pean Union”, September-November 2007.
Notes: Question 1: On the whole, are you: 4 = very satisfied; 3 = fairly satisfied; 2 = not very satisfied; 1 = not at all satisfied
with the life you lead?
Question 2: What are your expectations for the next twelve months – will the next twelve months be better, worse, or the
same when it comes to…
a) Your life in general?
b) The economic situation in our country?
c) The financial situation of your household?
d) The employment situation in our country?
e) Your personal job situation?
Earlier work has found that happiness is U-shaped in age; it rises with the level
of education; and is higher for married people, women, and the self-employed
(Blanchflower and Oswald, 2004, 2008; Blanchflower, 2008). Di Tella and Mac-
Culloch (2005) found that right-wing voters are happier than left wing voters, and
the unemployed are especially unhappy. The results are very similar to earlier
work, where an equation is estimated for the EU10 countries alone (Table 7, col-
umn 2). The results show that life satisfaction rises with the level of education,
and is higher for married people, right-wing voters and the self-employed. It is U-
shaped in age and lower for the unemployed. Bulgarians, Romanians and Hun-
garians are the least happy; and Slovenians are the happiest.
Hence, one of the likely reasons that EU10 workers came to the UK is that they
were unhappy at home. And if the least happy EU10 workers left to come to the
UK, then happiness levels in the EU10 should rise (although there is no evidence
either way on this proposition).
Table 7
Life satisfaction, important issues facing the country and its economic situation (2007)
Life satisfaction Rising prices /
Life satisfaction Unemployment National economy
(A10 only) inflation
Bulgaria -2.5924 (31.73) -1.9220 (20.69) .2100 (8.35) .5615 (23.52) -3.0994 (33.82)
Czech Republic -.7988 (10.18) excluded category .1040 (4.37) .4590 (18.91) -1.6031 (17.74)
Estonia -1.0884 (13.63) -.3286 (3.57) -.0118 (0.49) .5865 (25.49) .1596 (1.67)
Hungary -1.9342 (24.33) -1.1318 (12.42) .4149 (16.94) .3665 (14.53) -3.2811 (35.98)
Latvia -1.4093 (17.56) -.7179 (7.84) .0774 (3.25) .6765 (32.10) -2.4374 (26.91)
Lithuania -1.4972 (18.54) -.7730 (8.33) -.0360 (1.54) .6039 (26.53) -1.7294 (19.27)
Poland -.8328 (10.09) -.0083 (0.09) .3410 (13.72) .2690 (10.29) -1.4638 (16.09)
Romania -2.1805 (27.05) -1.4535 (15.77) .1401 (5.73) .3947 (15.59) -2.5480 (27.97)
Slovakia -1.1815 (15.00) -.4230 (4.72) .3265 (13.53) .3691 (14.89) -1.0871 (12.27)
Slovenia -.1373 (1.68) .7956 (8.11) .2180 (8.83) .6032 (26.52) -.3330 (3.62)
Austria -.2412 (3.02) .3006 (12.31) .4363 (17.73) .5217 (5.41))
Belgium .1522 (1.89) .3008 (12.28) .4014 (16.06) -.4570 (4.97)
Croatia -.9222 (11.01) .5552 (23.04) .0923 (3.46) -3.1833 (34.47)
Cyprus -.2679 (2.56) .1467 (4.95) .4945 (17.41) n/a
Denmark 1.5762 (18.13) -.0709 (2.84) .0307 (1.12) 2.7302 (28.26)
East Germany -1.1793 (11.65) .5337 (19.05) .4750 (16.73) -.4443 (3.93)
Finland .1235 (1.54) .2112 (8.54) .2729 (10.50) .5961 (6.33)
France -.6035 (7.56) .4318 (17.78) .3148 (12.39) -2.2685 (25.36)
Greece -1.4308 (17.73) .4446 (18.11) .3945 (15.66) -3.0403 (32.82)
Ireland .1907 (2.34) .0273 (1.16) .2326 (8.99) 1.2565 (13.03)
Italy -1.0781 (13.21) .3176 (12.81) .3268 (12.72) -1.6365 (17.53)
Luxembourg .7526 (7.23) .3844 (13.10) .4187 (14.12) .8463 (7.02)
Macedonia -1.4983 (17.65) .6113 (25.62) .1635 (6.18) n/a
Malta -.3948 (3.74) .1308 (4.48) .4427 (15.16) -1.2017 (10.55)
Netherlands .8062 (9.94) -.0730 (2.97) .1298 (4.88) .6767 (7.16)
Portugal -1.8310 (22.89) .5228 (21.41) .3796 (14.88) -3.0912 (33.34)
Spain -.2872 (3.43) .1836 (7.40) .1343 (5.02) -.6500 (6.95)
Sweden .6327 (7.78) .2355 (9.39) -.0826 (2.93) 1.4711 (15.43)
Turkey -.7104 (7.85) .5405 (21.48) -.0917 (3.31)) -1.5580 (15.51)
West Germany -.2911 (3.60) .4695 (19.40) .3561 (14.12) .0626 (0.66)
Age -.0696 (14.57) -.1024 (11.96) -.0020 (7.61) -.0013 (4.66) -.0211 (4.17)
2
Age .0006 (13.64) .0009 (10.96) .0002 (5.21)
Male -.0554 (2.15) .0153 (0.35) -.0078 (1.37) -.0099 (1.59) .2808 (10.25)
ALS 16-19 .2679 (7.91) .4136 (6.46) -.0268 (3.67) -.0217 (2.70) .2213 (6.03)
ALS 20+ .5392 (13.42) .7217 (9.57) -.0534 (6.17) -.0605 (6.39) .4380 (10.19)
Still studying .7544 (9.56) 1.0197 (6.72) .0470 (1.65) -.0560 (1.99) .4614 (5.67)
No FT education -.3677 (1.57) .4521 (1.02) -.0104 (0.18) .0597 (0.92) .6596 (1.95)
Married .3493 (7.62) .4092 (4.77) -.0140 (1.46) .0286 (2.67) -.0331 (0.68)
Remarried .2635 (2.84) .2590 (1.80) -.0348 (1.68) .0384 (1.70) -.0584 (0.58)
Living together .0935 (1.67) .0754 (0.76) -.0226 (1.90) .0130 (0.96) -.1056 (1.78)
Past lived together -.3098 (4.32) -.1117 (0.83) -.0083 (0.54) -.0033 (0.19) -.1717 (2.28)
Divorced -.3293 (5.18) -.1865 (1.71) -.0020 (0.15) .0260 (1.71) -.2623 (3.88)
Separated -.3480 (3.37) -.0632 (0.33) -.0080 (0.35) .0859 (3.35) -.4000 (3.63)
Widowed -.3085 (5.02) -.0869 (0.81) -.0075 (0.55) .0362 (2.39) -.1709 (2.61)
Native .0918 (1.82) -.0963 (0.88) .0011 (0.10) .0024 (0.20) .0114 (0.34)
Centre voter .1389 (4.44) .0350 (0.60) -.0063 (0.92) -.0069 (0.92) .1585 (4.30)
Right-wing voter .3079 (8.84) .3239 (5.19) -.0379 (4.99) -.0231 (2.76) -.2492 (3.63)
Unemployed -.7591 (11.58) -.6413 (4.94) .2468 (8.22) .0327 (1.17) -.0336 (0.58)
Retired -.1006 (1.86) -.1764 (1.46) .0654 (2.57) .0485 (1.84) -.1142 (1.01)
Farmer -.3292 (3.17) -.3789 (1.87) .0891 (2.63) .0546 (1.54) -.1384 (0.25)
Fisher -1.1623 (1.80) -1.4893 (1.53) .2559 (1.52) -.0403 (0.23) .4097 (3.34)
Shop owner .3691 (3.50) .9339 (3.94) .0463 (1.57) .0276 (0.90) -.0719 (0.81)
Business proprietor .0328 (0.39) .3625 (1.94) .0539 (1.59) .0101 (0.30) .1278 (1.17)
Professional .4607 (4.47) .8127 (4.34) .0051 (0.17) .0224 (0.73) .1806 (2.00)
General management .3244 (3.73) .4303 (2.99) .0110 (0.30) .0143 (0.37) .4696 (3.60)
Middle management .5456 (4.25) 1.0065 (3.49) .0588 (2.19) .0276 (1.01) .2163 (3.12)
Employed at a desk .3444 (5.22) .4866 (3.57) .0574 (2.17) .0474 (1.75) .1923 (3.00)
Employed travelling .1686 (2.77) .1007 (0.80) .0496 (1.65) .0641 (2.04) -.0568 (0.64)
Service employee .1128 (1.31) .2186 (1.39) .0695 (2.56) .0715 (2.55) -.0582 (0.87)
Supervisor .0862 (1.36) .1226 (0.93) .0677 (1.66) .1035 (2.42) .0588 (0.42)
Skilled manual .1018 (0.73) .1865 (0.76) .0952 (3.53) .0647 (2.37) .0024 (0.04)
Unskilled manual -.0844 (1.42) -.0372 (0.31) .0978 (3.34) .0857 (2.86) -.0092 (0.12)
Cut1 -5.0512 -5.2958 -3.9913
Cut2 -3.1644 -3.1666 -1.2861
Cut3 -0.0695 0.1425 2.4924
N 28,759 28,759 27,158
2
Pseudo R 0.1242 0.1075 0.1526 0.138 0.2249
Source (Table 7): columns 1 to 4 are from Eurobarometer #68.1, “Public Opinion in the European Union”,
September-November 2007. Column 5 is from Eurobarometer #67.2, “European Union Enlargement, Per-
sonal Data Privacy, the National Economy, and Scientific Research”, April-May 2007.
Notes: excluded categories are UK; single; left-wing voter; responsible for shopping; and age left school is less than sixteen
years. Columns 3 and 4 are dprobits. Columns 1, 2 and 5 are ordered logits. Responses to life satisfaction questions were:
1 = not at all satisfied; 2 = not very satisfied; 3 = fairly satisfied; and 4 = very satisfied. Responses to national economy’s
situation were: 1 = very bad; 2 = rather bad; 3 = rather good; and 4 = very good.
4.3 Macro-Economy
Eurobarometer #68.1 (2007) asks respondents to name the “two most important
issues facing their country” from a (given) list of fourteen (including crime, taxa-
tion, terrorism, health care, pensions and education). The probability of a respon-
dent naming unemployment is given in column 3 of Table 7. The probability of
naming rising prices/inflation is given in column 4.13 Worries about unemploy-
ment are high in Macedonia, Croatia, Turkey, East Germany, Hungary and Po-
land. Worries about inflation are high in Latvia, Lithuania, Slovenia and Estonia.
Eurobarometer #67.2 (also 2007) asks respondents whether they believe the
situation in their national economy is good or bad (1 = very bad; 2 = rather bad; 3
= rather good; and 4 = very good). An ordered logit is estimated in column 5 of
Table 7. Responses are consistent with those relating to life satisfaction: EU10 re-
spondents are especially pessimistic about the situation in their national econo-
mies, and this is particularly strong for Hungarians and Bulgarians. Danes, and to
a lesser extent Swedes, are especially optimistic.
4.4 Propensity to Move Abroad
Table 8 explores the potential flow of workers from the EU8 countries using
Eurobarometer #65.1 for 2006 (columns 1 to 3), and Eurobarometer #64.1 for
2005 (columns 4 and 5). In column 1, respondents report on whether they are
happy in the country they live in, and this is estimated using an ordered logit. The
results show that happiness is highest in Denmark and Sweden; and it is lowest in
Hungary, East Germany and the Czech Republic (results similar to those shown in
Tables 6 and 7). Column 1 also shows that men are unhappier; those living in rural
areas are especially happy with the country they live in; and more young children
raises happiness.
13 The weighted mean responses were highest for the following two categories: unemploy-
ment (30.7%) and rising prices (23.8%).
Table 8
Respondents’ happiness living in own country and likelihood
of working or moving abroad (2005/2006)
Thinking of
Happy living in Move next
Work abroad working Move if unemployed
this country five years
abroad
Ordered logit Dprobit Dprobit Dprobit Dprobit
East Germany -1.2213 (11.86) .0463 (3.05) .0564 (2.79) .0071 (0.89) .0155 (0.75) -.0272 (2.64)
Estonia .4366 (4.78) .0676 (4.66) .1277 (6.79) .0424 (4.62) .1015 (5.25) .0166 (1.70)
Hungary -2.1661 (25.41) .0406 (3.11) .0645 (3.75) .0078 (1.10) .1007 (5.18) .0133 (1.38)
Latvia -.0667 (0.76) .0491 (3.70) .1524 (8.19) .0387 (4.44) .2195 (10.07) .0943 (7.98)
Lithuania -.5185 (6.02) .0592 (4.18) .1582 (8.17) .0550 (5.53) .2506 (10.85) .1084 (8.63)
Poland -.5193 (6.09) .0505 (3.73) .1322 (7.10) .0444 (4.73) .1462 (7.09) .0412 (3.86)
Slovakia -.9379 (11.33) .0733 (5.08) .1091 (6.28) .0234 (2.83) .0680 (3.68) -.0041 (0.45)
Slovenia -.2841 (3.37) .0682 (4.77) .0157 (1.05) .0011 (0.18) .0139 (0.82) -.0263 (3.21)
Czech Republic -1.0654 (12.78) -.0330 (4.08)
Austria .0954 (1.09) -.0340 (4.23)
Belgium -.2103 (2.44) .0049 (0.54)
Denmark 1.6098 (13.62) -.0187 (2.22)
Finland .6086 (6.44) -.0354 (4.47)
France .0162 (0.19) -.0384 (4.74)
Greece .2895 (3.20) -.0395 (5.02)
Ireland .3072 (3.40) -.0169 (1.96)
Italy -.7628 (9.05) -.0117 (1.31)
Luxembourg .8058 (6.32) -.0103 (1.15)
Malta .1180 (1.04) -.0163 (1.48)
Netherlands -.1022 (1.18) .0882 (5.74)
Portugal -.8642 (10.11) -.0157 (1.83)
Spain .1353 (1.53) -.0094 (1.00)
Sweden 1.2625 (11.37) -.0395 (5.06)
West Germany -.8035 (9.54) -.0177 (2.07)
Age .0150 (10.17) -.0006 (3.32) -.0056 (14.59) -.0009 (6.42) -.0017 (5.22) -.0007 (4.59)
Male -.1559 (5.33) .0202 (5.54) .0338 (4.90) .0058 (2.39) .0236 (3.70) .0157 (4.84)
15-19 years schooling .0634 (1.58) .0025 (0.39) -.0027 (0.19) .0086 (1.54) .0122 (1.18) .0047 (1.01)
≥20 years schooling .0577 (1.21) .0036 (0.50) .0158 (0.98) .0185 (2.47) .0356 (2.83) .0189 (3.39)
Student .3261 (3.74) -.0119 (1.29) .1409 (4.50) .0049 (0.58) .0429 (1.82) .0322 (2.96)
Unemployed -.4016 (5.63) .0066 (0.68) .0977 (4.21) -.0012 (0.23) .0001 (0.01) -.0031 (0.39)
Retired .0809 (1.34) -.0142 (1.57) -.0104 (0.51) -.0119 (1.91) .0240 (1.32) .0106 (1.47)
Farmer -.0897 (0.63) .0050 (0.29) .0220 (0.57) -.0074 (0.71) -.0005 (0.02) -.0119 (0.78)
Professional -.1964 (1.65) .0051 (0.32) .0738 (1.91) .0027 (0.26) .0198 (0.60) .0193 (1.40)
Shop owner -.0861 (0.92) -.0083 (0.69) .0381 (1.23) -.0084 (1.12) .0402 (1.38) .0217 (1.87)
Business proprietor -.0126 (0.11) .0093 (0.66) .1354 (3.69) -.0087 (1.21) .0269 (0.97) .0067 (0.51)
Employed professional .0968 (0.97) .0013 (0.12) .0899 (3.36) -.0058 (1.05) .0123 (0.61) .0069 (0.67)
General management .3257 (2.35) .0093 (0.49) .0895 (2.01) -.0016 (0.15) .0017 (0.05) .0021 (0.14)
Middle management .1654 (2.21) -.0016 (0.17) .0565 (2.35) -.0124 (2.71) -.0018 (0.10) .0005 (0.06)
Employed - desk .0645 (0.95) -.0071 (0.81) .0794 (3.47) -.0114 (2.50) .0148 (0.77) .0061 (0.77)
Travelling salesman .0446 (0.46) .0025 (0.21) .1063 (3.43) -.0108 (2.02) .0144 (0.64) .0050 (0.48)
Employed not at a desk .0865 (1.20) -.0092 (1.07) .0762 (3.21) -.0101 (2.16) -.0001 (0.01) .0007 (0.09)
Supervisor .2081 (1.31) -.0054 (0.33) .1689 (3.32) .0123 (0.77) .0050 (0.12) -.0070 (0.40)
Skilled manual -.1279 (1.93) .0017 (0.19) .0855 (3.93) -.0074 (1.50) .0089 (0.50) .0067 (0.84)
Unskilled manual .0852 (1.03) -.0067 (0.71) .0450 (1.83) -.0117 (2.06) .0073 (0.33) .0025 (0.26)
Married .0529 (1.01) -.0083 (1.38) -.0240 (2.01) -.0093 (2.25) .0093 (0.80) .0061 (1.08)
Remarried -.2888 (2.76) .0032 (0.27) -.0342 (1.52) .0025 (0.31) .0147 (0.66) .0313 (2.51)
Living as married -.0322 (0.52) .0093 (1.34) -.0047 (0.39) -.0001 (0.02) .0331 (2.36) .0051 (0.77)
Previously living as married -.3397 (4.50) .0004 (0.05) .0295 (1.61) .0145 (2.34) .0413 (2.29) .0052 (0.63)
Divorced -.3627 (5.26) .0053 (0.64) .0240 (1.41) .0032 (0.56) .0449 (2.65) .0186 (2.21)
Separated -.4579 (4.21) .0135 (0.90) -.0142 (0.54) .0151 (1.40) .0033 (0.13) .0052 (0.42)
Widowed -.1128 (1.59) -.0067 (0.77) .0071 (0.35) -.0055 (0.82) -.0138 (0.88) -.0102 (1.28)
Owns a computer .0440 (1.30)
Owns a car .2387 (7.01)
Private renter .0161 (3.79) .0027 (0.26) .0104 (2.22)
Rural area .1656 (4.65) -.0046 (1.15) -.0114 (1.39) -.0168 (6.24) -.0205 (2.82) -.0116 (3.11)
Small or middle sized town .1349 (3.92) -.0046 (1.19) .0170 (2.16) -.0070 (2.92) -.0209 (3.00) -.0126 (3.42)
# children aged <10 years .0776 (3.46) -.0025 (0.98) -.0032 (0.65) -.0036 (1.87) -.0019 (0.38) -.0037 (1.51)
cut1 -3.2865
cut2 -1.747
cut3 0.5242
East Europe only No Yes Yes Yes Yes No
N 24,539 8,544 8,268 8,685 8,189 23,317
2
Pseudo R 0.088 0.0795 0.252 0.2173 0.1295 0.0789
Source (Table 8): columns 1 to 3 are from Eurobarometer #65.1, “The Future of Europe, Consumer Protec-
tion in Transborder Purchases, Family Planning, and Opinions and Experiences in Transborder Pur-
chases”, February-March 2006, ICPSR 20321. Columns 4 and 5 are from Eurobarometer #64.1, “Mobility,
Food Risk, Smoking, AIDS Prevention, and Medical Errors”, September-October 2005, ICPSR 4641.
Notes: excluded categories are UK; responsible for shopping; single and never lived with a partner; and less than sixteen years’
schooling. The Czech Republic is excluded from columns 2 to 5. T-statistics in parentheses. Question 1 (column 1): Please tell
me if you agree or disagree with the statement, “You are happy living in your country.” 4 = totally agree; 3 = tend to agree; 2 =
tend to disagree; and 1 = totally disagree. Question 2 (columns 2 and 3): Have you ever considered living in another member
state in order to work? 1 = yes, I have already done it; 2 = yes, I am thinking about it, but I haven’t decided yet; 3 = yes, I have al-
ready thought about it but gave up the idea; and 4 = no, I have never thought about it. In this question, column 2 dependent vari-
able set to one if one and zero otherwise. Column 3 dependent variable set to one if thinking about working abroad, and zero
otherwise. Those who have previously worked abroad are set to missing. Question 3 (Column 4): Do you think that in the next
five years you are likely to move to another country in the European Union? Question 4 (Column 5): If you were unemployed and
had difficulties finding a new job, would you be ready to move to another country to find one?
Columns 2 and 3 estimate the likelihood that individuals from East Germany and
the EU10, in comparison to the Czech Republic, which is the excluded country,
have worked abroad, or are thinking of working abroad (but are undecided), respec-
tively. As might be expected, those more likely to have worked abroad are males,
the more highly educated, managers, and those living in urban areas. Column 4 es-
timates the likelihood that an individual will move to another EU country in the next
five years.14 This was particularly high for Lithuania and Poland. Column 5 (Eastern
Europe only, excluding the Czech Republic) and column 6 (countries in Eastern and
Western Europe) estimate the likelihood that workers will move to another country
if they become unemployed. Column 5 shows that the probability is highest in
Lithuania and Latvia; these countries have seen the highest proportion of their popu-
lation move to the UK. The probability of moving abroad if faced with unemploy-
ment is higher for men and declines with age. The proportion that said they would
move abroad if unemployed was especially high in Lithuania (20%), Latvia (19%),
Poland (12%), Estonia (10%) and Hungary (8%). Finland had the lowest proportion
who would be prepared to move abroad if unemployed (2%).
These findings are consistent with those reported by Blanchflower and Shadforth
(2009). The authors used data from the Candidate Eurobarometer 2002.1 (ZA#4153)
conducted between March and April 2002 to examine Eastern Europeans’ intentions
to live and work in the EU in the five years following the survey. They found that
willingness to move was higher for males, the young, the unmarried or divorced, the
most educated, and professionals and business proprietors. It was especially high for
residents of Latvia, Poland and Slovakia. Unsurprisingly, EU10 workers appear to
have come to the UK in order to improve their lives. A significantly high proportion
said they would be prepared to move abroad in the future.
Overall, evidence suggests that Eastern Europeans are unhappy with their lives
and their home countries more generally. Workers are dissatisfied with their sala-
ries, their working hours, and their work/life balance. They are concerned about
14 Respondents were also asked if they would consider moving to another country or region,
but we only model the question that asked if respondents would move to another EU
country.
the availability of good jobs and are insecure about their current jobs. All of these
“push factors” present genuine reasons to move abroad and are consistent with the
literature which focuses on the economic factors determining migration. Put sim-
ply, the literature states that an individual will compare the income benefits from
migration, with the economic and social costs of moving. If the benefits outweigh
the costs, the individual may choose to migrate. The net benefit of migration is the
expected income differential between the destination country and the country of
origin. This is, in turn, determined by the relative probability of getting a job in
the destination country as captured by the difference in unemployment (or em-
ployment) rates in the two countries.
Pollard et al (2008) interviewed 370 Poles who had returned home after work-
ing in the UK. They find that the desire to learn English is a significant attraction,
and it is a way of “broadening horizons” for many young Poles coming to the UK.
They also find that Poles are registered on part-time courses as a way of learning
new skills which are often not available at home. They find that:
“… the strength of the UK economy has also acted as a “pull” factor for
many migrants. High levels of spending, low unemployment, and high de-
mand for labor especially in sectors such as construction have acted as
draws for many post enlargement migrants. The strength of the British cur-
rency in recent years has acted as a particular pull, allowing earnings and
savings from the UK to go even further when spent in migrants’ home coun-
tries.” Pollard et al (2008), p. 43.
Working in the UK is thus an opportunity for many to save money. The Center
for International Relations in Warsaw conducted a survey of high-skilled Poles
working in Great Britain. They found that 65% were saving some of the money
they were earning, and of this group 60% were sending money home to Poland.
5 What Impact Have EU10 Workers Had

on the UK Labor Market?
The empirical literature from around the world suggests there is little or no evi-
dence that immigrants have had a major impact on native labor market outcomes
such as wages and unemployment. Recent work on the UK labor market by a
number of authors is consistent with this view. Studies have shown that immi-
grants have had little impact on wages (see Manacorda et al 2006, and Dustman et
al, 2005). Where there has been (small) impact, evidence suggests that it has been
on the wages of the least skilled. For example, Blanchflower and Shadforth (2009)
find that between 2005 and 2006 there was a small negative relationship between
the change in the annual rate of wage inflation for those in elementary occupations
(defined in the LFS as SOC 9) and the change in the share of EU8 workers, across
regions, in the previous year (sourced from 2004 and 2005 WRS data). A similar
result was reported by Nickell and Saleheen (2008): they find a small negative re-
lationship, primarily in the semi-skilled and unskilled services sector.
There is no credible statistical evidence that arrivals from the EU10 had any
perceptible impact on unemployment. Gilpin et al (2006) found “no discernible
statistical evidence to suggest that EU8 migration has been a contributor to the
rise in claimant count unemployment in the UK” (p. 1, 2006). Lemos and Portes
(2008) recently updated Gilpin’s work and found a similar result (overall, and for
sub-groups). However, the increasing numbers of EU10 workers in the UK may
have raised the “fear” of unemployment. This tends to have a downward impact
on pay, especially in the non-union sector (Blanchflower, 1991). Blanchflower
and Shadforth (2009) find that the fear of unemployment is higher the longer the
job tenure; and it is lower for the more educated, those on indefinite contracts,
full-timers and those who work in the public sector. The fear of unemployment is
found to be highest in Eastern European countries and lowest in Norway and
France. The authors also model the impact of the fear of unemployment on earn-
ings (defined as the respondents’ net monthly income from their main paid job).
They found that the fear of unemployment lowers wages.
This confirms the findings in Blanchflower (1991), which showed that the prob-
ability of job loss appears to have a powerful effect upon earnings. Workers who
stated that they expected to be made redundant did not receive a compensating dif-
ferential and were paid, on average, approximately 8% less, ceteris paribus. One
possibility is that bad workers have a relatively high fear of redundancy because of
their poor performance. However, Blanchflower (1991) argues that the fear of un-
employment itself, and not poor worker quality, is the explanation for the significant
coefficient on the redundancy dummy. One possible way around this problem is to
exploit the fact that when plants close, both good and bad workers lose their jobs.
Thus, as a check, the “redundancy expected” variable for the UK was replaced with
one relating to the expectation of plant closure. This also lowered pay by eight% and
seems to support the idea that fear of unemployment is primarily not a proxy for
worker quality. Campbell et al (2007) examine whether individuals’ reported per-
ceptions of unemployment risk are reliable indicators of the probability of becoming
unemployed. The authors find that an individual’s fear of unemployment is in-
creased by personal experience of unemployment and by encountering unemploy-
ment through the experience of close friends. They also find that the probability of
becoming unemployed is higher for those with a greater fear of unemployment, i.e.
the fear of unemployment is rational because workers are able to reliably assess their
chances of becoming unemployed. Finally, the authors find that a high fear of un-
employment is associated with lower rates of wage growth for men, but it has no de-
tectable link with wage growth for women.
Figure 1
Unemployment expectations (3m average, advanced 12 months) vs. actual unemployment
UK
Unemployment expectations (LHS)
Survey balance Unemployment expectations, 1985-2008 average (LHS)

%
Unemployment expectations, 2004-2008 average (LHS)
60 Unemployment rate (RHS) 14
50 12
40
10
30
8
20
6
10
4
0
-10 2
-20 0
Jan-85
Jan-86
Jan-87
Jan-88
Jan-89
Jan-90
Jan-91
Jan-92
Jan-93
Jan-94
Jan-95
Jan-96
Jan-97
Jan-98
Jan-99
Jan-00
Jan-01
Jan-02
Jan-03
Jan-04
Jan-05
Jan-06
Jan-07
Jan-08
Jan-09
Jan-10
Source: LFS, EC Consumer Survey.
EU15
Survey balance Unemployment expectations, 1985-2008 average (LHS) %
50
10
40
8
30
6
20
4
10
0 2
-10 0
Jan-85
Jan-86
Jan-87
Jan-88
Jan-89
Jan-90
Jan-91
Jan-92
Jan-93
Jan-94
Jan-95
Jan-96
Jan-97
Jan-98
Jan-99
Jan-00
Jan-01
Jan-02
Jan-03
Jan-04
Jan-05
Jan-06
Jan-07
Jan-08
Jan-09
Jan-10
Source: EC Consumer Survey, Thomson Datastream.

Ireland
Survey balance %
50 16
40
14
30
12
20
10
10
8
0
6
-10
4
-20
-30 2
-40 0
Jan-85
Jan-86
Jan-87
Jan-88
Jan-89
Jan-90
Jan-91
Jan-92
Jan-93
Jan-94
Jan-95
Jan-96
Jan-97
Jan-98
Jan-99
Jan-00
Jan-01
Jan-02
Jan-03
Jan-04
Jan-05
Jan-06
Jan-07
Jan-08
Jan-09
Jan-10
Sweden
Survey balance %
40
10
30
20 8
10
6
0
-10 4
-20
2
-30
-40 0
Jan-85
Jan-86
Jan-87
Jan-88
Jan-89
Jan-90
Jan-91
Jan-92
Jan-93
Jan-94
Jan-95
Jan-96
Jan-97
Jan-98
Jan-99
Jan-00
Jan-01
Jan-02
Jan-03
Jan-04
Jan-05
Jan-06
Jan-07
Jan-08
Jan-09
Jan-10

A monthly consumer survey conducted by the European Commission in all

EU and EU-applicant countries is also consistent with the view that the fear of
unemployment in the UK has risen (and, according to this survey, it has been
above its long run average since around 2005).15 In this monthly survey, con-
sumers are asked: “How do you expect the number of people unemployed in this
country to change over the next twelve months?”16 The answers obtained from
the survey are aggregated into a survey “balance” (these are constructed as the
difference between the proportion giving positive and negative responses). The
EC also calculates average responses in the EU and euro-areas as a whole, on
the basis of national results.
Figure 1 plots the national survey balances for unemployment expectations

(as a three-month average, advanced twelve months) against the actual unem-
ployment rate in the UK, the EU15, Ireland and Sweden respectively. The fear
of unemployment and actual unemployment has risen in both the UK and Ire-
land since around 2004, and both countries have had large influxes of workers
from the accession countries. The fear of unemployment declined in the EU15
between 2004 and the end of 2007.17 Interestingly, the survey balances over this
time period fell in Austria, Belgium, Denmark, Finland, France, Germany, Lux-
embourg, the Netherlands, Portugal and Sweden; and survey balances for
Greece, Italy and Spain were essentially flat. Since the beginning of 2008, the
fear of unemployment has picked up, and expectations have picked up markedly
for 2009, for the EU15 as a whole and for every country within the EU15. This
is despite the fact that unemployment had fallen in every EU15 country except
Spain and Portugal between 2004 and the end of 2007.
15 See http://ec.europa.eu/economy_finance/db_indicators/surveys9185_en.htm
16 Consumers are asked to choose an answer that best describes what they think will hap-
pen to unemployment, i.e. that it will: (i) increase sharply; (ii) increase slightly; (iii) remain
the same; (iv) fall slightly; (v) fall sharply; or (vi) don’t know.
17 The EU15 series for unemployment and unemployment expectations are weighted accord-
ing to the population of each country for each year. Due to the availability of the data, the
EU15 series for unemployment expectations includes:
Jan85-Mar86 -- UK, Belgium, Denmark, Germany, Ireland, Greece, France, Netherlands
and Italy
Apr86-Aug87 -- as above, plus Portugal and Spain
Sep87-July95 -- as above, plus Finland
Aug95-Oct01 -- as above, plus Sweden and Austria
Nov01 onwards -- as above, plus Luxembourg
And the EU15 unemployment rate includes:
Jan85-Dec94 -- UK, Spain, France, Italy, Netherlands, Austria, Portugal, Finland and
Sweden
Jan95-Dec96 -- as above, plus Belgium, Denmark, Germany and Ireland
Jan97-Mar98 -- as above, plus Luxembourg
Apr98-Dec06 -- as above, plus Greece
The fear of unemployment survey balances declined between 2004 and the end
of 2007 in all EU10 countries except Hungary, which saw a significant increase.
Throughout 2008 and looking ahead into 2009, unemployment expectations have
picked up across the EU10 countries, with particularly sharp rise in Latvia,
Lithuania and Estonia. The fear of unemployment is especially high in Hungary.
Of particular interest is unemployment expectations and actual unemployment

in Ireland, as it is the only other major EU country (after the UK) which has ex-
perienced a big increase in migration from EU10 countries. For example, Ireland’s
population increased by 313,000 (8.1%) between 2002 and 2006, and migration
accounted for nearly 70% (213,000) of this increase. Those migrating to Ireland
mostly came from Poland (more than 60,000) and Lithuania (more than 22,000);
the rest came from other EU25 countries (excluding Great Britain and Northern
Ireland).18 According to the 2006 Census (Table 29A), 129,000 people whose
birthplace was in Eastern Europe were living in the Irish Republic.19 These num-
bers are dramatically higher than they were in the 2002 Irish Census, when only
approximately 2,000 Poles and Lithuanians were living in Ireland.
Figure 2
UK earnings growth
Annual % grow th rate Annual % grow th rate
5.0 7.0
4.5
6.0
4.0
3.5 5.0
3.0
4.0
2.5
3.0
2.0
1.5 AEI: Whole economy w age settlements (LHS) 2.0

AEI: 10-year w hole economy mean (LHS)
1.0
LFS hourly pay (in main job only) (RHS)
1.0
0.5
0.0 0.0
Jul-07
Nov-07
Jul-08
Jul-06
Mar-07
Mar-08
Jul-02
Nov-02
Jul-03
Nov-03
Mar-04
Jul-04
Nov-04
Mar-05
Jul-05
Nov-05
Mar-06
Nov-06
Mar-99
Jul-99
Nov-99
Mar-00
Jul-00
Nov-00
Mar-01
Jul-01
Nov-01
Mar-02
Mar-03
Source: Bank of England, Labour Force Survey, Income Data Services, Industrial Relations Services,
Labour Research Department.
18 www.cso.ie/census/documents/PDR%202006%20Commentary.pdf
19 www.cso.ie/census/documents/Final%20Principal%20Demographic%20Results%202006.pdf
Interestingly, as the number of Eastern Europeans in Ireland rose, so did the

fear of unemployment (as happened in the UK). However, unemployment in Ire-
land did not rise: ILO unemployment remained steady at 4.4% between 2002
and 2006.20 Consistent with a rise in the fear of unemployment, average earnings
growth in Ireland has fallen from 6.4% to 3.1% since 2003 (Blanchflower and
Shadforth, 2009). However, unemployment rose to an average of 4.6% in 2007
and to 5.4% in May 2008. The fear of unemployment rose sharply during this
time (the survey balance rose from 39.7 in October 2007 to 55.0 in May 2008).
Despite historically low levels of unemployment in the UK, wage pressure

has been contained. Figure 2 provides evidence of this, and from a number of
sources (including the Labour Force Survey).
Overall, evidence suggests that the flow of workers from EU10 countries has
had little or no impact on unemployment, but it has had some direct effect on the
relative wages of the least skilled. But the main impact of the flow of workers
from the EU10 countries has been the rise in the fear of unemployment, and this in
turn has contained wage growth.
6 Will the Flow of EU10 Workers

to the UK Continue in the Future?
It is uncertain as to whether the flow of EU10 workers to the UK will continue in

the future. There are valid arguments for both a reduction and an increase in the
flow.
Economic activity and the outlook for the global economy suffered a marked
downturn at the end of the third quarter and start of the fourth quarter of 2008.
The global financial crisis that began in August 2007 intensified around the time
of the collapse of the investment bank Lehman Brothers (September 2008). And
so began unprecedented and large-scale government rescue packages for banking
and financial systems across the world. In its latest projections (November 2008),
the OECD forecast a “protracted economic slowdown” in the US, Japan and Euro
area. The OECD expect the 2009 GDP to contract by 0.9% in the US, 0.6% in the
euro area and 0.1% in Japan.
According to Eurostat, GDP growth in the EU15 contracted by 0.2% in the

third quarter of 2008, following a similar fall in the second quarter. Thus, the
EU15 (as a whole) moved into an official recession in 2008. And within this, na-
20 www.cso.ie/statistics/sasunemprates.htm
tional statistics show that Germany, Spain, Ireland and Italy also moved into an
official recession in 2008 (France fared slightly better, reporting a 0.1% increase
in GDP in the third quarter, after a 0.3% fall in the second quarter). However, it is
feared that the UK may suffer a worse recession than other EU countries: its
economy contracted by 0.5% in the third quarter of 2008 (following zero growth
in the second quarter), and the UK is widely expected to suffer an even bigger fall
in output in the fourth quarter of 2008 and continue to shrink in 2009.
Based on these figures, it may be reasonable to assume that the flow of EU10
workers to the UK will slow while the UK economy is in recession. But it is the
relative attractiveness of the UK that is an important factor determining the flow
of foreign workers from EU10 countries; and the economic slowdown is a global
one. This means that firstly, countries likely to “compete” with the UK to attract
EU10 foreign workers (i.e. other countries in the EU15) are also suffering from
the economic slowdown. Many other EU countries are reducing their restrictions
on workers from Eastern Europe (it is a legal requirement for them to do so), so
workers will have more destinations to choose from in the future. But secondly,
and perhaps a more important factor determining the flow of foreign workers to
the UK, is the health of the EU10 economies – and hence the relative attractive-
ness of the UK compared to the EU10 economies.
Of the EU10 economies, Hungary appears to be faring particularly badly. It has

a large current account deficit and a severe lack of market confidence in its capital
markets. At the time of writing (November 2008), it is in the process of receiving
one of the biggest rescue packages in global history – a £25 billion package from
the IMF, World Bank and European Union.
Bulgaria, Romania and the Baltic states (Estonia, Latvia and Lithuania) are also
seen as very vulnerable in the global financial crisis. These countries face similar
problems to Hungary, with heavy foreign ownership of banks (this is helping to
transmit the banking problems of Western Europe further east), large current ac-
count deficits, and heavy external indebtedness. Latvia and Estonia have been in
recession since the start of 2008; Estonia’s economy contracted by 3.3% in the
third quarter of 2008; and Latvia may also need financial assistance after its gov-
ernment was forced to rescue the country’s second largest bank (November 2008).
Lithuania also looks likely to go into recession in 2009.
The Czech Republic, Poland and, to a lesser extent, Slovakia are in a stronger
economic position than other Eastern European economies, but they are also vul-
nerable to contagion (especially a steep decline in demand from Western Europe).
And Slovakia, with its heavy dependence on a single industry (cars), is particu-
larly vulnerable.
Overall, the IMF (October 2008) is forecasting a sharp slowdown in economic

growth in the Central and Eastern European (CEE) economies, with overall
growth of 5% in 2008 and 3.5% in 2009.
Therefore, the UK may still appear relatively more attractive for the residents
of some EU10 countries. The flow of workers may continue, although the mix of
workers from different EU10 countries is most likely to change. But even if the
flow of workers does slow significantly, it is reasonable to assume the flow will
pick up again in the future as economies recover. The economies of the EU15 are
expected to have started to recover by 2010.
As Table 1 shows, the number of new registrations under the WRS slowed in
the first quarter of 2008. At this time, the UK economy had already started to
slow. House prices were falling; oil and commodity prices were rising; unem-
ployment had also started to rise; and the value of sterling had fallen against the
euro, zloty and other EU10 currencies. All these factors make the UK a less attrac-
tive place to work – relative rates of return fall. But other economies were also
starting to slow; labor markets were loosening elsewhere; and the rise in oil and
commodity prices was an international phenomenon. Thus, the relative attractive-
ness of the UK may not have changed significantly.
A recent special survey conducted by the Bank of England’s Agents in spring

