CHAPTER 1
23
Some of the trends and motivations for migration in the Europe and
Central Asia (ECA) region are similar to those found elsewhere in the
world. However, many of the migration movements that have taken
place since 1990 are unique to the region, given the circumstances of
economic transition, political and social liberalization, and the
breakup of three federal states. Figure 1.1 shows how the factors
influencing migration have changed from the communist period to
the present. This chapter provides an overview of some of the main
migration trends that have taken place across the region over the past
15 years, with a focus on international movements among countries.
Migration in the ECA region is both large by international standards
and unique in that the region is both a major receiver and
sender of migrants. Figure 1.2 exhibits the ECA region and selected
ECA countries in terms of their shares of foreign-born populations.
Excluding movements between industrial countries, ECA accounts
for over one-third of world emigration and immigration. There are 35
million foreign-born residents in ECA countries, including 13 million
in the Russian Federation, 7 million in Ukraine, 3 million in Kazakhstan,
3 million in Poland, and 1.5 million in Turkey. Furthermore,
several ECA countries are among the top 10 sending and receiving
countries of migrants worldwide. Russia is home to the second largest
number of migrants in the world after the United States; Ukraine is
Overview of Migration Trends in
Europe and Central Asia, 1990–2004
01-ECA_Migration.qxd 11/10/06 11:58 AM Page 23
24 Migration and Remittances: Eastern Europe and the Former Soviet Union
fourth after Germany; and Kazakhstan and Poland are respectively
ninth and tenth.
Migration patterns in the region follow a broad biaxial pattern: on
one axis a migration system developed among the countries of Western,
Central, and Eastern Europe and on the other a system of movement
arose among the countries of the Commonwealth of
Independent States (CIS). However, this system is not exclusively bipolar.
Though the majority of migrants from Central and Eastern
European countries move into Western Europe, the same is true for
many migrants from the poorer CIS economies, particularly Moldova.
While the majority of migrants from Central Asia travel to the
resource-rich CIS countries (particularly Russia and Kazakhstan)
many move west in search of higher earnings, toward the European
Union (EU) and Turkey.
The creation of many new countries following the breakup of the
Soviet Union produced “new” migrants (long-term, foreign-born residents)
who may not have physically moved, but were defined as
migrants under UN practice. In addition to the issue of these “statistical”
migrants, there are numerous other problems in analyzing migration
trends across the region based on available data. This chapter and
the report in general are an attempt to pull together and analyze all
available migration data to gain as complete a picture as possible of
migration trends over the past 15 years; thus, the issue of the veracity
of migration data is a constant theme.
The chapter begins with a description of some of the problems
involved in measuring migration among the ECA countries during
FIGURE 1.1
Transition of the Migration System in the Europe and Central Asia Region
Migration under Central Planning in the Europe and Migration during the Transition Period in the Europe
Central Asia Region and Central Asia Region
Eight countries in the region (only five remain in their pretransition borders) Twenty-seven countries following the breakup of three federal states
Migration was very tightly controlled Much less control over migration
Prices were administratively set and wages and income were not very Prices are market determined and income is increasingly distributed
differentiated across sectors or regions among people, sectors, and regions
A massive and elaborate system of subsidies caused certain sectors and Wages and prices have adjusted to their market-clearing value
regions to be “over-valued” and others to be “under-valued”
Migration control efforts were aimed mainly at keeping people in a country Migration control is aimed at both keeping people in and outside a
country and, in general, migration control systems are poorly developed
Little involvement in international institutions and foreign trade Open economies, involvement with international institutions, and
“globalization”
Source: World Bank staff.
01-ECA_Migration.qxd 11/10/06 11:58 AM Page 24
Overview of Migration Trends in Europe and Central Asia, 1990–2004 25
FIGURE 1.2
Migration in Top 10 Sending and Receiving Countries and by Region, 2003
thousands
ECA Western Europe others
per thousand population
thousands per thousand population
Immigration into the top 10 receiving countries and by Region, 2003 (stock of immigrants)
Emigration from the top 10 sending countries and by Region, 2003 (stock of immigrants)
0 50 100 150 200 250 300 350 400
United States
Russian Fed.
Germany
Ukraine
France
India
Saudi Arabia
Australia
Kazakhstan
Poland
Countries
North America
ECA
Europe
Middle East and North Africa
South and West Asia
Sub-Saharan Africa
East and South East Asia
Others
Latin America
Regions
0 100 200 300 400 500
Russian Fed.
Mexico
Ukraine
India
China
Germany
Kazakhstan
Turkey
Philippines
Uzbekistan
Countries
ECA
Latin America
Europe
East and South East
South and West Asia
Middle East and North Africa
Others
Sub-Saharan Africa
North America
Regions
0 100 200 300 400 500 600 700
United Arab Emirates
Estonia
Latvia
Switzerland
Australia
Saudi Arabia
Kazakhstan
Ukraine
Belarus
United States
Countries
ECA
Middle East and North Africa
Sub-Saharan Africa
Latin America
South and West Asia
East and South East Asia
Developing Regions
0 50 100 150 200 250 300 350 400
Jamaica
Bosnia and Herzegovina
Albania
Slovenia
Armenia
Kazakhstan
Belarus
El Salvador
Georgia
Moldova
Countries
ECA
Latin America
Middle East and North Africa
Sub-Saharan Africa
East and South East Asia
South and West Asia
Developing Regions
Sources: UN Population Division 2003 and Walmsley, Ahmed, and Parsons 2005.
01-ECA_Migration.qxd 11/10/06 11:58 AM Page 25
26 Migration and Remittances: Eastern Europe and the Former Soviet Union
the transition period. Then using the data that are available, it analyzes
the impact of migration on overall levels of population change
in the ECA countries. The next section provides a broad overview of
migration flows across the ECA region during the period 2000–03, a
recent period when most of the ethnic migration had already taken
place and flows were dominated by the economic migration flows
that are expected to predominate in the future. Following this are discussions
of refugee and internally displaced population movements,
and transit and irregular migration. A further section looks at the
main migration partners of each ECA country. Finally, the chapter
looks at possible future migration trends in the region.1
Problems with Measuring Migration in ECA
There are three main sources for migration data in the ECA countries,
as well as in countries outside the region. These are population censuses,
usually conducted once a decade; administrative statistics of
persons crossing international borders; and surveys. This final category
includes surveys targeted directly at migrant populations, as well
as surveys designed for other purposes where migration-related questions
are asked.
Population censuses usually include questions that measure lifetime
migration. For instance, the last Soviet census, conducted in January
1989, included questions on place of birth, whether the respondent
had been living in his or her present residence continuously since birth,
and if not, when he or she had migrated to that place. All of the ECA
countries conducted population censuses between the years 1989 and
1992 and most conducted another census between 1999 and 2002. The
more recent round of censuses typically included a question on citizenship,
though this question was frequently not posed in the censuses
conducted around 1990. Some also included questions about persons
temporarily absent. The 2002 Russian census also included a set of
questions for those persons temporarily residing in Russia, although
the total of a quarter million persons enumerated were thought to significantly
underestimate the true figure.
Whereas censuses attempt to count stocks of migrants, administrative
statistics are counts of flows of migrants. In most cases, data on
total international border crossings also record information on the
age, sex, and country of previous residence or intended destination,
and other characteristics of migrants. It is the change, and in some
cases breakdown, of systems for measuring migration flows where
the ECA countries have suffered the most.
01-ECA_Migration.qxd 11/10/06 11:58 AM Page 26
Overview of Migration Trends in Europe and Central Asia, 1990–2004 27
Surveys are useful for obtaining qualitative information about
migrants and to serve as a check on the veracity of flow statistics from
administrative sources. An increasing number of surveys of migrants
have been conducted across the ECA region, both by the countries
themselves as well as by international organizations such as the International
Organization for Migration (IOM).
Several reasons make migration flows in ECA challenging to capture.
First, the type, direction, and magnitude of the flows in the region
have changed dramatically since the beginning of economic transition,
liberalization of societies (including increased freedom of movement),
and the emergence of 22 new states. What had previously been internal
boundaries have now become international borders. Migration in
ECA, which was once subject to considerable state control within several
self-contained migration spaces, now rests in the hands of individuals
who have the ability to transit across new and rather porous
international boundaries. In the former Soviet Union, the propiska or
resident permit system required persons to register before being
allowed to migrate to a new location. However, the visa-free travel
among the CIS countries for most of the 1990s contributed to an environment
of porous borders, which made the recording of migration
flows difficult. The extent to which the successor states have instituted
systems to properly measure total migration flows and to disaggregate
these flows by age, gender, nationality, and other characteristics useful
for analysis and policy making varies considerably.
