Available via license: CC BY-NC-ND
Content may be subject to copyright.
13
Sanja Franc, Anita Čeh Časni, Antea Barišić
ABSTRACT
The Eastern enlargements of the European Union (EU) since the early 2000s have included post-transitional
economies at a lower level of development than the existing member states and thus, have signicantly af-
fected the East-West migration ows and labour markets on both sides. This has provided a distinctive op-
portunity to study the eects of liberalisation and to identify economic factors leading to migration ows
with the purpose of enabling better estimations of future migration trends. In this research, a panel data
analysis with pair of country xed eects and time xed eects is used to explore several pull and push factors
of the East-West EU migration ows in the period from 2000 to 2017. Results indicate that emigration rate
responds rather quickly to the changes in GDP per capita and unemployment rate of the youth population in
immigration country, with statistically signicant elasticity coecients, suggesting that international migra-
tion contributes signicantly to adjusting the labour supply to uctuations in economic activity.
Key words: migration determinants, youth unemployment, EU enlargement, panel data analysis.
JEL: F22, C01
1. INTRODUCTION
Sanja Franc, PhD
Assistant Professor
Faculty of Economics and Business
University of Zagreb
E-mail: sfranc@efzg.hr
Anita Čeh Časni, PhD
Assistant Professor
Faculty of Economics and Business
University of Zagreb
E-mail: aceh@efzg.hr
Antea Barišić, MA
Teaching and Research Assistant
Faculty of Economics and Business
University of Zagreb
E-mail: abarisic@efzg.hr
South East European Journal of Economics and Business
Volume 14 (2) 2019, 13-22
DOI: 10.2478/jeb-2019-0010
Copyright © 2019 by the School of Economics and Business Sarajevo
DETERMINANTS OF MIGRATION FOLLOWING
THE EU ENLARGEMENT: A PANEL DATA ANALYSIS
Migration has a signicant role in the global inte-
gration process, together with international trade and
foreign direct investment, but unlike the latter, migra-
tion still hasn’t experienced liberalisation on a global
level during the last few decades. Migration motives
are generally divided into three groups: economic, po-
litical and socio-cultural, where dierent ‘push’ factors
in the country of emigration and ‘pull’ factors in the
country of immigration can be recognised. The most
common push factors are poverty, unemployment,
low wages, high fertility rates, lack of basic health and
education. On the other hand, most common pull fac-
tors in the country of immigration are prospects of
higher wages, opportunities for an improved standard
of living and personal or professional development
(Fan and Stark 2007). The simplest economic models
of migration explain that motivation for migration
Unauthentifiziert | Heruntergeladen 01.01.20 12:47 UTC
DETERMINANTS OF MIGRATION FOLLOWING THE EU ENLARGEMENT: A PANEL DATA ANALYSIS
14 South East European Journal of Economics and Business, Volume 14 (2) 2019
comes from real wage dierentials across countries
that emerge from various degrees of labour market
rigidity, while there are also models showing that mi-
gration is driven by expected rather than real wage
dierentials (Mansoor and Quillin 2006). Depending
on the sample of countries, time period and used
methods, studies point out to dierent importance of
push and pull factors.
After 1989, Europe began to face signicant intra-
regional migratory movements. Freedom of move-
ment obtained by the Central and Eastern European
citizens has generated large migratory movements
from the East to the West (Panzaru 2013). Atoyan et al.
(2016) estimated that about 5.5% of the population
of Southeast and Central Europe have left this region
in the period from 1988 to 2012. Enlargements of the
European Union since the early 2000s, which have
included emerging economies, have considerably af-
fected the state of the labour markets and stock of mi-
grants in the EU. Exploring the direction, intensity and
determinants of those migration ows is necessary to
understand the potential eects on both the origin
and the destination country. Thus, the East-West EU
migration has become a signicant topic in migration
studies (Favell 2008).
Previous research has shown that the main migra-
tion determinants in the EU were more connected to
the labour market (Kahanec and Zimmerman 2010)
rather than social benets (Giulietti et al. 2011), thus
conrming the neoclassical theory of migration.
Kahanec, Pytlikova, and Zimmermann (2014) also em-
phasise the eects of business cycles in destination
countries on migration ows.
This study contributes to the literature by identi-
fying the importance of selected economic determi-
nants that lead to migration ows in the contempo-
rary context, given the fact that scholars emphasise
dierences among migration determinants in dier-
ent periods and settings. It contributes especially to
the research of migration eects deriving from the
latest EU enlargements where new member states
signicantly dier in economic and other properties
to the old ones. Given the range of economic cycle
periods and including all new member states, this
research can enable better estimations of the future
migration ows and the creation of policies that could
aect them. It also emphasises youth unemployment
(young population between 15 and 24 years old) in
both origin and destination countries as an important
migration determinant. Young people are seen as driv-
ers of change in their societies and emigration of this
group can particularly aect the emigration countries.
At the same time, there is a lack of empirical evidence
on the importance of their unemployment ratios to
migration decision.
The main objective of this paper is to empirically
examine several ‘push’ and ‘pull’ factors to determine
their eects on migration ows from the ‘new’ EU
member states to the ‘old’ EU member states that dif-
fer signicantly in economic and other characteristics,
which is precisely the aspect that makes these Eastern
enlargements dierent from the previous ones.