2008 finds that firms do not expect to reduce their use of migrant labor in the fu-
ture. However, as the labor market loosened, the Agents report that more UK na-
tionals are applying for low-end jobs. There is some evidence that Polish workers
are less prepared than previously to do unskilled work at the National Minimum
Wage. Also, the more recent Polish arrivals tend to be older and less qualified
than the earlier tranches, and they are less likely to speak English. Although there
is no obvious reduction in the supply of Polish workers, there is some evidence
that employment agencies are switching to workers from other countries. The mix
of nationalities may change in the future: relatively small numbers of workers
have come to the UK from the Czech Republic and Hungary. Given the poor
health of the economies of Hungary, Bulgaria, Romania and the Baltic states,
workers in these countries may have a bigger incentive to leave their home coun-
tries. On top if this, UK wages are still relatively higher than wages in EU10
countries. Hence, even workers from economies which have fared better, such as
the Czech Republic and Slovakia, may still find the UK a relatively more attrac-
tive place to live and work.
The UK also continues to be an attractive destination because English is an in-

ternational language. And the development of network effects has made it easier
for foreign workers to obtain work in the UK. Pedersen et al (2004) find that net-
work effects (measured as the coefficient of the stock of immigrants already resi-
dent in a country and of the same nationality as a new arrival) have a large posi-
tive effect on immigration flows. The authors find that linguistic closeness, former
colonial ties and current business ties have a significant impact on migration
flows; geographic distance has a negative impact, suggesting the costs of migra-
tion are also important. They conclude that networks play an important role in ex-
plaining current immigration flows.
Pollard et al (2008) note a number of factors which have made it easier for
Poles to move to and from the UK. For example, it was possible to fly from
twenty-two British airports to ten Polish cities by December 2007, and in that
month alone, there were almost 385,000 passengers between the two countries. In
2007, ten million air passengers traveled between the UK and EU10 countries: a
three-fold increase on pre-enlargement air traffic. The authors also note growth in
sales of EU8 (primarily Polish) goods and services in the UK. For example, sev-
eral hundred Polish delicatessens have been established throughout the country
over the last four years. And previously established suppliers have been catering
to the increased demand for Polish goods. In July 2007, Tesco announced that it
was doubling both its range of Polish products and the number of stores stocking
them. Tesco now sells more Polish food in its UK stores than it does in Poland
(where it has 280 stores). Similarly, leading brands of Polish beer (Lech and Ty-
skie) were not widely available in the UK prior to 2004. However since then, the
owner of these two brands (SABMiller) has said that their combined UK sales ex-
ceeded forty-four million pints in one year.
From interviews of Poles living in London, Pollard et al (2008) also find that
those who settle in the UK in the long term, and especially once their command of
the English language has improved, often move into jobs more suited to their
skills. The authors provide evidence that enterprising young people from acces-
sion countries are attracted to the UK in order to set up new businesses. The Brit-
ish-Polish Chamber of Commerce estimates on its website that there are currently
forty thousand Polish entrepreneurs who have set up businesses in the UK.21
Sachdev and Harries (2006) interviewed a number of foreign workers from

EU8 countries. They find the three main reasons for coming to the UK are: (i) to
learn English; (ii) to earn money; and (iii) to “gain a new experience”. Most re-
spondents are uncertain how long they would stay in England. Respondents are
more likely to stay long-term if they have settled down by forming relationships
and having families, and/or if they had invested in their career. Respondents are
also more likely to stay if they have undertaken training courses, or reached a cer-
tain position in their chosen profession. Respondents are more likely to leave if
they have been unable to adjust to life in England, or if they are planning on fin-
ishing full-time education and then returning home, or if they are seeking a job in
21 The first sentence on the BPCC website says: “UK-Polish trade is booming; there are
400,000 Poles living in London and the South East, and 40,000 Polish entrepreneurs have
set up businesses in the UK.” See: http://bpcc.org.pl/en/content/view/192
a multi-national company where they could use their existing qualifications and
competency in English.
It is hard to calculate the overall impact on the UK of the net flow of workers
from the accession countries. Other factors, such as the migratory flows of UK na-
tionals, also need to be taken into account.
7 Conclusions
In summary, workers from the EU10 countries were attracted to the UK between
2004 and 2007 because of the favorable macroeconomic climate (low unemploy-
ment), high standard of living and a general dissatisfaction with life in their home
countries. We have found a great deal of evidence from a large number of data
sources suggesting that residents of EU10 countries are unhappy with their lives
and the country they live in. They are dissatisfied with their jobs and would ex-
perience difficulties finding a good new job. Job insecurity is high and the
work/life balance is poor. Relatively high proportions of Eastern Europeans ex-
press a desire to move abroad, and many would move abroad if they were made
unemployed. Eastern Europeans’ expectations for the future of their domestic
economies and their personal situations remain low; however, they have improved
since 2004.
It may be reasonable to assume that the flow of EU10 workers to the UK will
slow while the UK economy is in recession. But it is the relative attractiveness of
the UK that is an important factor determining the flow of foreign workers from
EU10 countries: the economic slowdown is a global one.
Firstly, this means that countries likely to “compete” with the UK to attract
EU10 foreign workers (i.e. other countries in the EU15) are also suffering from
the economic slowdown. Many other EU countries are reducing their restrictions
on workers from Eastern Europe (it is a legal requirement for them to do so), so
workers will have more destinations to choose from in the future. However, net-
work effects have been established in the UK (and particularly for Poles), and
these tend to have a large positive impact on migration flows.
Secondly, and perhaps a more important factor, is the health of the EU10
economies, and hence, the relative attractiveness of the UK compared to the EU10
economies. There are large differences, with some faring well (such as Poland, the
Czech Republic and Slovakia), and some faring badly (Hungary, Bulgaria, Roma-
nia and the Baltic states). Therefore, the mix of nationalities of foreign workers
may change in the future. A large number of workers have come to the UK from
Poland, but relatively small numbers have come from the Czech Republic and
Hungary. Given the poor health of the economies of Hungary, Bulgaria, Romania
and the Baltic states, workers in these countries may have a bigger incentive to
come to the UK. On top if this, UK wages are still relatively higher than wages in
EU10 countries. Hence, even workers from economies which have fared better,
such as the Czech Republic and Slovakia, may still find the UK a relatively more
attractive place to live and work.
The fact the UK opened its borders to a flow of highly skilled, motivated, edu-
cated, low cost and mobile workforce upon enlargement of the EU was a stroke of
genius, and the UK government should be given credit. Indeed, the post-EU
enlargement arrivals to the UK from the EU10 countries have a higher probability
of being in work than migrants from non-EU10 countries who arrived in the UK
both before and after EU enlargement. Contrary to the fears of some, the post-EU
enlargement arrivals to the UK from the EU10 countries appear to have had no
difficulty assimilating into the native population, and they have been welcomed
with open arms. By the time countries such as Germany and Austria began to con-
sider opening their borders, it was too late. Networks had developed and large
numbers of Poles, Slovakians, Lithuanians and Latvians were already living and
working in the UK.
There has been some deterioration in the availability of jobs in the UK as the
economy moves into recession. However, the UK is an attractive place to live and
work for Eastern European workers. This is aided by highly developed networks.
We argue that rather than dissipate, flows of Eastern European workers to the UK
could remain strong well into the future. The enlargement of the European Union
has benefited the UK enormously.
References
Accession Monitoring Report, May 2004-June 2008, Home Office, UK Border Agency.
www.ukba.homeoffice.gov.uk/sitecontent/documents/aboutus/reports/accession_monitori
ng_report/
Blanchflower, D. G. (1991), Fear, unemployment and pay flexibility, in: The Economic Jour-
nal, March, 101 (406), 483-496.
Blanchflower, D. G. (2001), Unemployment, well-being and wage curves in Eastern and Cen-
tral Europe, in: Journal of the Japanese and International Economies 15 (4), 364-402.
Blanchflower, D. G. (2004), Self-employment: more may not be better, in: Swedish Eco-
nomic Policy Review 11 (2), 15-74.
Blanchflower, D. G. (2008), International evidence on well-being, in: A. Krueger (ed.) Na-
tional Time Accounting and Subjective Well-being, forthcoming.
Blanchflower, D. G./A. Bryson (2003), Changes over time in union relative wage effects in
the UK and the US revisited, in: J. T. Addison/C. Schnabel (eds.), International Handbook
of Trade Unions. Cheltenham et al.
Blanchflower, D. G./A. Bryson (2008), The wage impact of trade unions in the UK public and
private sectors, in: Economica, forthcoming.
Blanchflower, D. G./R. B. Freeman (1997), The attitudinal legacy of Communist labor rela-
tions, in: Industrial and Labor Relations Review 50 (3), 483-459.
Blanchflower, D. G./A. J. Oswald/A. Stutzer (2001), Latent entrepreneurship across nations,
in: European Economic Review 45, 680-691
Blanchflower, D. G./A. J. Oswald (2004), Well-being over time in Britain and the United
States, in: Journal of Public Economics 88 (7-8), 1359-1386.
Blanchflower, D. G./A. J. Oswald (2008), Is well-being U-shaped over the life cycle?, in: So-
cial Science and Medicine 66 (8), 1733-1749.
Blanchflower, D. G./C. Shadforth (2009), Fear, Unemployment and Migration, in: The Eco-
nomic Journal 119 (535), F136-F182.
Campbell, D./A. Carruth/A. Dickerson/F. Green (2007), Job insecurity and wages, in: The
Economic Journal 177 (518), 544-566.
Clark, A. E./A. J. Oswald/P. Warr (1996), Is job satisfaction U-shaped in age?, in: Journal of
Occupational and Organizational Psychology 69, 57-81.
Constant A./K. F. Zimmermann (2007), Circular migration: counts of exits and years away
from the host country’. IZA Discussion Paper No. 2999, Bonn.
DaVanzo, J. (1983), Repeat migration in the United States: who moves back and who moves
on?, in: Review of Economics and Statistics 65 (4), 552-559.
Di Tella, R./R. MacCulloch (2005), Partisan social happiness, in: Review of Economic Stud-
ies 72 (2), 367-393, April.
Drinkwater S./J. Eade/M. Garapich (2006), Poles apart? EU enlargement and the labour mar-
ket outcomes of immigrants in the UK. IZA Discussion Paper No. 2410, Bonn.
Dustmann, C. (1994), Return intentions of migrants: theory and evidence. CEPR Discussion
Paper No. 906.
Dustmann, C. (1996), Return migration: the European experience, in: Economic Policy 22,
215-250.
Dustmann, C. (2003), Return migration, wage differentials and the optimal migration dura-
tion, in: European Economic Review 47 (2), 353-369.
Dustmann, C./F. Fabbri/I. Preston (2005), The impact of immigration on the British Labour
Market, in: The Economic Journal 115 (507), 324-341.
Dustmann, C./O. Kirchkamp (2002), The optimal migration duration and activity choice after
re-migration, in: Journal of Development Economics 67 (2), 351-372.
Dustmann, C./Y. Weiss (2007), Return migration: theory and empirical evidence from the
UK, in: British Journal of Industrial Relations 45 (2), 236–256.
Fife Research Coordination Group (2008), Migrant Workers in Fife Survey 2007.
Gilpin, N./M. Henty/S. Lemos/J. Portes/C. Bullen (2006), The impact of free movement of
workers from Central and Eastern Europe on the UK labour market. Department for Work
and Pensions, Working Paper No. 29.
Home Office (2008), Bulgarian and Romanian Accession Statistics, April-June 2008. UK
Border Agency.
www.ukba.homeoffice.gov.uk/sitecontent/documents/aboutus/reports/bulgarianromanian/
LaLonde, R./R. Topel (1997), The economic impact of international migration and the eco-
nomic performance of migrants, in: M. R. Rosenzweig/O. Stark (eds.), Handbook of Popu-
lation and Family Economics, Amsterdam.
Lemos, S./J. Portes (2008), The impact of migration from the new European Union Member
States on native workers. Department for Work and Pensions Working Paper No. 51.
Manacorda, M./A. Manning/J. Wadsworth (2006), The impact of immigration on the structure
of male wages: theory and evidence from Britain. IZA Discussion Paper No. 2352, Bonn.
Nickell, S. J./J. Saleheen (2008), The impact of immigration on occupational wages: British
evidence, Federal Reserve Bank Boston Working Paper 2008: 6.
OECD (2006), International Migration Outlook, Annual Report. Paris.
OECD (2008), Economic Outlook No. 84. Paris.
Office for National Statistics (2008), International Migration, Series MN No. 33, 2006 data.
Pedersen, P. J./M. Pytlikova/N. Smith (2004), Selection or network effects? Migration flows
into 27 OECD countries, 1990-2000. IZA Discussion Paper No 1104, Bonn.
Pollard, N./M. Latorre/D. Sriskandarajah (2008), Floodgates or turnstiles? Post-EU enlarge-
ment migration flows to (and from) the UK, Institute for Public Policy Research, London.
Sachdev, D./B. Harries (2006), Learning and skills planning and provision for migrants from
the accession states: an exploratory study. Learning and Skills Council.
https://www.lsneducation.org.uk/user/order.aspx?code=062435&src=XOWEB
Sanz de Galdeano, A./J. Turunen (2006), The Euro area wage curve, in: Economics Letters 92
(1), 93-98.
Yang, D. (2006), Why do migrants return to poor countries? Evidence from Philippine mi-
grants' responses to exchange rate shocks, in: The Review of Economics and Statistics 88
(4), 715-735.
Zaiceva, A. (2006), Reconciling the estimates of potential migration into the enlarged Euro-
pean Union. IZA Discussion Paper No. 2519, Bonn.
.
9. Impact of the Post-Accession Migration
on the Polish Labor Market
Paweł Kaczmarczyk
Marta Mioduszewska
Anna Żylicz
Post-2004 international mobility of Poles turned out to be one of the most spec-
tacular population movements in the modern European history. Changes noted
since the EU enlargement relate to scale, dynamics as well as structural features
(socio-demographic characteristics of migrants and their strategies) of migration
from Poland. Most of these changes are to be linked to selective opening of the
EU15 labour markets and introduction of so-called transitory arrangements.
Massive outflow of population – as observed in the Polish case – is supposed to
create a supply shock and significantly impact the labour market situation in
sending economy. However, the analysis provided shows that labour market ef-
fects of recent migration from Poland are moderate – the dynamics of the labour
market phenomena (employment, unemployment) as observed in last few years
were determined predominantly by business cycle related factors. Nevertheless,
long-term effects should be also considered. As suggested, mass migration may
lead to the outflow of ‘economically redundant’ population and through this foster
the modernization process of Polish economy.
Migration, DOI 10.1007/978-3-642-02242-5_9,
220 Paweł Kaczmarczyk, Marta Mioduszewska and Anna Żylicz
1 Introduction
EU enlargement was a turning point in post-war Polish migration history. Even

though the international mobility of Poles was relatively substantial prior to acces-
sion, it gained an extremely strong impetus following May 2004. Observed
changes relate not only to the scale and dynamics of the outflow but also to its
structural features. The opening of a few new, absorptive labor markets has led to
significant shifts in the geography of outflow, with the UK and Ireland as the most
remarkable examples of the new migration magnets. It has also led to changes in
the socio-demographic characteristics of migrants and their strategies. We argue
that selective introduction of so-called transitory arrangements by the EU15 mem-
ber states had a serious impact on the scope and forms of the post-2004 interna-
tional mobility of Poles.
The large outflow seen in the post-accession period raises the question of
socio-economic impacts of mobility. The most commonly discussed issue is the
effect of recent migration on the Polish labor market. We argue that the labor
market situation in Poland is predominantly determined by business cycle related
factors. Although both pre- and post-accession migration are large in absolute
terms, they have had only moderate short and medium-term effects on the alloca-
tion of workforce and labor market equilibrium. However, gradual changes in the
structure of recent migration may have serious long-term impacts, in particular on
the local and regional, as well as on the sectoral level.
The chapter is structured as follows. The first section provides an overview of

statistical evidence on the scale and dynamics of post-accession outflow from Po-
land. Additionally, we analyze selected structural features of recent migration and
look at migrant selectivity measures. The second section discusses labor market
impacts of migration, and an attempt is made to distinguish between effects that
could be associated with the outflow of workers from Poland and those predomi-
nantly related to a given phase of the business cycle. The last section concludes.
2 EU Enlargement and Migration from Poland
Central and Eastern Europe has been an area of large scale population movements
for centuries. The same holds true for Poland, which is commonly described as an
‘emigration country’. Indeed, mass emigration from Poland started at the end of
the 19th century and continued until the Second World War. In the post-war pe-
riod from the beginning of ‘the cold war’ in 1948, the communist regime at-
tempted to control population movements by imposing restrictive exit policies.
Some of these restrictions were lifted in the mid-1950s when out-migration of
9. Impact of the Post-Accession Migration on the Polish Labor Market 221
thousands of so-called ethnic Germans was allowed. Mobility started to increase

with changes in political climate, including the normalization of Polish-German
relations and the liberalization of international mobility. The process peaked in the
1980s and in particular after the introduction of martial law in December 1981.
The scale of long-term mobility in this decade is estimated to have been between
1.1 and 1.3 million people, and an additional one million are estimated to have
spent less than 12 months abroad. The importance of migration in this period of
Polish history was clearly shown by the national census data from 1988, which in-
dicated that around 900 thousand permanent citizens of Poland (2% of the total
population) stayed abroad temporarily (Kaczmarczyk and Okólski 2002).
On the eve of the systemic transition, Western experts expected a significant in-
crease in the scale of international migration from Central and Eastern Europe, in-
cluding Poland (see Layard et al. 1994). This presumption was mainly based on
drastic differences in real wages and living standards. However, this expectation did
not come true: neither official data on registered migration nor LFS data showed an
increase in international mobility. On the contrary, the LFS data indicated a slow
decline in the scale of migration between 1994 and 1998 from over 200,000 to
150,000 people stayed abroad every quarter (see Figure 3)1. Migration from Poland
started to rise again in the late 1990s and was mostly as a consequence of a serious
slowdown of the Polish economy and poor labor market conditions. According to
the 2002 National Census, there were around 790,000 Polish citizens (1.8% of the
total population) living abroad, which shows that prior to the EU enlargement Po-
land was already one of the most important sending countries in Europe.
In May 2004, Poland joined the European Union and was once more expected to
create a significant migratory impetus. However, it is important to note that due to
decisions of particular EU15 countries, the institutional framework did not change
as significantly as expected. In fact, only three countries introduced the freedom of
movement as of May 2004: Ireland, the United Kingdom and Sweden. Most other
EU15 countries made use of transitory arrangements and only opened their labor
markets for Polish citizens over the following 2-3 years. A few countries, Germany
being the most important one, are still considering keeping barriers until 2011.
Before analyzing the post-enlargement migration of Poles in detail, it is neces-

sary to make a few important methodological remarks. Typically, the basic statis-
tical source of data on mobility is the population register. However, due to defini-
tions applied in Poland, the data provided by this system represents only a small
1 Note that chances to obtain reliable estimates of the scale of international mobility in the
transition period are limited. This is due firstly to a lack of migration data or its poor quality,
and secondly to the variety of movements and categories of migrants observed. Short-
term or even circular migration has become particularly important, but it is extremely diffi-
cult to trace in statistical terms. The few existing data sources can capture only a part of
the phe-nomenon and should be interpreted with caution (see also below).
fraction of the phenomenon.2 An international migrant is defined as someone leav-

ing Poland with an intention to settle abroad and deregisters him- or herself from
the previous permanent place of residence. Such deregistrations are relatively rare
as they are generally perceived as unnecessary; hence, a large number of Poles
abroad are still considered permanent residents of Poland in administrative terms,
even if they are de facto emigrants. As a consequence, the most reliable data on
the international mobility of Polish citizens can be obtained from national cen-
suses (the most recent was completed in 2002) or from the LFS3, or it can be a
combination of both sources (Kaczmarczyk and Okólski 2008a).
2.1 Scale and Dynamics of Recent Migration from Poland
Considering commonly recognizable problems with data on international migra-

tion, we propose to begin the discussion on post-accession migration from Poland
with an estimate presented by the Polish Central Statistical Office in April 2008.
The figures shown in Table 1 were computed based on data from the 2002 Na-
tional Census (stock data), Polish LFS data (stock data) as well as data originating
from destination countries (both stock and flows data), and it is considered the
most reliable estimate presented so far.
The data presented in Table 1 indicates that between 2004 and the end of 2007
the number of Polish registered as permanent citizens staying temporarily abroad
increased by almost 1,5 million and more than doubled since 2005, and it equaled
around 2,3 million (6.6% of the total population). Most recent migrants decided to
go to other EU countries: the share of those staying in the EU24 equaled 80%,
compared to 57% in 2002. There was an important shift in destination countries:
Germany, which was the most important target country prior to 2004, lost its top
position and hosted less than 25% of Polish migrants in 2008. On the other hand,
the UK experienced a spectacular increase from 24,000 in 2002 to almost 700,000
in 2008. However, equally strong changes were experienced in a few other coun-
tries, including Ireland, which experienced an extraordinary increase from 2,000
to 200,000, the Netherlands or Sweden. It is important to note that these shifts
among the top destination countries are not necessary a consequence of Polish
mobility between EU member states. The high position of the UK seems rather to
be an outcome of recent outflow, while there is still a numerous group of people
2 E.g. the number of emigrants from Poland between 1990 and 2005 was around 350,000.
3 Polish Labour Force Survey (Badanie Aktywnosci Ekonomicznej Ludnosci) is used as a
source of data to allow the monitoring of intertemporal changes in Poles’ mobility. Accord-
ing to the Central Statistical Office, the data is not fully representative with regard to migra-
tion: the data set was meant as a tool to analyze labour market phenomena and the sam-
pling method was subordinated to this task. However, as shown by a series of tests car-
ried out by the Centre of Migration Research, it can be used to analyze structural features
of migrant population (Kaczmarczyk and Okólski 2008a).
choosing Germany as destination. In fact, the number of Poles staying in Germany

increased between 2004 and 2008. Moreover, recent migration from Poland
should not be understood in terms of concentration (i.e. concentration in English
speaking countries) but rather as ‘spilling over’ (Kaczmarczyk and Okólski
2008b). The data show that Polish citizens turned out to be extremely mobile in
international terms.
Table 1
The stock of temporary migrants from Poland by major destination countries, in thousands
* * *
Destination May 2002 2005 2007 2008
Total 786 1000 1950 2270
EU24 451 750 1550 1860
Austria 11 15 34 39
Belgium 14 13 28 31
France 21 30 49 55
Germany 294 385 450 490
Ireland 2 15 120 200
Italy 39 59 85 87
Netherlands 10 23 55 98
Spain 14 26 44 80
Sweden 6 11 25 27
United
Kingdom 24 150 580 690
Source: CSO (2008).
Notes: * as of January 1st.
The spectacular increase in the scale of migration from Poland and other Cen-
tral and Eastern European countries (CEE) can also be seen from the European
LFS data. As shown by Brücker et al. (2009), the number of Polish migrants in
EU15 countries, as indicated by the European LFS, increased by more than
600,000 between 2004 and 2007 (from 593,000 to 1,271,000). The total number
of EU8 migrants staying in the ‘old EU countries’ changed significantly as well:
from 926,000 to 2,242,000. It follows that Poland is the most important sending
country not only in absolute terms but also in relative terms. The share of migrants
in the sending population increased from 1.55% in 2004 to 3.33% in 2007. A
higher proportion of migrants in the total population occurred only in case of
Lithuania, with 1.4% in 2004 and 3.7% in 2007.
The data presented in table 1 show the existence of an immense dynamic of the
outflow in the post-accession period: the stock of Polish temporary migrants
changed by almost 100% between 2005 and 2007. Similar conclusions can be
drawn from the data from registration systems introduced in the main destination
countries (see Figure 1 and Figure 2).
Figure 1
Number of Polish applicants to the WRS system and year-to-year quarterly changes
in number of applicants (2004, 2nd quarter - 2008, 3rd quarter)
50000
40000
30000
20000
10000
-10000
-20000
Q2 Q3 Q4 Q1 Q2 Q3 Q4 Q1 Q2 Q3 Q4 Q1 Q2 Q3 Q4 Q1 Q2 Q3
2004 2005 2006 2007 2008

Number of applicants Year-to-year change
Source: Own calculations based on Home Office data.
Figure 2
PPS numbers issued to Polish citizens and year-to-year monthly changes
(July 2004 – November 2008)
14000
12000
10000
8000
6000
4000
2000
-2000
-4000
-6000
7 9 11 1 3 5 7 9 11 1 3 5 7 9 11 1 3 5 7 9 11 1 3 5 7 9 11
2004 2005 2006 2007 2008
Number of PSS numbers issued Year-to-year change
Source: Own calculations based on the Department of Social and Family Affairs data.
The data on the number of applicants with the Worker Registration Scheme
clearly shows a steady increase in the scale of labor migration from Poland to the
UK between 2004 and 2006.4 The same holds true for Ireland; however in this
case, the design of the data base is different. Figure 2 presents data on the Per-
sonal Public Service numbers obtained by Polish citizens.
Monthly PPS numbers data suggest similar dynamics of inflow to Ireland but
additionally point to a very strong seasonal pattern of recent migration from Po-
land. Both data sources also indicate that the dynamics of recent migration from
Poland slowed down significantly in 2007 and to an even greater extent in 2008.
In Ireland, negative changes (as compared to the same month a year before) were
visible from early 2007. The same can be concluded from the Polish LFS data (see
Figure 3).
Figure 3
Migrants from Poland according to the Polish LFS: number of people staying
temporarily abroad and year-to-year quarterly changes in thousands (1993-2008)
600
500
400
300
200
100
-100
II 1993
II 1994
II 1995
II 1996
II 1997
II 1998
II 1999
II 2000
II 2001
II 2002
II 2003
II 2004
II 2005
II 2006
II 2007
II 2008
IV 1993
IV 1994
IV 1995
IV 1996
IV 1997
IV 1998
IV 1999
IV 2000
IV 2001
IV 2002
IV 2003
IV 2004
IV 2005
IV 2006
IV 2007
Number of persons staying temporarily abroad Year-to-year changes
Source: Own calculations based on the LFS data.
Contrary to previous data, the Polish LFS provides information on the stock of
migrants staying abroad temporarily. The data clearly indicate a significant in-
crease in the scale of mobility in the post-2004 period. However, it also suggests
that from early 2008 the number of Poles staying abroad has been decreasing. This
4 Note, however, that particularly in the very first phase after accession, a significant share
of applications (30-40%) were submitted by those who were in the UK prior to 2004 and
just used the opportunity to legalize their stay abroad.
is one of the very first signs showing an expected process of return migration; an-
other important sign is the number of Polish returnees registering at labor offices
in Poland and using E301 and E303 forms to prove they are employed or entitled
to social benefits abroad. Latest developments in the real sphere of economy are
expected to foster this tendency.
2.2 Structural Features of Polish Migration

in the Pre- and Post-Accession Period
Polish migration in the post-accession period used to be described in terms of con-

tinuity and change (Kaczmarczyk 2008, Kaczmarczyk and Okólski 2008a). This
section is devoted to the analysis of the main structural features of the outflow,
and LFS data will be used to juxtapose migrants’ characteristics in the pre- and
post-2004 phase.5
One of the most important aspects of continuity is the predominance of labor

migration. Based on the Polish LFS data, it can be argued that more than 80% of
migrants take up employment while staying abroad. This has not changed since
the early 1990s. A second important feature of migration from Poland in the tran-
sition phase is the relatively large share of those who were staying abroad on a
short-term basis i.e. in the first half of the 2000s, over 60% of all temporary mi-
grants stayed abroad for less than 12 months.6 However, long-term mobility has
been growing in importance since 2004. Data gathered in destination countries
(i.e. the UK) point to a growing propensity to prolong the stay abroad, but simul-
taneously, the behavior of migrants is commonly described in terms of ‘intentional
unpredictability’, which suggest rather uncertain/fluent migration strategies
(Drinkwater, Eade and Garapich 2006).
As suggested above, one of the most striking changes observed in the post-
accession period is the ‘portfolio’ of destination countries (see Table 1 and Figure
4). Figure 4 shows that shifts observed with regard to target countries reflect to a
5 The analyses presented in this section are based mainly on migrant databases created on
the basis of the LFS data. The first one includes temporary migrants from Poland (i.e.
permanent residents aged 15 or older who at the time of survey stayed in a foreign coun-
st
try for longer than two months (from 2007 three months) who left between the 1 quarter
st
of 1999 and 1 quarter of 2004 (‘pre-accession migrants’). The second encompasses
st th
those who left Poland between the 1 quarter of 2005 and 4 quarter of 2006 (‘post-
accession migrants’). Due to problems with categorization (period of departure could not
nd
be identified unambiguously), those migrants who were recorded between the 2 quarter
th
2004 and 4 quarter 2004 were not included in the sample.
6 It is necessary to note, however, that due to the very nature of the data base survey
among Polish households, LFS underestimates the number of long-term and permanent
migrants, particularly those who left with their whole family.
large extent the ‘migration regime’ of a given EU member state. Those countries
which decided to open their labor markets to EU8 migrants immediately after EU
enlargement saw the biggest increase in migration from Poland, with an increase
from 12% to 42%. The share of those which imposed long-term restrictions on la-
bor market access declined from more than 40% to 25%. However, as noted
above, this shift was not due to an intensification of intra-European mobility,
which was within the EU15, but rather to the differences in dynamics of recent
outflow.
Figure 4
Destination countries of Polish migrants in the pre- and post-accession period
by type of ‘migration regime’*, in %
100%
12.3
25.3
80%
25.3
60%
42.2 18
40%
20%
19.8 42.4
12.1
0%
pre-accession post-accession
No restrictions Short-term restrictions Long-term restrictions Outside EEA
Source: Fihel et al. (2008).

Notes: * No restrictions: Ireland, UK, Sweden, the new member states (excluding Bulgaria, Malta and Romania); short-term
restrictions: Finland, France, Greece, Iceland, Italy, the Netherlands, the new member states (including Bulgaria and Roma-
nia), Portugal and Spain; long-term restrictions: Austria, Belgium, Denmark, Germany, Luxembourg, Norway.
Important changes were also noted with regard to the geography of outflow.
Migration from Poland was traditionally strongly diversified in spatial terms. Prior
to EU enlargement, most numerous flows (in relative terms) were observed in re-
gions with the longest tradition of international migration and strongest migrant
networks: in the Opolskie region, with a significant share of ethnic Germans in the
total population; the Małopolskie region, with a long tradition of mobility to the
US; and the Podlasie region, with intense migration to the US and Belgium (see
Map 1).
Map 1
Intensity of the outflow* in the pre- and post-accession period, by Polish regions
Source: Kaczmarczyk and Okólski (2008a).

Notes: * number of migrants per 1,000 inhabitants
This particular pattern changed after EU enlargement. Opening of a few rela-

tively absorptive labor markets made migration less risky and more easily acces-
sible. As a consequence, the structure of migration broken down by the sending
regions is less diverse than before. In the post-accession phase, the most important
migrant sending regions include predominantly economically underdeveloped ar-
eas with relatively large shares of natural or semi-natural agriculture. Podkar-

packie is the best example. Additionally, there is a visible change regarding the
type of settlement migrants originate from. A general tendency both in the pre-
and post-accession period was an overrepresentation of migrants originating from
rural areas and, to some extent, from medium and small towns. In the post-
accession period, however, a decline in the proportion of residents of rural settle-
ments among the migrants is accompanied by growing shares of urban residents
among those who migrate (Kaczmarczyk and Okólski 2008a, see also below).
With regard to demographic characteristics, an important feature of the post-

accession outflow is the increasing masculinization of migration. Prior to EU
enlargement, migration from Poland was a dominated by men, who constituted
57% of migrants; however, it has become even more visible since then. The share
of male population equaled 65% in the post-accession period compared to 47% in
the total population. Such a structure is observed in most destination countries, but
in the case of a few, the structure of demand of foreign labor led to opposite re-
sults e.g. in Italy (Fihel et al. 2008).
Figure 5
The age structure of Polish migrants
100%
9.8 7.7
90%
7.5
8.5
80%
6.3
9.4
8.1
70%
10.3
13
60%
10.3
50%
27.3
40% 21.1
30%
20%
26.7 27.7
10%
3.8 2.3
0%
15-19 20-24 25-29 30-34 35-39 40-44 45-49 50+
Source: Fihel et al. (2008).

Typically, international labor migration is dominated by young people, mostly

of mobile age. The same holds true for Poland. Those who migrate are predomi-
nantly aged 20-39: their share equaled 68.4% in the pre- and 76.1% in the post-
accession period. However, a new tendency is a growing share of very young
emigrants which are people aged 20-29, who constituted 47.8% prior to the EU
enlargement and 55% thereafter (see Figure 5).
Data presented in table 2 shows that following EU enlargement, the migrants

became significantly younger: the mean age decreased by 1.5 years (from 33.9 to
32.4) and the median age by 2 years (from 30 to 28). Interestingly, changes ob-
served are closely connected to the shift in the structure of destination countries.
In the post-accession period, migrants to countries which did not impose labor
market restrictions were 5.1 years younger on average than those choosing desti-
nations with short- or long-term barriers. A striking difference was found between
the two most important destinations: for the UK, the mean and median age in the
post-2004 period equaled 28.2 and 26.0, respectively; whereas for Germany, the
mean age equaled 35.6 and median age 34.0. The data presented in Table 2 show
that ‘new destinations’ (UK and Ireland) are particularly attractive to very young
migrants. This feature is bound to influence other migrant characteristics as well
as the labor market effects of recent migration.
Table 2
Mean and median age of Polish pre- and post-accession migrants
by main groups of destinations and selected destination countries
Destination
country
Mean Median Mean Median
No labor
28.9 25 28.5 26
restrictions
Ireland 28.5 26 30 27
Sweden 34.8 32.8 29.6 26
UK 28.2 25 28.2 26
Short- and
long-term 32 29 33.6 31
restrictions
Germany 33 30.4 35.6 34
Italy 34.2 31 35.3 32
Norway 39.5 37 36.1 34
Spain 31.6 29 32.6 28
Outside EEA 36.1 32 34.9 30
Canada 44.9 46.3 45.3 46.1
US 39.6 38 39.6 35.3
Total 33.9 30 32.4 28
Source: Fihel and Kaczmarczyk (2008).
Another important characteristic is the level of education of recent migrants.