The previous systems for measuring migration in the centrally
planned countries of the ECA are wholly inadequate for capturing
movements across the newly independent states. In their initial years,
the newly independent states had to erect the elements of government
apparatus, including independent statistical systems to measure
social and economic trends such as migration movements. With other
elements of state building causing greater concern, building systems
for measuring migration often received low priority. Many of these
issues in migration measurement are unique to the newly independent
states of the ECA region.
A second set of problems with proper migration measurement is
endemic to all countries. Definitions, underlying concepts, sources,
and reporting systems differ significantly between countries, making
available migration statistics fragmentary. The boundaries between
extended travel, seasonal work, and economic migration are blurred.
In most cases it is not clear whether an individual reported as
“migrant” is a long-term mover, a temporary mover, a seasonal
worker, someone on the move to another destination, an individual
transitioning through a territory, a returning migrant, a member of a
01-ECA_Migration.qxd 11/10/06 11:58 AM Page 27
28 Migration and Remittances: Eastern Europe and the Former Soviet Union
family already residing abroad with no intention to work, a student
(who may or may not undertake part-time employment), a refugee,
a member of the staff of a foreign company in the country, or some
other category of migrant.
Third, undocumented migration plays an important role in today’s
migrant flows to, from, and within ECA, as well as in many other parts
of the world. Reported data refers to legal migrants, based most often
on residence or work permits. Even countries in the region with seemingly
well-developed statistical systems often are not able to record
migration completely. Decennial population censuses are used to adjust
and calibrate population totals. For instance, in Lithuania, there was a
downward adjustment of the population by over 200,000, or more
than 5 percent of the population, following the census conducted there
in April 2001. Roughly the same magnitude of adjustment took place
in Estonia following its March 2000 census, when it adjusted the population
total downward by 67,000, or about 5 percent. Similar postcensus
adjustments downsizing the resident population were made in
the Czech Republic, Poland, and the Slovak Republic. Among the surprises
in the Russian census conducted in October 2002 was that the
total population was 1.2 million higher than the previous estimate,
mainly because of an undercount of migration.
These differences between population estimates and census figures
in the ECA countries are worth comparing to the experience of the
United States, long a traditional migration destination. Before the
2000 census in the United States, the population was estimated at
275 million. That census revealed a count of 281 million, a difference
of 6 million, almost all attributable to an undercount of the huge
migration into the United States during the 1990s.2 The United States
has long grappled with an issue that the ECA states are only beginning
to deal with in trying to estimate temporary or circular migration.
Until recently, most of the ECA states recorded only long-term
or permanent moves and much of the movements over the past
decade are of a temporary or circular nature.
The breakup of the Soviet Union, Yugoslavia, and Czechoslovakia
created a large number of “statistical migrants.” The commonly
accepted UN definition describes a “migrant” as a person living outside
his or her country of birth. As used here, statistical migrants
refers to persons who migrated internally while those countries
existed, thus not qualifying as a migrant under the UN definition at
the time, but who began to be counted as migrants when those countries
broke apart even though they did not move again. Having a large
number of these statistical migrants has hampered analysis of migration
patterns across the ECA region because of the difficulty of sepa-
01-ECA_Migration.qxd 11/10/06 11:58 AM Page 28
Overview of Migration Trends in Europe and Central Asia, 1990–2004 29
rating those who moved during the communist period, before the
start of transition and independence, and those who moved later for
ethnic or economic reasons. However, with data that are available
from population censuses, it is possible to get a fairly good idea of the
total number of statistical migrants and changes in their numbers
since the breakup of the countries.
Table 1.1 shows the population of the Soviet Union by place of birth
in 1989, at the time of the last Soviet census. At that time, 2.4 million
persons or 0.8 percent of the population had been born outside the
Soviet Union.3 This low figure is not surprising because for most of the
period between the end of World War II and the breakup of the Soviet
Union, there was little migration either into or out of the Soviet Union
and little shifting of international borders. In fact, the listed figure of
the Soviet population being classified as migrants is likely a considerable
overestimate because it also includes those not indicating their
place of birth. If similar data from the 2002 Russian census is any
guide, about one-quarter had actually been born outside the former
Soviet Union and about three-quarters did not indicate their place of
birth. Thus, the true figure of the migrant population was likely less
than 1 million or only about 0.3 percent of the population.
TABLE 1.1
Population by Place of Birth in the USSR, 1989
(thousands)
Born in Born in
Place of republic of Born Born republic of Born Born
permanent current elsehwere outside Total current elsehwere outside
residence residence in USSR USSR population residence in USSR USSR
USSR 255,409 27,955 2,378 100.0 89.4 9.8 0.8
RSFSR 135,550 10,478 994 100.0 92.2 7.1 0.7
Ukrainskaia SSR 44,332 6,665 455 100.0 86.2 13.0 0.9
Belorusskaia SSR 8,883 1,213 55 100.0 87.5 12.0 0.5
Uzbekskaia SSR 18,108 1,649 53 100.0 91.4 8.3 0.3
Kazakhskaia SSR 12,715 3,536 214 100.0 77.2 21.5 1.3
Gruzinskaia SSR 5,039 349 13 100.0 93.3 6.5 0.2
Azerbaidzhanskaia SSR 6,604 398 19 100.0 94.1 5.7 0.3
Litovskaia SSR 3,299 356 19 100.0 89.8 9.7 0.5
Moldavskaia SSR 3,739 579 18 100.0 86.2 13.3 0.4
Latviiskaia SSR 1,975 678 14 100.0 74.0 25.4 0.5
Kirgizskaia SSR 3,586 638 34 100.0 84.2 15.0 0.8
Tadzhikskaia SSR 4,650 433 9 100.0 91.3 8.5 0.2
Armianskaia SSR 2,570 267 467 100.0 77.8 8.1 14.1
Turkmenskaia SSR 3,205 311 7 100.0 91.0 8.8 0.2
Estonskaia SSR 1,155 403 8 100.0 73.7 25.7 0.5
Source: Eastview Publications and CIS Statistical Committee; USSR Census Results 1989 CD-ROM.
Note: Data are as of January 1989.
01-ECA_Migration.qxd 11/10/06 11:58 AM Page 29
30 Migration and Remittances: Eastern Europe and the Former Soviet Union
However, there was considerable migration among the states of
the former Soviet Union. In 1989, there were 28 million persons who
were residing in a republic other than the one in which they were
born. This figure amounted to 9.8 percent of the Soviet population,
which should be regarded as the number of “statistical migrants” that
were created by the breakup of the Soviet Union, greatly contributing
to the increase in the world stock of migrants. The bulk of these individuals
were in the three Slavic states, Uzbekistan, and Kazakhstan.
In percentage terms, the countries with the largest migrant stock populations
were Estonia, Latvia, and Kazakhstan. All of these countries
were prime destinations for Russian and Russian-speaking migrants
during the period after World War II.
Migration and Population Change
An analysis of migration and population change among the ECA
states begins at a broad level by dividing the countries into groups
according to their recent patterns of migration and natural increase in
population (figure 1.3; data underlying this figure are in appendix 2).
Natural increase is the difference between the number of births and
deaths and is a function of the age structure of the population and
levels of fertility and mortality. As will be discussed below, differential
rates of natural increase among countries are a major driver of migration
within the ECA region and elsewhere. A positive natural increase
occurs where the number of births exceeds the number of deaths,
which is the situation in nearly all countries in the world. Negative
natural increase is where the number of deaths in a population
exceeds the number of births. The 14 ECA countries shown below
with a negative natural increase or a natural decrease, along with
Italy and Germany, are among a small group of countries where this
is occurring. So many ECA countries are part of this group because
fertility levels have fallen steeply during the transition period, to 1.3
children per woman or less; such levels are unsustainable for natural
population increase.4 These figures are compared to net migration,
which is the difference between the number of immigrants to a country
and emigrants from a country.
There are two countries in the ECA region that have both a natural
increase and positive net migration; however, neither truly
belongs in this category because both suffer from data problems that
affect their migration counts. Turkmenistan has some rather unrealistically
high population estimates, which cause net migration figures
to appear unrealistically high. Bosnia and Herzegovina suffers from
01-ECA_Migration.qxd 11/10/06 11:58 AM Page 30
Overview of Migration Trends in Europe and Central Asia, 1990–2004 31
incomplete and inconsistent counts of migration, with some years
showing emigration and some immigration. Furthermore, in recent
years, there has been an undetermined amount of return migration
of some of the refugee populations that left during the mid-1990s.