Unemployment is a persistent issue in most of the
new EU member states, but young people are aected
particularly hard as the EU youth unemployment rate
is more than double the overall unemployment rate
and diers considerably among countries. Thus, be-
sides examining GDP per capita, unemployment rate,
and EU membership as determinants of migration, we
also analyse the youth unemployment rate as a deter-
mining factor of migration ows. While some of the
previous studies (Mayda 2005; Kim and Cohen 2010)
have included the share of young people in the total
population as a migration determinant, the additional
contribution of this research is in examining the sig-
nicance of youth unemployment rate in determining
the emigration rate.
The research is divided into ve parts, as follows.
Literature review on migration determinants in sec-
tion 2 points out to dierent push and pull factors,
each ranging from economic to social and political
ones. Examining the signicance of selected econom-
ic push and pull factors within the two groups of the
EU member states was the primary motivation for this
paper. Section 3 includes the empirical analysis which
was conducted using a panel data analysis with pair
of country xed eects and time xed eect including
several pull and push factors of migration ows. More
precisely, data on yearly immigrant inows into 15 de-
veloped European countries by country of origin was
used to empirically test which determinants of migra-
tion ows aect emigration rate the most. The analysis
included the period from 2000 to 2017, thus encom-
passing the eects of the nancial crisis and the post-
crisis period. Fixed eect panel model is employed
with country and time specic eects to avoid biased
estimates. Also, regressions have robust standard er-
rors clustered by country pair (destination and origin
country), to address heteroscedasticity and allow for
correlation over time of country pair observations.
Section 4 presents results of the empirical analysis
which indicate that emigration rate responds rather
quickly to changes in GDP per capita and unemploy-
ment rate of young population in immigration coun-
try, with statistically signicant elasticity coecients,
suggesting that international migration contributes
signicantly to adjusting the labour supply to uc-
tuations in economic activity. The conclusions of the
Unauthentifiziert | Heruntergeladen 01.01.20 12:47 UTC
DETERMINANTS OF MIGRATION FOLLOWING THE EU ENLARGEMENT: A PANEL DATA ANALYSIS
15South East European Journal of Economics and Business, Volume 14 (2) 2019
analysis in section 5 lead to a better understanding
of the EU migration ows determinants, as an essen-
tial prerequisite for estimating future migration ows.
Emigration rate responds rather quickly to changes in
GDP per capita and unemployment rate of the youth
population in immigration country, with statistically
signicant elasticity coecients. Also, EU membership
as a dummy variable has shown to aect migration.
2. LITERATURE REVIEW
International migration represents an integral part
of global ows. Dierent determinants are leading to
the decision on migration, reaching from econom-
ic, social, cultural, political to ecological and other,
but one of the main motives is often the aspiration
of migrants to improve their livelihood (Hatton and
Williamson 2005). Even in the early works in this area
(for example, Lee 1966), special attention was given
to dierences in the level of development between
areas, dierences in population characteristics and
the ease of overcoming migration obstacles. Thus, we
can distinguish between ‘push’ factors in the country
of emigration and ‘pull’ factors or desirable factors in
the country of immigration (Dorigo and Tobler 1983).
Besides the already mentioned ones, these factors
may include dierences in wages in certain sectors,
unemployment rates, opportunities for personal and
professional advancement, better living conditions,
freedom, climate conditions and other factors that
push migrants from the country of emigration and
pull them to destination countries (Jurčić and Barišić
2018).
Migration determinants are recognized as an im-
portant research topic, but studies reveal dierent
results depending on the time-frame, sample of coun-
tries and methods applied in the research. Various
studies point out a set of economic and demograph-
ic factors as being the most important ones while
explaining the migration process among dierent
samples of developed and less developed countries
in dierent parts of the world. Also, migration costs
deriving from the geographical distance and cultural
dierences are shown to be important in determining
migration ows.
Mayda (2005) used annual panel data set on the
sample of the OECD countries from 1980 to 1995,
while focusing on both supply and demand deter-
minants of migration patterns, and found results
broadly consistent with the theoretical predictions of
the standard international migration model. Namely,
the results have shown that pull factors, which in-
clude improvements in the income opportunities in
the destination country, signicantly increase the im-
migration rates. Oh and Jung (2013) while investigat-
ing migration ows in South Korea in the period from
1993 to 2011, suggest that economic development
accelerates emigration ows as it decreases nan-
cial restrictions to migration. They also revealed that
volume of trade, as evidence of an economic link be-
tween countries, is an important predictor of the size
and composition of foreign migrant population, while
speculating that this inuence is due to information
eects and foreign labour policy channel.
Besides GDP per capita and real wage per hour,
as important economic determinants, Sulaimanova
and Bostan (2014) pointed to depreciation of local
currencies and labour force growth in Tajikistan and
Kyrgizstan, as countries of origin, in determining their
emigration to Russian Federation in the period from
1998 to 2011. Also, remittances that are usually inves-
tigated in the studies of migration eects have shown
to be an important incentive that encourages fur-
ther emigration in some cases (Wickramasinghe and
Wimalaratana 2016; Sulaimanova and Bostan 2014).
Furthermore, social remittances are shown to enable
mobility through sharing ideas, practices and narra-
tives (Levitt and Lamba-Nieves 2011) that can be ad-
dressed through micro-level studies.
Economic theory also emphasises the importance
of personal taxation on migration, especially among
groups of high-income workers and professionals, but
empirical studies covering this topic are very limited.