However, even if migration from Poland was commonly described in terms of a
‘brain drain’, this was not necessarily true. Particularly in the 1990s, Polish mi-
grants were on average low-qualified, and this fact corresponded with a high con-
centration in low-paid, insecure jobs, which did not demand higher skills in the
receiving labor markets.
Table 3
The education structure of Polish pre- and post-accession migrants by sex, in %
pre-accession1 post-accession2
Level of education Total Men Women Total Men Women
University degree* 14.7 12 18.3 19.8 15.6 27
Secondary 14 7.1 23.1 14.2 8.8 23.8
Secondary vocational 26.1 26 26.3 28.1 29.8 25.1
Vocational 34.8 45.4 20.9 30.9 39.2 16.2

Primary 9.9 9.3 10.9 7 6.6 7.8
Unfinished 0.4 0.2 0.5 0 0 0
Total 100 100 100 100 100 100
Notes: *Including bachelor, master and Ph.D. degree.
From Table 3, it follows that even if migration in the pre- and post-accession
period was dominated by those with vocational education (61% and 59% of all
migrants respectively) the share of migrants with tertiary education (a university
degree) significantly increased after EU enlargement. The share of those holding a
university diploma rose from 14.7% to almost 20%; this percentage was as high as
27% for women.
Reasons for this change are manifold. The most important is the gradual im-
provement in the level of education of Poland. However, rising shares of Polish
graduates who migrate are also linked to changes in the structure of recent migra-
tion. In particularly, the increase in the relative scale of mobility of the high-
skilled was caused by an increasing dominance of the UK and Ireland among des-
tination countries (see Figure 6).
The data presented here show an immense contrast between the educational
structure of those who migrated to those countries that did not introduce labor
market restrictions and those who chose traditional destinations, in particular
Germany. In the UK and Ireland, shares of migrants with a university diploma in
the post-accession period amounted to 25-30%. For young people aged 25 and be-
low, the share of high-skilled was even bigger (Fihel et al. 2008).
Figure 6
The education structure of Polish migrants by type of destination, in %
pre-accession period
tertiary
45
40
35
30
25
20
primary and unfinished 15 secondary
10
5
0
vocational secondary vocational
IE, SE, UK Other EEA, CH Other
post-accession period
tertiary
50
45
40
35
30
25
20
primary and unfinished 15 secondary
10
5
0
vocational secondary vocational
IE, SE, UK Other EEA, CH Other

The statistics already presented in this section show that accession brought sig-
nificant changes both in the structure and the scale of population outflow. How-
ever, according to the analysis of Migrants’ Selectivity Indices7 (MSI) for main
socio-demographic characteristics time series based on LFS data 2004-2008, the
post-accession change in the structure of Polish migrants seems to be neither uni-
directional nor static.
Previous analyses of selectivity of the latest migration from Poland (Okólski

and Mioduszewska 2008) showed that the main selective factors for migration
process before May 1st 2004 were: sex, age, educational attainment, type of set-
tlement and region of origin (see Table 4).
Table 4
Migrants’ Selectivity Indices for main socio-demographic characteristics
in the pre-accession period
Category
(selected MSI in the pre-
Variable characteristics) accession period
Region of Opolskie 1.63
origin Podkarpackie 1.69
Podlaskie 1.61
Educational Tertiary 0.02
attainment Vocational 0.34
Age Mobile (20-39) 0.97
Sex Male 0.2
Type of Cities over
Settlement 100,000 -0.3
Cities under
100,000 0.09
Rural areas 0.15
Source: Mioduszewska (2008).
The highest values of MSI were observed in specific regions of origin with
strong traditions of migration (Podkarpackie, Opolskie and Podlaskie) and in the
‘mobile’ age category with other factors being of less importance. Nevertheless, at
7 The Migrant Selectivity Index can be presented by means of the following formula:
M V PV = i
=i
−
MSI = M P
V =i
PV = i
P
where MSIV=i is the index for category i of variable V; MV=i and PV=i is the number of mi-
grants and the number of people in the general population, respectively, falling into cate-
gory (or value) i of variable V; and M and P are the overall number of migrants and people
in the general population, respectively. The selectivity of outflow takes place if the index
assumes a non-zero value for any category (value) of a given variable. Positive values of
MSI mean that migrants falling into a specific category (variable) of a given variable are
relatively more numerous than people in general population with the same characteristic;
whereas negative values (but equal to or higher than -1) mean the opposite. The higher
the positive value or lower the negative value of MSI, the stronger the selectivity.
the very beginning of the post-accession migration process, the most visible
change emerged in sex selectivity and was caused by rapidly growing overrepre-
sentation of males in the outflow to the UK (see Figure 7). The high volatility of
MSIsex = male for Polish migrants to the UK may illustrate that the selectivity
pattern of this destination had not been formed yet, as opposed to the rather stable
and predictable trend in MSIsex = male for the outflow to Germany, which has
been the traditional receiving country for Polish migrants.
Figure 7
Migrants’ Selectivity Indices for males
0.7
0.6
0.5
0.4
0.3
0.2
0.1
0
III 2004 IV 2004 I 2005 II 2005 III 2005 IV 2005 I 2006 II 2006 III 2006 IV 2006 I 2007 II 2007 III 2007 IV 2007 I 2008
Total Germany UK
Source: Own calculation based on LFS.
The age selectivity after EU enlargement remained very high: during this pe-
riod the MSIage = mobile oscillated around 1.2. Furthermore, the difference was
also high between selectivity indices for two major destination countries indicated
by the data presented in Table 2. Extremely high values of MSIage = mobile for
the outflow to the UK in the first quarters after accession, with a peak at the be-
ginning of 2005, were probably caused by the predominant presence of migrant
pioneers: young men with the greatest chances for success in a foreign labor mar-
ket and the lowest level of risk aversion. However, sex selectivity of the outflow
to Germany tended to be rather stable and slightly increasing, excluding regular
cycles.
Figure 8
Migrants’ Selectivity Indices for two types of educational attainment
Tertiary education:
0.7
0.6
0.5
0.4
0.3
0.2
0.1
0
Total Germany UK
Vocational education:
1.5
0.5
-0.5
-1
Total Germany UK

The last and probably most interesting variable that differentiates the outflow to the
most significant receiving countries is educational attainment. Poles who had a univer-
sity degree before May 1st, 2004, as well as after were more likely to choose the UK
rather than Germany as a destination country. The opposite holds true for migrants
who had only completed vocational education. However, an upward shift can be ob-
served in MSIeducation = tertiary for the Germany directed outflow in the last quarters
(Figure 8).8 This in turn may signal a change in migration selectivity patterns.
Figure 9
Migrants’ Selectivity Indices for each type of settlement
0.6
0.4
0.2
0.0
-0.2
-0.4
-0.6
Cities over 100 thous. Cities under 100 thous. Rural areas
Finally, let us focus on the type of settlement as a selective factor in the post-
accession period, as the changes in selectivity indices for its categories seem to be
the most striking. Living in a big city (i.e. with over 100,000 inhabitants) has never
been a push factor in the post-accession phase: a negative value of MSIsettlement =
big city oscillates around -0.4 (see Figure 9). During the first two years after acces-
sion, the population outflow was mainly driven by migrants originating from small
and medium size cities of up to 100,000 inhabitants, followed by those who live in
rural areas. Surprisingly, at the time of second anniversary of the EU enlargement,
8 A positive value of MSItertiary for Poles migrating to Germany (0.08) is actually statistically
equal to 0. The p-value for mean test equals to 0.57 i.e, the null hypothesis which states
that the frequency of observations with a university diploma is the same in the sample of
migrants and non-migrants cannot be rejected.
this trend reversed and remained relatively stable. Furthermore, selectivity indices
for all the types of cities started to behave in almost the same manner.
The explanation for this fact may be twofold. First of all, it may be an effect of
the existence of migration networks in rural areas which ‘triggered’ much more
intense outflow after reaching a ‘critical mass’. On the other hand, the observed
phenomena could be linked to the economic conditions in Poland, particularly to
relatively worse labor market situations in rural areas.
To conclude, post-accession migration of Poles in structural terms differs sig-

nificantly from the population movements observed in the pre-enlargement period.
Recent migrants are on average younger, better educated, more likely than before
to come from cities (small and medium size). However, there are two important
remarks to make. Firstly, recent outflow from Poland is not a homogenous proc-
ess. Even if migration to countries like the UK or Ireland is remarkable, there are
still relatively significant streams targeting more traditional destinations like Ger-
many or the USA. These changes are reflected in the structural characteristics of
recent migrants. Secondly, data on the MSI clearly suggest that post-accession
migration from Poland is an ongoing phenomenon and most of the hypotheses
proposed above lack unambiguous evidence.
3 Impacts of Migration on the Polish Labor Market
The labor market consequences of migration have long attracted the interest of
both theoretical and empirical researchers. However, the main weight of studies
lies on the receiving end: most researchers investigate the labor market effects on
labor importing countries; the number of studies on the impact of emigration on
the labor market of the sending countries is modest (especially in the case of Cen-
tral and Eastern Europe). This fact is not all that surprising given that immigration
is often publicly regarded as a potential threat to the labor market and any policies
of admitting foreign workers require thorough investigation. On the other hand,
emigration policies are much harder to evaluate and implement, and emigration it-
self is much harder to control in a democratic system.
The effects of a massive outflow of migrants who form part of a country’s la-
bor force is addressed in migration theory (Borjas 2004, IOM 2005) and may be
classified according to the labor market sector they affect: the supply of the labor
force, whose effects should be visible even in the short term as well as in a longer
perspective; the ensuing equilibrium adjustments visible in the medium term; and
demand side modifications, with effects visible mostly in the long term.
The impact of large-scale emigration on the supply of labor may be both quan-
titative and qualitative. The most important quantitative effect is the decline in un-
employment. Qualitative effects include changes in the composition of the labor
force due to the selectivity of migration, as discussed in the preceding section of
this chapter, and human capital. This latter effect may appear as a consequence of
the migration of the highly skilled. It has received a considerable amount of atten-
tion, with models discussing both the negative and positive effects of such forms
of mobility (Grubel and Scott 1966, Bhagwati and Hamada 1974, Stark et al.
1997, Dumont and Lemaître 2005, Docquier and Marfouk 2004).
From a theoretical point of view, the most important labor market equilibrium
adjustment due to a massive outflow of the labor force, which results in the de-
cline of the supply of labor force, should be wage pressure. Empirical research
devoted primarily to mobility from Mexico to the USA mostly confirms these
theoretical predictions (Mishra 2007, Hanson 2005, Aydemir and Borjas 2006).
The demand side of the labor market cannot remain unaffected in view of the
changes of labor force availability and wage pressure. The expected outcome is a
change in employer’s strategies and eventually an upward mobility of foreign
(immigrant) labor (Borjas 2004, IOM 2005). In the remaining part of this section,
we will try to address the above-mentioned theoretical predictions in the context
of post-accession Poland.
Commonly to other CEE countries, Poland struggled with high unemployment

in the pre-accession period: it was as high as 20% in 2002. Furthermore, there was
a large share of long-term unemployment, low participation and employment
rates, and structural shortages. The shortages were mainly a relic of the commu-
nist period and were caused by low levels of human capital and sustained by low
levels of internal mobility – the effect of the latter factor is the lack of alleviation
of educational and professional mismatches on the regional level. The labor mar-
ket gradually improved as the Polish economy grew. GDP growth in each of the
four years from 2000 to 2004, when accession took place, was 1.2, 1.4, 3.9 and
5.3%, respectively.
In 2004, after a two-year period of stability, the number of unemployed started

to gradually decrease: according to LFS data, the number of unemployed de-
creased from 3.2 million to 1.2 million (the unemployment rate decreased from
19.1% to 7.1%) from the second quarter of 2004 to the second quarter of 2008. On
the other hand, the number of migrants living abroad increased by approximately
200,000, from 240,000 in the second quarter of 2004 to 540,000 in the second
quarter of 2007. Afterwards, the stock of migrants stabilized and started to de-
crease slightly: the estimated stock of migrants fell to 510,000 in the second quar-
ter of 2008 (Figure 10).
Figure 10
Migration and the labor force in Poland (2000-2008)
4000 16000
3500 15500
Employed and out of labor force

3000 15000
Unemployed and emigrants
2500 14500
2000 14000
1500 13500
1000 13000
500 12500
0 12000
II 2000
II 2001
II 2002
II 2003
II 2004
II 2005
II 2006
II 2007
II 2008
IV 1999
IV 2000
IV 2001
IV 2002
IV 2003
IV 2004
IV 2005
IV 2006
IV 2007
Emigrants Unemployed Employed Out of labor force
One can conclude on the basis of the data above that, following EU accession, a
rising intensity of migration was accompanied by declining unemployment: the
Pearson correlation coefficient for the period from the second quarter of 2004 to the
second quarter of 2008 equals -0.96, which indicates an almost perfectly negative
linear relationship between the two time series.9 However, this observation alone
does not prove the causality between migration and unemployment or an “unem-
ployment export” hypothesis. As depicted in Figure 10, the fall of unemployment
during the post-accession period was also accompanied by a rise in employment
rates from 44% to 50.1% between the second quarter of 2004 and second quarter of
2008 and a decline in the number of those economically active. Activity rates during
this period varied, but overall they fell slightly from 54.4% to 53.9%. The general
trends in the labor market continued even once emigration rates had stabilized, and
then the trend reversed.10 This indicates that post-accession emigration could not
have been the primary cause of the changes in the labor market, which resulted
mainly from structural and business cycle changes in the whole economy, albeit the
9 This relationship is in stark contrast with the pre-accession relationship between unem-
ployment and migration, which for the period between the fourth quarter of 1999 and the
first quarter of 2004 showed a Pearson correlation of 0.77, indicating that prior to the ac-
cession, the causality may have run in the adverse direction, with high unemployment lev-
els being one of the push factors for migration.
10 Although preliminary data for the third quarter of 2008 show unfavourable changes in em-
ployment and unemployment in Poland.
favorable condition of the labor market (the increasing demand for labor) was, to a
large extent, a consequence of the integration with the EU economy.
Recent studies on the relationship between labor markets in the new member
states and EU15 show that in the post-accession period, the link between employ-
ment levels in particular EU countries has strengthened. For example, Lo Turco and
Parteka (2008) analyzed the link between labor markets and trade with EU partners
and found that in tradable sectors, domestic employment in new member states is
positively affected by employment in trade-partner states. This supports the hy-
pothesis that the correlation between labor emigration from Poland and a decline in
unemployment there results from the fact that both are affected by the same factor:
namely, the business cycle in an enlarged European Union. However, the causality
between the two phenomena themselves is not significant. When demand for labor
declined in the EU towards the end of 2007 and 2008 due to cyclical factors and fi-
nancial crisis, both emigration and employment in Poland were affected. This was
especially true in export sectors.
The stock of migrants rose by approximately 300,000; whereas unemployment

fell by 2 million. Therefore, it should be stressed that the absolute changes in the
volumes of migration flows and unemployment mean that even if emigration had a
direct impact on the level of unemployment, only a small proportion of changes in
the latter variable could be attributed to the first. Bukowski et al. (2008) investigated
the impact of three factors on unemployment: demographic, changes in economic
activity, and changes in employment. They conclude that in the post-accession pe-
riod, the factor with the highest impact on the level of unemployment was the level
of employment. In other words, the changes in the level of unemployment may be
attributed mainly to a rise in the level of employment.11
This effect is especially clear with individuals in the most mobile age groups, 15-
24 and 25-44 (Figure 11). Both the demographic and activity effects are very small
for these two age groups, and they sometimes even work in the opposite direction,
i.e. increase unemployment. Theoretically, both the changes in activity patterns and
demographic effects may be influenced to some extent by massive post-accession
emigration. However, even if this were the case, the impact of migration on unem-
ployment would be marginal given the magnitude of the effects of these components
on the total number of unemployed.12 The results obtained by Bukowski et al.
(2008) thus support the hypothesis that the impact of emigration on unemployment,
if it exists, is not substantial.
11 The impact of the demographic factor was negligible throughout the whole period. The im-
pact of activity was small and in the second half of 2004/ first half of 2005 was observed to
“work” in the opposite direction and increase unemployment.
12 It should be stressed that demographic effects for those aged 15-24 should be attributed
primarily to the effect of the “dying out” of the demographic pyramid hump of the baby
boom of the 1980s.
Figure 11
Impact of demographic factors, changes in economic activity and employment
on unemployment (2004-2007)*
20 15-24
10
0
IQ 2004
IIQ 2004
IIIQ 2004
IVQ 2004
IQ 2005
IIQ 2005
IIIQ 2005
IVQ 2005
IQ 2006
IIQ 2006
IIIQ 2006
IVQ 2006
IQ 2007
IIQ 2007
-10
-20
-30
-40
Demography
-50 Employment
Activity
-60
-70
40 25-44
20
0
IQ 2004
IIQ 2004
IIIQ 2004
IVQ 2004
IQ 2005
IIQ 2005
IIIQ 2005
IVQ 2005
IQ 2006
IIQ 2006
IIIQ 2006
IVQ 2006
IQ 2007
IIQ 2007
-20
-40
-60
Demography
-80
Employment
Activity
-100
-120
Source: Bukowski et al. (2008).

Notes: * The sign is positive for components which increase the level of unemployment and negative otherwise
It is not only the unemployed who migrate. Budnik (2007) analyzed gross
flows between four states in the labor market on the basis of Polish LFS data ad-
justed by the 2002 census data13: employment, unemployment, non-participation
and emigration (Table 5). The reported probability of transition into emigration
was the highest among individuals who were unemployed and amounted to 0.5%.
It was significantly lower for the employed and those not participating in the labor
market (0.1%). However, when one relates the reported transition probabilities to
the population size of the three groups of individuals on the Polish labor market,
the share of previously unemployed individuals among emigrants amounts to ap-
proximately 1/3. This share is more or less (3 p.p. difference) the same as the
other two groups of individuals in the pre-accession period. In the post-accession
period, however, the proportion of migrants who worked prior to emigration rose
to 42%, with the share of non-participation equal to 29%.14 This result is in line
with other data which suggest that in the post-accession period the labor market
status of an individual (his unemployment status) is of relatively less importance
than in the pre-accession period (Kaczmarczyk 2008).
Table 5
Average transition probabilities in the Polish labor market (2002-2006)
Employment Unemployment Non-participation Migration

Employment 97.2 1.3 1.4 0.1
Unemployment 9.1 85.1 5.3 0.5
Non-participation 1.1 1.3 97.5 0.1
Migration 1.7 1.5 1.5 95.4
Source: Budnik (2007).
The estimations derived by Budnik (2007) from the LFS are further supported
by survey data collected within various regional research projects by the CMR.15
Each household survey is representative at the local level and shows that the pro-
13 The probabilities of migration had to be adjusted due to the structure of the LFS, which
underestimates emigration of individuals who travelled with all household members (for
whom no household members are present in Poland for the survey) and long-term emigra-
tion. However, the methodology adopted by Budnik, who used the 2002 census data, may
be undermined by the changes in the structure of post-accession migration flows, with re-
spect to the situation observed in 2002.
14 Post accession – 2006.
15 The surveys were conducted within the following projects: “Migration Policy as an instru-
ment for employment promotion and unemployment reduction”, conducted in 2007 in four
selected regions of Poland (Bilgoraj, Monki, Slupca, Starachowice); “Transborder Centre
of the Labour Market Support” and “International migration in the border region of the Dol-
noslaskie voivoidship”, each conducted in 2008 in one selected region (the cities of Lublin
and Zgorzelec, respectively). The locations for the surveys were chosen on the basis of
local labour market and economic characteristics with the aim of researching locations
with a large degree of heterogeneity (albeit only among regions with high levels of out-
flow).
portion of individuals employed prior to their (last) post-accession migration var-

ies greatly by location: from 22% in the Monki region to 64% in the Starachowice
region; The variation also depends on local migration patterns and the local labor
market; however, in each case, the share of employed individuals is substantial
(Table 6).16
Table 6
Share of individuals employed prior to their last migration, post accession period
Share of employed
Region
among migrants
Bilgoraj 44%
Monki 22%
Slupca 45%
Starachowice 64%
Lublin 41%
Zgorzelec 54%
Source: Own calculations based on CMR databases 2007, 2008.
Budnik (2007) extended the investigation of the transition probabilities be-

tween different labor market states into a steady-state analysis. In order to evalu-
ate the effect of migration on the Polish labor market, she compared the observed
migration scenario vs. a counterfactual one in which a migration state was not ac-
counted for in the transition matrix. The calculations lead to the conclusion that in
the post-accession period, the steady-state unemployment rate is, in the case of the
no migration scenario, slightly higher than in the real-life scenario. In 2006, for
example, the two rates equaled 10.3% and 9.9%, respectively. The bias for the
employment and activity rates was similar in magnitude. This indicates that al-
though the outflow of workers in the post-accession period was large, it did not
have a substantial impact on the steady-state shares of individuals with a different
labor market status. It should be noted, however, that even if no mass effects are
observed on a macro-scale in the labor market, it is possible that migration has se-
vere impacts at the local level.
The selectivity patterns of pre- and post-accession migration discussed in the

previous section suggest that different effects may be visible for specific regions
and types of labor demand. The fact that the selectivity patterns of migration with
respect to education have changed between the pre-accession and post-accession
periods are likely to affect the type and scale of labor shortages on the labor mar-
ket.
The number of vacancies and job openings increased rapidly from 2005 until
the third quarter of 2007. Labor shortages have since declined, albeit still severe in
some cases. The number of companies experiencing labor shortages as a barrier to
16 On average, similar to that calculated by Budnik (2007).

growth varied from practically none prior to 2005 to 14.2% in the third quarter of
2007; in the third quarter of 2008, the number of firms experiencing labor short-
ages as a barrier to growth fell to 6.3%. The shortages of workers were the most
apparent in construction, where as many as 35% of firms operating in this sector
were affected, and manufacturing, with more than 15%. Throughout 2007, labor
shortages were declared the most important barrier to growth (see Figure 12, NBP
2008). However, as the business cycle phase changed in 2008, when the Polish
economy started slowing down, labor shortages ceased to pose a serious problem
for most firms. In the third quarter of 2008, most firms found on average 5 other
barriers more important than labor shortages. Therefore, the labor shortages ob-
served were primarily due to a favorable economic climate, and the impact of
emigration was less important.
Figure 12
Labor shortages as a barrier to growth, by firm sector
0.4
0.35
0.3
0.25
0.2
0.15
0.1
0.05
0
2005 IV 2006 I 2006 II 2006 III 2006 IV 2007 I 2007 II 2007 III 2007 IV 2008 I 2008 II 2008 III
Manufacturing Construction Trade Transport Other Total
Source: NBP (2008).
Besides sectoral differences in the shortages of labor, differences in the re-

quired qualifications of workers also varied throughout the post-accession period
(Figure 13). From the first quarter of 2006 to the second quarter of 2007, a signifi-
cant (20 p.p.) decrease in the number of firms not experiencing labor shortages is
observed (NBP 2008). The changes in 2008 were minor compared to 2007. The
labor shortages are comprised mainly of qualified workers (albeit not necessarily
highly-qualified). However, the breakdown of the demand for labor according to
the sought qualification level changed throughout the observed period: the most
notable increase was among the proportion and number of firms looking for both
qualified and unqualified workers. Their share among all firms seeking labor in-
creased from 15% to 30%. The distribution of qualification shortages varies be-
tween sectors: the service sector has the largest share of firms seeking unqualified
labor.
Combining the results of the analysis of labor shortages with the analysis of
migration selectivity, which showed that individuals having higher and vocational
education were overrepresented among migrants, leads to the conclusion that emi-
gration added to the observed business cycle labor market phenomena in Poland.
Emigration and the upswing of the economy worked in the same direction as far
as labor shortages are concerned. On the one hand, the impact of emigration in
most sectors and regions of the economy was relatively small in comparison with
cyclical factors. According to NBP 2008, the third quarter of 2008 saw firms ex-
perience a loss of workers: approximately 14% was due to emigration; while 45%
left for better paid jobs in Poland.17 On the other hand, the increased outflow of
workers may have significantly affected firms which are always especially vulner-
able: those requiring employees with high qualifications and specific education or
skills. To investigate this matter further, we will devote the following to assess the
impacts of the outflow of highly qualified workers from Poland.
Figure 13
Labor shortages by qualifications of workers (2006-2008)
II 2008 37.9 17.1 5.1 39.8
II 2007 36.5 18.8 4.8 39.8
III 2006 36.9 11.3 2.8 49
1.4
I 2006 34.2 6 58.4
0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100%
Lack of qualified w orkers Lack of both qualified and unqualified w orkers

Lack of unqualified w orkers No labor shortages
Source: NBP (2008).
In the phrasing of Beine et al. (2001), the effects of the outflow of skilled labor
(i.e., brain drain) on the sending country may be two-fold: one may observe both a
negative (an ex post, static) drain effect due to the migration of skilled workers
and a positive (an ex ante, dynamic) brain effect – the increased investment in
education induced by the prospect of migration. We have shown that the recent
outflow from Poland is indeed characterized by a selective mobility of the well-
17 The rest of the firms stated the reasons for the loss of workers were not due to wages or
emigration.
educated. Accordingly, we may say that in the case of Poland, a brain drain pre-
vails. However, the positive selection of well educated emigrants is mostly due to
demographic developments arising from changes in the age structure of the send-
ing population and changes with regard to educational attainment. Individuals in
the most mobile age groups are better educated than average, which means that
the brain drain is not as significant as “raw” statistics would suggest.
Moreover, Fihel et al. (2008) show that the outflow of skilled workers from Po-
land has the features of a brain overflow18. This finding supports the crowding out
hypothesis postulated by Kaczmarczyk and Okólski (2008a) which says that re-
cent mass emigration from Poland may be described to a large extent in terms of
outflow of the “economically redundant” population. It also signifies that any
evaluation of the positive and negative effects of the outflow in terms of the brain
and drain effects is greatly affected.
As far as the drain effect is concerned, Fihel et al. (2008) suggest it is hard to
assess due to the different scales of qualification mismatches in specific sectors
and regions in Poland. Furthermore, the structure of educational attainment is still
changing for reasons other than migratory, which have a far more significant and
positive impact on the number of highly skilled. The EU accession induced brain
effect is also hard to assess because the time available for analysis has been too
short. Indeed, it is shorter than the number of years necessary to obtain higher
education. Therefore, although a trend of growing popularity in education may be
observed in Poland, as shown in the preceding section, it is impossible to extract
any post-accession brain effects. Moreover, recent studies (see for example
Drinkwater et al. 2008, Fihel et al. 2008) suggest that the performance of highly
educated emigrants from Poland in the receiving labor markets is characterized by
severe overeducation: emigrants work in occupations which do not make efficient
use of the individual’s skills or competencies. This signifies that the outflow of
skilled workers from Poland has the characteristics of a brain waste, which un-
dermines the theoretical rationale for increased human capital formation: the pos-
sibility of brain waste should decrease the propensity to acquire higher skills for
those who consider migration as an option.
As we have tried to show in the preceding paragraphs, the macro-scale analysis

of the outflow of highly skilled workers is beset with difficulties and ambiguities.
Therefore, an analysis at lower levels – sectoral or regional – would be more ap-
propriate. In this case, however, the necessary data are even harder to obtain. The
best researched sector which is prone to emigration of professionals is health care.
According to Fihel et al. (2008), the scale of potential migration of medical pro-
18 A brain overflow occurs when there is an (intentional or unintentional) oversupply of edu-

cated professionals in the sending country, whose abilities cannot be matched to job of-
fers. In such a case, migration of the highly skilled occurs at low or zero opportunity costs
and reduces the labour market supply-demand inequality in the sending country.
fessionals can be assessed on the basis of the number of certificates confirming

qualifications and professional experience, which are required by employers in
Western European states prior to employment. The number of such certificates is-
sued to medical doctors in Poland after accession was 6,724 as of the end of De-
cember 2007: 5.7% of the total number of this group. Dentists were issued with
1,924 certificates (6.3% of the total); and 9,300 certificates were issued to semi-
skilled medical staff (nurses and midwives), which amounts to 0.3% of this group.
On the basis of the their data, Fihel et al. (2008) conclude that migration of the so-
called ‘white personnel’ is a striking phenomenon, but the scale is not so large as
to pose a threat to the health care sector in the short-term: the Polish educational
system ‘produces’ medical professionals at a rate higher than their potential out-
flow to other states. However, the outflow of certain types of medical doctors is of
some concern. The proportion of anesthesiologists, plastic surgeons and chest sur-
geons who were issued a certificate corresponds to 15.6, 14.7 and 12.8% of active
specialists in these fields respectively, and this is likely to lead to severe short-
ages.
In economic theory, a direct consequence of labor shortages is wage pressure.

The vacancy rate throughout the post-accession period in Poland remained rela-
tively low, albeit increasing, and it reached a maximum of approximately 2% at
the beginning of 2007. However, firms did report some wage pressure (NBP
2008) which increased in magnitude when labor shortages rose, and decreased
when labor shortages fell (Figure 14).
Figure 14
Labor shortages, wage pressure (seasonally adjusted) and emigration (2005-2008)
45 600
40
500
wage pressure, labor shortages (%)
35
emigration (thousands)
30 400
25
300
20
15 200
10
100
5
0 0
I 2005 II 2005 III 2005 IV I 2006 II 2006 III 2006 IV I 2007 II 2007 III 2007 IV I 2008 II 2008 III 2008 IV
2005 2006 2007 2008
Firms planning on increasing w ages in next period (seasonally adjusted, NBP)
Firms experiencing labour shortages as a barrier to grow th (NBP)
Emigration stock (LFS)
Source: NBP (2008) and LFS.

Throughout the post-accession period, the fraction of enterprises planning on

increasing wages remained higher for firms experiencing labor shortages than for
those not reporting this problem. For example, in the fourth quarter of 2008, 22%
of firms experiencing labor shortages planned to increase wages, compared to
only 12% of the remaining population (NBP 2008). These planned wage increases
did not initially translate to high increases in actual wage levels on the aggregate
level in the economy: real wages rose at a rather moderate 2-4% annually rate be-
tween 2004 and 2006. On the other hand, the average monthly salary in 2007 rose
by approximately 9%, and in the first half of 2008 by 10% compared to the second
half of 2007. The largest increases, at over 15%, were witnessed in the health care,
construction and agricultural sectors; above-average growth was also recorded in
trade and manufacturing sectors. The largest impact was noted in those sectors
where labor shortages were the most intense, following Grabowska-Lusińska and
Żylicz (2008).
The direct impact of migration on wage levels is hard to assess. Budnik (2008)
tries to address the issue by comparing the actual migration scenario and a
counter-factual scenario with migration rates fixed at the 2002 level, with the ef-
fects on a wage equation being derived from a search and matching model. Within
the model, higher transition probabilities of emigration when combined with a de-
cline in emigration cost may increase the permanent income of emigrants and the
non-working population. This would improve the fallback position of potential
workers in the wage bargaining process, thus allow the workers to impose an up-
ward pressure on wages. Budnik (2008) estimates that the steady-state impact on
the wage rate of an increase in outflow of workers of around 4.5% (as observed
between 2002 and 2006) was moderate and equal to around 1% in 2006.
A different effect of shortages due to the outflow of workers may be an ad-

justment in the demand for labor in the longer term: the tendency to substitute in-
existent national workers with a foreign labor force. Grabowska-Lusińska and
Żylicz (2008) analyzed the demand for foreign labor in the context of intensive
labor shortages with a representative survey conducted in the third quarter of
2007. They found that although the realized demand for foreign workers was mar-
ginal, finding that less than 1% of all firms registered in Poland employed foreign
workers, the potential demand was higher: 3.3% planned to hire a foreigner and
18.2% did not rule out the possibility of employing one. For both the realized and,
to an even greater extent, potential demand, a vast majority of firms indicated that
foreigners are, and will be, employed for newly created jobs or to fill in for Poles
who had left of their own free will. Therefore, they were being used to fill labor
market shortages.19 This tendency is even more apparent when one analyses the
19 For the realised demand, the percentages of firms that reported hiring foreigners for newly
created jobs varied between 44 and 54% – depending on the size of the firm; for the po-
tential demand, the same fraction was slightly above 50%. Similarly, the percentages of
firms that reported hiring foreigners to fill in for Poles who had left in the realised demand
motives for hiring foreigners: firm managers stated that the two dominating causes
in the realized demand were the lack of Polish laborers and the fact that citizens
hold specific qualifications not available in the Polish labor market. Each option
was chosen by approximately 40% of firms and was irrespective of size and sec-
tor. For potential (unrealized) demand, the lack of Polish workers was the only
dominating option, and it was chosen by 50-70% of respondents, depending on
firm size. The data reported by Grabowska-Lusińska and Żylicz (2008) clearly in-
dicate that in view of labor shortages, the demand for foreign workers increases,
and foreigners are expected to fill the gaps in the native labor force. Given that
some labor shortages are due to the emigration of Poles, we may conclude from
what we have shown above, that if indeed mass labor emigration from Poland per-
sists, a long-term effect may include a rising participation of foreigners in the Pol-
ish labor market.
In the preceding paragraphs, we have acknowledged the crowding out hypothe-

sis of Kaczmarczyk and Okólski (2008a), who state that the post-accession emi-
gration of workers from Poland may be described in terms of the outflow of ‘eco-
nomically redundant’ population originating from economically backward
regions, which are characterized by limited employment opportunities. In view of
this concept, other long-term effects that may be visible in the Polish labor market
could be related to this phenomenon. The Polish economy is still undergoing a
process of transition to a market economy, with the initial stage, in 1989, charac-
terized by strong regional and sectoral disparities, and with a significant share of
the subsistence and semi-subsistence sectors. The modernization of the economy
cannot be completed without significant structural population changes. If massive
emigration affects the backward regions, it will be a factor which allows allevia-
tion of population surpluses in the peripheral regions. In this context, emigration
may have a significant long-term impact on the Polish labor market.
4 Conclusions
Existing data on post-accession emigration from Poland gathered by Poland and

the receiving countries clearly show a spectacular increase in both migration
stocks and flows. The total number of temporary migrants, as assessed by the
CSO of Poland, amounted to 2.27 million in January 2008. This was an increase
of more than 100% compared to earlier post-accession periods. However, since
2007 and 2008, the number of migrants has seemed to be on the decline reversing
the post-accession upward trend.
fluctuated around 40%, and in the potential demand fell in the bracket between 54 and
59% – depending on the size of the firm.
An important aspect of the emigration, which hasn’t altered throughout the

years following the regime change in 1989, is the fact that it has been predomi-
nantly labor migration. The massive scale and its connection to the labor market
are the two main features of the outflow. In this setting, there seems to be two di-
mensions of the effects of international mobility of Poles which are the most im-
portant: the impact on the population structure and the impact on the labor market.
We argue that in both cases, short- and mid-term consequences are not as impor-
tant as the possible long-term consequences of the outflow.
Regarding the structural characteristics of the extensive outflow from Poland,

the most important consequence of the EU accession seems to be the fact that the
flows have lost their old selectivity patterns in the post-accession period, attracting
new groups of potential actors. For example, the role of rural areas as sending en-
vironments increases; the wide spectrum of receiving economies allows migrants
of different education levels to find their ‘niches’, etc. This allows a more flexible
allocation of the Polish population in the longer run when individuals originating
from backward regions, or possessing education which is not adequately valued
by the local labor market drift abroad, allowing the modernization of the econ-
omy.
As far as the labor market consequences are concerned, we find that the short-
term effects of the outflow – labor shortages and changes in the level of unem-
ployment – and mid-term effects of the outflow of workers – such as wage pres-
sure – are not very pronounced. Although the data do support theoretical predic-
tions that mass emigration leads to increased labor shortages, a decrease in
unemployment and wage pressure, the overall effect of the outflow on the men-
tioned variables on the macro-scale is fairly small. Particular effects may certainly
include severe shortages in the local labor market or in specific sectors, for exam-
ple, those requiring highly qualified workers, but lower than the overall economy
level.
It is important to stress that the effect of the extensive outflow ‘disappears’ in

