BookPDF Available

Demographic scenarios for the EU: Migration, population and education

Authors:
2060
EUR 29739 EN
DEMOGRAPHIC
SCENARIOS FOR THE EU
MIGRATION, POPULATION AND EDUCATION
This publication is a Science for Policy report by the Joint Research Centre (JRC), the European Commission’s
science and knowledge service, written in partnership with the International Institute for Applied Systems Analysis
(IIASA). It aims to provide evidence-based scientic support to the European policymaking process. The scientic
output expressed does not imply a policy position of the European Commission or of IIASA. Neither the European
Commission, IIASA nor any person acting on behalf of the Commission and IIASA is responsible for the use that
might be made of this publication.
Manuscript completed in April 2019
Contact information
IIASA
Wolfgang Lutz (Ed.)
International Institute for Applied Systems Analysis (IIASA), Schlossplatz 1, A-2361 Laxenburg, Austria
Email: lutz@iiasa.ac.at
Tel. +43(0) 2236807294
JRC
Delilah Al Khudhairy
European Commission, Joint Research Centre, Rue du Champ de Mars 21, 1050 Brussels, Belgium
Email: Delilah.AL-KHUDHAIRY@ec.europa.eu
Tel. +32 22999158
EU Science Hub
https://ec.europa.eu/jrc
JRC116398
EUR 29739 EN
PDF ISBN 978-92-76-03216-8 ISSN 1831-9424 doi:10.2760/590301
Print ISBN 978-92-76-03215-1 ISSN 1018-5593 doi:10.2760/751889
Luxembourg: Publications Oce of the European Union, 2019
© European Union, 2019
The reuse policy of the European Commission is implemented by Commission Decision 2011/833/EU of 12 December
2011 on the reuse of Commission documents (OJ L 330, 14.12.2011, p. 39). Reuse is authorised, provided the source
of the document is acknowledged and its original meaning or message is not distorted. The European Commission
shall not be liable for any consequence stemming from the reuse. For any use or reproduction of photos or other
material that is not owned by the EU, permission must be sought directly from the copyright holders.
The designations employed and the presentation of material on the maps do not imply the expression of any opinion
whatsoever on the part of the European Union concerning the legal status of any country, territory, city or area
or of its authorities, or concerning the delimitation of its frontiers or boundaries.
All content ©European Union, 2019 except: cover ©European Union, 2019 - graphic elaboration from ©hobbitfoot
2019 - AdobeStock, page 7, ©digitalmagus, 2019 - AdobeStock; page 9, ©Denys Argyriou, 2019 - Unsplash; page 11,
©blvdone, 2019 - AdobeStock; page 18, ©Africa Studio, 2019 - AdobeStock; page 24, ©everythingpossible,
2019 - AdobeStock; page 27, ©scharfsinn86, 2019 - AdobeStock; page 31, ©jchizhe, 2019 - AdobeStock;
page 35, ©Jan Kranendonk, 2019 - AdobeStock; page 36, ©anyaberkut, 2019 - AdobeStock; page 43, ©srckomkrit,
2019 - AdobeStock; page 44, ©freshidea, 2019 - AdobeStock; page 50, ©tobago77, 2019 - AdobeStock;
page 55, ©upixa, 2019 - AdobeStock; page 58, ©thakala, 2019 - AdobeStock; page 63, ©africa, 2019 - AdobeStock;
page 66, ©Richtsteiger, 2019 - AdobeStock; page 72, ©steendia, 2019 - AdobeStock; page 74, ©piyaset,
2019 – AdobeStock.
How to cite this report: Lutz W. (Ed.), Amran G., Bélanger A., Conte A., Gailey N., Ghio D., Grapsa E., Jensen K.,
Loichinger E., Marois G., Muttarak R., Potančoková M., Sabourin P., Stonawski M., Demographic Scenarios for the EU
- Migration, Population and Education, EUR 29739 EN, Publications Oce, Luxembourg, 2019, ISBN 978-92-76-
03216-8, doi:10.2760/590301, JRC116398.
DEMOGRAPHIC
SCENARIOS FOR THE EU
MIGRATION, POPULATION AND EDUCATION
4
Table of contents
TABLE
OF CONTENTS
Executive summary 6
Introduction 14
Part A: Demographic structures and migration in the EU 17
1 Population ageing in Europe 18
1.1 Population ageing and old-age dependency 19
1.2 The impact of the scenarios on the total EU population 20
1.3 The impact of the fertility scenarios on the
EU’s working-age population 22
1.4 The impact of the fertility scenarios on the
EU’s 65+ population 22
1.5 The impact of the high-immigration scenario on the
size of the EU’s working-age and 65+ populations 23
1.6 Conclusions 23
2 Size and education levels of the future EU labour force 24
2.1 The demographic transition 25
2.2 A smaller but more-educated labour force 25
2.3 How dierent rates of labour participation impact
on the size of the total labour force 26
2.4 How lower labour-force dependency ratios can reduce
the public spending burden 29
2.5 Conclusions 30
3 Impacts of migration on the EU labour force 36
3.1 Migration scenarios 37
3.2 Current labour-force participation of immigrants 37
3.3 The impact of dierent volumes of migrants on the size
of the future EU labour force 38
3.4 The impact of dierent volumes of immigrants on
the educational composition of the future EU labour force 40
3.5 The impact of dierent volumes of migration
on the labour-force dependency ratio 41
3.6 Conclusions 42
4 Impacts of internal EU mobility over time 44
4.1 Intra-EU mobility and EU Member State populations 45
5 Table of contents
4.2 Intra-EU mobility scenarios 46
4.3 The impact of intra-EU mobility on the EU population 47
4.4 Conclusions 49
5 Demographic consequences of ‘brain drain’ 50
5.1 The hypothetical scenario of an EU ‘brain drain’ 51
5.2 Consequences of the high-emigration scenario 52
5.3 Conclusions 54
Part B: Demographic trends and migration
in Africa and Western Asia 57
6 World population growth trajectories 58
6.1 The projected population growth of Africa 59
6.2 How female education leads to lower fertility rates 61
6.3
How the expansion of girls' education determines
the future population growth in Africa
62
6.4 Conclusions 65
7 Reasons for migration in Africa and Western Asia 66
7.1 Conict and instability as push-factors 67
7. 2 Youth unemployment as a driver of outmigration 69
7. 3 Conclusions 71
8 Evidence for climate-related migration 72
8.1
The relationship between environmental change
and migration 73
8.2 Conclusions 75
Report summary and conclusions 76
The way forward for demographic analysis 77
Glossary 81
Endnotes 82
References 85
List of boxes and gures 88
List of tables and maps 89
Acknowledgements 90
6
Executive summary
EXECUTIVE
SUMMARY
At the very heart of a changing society lies
the number and composition of its members.
Population growth has shaped the EU over
recent decades and now its population is ageing.
The slow-moving shi towards longer-living,
lower-fertility, higher-educated societies brings
the EU to new demographic frontiers, as it does
in North America and East Asia.
Facing these developments naturally prompts
the questions: Who will live and work in Europe
in the coming decades? How many, and with what
skills? To answer these, we consider key factors
that will inuence European demographics over
the coming decades.
To move beyond some common misconceptions,
this report brings relevant scientic analyses
to the forefront by presenting a series of
demographic scenarios for the future of the EU.
By examining not only the role of migration, fertility
and mortality, but also education and labour force
participation, we can outline a more comprehensive
view of possible futures than conventional
demographic projections.
While some of the scenarios indicate probable
developments, others are hypothetical and
meant to be instructive for understanding the full
spectrum of possible futures. The value of these
scenarios is to improve our ability to anticipate
the coming changes and to guide our responses
to them.
The following key messages express the core
ndings from the work of the Centre of Expertise
on Population and Migration.
We are living longer and can lead
more productive lives
One of the most important trends in the EU
is ageing. Thanks to advances in medicine and
quality of life, the average life expectancy at birth
in the EU is about 81 years. Shis to longer lives
come as part of the larger, nearly universal process
of transitioning from pre-modern to post-industrial
demographic patterns.
While it is an important achievement that we are
on average living longer, an increasing proportion of
people aged 65+ can also bring social and economic
challenges. There are concerns about how to provide
for the growing proportion of retired citizens who
are dependent on a smaller labour force. Yet, absent
from the oen supercial discussions are reections
on the increasing number of years of active and
healthy life, greater productivity, and possible loss
of jobs due to technological advancements. All of
which have profound implications on what societies
will require from their future working populations.
Automation alone may signicantly cut the need
for labour, calling into question conventional
assumptions about what policymakers should target
as the ‘ideal’ size of the labour force.
It is important to acknowledge that neither higher
fertility nor more migration will stop population