Challenges in measuring this relationship are mostly
regarded to limited data availability. Thus, most of the
existing studies are done at the micro-level, including
only specic groups (usually those with the highest in-
come) or only several developed countries that record
in-detail administrative data (Kleven et al. 2019).
Several studies also point out to negative eects of
migration costs. Migration cost is usually measured as
the distance between capital cities of origin and des-
tination country and is reported to be an important
determinant of migration ows (Mayda 2005; Mayda
2010; Kim and Cohen 2010).
Demographic factors are considered to be closely
related to economic ones, and some of the demo-
graphic factors are even used as proxies for economic
or living conditions (Kim and Cohen 2010). Studies in-
cluding dierent country samples call attention to de-
mographic characteristics as important in determin-
ing the migration ows. Oh and Jung (2013) suggest
that demographic factors, especially ageing popula-
tion of developed countries, have an impact on mi-
gration ows, while Kim and Cohen (2010) point out
demographics (log population of origin and destina-
tion and log infant mortality rate (IMR) of origin and
Unauthentifiziert | Heruntergeladen 01.01.20 12:47 UTC
DETERMINANTS OF MIGRATION FOLLOWING THE EU ENLARGEMENT: A PANEL DATA ANALYSIS
16 South East European Journal of Economics and Business, Volume 14 (2) 2019
destination) as one of the most signicant variables in
aecting the ows among selected developed coun-
tries between 1950 and 2007. Oh and Jung (2013) also
reveal dierent patterns for skilled and unskilled mi-
grant workers.
Mayda (2010) found that the share of the young
population in the origin country had a positive and
signicant impact on emigration rates in case of the
OECD countries in the period from 1980 to 1995.
Focusing on developed countries in the period from
1950 to 2007, Kim and Cohen (2010) found that in-
creased youth population share in the destination
country was associated with lower inows, while the
increasing youth population share in the origin coun-
try was associated with higher inows. Efendic (2015)
also showed the importance of being young as one of
the most signicant individual characteristics in deter-
mining future emigration in Bosnia and Hercegovina
in the period from 2002 to 2010, which is in line with
the prior literature.
While neoclassical migration theory emphasises
economic determinants and excludes social and po-
litical dimensions (Wickramasinghe and Wimalaratana
2016), other studies show that wage and employment
dierentials were statistically signicant predictors of
migration in the expected directions only about half
the time (Mansoor and Quillin 2006), which means
that other non-economic factors are also important
in explaining migration. Migration ows change with
the altering socioeconomic and geopolitical condi-
tions (Wickramasinghe and Wimalaratana 2016) and
the most important non-economic determinants
are social and political factors shaping the migration
process.
A broad stream of literature emphasises that social
relationships have a signicant eect on migration. At
the centre of sociological research are migration net-
works which can be dened as ‘set of interpersonal
ties that connect migrants, former migrants and non-
migrants in origin and destination areas through ties
of kinship, friendship and shared community origin’
(Massey et al. 1993, p. 448). Migrant networks have
been emphasised as an important factor in labour mi-
gration in both developed and developing countries
as they reduce migration costs (Zhao 2003). They are
thought to drive continuous migration ows, not de-
pending on economic and other factors that might
have caused the initial ows (Liu 2013; Garip and Asad
2015). Also, there is a potential brain gain through
these networks, especially in cases of forming expa-
triate knowledge networks (Meyer 2001). However,
migrant networks are usually researched using sur-
vey data that are not very common among countries.
They do not allow for a broader study as there is no
universal framework of collecting these data (Zhao
2003) and even if collected, only a limited amount
of data is available (Haug 2008). In-depth qualitative
studies on smaller samples are made to examine it,
but there is a need for structuring the process and in-
terviews in both origin and destination countries to
analyse these eects (Haug 2008). Morover, there are
studies which claim that migration network theory is
not able to explain large scale international migration
as it ignores a variety of factors leading to migration
while also focusing mostly on the supply side and ig-
noring the demand-side factors (Krissman 2005). Also,
network theory has shown not to be equally bene-
cial in all settings nor across all social groups (Garip
and Asad 2015), and it might be losing its importance
with the development of technology that leads to
more accessible information than in previous periods
(Wickramasinghe and Wimalaratana 2016).
Various studies examined the relationship between
dierent political factors and migration. Having in
mind East European countries, particularly interesting
might be the study made on the sample of respond-
ents in Bosnia and Hercegovina that has shown the
political situation as well as conict and post-conict
experiences as more important factors determining
emigration, even more signicant than the economic
ones (Efendic 2015). Evidence from Kosovo show that
political factors were also important in the case of re-
turning migration (Kotorri 2017). Ravlik (2014) nd-
ings upon analysing data containing 212 origin and
167 destination countries suggest that migrants are
more attracted by countries with common colonial
history and also those that show higher Rule of Law
index, as well as Human Development index.
Opening of Eastern European economies and the
Eastern enlargement has made a signicant impact
on migration ows from the ‘new’ to the ‘old’, more
developed member states. Several studies (Fouarge
and Ester 2007; Zaiceva and Zimmermann 2008) con-
rmed that the proportion of individuals intending
to emigrate after the 2004 enlargement was more
signicant in the new member states than in the old
member states, indicating to the relevance of the inte-
gration enlargement. It is estimated that in the period
from 1988 to 2012, about 5.5% of the population of
Southeast and Central Europe left this region (Atoyan
et al. 2016).