the overall effect of a very favorable business cycle phase experienced by Poland
in the post-accession period. The favorable economic climate has led to unprece-
dented low levels of unemployment, high levels of employment and labor short-
ages; employers also experienced substantial wage pressure; and there was an in-
crease in the demand for labor in other countries acting as pull factor for
migration. However, it only took a couple of months following a world crisis,
which affected Poland’s most important economic partner, the EU, for labor mar-
ket trends in Poland to reverse. The latest data suggest a decline in the demand for
labor and the concerns of labor shortage among entrepreneurs, not to mention a
cut in the wage pressure.
In the above light, even mass migration may not be as important to the labor
market in the short run, as it may prove in the long run. This is due to the fact that
in the longer perspective, emigration may lead to the outflow of the ‘economically
redundant’ population, which, according to some theories (Kaczmarczyk and
Okólski 2008a) is a necessary condition for the process of the modernization of an
economy to complete. However, for this effect to take place, the ‘reversing’ of this
process, i.e. mass return migration (especially in view of recent economic devel-
opments), can not happen in an unconditional manner; whether this will be the
case is an open question.
References
Aydemir, A./G. J. Borjas (2006), A Comparative Analysis of the Labor Market Impact of
International Migration: Canada, Mexico, and the United States. NBER Working Paper
No. 12327.
Beine, M./F. Docquier/H. Rapoport (2001), Brain drain and economic growth: theory and
evidence, in: Journal of Development Economics 64 (1), 275-289.
Bhagwati, J./K. Hamada (1974), The brain drain, international integration of markets for
professionals and unemployment, in: Journal of Development Economics 1 (1), 19-42.
Borjas, G. (2004), Labor Economics, Boston.
Budnik, K. B. (2008), Search Equilibrium with Migration: the Case of Poland. National
Bank of Poland Working Paper No. 45. Warsaw.
Budnik, K. B. (2007), Migration Flows and Labour Market in Poland. National Bank of Po-
land Working Paper No. 44. Warsaw.
Bukowski, M./G. Koloch/P. Lewandowski (2008), Labour market macrostructure in NMS8
– shocks and institutions, in: M. Bukowski (ed.) Employment in Poland 2007 – Safety
on the flexible labour market. Warsaw: MPiPS.
CSO (2008), Informacja o rozmiarach i kierunkach emigracji z Polski w latach 2004-2007
[Information on the scale and directions of emigration from Poland in 2004-2007]. War-
saw: Central Statistical Office Poland.
Docquier, F./A. Marfouk (2004), Measuring the international mobility of skilled workers
(1900-2000) – Release 1.0. Policy Research Working Paper No. 3381, World Bank.
Drinkwater, S./J. Eade/M. Garapich (2006), Poles Apart? EU Enlargement and the Labour
Market Outcomes of Immigrants in the UK. IZA Discussion Paper No. 2410, Bonn.
Dumont, J.-C./G. Lemaître (2005), Counting immigrants and expatriates in OECD coun-
tries: a new perspective, mimeo: OECD.
Fihel, A./P. Kaczmarczyk/J. Mackiewicz-Lyziak/M. Okólski (2008), Labour Mobility
within the EU in the Context of Enlargement and the Functioning of the Transitional
Arrangements, Country Study: Poland”. Centre of Migration Reserach, University of
Warsaw.
Fihel, A./P. Kaczmarczyk (2008), Migration: A Threat or a Chance? Recent Migration of
Poles and its Impact on the Polish Labour Market, in: K. Burrell (ed.), Polish Migration
to the UK in the 'New' European Union: After 2004. London 2009, 23-48.
Grabowska-Lusińska, I./A. Żylicz (eds.) (2008), Czy Polska gospodarka potrzebuje cud-
zoziemców? [Does Polish economy need foreign workers?], Warsaw: OBM UW.
Grubel, H. B./A. D. Scott (1966), The international flow of human capital, in: The Ameri-
can Economic Review 56 (1/2), 268-274.
Hanson, G. H. (2005), Emigration, Labor Supply, and Earnings in Mexico. NBER Working
Paper No.11412.
Home Office (2008), Accession Monitoring Report May 2004 – September 2007. London.
IOM (2005), Migration and Development, World Migration Report Series. Geneva: Inter-
national Organization for Migration.
Kaczmarczyk, P. (2005), Migracje zarobkowe Polaków w dobie przemian [Labour migra-
tion of Poles in the age of transition]. Warsaw: WUW.
Kaczmarczyk, P. (2008), Współczesne migracje Polaków – aspekty regionalne i lokalne
[Recent migration of Poles – regional and local aspects]. Warsaw: WNE UW.
Kaczmarczyk, P./M. Okólski (2008a), Demographic and labour market impacts of migra-
tion on Poland, in: Oxford Review of Economic Policy, 24 (3), 600-625.
Kaczmarczyk, P./M. Okólski (eds.) (2008b), Polityka migracyjna jako instrument promocji
zatrudnienia i ograniczania bezrobocia [Migration policy and labour market change].
Warsaw: WNE UW.
Kaczmarczyk, P./M. Okólski (2002), From net emigration to net immigration. Socio-
economic aspects of international population movements in Poland, in: R. Rotte/P. Stein
(eds.), Migration Policy and the Economy: International Experiences, München: Hanns
Seidel Stiftung.
KPMG (2008), Migracje pracowników – szansa czy zagrożenie? [Workforce Migration –
Opportunity or threat?]. Warsaw.
Layard, R./O. J. Blanchard/R. Dornbusch/P. Krugman (1994), East-West Migration: The
Alternatives. Cambridge, MA.
Lo Turco, A./A. Parteka (2008), EU Enlargement, Economic Interdependence and the La-
bor Markets in Old and New Member States (Italian Ministry of Education, University
and Research Scientific Research Programs of National Relevance PRIN2007 on "EU
policies, economic and trade integration processes and WTO negotiation"). Paper pre-
sented at the National Bank of Poland's conference "Migration, Labour Market and
Economic Growth in Europe after Enlargement" (Warsaw 8-9 December 2008).
Mioduszewska, M. (2008), Najnowsze migracje z Polski w świetle danych Badania Akty-
wności Ekonomicznej Ludności [Recent migration from Poland according to the Polish
LFS data], CMR Working Papers 36/94.
Mioduszewska, M./M. Okólski (2008), Direct demographic consequences of post-accession
migration for Poland, in: R. Black/G. Engbersen/M. Okolski/C. Pantiru (eds.), EU
enlargement and labour migration within the EU. Amsterdam (forthcoming).
Mishra, P. (2007), Emigration and Wages in Source Countries: Evidence from Mexico, in:
Journal of Development Economics 82 (1), 180-199.
NBP (2008), Informacja o kondycji sektora przedsiębiorstw ze szczególnym uwzględnie-
niem stanu koniunktury w IV kw. 2008. Warszawa: Narodowy Bank Polski. Departa-
ment Analiz Makroekonomicznych I Strukturalnych.
OECD (2005), Trends in International Migration: SOPEMI 2004. Paris: OECD.
Stark, O./C. Helmenstein/A. Prskawetz (1997), A brain gain with a brain drain in: Econo-
mics Letters 55 (2), 227-234.
10. The Post-Enlargement Migration Experience
in the Baltic Labor Markets
Mihails Hazans
Kaia Philips
In this chapter, we use Lithuanian, Latvian and Estonian LFS data (2002-2007)
complemented with several other surveys to compare the profile of Baltic tempo-
rary workers abroad before and after EU accession with that of stayers and re-
turn migrants. Determinants of migration and return, as well as selection issues
are discussed. Post-enlargement migrants from all three countries were signifi-
cantly less educated than stayers. Other things equal, medium-educated workers
were most likely to move after accession, and human capital became increasingly
less pro-migration over time. Return migrants differ from all movers in many
ways and, in particular, are more educated. Although brain drain was not a fea-
ture of post-accession Baltic migration, brain waste was: during 2006-2007, the
proportion of overqualified among high-educated movers ranged from five out of
ten for Latvia to seven out of ten for Lithuania, but it was around one fifth among
high-educated stayers in all three countries. We find that the free movement of la-
bor partially introduced in 2004 (and expanded in 2006) for EU citizens, although
excluding Baltic non-citizens, brought about significant changes in how ethnicity
and citizenship affect workers’ mobility. We conclude by discussing migration
perspectives in the context of recession.
Migration, DOI 10.1007/978-3-642-02242-5_10,
256 Mihails Hazans and Kaia Philips
1 Introduction
The four year period after the EU enlargement of 2004 was an unusual time for
the three Baltic countries. Economic growth had been fostered by several factors
including the benefits of the common market; substantial injections from EU
structural funds; and strong domestic demand driven by a credit boom in both the
real estate and consumer durables markets. Overall real GDP growth during 2004-
2007 was 36, 38 and 48% in Lithuania, Estonia and Latvia, respectively. Unem-
ployment fell and total employment rose in each of the three countries, although in
varying degrees.
One of the defining features of this period was the outflow of a significant part
of the labor force: thousands of workers were quick to use the right to work in the
United Kingdom, Ireland, Sweden, and later, other countries of the ‘old’ Europe.
This negative demographic shock was very painful for the Baltic countries, which
lost significant proportions of their populations (from 6% for Lithuania to 13-14%
for Latvia and Estonia) during the 15 years preceding accession because of nega-
tive net migration and, except for Lithuania, negative natural increase (World
Bank 2007: p. 115). Large and growing Lithuanian and Latvian diasporas have
emerged in Ireland, where they are among the top five groups of foreign nationals,
and the UK. Estonian diaspora in Finland is second only to the Russian one and
has been first by net migration since 2004. Flows of return migrants attracted by
improving possibilities at home, flows of new migrants triggered by the network
effect, as well as repeated and seasonal migration has strengthened the informal
links between the diasporas and home population.
Migration has contributed to a decline in unemployment and real wage growth

in general, as well as improvements in the labor market position of ethnic minori-
ties and the low-skilled. Through these channels and through remittances migra-
tion has thus helped to improve living standards. However, migration-induced la-
bor shortages in some sectors have been obstacles to growth. Migration has also
changed the distribution of bargaining power in the labor marker in favor of em-
ployees. While this change was often a necessary correction in countries with
weak unions (see Masso and Eamets 2006: Table 5.5), in some cases (notably
construction) it went too far and contributed to labor costs rising too fast.
Some of the migration effects, both good and bad, went beyond the labor mar-
ket. One example is deteriorating (at least temporarily) quality and in some cases
falling productivity in sectors affected by labor shortages, e.g. retail, catering,
construction and health. On the other hand, labor shortages motivate firms to in-
vest in new capital and improve human resource management, which leads to in-
creases in productivity.
10. The Post-Enlargement Migration Experience in the Baltic Labor Markets 257
One of the most intriguing questions about the post-enlargement migration

wave from the Baltic countries is the behavior of Russian-speaking minorities,
many of whom do not have citizenship in Estonia and Latvia; and hence, they are
not covered by the free mobility provisions.
The above discussion highlights the prominent role of post-enlargement migra-

tion for the Baltic countries (and for the countries which host Baltic migrants)
from demographic, economic and social perspectives. In this chapter, we discuss
in detail the dynamics and composition of the Baltic migration and return migra-
tion flows before and after EU accession. We also look at the labor market out-
comes of migrants and return migrants and at migration impacts on the Baltic
economies and societies.
We refer to Kahanec et al. (chapter 1 in this volume) for an overview of the lit-
erature on post-enlargement migration within EU. Few papers have looked spe-
cifically at the Baltic countries in this context. A study by Hazans (2003a) based
on cross-country comparison of inter-regional migration rates concludes that the
Baltic populations are more mobile than those of the Czech Republic, Slovakia
and Slovenia. Hazans (2003b) employs a survey of internet users to study their
migration intensions, as well as the determinants of temporary and permanent mi-
gration. Kallaste and Philips (2004), SKDS (2006a) and Järv (2007) study migra-
tion intensions in Estonia and Latvia. Kaczmarczyk and Okólsky (2008) discuss
the economic impact of migration in Poland and the Baltic countries. The Univer-
sity of Latvia (2007) analyze migration intentions and foreign work experience of
Latvian residents and their relatives abroad. Indans et al. (2006) provide an inter-
view-based account of life and thoughts of Latvians working in Ireland. Indans
and Kruma (2006), SKDS (2006a), Kadziauskas (2007) and Nurmela (2008) look
at migration policies and attitudes. Hazans (2008) studies labor market outcomes
of return migrants in Latvia.
The rest of the chapter is organized as follows. Section 2 describes the data
sources. The dynamics and intensity of the Baltic migration flows before and after
accession are presented in section 3. Sections 4 and 5 are devoted to migrant hu-
man capital and their labor market outcomes. Return migration and labor market
transitions are analyzed in section 6, followed by a discussion of the migration
impact on the Baltic economies in section 7. Section 8 deals with the demographic
composition of migrant flows and stocks, as well as with determinants of migra-
tion and returning. Section 9 discusses remittances. The main findings are summa-
rized in section 10. The concluding section looks into the future of Baltic migra-
tion in the context of economic crisis.
2 Data
In order to describe various aspects of the migration process before and after EU
accession, as well as to provide cross-country comparisons, we use several com-
plementary data sources. Administrative data on gross migration flows into the
UK (National Insurance numbers and Worker Registration Scheme), Ireland (Per-
sonal Public Service Numbers) and other EEA countries (international migration
statistics) are used to measure the intensity of the migration and outline a migrant
profile along basic demographic and other available dimensions. For Ireland, we
also use the 2006 Population Census data.
For a deeper analysis we exploit Lithuanian, Latvian and Estonian LFS data
files (2002-2007). LFS data include a sufficient number of observations on mi-
grant workers abroad: 1844 for Lithuania, 577 for Latvia, and 800 for Estonia.
These numbers increase further if those with some economic activity abroad
within the last year are considered. The data also has a sufficient number of ob-
servations on return migrants with foreign work experience: 313 for Lithuania,
185 for Latvia, and 136 for Estonia. By design, the data do not cover migrants
who have already become permanent, i.e. are not considered as household mem-
bers at home, and those with no family left behind. However, the distribution of
movers found in pre-accession and post-accession LFS data by host countries is
consistent with the geography of actual migration flows. The UK and Ireland have
become major destinations for Lithuanian and Latvian movers, and Finland plays
a similar role for Estonia (see Tables 1a, 1b). Table 1b also confirms that labor
migration to countries outside the EEA accounted for less than 10% of the total in
both 2002-2003 and after accession. This justifies our decision to restrict analysis
of administrative data to EEA countries. While formally sending countries’ LFS
data capture only temporary migrants, one should keep in mind that the post-
accession migration from EU10 countries has been predominantly temporary so
far (see Pollard et al., 2008: Figure 20 and pp. 39-40; Barrell et al., 2007: p. 5;
European Commission, 2008: pp. 121-122), and that temporary migrants tend to
become permanent (see e.g. van Baalen and Müller, 2009, for Germany). This
suggests that our LFS data are suitable for cross-country comparisons, for analyz-
ing accession-induced and post-accession changes in migrant characteristics, and
for studying migrant labor market transitions. Whenever possible, we also support
our findings with the Irish Census data, which include both temporary and perma-
nent migrants.
Table 1a
Gross flows of adult migrants from the Baltic countries to EU-25 and Norway
by destination (2003-2007)
Lithuania Latvia Estonia
2003 2005 2007 2003 2005 2007 2003 2005 2007
UK 25.5 50.6 51.3 11.9 51.4 53.7 5.5 37.0 24.6
Ireland 19.3 31.3 22.3 25.2 34.2 24.3 15.7 23.8 8.6
Germany 25.9 8.4 8.7 36.9 8.6 9.3 25.3 7.9 8.1
Sweden 1.7 1.0 1.9 2.9 0.7 1.8 7.4 3.8 5.4
Norway 2.0 1.2 5.0 2.7 0.6 2.5 2.4 2.2 5.3
Finland 0.2 0.1 0.2 0.9 0.3 0.6 29.4 18.3 38.6
Denmark 5.0 1.8 2.5 6.3 1.5 2.7 3.9 1.9 2.6
Spain 13.2 4.2 5.8 4.8 1.1 2.5 2.1 1.4 2.4
Other EU15 2.8 1.0 1.4 3.1 0.8 1.6 5.5 1.6 2.3
Other EU10 4.2 0.6 1.0 5.3 1.0 1.0 2.8 2.1 2.0
Total 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0
UK + Ireland 44.8 81.8 73.6 37.1 85.6 78.0 21.1 60.8 33.2
Nordic countries 9.0 4.0 9.6 12.7 3.0 7.5 43.1 26.2 51.9
Source: See Notes to Figure 1.
Notes: Posted workers are not included. See Notes to Table 1c.
Table 1b
Temporary migrant workers from the Baltic countries by destinaton (LFS data)
2002-2003 2004-2006 2002-2003 2004-2007 2005-2006
UK + Ireland 27.8 51.4 31.8 63.3 7.6
Nordic countries 11.8 10.6 12.9 9.5 75.4
Germany 15.7 11.4 9.0 5.4 2.0
Other EEA 24.1 12.1 14.4 11.9 4.5
US+Canada 9.2 6.6 11.9 2.5 1.6
Russia+Ukraine+Belarus 8.9 5.0 14.4 5.3 5.8
Other 2.5 2.9 5.5 2.1 3.0
Total 100 100 100 100 100
N obs. 703 1141 201 415 290
Source: Calculations with Lithuanian, Latvian and Estonian LFS data.
Notes: Data on country of work were not available for Estonian LFS 2002-2004 and 2007. EEA: as of 2006, i.e. including
EU10
When relevant, we also employ data from representative surveys conducted in

Estonia in 2003 and 2006 (see Kallaste and Philips, 2004; Järv, 2007) and Latvia
in 2005-2008 (see Notes to Tables 2, 4), which include information on migration
intentions and foreign work experience. Some of these are based on very large
samples and hence allow for a detailed analysis when LFS data are not sufficient.
For instance, one of the Latvian surveys includes data on more than a thousand
relatives economically active abroad, with more than half having moved after ac-
cession. Another sample of over 10 thousand respondents includes data on 467
persons with post-accession foreign work experience. Results presented here are
mostly based on our own calculations.
3 Intensity and Main Destinations

of the Post-Enlargement Migration
The removal of restrictions on labor mobility of those from the new member states
by the UK and Ireland in 2004 triggered an almost immediate migration response.
As is now well documented (see e.g. Baas et al., chapter 2 in this volume; Euro-
pean Commission 2008: ch. 3; UK Home Office 2008), the majority of migrants
were temporary, staying in the receiving countries from several months to several
years. In most cases, migrants from the Baltic countries did not withdraw from
population registers in their home countries. Therefore, standard emigration rates
evaluated in the sending countries are not suitable for analysis. Various surveys,
including LFS, conducted in sending countries tend to under-sample migrant
households. Receiving country statistics is more relevant. One approach (Baas et
al., chapter 2 in this volume; see also European Commission, 2008) is to look at
the dynamics of stocks of new member states nationals in the receiving countries
using LFS data. According to this methodology, during the four post-enlargement
years (2004-2007) migrants to other EU countries accounted for 3.1% of the
working-age population in Lithuania, 2% in Poland and Slovakia, 1.3% in Latvia,
1% in Estonia and about 0.5% in the Czech Republic and Hungary (European
Commission 2008: Chart 3).
From the sending countries’ perspective, however, gross flows of migrants are
also of considerable interest as a measure of the proportion of the population in-
volved in the intra-EU mobility of labor. The dynamics of gross migration flows
from the three Baltic countries and, for comparison, their Central European 2004
enlargement companions (Poland, Slovakia, the Czech Republic and Hungary) in
2003-2008 is summarized in Figure 1.1 The upper panel features flows to the UK
and Ireland. In both cases, flows are measured by the number of social security
numbers allocated to adult nationals of the new member states in question. These
statistics are perfectly comparable across the two largest destinations of post-
enlargement migrants2, as well as across sending countries. Moreover, when ac-
cumulated over years, it has a straightforward interpretation as the total number of
individuals with foreign experience. Bearing in mind the very high activity rates
among migrants from EU8 countries (see below), we use the sending country’s
1 Slovenia, which had much lower emigration rates, is not shown in Figure 1. For the Baltic
countries more details, including the absolute sizes of the flows, are reported in Table 1c.
2 This is one reason for choosing British national insurance numbers (NINO) over the
Worker Registration Scheme (WRS) data. In addition, WRS was launched in 2004, and
this year’s data include up to 40% of migrants who arrived in previous years, whereas
NINO statistics (which is also available for previous years) is less distorted by this effect.
Another advantage of NINO from the Baltic perspective is that Latvian and Estonian mi-
grants who do not hold citizenship of these countries (see Zimmerman et al. 2008 for de-
tails) are not eligible for WRS but are eligible for NINO if they have a work permit. Finally,
NINO also covers the self-employed.
labor force (rather than working-age population) before enlargement as a refer-

ence point when measuring these flows.3
Several noteworthy facts are evident from Figure 1 (upper panel). Pre-
accession (year 2003) migrant flows from the Baltic countries to the UK and Ire-
land were nonzero but quite small, accounting in total for a third of 1% of the la-
bor force in Lithuania, a sixth of 1% in Latvia, and a ninth of 1% in Estonia.
Flows from Poland, the Slovak and Czech Republics and Hungary were even
smaller. In response to opening the British and Irish labor markets in 2004, flows
from all of the above mentioned countries increased sharply, reaching 1.5% in
Lithuania, 0.9% in Latvia and 0.4 to 0.5% in Estonia, Slovakia and Poland. Flows
from Hungary and the Czech Republic increased from almost zero to the levels
observed in Estonia and Latvia before enlargement. The network effect seems to
work well: the summary flows to UK and Ireland from the Baltic countries and the
Czech Republic roughly doubled in 2005, and the flows from Poland, Slovakia
and Hungary tripled.
In the following three years, Baltic flows declined steadily, quite likely due to
rising earnings and falling unemployment at home. However, these flows re-
mained relatively high in Lithuania (1.4%) and Latvia (1%) in 2008; the Estonian
flow fell to 0.3%. Note, however, that the main destination of Estonian migrant
workers, unlike their Latvian and Lithuanian colleagues, was the Nordic countries
rather than the UK and Ireland (Figure 1, lower panel). Flows from Poland, Slo-
vakia and Hungary featured quite a different dynamics: they continued to increase
until a peak in 2007 and fell only in 2008 to a level below that of Poland and Slo-
vakia in 2005.
A total of 10.1% of economically active Lithuanians moved (at least temporar-

ily) to the UK and Ireland during 2004-2008; for Latvia, Poland and Slovakia, this
proportion was 6.7, 6.2 and 5.7%, respectively; for Estonia, Hungary and the
Czech Republic, it was 2.3, 1.6 and 1.3%. In other words, the Baltic nations
ranked 1, 2 and 5 among the EU8 countries in terms of mobility to the UK and
Ireland.
3 Using the working-age population would hardly affect the comparison between the Baltic
countries, Slovakia and the Czech Republic; however, Polish and Hungarian flows would
fall by 7 and 13% relative to Lithuanian ones.
Figure 1
Gross flows of adult migrants from the Baltic countries and Central Europe to UK and
Ireland (2003-2008, top), selected other EU countries and Norway (2003-2007, bottom),
per 1000 economically active population of sending country in 2004
30
25
20
migrants per 1000 active population
15
10
0
2003
2004
2005
2006
2007
2008
2003
2004
2005
2006
2007
2008
2003
2004
2005
2006
2007
2008
2003
2004
2005
2006
2007
2008
2003
2004
2005
2006
2007
2008
2003
2004
2005
2006
2007
2008
2003
2004
2005
2006
2007
2008
average 2004-2008
average 2004-2008
average 2004-2008
average 2004-2008
average 2004-2008
average 2004-2008
average 2004-2008
Lithuania Latv ia Estonia Poland Slov akia Hungary Czech R.
UK Ireland
12
10
8
migrants per 1000 active population
0
2003
2004
2005
2006
2007
2003
2004
2005
2006
2007
2003
2004
2005
2006
2007
2003
2004
2005
2006
2007
2003
2004
2005
2006
2007
2003
2004
2005
2006
2007
2003
2004
2005
2006
2007
average 2004-2007
average 2004-2007
average 2004-2007
average 2004-2007
average 2004-2007
average 2004-2007
average 2004-2007
Lithuania Latv ia Estonia Poland Slov akia Hungary Czech R.
DE SE+FI+NO+DK ES+AT+NL+FR+CY
Source: UK – allocation of National insurance numbers, www.dwp.gov.uk/asd/tabtool.asp;

Ireland – allocation of Personal Public Service Numbers, http://www.welfare.ie/topics/ppsn/ppsstat.html
Other countries – migration statistics provided by Eurostat or receiving countries.
Notes: Available (incomplete) data suggest that flows to EU countries not included in the Figure are negligible, except for
flow from Poland to Italy. Workers sent by their employers to work in another country for a period of less than 12 months
(posted workers) in most cases are not covered by migration statistics and not shown in Figure 1. In 2006 number of posted
workers from Estonia (mostly to Finland) and Czech R. was twice as big as the gross migration flow to continental Europe;
for Poland, Slovak R. and Hungary the two flows were roughly of the same size; number of postings from Lithuania and Lat-
via (mostly to Norway) was similar to migration flow to Germany. See Table 1c for more details.
The lower panel of Figure 1 features annual flows of adults (15+) to Germany,
Nordic countries (Sweden, Denmark, Finland, and Norway), Spain, Austria, the
Netherlands, Cyprus and France.4 It appears that migration from the Baltic coun-
tries, especially Lithuania, to non-English speaking countries has been more geo-
graphically diversified than migration from Central Europe, which is heavily con-
centrated towards Germany. This panel is based on immigration statistics which,
unlike British and Irish data, cannot be simply aggregated over time because the
same person can be counted more than once. However, the data allow comparison
of migration intensity across countries and over time. Moreover, assuming that the
proportion of repeated moves within the same year is small, these flows could be
roughly compared with the flows to the UK and Ireland.
The data suggest that immediately before enlargement, flows from the Baltic
countries to the continental part of the ‘old’ Europe were larger than flows to the
UK and Ireland. This pattern was even more pronounced for Central European
countries. Interestingly enough, when the UK, Ireland and Sweden opened their
labor markets for the EU8 countries in 2004, total flow to other countries of the
old Europe increased in size as well, albeit less sharply. A closer look reveals that
flow towards the Nordic countries increased steadily for each of the Baltic coun-
tries and Poland. Flows to Germany from Lithuania, Latvia and Slovakia in-
creased only temporarily and returned to levels observed in 2003 or lower by
2007. Estonian–German and Czech–German flows declined. By contrast, flows
from Poland and Hungary to Germany remained well above the pre-accession
level throughout the whole period between 2004 and 2007. In terms of overall in-
tensity of migration towards continental old Europe, Poland is the clear leader,
with an average annual gross outflow equivalent to 1% of the labor force during
2004-20075, followed by Lithuania, Slovakia, Hungary and Estonia, with annual
outflow between 0.67 and 0.57%.
The combined evidence from both panels of Figure 1 suggests that Lithuanians
have been the most mobile nation among those which joined the EU in 2004. Po-
land ranks second with Latvia and Slovakia slightly behind with very close re-
sults. Estonia, Hungary and especially the Czech Republic have seen much
smaller outflow rates. These results are broadly consistent with those by Baas et
al. (chapter 2 this volume) based on stock estimates and with countries’ pre-
4 Available (incomplete) data suggest that including flows from the Baltic and Central Euro-
pean countries to the remaining EU15 countries (Belgium, Luxemburg, Italy, Greece and
Portugal) and flows between new member states would not change the results qualita-
tively, although flows to Italy from Lithuania and especially Poland are substantial
(Kaczmarczyk and Okólsky, 2008 estimate the flow to Italy to be about 8% of the total out-
flow).
5 This rate may well be even higher because the number of Polish seasonal workers in
Germany (European Commission, 2008: Table 7) exceeds immigration flow reported by
Eurostat (used in our analysis) by at least a factor of 2 during 2004-2005 and by at least a
factor of 1.6 during 2006-2007.
accession income and unemployment levels (see Kahanec et al., chapter 1 in this
volume), however, flow-based ranking ‘favors’ Poland and Latvia and suggests
that migration from these countries has been of more temporary nature on average
than migration from Lithuania, Poland and Slovakia. As far as the three Baltic
countries are concerned, higher mobility by Lithuanians might have to do with a
very low share of unemployed receiving benefits (Hazans 2007a: Fig. 2) – on top
of unemployment rate being initially higher by two points than in Estonia and
Latvia.
Two more remarks should be made regarding the cross-country comparisons.

First, as concerns individual mobility, population covered by the free mobility
provisions (rather than the total population) might be a more appropriate base for
calculating the migration rates. This amounts to excluding Latvian and Estonian
non-citizens, resulting in an increase of emigration rates for these countries by a
quarter (Table 1c, col. [6]). Latvia thus outranks Poland and Slovak Republic in
terms of individual mobility under this approach. Second, if the interest is in the
total number of country’s nationals working abroad, then posted workers should
be considered as well (although they are not even temporary migrants in the sense
that they have moved not on their own initiative). For some sending countries (in-
cluding Estonia, see col. [3] in Table 1c) posted workers outnumber migrant
workers, yet usually they are not covered by migration statistics.6 Among the new
member states, Estonia and Poland have the highest posting rates: 1.2% of the la-
bor force in 2006 (European Commission 2008: Ch. 3, Chart 4); whereas for other
EU8 countries these rates ranged between 0.6% (Slovakia) to 0.2% (Lithuania and
Latvia). When posted workers are accounted for, it appears that during 2004-2007
Estonia, relative to its labor force, was sending abroad more workers than Latvia
(Table 1c, col. [8]).
Distribution of gross migration flows from the Baltic countries by destination is

summarized in Table 1a. After enlargement, combined flows to the UK and Ire-
land accounted for at least 80% of the total outflows from Lithuania and Latvia;
for Estonia this proportion reached 60% in 2005 but fell to 34% in 2007. The
Germany share was quite stable at 8 to 10% over time and across the three coun-
tries. Between 2005 and 2007, the Nordic countries share increased from 3 to 7%
for Latvia, from 4 to 10% for Lithuania and from 26 to 53% for Estonia. In the lat-
ter case, Finland played a prominent role due to a common border and similar lan-
guage: about 40% of gross migration flows from Estonia in 2007 ended up in
Finland, which lifted restrictions for workers from the EU10 in 2006. With the ex-
ception of Spain, which received 4 to 6% of total outflows from Lithuania, indi-
vidual shares of other countries after enlargement have been quite small.
6 This is clear for Estonia, in which case the number of posted workers is much larger than
the number of Estonian migrants to Finland in the same year. Similarly, number of Latvian
workers posted in Norway is about ten times larger than the corresponding migrants' flow.
Table 1c
Gross flows of adult migrants and posted workers from the Baltic countries
to EU25 and Norway (2003-2008)
[1] [2] [3] [4] [5] [6] [7] [8]
Adult migrants
Posted
UK+Ireland Other EEA workers a Total Adult migrants as % of Total as % of
1000 1000
Labor force (2004)
Excl. non- Working-age
All citizens population (2004)
Lithuania
2004 23.4 10.5 0.7 34.6 2.1 2.1 1.4 1.5
2005 47.1 11.0 1.9 60.0 3.6 3.6 2.5 2.6
2006 39.0 11.6 3.4 54.0 3.1 3.1 2.2 2.3
2007 31.9 11.2 4.7 47.8 2.7 2.7 1.8 2.0
2008 21.8 NA NA
b b b b b b b
Total 163.2
Average 32.6 11.1 2.7 49.1 2.9 2.9 1.3 2.1
Latvia
2004 9.9 3.9 0.3 14.1 1.2 1.5 0.8 0.9
2005 22.5 4.1 3.8 30.3 2.3 2.9 1.6 1.7
2006 18.7 3.9 2.3 24.9 2.0 2.5 1.4 1.5
2007 13.5 3.6 2.3 19.4 1.5 1.9 1.0 1.2
2008 11.0 NA NA
b b b b b b b
Total 75.7
Average 15.1 3.9 2.2 22.2 1.8 2.2 1.3 1.4
Estonia
2004 2.8 3.5 7.0 13.2 1.0 1.2 0.7 1.5
2005 4.9 3.3 7.0 15.1 1.3 1.6 0.9 1.7
2006 3.5 3.9 8.0 15.2 1.1 1.4 0.8 1.7
2007 2.3 4.5 6.0 12.6 1.0 1.3 0.7 1.4
2008 1.9 NA NA
b b b b b b b
Total 15.3
Average 3.1 3.8 7.0 14.0 1.1 1.4 0.8 1.5
Source: Migration flows - see Notes to Figure 1. Posted workers: Administrative data of State Social Insur-
ance Fund Board of Lithuania, State Social Insurance Agency of Latvia and Estonian Ministry of Social Af-
fairs; Estonian data for 2004-2005 are estimates.
Notes: a) Posted workers are those sent by employers to work in another country for a period of less than 12 months. Major-
ity of Lithuanian and Latvian posted workers are in Norway, followed by Germany, Nordic countries (excl. Norway), Spain,
France and Italy. Almost all Estonian posted workers are in Finland. Posted workers usually are not covered by migration
statistics.
b) To avoid double counting, aggregation over time is not performed. The nature of the UK and Ireland's data excludes dou-
ble counting within each country. Aggregation of these flows performed on the assumption (based on evidence from the LFS
data) that the number of cases when the same migrant has received social security number in both countries is negligible.
It is instructive to compare the intensity of migration from the Baltic countries

and its dynamics with survey evidence on migration intensions presented for Es-
tonia and Latvia in Table 2 (p. 268). At the end of 2006, about 6% of working age
residents of Estonia revealed ‘having a specific plan’ to work abroad in the future,
while more than 10% of their Latvian counterparts confirmed ‘having made cer-
tain preparations’ towards working abroad in the next year in a survey conducted
between September 2005 and March 2006. One year later, 8% of economically ac-
tive respondents were ‘ready to go abroad [in the near future] in order to get a
suitable job.’ Taking into account the differences in time horizons, this indicates a
somewhat higher potential mobility in Latvia than in Estonia, as was the case in
reality. The dynamics of intentions after enlargement, which were available for
Latvia, seems to be consistent with reality as well: the proportion of potential mi-
grants peaked at the end of 2005 and fell a year later. Note that when migration
potential is measured without reference to specific steps made, the proportions of
potential movers in 2005-2006 reached 20 to 26%, which is well above the ob-
served mobility rates.
4 Human Capital of the Migrants
Proportions of former unemployed and inactive individuals among migrants serve

as important indicators of the type of labor market experience they bring to the
host country as well as provide lower bounds to the migration effect on unem-
ployment and participation rates in the sending countries. In the first two years af-
ter the EU enlargement of 2004, 11 to 13% of migrants from Lithuania and Esto-
nia and 15% of their Latvian counterparts were unemployed in the home country
in the previous year and about 7% were either students or pupils. These propor-
tions exceed the ones observed among stayers by a factor of three to four, indicat-
ing that work abroad has been an important coping strategy for the Baltic unem-
ployed or potential unemployed (see Figure 2). In the following two years, the
proportion of former unemployed stayed roughly unchanged among Lithuanian
migrants. It increased to 18% among Latvian migrants, which is almost double the
size of the level observed in 2002-2003, but decreased to just 5% among Estonian
migrants, which is the same as in 2002-2003. Given that both Latvia and Estonia
started with an unemployment rate of about 10% in 2003, one plausible explana-
tion for the different dynamics might be the fact (documented in Table 1a) that be-
tween 2005 and 2007, one quarter of Estonian migration flow switched from the
UK and Ireland to the Nordic countries.
From the labor market perspective (both in sending and receiving countries),
skill level and occupation are the most significant characteristics of the mi-
grants. Figure 3, based on Baltic LFS data for the years 2002-2007, compares
the skill composition of migrants from the Baltic countries, stayers and return
migrants. The Figure focuses on people aged 18 to 64 who have been economi-
cally active for some period of time during the last 12 months. Migrants were
not distinguished by destination country, but the vast majority of them had been
in Western or Northern Europe (see Table 1b), which is consistent with the re-
sults in Table 1a.
Figure 2
Employed migrants from the Baltic countries and stayers therein
by employment status 1 year ago (2002-2007)
Lithuania
100%
80%
60%
40%
20%
0%
2002-2003 2004-2005 2006 2002-2003 2004-2005 2006
migrants stayers stayers
Employed Unemployed Student Other
Latvia
100%
80%
60%
40%
20%
0%
2002-2003 2004-2005 2006-2007 2002-2003 2004-2005 2006-2007
Estonia
100%
80%
60%
40%
20%
0%
2002-2003 2004-2005 2006-2007 2002-2003 2004-2005 2006-2007

Notes: Number of migrant observations for the three sub-periods are 703, 788 and 350 for Lithuania; 170, 192 and 207 for
Latvia; 109, 168 and 523 for Estonia. Virtually all respondents who were unemployed or inactive a year ago have indicated
home country as their place of residence at that time. Posted workers, permanent migrants as well as those with no family
left behind are not covered due to survey design.
Figure 3
Economically active migrants from the Baltic countries, stayers therein
and return migrants by education level (2002-2007)
Lithuania
100%
80%
60%
40%
20%
0%
2002- 2004- 2006 2006, 2002- 2004- 2006 2002- 2004-
2003 2005 Ireland 2003 2005 2003 2006
temporary migrants all stayers return migrants

migrants
Low Medium High
Latvia
100%
80%
60%
40%
20%
0%
2002- 2004- 2006- 2006, 2002- 2004- 2006- 2002- 2004-
2003 2005 2007 Ireland 2003 2005 2007 2003 2007
temporary migrants all stayers return migrants

migrants
Low Medium High
Estonia
100%
80%
60%
40%
20%
0%
2002- 2004- 2006- 2002- 2004- 2006- 2004-
2003 2005 2007 2003 2005 2007 2007
temporary migrants stayers return

migrants
Low Medium High
Source: Calculations with Lithuanian, Latvian and Estonian LFS data; for Ireland –
census results in CSO Ireland (2008): Table A6.
Notes: Except for the Irish column, migrants are defined as household members who have been economically active (working or
seeking job) abroad for some time during the last year (hence, some overlap across periods is allowed); for return migrants the
reference period is two years. All data refer to persons aged 18 to 64. Number of migrant observations for the three sub-periods
is 802, 933 and 450 for Lithuania; 240, 258 and 273 for Latvia; 131, 189 and 589 for Estonia. Number of return migrant observa-
tions for the two sub-periods is 99 and 214 for Lithuania; 66 and 119 for Latvia; 20 (not shown) and 116 for Estonia. Data on mi-
grants in Ireland (N= 13087 for Lithuania, N = 8333 for Latvia) are based on Irish Population Census and refer to April 23, 2006.
Classification of education levels follows the approach of EU LFS: High refers to tertiary education, Medium – to upper secondary
education, and everything less is classified as Low.
Three messages are evident from the Figure. First, most of the migrants (two
thirds to three quarters) had secondary education. Second, enlargement changed the
skill composition of migrants: before enlargement, Lithuanian migrants had the
same skill distribution as stayers, while Latvian and Estonian migrants were more
educated on average than stayers. Post-accession migrants from all three countries
are significantly less educated than stayers, and the gap tends to increase over time.7
Third, post-accession return migrants are more educated than migrants, stayers and
pre-accession returnees.
Note that sending countries’ LFS data include only migrants still considered as
household members at home, i.e. mostly temporary migrants. However, the above
findings are also supported by other data sources. First, we also include in Figure 3
results from the Irish population Census 2006 based on both temporary and perma-
nent migrants (whose education has ceased) from Lithuania (more than 13 thou-
sands) and Latvia (more than 8 thousands). The low educated accounted for 20% of
Lithuanian migrants and 18% of their Latvian counterparts. Among economically
active stayers in the same group, these proportions were just 9% and 15%, respec-
tively. On the other hand, the incidence of high education was substantially lower
among migrants: 23 vs. 29.5% for Lithuania and 18 vs. 22% for Latvia.
The fact that incidence of tertiary education among Baltic migrants is lower than
among stayers contrasts with an opposite finding for migrants from the EU10 in
general (European Commission, 2008: Ch. 3, Chart 22).
Available data on migrants’ occupation before moving point in the same direc-
tion. Figure 4 (upper panel) compares the last occupation at home of employed mi-
grants from Latvia arriving in EEA countries in 2004-2005 and their counterparts ar-
riving before EU enlargement with the occupations of employed stayers in Latvia in
the respective periods (for better comparability the main occupation one year ago is
used for stayers). Among post-enlargement migrants, one finds more former manual
workers and fewer former highly skilled non-manual workers and students than
among migrants arriving in EEA countries during the previous years. When com-
pared to stayers, post-enlargement migrants feature higher proportions of former un-
skilled manual workers, low skilled non-manual (service and shop) workers and
former pupils or students, and a much lower proportion (one eighth as opposed to
one third) of former highly skilled non-manual workers. Before enlargement about
two fifths of migrants went to countries at that time outside the EEA. This was the
7 Fouarge and Ester (2007, Table 2 and 3) find the opposite when analyzing plans to move
to another EU country within next five years using the mobility module of the 2005 Euro-
barometer. This is not surprising because the wording of their question seems to imply
permanent migration and does not refer to working abroad. More than one third of poten-
tial movers from the Baltic countries and Poland in their samples are students. According
to LFS data, students who have actually worked abroad are a tiny minority among post-
accession migrant workers.
former Soviet Union in most, but not all, of the cases. These migrants were signifi-
cantly more skilled than those heading to Western Europe as well as than those who
stayed. The same pattern is found also in 2004-2005, although the number of such
migrants has fallen dramatically.
Figure 4
Migrants from Latvia by occupation in Latvia and abroad
Employed migrants from Latvia and stayers therein by occupation in Latvia
100%
80%
60%
40%
20%
0%
EEA (59%) RoW (41%) EEA (95%) Latvia Latvia
up until 2003 2004-2005 2002-2003 2004-2005
migrants stayers
Unskilled manual Skilled manual Low skilled nonmanual Highly skilled nonmanual Student Other
Migrants arrived from Latvia in 2004-2005 by occupation in Latvia and abroad