Based on this evidence, both of these countries should probably be
grouped in the category of countries with positive natural increase
and negative net migration.
There are 10 ECA countries that combine natural increase and net
emigration (12 if the two mentioned above are included). This is the
pattern for most of the world’s countries. This includes the countries
of Central Asia, the Caucasus, and many of the former Yugoslav
states. With their faster-growing populations, especially youth populations,
migration pressures in these countries will likely persist into
the future.
A third group of countries comprises those that combine having
more deaths than births and more immigrants than emigrants. These
are Russia and Belarus in the CIS and four of the smaller new EU
member states. While all have had more immigrants than emigrants
over recent years, in all but Russia, the population increases as a result
FIGURE 1.3
Natural Increase (Decrease) and Net Migration in the ECA Region, 2000–03
Ϫ4
Ϫ2
0
2
4
6
8
10
Ϫ6 Ϫ5 Ϫ4 Ϫ3 Ϫ2 Ϫ1 0 1 2
net migration (percent change)
naturalincrease/decrease(percentchange)
positive natural increase,
negative net migration
negative natural increase,
negative net migration
negative natural increase,
positive net migration
positive natural increase,
positive net migration
Albania
Tajikistan
Hungary
Turkmenistan
Bosnia and
Herzegovina
Russian Fed.
Uzbekistan
Kyrgyz Rep.
Azerbaijan
Kazakhstan
Macedonia
Armenia
Serbia and Montenegro
Slovenia
Belarus
Czech Rep.
Georgia
Slovak Rep.Poland
Moldova
Romania
Croatia
Lithuania
Latvia
Ukraine
Bulgaria
Source: National statistical office of the ECA countries and UNICEF, TransMONEE Database.
01-ECA_Migration.qxd 11/10/06 11:58 AM Page 31
32 Migration and Remittances: Eastern Europe and the Former Soviet Union
of net migration are small, amounting to less than 1 percent of their
populations. As pointed out elsewhere in this report, Russia has
become a major migration magnet within the CIS, with a measured
population increase from migration of 4 percent since 1990 and perhaps
an equal amount of undocumented migration.
A fourth group are nine ECA countries where populations are
declining because they experience both more deaths than births and
more emigrants than immigrants. This includes Ukraine and
Moldova, the three Baltic states, and four Central European countries,
including the largest, Poland. In all of these countries, both
trends are expected to continue well into the future, causing large
population declines as well as rapid aging of their populations.
Figures 1.4a and 1.4b show the net population change from migration
over the period 1989–2003 for the CIS and western ECA countries,
respectively.5 From this figure, one part of the region’s bipolar
migration story of the past decade and a half can be clearly seen, with
Russia showing by far the largest population gain from migration. The
impact on those other few countries with population gains from
migration has been minimal. Most of the migrants into Russia consist
of persons migrating from the other states of the former Soviet Union,
which show large population declines from migration. There have
been several countries in the region that have transitioned from net
emigration to net immigration including Belarus, Slovenia, Hungary,
Croatia, and Serbia and Montenegro.
Of the five ECA countries with population declines of over 15 percent,
four are in the southern tier of the former Soviet Union. The
three Baltic states have had considerable out-migration in large part
because of the emigration of large numbers of Russians and Russianspeakers
in the years immediately following the breakup of the Soviet
Union. In southeast Europe, Albania and Bulgaria have also had emigrations
of large portions of their populations.
These figures are based on counts of the long-term, permanent
migration of the populations and do not include short-term or undocumented
counts of population movements. These figures also understate
the potential impact of migration because it is usually the
better-educated segments of the population and those in the early
stages of their working lives who migrate in the largest numbers.
Major Migration Flows in the ECA Region
As mentioned often throughout this report, proper measurement of
migration is difficult, even for high-income countries with well-devel-
01-ECA_Migration.qxd 11/10/06 11:58 AM Page 32
Overview of Migration Trends in Europe and Central Asia, 1990–2004 33
oped statistical systems. For the ECA countries, measuring migration
during this period of rapid social, economic, and political change has
been especially difficult. However, by compiling migration data from
several different sources and triangulating, a fairly complete picture
of the major flows taking place within the region can be obtained. It
FIGURE 1.4
Net Migration in Western ECA and the CIS
a. Net migration in the CIS, 1989–2003
Ϫ25 Ϫ20 Ϫ15 Ϫ10 Ϫ5 0 5 10
Kazakhstan
Georgia
Armenia
Tajikistan
Estonia
Kyrgyz Rep.
Latvia
Uzbekistan
Lithuania
Moldova
Azerbaijan
Ukraine
Belarus
Russian Fed.
net migration as percent of population at beginning of period
population losses
from migration
population gains
from migration
b. Net migration in Western ECA, 1989–2003
Ϫ30 Ϫ25 Ϫ20 Ϫ15 Ϫ10 Ϫ5 0 5
Albania
Bulgaria
Romania
Serbia and Montenegro
Croatia
FYR Macedonia
Hungary
Slovenia
Slovak Rep.
Poland
Czech Rep.
net migration as a percent of population at beginning of period
1989–99 2000–03
population losses
from migration
population gains
from migration
Source: National statistical offices and UNICEF, TransMONEE Database.
01-ECA_Migration.qxd 11/10/06 11:58 AM Page 33
34 Migration and Remittances: Eastern Europe and the Former Soviet Union
is helpful to keep in mind that international migration involves a flow
between two countries and that when a person migrates, that person
ideally should be recorded twice, by both the sending and receiving
country. Even so, there is considerable variation in how countries
record migrants; some countries track movements of people by place
of previous or next residence, some by citizenship, and some by various
other methods.
Table 1.2 shows the migration flows among major blocs of ECA
countries and origins and destinations of flows outside the region for
the years 2000 to 2003. This was a period after much of the ethnicinduced
migration associated with the breakup of the Soviet Union,
Yugoslavia, and Czechoslovakia had already taken place and the magnitude
of migration flows had settled into a more “normal” pattern
influenced primarily by economic incentives. The table was compiled
by collecting all available data on migration by origin and destination
country according to both residence and citizenship definitions; this
was followed by calculating a “maximum” matrix of the highest of
each pair of flows. Migration data for 52 countries were collected,
comprising the 28 ECA countries, 21 countries in Western Europe,
plus Canada, Israel, and the United States. Sufficient data were available
to fill about 90 percent of the matrix. Most of the cells that were
not able to be filled represented flows between pairs of countries for
which there is not known to be substantial migration (for example,
between Iceland and Turkmenistan). Thus, the assembled data are
thought to be a fairly complete representation of migration involving
ECA countries during this period.
The data partially support the story that two major migration blocs
have developed involving migration of the ECA countries. As suspected
by other and anecdotal evidence, there has been considerable migration
from western ECA to Western Europe and considerable migration from
the rest of the CIS into Russia. At the same time, there are other flows
developing that were not suspected and not that readily apparent from
other data. About equal percentages of migrants from the CIS countries
other than Russia (other CIS) travel to Russia as to Western Europe,
with Ukraine and Kazakhstan being the major sending countries and
Germany the major receiver. Over 70 percent of migrants from western
ECA go to Western Europe. At the same time, there is also considerable
flow from Western Europe to western ECA. Flows between Germany
and three countries make up the bulk of this overall total, that is, flows
from Germany to Poland, Serbia and Montenegro, and Turkey. These
figures not only represent the return of persons who had previously
migrated but also indicate considerable “churning,” as for each of these
three flows, there are also large flows in the opposite direction.
01-ECA_Migration.qxd 11/10/06 11:58 AM Page 34
Overview of Migration Trends in Europe and Central Asia, 1990–2004 35
TABLE 1.2
Migration Flows Involving ECA Countries, 2000–03
TABLE 1.2A
Total Migration Flows Involving ECA Countries and Major Partners, 2000–03
To
U.S., Canada, Total
From Russia Other CIS Western ECA Western Europe Israel (emigration)
Russia 0 272,929 17,882 85,468 53,539 429,818
Other CIS 319,514 159,652 85,104 280,843 90,265 935,378
Western ECA 22,896 32,820 274,762 1,300,289 149,045 1,779,812
Western Europe 74,460 82,705 640,052 2,808,366 269,253 3,874,837
U.S., Canada, Israel 8,466 6,342 16,973 457,664 142,762 632,207
Total (immigration) 425,336 554,448 1,034,773 4,932,630 704,864
TABLE 1.2B
Percent of Total Emigration
To
U.S., Canada, Total
From Russia Other CIS Western ECA Western Europe Israel (emigration)
Russia 0 63 4 20 12 100
Other CIS 34 17 9 30 10 100
Western ECA 1 2 15 73 8 100
Western Europe 2 2 17 72 7 100
U.S., Canada, Israel 1 1 3 72 23 100
TABLE 1.2C
Percent of Total Immigration
To
U.S., Canada,
From Russia Other CIS Western ECA Western Europe Israel
Russia 0 49 2 2 8
Other CIS 75 29 8 6 13
Western ECA 5 6 27 26 21
Western Europe 18 15 62 57 38
U.S., Canada, Israel 2 1 2 9 20
Total (immigration) 100 100 100 100 100
Source: See text for explanation of how data were compiled.