In the case of East-West EU migration, ethnical
similarity and cultural (as well as geographical) prox-
imity makes the migrants from Eastern Europe more
desirable in the western countries (Favell 2008). Wage
dierences of workers with almost the same quali-
cations in dierent countries are signicantly higher
than dierences in product prices and the cost of
Unauthentifiziert | Heruntergeladen 01.01.20 12:47 UTC
DETERMINANTS OF MIGRATION FOLLOWING THE EU ENLARGEMENT: A PANEL DATA ANALYSIS
17South East European Journal of Economics and Business, Volume 14 (2) 2019
capital, which is partly due to the smaller volume
of the international labour movement as opposed
to capital and product movement (Freeman 2006).
Research have shown that in the past decade migra-
tions in Europe were connected to the labour market
conditions (Kahanec and Zimmerman 2010) and that
the primary migration determinant were not social
benets (Giulietti et al. 2011), which conrms the neo-
classical predictions according to which the dierence
in wages and employment, that is, demand and job
oer, along with employment conditions, are key de-
terminants in making individual migration decisions
(Massey et al. 1993).
Kahanec, Pytlikova, and Zimmermann (2014) esti-
mated the eects of the EU accession and economic
opportunities on migration based on data incorporat-
ing immigration ows and foreigner stocks collected
by all countries worldwide for 42 destination countries
in the period 1980-2010. Applying the dierence-in-
dierences and triple dierences estimator, they sub-
sequently nd that East-West migration ows in the
EU responded positively to the EU entry and econom-
ic opportunities in receiving labour markets. However,
the authors mainly focus on pull factors such as dis-
tance, opening of the labour market and GDP, and do
not take into account other economic, social or demo-
graphic factors nor do they capture the period after
the crisis.
While analysing determinants and shaping fac-
tors of labour emigration within the European Union,
Son and Noja (2012) developed double-log econo-
metric models that combine cross-section and time-
series in a panel structure by using a set of indicators
specic for the emigration process, as well as for the
economic activity, labour market, and education, as
main explanatory variables. The results of their study
show that high unemployment reduces the emi-
grant stock, mainly due to the loss of associated in-
come and to the reduction of the migrants’ capacity
to move and integrate into another country. At the
same time, a positive selection of emigrants at desti-
nation according to their educational level was iden-
tied, while an increase in education in the source
country downsizes the stock of emigrants mainly due
to an improvement in employment perspectives. On
the other side, Ganguli (2018) with the micro-level
analysis using RoyModel framework for exploring se-
lection of migrants from Russia, Ukraine and Bulgaria
to the USA, Spain and Greece, pointed out mostly
positive selection in communist and post-communist
periods among East European immigrants in the US,
while negative selection of these immigrants in the
European Union. These dierences might be primarily
due to the set of countries included in these studies.
Panzaru (2013) analysed several alternative eco-
nomic factors such as doing business index and the
labour market regulation index, as well as indicators
that reect a certain level of freedom and democracy,
such as indicators that characterise judicial independ-
ence and legal system, but such factors have shown a
limited inuence on migration in Central and Eastern
Europe from 2000 to 2010. This might be due to po-
tentially non-permanent plans of residing in destina-
tion countries but only reaching them to achieve a
higher personal wealth in short or medium term or to
shortage of this kind of detailed information on desti-
nation countries.
Some micro-level studies reveal unmet high ex-
pectations of migrants from Eastern Europe to more
developed Western countries such as Lithuanian-
Iceland study of temporary migration (Minelgaite,
Christiansen, and Kristjánsdóttir 2019), is what can
aect further developments of migration through
return ows even in cases where migrants were plan-
ning more temporary migration.
It is often debated that push factors such as unem-
ployment or low wages aect young people the most.
However, many young people also choose, or are
forced to migrate to escape poverty, violence, conict,
or are displaced due to the eects of war or climate
change. As such, young people are heavily represent-
ed in migration for humanitarian reasons, including
refugees, asylum-seekers and as unaccompanied mi-
nors (United Nations 2016), but these reasons are not
seen as important in case of the intra-EU migration as
in some of the other parts of the world. Nonetheless,
there is an overall shortage of empirical research on
youth migration determinants.
Young people are a social group that can be par-
ticularly aected by dierent push and pull factors.
Recent recession and higher unemployment have
shown to cause increasing depression, poor health,
higher criminality and suicide rates among the young
generation, while at the same time, young and ambi-
tious people are looking for opportunities in foreign
countries, creating families there, and thereby, aect-
ing demographic prospects of their home countries
(Rakauskienė and Ranceva 2014). In Europe, Baltic and
the Mediterranean countries can be denoted as af-
fected the most because young people are emigrating
immediately after graduation, which means a loss not
only of investment in their education, but also a det-
riment for the future competitiveness of the country
(Rakauskienė and Ranceva 2014).
Van Mol (2016) investigated the inuence of mi-
cro-and macro-level characteristics on migration
aspirations of young people across the EU member
states. The results reveal the importance of individual
Unauthentifiziert | Heruntergeladen 01.01.20 12:47 UTC
DETERMINANTS OF MIGRATION FOLLOWING THE EU ENLARGEMENT: A PANEL DATA ANALYSIS
18 South East European Journal of Economics and Business, Volume 14 (2) 2019
characteristics and feelings of discontent with the
current climate in explaining emigration aspirations.
Furthermore, the author detected a negative relation-
ship between relative welfare levels with emigration
aspirations and a positive relationship with the youth
unemployment ratio. Together, the results suggest
that potential young intra-EU movers are positively se-
lected from the population.