100%
80%
60%
40%
20%
0%
unskilled skilled manual low skilled highly skilled student other
manual nonmanual nonmanual
last occupation in Latvia
Unskilled manual abroad Skilled manual abroad Low skilled nonmanual abroad
Highly skilled nonmanual abroad Student abroad
Source: Migrants: calculations with (un-weighted) data of 'Relatives abroad' module of the survey on geo-
graphical mobility of population conducted by Data Serviss Ltd for the University of Latvia in 2005/Q1-2006/Q4
in the framework of the National Program of Labor Market Studies undertaken by the Latvian Ministry of Wel-
fare. Results should be treated as qualitative only due to: (i) unsolved weighting issue; (ii) limited number of
observations in columns; (iii) custom (rather than ISCO) classification of occupations used. Stayers: LFS data.
Notes: Upper panel: Last occupation in Latvia is shown for migrants. Main occupation one year ago is shown for stayers.
'Other' include unemployed. EEA: European Economic Area; RoW: Rest of the World. Only employed stayers aged 18 to 64
are included. Middle and lower panel: migrants whose occupation in Latvia or abroad is unknown as well as those not em-
ployed abroad are excluded. Number of migrant observations by panel and column:
Panel Col. 1 Col. 2 Col. 3 Col.4 Col.5 Col.6
Upper 260 203 505 - - -
Middle 80 147 113 72 99 41
Lower 30 123 66 120 126 32
Migrants from Latvia by occupation in Latvia and abroad
Migrants arrived from Latvia before 2004 by occupation in Latvia and abroad
100%
80%
60%
40%
20%
0%
unskilled skilled manual low skilled highly skilled student other
manual nonmanual nonmanual
last occupation in Latvia
Unskilled manual abroad Skilled manual abroad Low skilled nonmanual abroad
Highly skilled nonmanual abroad Student abroad
Yet another approach is to look at the composition of flows rather than stocks,
and this leads to similar conclusions. For instance, according to a survey on unde-
clared emigration conducted by Statistics Lithuania, the share of low educated mi-
grants aged 16+ increased from 8% during 2001-2003 to 18% during 2006-2007.8
Among economically active stayers of the same age, it decreased from 11 to 8%
over the same period. For Latvia and Estonia, we rely on the results of surveys of in-
tentions to work abroad conducted in 2005-2007 in Latvia and in 2003 and 2006 in
Estonia (Table 2). The majority of post-enlargement potential movers (57 to 70% in
Latvia, 64 to 74% in Estonia) have secondary education (Table 2, panel B). In Lat-
via, the structural break in composition of potential movers (from more to less edu-
cated than stayers) seems to have taken place in the second part of 2005 (see col.
[5]-[8] in Table 2, panel C). According to a large-scale survey conducted on the eve
of 2007, the proportion of low-skilled among potential movers exceeds that among
stayers by 5.3 points, whereas it is the opposite for the proportion of highly skilled.
This pattern also holds, albeit in a weaker form, when only potential movers who
have made some preparations are considered (see col. [9], [10] in Table 2, panel C).
A similar pattern was already found in Estonia in 2003, but became much stronger
in 2006: among potential movers aged 15-64 with a ‘specific plan’, just 9% are
highly educated. This proportion is twice as big among stayers (col. [1]-[3] in Table
2, panels B, C).
8 See http://db1.stat.gov.lt/statbank/SelectVarVal/Define.asp?Maintable=M3020109&PLanguage=1;
simple calculations have been made to arrive at the reported results; non-responses to the educa-
tion questions (13% in 2001 and 10% in 2006-2007) have been ignored.
Table 2
Intentions to work abroad, Estonia (2003, 2006) and Latvia (2005-2007)
Survey I II III IV V VI
[1] [2] [3] [4] [5] [6] [7] [8] [9] [10]
Country Estonia Latvia Latvia Latvia
Period 2003 2006/Q4 Jul Dec 2005/Q4- 2006/Q4-
2005 2005 2006/Q1 2007/Q1
Horizon Future Future Future Two One year Near future
years
Wording Plans Consider Have Ready Big Would like Have Ready
a plan prob. made (to get a
steps suitable job)
Age 15-64 15-64 18-65 18-65 18-65 15-65 15-65 15-65 18-65
Status All All All All Econ. active
b All All Econ. active
b
N obs. 1000 1505 884 796 5292 8005 9696 9517

A. Proportion of potential movers by socio-economic characteristics, %
All 42.2 25.3 5.7 15.4 26 19.8 19.7 10.5 8.5 8.2
Men 52 28.2 7.5 15.6 31.2 22.8 22.4 11.8 9.1 8.8
Women 33.1 22.7 4 15.3 21.3 16.4 17.1 9.3 7.9 7.6
Age
15(18)-24 70 51.3 8.8 29.3 52.3 42.3 36.6 18.1 18.5 18.5
25-34 46.6 29.2 8.3 17.1 34.1 22.5 21.8 12.5 11.3 11.3
35-44 47.5 20.9 4.9 13.7 22.9 17.8 17.4 10.9 6.7 6.7
45-54 25.3 13 3.3 12.3 17.7 14.6 14.1 7.3 5.3 5.3
55-64 11 4.6 2.3 5.6 6.3 8.4 5.8 2.5 1.7 1.7
Ethnicity
Titular 40.5 26.1 5.1 13.7 24.3 19.2 19.6 10.9 8.6 8.4
Minority: All
45.5 17.3 8.2 17.8 28.3 20.7 19.9 10 8.3 7.9
Citizens 47 21.2 13.7 18.7 30.4 20.9 21.6 11.9 9.7 9.2
Noncitizens 44 13.8 3 16.4 25.7 20.5 17.6 7.6 6.2 6
Education
low 42.9 31.7 5.2 10.6 31.6 25.6 23.9 11.6 10.7 9.3
medium 43.3 25.1 6.6 15.6 26.7 19.7 20.3 10.7 8.2 8.2
high 36.6 17.7 3.1 17.3 21.4 17.6 17.1 9.3 7.2 7.2
B. Potential movers by education, %
low 27.3 23.4 16.8 9.2 11.6 14.3 23.1 21.9 23 18.5
medium 59.5 64.2 73.8 56.8 69.7 60.7 58.1 58.7 59.8 63.2
high 13.2 12.4 9.4 34 18.7 25.1 18.8 19.4 17.2 18.3
100 100 100 100 100 100 100 100 100 100
C. Difference in composition between potential movers and stayers, % points
Low educ. 2 6.7** -1.2 -5.0* 2.8 4.0*** 3.9** 2.2 5.3** 2.5
High educ. -4.3* -7.1** -8.8** 4.3 -5.4* -3.8* -3.9*** -2.9 -3.5** -2.7*
Source: Calculations with survey data. Surveys conducted: I and II – by SaarPoll Ltd. and Faktum & Ariko
respectively, both for Estonian Ministry of Social Affairs; III – by SKDS for University of Latvia; IV – by
SKDS for Latvian Ministry of Foreign Affairs; V – by SIA „Data Serviss” for University of Latvia; VI – by GfK
Baltic for Latvian University of Agriculture (V and VI were commissioned by the Ministry of Welfare in the
framework of the National Program of Labor Market Studies).
Notes: a) Potential movers are defined as follows: Surveys I, II – answers ‘Yes, definitely’ or ‘Yes, with high probability’ to the
question about plans to work abroad (Col [3] – also a specific plan exists). Surveys III and VI – have ticked option ‘Move
abroad’ when answering the question ‘What are you ready to do in order to get a better /desirable job?’; Survey IV – ‘very
big’ or ‘big’ probability to go for a work a broad. Survey V – answers “Yes” or “Rather yes” to the question “Would you like
to go to work abroad?” (Col. [8] – also have made some preparations).
b) Excluding long-term (one year or more) unemployed.
The above results do not imply that the low skill level as such has been a factor
promoting migration, other things equal. The skill composition of migrants is cer-
tainly affected by age and geographical origin. When these and other relevant fac-
tors (including selection into not living in a one-person household) are controlled
for, medium-educated workers are most likely to move. During the post-accession
period, the effect of low (vs. medium) education on mobility shows positive trends
in Latvia and Estonia; Lithuania features a negative trend in the effect of higher
education on mobility. In summary, human capital in all three countries has be-
come less pro-migration after accession than it was before, and was becoming
even less pro-migration during the post-accession period.
5 Labor Market Outcomes
Labor market status: According to host country LFS results, migrants from the
EU10 to the EU15 ‘show higher labor market participation and employment rates
than the overall populations both in the sending and receiving countries; in fact,
78% of those migrants aged 15-64 are employed (European Commission, 2008:
Ch. 3, Chart 17; p. 129). In the UK, the major destination country for Baltic mi-
grants, the employment rate of migrants from EU8 countries (men aged 16-64 and
women aged 16-59) was above 80% in 2005/Q3 and above 84% in 2008/Q3
(ONS, 2008). Specifically for Lithuanian migrants in the UK, Khan (2008) reports
an employment rate of 84% for men and 80% for women, an unemployment rate
2% for men and 6% for women, and an inactivity rate of 14% for both genders in
2008/Q2. Employment rates of Lithuanian and Latvian nationals in Ireland were
between 82 and 83% according to the 2006 Population Census data. The unem-
ployment rates (9 to 10%) were higher than in the UK; and the inactivity rates at 8
to 9% were lower (see CSO Ireland, 2008). Estonian migrants in Finland have
seen much better labor market outcomes that immigrants from other countries be-
fore accession due to a good command of the Finnish language as well as the
similarity of the two cultures: Sutela (2005) finds that “the quality of their jobs as
well as the risk of unemployment reported are both almost similar to those experi-
enced by Finnish employees.” However, opening the Finnish labor market for
EU10 workers fostered further improvements: the unemployment rate among Es-
tonian immigrants fell from 16% in 2005 to 10% in 2007, and it was as high as
30% among immigrants from Russia (Employment and Economic Development
Office of Finland, 2008).
Occupations and industries: Figure 5 is based on the Baltic LFS 2002-2007

and compares occupational distribution of employed Baltic migrants across the
three countries and three bi-annual sub-periods. The share of manual workers
among migrants had been rising steadily. In 2006-2007 and compared to the pre-
accession period, it increased from two thirds to four fifths for Lithuania, from
two fifths to two thirds for Latvia, and from less than three fifths to four fifths for
Estonia. This is well above the average of about two thirds for all mobile EU10
nationals in 2007 (reported in European Commission, 2008: Ch. 3, Table 11).

Moreover, the share of unskilled manual workers has also increased among Lat-
vian and Estonian migrants. This has not been the case for Lithuanian migrants,
where this share was initially much higher. The number of migrants who were
employed as an unskilled manual worker in 2006-2007 was one in three Lithua-
nian migrants, one in four Latvian migrants, and one in seven Estonian migrants.
The proportion of low-skilled non-manual workers (mostly waiters, sales and re-
tail assistants, hotel housekeepers and personal care workers) has been declining
for all three countries, and it varied from 6% in Estonia to 17% in Latvia in 2006-
2007. Finally, the share of highly skilled non-manual workers (managers, profes-
sionals and technicians) has dropped sharply compared to the pre-accession pe-
riod: from 15 to 6% among Lithuanian migrants, 43 to 17% among Latvian mi-
grants and 27 to 13% among Estonian migrants. Results from the Irish Population
Census 2006 based on almost 16 thousand Lithuanian and over 9 thousand Lat-
vian migrants are presented in the same figure and provide a picture broadly simi-
lar to the LFS-based one. The Irish data feature fewer unskilled manual and more
low-skilled non-manual workers than the LFS data for Lithuania; and more skilled
manual and low skilled non-manual workers at the expense of highly skilled non-
manual workers for Latvia. These differences may reflect differences between
sectoral distribution of migrants in Ireland and the UK (discussed below), as well
as specifics of the sending countries LFS data (which do not cover permanent mi-
grants, as well as those with no family left behind) and sampling errors.
Distributions of Lithuanian and Latvian workers in Ireland by the four broad

groups of occupations are statistically indistinguishable. This similarity also holds
for economic activities: over half of the males worked in construction and manu-
facturing industries, while retail work, hotels and restaurants accounted for almost
half of the female employment (CSO Ireland, 2008: Table A5). Figure 6 (upper
panel) provides details on the distribution of both Baltic migrants and Polish for
comparison, according to economic activities in UK and Ireland. The high share
of agricultural employment is the most striking feature of Latvian and Lithuanian
migrants in the UK. After excluding workers employed by recruitment agencies
and others in unknown sectors, agriculture accounted for one third of Lithuanian
employment and two fifths of Latvian employment in the UK, without a signifi-
cant variation over the period from May 2004 to September 2008. Among Esto-
nian workers it decreased from one fifth in 2004-2005 to one tenth in 2006-2007.
By contrast, only 5% of Lithuanians and 10% of Latvians were employed in agri-
culture in Ireland. This difference might be partly due to the stock nature of the
Irish data which refer to April and, therefore, do not capture seasonal farm work-
ers. The share of Polish migrants employed in agriculture was just half of that ob-
served among their Lithuanian counterparts in both the UK and Ireland.
Figure 5
Employed migrants from the Baltic countries, stayers therein and return migrants
by occupation abroad (2002-2007)
Lithuania
100%
80%
60%
40%
20%
0%
2002-2003 2004-2005 2006 2006, Ireland 2006 2004-2006
temporary migrants all migrants stayers return

migrants
Unskilled manual Skilled manual Low skilled non-manual Highly skilled non-manual
Latvia
100%
80%
60%
40%
20%
0%
2002-2003 2004-2005 2006-2007 2006, Ireland 2006-2007 2004-2007
temporary migrants all migrants stayers return

migrants
Estonia
100%
80%
60%
40%
20%
0%
2002-2003 2004-2005 2006-2007 2006-2007 2005-2007
temporary migrants stayers return migrants
Source: Calculations with Lithuanian, Latvian and Estonian LFS data; for Ireland – census results in CSO
Ireland (2008), 36, 42.
Notes: Except for the Irish column, migrants are defined as household members employed abroad during the survey week;
hence, permanent migrants as well as those with no family left behind are not covered. Number of migrant observations for
the three sub-periods are 703, 788 and 350 for Lithuania; 170, 192 and 207 for Latvia; 109, 168 and 518 for Estonia. Data
on migrants in Ireland (N= 15715 for Lithuania, N = 9256 for Latvia) are based on the Irish Population Census and refer to
April 23, 2006. When calculating the Irish results, persons with unknown occupations (10% of Lithuanian sample, 8% of Lat-
vian sample) were excluded. Highly skilled non-manual occupations include ISCO groups 1-3; low skilled non-manual –
ISCO groups 4-5; skilled manual – ISCO groups 6-8; unskilled manual – ISCO group 9 (elementary occupations).
Figure 6
Baltic nationals working abroad by sector of employment before and after EU accession
Baltic and Polish migrants (excluding those working for recruitment agencies)
by sector of employment in UK (flow data, WRS), Ireland (stock data, Census)
100%
Other
80%
Health
Trade
60%
Hotels and
Restaurants
40%
Construction
Manufacturing
20%
Agriculture
0%
UK 2004- Ireland UK 2004- Ireland UK 2004- UK 2006- UK 2004- Ireland
2008 2006 2008 2006 2005 2008 2005 2006
Lithuania Latvia Estonia Poland
Baltic nationals working abroad by sector of employment

before and after EU accession (stock data, LFS)
100%
80%
Services
60%
Industry
40% Agriculture
20%
0%
2002-2003 2004-2006 2002-2003 2004-2007 2002-2003 2004-2007

Notes: The upper panel is based on the Worker Registration Scheme (WRS) data for the UK and the 2006 Population Cen-
sus data for Ireland. Workers whose actual sector of employment was unknown are excluded. In the UK, the excluded group
accounts for 35 to 40% of Baltic workers and 44% of Polish workers (mostly working for recruitment agencies), hence the re-
sults should be used with caution. By design, the WRS excludes also self-employed (a small group among Baltic migrants in
the UK, see Figure 8), as well as Baltic nationals who do not have citizenship of their home country (up to 15% of Latvian
workers abroad and up to 18% of Estonian workers abroad belong to this category, see Figure 10). In Ireland, the excluded
(due to non-response) group is rather small: 10 to 12% for Lithuania and Latvia, 7% for Poland. The lower panel is based on
sending countries’ LFS data; see Notes to Figure 5 for the number of migrant observations per country and period.
Figure 7
Over-qualification of employed migrants from the Baltic countries, stayers therein
and return migrants (2002-2007)
Lithuania
80
% of over-qualified among employed
70
60
50
40
30
20
10
0
2002-2003 2004-2005 2006 2006 2004-2006
migrants stayers return migrants
Share of over-qualified among medium-educated Share of over-qualified among high-educated

Share of over-qualified in total employment
Latvia
50
% of over-qualified among employed
45
40
35
30
25
20
15
10
5
0
2002-2003 2004-2005 2006-2007 2006-2007 2006-2007

Estonia
70
60
% of over-qualified among
50
40
30
employed
20
10
0
2002-2003 2004-2005 2006-2007 2006-2007 2006-2007


Notes: See Notes to Figure 5 for definition and other details. Over-qualified workers include: (i) Medium-educated workers
(ISCED 3-4) employed in low-skilled (unskilled manual) occupations (ISCO 9); (ii) High-educated workers (ISCED 5-6) em-
ployed in medium-skilled (ISCO 4-8) or low-skilled (ISCO 9) occupations.
Figure 8
Incidence of job-related characteristics among employed migrants
from the Baltic countries, stayers therein and return migrants (2002-2007)
Lithuania
80
70
60
50
40
%
30
20
10
0
self- temporary part-time supervisory tenure up to w eekly seeks other
employed 1 year hours ≥ 48 job
Migrants 2002-2003 Migrants 2004-2005

Migrants 2006 Stayers 2002-2003
Stayers 2006 Return migrants 2004-2006
Latvia
70
60
50
40
%
30
20
10
0
self- temporary part-time supervisory tenure up to w eekly seeks other
employed 1 year hours >=48 job

Migrants 2006-2007 Stayers 2002-2003
Stayers 2006-2007 Return migrants 2006-2007
Estonia
70
60
50
40
%
30
20
10
0
self- no contract temporary part-time supervisory tenure w eekly seeks other
employed ≤ 1 year hours >=48 job

Stayers 2006-2007 Return migrants 2005-2007

Notes: Migrants and stayers aged 18 to 64 years are included.
Agriculture, manufacturing and construction accounted for between one half to

two thirds of Lithuanian and Latvian workers in the UK and Ireland. This propor-
tion was just short of one half among Polish migrants and fell from one half in
2004-2005 to one third in 2006-2008 among Estonians in the UK. The total share
of the hospitality and retail industry varied from one quarter to one third of em-
ployment among Latvian and Lithuanian workers in both countries, as well as Es-
tonians in the UK (2004-2005) and Poles in Ireland. It was as high as two fifths
among Poles in the UK and reached even higher levels among Estonians in the
UK during 2006-2007. Baltic migrants in the UK were more often employed by
hotels than shops; the opposite was true for Ireland. One out of nine Estonian mi-
grants in the UK was employed in health and social work in 2006-2007. Individ-
ual shares of other sectors were small.
Despite the fact that the UK and Ireland host over three quarters of Lithuanian
and Latvian migrants, the sectoral distribution described above might be distorted
by a rather big share of 35 to 40% of Baltic workers in the UK in an unknown sec-
tor of employment because most of them work for recruitment agencies (see the
UK Home Office, 2008: Table 11). Moreover, the British Worker Registration
Scheme excludes the self-employed, which is a small group among Baltic mi-
grants (see Figure 8), as well as Baltic nationals who are not citizens of their home
country. Up to 15% of Latvian workers abroad and 18% of Estonian workers
abroad belong to the latter category (see Figure 11). Finally, only one third of Es-
tonian migrants went to the UK and Ireland during 2006-2007. Therefore, we pro-
vide alternative data based on Baltic nationals working abroad and observed in
sending countries’ LFS in Figure 6 (lower panel). The data do not cover perma-
nent migrants nor those with no family left behind.
As one would expect, the share of agricultural employment (17, 12 and 8%

among Latvian, Lithuanian and Estonian migrants, respectively) according to the
annual average stock data is much lower than the British flow data but higher than
in the Irish April stock data. However, the total share of agriculture and industry
among post-accession migrants (from one half to three fifths) is well in line with
the previous data sources and significantly higher than among all migrants from
the EU10 (about two fifths according to European Commission, 2008: Ch. 3, Ta-
ble 9). A comparison with the pre-accession period reveals that the share of indus-
trial employment among Baltic migrants has significantly increased for all three
countries. This has been at the expense of agriculture in the Lithuanian case and at
the expense of services for Latvia and Estonia.
Over-qualification and occupational mobility: Immigrant skills might be not

immediately utilized to their full extent in the host country labor market, either
because of the specifics of the latter or the lack of complementary skills, such as
fluency in the local language (Dustmann et al 2008). In such cases, as well as
when qualifications obtained abroad are not fully recognized or simply not ac-
cepted by employers, migrants work in occupations for which they are (formally)
overqualified. This phenomenon, also known as ‘brain waste’, is well documented
in the literature. In the EU, over 40% of recent high-educated immigrants from
other EU countries and almost 30% of their medium-qualified counterparts were
overqualified in 2007. Among the native-born, this rate was below 20% for high-
educated and 8% for medium-qualified (European Commission, 2008: Ch. 2, sec-
tion 8.3; these data, however, exclude the UK, Ireland and Germany – countries
which host most of the recent migrants from the EU10).
Figure 7 sheds light on the extent of over-qualification of Baltic migrants be-

fore and after EU accession. The proportion of overqualified among high-
educated movers reached extremely high levels in 2006-2007: 47% for Latvia,
60% for Estonia and 70% for Lithuania, while it was around 20% among high-
educated stayers in all three countries. A sharp increase in the over-qualification
rate of the high-educated migrants after EU accession is evident for Latvia and Es-
tonia, but Lithuanian migrants with tertiary education were overqualified almost
to the same extent before accession. The story is different for medium-educated
movers: among those of Lithuanian or Latvian origin, the over-qualification rate
was about 30% in 2006-2007- three times higher than among stayers. This rate
was just above 10% among both stayers and movers for Estonia. Compared to the
pre-accession period, the over-qualification rate among medium-educated movers
decreased for Lithuania but increased somewhat for Latvia and Estonia. During
2006-2007, the proportion of overqualified people reached almost two fifths of all
Lithuanian migrants and almost one third of Latvian migrants, while it was just
one fifth among Estonian migrants.
In order to check whether over-qualification of migrants is caused by moving,

we look at the migrant’s last occupation at home using the 'Relatives Abroad'
module of a large survey of 'Geographical Mobility of Population' conducted in
Latvia between September 2005 and March 2006 (see Notes to Figure 4 for de-
tails). Three out of five post-accession migrants who worked in high-skilled occu-
pations at home switched to low or medium-skilled occupations abroad; while
among pre-accession migrants such downward mobility rate was less than two
fifths (Figure 4, middle and lower panels). This is very much in line with the over-
qualification rates reported above. As far as movers who held medium-skilled jobs
at home are concerned, their downward occupational mobility rate is similar to the
over-qualification rate among those who moved before accession. It exceeds 50%
for those who moved in 2004-2005, which is well above the over-qualification
rate documented in Figure 7. This could be because the latter group was to some
extent under-qualified at home or because of custom occupation classification
used in the survey. In any case, there is no evidence that overqualified movers
were already overqualified at home.
Type of work: Work-related characteristics of movers and stayers by country

and period are summarized in Figure 8.
The share of wage earners among employed Baltic migrants has been very high
both after and (except for Latvian migrants) before accession: 94 to 96% for
Lithuanians, 95 to 98% for Estonians, and above 98% for Latvians: The average
for mobile EU10 nationals was 91% in the EU at large in 2006 and 92% in the UK
in 2007 (European Commission, 2008: Ch. 3, Charts 18-19). The low share of
self-employment is also characteristic of the domestic labor market in Latvia and
Estonia. It was 9% in 2006-2007, compared to 16% in Lithuania and in the EU10
in 2006. Comparisons across host countries suggest a somewhat higher share of
those self-employed (between 10 and 20%) among Lithuanian migrants in the US
and Canada, as well as in Russia, Ukraine and Belarus; whereas it is just 9% in
Germany – a striking difference with the EU10 average of 38% (European Com-
mission, 2008: Ch. 3, Chart 19).
Such outcomes as fixed-term contracts, work without contract and long hours
are found among movers much more often than among stayers. A high incidence
of temporary contracts (60% among Lithuanian movers in 2004-2006 and Latvian
movers in 2004-2005) and job tenure not exceeding one year (50 to 70% in all
cases, except Estonian movers in 2006-2007) also highlights the temporary nature
of the recent migration. During 2006-2007, however, the incidence of fixed-term
contracts and short tenure among Latvian movers dropped by 14 points for each.
This indicates indirectly that many Latvians have earned a good reputation with
employers within a year. Why there is no similar evidence for Lithuanians is an
open question. One explanation might be related to the fact that Latvian workers
were more willing to accept long hours: more than two fifths of Latvian post-
accession movers worked 48 hours a week or more. This was the case for only one
in five Lithuanians in 2004-2005 and one in nine in 2006.
Estonian movers saw even stronger progress in the quality of jobs in 2006-
2007: the incidence of fixed-term contracts dropped from 30 to 10%, while the
proportion of workers without a contract fell from 22 to 9%. Lifting restrictions
for EU10 workers by Finland in May 2006 may well have been an important fac-
tor in these developments.
The incidence of long hours among Estonian migrants fell steadily: from two
fifths in 2002-3003 to one third in 2004-2005 to one quarter in 2006-2007, re-
maining higher than Lithuanians but lower than Latvians. This is an interesting
fact in the context of the transfer of norms: just 3% of Lithuanian stayers work
48+ hours a week. This proportion is 10% in Estonia and 16% in Latvia.
A small, but not negligible, proportion of movers has been assigned supervi-
sory responsibilities: 3% among Lithuanians, up to 5% among Latvians and up to
10% among Estonians. Part-time work and on-the-job search have been virtually
absent among the migrant workers.
6 Length of Stay Abroad, Return Migration

and Labor Market Transitions
It is well documented that the post-accession migration from EU10 countries has
been (so far!) predominantly temporary or circulatory. About two thirds of work-
ers registered in the UK WRS intended to stay for no more than one year (Pollard
et al., 2008: Figure 20). Barrell et al. (2007: p. 5) compare stock and flow dynam-
ics of EU10 nationals in the UK and Ireland and conclude that
‘N[ew]M[ember]S[tates] migration to Ireland and the UK has been very much of a
temporary nature, with relatively short stays before return’ (see European Com-
mission, 2008: pp. 121-122; Pollard et al., 2008: pp. 39-40 for further evidence).
Blanchflower and Shadforth (2009) find that workers from the EU10 return more
rapidly than those who have arrived in the UK from other countries and suggest
treating them as temporary workers rather than migrants. Kaczmarczyk and Okól-
sky (2008: Figure 2.7) provide evidence that after EU accession, the proportion of
Polish migrants staying abroad for less than a year increased sharply and stayed
above three fifths for two years, then fell somewhat but remained above one half.
Hazans (2008) finds that by the beginning of 2007 almost 5% of the economically
active Latvian population had worked abroad over the last three years, and 12%
had family members with such experience.
We refer to Dustmann (1997, 2003), Dustmann and Weiss (2007), World Bank
(2007: pp.179-180), Hazans (2008), van Baalen and Müller (2009) and the litera-
ture therein for theoretical models, discussions and empirical results on return mi-
gration. Returning might be driven by a higher marginal utility of consumption in
the home country than in the host country (put more simply, everything tastes bet-
ter at home…); by changing prices and exchange rates; by improving employment
possibilities, working, living and social conditions in the home country and/or ris-
ing expected earnings there; and by expected gains at home from human capital
accumulated abroad. On the other hand, migrants might have pre-defined (or
endogenously determined) maximal duration or a target savings level frequently
related to a planned purchase of a car or an apartment, or repaying a loan. Once
this level is reached, they return. More generally, the achievement of any pre-
defined objective (e.g. gaining new experience, increased self-confidence and self-
respect, working out a business plan) might be a sufficient reason to return. Mi-
grants facing bad labor market outcomes in the first few months abroad tend to re-
turn quickly. By contrast, the occurrence of a new family or partnership while
abroad leads to permanent emigration in most cases. There is survey and interview
based evidence in Latvia which confirms the existence of all these motives (see
University of Latvia, 2007: pp. 90-112). It should be noted that non-economic mo-
tives play an important role. Return migration is closely related to migrant transi-
tions between employment and other labor market states: the latter often is either a
reason for or a result of the former. In this section, we provide survey-based evi-
dence on both issues with regards to recent mobile Baltic nationals.
Figure 9 is based on a survey of relatives abroad which was conducted in Lat-

via between September 2005 and March 2006 (see Notes to Figure 4 for details).
It shows how the intended duration of further stay in the host country increases
with the time already spent there. This is in line with van Baalen and Müller
(2009), who find that 70% of temporary migrants in Germany tend to increase
their intended duration of stay over time. Almost half of the post-accession mi-
grants, of whom over 90% are in EEA countries, did not plan to stay for more than
a year, starting from the survey date. Three quarters of those who arrived before
May 2004 planned to stay for more than two years, with almost no difference be-
tween EEA and other host countries: pre-accession migrants were almost evenly
split between these destinations. According to a population survey conducted
within the same study, among those with some, but not necessarily post-accession,
foreign work experience, about three fifths had spent no more than a year abroad,
and one fifth one to two years (University of Latvia, 2007: Figure 4.3). By con-
trast, among return migrants found in an Estonian survey carried out at the end of
2006 (see Table 2 for details) only 35% stayed abroad for up to one year and 28%
for one to two years.
Figure 9
Latvian migrants’ expected duration of further stay in the host country (2005/Q4-2006/Q1)
100%
80%
60%
40%
20%
0%
less than 3 3 to 11 months 1 to 2 y ears total post- more than 2 more than 2
months accession y ears (in EEA y ears (in non-
migrants countries) EEA countries)
elapsed duration of stay
Less than 3 months 3 to 11 months 1 to 2 years More than 2 years
Source: Calculations with (un-weighted) data of 'Relatives abroad' module of the survey on geographical
mobility of population conducted by Data Serviss Ltd for the University of Latvia in 2005/Q1-2006/Q4 in the
framework of the National Program of Labor Market Studies undertaken by the Latvian Ministry of Welfare.
LFS-based data on return migration and labor market transitions of employed

Baltic movers are summarized in Figure 10. While more than half of Latvian
movers return to home country within a year, this is the case for less than two
fifths of their Lithuanian counterparts and just over one quarter of Estonian mi-
grant workers. The relatively low propensity to return among Estonians is consis-
tent with longer stays reported above, as well as with the findings on the incidence
of short tenure in Figure 8.
Figure 10
Return migration and transition out of employment among employed migrants
from the Baltic countries and stayers therein (2002-2007)
% of employed movers or stayers
60
50
40
30
20
10
0
2002-2003 2004-2006 2002-2005 2006-2007 2002-2003 2004-2007
Movers return home w ithin a year Movers unemployed or inactive after 1 year
Stayers unemployed or inactive after 1 year

Notes: Migrants and stayers aged 18 to 64 years are included.
One out of five Lithuanian and Latvian workers abroad and one out of ten Es-
tonian employed movers become unemployed or inactive a year later, most of
them in the home country. Among employed stayers, these proportions are much
lower: 5 to 7% in Lithuania and Latvia, 4% in Estonia (Figure 10). Among return
migrants with some foreign work experience during the last two years, inactivity
rates are even higher: almost 40% in Lithuania, more than 20% in Latvia and al-
most 20% in Estonia (Figure 11). This should not be interpreted as evidence for a
weak labor market position of returnees. One reason for high inactivity and unem-
ployment among return migrants is the fact that due to accumulated savings, they
can afford extended periods of inactivity and a longer job search (according to a
recent study, 70% of return migrants in Latvia spend part of the earnings from
abroad for everyday needs (University of Latvia, 2007, Table 4.15). Another rea-
son is that many return migrants plan to work abroad again and take no job (or
only temporary job) while at home. Indeed, according to a survey conducted in
Latvia on the eve of 2007 (see Notes to Table 2 for details), 27% of return mi-
grants and just 7% of stayers are ready to go to work abroad.
Figure 11
Demographic characteristics of economically active migrants
from the Baltic countries, stayers therein and return migrants (2002-2007)
Lithuania
80
70
60
50
40
%
30
20
10
0
female has a has kids age 35+ big city non- born student currently
partner under 15 citizen abroad inactive

Migrants 2006 Stayers 2006
Return migrants 2002-2003 Return migrants 2004-2006
Latvia
70
60
50
40
%
30
20
10
0
female has a has kids age 35+ rural minority noncitizen born student currently
partner under 15 abroad inactive