ageing because the momentum has been decisively
set by past, long-term demographic trends.
However, reaching the age of 65 is no longer
synonymous with the end of a person’s productive
life. Overall, we are living more productive lives
which, together with exible retirement ages,
can help to alleviate the economic challenges
arising from changes in the conventionally dened
7 Executive summary
working-age population and the increasing
ratio of workers to non-workers (known as the
dependency ratio).
A smaller, better-educated labour
force on the horizon
Another reason for optimism is the fact that
the EU’s labour force is clearly becoming better
educated over time. In almost all scenarios, the
EU’s future labour force will be more highly skilled
and therefore likely to be more productive and
adaptive to an ever-changing job market.
The EU labour force is transforming through social
development and the ageing process. The total size
of the labour force is projected to get smaller over
the next four decades, based on the assumption
that the current patterns continue. According to this
assumption, the projected decrease in the size
of the labour force will come from the population
with low (from 50.7 to 14.0 million) and middle
(from 108.2 to 74.2 million) levels of education.
At the same time, the number of workers with
a short post-secondary education (such as
technical training), a bachelor’s, a master’s
degree or higher, is rising, not just as regards their
proportion of the overall labour force, but also
in absolute terms. In every country, since the
younger cohorts are better educated than the older
ones, these post-secondary groups are expected
to almost double (+45%) over the next 40 years.
Considering the future context, any possible
reduction in the size of the labour force may
not be an economic problem if future jobs are
fewer in number and require more skills. Already
today unemployment is generally higher among the
lower skilled and there may be even fewer low skills
jobs in the future. Regardless of changes in the
population’s age structure and labour force size,
it can be expected that the human capital of future
workers of any age, measured by the highest level
of their educational attainment, will be higher than
it is today.
Labour force participation
as a remedy for the challenges
of population ageing
Using a set of demographic scenarios of the future
that highlight the impact of changes to a variety
of factors, it becomes clear that the most feasible
and eective remedy to negative consequences
of population ageing is neither focusing on higher
fertility nor more migration, but rather increasing
labour force participation.
Extending constant labour-force participation
rates into the future shows the likely path
to a smaller labour force and increasing
dependency ratio. Scenarios that deviate from
this course, following either 1) the equalisation
of labour-force participation between men and
women, or 2) a gradual convergence of all Member
States to the participation rates of men and women
already seen in Sweden today, demonstrate the
power that improving labour-participation rates
has to nullify potential increases in dependency.
In fact, the momentum is on the side
of increasing labour-force participation.
If it continues to grow in the future, the labour
force size and dependency ratio could stabilise
at current levels. In other words, a strong but
realistic increase – as is already the reality
in Sweden – of labour-force participation over
time could compensate for a large part of the
anticipated negative economic consequences
of population ageing in the future.
Higher immigration volume
would increase labour force size,
but much less the essential ratio
of workers-to-non-workers
To the extent the EU is a destination region
for international migration, immigration becomes
an inuential factor on demographic
developments. Policymakers regularly balance
migration policy with the best interests
of the Member State. Doing so while being mindful
of a long-term understanding of demography
is crucial. At the core of such considerations are
factors including: the volume of immigrants from
third countries entering the Member State, their
levels of education, how well they will integrate
into the labour market and society at large,
and how eective the Member State
is at enforcing their migration policy.
Migration levels can have a large inuence
on the total population size and the size of the
labour force. With no third country immigration,
the natural decline resulting from lower fertility
would bring the EU population to 466 million
by 2060, the level observed in the 1980s.
However, migration levels have a limited eect
on changing the EU’s age structure, in part
because migrants tend to settle for the long
term and to age just as the native population
does. As such, irrespective of various levels
of immigration, the ndings of this report show
an almost inevitable trend towards continued
demographic ageing in the EU.
While a high volume of immigration would increase
the overall size of the EU labour force, it would
have a limited impact on the proportion
of workers to non-workers in the long-run.
If higher immigration volumes were to coincide with
deteriorating economic integration of migrants,
it would actually result in a labour force situation
that is worse than with medium or low volumes
of immigration, which highlights the importance
of eective eorts for economic integration.
Westward movement inside the
EU has substantial impacts on the
population sizes of Member States
In recent decades, intra-EU mobility – the free
movement between EU Member States – has
facilitated population changes within the EU.
Over the past 25 years, some of the Eastern
European Member States have lost a large share
of their population through a combination of low
fertility and, most notably, sizeable emigration.
Greater
labour force
participation
has more impact
on alleviating
pressure from
ageing than
either fertility
or migration
8
Executive summary
Dramatically, Bulgaria and the Baltic States lost
between 16% and 26%.
Intra-EU mobility has the potential to produce large
population shis within the EU over time. If the
movements of recent years persist as they have,
the population of Romania would reduce from
19.9 million in 2015 to 13.8 million by 2060 (-30%
of the population). Conversely, the losses would be
less than half without intra-EU mobility (only -14%).
The receiving Member States rely on these ows to
help compensate for their own ageing populations,
but the eect on their total populations is more
limited because they are generally more populous.
Pre-existing economic disparities between Member
States have encouraged many citizens to search
for work in places other than their country of origin.
These developments have likely been to the economic
benet of the union as a whole, but not necessarily
for all sending Member States. This contributes to
slowing the convergence between Member States,
and impacts areas such as infrastructure, education
and even population ageing. This in turn has
implications for the goals of economic development
and Cohesion Policy, in particular when the movement
is disproportionately highly skilled workers educated
in the sending Member States.
Dierences in wages and living standards
continue to drive westward migration within
the EU. Targeting economic inequality between
Member States can encourage greater cohesion
and integration and can help those Member States
facing disproportionate population decline, a loss
of working-age population, brain drain and more
pronounced population ageing. Policies should
address practical, labour-force-oriented skills
and try to reverse the education selectivity of
emigration by oering competitive employment
opportunities to the highly skilled and possibly
facilitating return migration of some of the talent
that has le.
Losing large numbers
of highly skilled workers leads
to lower productive potential
and accelerated population ageing
The loss of talent to comparatively higher-
income countries continues to confront some
societies within the EU, and many others around
the globe. This has demographic implications,
whereby sending countries nd themselves with
a smaller and less-educated workforce. Such
changes would coincide with a more rapidly
ageing population because emigrants tend
to be early-career adults. The high emigration
of talent may also negatively impact innovation
and economic growth.
As an illustrative example of how high emigration
impacts demographics, we use a hypothetical
case of the EU falling far behind in future global
9 Executive summary
competition. By simply extending Spain’s nancial
crisis emigration rates into the future for the EU
as a whole, we can see how fast high emigration
depletes the working-age population of the EU –
50% by 2050 in this hypothetical case. While not
a scenario to be viewed as a likely future