Note: “Other CIS” consists of Armenia, Azerbaijan, Belarus, Georgia, Kyrgyz Republic, Kazakhstan, Moldova, Tajikistan, Turkmenistan, Ukraine, and Uzbekistan.
“Western ECA” consists of Albania, Bosnia and Herzegovina, Bulgaria, Serbia and Montenegro, the Czech Republic, Estonia, Croatia, Hungary, Lithuania, Latvia, FYR
Macedonia, Poland, Romania, the Slovak Republic, Slovenia, and Turkey. “Western Europe” consists of Austria, Belgium, Switzerland, Cyprus, Germany, Denmark,
Spain, Finland, France, Greece, Ireland, Iceland, Italy, Liechtenstein, Luxembourg, Malta, the Netherlands, Norway, Portugal, Sweden, and the United Kingdom.
On the immigration side, Russia receives 75 percent of its immigrants
from other CIS countries. There are minimal flows from the CIS
states in the western ECA, with over half consisting of migrants from
Ukraine to the Czech Republic and from Moldova into Romania;
Ukraine and Moldova are thus unique in having significant migrant
01-ECA_Migration.qxd 11/10/06 11:58 AM Page 35
36 Migration and Remittances: Eastern Europe and the Former Soviet Union
flows both to Western Europe and to resource-rich CIS countries. The
largest flows into Western Europe are from other Western European
countries, making up about half of the total. However, flows into Western
Europe from western ECA make up about one-third of the total.
Figure 1.5 shows the largest country-to-country migration streams
involving a CIS country for the period 2000–03. Much of this is driven
by the gravity of proximity and population size; thus, it is not surprising
that Russia is either a source or destination of most of these flows.
The largest flows that do not include Russia are flows from Kazakhstan
to Germany and Ukraine to Germany. The flow from Ukraine
to Germany can be explained by proximity, population size, and large
differences in per capita income, while the flow from Kazakhstan to
Germany can be explained by the fact that Kazakhstan was home to
the largest concentration of Germans in the former Soviet Union and,
initially, Germany had a rather liberal law for the return of the
Aussiedler. The pull of Russia from the other CIS countries is clearly
evident from the map.
Figure 1.6 shows the largest country-to-country migration streams
involving a western ECA country for the same period. A quite different
pattern emerges than among CIS states, with a country outside
FIGURE 1.5
Largest Migration Flows Involving a CIS Country, 2000–03
Source: World Bank staff estimates based on analysis of migration statistics from a variety of sending and receiving countries.
01-ECA_Migration.qxd 11/10/06 11:58 AM Page 36
Overview of Migration Trends in Europe and Central Asia, 1990–2004 37
the region, Germany, being the major driver of migration for these
countries. Again the gravity of migration encompassing proximity,
population size, and the size of the German economy explains many
of the notable patterns. None of the largest flows involved two countries
within the region because there are only two countries, Turkey
and Poland, that can be considered sizable (or at least medium-sized
comparable to the largest Western European countries). What is
interesting is that all of the largest flows involving Germany are twoway
flows with large amounts of return migration.
Refugees and Internally Displaced Persons
Each of the ECA countries is an ethnic homeland. However, many
other ethnic homelands exist at the subnational level. The boundaries
of many of these were drawn arbitrarily by outside authorities and do
not necessarily coincide with what different ethnic groups regard as
their rightful homelands. During the communist period, there was
considerable migration of different ethnic groups to regions or countries
outside of their homelands. When Yugoslavia and the Soviet
FIGURE 1.6
Largest Migration Flows Involving a Western ECA Country, 2000–03
Source: World Bank staff estimates based on analysis of migration statistics from a variety of sending and receiving countries.
01-ECA_Migration.qxd 11/10/06 11:58 AM Page 37
38 Migration and Remittances: Eastern Europe and the Former Soviet Union
Union broke apart, they did so along their ethnic seams. Most of this
occurred peacefully but was accompanied by some diaspora migration.
However, in some cases these cleavages instigated considerable
ethnoterritorial conflict; as a result, forced migration became the predominant
form of migration in some parts of the region. Figure 1.7
shows the major displacements that took place in the former
Yugoslavia in 1995 at about the peak of the conflict there. Figure 1.8
shows the same for the former Soviet Union for the mid-1990s.
Figure 1.9 shows the temporal trends in the numbers of refugees
and internally displaced persons (IDPs) across the ECA region
between 1989 and 2003.6 The figure shows a combination of actual
and statistical trends. During the late communist period, the numbers
of refugees and IDPs were rather small. However, estimates rely on
imperfect data counting measures; none of these countries had
acceded to the 1951 Geneva Convention on Refugees and hence did
not have mechanisms in place for recognizing and counting refugees.
As the newly independent states in the region and others began to
erect institutions capable of enumerating refugees and asylum seekers,
their numbers began to increase. Thus, part of the rise from 1989
to the mid-1990s is statistical. However, a large part of the increase is
real, brought about by the increase in the number of persons disFIGURE
1.7
Main Displaced Populations from the Former Yugoslavia, December 1995
Source: Humanitarian Issues Working Group HIWG06/6, December 11, 1996.
01-ECA_Migration.qxd 11/10/06 11:58 AM Page 38
Overview of Migration Trends in Europe and Central Asia, 1990–2004 39
FIGURE 1.8
Main Displaced Population from the Former Soviet Union, Mid–1990s
Source: Based on IOM, CIS Migration Report 1996.
Note: Map is designed to broadly illustrate major refugee and IDP flows at the time, based upon best available information,
and is not intended to be authoritative or precise.
FIGURE 1.9
Refugees and Internally Displaced Persons in the ECA Region, 1989–2003
0
200,000
400,000
600,000
800,000
1,000,000
1,200,000
1,400,000
1989 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003
refugees IDPs
persons
Source: UNICEF, TransMONEE database.
01-ECA_Migration.qxd 11/10/06 11:58 AM Page 39
40 Migration and Remittances: Eastern Europe and the Former Soviet Union
placed as a result of the breakup of the Soviet Union and Yugoslavia
and the resultant ethnoterritorial disputes.
The number of refugees increased from 145,000 in 1989 to over a
half million during the years 1992 to 1997 (with the exception of
1994), but fell to about 237,000 in 2003. It should be kept in mind that
these figures refer to the numbers of refugees and IDPs within each of
the ECA countries, not from the countries. Refugees, by definition,
have crossed an international border, whereas IDPs have not. If the
number of refugees from the ECA countries were counted instead, the
number would certainly be higher because many of those from the
former Yugoslav states fled to Western Europe. Partly for these reasons,
the number of IDPs is comparatively much higher than for the
number of refugees, rising from about 100,000 in 1989 to over a million
during the years 1993 to 1997 before declining slightly to 927,000
in 2003. In 2003, the largest concentrations of IDPs were in Azerbaijan
(576,000) and Georgia (262,000). These numbers are down only
slightly from peaks in the mid-1990s because the conflicts that gave
rise to them continue to persist without any permanent settlement.
Figure 1.10 shows the countries in the ECA region with the largest
concentrations of refugees, IDPs, and “others of concern” at the end of
2004, according to the UNHCR. Overall, the ECA region accounts for
7.4 percent of the world population in total but contains 19 percent of
FIGURE 1.10
Largest Numbers of Refugees, IDPs, and Others of Concern in the ECA Region, 2004
0 100 200 300 400 500 600 700
Russian Fed.
Serbia and Montenegro
Azerbaijan
Latvia
Bosnia and Herzegovina
Georgia
Armenia
Estonia
Ukraine
Kazakhstan
thousands
refugees IDPs other
Source: UNHCR, 2004 Global Refugee Trends (http://www.unhcr.org).