As it can be seen from the presented literature
review, there is no consensus on the modelling ap-
proach or the variable selection when studying mi-
gration determinants. Moreover, migration studies
reveal dierent results depending on the time-frame,
sample of countries or methods employed. This paper
explores several economic determinants of the con-
temporary migration ows between new EU member
states and older EU member states, covering a period
of 18 years, from 2000 to 2017, thus capturing the pe-
riod prior to and post-global nancial crisis. Therefore,
the contribution of this paper to the existing empiri-
cal literature on determinants of migration is twofold.
Namely, the research points out youth unemployment
(young population between 15 and 24 years old) as a
determinant of migration ows, since young people
are a social group that can be particularly aected by
dierent push and pull factors. Also, in the empirical
analysis, we use the traditional panel data estimator
with time-and-entity-xed eects to explore the rela-
tionship between predictor and outcome variables
3. DATA AND METHODS
In this research data form the Eurostat and OECD
databases was used. Yearly data on immigrant inows
from 13 new member states (Cyprus, Czech Republic,
Estonia, Hungary, Latvia, Lithuania, Malta, Poland,
Slovakia, Slovenia, Bulgaria, Romania and Croatia) to
15 developed European countries (Austria, Belgium,
Denmark, Finland, France, Germany, Greece, Ireland,
Italy, Luxemburg, Netherlands, Portugal, Spain,
Sweden and United Kingdom) was used for the re-
search. More precisely, data on yearly immigrant
inows into 15 developed European countries by
country of origin was used for the period from 2000
to 2017. The variable of interest is the emigration rate.
Namely, it was empirically tested which determinants
aect emigration rate the most. Therefore, explanato-
ry variables are the following: GDP per capita (in PPP)
in both destination and origin countries, unemploy-
ment rate (in destination and origin countries), the
rate of unemployment of young population (in both
origin and destination countries) and dummy variable
EUmember indicating whether the origin country was
a member of the EU in particular year of the analysis.
All variables except dummy variable are expressed in
natural logarithms, so their coecient estimates are
interpreted as elasticities. Table 1 shows descriptive
statistics of analysed variables.
As evident from Table 1, this is an unbalanced pan-
el data set, since the number of observations is not
the same for all of the analysed periods.
In a manner of Mayda (2005), we have estimated
empirical model that includes emigration rate as the
dependent variable, minding the pull and push fac-
tors that are on average positive and negative de-
pending on the country of origin or the destination. As
the proxy variable for wage in destination and origin
countries, we have used GDP per capita in Purchasing
Power Parities. In addition, as a determinant of migra-
tion ows we have used the rate of unemployment of
young population (between 15 and 24 years old) in
Table 1: Descriptive statistics of the variables included in the baseline model (2000-2017)
Variable Number of
observations
Mean Standard
deviation
Minimum Maximum
Emigration rate 2,325 -9.39 1.86 -14.51 -4.21
GDP pc in immigration country 3,510 11.40 0.28 10.87 12.34
GDP pc in emigration country 3,510 10.82 0.23 9.99 11.37
Unemployment rate in immigration country 3,510 1.99 0.47 0.59 3.31
Unemployment rate in emigration country 3,480 2.17 0.42 1.06 2.99
Unemployment rate of young population
immigration country
3,509 2.84 0.51 1.72 4.07
Unemployment rate of young population in
emigration country
3,510 2.99 0.40 2.07 3.91
Dummy variable EUmember 3,492 0.71 0.45 0 1
Note: all the variables are expressed in natural logarithms, except the dummy variable EUmember
Unauthentifiziert | Heruntergeladen 01.01.20 12:47 UTC
DETERMINANTS OF MIGRATION FOLLOWING THE EU ENLARGEMENT: A PANEL DATA ANALYSIS
19South East European Journal of Economics and Business, Volume 14 (2) 2019
both origin and destination countries. Our model is a
xed eect panel model, so we have employed coun-
try and time specic eects to avoid biased estimates.
Also, our regressions have robust standard errors clus-
tered by country pair (destination and origin country),
to address heteroscedasticity and allow for correlation
over time of country pair observations. The baseline
empirical model of the determinants of immigration
ows is the following:
(1)
where i is the origin country, j is the destination coun-
try, and t is the time. lnemrateijt is the logarithm of emi-
gration rate from i to j at time t. lnGDPpc is the (ln) per
worker GDP, PPP-adjusted. lnunemployment is the (ln)
unemployment rate and lnunempoyoung is the (ln) un-
employment rate of young (15-24). DEUMember is the
dummy variable that equals 1 if the country of origin
is the member of EU at the analysed time (2000-2017).
Finally, the baseline model also includes destination
and origin countries’ xed eects and year eects. In
order to account for endogeneity in time series di-
mension, we used lagged values of (ln) GDP per capita
in both, destination and origin country.
4. RESULTS OF THE EMPIRICAL ANALYSIS
Presented Table 2 contains the estimation results
of the baseline model given by the equation 1. The
estimates are broadly consistent with the theoretical
predictions of the international migration model.
According to the presented results, the elasticity of
emigration rate to changes in GDP per capita in des-
tination country is statistically signicant and posi-
tive with the coecient being 3.41. Furthermore, the
elasticity of emigration rate to changes in GDP per
capita in origin country is also statistically signicant
and positive, but the coecient is much smaller (0.96).