for the EU as a whole, it helps demonstrate what
some countries are facing and how the currently
observed pattern of net migration inows should
not be automatically assumed to continue
in the long term.
The question of high emigration by educated
citizens, or ‘brain drain’, highlights the
interconnectedness between global migration ows
and local demographics. Therefore, to understand
the context of the EU’s future population,
it is also necessary to look beyond the borders
of the EU. Neighbouring regions such as Africa
and Western Asia face very dierent demographic
and migratory trends.
Actual realised migration to the EU will, in part,
depend on its attractiveness as a destination –
that is, the pull factors created by Member States.
At the same time, socio-economic development
in third countries may also increase the propensity
to undertake migration journeys. Creating
a system which does not undermine
the development of human capital in countries
of origin by making it harder for third countries
to retain their highest-skilled workers
is one of many dilemmas.
Pressure from continued population
growth in Africa and Western Asia
can lead to an increase in
push factors for migration
Africa’s population is projected to increase by a factor
of two to three over the coming decades, or possibly
even more in the event of stalled socio-economic
development. In addition, population growth in
Western Asia will also be signicant. For example,
the population of Afghanistan is likely to triple
and that of Pakistan to almost double. Such massive
projected population increases require consideration
of the future job opportunities and economic growth.
The peace and stability necessary to facilitate such
opportunities may also be challenged by population
growth. These challenges can contribute to creating
push factors for migration, that is, conditions that
drive people to emigrate from their home countries.
Conict and insecurity are central push factors.
Emigration is also linked to people deciding
to migrate because of a lack of opportunities to
thrive in their own country, for instance due to high
youth unemployment and too few appropriate jobs.
Similarly, climate change can inuence migration
by aecting other push factors of migration such
as political and economic conditions. For example,
climate change can be a possible trigger of conict,
which in turn would lead to migration. When direct
climate-change-related migration does occur,
we nd that it tends to be mostly intra-regional.
Regardless
of changes in the
size of the labour
force, the human
capital of future
workers in the EU
will be higher than
it is today
10
Executive summary
This is in line with the fact that intra-regional
migration, for example within Africa, makes up
by far the largest share of migration globally.
The prospect of climate change and the expected
population growth in Africa and Western Asia
are likely to present important challenges. Thus,
pursuing policies that enhance general resilience
and local employment opportunities and foster
stability and security are critical to building
sustainable alternatives to migration
in the countries of origin.
Girls’ education matters greatly
for the future of population growth
in Africa
While Africa and Western Asia are facing high
population growth, many regions of the world have
progressed towards late stages of the demographic
transition with low rates of both mortality and
fertility, notably East Asia, North America and
Europe. During the second half of the century,
world population may even eventually peak and
start to reduce slightly, depending largely on how
fast fertility levels in Africa fall to moderate levels.
Rapid population growth creates an urgent need
for expanding education in Africa. Education
expansion must keep pace with the pressure
coming from rapid population growth, as it holds
the key to accelerating the demographic transition
and bringing development successes within reach.
Achieving such goals depends, in particular,
on giving girls access to education, as education
and family planning are closely intertwined.
Education broadens horizons and helps bring
fertility into the realm of conscious choice for both
women and men. Evidence from educational sub-
populations within countries indicates that higher
living standards and decisions for moderate fertility
levels accompany higher education
and the associated wider range of life choices.
Demographic changes are long term and occur
at a steady and predictable pace. This provides
us with a unique opportunity for foresight
because it enables us to study how certain
trends are likely to impact the future populations
of EU Member States and the world. In this way,
the scenarios examined in this report help to
enable evidence-based planning for the future.
11 Executive summary
- 38 %
- 20 %
+ 7 %
- 1 %
- 30 %
- 14 %
+ 17 %
+ 5 %
- 37 %
- 23 %
+ 26 %
+ 21 %
2
0
1
5
2
0
1
5
2
0
1
5
AFRICA AFRICA AFRICA AFRICA
DEMOGRAPHIC SCENARIOS FOR THE EU
MIGRATION, POPULATION AND EDUCATION
How can we avoid overburdening
our social system with population ageing?
increase migration?
increase fertility?
2060
or
34 %
29 %
30 %
467 million
632 million
523 million
no migration
double migration
no migration,
+25% birth rate
Population ageing and a smaller labour force means that
European workers will need to support more dependents in the future
In 2015, 54 % of EU-28 population lived
in western Member States. If trends continue,
by 2060 that share will be 59 %
In 2015, the share of people
aged 65+ was the same in both
country groups, but it will grow
faster in the eastern Member
States by 2060
If trends continue, by 2060 the world population
will be 2.3 billion larger than today, with Africa
contributing 57 % (1.3 billion) of this growth
The education of girls has the strongest and most consistent connection to moderate fertility rates
Total fertility rate - number of children per woman in Africa
A smaller, better-educated labour force may prove
to be more adaptable to the changing nature of work,
automation and artificial intelligence
Even doubling migration has little effect compared
to continuing current trends. In fact, the effect is the same
as if immigrants were better integrated and participated
at the same rate as Europeans
226 million 245 million
lowest ratio
of dependents
to workers
245 million 215 million
The movement of workers towards wealthier EU states speeds up
ageing and population decline in eastern Member States
Girls’ education matters greatly for the future
of world population growth
The EU's future labour force
will be smaller and better educated
Increasing labour force participation
is the most effective way to cope
with population ageing
Migration increases the EU's total labour force size, but
has a limited effect on the ratio of dependents to workers
2015
2015
2015
2015
if trends continue, labour
force participation rate
as today
2060
2015
if men and women
equally participated
if participation rate as
high as in Sweden today
1980
2015
2060
if trends continue
Women with less than primary education in Africa
EU labour force over total population
13 % 466 million
509 million
521 million
19 %
32 %
35 % 59 %
7.3
billion
8.9
billion
9.6
billion
11
billion
1.2
billion
1.9
billion
2.5
billion
3.2
billion
2060
double
migration
and higher
integration
if trends
continue
2060
2060
2060
4.7
1.51 2.26 3.16
44 % 0 % 10 % 37 %
rapid education
development
education trends
continued
stalled education
development
2015 rapid education
development
education trends
continued
stalled education
development
2060 if intra-EU mobility
2060
2015
stopped
continued
UK AT DE LT ROLV
in western countries in eastern countries
100
108
age 65+ total EU population
life expectancy in the EU today
A longer active life can help
address the challenges related
to population ageing
81
100
106
100
133
100
126
100
127
dependents
workers
2015
2015
100
123
Even much higher immigration or fertility
will not change the pace of population
ageing significantly by 2060
The share of the population over
age 65 is bound to increase
WORLD WORLD WORLD WORLD
KEY MESSAGES
30 %
19 % 19 %
35 %
age 65+
- 38 %
- 20 %
+ 7 %
- 1 %
- 30 %
- 14 %
+ 17 %
+ 5 %
- 37 %
- 23 %
+ 26 %
+ 21 %
2
0
1
5
2
0
1
5
2
0
1
5
AFRICA AFRICA AFRICA AFRICA
DEMOGRAPHIC SCENARIOS FOR THE EU
MIGRATION, POPULATION AND EDUCATION
How can we avoid overburdening
our social system with population ageing?
increase migration?
increase fertility?
2060
or
34 %
29 %
30 %
467 million
632 million
523 million
no migration
double migration
no migration,
+25% birth rate
Population ageing and a smaller labour force means that
European workers will need to support more dependents in the future
In 2015, 54 % of EU-28 population lived
in western Member States. If trends continue,
by 2060 that share will be 59 %
In 2015, the share of people
aged 65+ was the same in both
country groups, but it will grow