01-ECA_Migration.qxd 11/10/06 11:58 AM Page 40
Overview of Migration Trends in Europe and Central Asia, 1990–2004 41
the total number of asylum seekers, refugees, and others of concern. In
particular, the ECA region accounts for a disproportionate share of the
world’s total number of IDPs (32 percent), because of past or ongoing
conflicts in Russia, Georgia, and Azerbaijan in the CIS, and in Serbia
and Montenegro and Bosnia and Herzegovina in the former Yugoslavia.
Substantial proportions of ECA migrants also fall into the category of
“others of concern,” which generally includes asylum seekers, returned
refugees, returned IDPs, and various other categories of (usually forced)
migrants. In ECA countries, this includes various categories of stateless
persons and, in Latvia and Estonia, the large Russian-speaking groups
of noncitizens. Aside from those two countries, the ECA countries with
the largest numbers of persons of concern are mostly those where there
has been or continues to be conflict. The region also accounts for a disproportionate
share of others of concern because of the large number
of stateless or noncitizens living in various countries. Many of the original
ethnoterritorial conflicts that gave rise to these groups of forced
migrants remain unresolved more than a decade after they first arose.
Transit and Undocumented Migration in the ECA Region
With the opening up of the ECA countries to the rest of the world and
the liberalization of migration, transit, illegal, and undocumented
migration has become an issue for countries in the region, and particularly
for those that were not previously under communist rule.
Some migrants (from within and outside ECA) hoping to migrate to
the United States, Japan, or Western Europe seek transit through
ECA countries. Some transit migrants then conclude that this hope is
unrealistic and settle in the transit country, which typically is poorer
than the West but more developed than their home country. Russia is
emerging as a transit as well as a key sending and receiving country.
Ukraine, Romania, and Azerbaijan are examples of other countries in
the ECA region that have significant transit migrant populations. This
section first considers the motivations of migrants who come to the
ECA. It then considers the experience of the host countries from two
perspectives: the statistical frequency of undocumented migration (a
figure notably difficult to calculate), and the policy decisions of ECA
states for regulating this phenomenon.
Migration Experiences
The decision to migrate, as well as the choice of destination, reflects a
careful calculation of relative risks and income-earning potential for
01-ECA_Migration.qxd 11/10/06 11:58 AM Page 41
42 Migration and Remittances: Eastern Europe and the Former Soviet Union
those who end up in ECA countries. A U.K. Economic and Social
Research Council survey of Fujianese Chinese finds that Europe was
the second choice for refugees unable to get to Japan or the United
States but who wanted to make money abroad within a set time.
Fujianese migrants choose their preferred migration destination
based on the likelihood of successfully getting there, expected
income, and the presence of relatives or friends. Availability of legal
residence status seems to be less important, although visa requirements,
perceived ease of obtaining refugee status, and amnesties for
undocumented migrants are all important in directing Fujianese (and
other Chinese) migrants to particular countries at particular times
(Economic and Social Research Council 2002).
A survey conducted from May to October 2003 of transit migrants
in Azerbaijan (IOM 2004) also determines that the motivations for
migration are the result of careful contemplation. Most such transit
migrants depart from developing countries in Asia and the Middle
East and aim to settle in North America or Western Europe. Some
would like to return home when the political and economic situations
in their home countries stabilize. Some entered and reside in
Azerbaijan legally, while others migrated illegally. Most undocumented
entries were through Iran, and were frequently assisted by
middlemen. “Push factors”—including conflict and economic difficulties
in the countries of origin—were the main motivations for migration.
For many, Azerbaijan was attractive owing to its geographical
proximity to and cultural similarities with their homeland.
Countries with generous immigration provisions, such as Ukraine,
also have the potential to become important crossroads for the transportation
of undocumented migrants. A Kennan Institute study
(Kennan Institute 2004) focusing on nontraditional immigrants from
Asia and Africa identified a set of migrants heading for Western
Europe who took advantage of the relatively open immigration system
in Ukraine (at least before 1999). They entered both legally and
illegally, and hoped to stay a short time before crossing to Western
Europe. Some had been duped by traffickers who promised safe passage
to Western Europe and then dumped them in Ukraine. In this
case as well, migration decisions were greatly influenced by available
information from government, extended family, business ties, friends
who had studied in Ukraine during Soviet times, communities of
compatriots in Ukraine, and organizers of undocumented migration.
The majority of Chinese immigrants stated that they relied primarily
on small business owners and traders, individuals who were first to
take advantage of favorable conditions for entering Ukraine after the
breakup of the Soviet Union. Many such migrants had legalized their
01-ECA_Migration.qxd 11/10/06 11:58 AM Page 42
Overview of Migration Trends in Europe and Central Asia, 1990–2004 43
status and launched businesses, especially in the food industry and
trading at Kiev markets. In contrast, many African migrants were
informed about Ukraine as an “easy” transit country to Western
Europe by countrymen who had studied in Ukraine during Soviet
times.
Profiles of undocumented migrants demonstrate that young, middle-level
educated men are more likely to migrate illegally. Most
respondents to the Azerbaijan survey were between the ages of 18
and 34 and the majority had completed secondary or vocational
schools (with legal migrants having more education on average than
irregular migrants) and had worked as low-skilled workers. Among
legal migrants, men and women were about equally numerous,
whereas most irregular migrants were men (Economic and Social
Research Council 2002). The survey of undocumented transit
migrants in Ukraine found that about 15,000 such migrants, many
young Muslim men, may be located in Kiev. Many were married to
Ukrainian women. Two-thirds had a high level of education and had
lived in large cities or capitals in their home countries before migrating
(Kennan Institute 2004).
Despite the careful calculations made in decisions to migrate, the
migration process is long and difficult for most transit migrants. Those
interviewed in Azerbaijan had all spent at least one year there, and
most were uncertain how much longer they would stay in transit.
Few expected to depart for their final destinations within the next
year and 11 percent had decided to stay in Azerbaijan if possible.
Transit migrants faced a number of difficulties—including shortages
of finance, unemployment, poor access to housing and health care,
and language barriers—yet were largely satisfied with the overall attitudes
of government officials and the local population. More irregular
migrants had employment in Azerbaijan than did legal migrants
(Economic and Social Research Council 2002).
A major factor inhibiting the further movement of so-called transit
migrants was their lack of information. The intended final destinations
of most irregulars were the United States, Canada, and Western
Europe, whereas most legal migrants intended either to return home
(especially to Russia) or to continue on to Western Europe. Most were
poorly informed about the rules and regulations for entry to their
planned destination countries and living conditions there. Furthermore,
illegal migrants who intended to return home were often
dependent on outside assistance to do so. Most legal migrants planned
to leave Azerbaijan on their own, while most irregular migrants were
hoping for assistance from humanitarian organizations, travel agencies,
and middlemen (Economic and Social Research Council 2002).
01-ECA_Migration.qxd 11/10/06 11:58 AM Page 43
44 Migration and Remittances: Eastern Europe and the Former Soviet Union
Thus, clearly the migration experience is substantially influenced by
the legal status of those who undertake it.
ECA Country Experiences and Policies
Undocumented immigration is by definition difficult to quantify. Currently,
there are estimated to be upward of 3 million undocumented
immigrants in the EU, and between 1,300 and 1,500 in Russia. The
International Organization for Migration reports that “99 percent of
labor migration in the Eurasian Economic Union formed of Tajikistan,
Kyrgyz Republic, Kazakhstan, the Russian Federation, and Belarus is
irregular. Due to their irregular situation, most labor migrants do not
benefit from the same protection rights other regular citizens enjoy
and are thus more vulnerable to exploitation by underground
employers” (IOM 2001, p. 11). Legal status not only affects the relative
migration costs and expected benefits, but also changes the
underlying economic incentives. Table 1.3 provides a range of estimates
of undocumented migration in selected ECA countries, Western
Europe, and the United States.
ECA countries act as source, host, and transit countries for undocumented
migrants. The concerns associated with the illicit movement,
transit, and trade in people are therefore salient across the region. The
major host is Russia, most of whose undocumented workers are from
the rest of the CIS. However, following accession of the EU-8 to the EU,
undocumented migration from western CIS, Russia, the Balkans, and
Turkey is becoming an increasing issue for the EU-8 and other countries
along its borders. Demographic change is generating a demand for workers
in certain sectors and regions, while other migrants are becoming
“stuck” as they fail to cross the EU-15 borders. The status of the EU-8 is
in transition, but the slowdown in westward emigration in most countries,
as well as the opening up of labor markets in some parts of the EU-
15, is increasingly regularizing flows. In fact, the expansion of the
Schengen Agreement to cover the EU-8 is extending the problem eastward,
as irregular migrants are now becoming stuck in the Ukraine.7
Turkey hosts a number of undocumented workers mainly from ECA,
but also from the Middle East. Taking into account these factors and the
role of the ECA as the main overland route to Western Europe, the
whole of the region is a major transit route. Transit migrants may come
from the region itself, or from the Middle East, Africa, or Asia. It is
thought that of the 500,000 trafficked women in Eastern Europe,
300,000 originated in or were transported through the Balkans.