The unemployment rate in origin country is positive
and statistically signicant (on 5% signicance level)
with the elasticity of 0.4563. Youth unemployment
is often examined separately because it tends to be
higher than unemployment in older age groups. It
usually comprises of labour force aged 15 to 24 years
old. According to our baseline model, the elastic-
ity of emigration rate to changes in unemployment
rate of young population is negative and statistically
signicant in the case of destination country (coef-
cient is -0.89), and it is negative, but not statistically
signicant in the case of origin country. It is also im-
portant to emphasise that the used Eurostat data on
youth unemployment includes only those young peo-
ple that are in the labour market and not the propor-
tion of all unemployed young adults. Furthermore, the
dummy variable EUmember is statistically signicant
and positive (in the case that the origin country is a
member of the EU).
Empirical results are due to specic statistical
methodology that was used, and they are somewhat
exploratory in their nature. However, they are partially
01 1
2 13
45
6
ln ln
ln ln
ln ln
ln
ijt it
jt i
ji
j t ijt
emrate GDPpc
GDPpc unemployment
unemployment unempyoung
unempyoung DEUmember I Table 2: Fixed eects panel model of the determinants
of immigration ows from the new member states (2000-
2017) 1
Dependent variable:
Ln Emigration rate
NMS
Independent variables:
Ln GDP pc in immigration country (t-1) 3.409 ***
[0.685]
Ln GDP pc in emigration country (t-1) 0.9559 ***
[0.3466]
Ln Unemployment rate in immigration
country
0.3478*
[0.1789]
Ln Unemployment rate in emigration
country
0.4563**
[0.2245]
Ln Unemployment rate of young
population immigration country
-0.8945***
[0.2413]
Ln Unemployment rate of young
population in emigration country
-0.1005
[0.2143]
Dummy variable EUmember 0.6681***
[0.0753]
Time xed eects 0.0326***
[0.0089]
Constant -123.449***
[14.2702]
Number of observations 2,171
Number of groups (countries) 157
R square 0,5369
F(8,156) 81.81***
Note: the estimated empirical model is a panel model with
pair of country xed eects and time xed eects; robust
standard errors clustered at the country-pair level are given
in parenthesis; ***, ** and * denote 1%, 5% and 10% signi-
cance level, respectively.
1 All relevant diagnostic tests for the estimated baseline model
were conducted. They are not shown here in order to save space,
but are available from the authors upon the request.
Unauthentifiziert | Heruntergeladen 01.01.20 12:47 UTC
DETERMINANTS OF MIGRATION FOLLOWING THE EU ENLARGEMENT: A PANEL DATA ANALYSIS
20 South East European Journal of Economics and Business, Volume 14 (2) 2019
consistent with Mayda (2005) where the emigration
rate is positively related to the destination country per
worker GDP, but the impact of income opportunities
change at home on the emigration rate was found to
be insignicant. The share of the young population in
the origin country has a positive and signicant im-
pact on emigration rates, but Mayda dened young
people as a group from 15-29 years old. Grau Grau and
Ramirez Lopez (2017) found that GDP per capita and
GDP growth behave similarly, with reasonably high
signicance levels (5%). The positive nature and size of
the coecients indicate that they are decisive for mi-
gration ow and positively aect growth in the num-
bers of immigrants entering Europe.
In the EU, unemployment has been on the rise
since 2008, which is due to the economic crisis which
caused considerable job loss, fewer job oerings, and
consequently, a rise of the unemployment rate. Older
workers are struggling to nd new jobs despite their
experience, and young graduates are struggling to
nd new jobs because there are no new workplaces
created (Statista 2018). Following the results of this
research which indicate that young population un-
employment rate in the immigration country is sig-
nicant and negatively correlated with the emigration
rate, it can be concluded that lower unemployment
rates are an important pull factor of migration. If there
are no jobs for young people in the country of origin,
they will look for better opportunities abroad, where
youth unemployment is decreasing.
The estimation results from our migration model
suggest that emigration rate responds rather quickly
to changes in GDP per capita and the youth unem-
ployment rate in immigration country. Thus, interna-
tional migration contributes signicantly to adjusting
the labour supply to uctuations in economic activity.
5. CONCLUDING REMARKS
The primary objective of this paper was to study
determinants of migration ows from the ‘new’ EU
member states to the ‘old’ member states that dier
signicantly in economic and other properties. The
emphasis was given to youth unemployment in both
origin and destination countries since young people
are a social group that can be particularly aected by
dierent push and pull factors. However, there is a lack
of empirical evidence on the importance of this factor
to migration ows.
In the empirical part of this analysis, the traditional
panel data estimator with time and country xed ef-
fects was used to explore the relationship between
emigration rate and the set of economic explanatory
variables: GDP per capita (in PPP) in both destination
and origin countries, unemployment rate (in destina-
tion and origin countries), the rate of unemployment
of young population (in both origin and destination
countries) and dummy variable EUmember. The data
set covered the period from 2000 to 2017, capturing
the period of the global nancial crisis as well as post-
crisis period for all analysed countries.
The results have shown that GDP per capita is a
signicant migration determinant with a positive sign
in both the emigration and immigration countries, but
it can be concluded that migrants are more motivat-
ed by the increase in the GDP per capita in the immi-
gration country than in the country of origin. Results
also reveal that the increase in the overall unemploy-
ment rate in the emigration country will increase the
emigration rate. Furthermore, research indicates that
young people are motivated by the perceived la-
bour market opportunities in the immigration coun-
try measured by the youth unemployment rate. Also,
dummy variable EUmember has shown to aect mi-
gration, although not all EU countries allowed free
movement of labour from the new member states im-
mediately after their accession.