faster in the eastern Member
States by 2060
If trends continue, by 2060 the world population
will be 2.3 billion larger than today, with Africa
contributing 57 % (1.3 billion) of this growth
The education of girls has the strongest and most consistent connection to moderate fertility rates
Total fertility rate - number of children per woman in Africa
A smaller, better-educated labour force may prove
to be more adaptable to the changing nature of work,
automation and artificial intelligence
Even doubling migration has little effect compared
to continuing current trends. In fact, the effect is the same
as if immigrants were better integrated and participated
at the same rate as Europeans
226 million 245 million
lowest ratio
of dependents
to workers
245 million 215 million
The movement of workers towards wealthier EU states speeds up
ageing and population decline in eastern Member States
Girls’ education matters greatly for the future
of world population growth
The EU's future labour force
will be smaller and better educated
Increasing labour force participation
is the most effective way to cope
with population ageing
Migration increases the EU's total labour force size, but
has a limited effect on the ratio of dependents to workers
2015
2015
2015
2015
if trends continue, labour
force participation rate
as today
2060
2015
if men and women
equally participated
if participation rate as
high as in Sweden today
1980
2015
2060
if trends continue
Women with less than primary education in Africa
EU labour force over total population
13 % 466 million
509 million
521 million
19 %
32 %
35 % 59 %
7.3
billion
8.9
billion
9.6
billion
11
billion
1.2
billion
1.9
billion
2.5
billion
3.2
billion
2060
double
migration
and higher
integration
if trends
continue
2060
2060
2060
4.7
1.51 2.26 3.16
44 % 0 % 10 % 37 %
rapid education
development
education trends
continued
stalled education
development
2015 rapid education
development
education trends
continued
stalled education
development
2060 if intra-EU mobility
2060
2015
stopped
continued
UK AT DE LT ROLV
in western countries in eastern countries
100
108
age 65+ total EU population
life expectancy in the EU today
A longer active life can help
address the challenges related
to population ageing
81
100
106
100
133
100
126
100
127
dependents
workers
2015
2015
100
123
Even much higher immigration or fertility
will not change the pace of population
ageing significantly by 2060
The share of the population over
age 65 is bound to increase
WORLD WORLD WORLD WORLD
KEY MESSAGES
30 %
19 % 19 %
35 %
age 65+
Introduction
The Centre of Expertise on Population
and Migration (CEPAM)
1
was founded by the
European Commission and the International
Institute for Applied Systems Analysis (IIASA)
in response to the migration events of 2015.
Despite its origins in the dramatic scenes that
seized global attention, CEPAM’s mission has been
to focus on the long term, providing analyses
of the gradual, but consequential demographic
changes taking place across the EU and wider world.
The sections
The rst ve sections focus on demographic
challenges inside the EU, such as population
ageing, a shrinking labour force, more non-working
people being dependent on working people,
and showing the impact of high levels
of emigration in some EU Member States.
With these challenges in mind, the report looks
towards 2060 to understand the long-term
eects of alternative scenarios for the future,
and whether undesirable consequences can be
limited or counteracted.
As the EU and its demographics do not exist
in isolation, Sections 6 to 8 explore relevant world
demographic trends. The topics covered include
the future of global population growth and the
potential divergent trajectories. Then, moving from
human numbers to qualitative decisions about
migration, the reports looks at push factors such
as conict, instability and climate change.
Finally, the report covers the current limitations
of demographic data and research capacity
for demographic modelling. Overcoming these
limitations is essential to providing a solid scientic
basis for the development of future policies.
The scenarios and narratives
Due to the inertia of demographic processes,
d emographers are able to project changing
population sizes and structures for decades into
the future. We can consider the impact of future
EU demographic changes by using demographic
scenarios. The projections are based on a few
clearly identiable quantitative assumptions
which refer to future forces of change, such
as fertility, mortality, migration or changes in
education and labour-force participation. Dierent
combinations of such assumptions are called
demographic scenarios.
In this report, we use a number of demographic
scenarios to study dierent, possible migration
ows into the EU. The scenarios are described
in each section. Reference points for all these
scenarios are the ‘central’, ‘medium’, or ‘SSP2’
2
scenarios which make middle-of-the-road
assumptions on future trends, closest to what
is considered likely from today’s perspective.
14
INTRODUCTION
p. 18
Population ageing
in Europe
p. 24
Size and education
levels of the future
EU labour force
p. 44
Impacts
of internal
EU mobility
over time
p. 58
World population
growth trajectories
p. 72
Evidence for
climate-related
migration
p. 36
Impacts
of migration
on the EU
labour force
p. 50
Illustrating the
consequences
of ‘brain drain’
p. 66
Reasons for
migration in Africa
and Western Asia
Introduction
15
In addition to the central or medium scenario, we
have developed a set of scenarios, which are used
in Sections 1-6, based around certain parameters
(e.g. ‘zero’ or no migration; a 25% increase in the
fertility rates, or the equal participation of men and
women in the labour force). Such parameter-based
scenarios are useful for unpacking the forces that
shape future demographic trends and thus also
have immediate policy relevance.
For Section 3, we use alternative narratives
concerning the characteristics of migration
to the EU, and build scenarios around them
to show the dierent eects of changing variables.
The scenarios range from high and low volumes
of migration to high and low educational
attainment of migrants. These migration narratives
have been formed around real-life examples by
looking at migration to specic countries, such
as Canada or Japan. The narratives have been
developed over the past year in a consultation
process with dierent European Commission
Directorates-General and external experts.
The report also builds on the research on migration
conducted by the Joint Research Centre (JRC).
3
We used multidimensional macro models
of population dynamics as well as micro-simulation
models to be able to project such scenarios
for multiple human characteristics over many
decades in the future. The macro models divide
populations by age, sex, levels of education,
labour-force participation and place of residence.
However, beyond a certain number of subdivisions,
it becomes more feasible to use micro-simulation
rather than macro-simulation. Such simulations
generate large numbers of virtual individuals
for whom many more dierent characteristics can
be captured and their future distribution modelled.
For this report, an important example of analysis
that could only be carried out by micro-simulation
is immigrants’ duration of stay in the country,
which is essential for forecasting their integration
into the labour market (see Section 3).
The terminology
A glossary of demographic terms is listed
at the end of the report.
p. 18
Population ageing
in Europe
p. 24
Size and education
levels of the future
EU labour force
p. 44
Impacts
of internal
EU mobility
over time
p. 58
World population
growth trajectories
p. 72
Evidence for
climate-related
migration
p. 36
Impacts
of migration
on the EU
labour force
p. 50
Illustrating the
consequences
of ‘brain drain’
p. 66
Reasons for
migration in Africa
and Western Asia
PART A:
DEMOGRAPHIC
STRUCTURES
AND MIGRATION
IN THE EU
18
1: Population ageing in Europe
SUMMARY
The dominant demographic challenge facing the EU comes from population ageing,
commonly understood as an increasing proportion of people aged 65 and above. This
section looks at possible measures to counteract associated concerns – particularly
that more elderly people will depend on a smaller working population – by looking
at how dierent scenarios change the EU’s total population, the size of its potential
labour force, and the population that will be 65+ by 2060.
Neither higher fertility nor more immigration will stop population ageing, as the
momentum has been decisively set from past, long-term trends. At the same time, we
are living more productive lives which, together with exible retirement ages, will help
to cope with ageing-related challenges.
18
BOX 1: SCENARIOS FEATURED IN SECTION 1
Central scenario: This represents the future of the EU-28 population assuming slightly increasing fertility
4
          