The growth of undocumented migration in the ECA region may be
closely tied to the migration policies used to regulate it, and particu-
01-ECA_Migration.qxd 11/10/06 11:58 AM Page 44
Overview of Migration Trends in Europe and Central Asia, 1990–2004 45
larly policies in the EU-15 that cap supply of labor below demand.
The flow of labor under existing migration agreements is regulated
through quotas, as well as a maximum residency period allowed in
the receiving country. Quotas often appear small both in relation to
the perceived need for labor and in relation to the actual flow of labor
migrants. Thus, for instance, Jandl (2003) notes that while 1.11 million
foreigners had valid residence permits in Spain in 2000, the 2001
census counted 1.57 million foreigners and the Organisation for EcoTABLE
1.3
Estimated Irregular Migrants
(thousands)
Total Estimated number of Average %
number of irregular migrants Year of of total
Country migrants Max Min estimation migrants
North America and Canada
United States 34,988 10,300 — 2004 29.44
Canada 5,826 200 100 2003 3.43
High-income Europe
Greece 534 320 — 2003 59.87
Portugal 233 100 — 2003 42.96
Italy 1,634 500 — 2003 30.59
United Kingdom 4,029 1,000 — 2003 24.82
Spain 1,259 280 — 2003 22.24
Belgium 879 150 — 2003 17.06
Germany 7,349 1,000 — 2003 13.61
Switzerland 1,801 180 — 2003 9.99
Netherlands 1,576 163 112 2003 8.72
France 6,277 400 — 2003 6.37
Ireland 310 10 — 2003 3.23
Finland 134 1 — 2003 0.75
Total 26,015 4,104 — 15.78
ECA countries
Poland 2,088 600 — 2000 28.73
Ukraine 6,947 1,600 — 2000 23.03
Tajikistan 330 60 — 2002 18.16
Czech Republic 236 40 — 2003 16.98
Slovak Republic 51 8 — 1998 15.69
Turkey 1,503 200 — 2001 13.31
Russia 13,259 1,500 1,300 2000 11.31
Kazakhstan 3,028 300 220 2002 9.91
Belarus 1,284 150 50 2000 11.68
Kyrgyz Republic 572 30 — 1998 5.24
Uzbekistan 1,367 30 — 2000 2.19
Lithuania 339 2 — 1997 0.59
Sources: Pew Hispanic Center; IOM; ILO; World Bank; ISTAT; Home Office in United Kingdom; Jimenez (2003); Center on
Migration, Policy and Society of the University of Oxford; EU Business Council of Europe; Ministry of Labor in Finland;
Sadovskaya (2002); Migration Policy Group; Jandl (2003).
Note: — = not available. Estimation methods are different for each country. Total number of migrants is at the point in
2000 and is estimated by UN (2003).
01-ECA_Migration.qxd 11/10/06 11:58 AM Page 45
46 Migration and Remittances: Eastern Europe and the Former Soviet Union
nomic Co-operation and Development (OECD) (OECD 2005) estimates
that roughly 1 million irregular migrants (around 6 percent of
the labor force) will be affected by the recent amnesty. In the United
Kingdom, Migration Watch estimates that the number of irregular
migrants—including disappeared asylum seekers, visa overstayers,
and clandestine entries—is over 100,000 a year; other sources put the
figure as high as 500,000.8 Jandl (2003) estimates that the stock of
irregular migrants in Europe is somewhere between 2.6 million and
6.4 million and the annual number of border apprehensions in EU-15
is close to 300,000.
In light of these numbers, and assuming that most clandestine
migrants succeed in finding work, the quotas for labor migration in
the bilateral agreements between EU-15 and Central Europe and the
Balkans are very small. For example, the Italian agreement on seasonal
migration concluded in 1997 with Albania allows 3,000
migrants a year; Germany’s quota for guest workers is 15,500 a year
(though there are approximately 200,000 seasonal agricultural workers),
and the United Kingdom allows an annual inflow of 25,000 from
all countries outside of the European Economic Area (OECD 2004).
Between the time of EU enlargement in May 2004 and November
2005, there has been an inflow of 156,165 workers from the EU into
the United Kingdom and 107,024 into Ireland. Through December
2004, there was a flow of 3,514 workers into Sweden.
Major Migration Partners of the ECA Countries
An important aspect of migration management is understanding the
patterns of migration for any particular country. Such an exercise is
similar to investigating a country’s major foreign trade partners,
though usually fewer countries are major senders and receivers of
migrants to any particular country than are significant trade partners.
Furthermore, the problems with obtaining migration data in many
countries in the region make this a somewhat inexact exercise. Fortunately,
the largest country in the region, Russia, which is also the
main migration partner of most of the other former Soviet Union
(FSU) states, has a fairly complete set of migration data, although it
does not include the undocumented migrants in the country. Figure
1.11 shows that Russia has been a net recipient of migration from all
of the other FSU states except for Belarus, and a net sender to the “far
abroad” or to countries outside of the FSU (data underlying these figures
are in table 1.6 of appendix 1). The countries from which Russia
has received the largest numbers of migrants are those from which
01-ECA_Migration.qxd 11/10/06 11:58 AM Page 46
Overview of Migration Trends in Europe and Central Asia, 1990–2004 47
there has been a large return of ethnic Russians—Kazakhstan,
Ukraine, and Uzbekistan. However, since 1994, there has been a net
immigration to Russia of many other nationalities. If undocumented
migrants were included, the numbers representing non-Russians
would be even larger.
Three countries outside Russia are the primary destinations for
Russian migrants: Germany, Israel, and the United States. Those who
migrate consist primarily of Germans, Jews, and Russians, reflecting a
combination of ethnic and economic factors driving their decisions to
migrate.
The trends shown in the data from Belarus, Moldova, and Ukraine
(see figure 1.12) are roughly consistent with the data that appear in
the data from Russia. Ukraine had net migration losses to Russia
while Belarus overall gained migrants. Moldova had net overall losses
and net migration losses to the FSU countries, though it did gain
migrants from all FSU countries except Russia, Ukraine, and Belarus.
All three of these countries are net recipients of migrants from all of
the other FSU states. As was the case for Russia, the same three countries
outside the FSU—Germany, Israel, and the United States—are
the primary destinations of migrants from Ukraine, Belarus, and
Moldova. There is anecdotal evidence that an increasing number of
FIGURE 1.11
Russia, Net Migration by Country, 1989–2003
Ϫ1,000 Ϫ500 0 500 1,000 1,500 2,000
net migration (thousands)
Migration losses to all
states outside FSU
Migration gains from all FSU states
(except Belarus)
Germany
Israel
United States
Other
Belarus
Canada
Finland
Australia
Poland
Sweden
Lithuania
Estonia
Moldova
Latvia
Turkmenistan
Armenia
Kyrgyz Republic
Ukraine
Tajikistan
Azerbaijan
Georgia
Uzbekistan
Kazakhstan
Source: Goskomstat Rossii (selected publications).
01-ECA_Migration.qxd 11/10/06 11:58 AM Page 47
48 Migration and Remittances: Eastern Europe and the Former Soviet Union
labor migrants from Ukraine and Moldova are departing for the countries
of Western Europe.
For the three Baltic states (Latvia, Lithuania, and Estonia), mainly
titular members of these states have migrated to Russia and the other
Slavic states. The data do not demonstrate the fact that this ethnic
migration peaked in 1992–93, just after the breakup of the Soviet
Union, or that it has declined substantially since then as Russians and
other minorities in the Baltics have remained as a result of faster
growing economies and impending EU membership. As in other FSU
countries, Germany, Israel, and the United States are the primary destinations
for migrants from the Baltic states to countries outside the
FIGURE 1.12
Major Migration Partners of the CIS Countries
Ukraine, Belarus, and Moldova: net migration by country
Ϫ250 Ϫ200 Ϫ150 Ϫ100 Ϫ50 0 50 100 150
Russian Fed.
Ukraine
Belarus
Moldova
Latvia
Lithuania
Estonia
Armenia
Azerbaijan
Georgia
Kazakhstan
Kyrgyz Rep.