Accordingly, the estimation results suggest that
emigration rate responds quickly to changes in GDP
per capita and unemployment rate of the young pop-
ulation in the immigration country. Thus, international
migration contributes signicantly to adjusting the la-
bour supply to uctuations in economic activity. The
results of the analysis enable a better understanding
of migration ows determinants, as an essential pre-
requisite for estimating future migration ows and
their overall potential eects on origin and destina-
tion countries.
Given the scale of emigration that new member
states have experienced since joining the EU, some
policy recommendations for these countries and
other candidate countries can be drawn down from
this study. As low levels of young population unem-
ployment have shown to be an important pull factor,
which can be a result of the perceived employment
opportunities in more developed countries, it is es-
sential not only to address the youth unemployment
levels, but also their status and development pros-
pects in countries of origin. Young people are drivers
of change in the society and are especially important
in ageing societies of Europe. Upon their emigration,
countries of origin lose their investment in educating
them and also lose a part of innovation capabilities
that younger population takes to destination coun-
tries upon migration.
In order to reveal key determinants leading young
people to migrate in detail, more research on this
Unauthentifiziert | Heruntergeladen 01.01.20 12:47 UTC
DETERMINANTS OF MIGRATION FOLLOWING THE EU ENLARGEMENT: A PANEL DATA ANALYSIS
21South East European Journal of Economics and Business, Volume 14 (2) 2019
topic is needed. Therefore, further research on push
and pull factors within the youth population could
be of great interest in both origin and destination
countries. Availability of the statistical data on young
population immigration by countries of origin would
enable further research of migration determinants
of this age group. Since in this study, a rather simple
panel data model was used, implications for policy
makers are not detailed nor exhaustive. However, the
taken panel approach could inspire future studies that
would reveal more details on the migration ows de-
terminants with special attention given to immigra-
tion of young population.
REFERENCES
Atoyan, M. R., Christiansen, L. E., Dizioli, A., Ebeke, M. C., Ilahi,
M. N., Ilyina, M. A., and Raei, M. F. 2016. Emigration and its
economic impact on Eastern Europe. International mon-
etary fund. Discussion Note SDN 16/07. https://www.imf.
org/external/pubs/ft/sdn/2016/sdn1607.pdf (accessed
January 15, 2019).
Doğan, D., Gizem, M., and Kabadayı, A. 2015. Determinants
of Internal Migration in Turkey: A Panel Data Analysis
Approach. Border Crossing 5 (1-2): 16-24.
Dorigo, G. and Tobler, W. 1983. Push-pull migration laws.
Annals of the Association of American Geographers 73
(1): 1-17.
Efendic, A. 2016. Emigration intentions in a post-conict
environment: evidence from Bosnia and Herzegovina.
Post-Communist Economies 28 (3): 335-352.
Eurostat. 2019. (database online). https://ec.europa.eu/eu-
rostat/ (accessed January 7, 2019).
Fan, S. and Stark O. 2007. The brain drain, educated unem-
ployment, human capital formation, and economic bet-
terment. Economics of transition 15 (4): 629-660.
Favell, A. 2008. The new face of East–West migration in
Europe. Journal of ethnic and migration studies 34 (5):
701-716.
Fouarge, D. and Ester, P. 2007. Factors determining inter-
national and regional Migration in Europe. European
Foundation for the Improvement of Living and Working
Conditions. Dublin.
Ganguli, I. 2018. Immigrant selection before and after com-
munism. Economics of Transition 26 (4): 649-694.
Garip, F. and Asad, A. L. 2015. Migrant networks. Emerging
Trends in the Social and Behavioral Sciences: An
Interdisciplinary, Searchable, and Linkable Resource.
Harvard University.
Giulietti, C., Guzi, M., Kahanec, M., and Zimmermann, K.
F. 2011. Unemployment benets and immigration:
Evidence from the EU. IZA Discussion Papers No. 6075.
Institute for the Study of Labor (IZA), Bonn. http://ftp.iza.
org/dp6075.pdf (accessed January 22, 2019).
Grau Grau, A. and Ramirez Lopez, F. 2017. Determinants of
Immigration in Europe. The Relevance of Life Expectancy
and Environmental Sustainability. Sustainability 9 (1093):
1-17.
Hatton, T. J. and Williamson, J. G. 2005. Global migration and
the world economy: Two centuries of policy and perfor-
mance. Cambridge, MA: MIT press.
Haug, S. 2008. Migration networks and migration decision-
making. Journal of Ethnic and Migration Studies 34 (4):
585-605.
Jurčić, Lj. and Barišić, A. 2018. Determinants, trends and
implications of modern migration. Paper presented
at The traditional consultation of the Croatian Society
of Economists: Economic Policy of Croatia, Opatija,
November.
Kahanec, M. and Zimmermann, K. F. 2011. High-Skilled
Immigration Policy in Europe. Discussion Paper No. 1096.
Institute for the Study of Labor (IZA), Bonn. http://ftp.iza.
org/dp5399.pdf (accessed January 15, 2019).
Kahanec, M., Pytlikova, M., and Zimmermann, K. F. 2014.