assumptions on mortality and education levels, includes intra-EU mobility and uses the recent average for

Zero international migration scenario (ZIM):


High immigration scenario: Immigration from third countries to the EU is doubled compared to the central


ZIM plus 10%, 25% and 50% fertility:
This is the zero international migration scenario, plus an increase

POPULATION
AGEING
IN EUROPE
 Population ageing and old-age
dependency
Population ageing is widely considered a challenge
for the future of the EU’s economic and social
security systems. Specically, the EU would face
a growing burden from care and pensions as well
as a projected decrease in the labour force size.
Population ageing results from the combined
eects of trends in fertility, mortality and
migration. Increasing longevity is unquestionably
a valuable achievement of human development,
and continuing improvement in life expectancy
is without doubt a universally supported target.
Life expectancy at birth in the EU averages about
81 years, nine years more than the global average.
While measures of population ageing typically use
age 65 as the ‘elderly’ threshold, it has recently
been suggested that the age range in which people
are considered active should be raised as people
are living longer and are more productive for
longer. Moreover, in several Member States, the
statutory retirement age is already over 65 and it is
expected to increase further on the basis of current
legislation. In line with this approach, people would
not be considered ‘elderly’ when they reach 65, but
rather when they reach the age at which remaining
life expectancy is 15 years or less (Sanderson and
Scherbov 2005, 2007, 2010, 2013). In this report,
for simplicity’s sake, we still refer to the conventional
denition of working age as being 20-64 years.
Could either more migrants or more children slow
down the process of population ageing in the long
run? In political discussions, some see migration as
a remedy for the perceived problems of population
ageing, whilst others believe the remedy lies in
higher birth rates and support for larger families.
As we will show in this section, these two remedies
appear to have dierent consequences
on population size and age structure. For example,
although substantial immigration has an
immediate impact on the size of the working-age
population, its impact on the age structure
is less signicant in the long run because
immigrants also grow older. While some
immigrants may return to their countries of origin,
past experience indicates that a majority of retiring
immigrants in the EU will remain in the long term.
19 1: Population ageing in Europe
Neither higher fertility
nor more migration
will stop population
ageing, but more
productive lives are
helping to meet the
related challenges.
Figure 1.1: Population size of EU-28 in 2015-2060, by scenario
Source: CEPA M
In contrast, policies aimed at directly or indirectly
increasing fertility (for example, by promoting
work-life balance), if eective, would oer the
labour force long-term benets taking eect
in 20-25 years. In the short-term, higher fertility
leads to greater dependency on the labour force
by increasing the share of dependent children in
the population. This can create a temporary decline
in labour-force participation if parents leave work
to take on childcare. For comparison,
an increase in the total fertility rate (TFR) by 0.1
child per woman in the EU-25 would have nearly
the same eect on the old-age dependency ratio
as an additional 375 000 immigrants per year
by 2050 (Lutz and Scherbov 2007).
Bearing in mind these considerations, we use six
scenarios to demonstrate the impact of dierent
fertility and migration trajectories on the EU’s
total and working-age population until 2060.
These scenarios are: a central scenario, a zero
international migration scenario (ZIM), a high
immigration scenario, and three fertility scenarios
(ZIM +10%, ZIM +25%, ZIM +50%), as described
in box 1. By juxtaposing the three assumptions
of fertility: 10%, 25% and 50% higher fertility than
the central and ZIM scenarios, we can measure
the eect of increased fertility and distinguish it
from the eect of migration. To measure the eect
of migration, we compare the ZIM scenario with
the central and high immigration scenarios. In this
section, the focus is on the impact of the amount
of immigration and the fertility rates. Other factors,
such as the immigrants’ educational composition
and labour-market participation, are examined later
in Sections 2 and 3.
 The impact of the scenarios
on the total EU population
Despite perceptions of a shrinking population,
the EU is expected to grow further if recent trends
continue. As shown in gure 1.1, under the central
scenario, the EU’s population would grow from
508.5 million in 2015 to around 521 million by
2060, which is an increase of 2.5%.
5
The working-
age population would decrease from 306
to 256 million people – a decrease of 16% –
450
470
490
510
530
550
570
590
610
630
650
2015 2020 2025 2030 2035 2040 2045 2050 2055 2060
Population in millions
High immigration scenario ZIM+50% fertility
Zero international migration (ZIM) ZIM+25% fertility
ZIM+10% fertilityCentral scenario
1: Population ageing in Europe 20
Figure 1.2: Working-age population (20-64) of EU-28 in 2015-2060, by scenario
Source: CEPA M
and the proportion of the population aged 65
and over would grow from 19% to 32%. This
means that by 2060, for every 100 people of
working age there would be 114 of non-working
age, predominantly 65 years and older. Under
the central scenario, the number of births would
decrease from 26 million in 2015-2020 to 24
million in 2055-2060, given the lower number
of women of childbearing age.
In the ZIM scenario, the EU population would reduce
by roughly 9%, to 466 million people by 2060,
bringing it to the same size as in the 1980s. In the
short term, the scenario would produce a smaller
working-age population by almost 30% (to 222
million by 2060), as shown in gure 1.2. However,
in the long term, the dierence between the central
and the ZIM scenario is only minor because migrants
also age. With zero international migration the share
of 65+ would be 34% in 2060, which is only
2 percentage points higher than the central scenario.
For every 100 people of working age, there would
be 118 of non-working age, which is 4 percentage
points higher than in the central scenario.
200
220
240
260
280
300
320
340
2015 2020 2025 2030 2035 2040 2045 2050 2055 2060
Population 20-64 in millions
High immigration scenario ZIM+50% fertility
Zero international migration (ZIM) ZIM+25% fertility
ZIM+10% fertility
Central scenario
As regards the fertility scenarios, increasing
fertility by 10% (approximately 0.2 children per
woman) while assuming zero migration generates
a smaller total population than in the central
scenario. A 25% increase in fertility (approx.
0.4 children per woman), bringing the TFR to
replacement level
6
, means that the EU population
would increase by 2.8% to reach 521 million
people in 2060 (a similar trend to the central
scenario). A 50% increase in fertility (about 0.85,
increasing the TFR to a rather unlikely 2.6 children
per woman) would generate a faster population
increase than the central scenario and would
encounter a steeper increase aer 2050 when
the larger cohorts of births would reach
reproductive age. By 2060, this scenario would
see around 580 million people living in the EU.
21 1: Population ageing in Europe
 The impact of the fertility scenarios

In terms of the eect on the size of the working-age
population, the three fertility scenarios in gure
1.2 show that the size would decline at the same
pace until 2040 – that is, until babies currently being
born start to reach working age. Only the ZIM +50%
scenario would have a larger working-age population
than the central scenario, and then only aer 2050.
Under this scenario, the total EU population would
grow by nearly 75 million people by 2060. However,
such a fertility rate has not been seen in most EU
countries since the 1970s.
 The impact of the fertility scenarios