Return migrants 2002-2003 Return migrants 2004-2007
Return migrants 2006-2007
Estonia
80
70
60
50
40
%
30
20
10
0
female has a has kids age 35+ rural minority noncitizen born student currently
partner under 15 abroad inactive
Return migrants 2006-2007
Source: Calculations with LFS data and one ad hoc Latvian survey (survey VI in Sources of Table 2).
Notes: Persons aged 18 to 64 years who have been economically active for some time during the last year are included. Re-
turn migrants have worked abroad during the last two years (three years for Latvian column ‘2006-2007’).
Characteristics of return migrants differ from those of all migrant workers in

many respects. As documented in Figure 3, return migrants are more educated. In
terms of occupations, they are clearly more skilled than all movers on average but
somewhat less skilled than stayers (Figure 5). Returnees are less often than mov-
ers ‘too educated’ for their jobs (Figure 7), although in Lithuania their over-
qualification rate is higher than that of stayers (in Latvia and Estonia the latter ef-
fect is within the margin of error). Hazans (2008) shows that after controlling for
workers’ demographic characteristics and education, as well as for foreign and
unemployment experience of their family members, returnees command a sub-
stantial earnings’ premium: 15% on average, with more than 20% among men and
6% among women. Accounting for a variety of job characteristics leaves the gap
almost unchanged. The earnings gain of the average return migrant is shown to be
entirely caused by foreign experience. There is, however, a great deal of individ-
ual heterogeneity: in the upper quartile of the earnings distribution, the unex-
plained gap in favor of return migrants exceeds 40%, while in the bottom quartile
it is negative. This is consistent with the coexistence of success, failure and oppor-
tunity stories behind return decisions. Preliminary results suggest that Estonian re-
turnees earn about 30% more than otherwise similar stayers despite concentrating
in the bottom part of the earnings distribution.
Job-related characteristics of employed return migrants are compared with that

of movers and stayers in Figure 8. One in five employed return migrants in
Lithuania and one in eight in Latvia are self-employed. These rates are in line with
those found among stayers and are much higher than among migrant workers
abroad. By contrast, as LFS data suggest, Estonian returnees are not likely to be-
come self-employed. Lithuanian and Estonian returnees are much less likely than
stayers to hold supervisory jobs, while it is the other way around in Latvia. The
incidence of temporary contracts and short tenure among returnees is much higher
than among stayers but not as high as among movers, except for short tenure in
Estonia. Returnees have similar working hours to stayers. One in ten employed
Lithuanian and Estonian returnees look for another main job, which is a much
higher rate than among stayers.
7 Migration Impact on the Baltic Labor Markets

and Economies
Economic theory suggests several channels through which migration possibilities

and actual migration affect the sending country’s labor market. New emigration
options reduce labor supply and make it more elastic, thus increasing wages and
narrowing the gap between the marginal productivity of labor and pay (see Fig-
ures 10.1, 10.5 and the discussion in Ehrenberg and Smith, 2006). Unemployment
goes down because those unemployed or inactive move abroad (up to 20% of
post-accession movers were not employed a year ago according to Figure 2) or fill
the vacancies left behind by previously employed migrants.9 The latter process, of
course, results in a rising vacancy rate because it takes time for vacancies to be
filled. Labor shortages may develop in some segments of the labor market when
the unemployed lack the skills required by employers and cannot replace the mov-
ers (see Rutkowski, 2007 for discussion and evidence). Note, however, that such
cases were clearly not the dominant ones across the whole of the economy, other-
wise unemployment rates would have increased, while employment rates would
have fallen – a scenario opposite to the one observed in reality.
Figure 12 supports the above story for each of the Baltic countries. During
2004-2007, the unemployment rate declined by eight points in Lithuania, by four
and a half points in Latvia and by more than five points in Estonia, and the dy-
namics of this decline was consistent with the one of migration flows. It is impor-
tant to note, however, that migration was not the only cause for falling unem-
ployment. Total employment, driven by growing domestic demand, intensifying
trade, injections from EU structural funds and predatory credit expansion (mainly
by Scandinavian-owned banks10), increased in absolute terms by much more than
unemployment decreased. While LFS total employment figures are not reliable
due to population estimates which fail to account for the largely unregistered emi-
gration, increase in domestic employment is confirmed by data from various
sources independent from the LFS. For instance, the increase in the number of so-
cial security contributors in Latvia between 2003 and 2007 amounts to 13.5% of
the 2004 labor force; similar estimates for Lithuania and Estonia based on enter-
prise statistics are 7.5% and 11.1%, respectively.
Figure 12 also documents that all three countries, especially Lithuania, experi-
enced a substantial increase in overall vacancy rates during the period observed.
More specifically (not shown in the figure), the vacancy rate in the construction
sector tripled in Latvia between 2005 and 2007; while in Lithuania it increased by
a factor of more than ten between 2004 and 2007.
9 Ignoring the small proportion of migrants who move as ‘immediately permanent’, the mov-
ers in the short run should be still counted in the sending country’s population, while those
who stay for more than a year should not be counted. Hence in some cases, immediate
and medium-term effects of a person moving abroad on the home-country’s unemploy-
ment rate might differ but they never have opposite signs. Some ‘good’ flows might be-
come unemployment-neutral but still foster employment and participation rates.
10 See Andrusz (2009), Vistesen (2008).
Figure 12
Migration and labor market developments in the Baltic countries (2004-2007)
Migrants/Labor force, %
Lithuania
(LHS)
4 20
18 Decline in unemployment
rate, % points (LHS)
16
3 Change in
14 employment/Labor force,
12 % (LHS)
Vacancies/Employment, %
2 10 (LHS)
8 Inflation (RHS)
6
1
4 Real w age grow th (RHS)
2
Cumulated migration
0 0
flow s/Labor force, %
2004 2005 2006 2007 (RHS)
Latvia
4 20 (LHS)
16
3
14 Change in
employment/Labor force,
12 % (LHS)
Vacancies/Employment,
2 10 % (LHS)
8
Inflation (RHS)
6
1
2
0 0 Cumulated migration
2004 2005 2006 2007 flow s/Labor force, %
(RHS)
Estonia
(LHS)
20
4
16
3 Change in
14
employment/Labor force,
12 % (LHS)
Vacancies/Employment,
10 % (LHS)
2
8 Inflation (RHS)
6
1
2
Cumulated migration
0 0
flow s/Labor force, %
2004 2005 2006 2007
(RHS)
Source: Migration – see Figure 1. Other data – Eurostat and National Statistical Offices.
Notes: Labor force refers to 2004. Employment – administrative data independent of LFS and population estimates.
The vacancy rate for manual jobs more than doubled in Latvia and more than
quadrupled in Lithuania between 2005 and 2007. This led to growth in unit labor
costs caused by rising wages and hiring expenses. See Karu and Nurmela (2006a-
d), Ministry of Social Security and Labour of Lithuania (2007: p. 25; 2008: pp. 26,
218-219), Ministry of Welfare of Latvia (2008: Table 2.3.4) for evidence on labor
shortages in construction, manufacturing, land transport, health and police, the re-
sulting wage pressure and increasing numbers of workers from non-EEA coun-
tries.11 Based on official estimates presented in Figure 12, real wage growth dur-
ing the four years (2004-2007) was 56% in Latvia, 47% in Lithuania and 41% in
Estonia.12 Real wage growth was largely in line with accumulated migration flows
in Lithuania, and it was faster in Estonia and especially Latvia. One likely reason
for this difference is an initially higher unemployment rate in Lithuania.13 Wage
growth combined with remittances and credit expansion contributed to inflation.
Consumer prices grew in line with migration outflow in Estonia but more slowly
in Lithuania. In Latvia, they started to outpace emigration in 2007. Overall infla-
tion in 2004-2007 was 14.1% in Lithuania, 19.6% in Estonia and 33.2% in Latvia.
Tight labor markets forced employers to lower hiring standards as an alterna-

tive or complement to further wage increase: media frequently presented anecdo-
tal evidence of falling quality in services and construction. Higher wages and
lower standards led to higher participation14 and employment rates, both in general
and disadvantaged groups, when formerly inactive and unemployed filled vacan-
cies, including those not caused by migration. Hazans (2007b-d; 2009) documents
improvements in the labor market position of ethnic minorities, the elderly and the
low-skilled in Latvia; similar results for Estonia are found in Ministry of Social
Affairs of Estonia (2008: pp. 31, 33).
We refer to Barrel et al. (2007) and Baas et al. (chapter 2 in this volume) for
specific estimates of the migration effects on wages, unemployment, inflation,
productivity, GDP and working-age population in sending countries. In particular,
11 Compared to the previous year, the number of work permits issued to such workers in
Latvia doubled in 2006 and tripled in 2007, whereas in Lithuania it nearly doubled in 2007.
Most of these workers come from Ukraine, Russia, Belarus, Moldova and Uzbekistan.
12 Official data are distorted by ignoring envelope wages. LFS-based estimate of real wage
growth for the same period is just 39% for Latvia, whereas it is 57% for Estonia. Williams
and Renooy (2009) indicate that in Latvia and Lithuania envelope wages are more com-
mon and paid more for regular hours, while in Estonia they are less common and paid
more for overtime or extra work.
13 Other reasons might be related to country-specific features such as the real estate bubble
being less pronounced in Lithuania.
14 This effect is reduced by the negative impact of remittances and savings from earnings
abroad on the labor supply of migrants' household members left in the home country, as
well as return migrants. Between 2003 and 2007, participation rates have increased by
about four percentage points in Latvia and by about three percentage points in Estonia,
but felt by almost two percentage points in Lithuania.
Barrel et al. (2007: Tables 3, 4) estimate that migration contributed to a fall in the
unemployment rate over the four year period of 2005-2008 of 3.1 points for
Lithuania, 2.4 points for Latvia and 1.0 point for Estonia; whereas the estimated
impact on inflation is 0.8, 0.8 and 0.2 points, respectively. In our opinion, the true
effects on unemployment and wages might be higher than the ones based on
macro-models. For instance, evidence from Figure 2 suggests that the direct ef-
fects of flows from unemployment and inactivity to foreign countries might alone
explain the declines in the unemployment rates of the sizes quoted above, without
accounting for replacement flows. Macro-models do not account for the mo-
nopsonistic structure of the labor market, in particular the threat of a substantial
fall in labor productivity when a firm loses not just a marginal worker but, say,
half its workforce. Scale effect, work organization problems, and the inability to
compete for publicly financed projects can all be underlying factors. Underesti-
mated wage effects and ignoring the deterioration of hiring standards lead to an
underestimate of the resulting return migration and an increase in labor force par-
ticipation, especially if assumed labor supply elasticities are derived from data
with modest wage changes.
Assuming that emigration from high-unemployment regions (which, according

to the wage curve are also low wage regions, see Blanchflower and Oswald, 1994;
Brücker and Jahn, 2008) is more intensive, regional disparities in unemployment
and wages should be reduced by external mobility. This is found for Latvia in
Hazans (2007b) and for Estonia in Republic of Estonia (2007: Table 1), Ministry
of Social Affairs of Estonia (2008: p. 34). Similarly, given that emigration was
predominantly medium and low-skilled, one should expect a reduction of the
wage differentials between high-educated and other workers. In fact, according to
LFS-based Mincerian earnings function augmented with gender, ethnic and re-
gional controls, between 2003 and 2007 net wage differential between full-time
employees with tertiary and secondary general education decreased by 15 points
in Latvia and by 11 points in Estonia.
8 Demographics of the Baltic Migration
Composition of the flows and stocks in the host countries: Table 3 is based on
administrative data and presents demographic profiles of post-enlargement Baltic
and Polish migration flows to the UK, Estonian migration flow to Finland, as well
as Baltic (excluding Estonia) and Polish nationals living in Ireland in April 2006
(Population Census data).
Table 3
Demographics of Baltic and Polish migrants in UK, Ireland and Finland (%)
Poland Lithuania Latvia Estonia
UK (flow data)
2004-2007
Age 15-24 42.1 43.8 44.2 48.5
Age 25-34 41.0 32.0 29.4 29.9
Age 35-54 16.1 22.9 24.6 19.4
Age 55-64 0.9 1.4 1.7 2.3
Total 100.0 100.0 100.0 100.0
Female 42.6 50.1 53.4 57.6
Ireland (stock data) Finland

2006 2004-2007
Age 15-24 30.3 30.9 30.6 22.3
Age 25-34 50.8 42.5 40.8 30.4
Age 35-54 18.0 25.3 27.1 42.2
Age 55-64 1.0 1.3 1.5 5.1
Total 100.0 100.0 100.0 100.0
Female 35.4 43.5 45.7 48.3
(Age 0-14)/Age 15-64 8.2 12.8 10.3 19.2
Age 65+/Age 15-64 0.2 0.3 0.3 1.7
Household type
One person 3.1 2.8 3.8
Couple without children (with or 28.7 24.9 23.9
without other members)
Couple with children (with or 21.1 32.4 25.1
without other members)
Single parent with children 3.8 7.2 9.1
Two or more family units 4.8 6.8 5.1
Non-family household 38.5 25.9 32.9
Total 100 100 100
Mixed with Irish 9.1 11.5 11.9
Mixed with non-Irish 8.1 15.2 27.0
Source: UK – Calculation with data on allocation of National insurance numbers
(www.dwp.gov.uk/asd/tabtool.asp). Ireland – elaboration on Population Census data published in CSO Ire-
land (2008), 29, 33, 41. Table A1. Finland – elaboration on immigration data of Statistics Finland.
One can observe significant differences between the Baltic flows on one side
and the Polish ones on the other; between Estonian flows to the UK and Finland;
and between the UK and Irish profiles. The latter differences might be, at least in
part, caused by flow-stock distinction.
Baltic migrants are in general not as young as Polish ones: 22 to 26% of the
Baltic movers to the UK and 27 to 29% of their counterparts in Ireland are at least
35 years old, whereas these proportions are 17 and 19% for the Polish migrants.
Furthermore, Estonian migrants to Finland are even older: almost half of them are
aged 35 and above. The proportion of females among the Baltic migrants is 50 to
58% in the UK and 44 to 46% in Ireland and is about on average 10 points higher
than for Poles in the same country. The highest youth dependency ratio, which is
migrants under 15 years of age as a proportion of 15-64 year old migrants, is
found among Estonian migrants in Finland (likely due to the geographical prox-
imity and the low language barrier), and reaches 19%, not much lower than
among stayers in Estonia (23%). For Lithuanian and Latvian migrants in Ireland,
this ratio is 13 and 10%, respectively. It is even lower (8%) among their Polish
counterparts. The old age dependency ratio is very low in all cases: 0.2 to 0.3% in
Ireland and 1.7% in Finland.
The Irish data also provide a description of migrant household types for 2006.
One quarter of Lithuanian and one third of Latvian migrants lived in non-family
households. This proportion was even higher for Polish migrants at almost two
fifths. A further 5 to 7% lived in households with two or more family units. About
half of Latvians and Poles and 57% of Lithuanians lived in households based on a
couple with or without other members. This couple had children under 15 in about
three out of five such cases among Lithuanians, one out of two among Latvians
and two out of five among Poles. Single parents with children accounted for 7 to
9% among Latvians and Lithuanians but just 4% among Poles. Only 3 to 4% lived
alone. One in eight Baltic migrants lived in households with some Irish members.
Furthermore, one in seven Lithuanians and one in four Latvians lived together
with people of non-Irish nationalities.
Demographic profiles and determinants of temporary and return migra-

tion: This sub-section is based on Figure 11, which compares demographic pro-
files of Baltic temporary migrants found in the LFS data with that of stayers and
return migrants. In addition, we discuss results from Hazans and Philips (2009) on
the determinants of mobility15 and data on migration intensions from Table 2.
Females account for one fifth to one third of post-enlargement temporary

workers – much less than according to host country statistics. The following offer
some explanations: (i) men are more likely to stay abroad permanently or for pro-
longed periods; (ii) male movers are more likely to be considered as gone from
their households of origin; (iii) female movers are less likely to leave family be-
hind (i.e. more likely to be tied movers or to live alone before moving). The pro-
portion of females among return migrants in Lithuania is similar to that among
migrants, while in Latvia, and especially Estonia, it is higher. This suggests that
men are less likely to return on average. After controlling for other factors, a posi-
tive female effect on returning is indeed found in Estonia; whereas in Lithuania
and Latvia, the effect is negative for childless women. Latvian women with chil-
dren under 15 are more likely to return than men, other things equal. In line with
data on migration intensions reported in Table 2 and results of other studies (see
Blanchflower and Shadforth, 2009: Table 10; Fouarge and Ester, 2007: Table 7),
15 To correct for the fact that movers with no family left behind are not covered by LFS, we
estimate probit models with sample selection into not living in a one-person household us-
ing age-squared as an instrument.
we find that Baltic females are significantly less likely to work abroad than other-
wise similar males; this effect is less pronounced in Lithuania.
In each of the Baltic countries, about three out of five stayers have a spouse or
partner. The same proportion is found among Estonian movers, which confirms
once again that they are more likely to stay abroad permanently). However, it is
just two fifths among Lithuanian and Latvian movers. One quarter to one third of
movers have children under 15, who may, or may not, stay at home. The effect of
the migration experience on cohabiting is ambiguous: some partnerships might not
survive the time apart, but new ones might be created abroad. In the latter case,
however, return is less likely. Hazans (2008: Appendix 3 and Table 6) finds a
slightly negative (significant for women) causal effect of foreign experience on
having a partner among return migrants in Latvia. As shown in Figure 11, how-
ever, they are on average somewhat more likely to have a partner than migrants in
general but less likely than stayers. By contrast, post-accession Lithuanian and es-
pecially Estonian returnees are less likely to have a partner compared to migrants
as a whole. This raises the question of the causal effect of marital (cohabiting)
status of migrants on the likelihood of returning. This question awaits a more
thorough investigation, but simple models (not accounting for endogeneity of
partnership) find a negative effect among all Lithuanians, as well as among child-
less Latvian men.
An important question from the social perspective concerns negative impacts

(including school drop-out) on children left behind by parents working abroad
(Soros Foundation Romania, 2008; World Bank, 2007: pp. 177-178). Such chil-
dren, who are called ‘the Irish kids’ in Latvia, are raised by ‘quasi-single’ parents
or even by grandparents if both parents work abroad. The National Strategy Re-
port of Lithuania on social protection and social inclusion acknowledges that ‘in a
few recent years, some children, though formally cared by parents, in reality have
been left for some time without parental care while they are in emigration abroad.
Such children are left alone or placed under guardianship of relatives (sometimes
of non-relatives). This creates unfavorable conditions for children’s development
and social integration.’ (Republic of Lithuania, 2008: p.23). The Lithuanian gov-
ernment plans to reform the rules on guardianship as part of its response to the
situation.
Other things equal, the propensity to move declines with age in all three Baltic
countries. However, about one third of Lithuanian and Latvian temporary mi-
grants, as well as return migrants, are older than 35. The Estonian picture is dif-
ferent: about half of the migrants but just a quarter of the returnees are older than
35, suggesting that younger migrants are more likely to return. Indeed, after con-
trolling for other factors, the likelihood of return reaches its minimum at 50 years
of age (note that nine out of ten migrants are younger than 50). The age effect on
Latvian migrant propensity to return is also negative, while it has an inverse U

shape with maximum at 40 for Lithuanian migrants.
One of the most intriguing questions about the post-enlargement migration

wave from the Baltic countries is the behavior of Russian-speaking minorities,
many of whom in Estonia and Latvia do not have citizenship; and hence, they are
not covered by the free mobility provisions (see Kahanec and Zaiceva 2009 for a
study on the citizenship effects on labor market outcomes in Europe). Regarding
migration intentions ‘in the near future’ or supported by specific steps, the propor-
tion of potential movers among minority populations in general is somewhat
lower than among the titular population in Latvia but higher in Estonia (Table 2,
col. [3] and [8-10]). In both countries minority citizens were, however, more in-
clined to move than the titular population, but it was the other way around for
non-citizens. This citizenship effect is much more pronounced in Estonia.
The proportion of minorities among Latvian migrant workers dropped from

56% before accession to 45% after (still being slightly over-represented). The
proportion of non-citizens dropped more steeply from 33% before accession to
15% in 2004-2005. In other words, non-citizens were over-represented among
pre-accession migrants but under-represented among post-accession ones. The
profile of Estonian migrants underwent similar changes after the lifting of restric-
tions by Finland in 2006. The proportions of minorities, non-citizens and foreign-
born among return migrants in both Latvia and Estonia were in line with those
among migrants. Lithuanian non-citizens were over-represented among pre-
accession migrant workers but became under-represented by 2006; however, their
proportion among return migrants is much higher than among stayers.16
Given that ethnicity, citizenship and being foreign-born are all inter-related
with each other (although not multi-collinear), as well as with regions and degree
of urbanization, obtaining a clear picture of the effects of these factors on mobility
is only possible in a multivariate framework. The ethnic minority effect on the
propensity to work abroad before accession was strongly positive in Latvia and
Estonia. Following accession, it has become much smaller in size in both cases
but not significant for Estonia (also for Latvia when minorities are not sub-divided
by citizenship). Moreover, during the 2006-2007 (i.e. when Finland has opened its
labor market) minorities were significantly less likely then otherwise similar eth-
nic Estonians to become economic migrants. The effect of being a non-citizen on
the likelihood of working abroad is strongly negative for Latvia and Estonia in the
post-accession period, while the effect of being foreign-born has the opposite
sign.17 When all effects are combined, foreign-born minority citizens appear the
16 We are not able to track changes with regard to ethnicity of Lithuanian migrants, due to
data limitations.
17 Local-born and foreign-born minorities in the Baltic countries differ also in terms of educa-
tional attainment, see Hazans et al. (2008: Figure 4).
most mobile; local born minority non-citizens appear the least mobile. In Estonia,
non-citizens are also more likely to return if having moved, other things equal. In
Latvia, minority migrants are less likely to return, other things equal.
Residents of big cities are over-represented among both movers and returnees
in Lithuania; in Estonia, the same is true for rural residents.
9 Remittances and Use of Foreign Earnings
Figure 13 illustrates the evolution of officially recorded remittance flows gener-

ated by Baltic nationals working abroad (including posted workers). A sharp in-
crease in these flows after the EU accession in 2004 is evident for all three coun-
tries, but especially for Estonia. Moreover, starting from 2004, remittances to
Estonia exceed those to Latvia and Lithuania in per capita terms; the difference
was as large as 40 to 70% in 2006-2007. This might seem surprising given that the
largest labor outflow was registered from Lithuania and the smallest from Estonia
(Figure 1). One explanation has to do with posted workers: they are mostly not
covered by migration statistics, but their earnings are included in the remittances
statistics (World Bank, 2008). When posted workers are included, number of Es-
tonians working abroad as a share of country’s working-age population is some-
what higher than the Latvian figure but smaller than the Lithuanian one (Table 1c,
last column; we assume similar employment rates among adult migrants from all
Baltic countries).
The question remains, however, why Estonian and Latvian movers remit more
than they Lithuanian counterparts. One can offer at least three (not competing)
explanations. First, such an outcome is possible if Lithuanians rely on informal
channels of sending money home (not captured by the World Bank statistics) to a
much larger extent than Estonians and Latvians. Second, as shown in Figure 10
and confirmed by tenure data in Figure 8, the intensity of return migration among
Estonian movers is much lower than among Latvians and Lithuanians. In other
words, Estonian migrants send money home, whereas other Baltic migrants tend
to bring money in person. Finally, it is possible that Estonian migrants in Finland
enjoy higher earnings on average than other Baltic migrants located mainly in Ire-
land and the UK. The reasons might include longer tenures and relatively good
Finnish language skills.
Figure 13
Remittances to the Baltic countries (USD per capita), 2000-2007
350.0
remittances, USD per capita
300.0
250.0
200.0
150.0
100.0
50.0
0.0
2000 2001 2002 2003 2004 2005 2006 2007
Estonia Latvia Lithuania
Source: World Bank (2008) and own calculations.

Notes: Remittances include workers’ remittances (current transfers by migrants who are considered residents in the host
country) and earnings of posted, seasonal and other short-term workers who have centers of economic interest in their own
economies.
In 2007, Estonian migrant workers sent back $329 per every Estonian resident,
while the corresponding figures for Latvia and Lithuania were $219 and $192
(World Bank, 2008 warns that ‘the true size of remittances... is believed to be lar-
ger’). Even these under-estimated amounts, while being a relatively small part of
Baltic countries’ GDP at 2 to 2.5%, are substantial from the receiving families’
perspective. For instance, according to two independent surveys conducted in Lat-
via, 7.2% of population in 2007 and 6.5% in 2008 had family members working
abroad at the time of the survey (Table 4). In 2007, therefore, the $219 meant
$3000 for each person with a migrant worker in the family. Overall, about two
thirds of Latvian migrant workers were sending money home between 2004-2006
and 2006-2008 (Table 4). About two thirds of return migrants from the UK and
Ireland used foreign earnings for everyday needs; almost 30% invested in hous-
ing; 19% in a car; 22% in other durables; and 11% in education.
Figure 14 confirms the role of remittances in improving the standards of living in

Estonia: both before and after accession, the proportion of households facing financial
difficulties is significantly lower among those with family members working abroad.
Table 4
Migration networks, foreign work experience and remittances (Latvia, 2005-2008)
Survey: 2005/Q4-2006/Q1 a Survey: 2006/Q4-2007/Q1 a Survey: 2008/Q3 b

N = 8005 N = 10177 N = 4000
Networks % Networks % Networks %

Another family
member has worked
Another family member or Another family member has abroad during the
relative works abroad (left less worked abroad during the last last two years and
than 2 years ago) 6.8 three years 10.3 returned 6.1
Another family
Another family member or Another family member member works
relative works in some EU works abroad at the time of abroad at the time of
country (left in the last 2 years) 6.2 the survey 7.2 the survey 6.5
Another family member or
relative works abroad (left
more than 2 years ago) 7.2 Any of the above 12.0 Any of the above 12.6
A friend or neighbor works
abroad 71.5
Foreign work experience of the respondent
Has worked abroad
Has worked abroad during during the last two
Has worked in UK or Ireland 2.6 the last three years 5.0 years 3.3
Living abroad
With a family member or
partner 21.7
With another relative/friend 44.4
Alone 33.9
Remittances during
Remittances/Earnings the last two years
Incidence among
respondents with
own or family foreign
More than 50% 16.9 work experience 65.1
20 to 50% 15.7
Up to 20% 29.3
None 38.1
Spending abroad/ Earnings
All 6.9
More than 50% 60.0
20 to 50% 15.7
Up to 20% 16.9
Use of foreign earnings by
return migrants: Incidence by
purpose
Everyday needs 63.4
House 28.7
Car 19.1
Other durables 22.0
Education 11.3
Invest in own business 0.7
Source: Calculations with survey data.
Notes: a) See Sources of surveys V and VI in Table 2. Respondents aged 18 to 64 are included. b) Survey on drug use
conducted by SPI.
Figure 14
Estonian households with and without members working abroad by answers
to the question “How did you manage during this period?” (2002-2007)
Estonia
100%
80%
60%
40%
20%
0%
no HH members some HH no HH members some HH
abroad members abroad members
abroad abroad
2002-2003 2004-2007
Well Some difficulties Great difficulties
Source: Calculations with Estonian LFS data.
10 Conclusions
The onset of the worldwide economic crisis in 2008 raised the question whether
shrinking employment opportunities in the host countries combined with the de-
preciation of the British pound would trigger a sharp increase in return migration
(see e.g. Devine, 2009; Tapinsh, 2008). However there is a growing stock of evi-
dence that things are going to develop in the opposite direction. Most Baltic mi-
grants in the UK, Ireland and other host countries see staying as a much safer path
than returning, and some are thinking about Canada or Australia (see LETA 2008,
Meiden 2009, Sloga 2008 – the titles speak for themselves). The reasons are un-
derstandable. First, unemployment rates in the Baltic countries on the eve of 2009
were returning to the pre-accession levels with embarrassing speed. The season-
ally adjusted rates in January ranged from 8.6 in Estonia to 12.3% in Latvia, at
least twice as high as a year ago. Although the unemployment rates in the host
countries were rising as well, they were lower: about 6.5% in the UK and 3% in
Norway for example. Combined with more generous social protection systems,
risk considerations did not favor returning, let alone still substantial (price-
adjusted) wage differentials. Post-accession migrants from the new member states
in the UK and Ireland work longer hours and receive about a 10% lower hourly
wage than similar native workers in the same region and sector (Blanchflower and
Lawton, chapter 8 in this volume: Table 3; Blanchflower and Shadforth 2009: Ta-
ble 13; Barrett et al. 2008: Table 7)18, which certainly adds to their competitive-
ness, especially amid crises. Recent measures by the British government to restrict
access of non-EEA nationals to the UK labor market (see UK Border Agency
2009) will help Baltic workers as well. Baltic migrants interviewed acknowledge
that life has become more difficult since the start of recession, but they cope well
and look into future with optimism – something most of the stayers lack nowa-
days. The essence of what they say is: if you are not afraid of hard and in some
cases unpleasant work, you will finally find a job - or won’t lose one in the first
place. They also emphasize the increasing importance of local language skills.
Post-accession migrant flows from the Baltics were predominantly temporary

and circulatory. This pattern was most pronounced in the Latvian case, while Es-
tonian migrants showed relatively low propensity to return and longer periods of
stay in the host countries. Moreover, the flows to the UK and Ireland declined af-
ter 2005, which is unlike similar flows from Poland and Slovakia. However, big
and growing Lithuanian and Latvian diasporas have emerged in the UK and Ire-
land, and a big Estonian diaspora in Finland.19 Media and case studies evidence20
describe a rich social infrastructure within these diasporas: weekly newspapers
published in Lithuanian, Latvian and Russian; Sunday schools in all three Baltic
languages and Russian; regular Lithuanian and Latvian schools in Ireland, and Es-
tonian schools in Finland and Sweden; language courses, sport teams, discos, con-
certs and other activities; numerous NGOs (including religious communities); pro-
fessional networks with informal meetings at certain bars or Eastern-European and
Russian shops etc. In each host country, the Baltic diasporas interact with other
East Europeans, using Russian to communicate.
The above considerations suggest, moreover, that a new wave of emigration is

likely to emerge. The director of a recruiting company in Latvia says: “We have
returned to the situation of 2004, when thousands were leaving Latvia. We are
18 This wage gap is much less pronounced in the Nordic countries, see Gerdes and Waden-
sjö (chapter 7 in this volume).
19 The latest official estimates (Office for National Statistics, 2009) put the number of Lithua-
nian and Latvian nationals in the UK at 58 (s.e. 10) and 23 (s.e. 6) thousand. The true
numbers are likely to be higher. Bauere et al. (2007) report about 29 thousand Latvians
and about 5.5 thousand for Estonians for the period ending 2006. SKDS (2006b) reports
35 to 40 thousand Latvians in the UK. The unofficial estimates for Ireland are 50 to 100
thousand Lithuanians and 30 to 50 thousand Latvians, well above the 2006 Census esti-
mates of 25 and 13 thousand. According to Statistics Finland, 20 thousand Estonian na-
tionals lived in Finland in 2007.
20 See e.g. www.labrit.co.uk; www.mfa.gov.lv/lv/ireland/Vestnieciba/kopiena/;
www.lietuvis.ie/; www.saloje.lt/; www.newtotown.ie/forum/living-ireland/ireland-lithuanian-
perspective-517.html; www.mfa.gov.lv/en/ireland/news/embassy-
news/template/?pg=7256; www.sveiks.ie/; www.nashagazeta.com/; www.lti.ie/;
www.latviesi.org/ra/?slang=lv; http://baltic-ireland.eu/; www.lbi.ie/; SKDS(2006b); LETA
(2008); Mazan (2009); Meiden (2009).
struggling to serve all clients that knock at the doors, calls or send us e-mails. A
typical client is a man aged 30 to 40 ready for whatever type of work and nor very
demanding in terms of pay” (Mazan 2009). This is echoed by a colleague at a
company which offers placements in the UK, the Netherlands and Scandinavia in
sectors such as agriculture, shipping and distribution: “Interest in working abroad
had skyrocketed in recent months, particularly amongst people aged 20 to 30, both
unskilled workers and professionals, many of whom have been laid off or strug-
gling to repay debts” (McIntosh, 2009).
Note that the psychological cost of moving is much lower now than five years
ago. According to surveys conducted in Latvia between 2006-2008 (see Table 4),
at least 5% of working-age population are post-accession return migrants; 12%
have a family member with recent foreign work experience; and more than 70%
have a friend or neighbor with such experience. Moreover, adjustment is much
easier inside already existing diasporas, which also offer job opportunities for pro-
fessionals and personal service workers. The chances are high that Baltic migra-
tion will become more long-term or permanent in nature, and that the proportion
of high-educated migrants will grow. McIntosh (2009) reports a 26-year-old IT
worker holding a master degree saying: “I don’t see the way out right now actu-
ally. I am at point zero. I am just starting my career, but I don’t see the structure
here to develop myself in the labor market”; while a civil servant of the same age
says: “Some of my friends who have no work say that they don’t feel that they are
needed here in their country...We all thought that we are the next generation that
will change the future of the country, but now we have learnt more, we understand
that nothing will change”. Of course the actual size of the new emigration wave
will depend on how the crisis unfolds in the Baltics and elsewhere. There is a real
risk, however, that when entering the recovery, the Baltic economies might ex-
perience more severe labor shortages than before – and earlier than it was ex-
pected based on previous population projections. This highlights the relevance of
recommendations made by Boeri (2008) before the crisis started: “[The] Baltics
should plan on attracting migrants from non-EU Eastern countries and invest in
return migration”21.
21 The first steps to simplify procedures related to the employment of non-EU migrant work-
ers were made by Estonia in 2008 (see Nurmela, 2008) and are under consideration in
Latvia and Lithuania. According to the Lithuanian Ministry of Social Security and Labour
(2008: pp. 201-203), a comprehensive publication for migrants willing to come back to
Lithuania has been disseminated in 2007 and a project „Building-up a Positive Image of
Lithuania and Lithuanian Economic Migrants Seeking Their Return” has been launched in
2008.
References
Andrusz, K. (2009), Lithuania, Estonia could be on thin ice, IMF says.

www.bloomberg.com/apps/news?pid=20601095&sid=aWeB6GSOFbBM
van Baalen, B./T. Müller (2009), Return intentions of temporary migrants: The case of Ger-
many. Paper presented at Second Conference of Transnationality of Migrants (Marie Curie
RTN), Louvain-la-Neuve, 22-24 January 2009.
Baas, T./H. Brücker/A. Hauptmann (2009), Labor Mobility in the Enlarged EU: Who Wins,
Who Loses?, in: M. Kahanec/K. F. Zimmermann (eds.), EU Labor Markets After Post-
Barrell, R./J. FitzGerald/R. Riley (2007), EU enlargement and migration: Assessing the mac-
roeconomic impacts. NIESR Discussion Paper No. 292.
Barrett, A. (2009), EU Enlargement and Ireland’s Labor Market, in: M. Kahanec/K. F.
Zimmermann (eds.), EU Labor Markets After Post-Enlargement Migration. Berlin et al.,
141-157.
Barrett, A./S. McGuinnes/M. O’Brien (2008), The Immigrant Earnings Disadvantage across
the Earnings and Skills Distributions: The Case of Immigrants from the EU’s New Mem-
ber States in Ireland. IZA Discussion Paper No. 3479, Bonn.
Bauere, V./P. Densham/J. Millar/J. Salt (2007), Migrants from Central and Eastern Europe:
Local geographies, in: Population Trends No. 129, 7-19.
Blanchflower, D. G./H. Lawton (2009), The Impact of the Recent Expansion of the EU on the
UK Labour Market, in: M. Kahanec/K. F. Zimmermann (eds.), EU Labor Markets After
Post-Enlargement Migration. Berlin et al., 181-215.
Blanchflower, D. G./C. Shadforth (2009), Fear, unemployment and migration, in: The Eco-
nomic Journal 119 (535). F136-F182.
Blanchflower D. G./A. J. Oswald (1994), The wage curve. MIT press.
Boeri, T. (2008), Recent migration trends: Lessons for the Baltics. Presentation at the high-
level conference on "Convergence in the Baltics" organized by IMF and Eesti Pank in
Brussels, 1 February 2008.
Brücker, H./E. J. Jahn (2008), Migration and the Wage Curve: A Structural Approach to
Measure the Wage and Employment Effects of Migration. IZA Discussion Paper No.
3423, Bonn.
CSO Ireland (2008), Census 2006 – Non-Irish Nationals Living in Ireland. Dublin: The Sta-
tionery Office.
Devine, D. (2009), Many migrant workers from eastern Europe return home.
www.walesonline.co.uk/news/wales-news/2009/02/20/
Dustmann, C. (1997), Return Migration, Uncertainty, and Precautionary Savings, in: Journal
of Development Economics 52, 295-316.
Dustmann C. (2003), Return Migration, Wage Differentials, and the Optimal Migration Dura-
tion, in: European Economic Review 47 (2), 353-367.
Dustmann C./Y. Yoram Weiss (2007), Return Migration: Theory and Empirical Evidence
from the UK, in: British Journal of Industrial Relations 45 (2), 236-256
Dustmann, C./A. Glitz/T. Frattini (2008), The labour market impact of immigration. in: Ox-
ford Review of Economic Policy 24 (3), 477-494
Ehrenberg, R. G./R. S. Smith (2006), Modern Labor Economics, Theory and Public Policy.
9th edition, New York.
Employment and Economic Development Office of Finland (2008), Unemployment rate
among foreigners in 2001-30.4.2007. Helsinki.
European Commission (2008), Employment in Europe. Brussels 2008.
Fouarge, D./P. Ester (2007), Factors determining international and regional migration in
Europe. European Foundation for the Improvement of Living and Working Conditions.
Dublin.
Gerdes, C./E. Wadensjö (2009), Post-Enlargement Migration and Labor Market Impact in
Sweden, in: M. Kahanec/K. F. Zimmermann (eds.), EU Labor Markets After Post-
Hazans, M. (2003a), Determinants of Inter-Regional Migration in the Baltic Countries, ERSA
conference papers ersa03p354, 43rd Congress of European Regional Science Association.
http://ssrn.com/abstract=742184
Hazans, M. (2003b), Potential emigration of Latvian labour force after joining the EU and its
impact on Latvian labour market, BICEPS working paper, pp. 1-56.
Hazans, M. (2007a), Looking for the Workforce: The Elderly, Discouraged Workers, Minori-
ties and Students in the Baltic Labour Markets, in: Empirica 34 (4), 319-349.
Hazans, M. (2007b), Coping with growth and emigration: Latvian labour market before and
after EU accession.
Hazans M. (2007c), Latvian labour market before and after EU accession, in: Beyond Transi-
tion Newsletter 18 (1), 10 -11
Hazans, M. (2007d), Social and labour market integration of ethnic minorities in Latvia.
Background study for K. F. Zimmermann/M. Kahanec/A. Constant/D. J. DeVoretz/L. Ga-
taullina/A. Zaiceva (2008), Study on the Social and Labour Market Integration of Ethnic
Minorities. IZA Research Report 16, Bonn (unpublished manuscript).
Hazans, M. (2008), Post-enlargement return migrants’ earnings premium: Evidence from Lat-
via. Paper presented at EALE Conference in Amsterdam, 18-20 September 2008.
Hazans, M. (2009), Labour market integration of ethnic minorities in Latvia, in: M. Ka-
hanec/K. F. Zimmermann (eds), Ethnic Minorities in European Labour Markets: Chal-
lenges and Solutions. Cheltenham (forthcoming).
Hazans, M./I. Trapeznikova/O. Rastrigina (2008), Ethnic and parental effects on schooling
outcomes before and during the transition: Evidence from the Baltic countries, in: Journal
of Population Economics 21 (3), 719-749.
Hazans, M./K. Philips (2009), Baltic migrant workers before and after EU accession: Selectiv-
ity, ethnic and human capital effects (forthcoming).
Indans, I./A. Lulle/M. Laizane-Jurkane/L. Znotina (2006), Latvija and free movement of la-
bor: The case of Ireland (in Latvian). The Commission for Strategic Analysis: Riga (in
Latvian).
www.politika.lv/index.php?id=6133
Indans, I./K. Kruma (2006), Latvia’s immigration policy: Problems and perspectives. Latvian
Foreign Policy Institute – Konrad Adenauer Foundation: Riga (in Latvian).
www.liic.lv/uploads/filedir/File/LV_imigr_pol.pdf
Järv, K. (2007), Migration intentions of working-age population in Estonia, Policy Analysis
Series of the Ministry of Social Affairs No 8/2007.
http://213.184.49.171/eng/HtmlPages/series_20078eng/$file/series_20078eng.pdf
Kaczmarczyk, P./M. Okólsky (2008), Economic impact of migration on Poland and the Baltic
states. Fafo-Paper 2008:1, Oslo.
Kadziauskas, G. (2007), Lithuanian migration: Causes, impacts and policy guidelines, in: J.
Smith-Bozek (ed.), Labor mobility in the European Union: New members, new chal-
lenges. Center for European Policy Analysis, Washington, DC, 80-100.
Kahanec, M./A. Zaiceva (2009), Labor Market Outcomes of Immigrants and Non-Citizens in
the EU: An East-West Comparison, in: International Journal od Manpower 30 (1+2), 97-
115.
Kahanec, M./A. Zaiceva/K. F. Zimmermann (2009), Lessons from Migration after EU
Enlargement, in: M. Kahanec/K. F. Zimmermann (eds.), EU Labor Markets After Post-
Kallaste, E./K. Philips (2004), Potential of Estonia’s working age population to work in the
countries of the European Union. PRAXIS Centre for Policy Studies.
http://213.184.49.171/eng/HtmlPages/IVL2004_kokkuvote_eng/$file/IVL2004_kokkuvot
e_eng.pdf
Karu, M./K. Nurmela (2006a), Poor salaries lead to shortage of workers in the civil defence
sector.
www.eurofound.europa.eu/eiro/2006/05/articles/ee0605029i.htm
Karu, M./K. Nurmela (2006b), Negotiations over new wage agreement in healthcare.
Karu, M./K. Nurmela (2006c), Protest over wage discrimination in public transport sector.
Karu, M./K. Nurmela (2006d), Problem of acute labour shortage and migrant workers.
www.eurofound.europa.eu/eiro/2006/12/articles/EE0612019I.htm
Khan, K. (2008), Employment of foreign workers: Male and female labour market participa-
tion. Office for National Statistics.
www.statistics.gov.uk/articles/nojournal/MFMigAug08.pdf
LETA (2008), Latvians in Ireland feel crisis but don’t consider returning home.
www.nra.lv/zinas/11147-latviesi-irija-izjut-krizi-bet-negrasas-atgriezties-dzimtene.htm
Masso, J./R. Eamets (2006), Macro-level labour market flexibility in the Baltic States, in: T.
Paas/R. Eamets (eds), Labour market flexibility, flexicurity and employment: Lessons of
the Baltic States. New York.
Mazan, T. (2009), Ready to accept any job, just not in Latvia. Telegraf, 3 February 2009 (in
Russian).
www.telegraf.lv/index.php?act=archive&date=20090203&gid=23&id=43358
McIntosh, K. (2009), The great Latvian brain drain. The Baltic Times, 21 January 2009.
www.baltictimes.com/news/articles/22182/
Meiden, I. (2009), Despite crisis [Latvian ex-patriots] are not going to leave the Emerald Is-
land. Vesti Segodnja, 14 March 2009 (in Russian). www.ves.lv/article/73519
Ministry of Social Affairs of Estonia (2008), Health, labour and social sector in 2007. Tallinn,
www.sm.ee/fileadmin/meedia/Dokumendid/Sotsiaalvaldkond/inglisekeelsed/sotsmin_EN
G_trykki_1_.pdf
Ministry of Social Security and Labour of Lithuania (2007), Social report 2006-2007. Vilnius.
Ministry of Social Security and Labour of Lithuania (2008), Social report 2007-2008. Vilnius.
Ministry of Welfare of Latvia (2008), Latvian labour market in 2007. Riga.
www.lm.gov.lv/upload/darba_tirgus/darba_tirgus/latvijas_darba_tirgus_parskats_2007_ga
ds.doc (in Latvian)
Nurmela, K. (2008), Better provisions for non-EU migrant workers.
Office for National Statistics (2008), Employment of foreign workers. Quarterly LFS migrant
worker estimates, November 2008.
Office for National Statistics (2009), Population by country of birth and nationality, Jul07 to
Jun08.
www.statistics.gov.uk/StatBase/Product.asp?vlnk=15147
Pollard, N./M. Lattorre/D. Sriskandarajah (2008), Floodgates or turnstiles? Post-EU enlarge-
ment migration flows to (and from) the UK. Institute for Public Policy Research, London.
Republic of Estonia (2007), Operational programme for the development of the living envi-
ronment.
www.struktuurifondid.ee/public/elukeskond_ENG.pdf
Republic of Lithuania (2008), National report of Lithuania on social protection and social in-
clusion strategies 2008–2010.
http://ec.europa.eu/employment_social/spsi/docs/social_inclusion/2008/nap/lithuania_en.p
df
Rutkowski, J. (2007), From the Shortage of Jobs to the Shortage of Skilled Workers: Labor
Markets in the EU New Member States. IZA Discussion Paper No. 3202, Bonn.
SKDS (2006a), Society’s attitudes towards labor force migration: A survey of Latvian popula-
tion (in Latvian).
www.politika.lv/en/index.php?id=6134
SKDS (2006b), A study of Latvian diasporas. Part 1. 173 pp. (in Latvian).
www.integracija.gov.lv/doc_upl/petijums_diasporas07092006.doc
Sloga, G. (2008), It is better to be unemployed in Ireland. Diena, 31 October 2008 (in Lat-
vian).
www.diena.lv/lat/politics/sabiedriba/krize-2008-10-30-1
Soros Foundation Romania (2008), Children left at home. Risks and solutions.
www.osf.ro/en/
Sutela, H. (2005), Maahanmuuttajat palkkatyössä, in: S. Paananen (ed.) Maahanmuuttajien
elämää Suomessa. Statistics Finland, Helsinki, 83-109.
Tapinsh, A. (2008), Eastern Europe’s migrant workers weigh return home.
www.thaindian.com/newsportal/uncategorized/eastern-europes-migrant-workers-weigh-
return-home_10042107.html
UK Border Agency (2009), Migrant workers face tougher test to work in the United Kingdom
www.ukba.homeoffice.gov.uk/sitecontent/newsarticles/migrantworkerstoughertest
UK Home Office (2008), Accession Monitoring Report May 2004 - September 2008 (A8
Countries). A joint online report between the UK Border Agency, Department for Work
and Pensions, HM Revenue and Customs and Communities and Local Government.
University of Latvia (2007), The geographical mobility of the labour force,
www.darbatirgus.gov.lv/doc_upl/DGM_petijums_gala_eng_1_dala.pdf
Varblane, U. (Ed.) (2008), The Estonian economy current status of competitiveness and future
outlooks, in: Estonia in focus, 2008, No 1.
www.arengufond.ee/upload/Editor/English/ty_raport_eng.pdf
Vistesen, C. (2008), The CEE and the Baltics – Moving towards the Center of the Storm?, in:
Baltic Economy Watch.
http://balticeconomy.blogspot.com/2008/10/cee-and-baltics-moving-towards-center.html
Williams, C. C./P. Renooy (2009), Measures to combat undeclared work in 27 European Un-
ion member states and Norway: Overview report. European Foundation for the Improve-
ment of Living and Working Conditions, Dublin.
World Bank (2006), International migration, remittances and the brain drain.
World Bank (2007), Migration and remittances: Eastern Europe and the former Soviet Union.
World Bank (2008), Migration and remittances factbook 2008.
Zimmermann, K. F./M. Kahanec/A. Constant/D. J. DeVoretz/L. Gataullina/A. Zaiceva
(2008), Study on the Social and Labour Market Integration of Ethnic Minorities. IZA Re-
search Report 16, Bonn.
11. The Case of Albania
Zvezda Dermendzhieva
Randall K. Filer
The unique opening up of a previously closed society in post-communist Albania