The Free Movement of Workers in an Enlarged European
Union: Institutional Underpinnings of Economic
Adjustment. IZA Discussion Papers No. 8456. Institute
for the Study of Labor (IZA), Bonn. https://pdfs.seman-
ticscholar.org/36bd/759457ed9a37dd52580e3e9258cfa
8ce7d68.pdf (accessed January 15, 2019).
Kim, K. and Cohen, J. E. 2010. Determinants of International
Migration Flows to and from Industrialized Countries: A
Panel Data Approach Beyond Gravity. International mi-
gration review 44 (4): 899-932.
Kleven, H., Landais, C., Muñoz, M., and Stantcheva, S. 2019.
Taxation and Migration: Evidence and Policy Implications.
National Bureau of Economic Research No. w25740.
Kotorri, M. 2017. The probability of return conditional on
migration duration: evidence from Kosovo. South East
European Journal of Economics and Business 12 (2):
35-46.
Krissman, F. 2005. Sin coyote ni patron: why the “migrant
network” fails to explain international migration.
International migration review 39 (1): 4-44.
Lee, E. S. 1966. A theory of migration. Demography 3 (1):
47-57.
Levitt, P. and Lamba-Nieves, D. 2011. Social remittances re-
visited. Journal of Ethnic and Migration Studies 37 (1):
1-22.
Liu, M. M. 2013. Migrant networks and international migra-
tion: Testing weak ties. Demography 50 (4): 1243-1277.
Mansoor, A. and Quillin, B., eds. 2006. Migration and remit-
tances: Eastern Europe and the Former Soviet Union
(chapter 3). Washington: World Bank.
Massey, D., Arango, J., Hugo, G., Kouaouci, A., Pellegrino, A.,
and Taylor, J. 1993. Theories of International Migration:
Unauthentifiziert | Heruntergeladen 01.01.20 12:47 UTC
DETERMINANTS OF MIGRATION FOLLOWING THE EU ENLARGEMENT: A PANEL DATA ANALYSIS
22 South East European Journal of Economics and Business, Volume 14 (2) 2019
A Review and Appraisal. Population and Development
Review 19 (3): 431- 466.
Mayda, A. M. 2005. International Migration: A Panel Data
Analysis of Economic and Non-Economic Determinants.
Discussion Paper No. 1590. The Institute for the Study
of Labor, IZA, Bonn. http://ftp.iza.org/dp1590.pdf (ac-
cessed January 15, 2019).
Mayda, A. M. 2010. International migration: A panel data
analysis of the determinants of bilateral ows. Journal of
Population Economics 23 (4): 1249-1274.
Meyer, J. B. 2001. Network approach versus brain drain: les-
sons from the diaspora. International migration 39 (5):
91-110.
Minelgaite, I., Christiansen, Þ. H., and Kristjánsdóttir, E. S.
2019. Lithuanian temporary workers in Iceland in anoth-
er economic boom: Expectations and experiences. The
South East European Journal of Economics and Business
14 (1): 101-114.
OECD. 2019. International Migration Database (da-
tabase online). https://stats.oecd.org/Index.
aspx?DataSetCode=MIG (accessed January 7, 2019).
Oh, Y. and Jung, J. 2013. Determinants of International
Labor Migration to Korea. KIEP Research Paper No. 13-
08. http://dx.doi.org/10.2139/ssrn.2452880 (accessed
February 25, 2019).
Panzaru, C. 2013. The Determinants of International
Migration. A Panel Data Analysis. Journal of Politics and
Law 6 (1): 142-148.
Rakauskienė, O. G. and Ranceva, O. 2014. Youth unemploy-
ment and emigration trends. Intellectual Economics 8
(1): 165-177.
Ravlik, M. 2014. Determinants of international migration:
a global analysis. Working papers series WP BRP 52/
SOC/2014. National Research University Higher School
of Economics. https://www.hse.ru/data/2014/10/02/11
00265067/52SOC2014.pdf (accessed February 15, 2019).
Son, L. and Noja, G. G. 2012. A macroeconometric panel
data analysis of the shaping factors of labour emigra-
tion within the European Union. Theoretical and Applied
Economics 11 (576): 15-30.
Statista. 2018. Youth unemployment rate in EU member
states as of December 2018. https://www.statista.com/
statistics/266228/youth-unemployment-rate-in-eu-
countries/ (accessed February 25, 2019).
Sulaimanova, B. and Bostan, A. 2014. International Migration:
A Panel Data Analysis of the Determinants of Emigration
from Tajikistan and Kyrgyzstan. Eurasian Journal of
Business and Economics 7 (13): 1-9.
United Nations. 2016. Youth and migration. UN policy brief.
UN Department of Economic and Social Aairs. https://
www.un.org/esa/socdev/documents/youth/fact-sheets/
youth-migration.pdf (accessed 25 February, 2019).
Van Mol, C. 2016. Migration aspirations of European youth in
times of crisis. Journal of Youth Studies 19 (8): 1303-1320.
Zaiceva, A. and Zimmerman, K. F. 2008. Scale, diversity, and
determinants of labour migration in Europe. Oxford
Review of Economic Policy 24 (3): 427–451.
Zhao, Y. 2003. The role of migrant networks in labor migra-
tion: The case of China. Contemporary Economic Policy
21 (4): 500-511.
Wickramasinghe, A. A. I. N. and Wimalaratana, W. 2016.
International migration and migration theories. Social
Aairs 1 (5): 13-32.
Unauthentifiziert | Heruntergeladen 01.01.20 12:47 UTC