Table 1.1 compares the population structure
indicators in all six scenarios: the share of 65+
in the total population and the total age-
dependency ratios. For ZIM +50%, the share
of 65+ would increase from 19% to 27% in 2040
and remain stable until 2060. The total age-
dependency ratio would reach 130, which means
that for every person of working age there would
be 1.3 dependants. This is because it also includes
an increase in the number of births by almost 70%
to 43.5 million in 2060.
Year Central
scenario ZIM ZIM +10%
fertility
ZIM +25%
fertility
ZIM +50%
fertility
High immigration
scenario
% POPULATION 65+
2015 19 19 19 19 19 19
2040 28 30 29 28 27 26
2060 32 34 33 30 27 29

2015
76 76 76 76 76 76
2040 99 103 106 112 121 93
2060 114 118 122 125 130 104
Tab le 1.1: Share of population aged 65 and above and the total age-dependency ratio in 2015,
2040 and 2060 in the EU-28, by scenario.
ZIM = Zero International Migration
Total age dependency ratio = a ratio bet ween the economically inactive (age 0-19 and 65+) and working age population (age 20-64)
Source: CEPA M
1: Population ageing in Europe 22
 Impact of the high immigration
scenario on the size of the EU’s

Therefore, it appears that even with higher fertility,
the non-working-age population would be above
the current levels while the working-age population
would continue to decline. Can more migration
counteract such a decline? The high immigration
scenario, which is equivalent to doubling 2013-2016
migration ows to 20 million coming into the EU
every 5 years, would increase the EU’s population
size by almost 25%, reaching about 632 million
people by 2060 (see gure 1.1). There would also
be a higher number of births due to more people
of reproductive age, from 27 million in 2015-2020
to almost 30 million in 2055-2060. With high
immigration, the total age-dependency ratio would
reach 104 people by 2060 compared to 114 in the
central scenario. However, the high immigration
scenario would only moderately reduce the pace
of ageing. The proportion of the 65+ population will
increase, from 19% in 2015 to 29% in 2060. This
is only 2 percentage points lower than the proportion
obtained under the central scenario assumptions.
 Conclusions
A number of conclusions can be made from our
ndings. First, increasing the number of births,
even by means of successful government initiatives,
is not an instant remedy for population ageing
in the EU. An increase in fertility to the replacement
level of about 2 children per woman from 2020
onwards would not completely prevent the EU
working-age population from declining in size.
Although the pace of ageing would be slower and the
total population size would stabilise, such an increase
would still lead to higher total dependency ratios.
Secondly, more immigration from third countries
into the EU would have an immediate eect on the
total size of the EU population and the working-
age population. However, even with double the
immigration levels of 2013-16, immigration
would make little improvement in the long-term
on reducing the proportion of people of non-
working age relative to those of working age (as
conventionally dened by ages 20-64). This is
due to the fact that immigrants inevitably grow
older, retire from work and require social benets
(e.g. pensions), just like non-immigrants. In short,
increasing immigration or fertility will not stop the
population from ageing, reecting a trend that is
happening not only in the EU, but in almost every
country worldwide.
These two conclusions can help to illustrate
the impact of possible policy options. For example,
substantial investment in work-life balance
initiatives (including tax incentives, child allowances,
day-care access and family-friendly work
arrangements) sustained over time have created
an environment conducive to higher fertility close
to the replacement level in some EU Member States,
such as France. Similarly, increasing the number
of immigrants would theoretically be possible.
For this, there would be political, social and
economic issues to consider, and well-designed
integration policies would be required to enable
smooth immigration and successful job matching.
Bearing in mind that the above options are based
on scenarios, it is important for policymakers
today to consider the possibility that neither
increased fertility rates nor migration are the
panacea for the perceived problems associated
with Europe’s ageing population. Solutions must
have a broader, multidimensional demographic
perspective, taking into account levels of education
and labour-force participation. In fact, the most
eective remedy would be to increase labour-force
participation, rather than looking to either migration
or fertility. The next section goes in this direction
by showing how higher labour-force participation
can compensate for a decreasing working-age
population – an unavoidable demographic reality
EU Member States will have to face in the
coming decades.
23 1: Population ageing in Europe
SIZE AND
EDUCATION LEVELS
OF THE FUTURE
EU LABOUR FORCE
2: Size and education levels of the future EU labour force
SUMMARY
The EU labour force is transforming amid social development and the ageing process.
This section projects a smaller and more educated labour force in the future. In light
of the context, a possible decline in labour force size may not be an economic problem
if future jobs are fewer in number and require higher skills. At the same time, this
section considers scenarios that test how greater labour-force participation could
not only stem the tide of a decreasing working-age population, but also reduce the
expected increase in the dependency ratio.
The constant participation scenario acts as the reference trajectory, which charts
the expected path of a shrinking labour force and a rising dependency ratio. We
then introduce the equalisation and Swedish scenarios to see how increasing
labour-participation rates in the EU can help oset this trajectory. An increase in
female labour-force participation and longer productive lives could stabilise the size
of the labour force and dependency ratio at current levels.
24
BOX 2: SCENARIOS FEATURED IN SECTION 2, BASED ON CEPAM’S SSP DEVELOPMENT
PATHWAYS (PAGE 58)
Constant participation scenario:-


Equalisation scenario:          

Swedish scenario: All labour-force participation rates in all EU Member States gradually converge to those

SIZE AND
EDUCATION LEVELS
OF THE FUTURE
EU LABOUR FORCE
 The demographic transition
Historically, the EU population has gone from
having high fertility and high mortality rates
to stable, below replacement fertility and low
mortality, as part of the universal process known
as the demographic transition. At the end of the
transition, population growth rates either stabilise
or reverse. While there are fewer births and longer
life expectancies, young people are becoming more
educated and entering the labour market later.
Such developments will inevitably shape the size
of the EU’s future labour force and its distribution
by level of education.
Despite population ageing, a smaller, higher-
skilled labour force may be poised to complement
the possible changing needs of the future labour
market (Cedefop 2016). Furthermore, getting more
women into the labour force and having longer
working lives as Swedes already have today, has
the potential to nullify the shrinking
of the working-age population over time.
 A smaller but more-educated
labour force
Under the constant participation scenario,
the total size of the labour force is projected
to change from 245.8 million to 214.1 million over
the next four decades, as illustrated in gure 2.1.
This decrease is solely attributable to there being
fewer workers with low (primary education)
and middle education (upper secondary and below).
Projected decreases in the size of the labour force
come from the population with low (from 50.7
to 14.0 million) and middle (from 108.2 to 74.2
million) levels of education.
At the same time, the number of workers with
a short post-secondary education (such as
25 2: Size and education levels of the future EU labour force
Higher female
labour-force
participation
and longer productive
lives have the power
to stabilise the
labour-force size
and dependency ratio.
Figure 2.1: EU labour force by education level, the constant participation scenario, 2015-2060
Source: CEPA M
technical training), a bachelor’s, a master’s degree
or higher, are rising, not just in relation to their
proportion to the overall labour force, but also
in absolute terms. Signicantly, these post-
secondary groups are expected to increase
by almost a half (+45%) over the next 40 years.
Therefore, irrespective of changes in the
population’s age structure and labour force sizes,
it can be expected that the human capital of future
workers of any age, measured by the highest level
of their educational attainment, will be higher than
today. This is because in all Member States, young
people are already better educated than their
elders and will gradually replace them through
the process called ‘demographic metabolism’.