was utilized to investigate the motivations for and impacts of emigration across
geographically close but economically diverse borders. As would be anticipated,
labor flows were very large across all segments of society. There is, however, ex-
tensive evidence that these flows may have significantly improved life for those
who remained behind. The consequences of Albania’s accession and those of simi-
larly affected countries in the European Union is discussed.
Migration, DOI 10.1007/978-3-642-02242-5_11,
306 Zvezda Dermendzhieva and Randall K. Filer
1 Documenting the Importance
The Albanian economy has some of the highest levels of migration and remittance
flows among the major post-communist economies of Central and Eastern Europe.
According to recent estimates, in 2005 Albania ranked fourth in the world in emi-
grants as a percentage of the population, with almost one third of Albanians
(27.5%) living abroad (World Bank, 2008). At the same time, remittance flows
have consistently amounted to a large share of the country’s GDP, exceeding by
several fold both net foreign investment and the amount of development aid re-
ceived by Albania. In 2006, for example, recorded remittances (generally admitted
to be only a fraction of the total) amounted to 13% of GDP and almost 4 times ei-
ther FDI or Development Assistance (World Bank, 2008).
2 Push and Pull Factors
Mass emigration and poverty go hand in hand in Albania. Heavy reliance on agri-
culture1 combined with a high level of land fragmentation, especially in rural ar-
eas, has induced many Albanians to search for employment abroad since the end
of communism. The beginning of the economic transition in Albania was also
marred by high inflation rates reaching 350% at the end of 1990, and high unem-
ployment due to the restructuring of the public sector and the inability of the Al-
banian economy, and especially its private sector, to grow fast enough to provide
jobs for the relatively young Albanian population (19.5% of the Albanian popula-
tion in 1989 was between the ages of 15 and 24) (Barjaba, 2000). In addition to
poverty and high unemployment rates, the lack of basic services and infrastruc-
ture, such as running water and electricity, is a strong push factor creating mass
emigration from post-communist Albania (Carletto et al., 2005; Stampini, et al.,
2008).
Short-term migration of a household member, mostly to neighboring Greece

and lasting between days and months, has become a common strategy for many
Albanian households to make ends meet (Carletto et al., 2006). Apart from being a
means of simple survival, emigration of a household member is often perceived to
make the difference between being poor and being relatively prosperous (De Soto,
et al., 2002).
Albania’s unemployment rate is significantly higher than that of neighboring

European Union members Greece and Italy, while the per capita income is ap-
1 More than half of the Albanian population still works in the agricultural sector (King and
Vullnetari, 2003).
11. The Case of Albania 307
proximately ninety% lower (De Soto et al., 2002). In addition to considerable gaps
in earnings and unemployment between Albania and its two EU neighbors, expo-
sure during the communist period is an important pull factor for the migrants from
Albania (Barjaba, 2000; Carletto et al., 2006). Many Albanians in the south of the
country had historic ties to Greece and spoke Greek at home. Exposure to Italian
television enabled Albanians, especially the younger portion of the population, to
become familiar with both the Italian language and mass culture, eradicating lan-
guage barriers and portraying an attractive life style abroad. As a result, almost
one half of the children who have left the home of their parents since 1990 had
lived abroad by 2002, primarily in Greece and in Italy (Carletto et al., 2006).
3 Patterns and Outliers
Current migration trends are likely to continue in the future because recent Alba-
nian migration has been driven by economic reasons (King, 2005) combined with
the fact that many Albanian households continue to live in poverty. According to a
poverty assessment study in Albania from 2002, an estimated 40% of the Albani-
ans have become worse off in terms of food intake and general socioeconomic
conditions when compared to the beginning of the transition period (De Soto et
al., 2002).
Two peaks in both temporary and permanent emigration from Albania can be
distinguished. Emigration from Albania increased sharply between 1990 and
1993. During this period the Albanian government liberalized the issuing of pass-
ports, giving the Albanians the right to leave the country and return freely. The
mass emigration during this period was preceded by smaller-scale emigration
starting in the summer of 1990 when around 5,000 Albanians sought refuge at
Western embassies in Tirana and were allowed to leave the country, most finding
refuge in Italy (King and Vullnetari, 2003). The first democratic elections in Al-
bania in 1991, however, brought political turmoil that led to the mass emigration
of Albanians either by boats to Italy, or through illegal crossing of the Albanian-
Greek border. Estimates are that between 1991 and 1993 approximately 300,000
Albanians, or ten% of the Albanian population, left the country, with most going
to Greece and Italy.
The second peak in emigration corresponds to the aftermath of the collapse of

the pyramid investment schemes in 1996-1997. An underdeveloped banking sec-
tor led many Albanians to put their savings, which were mostly from remittances
and retained earnings from abroad, in pyramid investment schemes that offered
monthly interest rates in excess of ten%. Contributions to these schemes are be-
lieved to have reached half of the country’s GDP in 1995-1996 (King and Vull-
netari, 2003; King, 2005). Most Albanians lost their savings, including savings
generated through migration abroad, after the investment schemes collapsed in
early 1997. The political and economic chaos that followed created a new wave of
emigration.
4 Estimates of Migrant Outflows, Return Flows

and Stock Abroad over Time
Thus, from a level of virtually zero emigration during the communist period Al-
bania rapidly became one of the top migration source countries in the world. Ac-
cording to estimates provided in Barjaba (2000), fifteen% of the Albanian popula-
tion lived abroad in 1998. By 2005, the World Bank (2008) estimated that 27.5%
of the Albanian population, or 860,485 Albanians resided abroad. In comparison,
the Bank estimated that the stock of emigrants as a percentage of the population of
all developing countries in 2005 was 2.7%.
Since most emigration from Albania has been illegal and unrecorded, these and
other estimates of the emigration flows taking place over time are speculative
(King, 2003). Due to the irregular character of Albanian emigration, there are also
no accurate estimates of the location of Albanian emigrants (Government of Al-
bania, 2005). Nevertheless, approximate figures on the stock and the distribution
by the host country of Albanian emigrants at different points in time are provided
by several different sources and summarized in Table 1.
Estimates from the Albanian Ministry of Labor and Social Affairs and Eurostat
suggest that the number of Albanian emigrants abroad in 1999 was approximately
800,000 (Barjaba, 2000). Estimates for 2001 from the Albanian Institute for Sta-
tistics (INSTAT) are based on the 2001 Housing and Population Census. These
official figures, however, are likely to be underestimated since they do not take
into consideration temporary migrants and the Albanians who have spent less than
one year abroad2 (King, 2003). The figures, however, are close to those provided
by the OECD for the same period based on country censuses, labor force surveys
and socio-economic surveys from different countries. The most recent estimates
from the Ministry of Labor and Social Affairs from 2004 report that the number of
Albanians living abroad exceeded one million (King, 2005; Government of Alba-
nia, 2005). Recent World Bank (2008) figures on migration combining various
data sources, however, suggest that the number of Albanians living abroad in 2005
was much lower at 860,485.
2 These migrants are still considered as household members by the census.

Table 1
Estimates of the number of Albanians living abroad: Various sources
Country 1999 2001 2005
1a 2b 3c 4d 5e 6f
Australia 1,451
Austria 856 1,834 2,000
Belgium 368 2,500 1,413 5,000
Canada 5,000 6,280 11,500
Czech Republic 178
Denmark 76 114
Finland 30 40
France 2,000 2,000 2,666 2,000
Germany 11,343 12,000 15,000
Greece 500,000 500,000 403,852 600,000
Hungary 74 141
Ireland 10 208
Italy 200,000 200,000 159,207 250,000
Japan 31
Luxembourg 33 223
Mexico 8
New Zealand 63
Norway 208
Poland 60
Portugal 8 41
Romania 16
San Marino 5
Slovak Republic 17
Slovenia 36
Spain 111 467
Sweden 145 538
Switzerland 791 1,504 1,500
The Netherlands 449 437 1,000
Turkey 2,000 3,313 5,000
United Kingdom 5,000 2,314 50,000
USA 12,000 39,780 150,000
Total 721,351 735,500 663,000 626,388 860,485 1,093,000
Source: Barjaba (2000), King (2003, 2005), OECD (2006), World Bank (2008).
Notes: The figures above are based on estimates originally provided by: aEurostat (1999); bThe Albanian Ministry of Labor
and Social Affairs (2001); cINSTAT(2002); dOECD (2006); eThe World Bank (2008); fThe Albanian Ministry of Labor and
Social Affairs (2004).
Figure 1 contains estimates of the temporary and permanent emigration flows

for the period 1990-2002 (Careletto et al. 2005). These estimates are based on the
2002 and 2003 Albanian LSMS and are lower than the true figures because they
do not take into account migrants who have left no household members behind.
Finally, Konica and Filer (2008) is the only study which provides estimates on re-
turn migration flows (see Figure 2). Unfortunately, since the survey used for this
study was conducted in early 1996, their estimates only cover the early stages of
the transition period before the collapse of the pyramid schemes. One can observe,
however, that during 1993-1994 there was a distinct slowing of the upward net
emigration trend and some increase in return migration. One can only speculate,
however, whether this trend has continued or has been reversed by events such as
the collapse of the pyramid investment schemes. Clearly the extent and behavior
of return migrants is an important area for future research.
Figure 1
Albanian Emigration (1990-2002)
500000
450000
400000
350000
300000 Temporary
migration
250000
Permanent
200000 emigration
Cum. Perm.
150000
emigration
100000
50000
0
1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002
Source: Carletto et al. (2005).

Notes: Based on the 2002 and 2003 Albania LSM.
Figure 2
Emigration and return migration (1989-1996: Konica and Filer)
200000
180000
160000
140000
120000
Emigrant
100000 Outflow s
80000 Returned
Migrants
60000
Net
40000 Emigration
20000
0
1989 1990 1991 1992 1993 1994 1995 1996a
Source: Konica and Filer (2008).

5 Estimates of Remittances
According to recent data, Albania is among the countries which receive the largest
amount of remittances relative to its GDP (World Bank, 2008). Recorded remit-
tances were approximately 1.5 billion US$ or 15% of the country’s GDP in 2006
and 2007. Since remittances are frequently transferred to households in Albania
through informal channels, these estimates are likely to understate the actual level
of inflows. According to the official statistics of the Bank of Albania, however,
remittances increased substantially between 1992 and 2006, although they fell as a
share of GDP and other measures of economic activity (Figure 3 and Table 2). Of-
ficial statistics substantially, however are likely to understate the amount of remit-
tances. Respondents interviewed in 1996 by Konica and Filer (2008) reported re-
ceipts of remittances that were 70% greater than the official figures in Figure 3.
Korovalis (1999) reports a similar figure and estimates actual remittances to Al-
bania are 75% higher than officially recorded flows. The underestimation of re-
mittances by official figures is reinforced by the fact that, in addition to informal
cash flows, it is hard with official data to detect the substantial flow of remittances
in the form of physical goods (see Konica and Filer, 2008 and Mancellari et al.,
1996).
Figure 3
Workers’ remittances (USD in millions), 1992-2006
1400
1200
1000
800
600
400
200
0
1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006
Source: World Bank World Development Indicators, 2008.

Table 2
Remittances* as share of GDP, FDI, and Official Development Assistance
and Official Aid (1992-2006)
Year Share of GDP Share of FDI Share of ODA and OA
1992 0.21 7.5 0.37
1993 0.22 4.74 0.93
1994 0.13 4.99 1.62
1995 0.16 5.49 2.13
1996 0.17 5.54 2.21
1997 0.12 5.62 1.61
1998 0.17 10.05 1.68
1999 0.10 8.66 0.73
2000 0.14 3.71 1.67
2001 0.15 2.97 2.28
2002 0.14 4.77 2.09
2003 0.14 4.37 2.23
2004 0.14 3.14 3.44
2005 0.14 4.49 3.64
2006 0.13 3.74 3.66
Source: World Bank World Development Indicators, 2008.
Notes: *Total workers' remittances in millions of current USD, official BoP statistics.
The cyclical pattern of emigration flows and remittances suggests that emigra-
tion from Albania is a strategic response to the need to find alternative income
sources in order to alleviate poverty at home. An increase in emigration in the af-
termath of the pyramid crisis in 1996-1997, for instance, was succeeded by an in-
crease in the amount of remittances received by the Albanian households in sub-
sequent years (Figure 3).
6 Data Sources Used in Previous Studies
A lack of individual and household data on the Albanian migrants and non-
migrants, largely due to the absence of appropriate sampling frames, has ham-
pered empirical research on migration from Albania and its consequences. Before
data from the first Albania LSMS from 2002 became available, a number of stud-
ies were conducted using self-collected surveys or convenience samples of mi-
grants and non-migrants and their households that achieved various degrees or
representativeness.
Table 3
Data Sources Used in Previous Studies
Table 3 (continued)
Data Sources Used in Previous Studies
Source: Own illustration.

This problem is not specific to Albania. Migrants are “rare elements” in the
population and representative surveys of migrant households are rare (McKenzie
and Mistiaen, 2007). Researchers tend to use specialized surveys of non-
probability samples of migrants, non-migrants, and their families which are often
based on the snowball method of collecting data, or data collected from regions
with high emigration. With few data sets using fully appropriate survey methods,
many studies are done on non-representative samples of the population of mi-
grants and their families and, therefore, provide biased results.
In the case of Albania, the small number of datasets available makes this prob-
lem even more severe. Many studies make use of the same data implying that their
results may not be independent. Table 3 summarizes the official statistics and
sample surveys used in the studies of Albanian migration to date. The three waves
of the Living Standards Measurement Survey have clearly done a professional job
of insuring randomness and representativeness, but they only provide data for the
period ten or more years after the start of the transition. Among earlier data
sources, the sample survey conducted by Konica and Filer appears to be the most
scientific.
7 Composition of Migrants
Despite the paucity of data, there are several studies of the characteristics of mi-
grants and the micro-level correlates of migration in Albania (see Table 4). These
results often need to be interpreted with caution in view of data limitations (Carletto
et al., 2005; Stampini et al., 2008). To illustrate, Gedeshi (2002), Kule et al. (2002)
and de Coulon and Piracha (2005) are based on data of unknown representativeness.
Labrianidis and Lyberaki (2004), and Labrianidis and Kazazi (2006) use data col-
lected from a snowball sample of migrants who returned from Greece or Italy. Ca-
vounidis (2004) bases her study of Albanian migrants in Greece on data from the
2001 Greek census and the Albanian migrants who have chosen to undergo legaliza-
tion in Greece in 1998.3 Castaldo et al. (2005, 2007) use data on intentions to mi-
grate, which may not be a good indicator of actual future migration, and is limited to
those who had not already migrated by 2002. Konica and Filer (2008), Carletto et al.
(2005), Carletto et al. (2006) and Stampini et al. (2008) use surveys that only con-
tain data on migrants who have left no household members behind. The sample in
Arrehag et al. (2006) is limited to one district in Albania characterized by very high
out-migration rates while the sample analyzed by Germenji and Swinnen (2005)
consists of rural households only.
3 Studies using Greek data must always be interpreted with caution because it is often difficult
to differentiate between Albanians and Kosovars and Macedonians of Albanian ethnicity.
Table 4
Composition of migrants
Table 4 (continued)
Table 4 (continued)
Table 4 (continued)
11. The Case of Albania
Source: Own illustration.
Notes: a) The study refers to the Albanian migrants in Greece only; b) The authors use intentions data to examine the factors that influence the propensity to consider migration among those
who have not migrated in the preceding five years; c) The study is based on a survey conducted in the Korce district; d) The study is on temporary work migration from rural areas only.
319
Education most likely affects the migration decision in a non-linear way. In the
rural regions of Albania individuals with secondary education are more likely to
emigrate than those with only primary education (Germenji and Swinnen, 2005).
Having a university degree, however, is not associated with a higher propensity to
emigrate. These results may be due to the better income and employment oppor-
tunities in Albania for college graduates combined with a low return on Albanian
education in foreign labor markets.
There is, however, evidence of positive selection on education at all levels

among female migrants (although not among the male migrants), but that the im-
portance of education declines over time for both females and males, and the de-
cline is more pronounced for males (Stecklov et al., 2008). This finding might be
a result of increasing labor market inequality within Albania, which increases the
returns to migration for better educated females in comparison to the males.
Greece appears to attract less skilled and lower income migrants than Italy
(Labrianidis and Lyberaki, 2004), perhaps due to both higher migration costs to
Italy, as well as the peculiarities of the Greek labor market given that Greece has
the largest shadow economy among EU countries, making it easier for even illegal
immigrants to find an unskilled job (Baldwin-Edwards, 2004).
The characteristics of Albanian migrants in different countries may depend on

the type of migration in which they are involved. During the summer holidays, for
instance, a number of teachers and public sector employees migrate temporarily to
Greece for employment in the agricultural sector to supplement low public sector
wages. Earnings from such seasonal work may well exceed the annual salary of a
secondary school teacher in Albania (King and Vullnetari, 2003). Thus, there may
be a lower than expected correlation among education and job level for migrants
to Greece.
Albanian migrants originate primarily from the coastal regions and the regions
close to the border with Greece. The Greek border can be crossed on foot, which
significantly reduces migration costs for those who live in the south of Albania
(Germenji and Swinnen, 2005). Older individuals, however, may have more trou-
ble coping with the physical difficulties encountered crossing the Albanian-Greek
border (De Soto et al., 2002).
Migrants typically come from lower-income households. Migrants’ networks in

the destination country are likely to have a significant role in reducing the cost of
migration for these individuals. The poorest households, however, are often not
able to send migrants abroad due to an inability to finance the original move (De
Soto et al., 2002; Germenji and Swinnen, 2005). Indeed, according a recent World
Bank poverty assessment report on Albania cited by De Soto et al. (2002), one
definition of being poor in Albania is not being able to send a migrant abroad.
Furthermore, considerable anecdotal evidence exists, at least from rural areas, that
access to different types of income affects the decision to send a migrant abroad.
Access to a non-farming income is associated with a lower propensity to send a
migrant abroad. This pattern implies that migration is one of several ways used to
diversify income sources for the rural population (Germenji and Swinnen, 2005).
While the first emigration wave from Albania in the early 90s consisted pri-
marily of illegal migrants, a large number of those currently abroad are legal.
While this tendency may have resulted in part from differential return rates, espe-
cially due to capture by foreign immigration authorities, many Albanians in
Greece and Italy (the two predominant destinations for Albanian migrants) have
taken advantage of the opportunities given by the Greek and Italian governments
to legalize their status. Since 1999 the balance between the number of illegal and
legal Albanian migrants has been changing in favor of the legal ones (Barjaba,
2000). After the regularization procedures in Greece and in Italy, the estimated ra-
tios of irregular to regular migrants in 2000 were 1:4 in Italy and 1:1 in Greece.
Following legalization of their status, many emigrants have also chosen to re-
unite with their families in the host countries (de Zwager et al., 2005). This has led
to an increase in female emigration from Albania in recent years, with the number
of female migrants in Greece increasing two-fold in 2001 compared to the first
emigration wave (King and Vullnetari, 2003). Illegal crossings of the Albanian-
Greek border are typically undertaken by groups of Albanian men. Thus, when
men and women are compared, a larger share of the female migration is legal
(Baldwin-Edwards, 2004; Arrehag et al., 2006). In addition to moves for family
reunification in the host country after regularization, female migration has risen
relative to male migration due to the USA’s Diversity Visa programs and Can-
ada’s Skilled Worker Visa program. Finally, the ratio of female to male migrants
in recent years has increased because it has become more common for young Al-
banians to emigrate to obtain an education (King and Vullnetari, 2003).
The size of temporary migration flows towards more distant destinations, such
as Italy and Germany, is also slowly increasing. This may be due to accumulating
household migration experience and improving migration networks, which reduce
the cost of more distant migrations (Carletto et al., 2005; Carletto et al., 2006;
Stampini et al. 2008).
Two types of returning migrants can be distinguished – those who return to re-
settle in Albania and circular migrants for whom a return precedes another migra-
tion to the same or another country King and Vullnetari (2003). Circular migra-
tion, whereby migration episodes and temporary returns to Albania alternate, is
widespread among Albanians (Labrianidis and Lyberaki, 2004; Labrianidis and
Kazazi, 2006). Circular migrants tend, however, to spend only short periods
abroad. Labrianidis and Lyberaki (2004) report that about 40% of the respondents
in their sample of migrants return to Albania once a year and another 28.5% return
even more frequently. Because of the geographic proximity and the land border
between Albania and Greece, which is easier and less costly to cross than the sea
border with Italy, migrants to Greece return to Albania substantially more fre-
quently than those to Italy. For the sample of migrants in Labrianidis and Lyberaki
(2004), the percentage of migrants who have migrated more than once to Greece
is 28% while and the percentage of migrants who have migrated to Italy more than
once is 3.6%. Furthermore, their study indicated 23% of the return migrants in the
sample move back and forth between Albania and Greece or Italy, thus “living in
two countries” and taking advantage of the seasonal and temporary employment
opportunities in the Greek and Italian labor markets.
Permanent return migrants may come back to Albania for diverse reasons. For
many Albanians the purpose of emigration is not to settle permanently in the host
country, but rather to accumulate sufficient earnings abroad to establish a small-
scale, household business (Labrianidis and Liberaki, 2004; Nicholson, 2001,
2004; Cuka et al. 2003). In samples of returned migrants, the average stay abroad
is seven years (Labrianidis and Lyberaki, 2004); and remittances, which are fre-
quently used to establish small enterprises, tend to be saved until migrants return
from abroad (Konica and Filer, 2008).
Greece might be the most feasible destination country because of its close
proximity, but it is not the most desirable. The money saved from working there
may also be used to finance onward migration to more distant countries (King and
Vullnetari, 2003; Baldwin-Edwards, 2004; Labrianidis and Lyberaki, 2004). Oth-
ers may return to Albania because they are disappointed with the jobs they can ac-
tually obtain abroad or discover improved opportunities in Albania. Furthermore,
many returns to Albania are involuntary and result from enforcement actions by
destination countries’ police forces (Konica and Filer, 2008).
Temporary migrants can be distinguished from permanent ones. Those in Italy

and more distant countries tend to be better educated than either non-migrants or
migrants in Greece (Carletto et al., 2005; Carletto et al., 2006; and Stampini et al.,
2008). Temporary migrants to Greece also come from relatively lower income
households. This may be explained by the fact that relatively more wealthy
households or more educated migrants are more likely to afford the higher cost of
a more distant migration, as well as by the fact that it might be easier for the better
educated Albanians to adapt and find employment in Italy, where the wages are
higher and more non agricultural employment is available than in Greece. Tempo-
rary migrants come mostly from mountainous rural areas and the central regions
in Albania. Permanent migrants to Italy originate mostly from Tirana and the ur-
ban coastal and mountain areas. Temporary migrants to Greece mostly originate
from the rural areas in the northern and central parts of Albania, whereas tempo-
rary migrants to Italy and other countries in Europe mostly originate from the ur-
ban coastal areas, agricultural communities and the mountain regions in Albania
(Carletto et al., 2005). The rural areas of the district of Korce are more likely than
the urban areas to send seasonal migrants abroad. This is probably due to the
fewer sources of income available to rural households (Arrehag et al., 2005).
8 Potential Migration
Evidence from data regarding stated intentions to migrate is largely consistent

with the evidence based on data on actual migration. Younger individuals (mostly
in the 26-35 age group) and males are more likely to be considering emigration,
and the unemployed and individuals with secondary and vocational education ex-
hibit the greatest desire to migrate. Outside of Tirana, and especially in rural areas,
those who are relatively less poor are more likely to consider migration. Females
in households with permanent migrants are more likely to be planning to migrate
than those in households without migrants (Castaldo et al., 2005, 2007).
The gaps in earnings and employment between Albania and its EU neighbors
continue to be large, implying that the migration patterns of recent years are
unlikely to change in the near future (Barjaba, 2000; Castaldo et al., 2007). Al-
though many Albanians believe that migration to Greece is less costly and that it
is easier to find work in Greece than in other countries, potential migrants from
Albania state that the most desirable destinations are the USA, Germany and Italy
(Baldwin-Edwards, 2004; Cavounidis, 2004). Migration to Greece is often re-
garded as temporary and only a precursor to further and more permanent migra-
tion to another EU country or to North America (King and Vullnetari, 2003;
Labrianidis and Lyberaki, 2004). Relaxing the visa regime in the EU for Albani-
ans by the EU will further reduce the cost of emigration to EU destinations other
than Greece, enticing more Albanians to migrate directly to those countries. Such
visa relaxation would also lead to an increase in the migration of relatively better
educated Albanians. Past evidence suggests that more educated workers (with bet-
ter options in Albania itself) are less willing to migrate illegally and subsequently
be forced to work in the irregular economy upon arrival in the destination country
(Labrianidis and Lyberaki, 2004; Labrianidis and Kazazi, 2006; Carletto et al.
2005; Carletto et al. 2006; Stampini et al. 2008). Thus, an EU visa policy change
could have implications for the potential brain drain from Albania.
9 Impacts of Migration
As one of the countries most affected by migration after the end of the communist
era in Central and Eastern Europe, Albania has been referred to as a “laboratory
for the study of migration and development” (King, 2005). The direction of the
causal relationship between migration and economic development in Albania,
however, is not necessarily obvious (King, 2005). Several factors are at work, in-
cluding: brain drain effects, the impact of remittances on consumption demand
and business formation, and the joint impact of migration and remittances on labor
supply.
Evidence on the brain drain from Albania is both scarce and controversial. Ac-
cording to World Bank estimates (2008), the cumulative migration rate of the ter-
tiary educated from Albania was 20% by 2000. Since the majority of the Albani-
ans who complete a doctoral degree in Western Europe or in the US do not return
to Albania (Germenji and Gedeshi, 2008), one may argue that this percentage is
much higher. According to data from UNDP and the World Bank, over 25,000
Albanian university-level students were studying abroad in 2006, whereas the es-
timated number of Albanian students enrolled in public universities in Albania
was only around 43,600. Many of these locally-educated Albanians eventually en-
roll in graduate schools abroad (Germenji and Gedeshi, 2008). Thus, with increas-
ing migration for education purposes, much of the current danger of brain drain
from Albania likely comes not from the fact that the more educated emigrate, but
rather from the fact that the Albanians who obtain their education abroad find no
reason to return to Albania.
The USA and Canada are currently the primary destinations for Albanian mi-
grants with university educations (Germenji and Gedeshi, 2008). The interviews
Germenji and Gedeshi conducted with representatives of universities and research
institutions in Albania revealed that between 1991 and 2005 approximately 1,295
scientific workers emigrated, even though the total number of persons employed
at universities and research institutes in Albania in 2005 was only about 2,500.
Thus, about half of the scientific work force has emigrated, most of them with
their families. The majority of those who left were in the 25-34 age group at the
time of their emigration, representing a serious loss of the country’s most capable
scientific workers.
A problem that has only recently been addressed by the literature on Albanian
migration is that emigration affects elderly people through the so-called “care
drain.” Although partially offset by the beneficial role of remittances for poverty
reduction in Albania, emigration of the young puts substantial pressure on the in-
adequately developed elderly care system in Albania (De Soto et al., 2002; Vull-
netari and King, 2008).
Migration significantly affects the home economy and the household members
left behind through the remittances that migrants send and the savings the mi-
grants accumulate abroad. The majority of the Albanian migrants send remit-
tances, generally to immediate family members (de Zwager et al., 2005). The
amount of remittances sent to wives and children of emigrants is much larger than
the amount sent to other family members (Gedeshi, 2002).
Remittances to Albania are typically sent irregularly and through informal

channels due to the underdevelopment of the Albanian banking system and its
limited geographical scope (Gedeshi, 2002) and the proximity of the major host
countries that allows for frequent returns of both migrants themselves and mem-
bers of their networks (Arrehag et al., 2005; de Zwager et al., 2005).
Recent empirical evidence from developing countries in general suggests that

remittances reduce poverty (Adams, 2007). There is also evidence of a correlation
between migration and poverty in Albania. A majority of Albanian households
who identify themselves as relatively more prosperous have had at least one emi-
grant family member who spent between 13 and 60 months abroad during the 90s
De Soto et al. (2002). Not surprisingly, most remittance income is spent on con-
sumption and necessities, while the second most common use of remittances is re-
furbishing or building a house (De Soto et al., 2002; Gedeshi, 2002; King and
Vullnetari, 2003; Konica and Filer, 2008). Albanian households that receive re-
mittances have higher consumption patterns with respect to basic necessities such
as food than households without remittances (Castaldo and Reilly, 2007). As
summarized by Mancellari et al. (1996):
“[remittances] raised considerably disposable incomes and spending within

the country, and the demand for imported consumer and production goods.
This in turn has stimulated a rapid rise in small-scale trade and investment
projects. Non-cash remittances include not just consumer goods, such as
electronics and second-hand cars, but also second-hand production goods
such as vans, tractors, and other machinery. Therefore, not only do remit-
tances give the economy a new, prosperous look, with many new shops and
restaurants, but they also have a direct and significant effect on production,
by easing the foreign exchange constraint on buying key inputs” (p. 483).
Interviews with migrants’ households in Albania indicate that urban house-

holds also tend to use remittances to finance education-related expenditures while
rural households pay debts or invest in agriculture (Arrehag et al., 2005; de Zwa-
ger et al., 2005). Furthermore, permanent migration of former household members
leads to more investment in livestock production and to higher agricultural and to-
tal household incomes McCarthy et al. (2006). Previous migration of a current
household member reduces livestock holdings but increases fruit cultivation (es-
pecially in the case of previous migration to Greece). This difference might be a
result of the experience the migrants acquire as agricultural workers in Greece.