Tajikistan
Turkmenistan
Uzbekistan
Germany
Israel
United States
Other
net migration (thousands) net migration (thousands)
net migration (thousands) net migration (thousands)
Ukraine Belarus Moldova
Baltics: net migration by country
Ϫ90 Ϫ80 Ϫ70 Ϫ60 Ϫ50 Ϫ40 Ϫ30 Ϫ20 Ϫ10 0 10
Russian Fed.
Ukraine
Belarus
Moldova
Latvia
Lithuania
Estonia
Germany
Israel
United States
Poland
Finland
Other
Latvia Lithuania Estonia
Transcaucasus states: net migration by country
Ϫ300 Ϫ250 Ϫ200 Ϫ150 Ϫ100 Ϫ50 0 50 100
Russian Fed.
Ukraine
Belarus
Moldova
Armenia
Azerbaijan
Georgia
Kazakhstan
Kyrgyz Rep.
Tajikistan
Turkmenistan
Uzbekistan
Germany
Israel
United States
Armenia Azerbaijan Georgia
Central Asia: net migration by country
Ϫ1,200 Ϫ1,000 Ϫ800 Ϫ600 Ϫ400 Ϫ200 0 200
Russian Fed.
Ukraine
Belarus
Azerbaijan
Georgia
Kazakhstan
Kyrgyz Rep.
Tajikistan
Turkmenistan
Uzbekistan
Germany
Israel
United States
Other
Kazakhstan Kyrgyz Rep. Tajikistan
Turkmenistan Uzbekistan
Source: National statistical offices of the ECA countries.
01-ECA_Migration.qxd 11/10/06 11:58 AM Page 48
Overview of Migration Trends in Europe and Central Asia, 1990–2004 49
FSU, although there may have been a broader dispersion of destinations
after these states became EU members in 2004.
For the three Caucasus countries (Armenia, Azerbaijan, and Georgia),
Russia has been the dominant migration destination. There is
considerable evidence that these figures represent only a fraction of a
much larger undocumented and circular migration from these countries
to Russia. This is especially the case with Georgia, where the data
on net migration by country only cover the period 1990 to 1992. In
contrast, the 2002 population census in Georgia revealed a net migration
loss of 1.1 million persons or 20 percent of the population. The
migration of Armenians from Nagorno-Karabakh and the surrounding
regions in Azerbaijan is shown in this data set, although such
movement was confined to the early 1990s. The United States is the
primary destination outside the FSU for migrants from Armenia, with
most of these joining the already large Armenian diaspora community
there, while Israel remains a top Azerbaijani destination.
For the five Central Asian countries (Kazakhstan, the Kyrgyz
Republic, Tajikistan, Turkmenistan, and Uzbekistan), Russia again
dominates as a migration destination, as migration turnover to other
FSU states is rather minimal. There is, however, some tentative evidence
that Kazakhstan is becoming a favored migration destination
for persons from the other Central Asian countries. From both Kazakhstan
and the Kyrgyz Republic, there were large migrations of ethnic
Germans to Germany. From Kazakhstan, over 800,000 Germans
left and from the Kyrgyz Republic, nearly 100,000. These movements
were the remnants of both voluntary and forced migrations of Germans
to Central Asia during the Soviet period.
Figure 1.13 shows the major migration patterns of the largest western
ECA country, Poland. As can be seen, Poland is losing people to
many developed countries (albeit to varying degrees) and remains a
net emigration country. Its largest losses are to neighboring Germany,
the United States, and Canada, where there are already large Polish
diaspora populations as a result of past migrations. The figure for Germany
is likely an underestimate because many Poles can travel rather
easily to Germany. This figure encompasses the period before Poland
became an EU member and thus does not include Poles working in
the United Kingdom, Ireland, and Sweden. Many of them would not
likely be included in these totals, because such labor migrants generally
do not view their departure from Poland to be permanent.
Figure 1.14 provides data on the main migration partners of Hungary,
Romania, the Czech Republic, and the Slovak Republic. According
to these data, Hungary is a net recipient of migrants from nearly all
listed countries, with especially large numbers coming from Romania,
01-ECA_Migration.qxd 11/10/06 11:58 AM Page 49
50 Migration and Remittances: Eastern Europe and the Former Soviet Union
Yugoslavia, and other countries that housed ethnic Hungarians after
present-day Hungary was carved out of the Austro-Hungarian Empire.
Romania shows population losses to nearly every other country, with
especially large losses to Germany, where many Romanians have gone
for work. The only country from which Romania is gaining migrants
is its close ethnic neighbor, Moldova. The Czech Republic has been a
net recipient of people from other countries, with the bulk of in-migration
coming from the Slovak Republic (which had been a part of
Czechoslovakia until 1993). The Slovak Republic itself is a net recipient
from all listed countries except the Czech Republic.
Future Migration Trends in the Region
One of the themes of this report is that both economic and demographic
incentives affect the motivation to migrate for ECA and
neighboring countries. This section describes the demographic implications
for future migration flows in this region.
Future Migration Patterns in the EU and Neighboring
Countries
A combination of income convergence and demographic change suggests
that the potential for large-scale migration from western ECA to
FIGURE 1.13
Poland: Net Migration by Country, 1992–2003
Ϫ200,000 Ϫ150,000 Ϫ100,000 Ϫ500,00 0 50,000
Australia
Austria
Belgium
Canada
Czech Rep.
Denmark
Finland
Germany
Hungary
Greece
Italy
Netherlands
Norway
Russian Fed.
Sweden
United Kingdom
United States
net migration
Source: Migration Policy Institute; OECD SOPEMI 2003; and German Federal Statistical Office.
01-ECA_Migration.qxd 11/10/06 11:58 AM Page 50
Overview of Migration Trends in Europe and Central Asia, 1990–2004 51
the EU and other neighboring countries is limited. The richest countries
in western ECA have already begun to be net immigration countries.
This suggests that the experience of most Western European
countries that are net recipients of migrants is likely to become the
norm in most western ECA countries with income convergence and
EU membership. Even with no convergence, changes in migration
patterns appear inevitable.
FIGURE 1.14
Major Migration Partners of Selected Western ECA Countries
Czech Republic, net migration by country, 1988–98
0 5 10 15 20 25 30 35
Slovakia
Ukraine
Germany
Russia
United States
Canada
Poland
Yugoslavia
Bulgaria
Austria
Romania
Others
Slovakia, net migration by country, 1988–98
Ϫ10 Ϫ8 Ϫ6 Ϫ4 Ϫ2 0 2 4 6
Ukraine
Yugoslavia
Romania
Russia
United States
Poland
Bulgaria
Germany
Canada
Austria
Others
Czech Rep.
Hungary, net migration by country, 1989–98
0 10 20 30 40 50 60 70 80 90
Bulgaria
Czech Rep.
Slovakia
Romania
Russia
Ukraine
Uzbekistan
Croatia
Yugoslavia
Austria
France
Germany
Greece
Italy
United Kingdom
Israel
United States
Others
net migration (thousands)
net migration (thousands)
net migration (thousands)
net migration (thousands)
Romania, net migration by country, 1990–98
Ϫ160 Ϫ140Ϫ120Ϫ100Ϫ80Ϫ60 Ϫ40 Ϫ20 0 20 40
Moldova
Ukraine
Slovakia
Yugoslavia
Czech Rep.
Denmark
United Kingdom
Switzerland
Greece
Australia
Belgium
Sweden
Israel
France
Others
Italy
Canada
Austria
United States
Hungary
Germany
Source: Walmsley, Ahmed, and Parsons (2005).
01-ECA_Migration.qxd 11/10/06 11:58 AM Page 51
52 Migration and Remittances: Eastern Europe and the Former Soviet Union
With the exception of Albania, all western ECA countries are forecast
to experience population declines between now and 2050. The
total population of these countries peaked in 1990 at 130 million and
is projected to decline by 19 percent to 104 million by mid-century.
As shown in figure 1.15, western ECA source countries are often projected
to have larger population declines than those in Western
Europe. The population of Western Europe is expected to increase
from its current size of 397 million to a peak of 407 million in 2030
before declining to 400 million in 2050. For western ECA, a decline in
the working-age population and a corresponding increase in those
over age 65 will create a demand for workers from abroad. The more
prosperous western ECA countries may be able to source some of
these workers from the rest of the region. However, for the region as
a whole, demand will have to be met from elsewhere, probably CIS,
Africa, and Asia. Whether these flows are legal or undocumented will
depend on immigration legislation.
While the total population of Western Europe is expected to rise
slightly between now and mid-century as a result of the current age
structure of these countries and expected demographic trends, the
working-age population in these countries is expected to decline substantially.