participation impact on the size
of the total labour force
The three scenarios in gure 2.2 illustrate how
the divergent labour force participation rates
inuence the size of the labour force over time.
All three assume the same demographic future
as the constant participation scenario (see gure
2.1). Under the constant participation scenario,
the labour force size would be about 13% smaller
than it is today (from 245 million to just under 215
million). Under the equalisation scenario, the size
of the labour force would decline to 227.2 million
in 2060, which represents about 13 million more
workers than under the constant participation
scenario. Furthermore, if participation rates rise
to the levels currently observed in Sweden
(the Swedish scenario), the labour force size
would stabilise at 245 million workers, avoiding
any decline. Both the equalisation
and the Swedish scenarios express the high
capacity that participation rates – in particular
female participation – have to aect the labour
force size. In fact, participation rates can impact
the size of the labour force notably more than
increasing migration ows.
7
Looking at EU Member States more closely, under
the constant participation and Swedish scenarios,
map 2.1 shows that under the former, the size
of the labour force in many Western Member
States would remain stable, due in part to gaining
working-age nationals from Eastern Member
States. In turn, Eastern and Southern EU Member
States would continue to see reductions in the size
of their labour force.
8
0
50
100
150
200
250
2015 2020 2025 2030 2035 2040 2045 2050 2055 2060
Labour force in millions
Total labour force by education level in the European Union
according to the constant participation - CEPAM Medium (LT),
2015-2060
Low Middle Short post-secondary Bachelor Master or higher
2: Size and education levels of the future EU labour force 26
Figure 2.2: The three scenarios inuencing the size of the EU’s total labour force, 2015-2060
Source: CEPA M
200
205
210
215
215
220
230
235
240
245
250
2015 2020 2025 2030 2035 2040 2045 2050 2055 2060
Population in millions
Constant scenario Equalisation scenario Swedish scenario
Increasing
labour-force
participation rates
– particularly
of females –
has a high capacity
to affect labour-
force size
27 2: Size and education levels of the future EU labour force
Map 2.1: Changes in the size of the labour force 2015-2060 under the constant participation
and Swedish scenarios
Source: CEPA M
Created with mapchart.net©
2: Size and education levels of the future EU labour force 28




















ratios can reduce the public
spending burden
When it comes to considering future public
spending on pensions, social services, etc.
for the non-working population, it is important
to look at the labour-force dependency ratio (LFDR).
This measure is important for understanding the
potential economic impact of population ageing,
since the higher the ratio, the more challenges
there are to labour markets, government tax,
government spending and the wider economy.
In 2015, the LFDR was 1.05, meaning that
there were about 105 inactive people for every
100 workers. With the constant participation
scenario, in gure 2.3, the LFDR would reach
1.36 in 2060. However, ratios would vary a lot
by Member State, as seen in map 2.2. They would
be much higher in Member States experiencing
low fertility, higher life expectancy and low
labour-force participation, such as Greece (1.69)
and Italy (1.72). On the other hand, Member
States with medium or high fertility and a higher
labour-force participation rate would have more
favourable ratios, such as Sweden (1.04)
and Denmark (1.05).
Changes in labour-force participation rates could
signicantly impact the projected LFDR. In the
equalisation scenario in gure 2.3, the ratio
would stabilise at about 1.2, which reduces to half
the expected increases compared to the constant
participation scenario, meaning it is a more
favourable scenario. In the Swedish scenario,
the ratio would stabilise to the level of what
it was in 2015. In other words, a strong but realistic
increase – it is already reality in Sweden – of labour
force participation over time could nullify a large
part of the negative future economic consequences
of population ageing. So, it is clear that labour force
participation rates have the highest potential
to mitigate the challenges of population ageing.
Figure 2.3: EU labour-force dependency ratio, by scenario, 2015-2060
Source: CEPA M
1
1.05
1.1
1.15
1.2
1.25
1.3
1.35
1.4
2015 2020 2025 2030 2035 2040 2045 2050 2055 2060
Labour-force dependency ratio (LFDR)
Constant scenario Equalisation scenario Swedish scenario
29 2: Size and education levels of the future EU labour force
 Conclusions
The EU’s future labour force is likely to be smaller
but better educated. The impact of the decline
in the size of the labour force could be lessened
by an increase in productivity thanks to a
more educated labour force. Plus, incentives
to encourage more women to participate in the
labour force and to get everybody to participate
in the labour force, as observed in Sweden, could
also prevent its decline and could stabilise the
LFDR. The combined eect of increasing labour-
force participation and worker productivity could
enable economic growth to continue, thereby
osetting public spending on dependant people.
Overall in the EU, the total cost of ageing (public
spending on pensions, health care, long-term care,
education and unemployment benets)
is expected to increase by 1.7 percentage points
to 26.7% of GDP between 2016 and 2070
(European Commission, 2018a). Thus, the scenarios
projected in this section are important for EU
policymakers to consider. In the context of this
challenge and covering the extra 1.7% of GDP,
a higher educated labour force will need
to be utilised. Policymakers should ensure that
the education system is responsive to the needs of
the job market and the changing nature of work, so
that increasingly highly educated young people can
contribute to the economy and actualise the higher
potential productivity that comes with education.
Map 2.2: Labour-force dependency ratio, constant participation scenario, 2060
Source: CEPA M
Labour-force dependency ratio (LFDR) = a ratio between people in the labour force and people not in the labour force.
Created with mapchart.net©
2: Size and education levels of the future EU labour force 30

ratio, constant scenario





31 2: Size and education levels of the future EU labour force
In the
context of
ageing and extra
public spending,
policymakers should
ensure that
the education system
is responsive
to the job market
Migration narratives for Section 3 scenarios 32
Low VolumeHigh
Human Capital of
Immigrants (similar to
Japan in recent years)
EU Member States adopt more selective
migration policies to both emphasise
skills and substantially reduce the overall
volume of ows. Immigrants are admitted
primarily based on their potential
for contributing to the economy, using
a points system to target highly specialised
and needed skills.
The number of immigrants settling
in the EU amounts to approximately 1.2
million every 5 years (corresponding to the
recent immigration rate in Japan, which
would mean that ows reduce to about
12% of the volume of the past 5 years).
Integration into the labour market is
facilitated by the highly skilled proles and
screening for other criteria. Member States
maintain the system’s integrity by enforcing
rules on employment and penalising
businesses for unlawful hiring practices.
Dealing with an ageing workforce,
policymakers in this narrative address
challenges by investing in automation
and mechanisation. The new focus intends
to use emerging technologies to improve
production eciency and reduce the
dependence of various sectors (agriculture,
heavy industry, etc.) on repetitive, low-skilled
labour, suitable for automation. Similarly,
the health-care system reduces the risks
of labour shortages by using technologies
to improve health-care services, emulating
existing best practices used abroad.
MIGRATION NARRATIVES