Migrants’ households make fewer agricultural investments in crop production,
such as fertilizer and equipment, but invest more in livestock production (Miluka
et al., 2007). This pattern may be a result of the differing labor intensity of various
agricultural activities combined with the impact of prime age males being missing
from households through migration.
Albanian households that receive remittances from abroad have significantly

higher than average budget shares spent on durables and utilities (Castaldo and
Reilly, 2007). Migrants and their households in Albania typically invest in small
retail or hospitality household businesses, such as shops, bars, restaurants or small
hotels (King and Vullnetari, 2003). Thus, taking into account the small scale of
household entrepreneurial projects in Albania, durable goods such as refrigerators,
cars, or agricultural equipment might be considered productive investment as they
are often used by the Albanians in their household business (Nicholson 2001,
2004; Castaldo and Reilly, 2007).
Anecdotal evidence suggests that the effect of migration on the labor supply of
the non-migrant family members left in Albania can be positive since other mem-
bers of households where the males are abroad need to compensate for the ab-
sence of the migrants by working more. This is particularly true in rural areas,
where it is common for the households to work on their own land (De Soto et al.,
2002; Vullnetari and King, 2008). Remittances, however, are a source of non-
labor income for those household members who remain in Albania, and may re-
duce the probability of Albanian women working (Konica and Filer, 2008). A dis-
tinct negative effect can be found on the labor force participation for women, al-
though the effect is limited to households with a migrant in Italy (Carletto et al.,
2006). This pattern may be because the earnings of the migrants in Italy are
higher. Such an effect is important in the case of Albania as unemployment pres-
sures on the local labor market are decreased not only through emigration, but also
through the reduction in female labor supply.
Data from 2005 show that members of households with migrants spend fewer
hours working in agriculture, both in total and in per capita terms (Miluka et al.,
2007). In order to partially offset missing male labor, however, women in such
households work more than females in households without migrants. There is also
evidence that migrants’ households make fewer agricultural investments in crop
production, but they do invest in livestock production. This evidence leads to the
conclusion that migration in Albania may be a strategy to divest from crop pro-
duction and invest into livestock. One can expect that this may eventually lead to
land sales, land consolidation, and an agricultural sector with better investment
and return prospects in Albania.
Macroeconomic data from a large number of countries for the period 1975-
2003, indicate that the amount of remittances has a positive impact on the ratio of
bank credit to the private sector and the share of bank deposits as a percentage of
GDP (Aggarwal et al., 2006). There is, however, no evidence of this effect in Al-
bania. One suggestion is that the Albanian financial sector is currently not able to
effectively allocate remittances to productive activities and that, while the banking
sector has been somewhat more successful recently in attracting the savings of the
migrants, the majority of Albanian households still place little trust in the formal
financial sector (Uruci and Gedeshi, 2003).
Leon-Ledesma and Piracha (2004) studied the impact of return migration on

economic development during the transition from communism from a macroeco-
nomic perspective. Due to the fact that migrants’ remittances and savings can be
used to finance consumption and productive investment that lead to job creation,
they find evidence from the Central and Eastern European countries (unfortu-
nately from a sample that does not include Albania) that a larger number of return
migrants increases labor productivity. Although there is little evidence of large in-
vestments resulting solely from migration and remittances in Albania (Barjaba,
2000), there is more than anecdotal evidence that they support the private sector
development by allowing the migrants’ households to invest in small-scale house-
hold business projects (Konica and Filer, 2008, Kilic et al., 2007).
It is the return of migrants themselves that is critical in establishing small en-

terprises, rather than the simple sending of funds to family members left behind
(Konica and Filer, 2008). A specialized survey revealed that although most Alba-
nian migrants work illegally and in low-skilled occupations, the savings and skills
acquired abroad help to establish their own business upon return. Thus, the tempo-
rary migration in Albania may be motivated by the possibility of improving the
migrants’ chances of finding better employment opportunities upon return (Kule
et al., 2002). The same data also revealed that return migrants are almost twice as
likely to be self-employed as those who never migrated (de Coulon and Piracha,
2005). Return migrants from Greece and Italy are better off in terms of both em-
ployment and living conditions than they were prior to migration, and return mi-
grants from Greece perform relatively better in the Albanian labor market upon re-
turn than those who migrated to Italy (Labrianidis and Lyberaki, 2004; Labrianidis
and Kazazi, 2006). A positive correlation between the length of stay abroad and
the likeliness of owning a business upon return is also evident,4 More recent na-
tionally representative data from the 2005 Albanian LSMS show that men with
previous migration experience are more likely to work off-farm and be self-
employed (Carletto et al., 2006). A study that attempts to control for possible en-
dogeneity of the migration decision and length of stay abroad finds that past
4 All the findings, however, are not based on nationally representative survey data and,
therefore, should be interpreted with caution.
household migration experience has positive impact on the probability of owning

a business and that work experience in Italy is relatively more important than a
migration period in Greece (Kilic et al., 2007). The positive effect is seen, how-
ever, only for migration experience during 1990-2000. Based on the findings, the
authors suggest that more recent migrants may still need to acquire additional
skills and/or savings in order to complete their “migration cycle” and establish a
business upon return.
Although significant return migration may be desirable for a number of rea-

sons, the return of those migrants who emigrated in the mid-1990s during the first
wave of emigration to Greece and Italy was interrupted by the pyramid scheme
crisis that sent a new wave of Albanian migrants abroad (de Zwager et al., 2005).
During recent years there has been little evidence of voluntary return migration,
and those who do return are mostly involuntarily returned by foreign immigration
authorities, seasonal workers in Greece, or household members who spend short
periods abroad to augment the insufficient incomes they earn in Albania or to ac-
cumulate savings that would enable them to invest in small household business at
home. It may also be that return migration is motivated partially by “push” factors
such as a failure to integrate abroad or to earn enough to be able to send remit-
tances (Barjaba, 2000; King, 2005) rather than the attraction of increased opportu-
nities in Albania.
Although the Albanian government and the International Organization for Mi-
gration (IOM) in Tirana have recently made efforts to attract return migrants by
providing material and financial assistance to those who choose to return and rein-
tegrate, the number of emigrants who have returned this way has been insignifi-
cant. Between 2000 and 2004 just over 300 emigrants took advantage of these
programs (de Zwager et al., 2005).
More than half of Albanian migrants abroad who have been interviewed by
IOM Tirana would like to return to Albania in the future, possibly during the pe-
riod 2010-2015. These predictions, however, can only be based on stated inten-
tions to return rather than what actual happens. It is unrealistic to expect that sig-
nificant return migration will occur in the near future, particularly among the
economically active population abroad, as long as the large income gaps between
Albania and the destination countries persist (Vullnetari, 2007).
A further impediment to potential return flows is the increase in family reuni-

fication emigration during recent years, especially after the regularization proc-
esses in Greece and Italy (Vullnetari 2007). The return of children of current
migrants who have settled abroad seems unlikely. Some return after retirement
among Albanians who emigrated during the first waves may, however, start to
occur due to cost of living differentials. Such return migration, if it occurs, is
likely to be directed mostly towards Tirana or urban areas. It will further exac-
erbate housing shortages, and is unlikely to inject significant investment into the
Albanian economy.
10 Analysis, Trends and Policy Recommendations
The prospects of significant return migration to Albania are bleak, largely due to a
lack of investment opportunities. This situation is aggravated by the persistently
low incomes and spending capacity in Albania, the lack of long-term credit, poor
basic services, and underdeveloped infrastructure.
The trends towards regularization and integration in the host countries also
make future return migration of the younger generation unlikely. Albanians cur-
rently account for over half of the total number of immigrants in Greece and,
compared to other immigrant groups, a large number of Albanian children attend
Greek schools (Cavounidis, 2004). The growth in remittances to Albania may also
decrease in the future with family reunion and integration trends becoming
stronger in the host countries (Gedeshi, 2002).
While investment in infrastructure, especially in the rural areas, may decrease

emigration and attract return migration through increasing the investment oppor-
tunities in the Albanian private sector, access to alternative income sources, lower
migration costs and higher incomes for the poorest households may lead to more
households being able to send migrants abroad (Germenji and Swinnen, 2005).
Furthermore, the potential increase in emigration towards other EU countries may
lead to further human capital flight from Albania. Overall, it seems unrealistic to
expect major changes in the pattern of Albania being a major exporter of labor in
coming years.
The possibility of future EU enlargement to include Albania, as well as coun-

tries that are similar to Albania in terms of economic conditions and recent emi-
gration patterns, such as Moldova, Macedonia (already a candidate) and other
countries in the Western Balkans and the South Caucasus, calls for a comparison
with previous EU enlargements in order to speculate about the potential impact on
migration in Europe following further EU enlargement.
Currently, Albania is at the upper end of the distribution among the potential
accession countries and significantly above the recent EU joiners (see Table 5) in
both the number of migrants and remittances as a share of GDP.5 Official esti-
5 The comparison with Cyprus and Malta is not indicative as both economies are relatively
small and have simultaneously experienced large immigration inflows.
mates of the remittances as a percentage of GDP in 1994 which are available for
some of the transition countries6 indicate that, unlike the earlier joiners from 2004,
the economies of the next wave of potential accession countries have been heavily
and consistently reliant on remittances. In 1994, for instance, remittances ex-
ceeded 15% of Albania’s GDP, while, with the exception of Cyprus and Slovenia,
remittances in all other countries were below 1% of the GDP.
Table 5
Albanian emigration and remittances compared to neighboring transition countries
Emigrants as %
of Population Remittances as
(2005) % of GDP (2006)
Albania 27.5 14.9
Non-Accession Transition Countries
Armenia 26.9 18.3
Azerbaijan 16.2 4.0
Belarus 18.4 0.9
Bosnia & Herzegovina 37.7 17.2
Croatia 16.0 2.9
Georgia 22.9 6.4
Macedonia 18.2 4.0
Moldova 16.8 36.2
Russia 8.0 0.3
Serbia & Montenegro 21.9 13.8
Ukraine 13.1 0.8
Recent EU Accession Countries
Bulgaria 12.1 5.4
Cyprus 19.2 n.a.
Czech Republic 4.1 0.8
Estonia 13.7 2.4
Hungary 4.7 0.3
Latvia 10.1 2.4
Lithuania 9.3 2.1
Malta 26.5 0.6
Poland 6.0 1.3
Romania 5.7 5.5
Slovak Republic 9.6 0.8
Slovenia 6.8 0.8
Source: World Bank (2008).
Furthermore, a comparison of the change in the official remittances between

2002 (one year before the accession) and 2005 (one year after the accession) for
the group of countries that joined the union in 2004 and two control groups of
countries during the same period (low income prior EU members7 and non-
6 Data on remittances as a percentage of GDP in 1994 is available for Cyprus, the Czech
Republic, Estonia, Lithuania, Malta, Poland, Romania, Slovakia, and Slovenia from the
World Bank WDI 2008.
7 Greece, Spain and Portugal.
accession countries in Central and Eastern Europe8) shows that remittances for the
2004 accession wave increased substantially more than the remittances for the
other two groups. Remittances to joiners increased by 160%, compared with an
increase of 85% for non-accession CEE countries and a fall of 5% for remittances
to low-income current EU members.9 The larger increase in remittances after ac-
cession, despite restrictions on labor migration imposed by some old EU mem-
bers, suggests that a potential EU accession is unlikely to undermine the impor-
tance of remittances as a source of household income not only for the Albanian
economy, but also for the economies of other potential joiners which currently ex-
hibit migration patterns similar to those of Albania.
In short, the issue of potential EU accession and its effect on Albania is com-
plex. There is no reason to believe that emigration will not continue to play an im-
portant role for Albania. Indeed, accession may even increase outflow in the ab-
sence of severe restrictions on labor mobility, especially that of highly trained
workers, that would deny the principle of Europe being a single economic space.
On the other hand, in a global economy it is possible that Albania’s comparative
advantage may well be a skilled, motivated labor force, and that the long-run de-
velopment of the country can best be promoted by its neighbors recognizing that
easy access to improved opportunities abroad may well be the best form of devel-
opment assistance that can be provided to Albania and other, similarly situated,
countries.
8 Albania, Bosnia, Bulgaria, Croatia, Macedonia, Moldova and Serbia.

9 These figures exclude Estonia, Slovakia and Romania, where data inconsistencies cause
us to doubt the reported magnitude of remittance increases.
References
Adams, Jr., R. (2007), International Remittances and the Household: Analysis and Review
of Global Evidence. World Bank Policy Research Working Paper 4116.
Aggarwal, R./A. Demirguc-Kunt/M. Peria (2006), Do Workers’ Remittances Promote Fi-
nancial Development? World Bank Policy Research Working Paper 3957.
Arrehag, L./Ö. Sjöberg/M. Sjöblom (2005), Cross-Border Migration and Remittances in a
Post-Communist Society: Return Flows of Money and Goods in the Korçë District, Al-
bania, in: South Eastern Europe Journal of Economics 1, 9-40.
Arrehag, L./Ö. Sjöberg/M. Sjöblom (2006), Post-Communist Cross-Border Migration in
South-Eastern Albania: Who Leaves? Who Stays Behind?, in: Journal of Ethnic and
Migration Studies 32 (3), 377-402.
Baldwin-Edwards, M. (2004), Albanian emigration and the Greek labour market: Economic
symbiosis and social ambiguity, in: South-East Europe Review 1, 51-66.
Barjaba, K. (2000), Contemporary patterns in Albanian emigration. South-East Europe Re-
view 3 (2), 57-64.
Carletto, G./B. Davis/M. Stampini (2005), Familiar Faces, Familiar Places: The Role of
Family Networks and Previous Experience for Albanian Migrants. FAO, ESA Working
Paper 05-03, Rome.
Carletto, G./B. Davis/M. Stampini/A. Zezza (2006), A country on the move: international
migration in post-communist Albania, in: International Migration Review 40 (4), 767-
785.
Castaldo, A./J. Litchfield/B. Reilly (2005), Migration and poverty in Albania: what factors
are associated with an individual’s predisposition to migrate?, in: Journal of Southern
Europe and the Balkans 7 (2), 157-173.
Castaldo, A./J. Litchfield/B. Reilly (2007), Who is Most Likely to Migrate from Albania?
Evidence from the Albania Living Standards Measurement Survey, in: Eastern Euro-
pean Economics 45 (5), 69-94.
Castaldo, A./B. Reilly (2007), Do Migrant Remittances Affect the Consumption Patterns of
Albanian Households? in: South-Eastern Europe Journal of Economics 1, 25-54.
Cavounidis, J. (2004), Migration to Greece from the Balkans, in: South Eastern Europe
Journal of Economics 2, 35-59.
Çuka, E./H. Papapanagos/N. Polo/P. Sanfey (2003), Labor Market Developments in Alba-
nia: An Analytical Overview, in: Review of Development Economics 7 (2), 217-227.
de Coulon, A./M. Piracha (2005), Self-selection and the performance of return migrants:
the source country perspective, in: Journal of Population Economics, 18 (4), 779-807.
De Soto, H./P. Gordon/I. Gedeshi/Z. Sinoimeri (2002), Poverty in Albania. A Qualitative
Assessment. World Bank Technical Paper 520.
de Zwager, N./I. Gedeshi/E. Germenji/C. Nikas (2005), Competing for Remittances. Ti-
rana: IOM.
Gedeshi, I. (2002), Role of Remittances from Albanian Emigrants and Their Influence in
the Country’s Economy, in: Eastern European Economics 40 (5), 49-72.
Germenji, E./I. Gedeshi (2008), Highly Skilled Migration from Albania: An Assessment of
Current Trends and the Ways Ahead. Development Research Centre on Migration,
Globalisation and Poverty Working Paper T-25.
Germenji, E./J. Swinnen (2005), Human Capital, Market Imperfections, Poverty, and Mi-
gration: Evidence from Rural Albania. LICOS Centre for Transition Economics Discus-
sion Paper 157/2005.
Government of Albania (2005), National Strategy on Migration and National Action Plan
on Migration. Tirana: Albanian Government in cooperation with the International Or-
ganization for Migration.
Kilic, T./G. Carletto/B. Davis/A. Zezza (2007), Investing Back Home: Return Migration
and Business Ownership in Albania. World Bank Policy Research Working Paper 4366.
King, R. (2005), Albania as a laboratory for the study of migration and development, in:
Journal of Southern Europe and the Balkans 7 (2), 133-56.
King, R. (2003), Across the sea and over the mountains: documenting Albanian migration,
in: Scottish Geographical Journal 119 (3), 283-309.
King, R./J. Vullnetari (2003), Migration and Development in Albania. Development Re-
search Centre on Migration, Globalisation and Poverty Working Paper C5.
Konica, N./R. Filer (2008), Albanian Emigration: Causes and Consequences, in: South-
Eastern Europe Journal of Economics (forthcoming).
Korovilas, J. (1999), The Albanian Economy in Transition: The Role of Remittances and
Pyramid Investment Schemes, in: Post-Communist Economies 11 (3), 399–415.
Kule, D./A. Mançellari/H. Papapanagos/S. Qirici/P. Sanfey (2002), The Causes and Conse-
quences of Albanian Emigration During Transition: Evidence from Micro Data, in: In-
ternational Migration Review 36 (1), 229-239.
Labrianidis, L./A. Lyberaki (2004), Back and Forth and in Between: Returning Albanian
Migrants from Greece and Italy, in: Journal of International Migration and Integration 5
(1), 77-106.
Labrianidis, L./B. Kazazi (2006), Albanian Return-migrants from Greece and Italy: Their
Impact upon Spatial Disparities within Albania, in: European Urban and Regional Stud-
ies 13 (1), 59-74.
León-Ledesma, M./M. Piracha (2004), International Migration and the Role of Remittances
in Eastern Europe, in: International Migration 42 (4), 65-83.
Mançellari, A./H. Papapanagos/P. Sanfey (1996), Job Creation and Temporary Emigration:
The Albanian Experience, in: Economics of Transition 4 (2), 471-491.
McCarthy, N./G. Carletto/B. Davis/I. Maltsoglou (2006), Assessing the Impact of Massive
Out-Migration on Agriculture. FAO, ESA Working Paper 06-14, Rome.
McKenzie, D. J./J. Mistiaen (2007), Surveying Migrant Households: A Comparison of
Census-Based, Snowball, and Intercept Point Surveys. World Bank Policy Research
Working Paper No. 4419.
Miluka, J./G. Carletto/B. Davis/A. Zezza (2007), The Vanishing Farms? The Impact of In-
ternational Migration on Albanian Family Farming. World Bank Policy Research Work-
ing Paper 4367.
Nicholson, B. (2004), Migrants as Agents of Development: Albanian Return Migrants and
Microenterprise, in: D. Pop (ed.), New Patterns of Labour Migration in Central and
Eastern Europe. Cluj Napoca, Romania, 94-110.
Nicholson, B. (2001), From migrant to micro-entrepreneur: do-it-yourself development in
Albania, in: South-East Europe Review 3, 39-42.
Stampini, M./C. Carletto/B. Davis (2008), International Migration from Albania. The Role
of Family Networks and Previous Experience, in: Eastern European Economics 46 (2),
50-87.
Stecklov, G./C. Carletto/C. Azzari/B. Davis (2008), Agency, Education, and Networks:
Gender and International Migration from Albania. World Bank Policy Research Work-
ing Paper 4507.
Uruçi, E./I. Gedeshi (2003), Remittances Management in Albania. CeSPI Working Paper
5/2003, Rome.
Vullnetari, J. (2007), Albanian Migration and Development: State of the Art Review.
IMISCOE Working Paper 18.
Vullnetari, J./R. King (2008), ‘Does your granny eat grass?’ On mass migration, care drain
and the fate of older people in rural Albania, in: Global Networks 8 (2), 139-171.
World Bank (2007), Albania – Urban Growth, Migration and Poverty Reduction: A Pov-
erty Assessment. Washington DC: The International Bank for Reconstruction and De-
velopment/World Bank. Report No. 40071-AL.
World Bank (2008), Migration and Remittances Factbook 2008. Washington DC: The In-
ternational Bank for Reconstruction and Development/World Bank.
Editors and Authors
Martin Kahanec
Martin Kahanec is Deputy Director of Research at the Institute for the Study of
Labor (IZA), Bonn. Furthermore he serves as the institute’s Deputy Program Di-
rector of the research area “Migration”, and as project leader of the sub-area “EU
Enlargement and the Labor Markets”. He has held several advisory positions and
participated in a number of scientific and policy projects with the World Bank, the
European Commission, OECD, and other international and national institutions,
including coordination of the “Study on the Social and Labour Market Integration
of Ethnic Minorities” and “Study on Active Inclusion of Migrants” funded by the
European Commission. He obtained his doctoral degree in Economics at the Cen-
tre for Economic Research (CentER), Tilburg University, the Netherlands. Ka-
hanec has published in refereed journals, co-authored a chapter in the Oxford
Handbook of Economic Inequality, and edited scientific volumes and a journal
special issue. His main research expertise is in labor and population economics,
ethnicity, and migration.
Klaus F. Zimmermann
Since 1998 Klaus F. Zimmermann has been Director of the Institute for the Study
of Labor (IZA) and Full Professor of Economics at the University of Bonn. More-
over, he serves as President of the German Institute for Economic Research (DIW
Berlin, since 2000), Honorary Professor of Economics at the Free University of
Berlin (since 2001) and Honorary Professor at the Renmin University of Peking
(since 2006). He engages in a variety of policy advice activities, e.g. as advisor to
the President of the EU Commission (2001-2003 and since 2005), as economic
advisor to the Prime Minister of North Rhine-Westphalia (since 2008), and as
member of the World Economic Forum's Global Agenda Council on Migration.
Since 1988 Klaus F. Zimmermann has been Editor-in-Chief of the Journal of
Population Economics. He serves as Associate Editor for various scientific jour-
nals and is author or editor of 33 books and over 190 papers in refereed journals
and collected volumes. His special research interests center on labor economics,
population economics, migration, industrial organization, and econometrics.
335
336 Editors and Authors
Timo Baas
Timo Baas is a Researcher at the Institute for Employment Research (IAB) in

Nuremberg. He holds a doctoral scholarship at the University of Potsdam and has
been visiting fellow at the German Institute for Economic Research (DIW), Berlin
from 2004 to 2006 and, from 2004 onwards, a lecturer at the Free University of
Berlin. His research centers on international migration, European integration, and
the construction of macroeconomic models.
Alan Barrett
Alan Barrett is Research Professor at the Economic and Social Research Institute
(ESRI), Dublin. Between 2001 and 2003 he worked for the Irish Finance Ministry.
He has been an IZA Research Fellow since 1998. His main research interest is
migration, and he has published on the topic in journals such as the Journal of
Population Economics, Labour Economics and the Oxford Review of Economic
Policy. Barrett also writes on developments in and prospects for the Irish econ-
omy, through his co-authorship of the ESRI's Quarterly Economic Commentary.
David G. Blanchflower
David Blanchflower holds a chair as Professor of Economics at Dartmouth Col-

lege, acts as Program Director of the “Future of Labor” research program at IZA,
and is a Research Associate at the NBER. He was an External Member of the
Monetary Policy Committee (MPC) at the Bank of England from 2006-2009.
Blanchflower has published in numerous refereed journals and has been a member
of the Editorial Board of Small Business Economics, of the Editorial Board of
Scottish Journal of Political Economy and of the Editorial Board of Industrial &
Labor Relations Review.
Karl Brenke
Karl Brenke has been a Research Associate at the German Institute for Economic
Research (DIW Berlin) since 1985. From 1983 to 1985 he worked as a Research
Associate at the Research Institute “Social Economy of Labor” at the Free Univer-
sity of Berlin. He has studied sociology, economics, statistics and philosophy and
holds a M. phil. in sociology. Among his broad research interests are demographic
change and migration.
Editors and Authors 337
Herbert Brücker
Herbert Brücker is Professor of Economics at the University of Bamberg and head

of the Department for International Comparisons and European Integration at the
Institute for Employment Research (IAB) in Nuremberg. Previously, he was a Re-
search Associate at the German Institute for Economic Research (DIW Berlin). In
2000 he joined IZA as a Research Fellow. He has received his habilitation in eco-
nomics from the Technical University of Berlin in 2005 and his Ph.D. in Econom-
ics from the University of Frankfurt a.M. in 2004. His research interests focus on
international migration, labor markets and international economics.
Marcel Canoy
Marcel Canoy is Professor in the Economics of Healthcare at the University of

Tilburg and at the same time chief economist of ECORYS, an international com-
pany specializing in economic, spatial and social development. He has held sev-
eral post-doctoral positions at the universities of Paris, Maastricht and Leuven,
and worked for the Netherlands Bureau for Economic Policy Analysis (CPB) from
1996 to 2005. From 2005 to 2009 he worked as economic adviser to the Bureau of
European Policy Advisers (BEPA) of the European Commission in Brussels,
where he advised inter alia on the internal market, the future of the welfare state
and migration.
Zvezda Dermendzhieva
Zvezda Dermendzhieva is a Ph.D. candidate in Economics at the Center for Eco-

nomic Research and Graduate Education (CERGE-EI) in Prague and a Junior Re-
searcher at the Economics Institute of the Czech Academy of Sciences. She has
lectured Macroeconomics, Microeconomics, and Labor Economics at the Prague
School of Economics and held a research intern position at the United Nations
University in Helsinki, Finland.
Randall K. Filer
Randall K. Filer is Professor of Economics at Hunter College and the Graduate

Center of the City University of New York. In addition he acts as a Visiting Pro-
fessor of Economics and Senior Scholar at the Center for Economic Research and
Graduate Education (CERGE-EI) in Prague. Furthermore he is President of the
CERGE-EI Foundation, the Eastern European Coordinator of the Global Devel-
opment Network, Chair of the International Faculty Committee at the Interna-
tional School of Economics in Tbilisi, Georgia (ISET), and a Research Fellow at
IZA. Filer’s research findings have been published in leading professional jour-
nals including American Economic Review, Journal of Political Economy, Re-
view of Economics and Statistics, and European Economic Review. His areas of
expertise include financial and capital markets, labor markets, urban economics,
demography and development economics, including the economic transition in the
post-communist countries of Central and Eastern Europe.
Christer Gerdes
Christer Gerdes is Assistant Professor at the Swedish Institute for Social Research
(SOFI) in Stockholm and is affiliated to the Stockholm University Linnaeus Cen-
ter for Integration Studies (SULCIS). Her research interests are focused on various
aspects of immigration, mostly by doing empirical research on immigration into
Nordic countries. In autumn 2008 she completed her Ph.D. thesis on “Studying
the Interplay of Immigration and Welfare States”.
Andreas Hauptmann
Andreas Hauptmann has been a Researcher at the Institute for Employment Re-
search in Nuremberg (IAB) since 2008 in the area “International Comparisons and
European Integration.” He received his diploma in economics at the University of
Wuerzburg and started his Ph.D. in economics at the University of Glasgow in
September 2008. His research focuses on migration, macroeconomic modeling
and labor market structures.
Mihails Hazans
Mihails Hazans is Professor of Econometrics at the University of Latvia and a Re-

search Associate at the Baltic International Centre for Economic Policy Studies
(BICEPS). He has served as an expert or consultant for OECD, World Bank, ILO,
European Commission and European Training Foundation. Focusing on the Baltic
and Balkan countries, he has published on migration, commuting, education, eco-
nomics of minorities, labor supply, determinants of earnings, job search, working
conditions, and policy evaluation.
Anna Horvath
Anna Horvath is a social science adviser at the Eurydice European Unit at the
Education, Audiovisual and Culture Executive Agency in Brussels, and a doctoral
candidate at the Department of Political Science of the Central European Univer-
sity in Budapest. Her research interests include social inclusion, immigration and
education policies in the European Union.
Agnès Hubert
Agnès Hubert is an adviser in the think tank of the European Commission. She
holds degrees in Economics and Political Science from the University of Paris 1-
Sorbonne. She has held positions of responsibilities in the European Institutions in
the last 20 years, in international commodity agreements; information and com-
munication and in gender issues. From 1992 to 1996, she directed the unit for
gender equality and has continued since to work on gender issues in the Forward
Studies Unit and when seconded for two years to the Women’s rights committee
of the European Parliament. She also contributed to the White paper on European
Governance (2001) as a member of the governance team of the Commission.
Among many other contributions, she authored books on “L'Europe et les
femmes, identités en mouvement”, and “Democracy and Information Society in
Europe”.
Paweł Kaczmarczyk
Paweł Kaczmarczyk is Vice Director of the Centre of Migration Research at the

University of Warsaw and Assistant Professor at the Faculty of Economic Sci-
ences, University of Warsaw. He is also a member of the Board of Strategic Ad-
visers to the Prime Minister of Poland. His main research areas include causes and
consequences of labor migration (with special attention paid to analysis of migra-
tion processes in central and east European countries), labor economics, interna-
tional economics and migration policy. Kaczmarczyk is author of many publica-
tions in the field.
Helen Lawton
Helen Lawton has been working as an Economist at the Bank of England since
2002. Together with David G. Blanchflower in his position as an External Mem-
ber of the Bank of England's Monetary Policy Committee she authored several
studies on the economics of migration. Helen Lawton studied economics at the
University of Cambridge.
Frédéric Lerais
Frédéric Lerais was a member of the Bureau of European Policy Advisors (BEPA)
from 2005 to 2009, and has dealt with various European social issues. He started
his career at the French Observatory of Economic Conjunctures (OFCE) in Paris.
Afterwards, he managed the mission for economic analysis of the Directorate for
animation of research, studies and statistics (DARES), at the French Ministry of
Labour. This mission carried out economic and empirical studies of the labor
market. Lerais is graduated from the École nationale de la statistique et de
l’administration économique (ENSAE) and from the University of Paris I (Pan-
theon Sorbonne).
Marta Mioduszewska
Marta Mioduszewska graduated from the Department of Economic Sciences, Uni-

versity of Warsaw in 2008. She is a Ph.D. candidate at the Department of Eco-
nomic Sciences and Research Assistant at the Centre of Migration Research
(CMR), University of Warsaw.
Kaia Philips
Kaia Philips is Associate Professor of Statistics and Vice-Dean at the University

of Tartu, Faculty of Economics and Business Administration. Her research include
human capital, workers’ flows and job flows, free movement of labor, industrial
relations and trade unions in transition countries. She has served as expert for
European Foundation, European Commission and other international organisa-
tions in numerous international research projects.
Myriam Sochacki
Myriam Sochacki has been a member of the European Commission’s Bureau of

European Policy Advisers (BEPA), Brussels since 2005. She obtained a Master’s
degree in French literature at Paris Sorbonne, followed by a post-graduation in
communication at CELSA/Sorbonne. She joined the French Navy in 1984 and
served as a public relations and communication expert, including from 1994 to
1996 as a spokesperson in Sarajevo. In 1996 she became Head of Press and In-
formation of the Western European Union, Brussels. From 2001 to 2004 she was
project manager of a program on security issues and the Euro-American relation-
ship developed with the support of the Luso-American Foundation, Lisbon.
Eskil Wadensjö
Eskil Wadensjö is Professor of Labor Economics at the Swedish Institute for So-
cial Research, and Dean of the Faculty of Social Sciences at Stockholm Univer-
sity. He was President of the European Association of Labour Economists (EALE)
from 1993 to 1999 and Chairman of the Swedish Economic Association from
1992 to 1993 and has been member of several governmental committees. He has
been IZA Research Fellow since 2000. His main research interests include the
economics of international migration, labor market policy and social security. He
is director of the Stockholm University Linnaeus Center for Integration Studies
(SULCIS), a large ten-year research program founded by the Swedish Research
Council.
Mutlu Yuksel
Mutlu Yuksel has been a Research Associate at the Institute for the Study of La-
bor (IZA), Bonn, since 2007, after completing his Ph.D. at the University of Hous-
ton, USA in May 2007. His research interests include Labor Economics, Devel-
opment Economics, and Migration. His current research focuses on
discrimination, assimilation, intergenerational mobility, and the effects of immi-
gration on natives.
Anzelika Zaiceva
Anzelika Zaiceva joined the Institute for the Study of Labor (IZA) as a Research
Associate in 2006. Since 2007 she is also a Research Fellow at the University of
Bologna, where she takes part in a research project on “The Political Economy of
Unemployment in Transition Economies”. She received her Ph.D. in Economics
in 2007 at the European University Institute in Florence. As a member of IZA’s
research programs on “Migration” and “Transition and Emerging Economies” her
research interests focuses on these topics but also includes labor and population
economics als well as gender research.
Anna Żylicz
Anna Żylicz is a Ph.D. student at the Faculty of Economic Sciences, University of

Warsaw, where she graduated in economics and mathematics. She works as a doc-
toral fellow and Research Assistant at the university’s Centre of Migration Re-
search (CMR). Her research interests include topics in the supply-side economics
of migration, especially microeconomic econometric modelling. Additional inter-
ests include demography and game theory.
Index
A 296, 306, 307, 311, 312, 327,

Assimilation, 52, 98, 125, 131–135, 329, 330
139–143 Gender, 75, 89, 92, 93, 116, 120, 140, 153,
Asylum, 74–76 155, 163, 172, 189, 290
B Guest Workers, 78, 118, 186
Brain Drain, 32, 36, 95, 231, 245, 246, H
255, 323, 324 Human Capital, 6, 8, 9, 21, 36, 37, 40, 41,
Brain Waste, 36, 54, 246, 255, 280 55, 62, 135–139, 153–155, 159,
Business Cycle, 219, 220, 239, 240, 244, 187, 191, 238, 246, 255, 257,
245, 250 266, 273, 282, 329
C I
Citizenship, 82, 115, 118, 124, 126, 165, Identity, 81, 89–92
166, 255, 257, 260, 276, 294 Integration, 41, 57, 58, 72–78, 81, 84, 89–
Complementarity, 7, 8, 48, 55, 57, 289 93, 101, 125, 139, 155, 240,
D 293, 329
Data J
Administrative, 150, 151, 158, 258, Job Satisfaction, 192–196
288, 290 L
Census, 21, 221, 242, 258, 273, 276, Labor
290 Demand, 55, 56, 156–159, 243
Survey, 19, 21, 150, 151, 242, 272, Shortages, 32, 33, 73, 95, 243–250,
297, 327 256, 287, 289, 300
Diversity, 73, 78, 321 Supply, 37, 47–60, 64–67, 145, 156,
Downgrading (of Qualifications), 21, 36 157, 286, 289, 290, 324, 326
E Labor Market
Economic/Financial Crisis, 3, 6, 38–41, Effects, 47, 60, 66, 230, 237
50–52, 67, 71, 78, 104, 159, Equilibrium, 220, 238
208, 209, 240, 250, 257, 298, Outcomes, 41, 112, 124, 126, 128, 153,
300, 312, 328 154, 158, 202, 257, 273,
Education Level, 20, 142, 268 282, 294
Employment Language, 8, 16, 22, 36, 39, 74, 81, 88, 96,
Effects, 50, 55 125, 141, 167, 210, 211, 264,
Rate, 20, 96, 118, 131, 142, 152, 168, 273, 279, 292, 295, 299, 307
169, 177, 273 Life Satisfaction, 193, 196–199
Entrepreneurship, 41, 190, 211, 250 M
Ethnic Minorities, 256, 294, 289 Migrant
European Commission, 5, 9, 10, 17–22, Characteristics, 230, 258
31, 36, 73, 79, 83, 92, 94, 97, Networks, 5, 228
100, 101, 206, 258–264, 269, Selectivity Index, 233, 237
273, 274, 279–282 Migration
F Circular, 9, 41, 186, 221, 321
Family Reunification, 321, 328 Decision, 320, 321, 327
Free Mobility of Workers, 48, 50, 114 Determinants of, 19, 255, 257
G Flows, 6, 9, 13, 17, 19, 22, 26, 40, 41,
GDP, 18, 19, 31, 36, 47, 50, 58–60, 63, 66, 48, 50, 51, 54, 66, 68, 102,
68, 177, 208, 238, 256, 289, 112, 117, 133, 166, 186,
343
344 Index
211, 212, 240, 242, 257– T

266, 287, 289, 290 Temporary Workers, 16, 186, 255, 282,
Illegal, 9, 74, 78 292
Impact of, 1, 49, 54, 58, 60, 62, 66, Transition, 32, 89, 91, 221, 226, 242, 243,
109, 156, 217, 240, 248, 248, 249, 284, 306–309, 313,
257, 324 327, 330
Intentions, 9, 22, 112, 167, 257, 259, Transitional Arrangements, 4, 5, 14, 22,
294 26, 39, 48, 54
Networks, 237, 297, 320, 321 U
Permanent, 257, 269, 323, 326 Unemployment, 6, 8, 18–23, 26, 27,
Policy, 5, 22, 242 30–33, 38, 42, 44, 47–51, 55,
Reasons for, 22 58–70, 79, 94, 95, 99, 117, 123,
Return, 10, 40, 67, 68, 159, 165, 186, 131, 136, 142, 146, 147, 152,
187, 226, 251, 257, 156–161, 167, 168, 186, 192,
282–284, 290, 292, 295, 193, 199–214, 219, 238–243,
298, 300, 309, 310, 320, 250, 252, 256, 261, 264, 266,
327–329 273, 284–290, 298, 301, 306,
Scenarios, 50, 53, 243, 248 307, 323
Seasonal, 19, 256 "Fear" of, 30, 203, 206–208
Stocks, 51, 249 Rate, 19, 26, 30, 33, 47, 55, 60–63,
Temporary, 259, 321, 322, 327 66–68, 117, 152, 167, 206,
Mobility 238, 243, 264, 266, 273,
Labor, 5, 47–49, 53, 68, 94, 260, 331 287–290, 306
Occupational, 279, 280
V
N Vacancy Rate, 26, 28, 247, 287, 289
Nationality, 82, 96, 114–119, 124, 126,
134, 142, 149, 150, 155, 185, W
210 Wage
Curve, 49, 55, 290
P Equation, 174–176, 248
Participation Rate, 20, 21, 54, 63, 117, Rigidities, 49, 55
266, 287, 289 Work Permit, 113, 114, 260, 289
Posted Workers, 259, 262–267, 295 Working Hours, 123, 169, 170, 177, 178,
Public Perceptions, 41, 71–78, 83– 86, 188, 191, 193, 201, 286
95–101, 104
Pull Factors, 39, 250, 306, 307
Push Factors, 202, 236, 239, 306
R
Remittances, 36, 37, 40, 73, 256, 257, 289,
295–297, 306, 307, 311, 312,
322–331
S
Security, 14, 33, 74, 75, 78–81, 88, 89, 98,
102, 103, 118, 142, 143, 150,
158, 195, 196, 260, 265, 287,
289, 300
Self-employment, 20, 27, 124, 125,
188–192, 281
Simulations, 48, 50, 53, 54, 56, 58, 61,
63–67, 151, 157, 159, 186
Social Disorganization Theory, 77

(Martin Kahanec, Klaus F. Zimmermann) EU Labor PDF

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

(Martin Kahanec, Klaus F. Zimmermann) EU Labor PDF

Uploaded by

Copyright:

Available Formats

EU Labor Markets After

EU Labor Markets After

ISBN 978-3-642-02241-8 e-ISBN 978-3-642-02242-5

Library of Congress Control Number: 2009937939

© Springer-Verlag Berlin Heidelberg 2010

Typesetting: Thomas Wasilewski, IZA, Bonn

Cover design: WMXDesign GmbH, Heidelberg, Germany

Printed on acid-free paper

Springer is part of Springer Science+Business Media (www.springer.com)

This volume results from a continuing dedication of IZA’s Migration Program

The migration research group at IZA produces numerous discussion papers,

Bonn, May 2009

Martin Kahanec Klaus F. Zimmermann

1 Lessons from Migration after EU Enlargement ...................................3

2 Labor Mobility in the Enlarged EU:

3 Post-Enlargement Migration and Public Perception

4 EU Enlargement under Continued Mobility Restrictions:

5 The Experience of Spain with the Inflows

6 EU Enlargement and Ireland’s Labor Market .................................145

8 The Impact of the Recent Expansion of the EU

9 Impact of the Post-Accession Migration

10 The Post-Enlargement Migration Experience

11 The Case of Albania ............................................................................305

Editors and Authors ..........................................................................................335

The Eastern enlargement of the EU was an institutional impetus to the migration

M. Kahanec and K.F. Zimmermann (eds.), EU Labor Markets After Post-Enlargement 3

Early theories of migration stress the significance of (expected) regional dispari-

By a similar argument, low-skilled immigration decreases the wages to wl2

This simple analysis has a straightforward message: skilled immigration bene-

a. Actual Migration Trends and Trajectories

To provide an updated picture of post-enlargement migration from the new mem-

Although the statistics on stocks of residents usually rely on registers of foreign

Source: Brücker and Damelang (2009), Tables 3b and 4b.

EU15 EU10 EU2 Non-EU27

Turning now to the available data on actual migration flows, it is important to

Total foreigners EU12

Total foreigners EU12

Total foreigners EU12

Total foreigners EU12

UK Worker Registration Scheme applications, May 2004-December 2008

UK National Insurance numbers (in thousands), January 2002-September 2008

Irish personal public service numbers, 2001-2008

Finally, valuable evidence is based on country-level studies. Gilpin et al. (2006)

b. The Source Country Perspective

To complete the picture of post-enlargement migration flows, we present available

Source: Brücker and Damelang (2009), Table 6b.

A similar upward trend in emigration is apparent from Polish Labour Force

Regarding potential macro-economic determinants of migration, Figure 7 is

c. Who Migrated, and How Do They Fare?

According to the European Commission (2008a, b), the post-enlargement mi-

It is also important to note that a significant proportion of these migrants is

Building on Brenke and Zimmermann (2007), Brenke et al. (chapter 4 in this

yes, you have

yes, you have

Source: Own tabulations from the Eurobarometer EB 65.1.

1996 1997 1998 1999

4 The Impact of Post-Enlargement Migration Flows:

a. The Destination Countries

18 Note, however, that these are nominal values.

EU15 Germany Ireland Sw eden UK

EU15 Germany Sw eden UK

While causal interpretations of these findings would be misleading, the stabil-

19 See also Doyle et al. (2006) and Wadensjö (2007).