Of course, the largest variable in future European migration
FIGURE 1.15
Population Size of Western Europe, Western ECA, and Turkey, 1950 to 2050
0
50
100
150
200
250
300
350
400
450
Western Europe Western ECA Turkey
1950 1955 1960 1965 1970 1975 1980 1985 1990 1995 2000 2005 2010 2015 2020 2025 2030 2035 2040 2045 2050
population(millions)
Source: United Nations Population Division, World Population Prospects: The 2004 Revision (http://www.un.org/esa/population/unpop.html).
01-ECA_Migration.qxd 11/10/06 11:58 AM Page 52
Overview of Migration Trends in Europe and Central Asia, 1990–2004 53
patterns in both Western Europe and western ECA is Turkey, in which
most of the future population growth and additions to the labor force
in Europe are expected to take place. Because of its younger age structure
and higher fertility rates, Turkey is expected to grow by 33 million
between now and 2050 to a total of 101 million, nearly the size of the
other western ECA countries combined. Turkey, with an increase of 16
million in its working-age population, could produce sufficient migration
to cover the 12 million person population deficit in the EU.
Future Migration Patterns in the Former Soviet Union
Economic factors such as differences in per capita income drive migration
patterns among the post-Soviet states in the short term. These
will continue to be important, but demographic factors also will play
an important role. Figure 1.16 shows the population and expected
population of the FSU states over the period 1950–2050. The countries
are grouped into the northern FSU—the Slavic and Baltic states
and Moldova, and the southern FSU states—Central Asia and the
Caucasus. The northern states as a group are characterized by continued
low fertility, aging populations, an excess of deaths over births,
and declining populations. The group’s population peaked in 1990
and is expected to decline over the next half century by about onethird
to 149 million. By contrast, the southern FSU states have
younger populations, above replacement-level fertility, and continued
growing populations. As a group, these countries nearly tripled in
size, from 25 million in 1950 to 72 million in 2000. While growth is
declining, the momentum built into the age structure of these populations
will cause their continued growth to 93 million in 2050.
Differential rates of population growth (or decline) do not necessarily
imply that there will be migration from the high-growth to
low-growth areas but do present a precondition to that effect. While
the northern FSU states will have declining working-age populations
in even greater numbers than their overall population declines, most
of the southern FSU states, with their “youth bulges,” will have growing
working-age populations with economies not growing fast
enough to supply jobs. Given their geographic proximity and common
historical legacy, it would be only natural that the youth of the
southern FSU would look north for jobs, and as mentioned above,
there is ample evidence that they are doing so. Furthermore, historical
legacy contributes to the selection of migration destinations. The
Soviet Union was an almost self-contained migration space; the interconnectedness
of FSU countries may cause people to favor destinations
in that area over others.
01-ECA_Migration.qxd 11/10/06 11:58 AM Page 53
54 Migration and Remittances: Eastern Europe and the Former Soviet Union
A recent United Nations study examined the issue of using
“replacement migration” as a policy measure to address declining and
aging populations.9 The EU and Russia were included in the study, as
were other countries—including France, Germany, Italy, Japan, the
Republic of Korea, the United Kingdom, and the United States—that
face similar trends of declining and aging populations. The population
declines projected by 2050 in these countries range from 17 percent
(Moldova) to 52 percent (Estonia). Countries with aging and declining
populations face a number of policy dilemmas, including appropriate
retirement ages, pension system reform, and health care for the
elderly; support levels and ratios between working and pension-age
populations; labor force participation; and possible replacement
migration and the integration of immigrant populations. In contrast
to these other possibilities, replacement migration refers to the principle
of using international migration to offset declines in total population,
working-age population, or population aging.
Figure 1.17 shows the combination of natural increase (the difference
between births and deaths) and net migration for Russia for the
period 1980–2015. During the 1980s, Russia’s population was growing
as a result of both demographic and migratory factors. Starting in
1992 and expected to continue for the foreseeable future, the number
of deaths has exceeded the number of births. Migration into RusFIGURE
1.16
Population Size of the Northern and Southern FSU States, 1950 to 2050
1950 1955 1960 1965 1970 1975 1980 1985 1990 1995 2000 2005 2010 2015 2020 2025 2030 2035 2040 2045 2050
population(millions)
0
50,000
100,000
150,000
200,000
250,000
Southern FSUNorthern FSU
Source: United Nations Population Division, World Population Prospects: The 2002 Revision Population Database (http://www.un.org/esa/population/unpop.htm).
Note: The northern FSU consists of Russia, Ukraine, Belarus, Moldova, Latvia, Lithuania, and Estonia. The southern FSU consists of Armenia, Azerbaijan, Georgia,
Kazakhstan, the Kyrgyz Republic, Tajikistan, Turkmenistan, and Uzbekistan.
01-ECA_Migration.qxd 11/10/06 11:58 AM Page 54
Overview of Migration Trends in Europe and Central Asia, 1990–2004 55
sia spiked sharply in the 1990s following the breakup of the Soviet
Union and has declined sharply since then (at least documented
migration). If these trends continue, Russia’s population will decline
and age rapidly. For Russia to maintain the size of its total and working-age
populations, allowing migration seems to be the only policy
option.
Under the medium-variant scenario used in the study, the EU is
projected to have a net migration of 13.5 million and Russia to have
a net migration of 5.4 million between 2000 and 2050. To maintain
the population size as it was in 1995 using migration alone would
require a net migration of 47.9 million into the EU and 24.9 million
into Russia during that period. Maintaining the same size workingage
population would require a net migration of 79 million into the
EU and 35.8 million into Russia. For comparison’s sake, there was a
net migration of about 8.8 million into the EU and about 3.3 million
into Russia during the 1990s. Furthermore, for Russia this was a
period of extraordinary change and unprecedented migration that is
not likely to be repeated.
FIGURE 1.17
Russia: Net Migration and Natural Increase, 1980–2015
Ϫ1,500
Ϫ1,000
Ϫ500
0
500
1,000
1,500
1980 1985 1990 1995 2000 2005 2010 2015 2020 2025
natural increase net migration
thousands
Source: Goskomstat Rossii.
Note: Data are actuals from 1980 to 2003 and projected from 2004 to 2015.
01-ECA_Migration.qxd 11/10/06 11:58 AM Page 55
56 Migration and Remittances: Eastern Europe and the Former Soviet Union
For the EU, Russia, and the other large aging and declining populations
in the UN study, it is obvious that the needed replacement
migration levels are far above levels that are politically and socially
plausible. Even low levels of migration will require very careful political
and social balancing acts in Russia, the other northern FSU countries,
and other major migration destinations. Policies must be
designed to accommodate these new migration realities in both destination
and originating countries, and, most importantly, the dynamic
fluctuations between the two. There is evidence that Russia and some
of the other FSU states are facing up to this new migration reality in
the region and taking steps to regularize it.
Endnotes
1. Much of the migration data upon which this chapter is based is contained
in appendix 1.
2. Estimates as of March 2004 are that there are 10.3 million undocumented
migrants in the United States and each year another 700,000 to
800,000 unauthorized enter the country, which is about the same size as
those who migrate legally to the United States (Passel 2005).
3. The figure for Armenia, which includes those not indicating their place
of birth, is likely a large overestimate because of the problems with the
census, which was conducted in January 1989, just after the devastating
earthquake in December 1988.
4. For more on the fertility decline in the ECA region, see Heleniak (2005).
5. Turkmenistan and Bosnia and Herzegovina are not included because of
the suspected migration data problems mentioned above.
6. To ensure comparability, the data are taken from one source, UNICEF’s
TransMONEE database, which collects data from the national statistical
offices of the 27 transition ECA countries, not including Turkey.
7. The Schengen Agreement originally was a state treaty to end internal
border checkpoints and controls among European countries. Today the
Schengen system is part of EU legislation regulating border control, visa
and admission and nonadmission standards, as well as the joint Schengen
Information System. The 15 current Schengen countries include
Austria, Belgium, Denmark, Finland, France, Germany, Iceland, Italy,
Greece, Luxembourg, the Netherlands, Norway, Portugal, Spain, and
Sweden. All these countries except Norway and Iceland are EU members.
The name “Schengen” originates from the small town in Luxembourg
where the agreement was signed in 1985.
8. Data from the U.K. Home Office. Source at http://www.timesonline.
co.uk/article/0,,2087-1572533,00.html, retrieved June 22, 2005.
9. United Nations Population Division 2001. The study uses the 1998 Revision
of UN population projections as a baseline. The European Union
defined in the report was the EU-15.
01-ECA_Migration.qxd 11/10/06 11:58 AM Page 56