Low VolumeLow
Human Capital of
Immigrants (low-skilled,
circular system)
A reform of the education system benets
EU Member States by successfully linking
university and technical programmes
to economic needs. With a more ecient
education-to-labour-force path, Member
States do not aim to supplement their
labour force by recruiting large numbers
of immigrants from third countries. Instead,
the primary focus of migration policy shis
to temporary refugee assistance and lling
remaining low-skilled jobs, with ows
at around 1.2 million every 5 years.
A new migration system is implemented
with a focus on work permits of a temporary
and seasonal nature. Such changes
to labour migration create more circular,
rather than permanent movement,
in part to address mounting concerns from
developing countries of ‘brain drain’ losses.
Member States also adjust asylum policies
accordingly.
The general orderly management and low
volume of immigrants do not overload
Migration narratives for Section 3 scenarios
33
relevant support resources. At the same
time, permits for low-skilled immigrants and
temporary workers are calibrated to match
unlled jobs, facilitating their entry into the
labour force. The new system depends on
regular monitoring of the economy, ecient
administration of set migration levels,
and ensuring that those employing
temporary workers full legal work
requirements as protection against fraud.
High Volume – High
Human Capital
of Immigrants
EU Member States begin recruiting
increasing numbers of highly skilled workers
from countries around the world to address
economic concerns. The introduction
of a new high-volume, highly skilled
immigration system brings approximately
20 million migrants into the EU every
ve years (corresponding to the recent
immigration rate in Canada, which would
mean that ows double the volume
of the past ve years). Maintaining its highly
selective nature, the system relies on the
enforcement of work permits and border
checks as the EU becomes an increasingly
prominent immigration destination.
Most of the new immigrants originate from
middle-income countries with relatively
strong education systems, although
the system also attracts the highly educated
elite from low-income countries. In turn,
sending countries struggle with the losses of
their most talented and educated citizens as
the EU’s appetite for skilled workers persists.
Integration into the economic system
comes without signicant barriers, since
the admission of migrants is a function
of scoring highly on criteria for education,
skills, language ability and work experience.
A large proportion of immigrants pursue
opportunities in the urban economic hubs
of Western and Northern EU, given
the greater demand for highly skilled
workers as well as their own personal
preferences and existing migrant networks.
This narrative is similar to the model
currently used in Canada.
The Canadian immigration programme
is a merit-based system that recognises
three broad classes of immigrants:
economic migrants, sponsored family
and refugees. Immigration targets
are established for all immigration
categories and subcategories, and are
set at the federal level, in agreement
with provincial authorities. Targets
are estimations of future immigration
inux and as such are not considered
as quotas.
11
Between 2007 and 2016, the Canadian
government granted permanent
residency to 2 600 000 immigrants,
60% of whom were economic migrants
chosen for their potential ability
to integrate in the Canadian labour
market.
12
In this context, dependants
of economic migrants are also classied
as economic migrants.
Most economic migrants are admitted
based on a selection grid assessing
their skills and human capital, such as
education level, previous work experience
and uency in Canada’s ocial
languages, French and English. Furthermore,
economic migrants may be admitted
as permanent residents under sub-
programmes targeted at individuals with
specic characteristics, such as investors,
entrepreneurs or individuals
with previous work experience in Canada.
Aside from programmes leading
to permanent residency, Canada has
schemes to provide temporary work permits
to foreign nationals. The number of holders
of temporary work permits increased
signicantly during the last decade, doubling
from 173 000 in 2007 to 340 000
in 2016 .
13
There are two programmes for refugees.
The rst is the Refugee and Humanitarian
Resettlement Program, for people requiring
protection from outside Canada (about
65% of refugees according to the 2019
target). The second one is the In-Canada
Asylum Program for people making refugee
protection claims from within Canada.
The United Nations Refugee Agency
(UNHCR), along with private sponsors,
identies refugees for resettlement.
A person cannot apply directly to Canada
for resettlement. In 2019, refugees are
expected to account for approximately
15% of the total number of migrants.
High Volume – Low
Human Capital
of Immigrants
In the face of persistent conict, instability
and poverty in the EU neighbourhood, and
increasing costs of securing the external
borders, EU Member States decide to expand
the intake of low-skilled labour and adopt
more generous asylum policies (in terms of
determining safe countries, allocating resources
and acceptance rates). Many Member States
aim for the highest migration-granting rates of
recent years, focusing less on skill selection or
border enforcement.
Intensifying push factors create conditions for
a constant supply of immigrants as tens of
millions of the developing world’s impoverished
look abroad to make their livelihoods.
Furthermore, the EU is seen as an ever-more
viable destination by prospective immigrants,
encouraging a sustained exodus. Overall,
migrant ows into the EU as a whole reach 20
million every 5 years (about double the volume
compared to the past 5 years), progressively
characterised by greater proportions of low-
skilled workers, as well as refugees and family
reunication cases.
Integrating them into the labour market
becomes an important topic in public discourse
as record numbers of immigrants are given
permission to live and work in the EU. Most
arriving immigrants tend to have a lower skills
prole compared to the EU average, presenting
barriers to their potential labour-market
participation. Due to more plentiful economic
opportunities, personal preferences, historical
links and existing communities, the majority of
migrants choose to live in Western, Northern,
and a couple of Southern Member States.
Migration narratives for Section 3 scenarios 34
Migration narratives for Section 3 scenarios
35
SUMMARY
Policymakers regularly balance migration policy with a Member State’s best interests
– doing so while being mindful of the fact that a long-term understanding of
demography is crucial. At the core of such considerations are factors including: the
volume of immigrants from third countries entering the Member State, their levels
of education, how well they will integrate into the labour market, and how eective
the Member State is at enforcing migration policy. This section shows how dierent
volumes of immigration and the integration by immigrants into the labour market
change the educational composition of the future EU labour force and the LFDR ratio.
Whilst a high volume of immigration would increase the size of the EU’s total labour
force, it would have a limited impact on the proportion of workers to non-workers in
the long-run. Larger inows combined with deteriorating economic integration could
actually result in a labour force situation that is worse than that with medium or low
volumes of immigration, highlighting the importance of eective integration eorts.
BOX 3: ADD-ON SCENARIOS TO THE MIGRATION NARRATIVES FEATURED IN SECTION 3
(SEE PAGES: 32-34)
Baseline labour-force integration scenario:

Low integration scenario:    

High integration scenario:

Baseline volume scenario: Uses the short-term average for migration to the EU; medium

3: Impacts of migration on the EU labour force 36
37 3: Impacts of migration on the EU labour force
IMPACTS
OF MIGRATION
ON THE EU
LABOUR FORCE
 Migration scenarios
First, to assess the dierent volumes of immigrants
and their human capital, we constructed four
migration scenarios which are based on dierent
qualitative narratives (see box 3) : (a) low volume
– high human capital of immigrants; (b) low
volume – low human capital of immigrants; (c)
high volume – high human capital of immigrants;
and (d) high volume – low human capital
of immigrants.
9
These scenarios are illustrated
in gure 3.1. Then we combined them with three
dierent labour-force integration scenarios:
baseline, low integration and the high integration
scenario (dened in box 3).
By combining these migration scenarios with
labour-force integration scenarios it is possible
to analyse the impact of the volume of immigrants,
their levels of education and of integration on the
EU’s labour force size, its educational composition,
and its LFDR
 Current labour-force participation
of immigrants
To understand how the integration scenarios aect
the EU labour force, we rst need to look at
the current levels of participation by immigrants
in the labour market, as shown in gure 3.2.
Recently arrived male immigrants tend
to participate much less than native-born
populations. Over time, their labour force
participation rate (LFPR) tends to increase
as the duration of stay in the host country
increases. On average, aer a 10-year stay,
participation rates for male immigrants are
only slightly lower than those of the native-born.
The dynamics of labour-force integration vary
widely among EU Member States. In Germany,
Belgium, Denmark and the Netherlands there
is a large gap in LFPR between the native-born
and immigrants, even for those who have been
in their host country for a decade. Conversely, LFPR
37
High migration
volumes increase
labour-force size but
have a limited eect
on the number of
dependent people
per worker in the EU.
3: Impacts of migration on the EU labour force 38
for the native-born and male immigrants, even
recent ones, are similar to one another in Greece,
Spain and Portugal.
Gender disparities in LFPR are more prominent
among immigrants than between native-born men
and women. Female immigrants are the least likely
to participate in the labour force as their LFPRs are
lower than both male immigrants and native-born
men. Moreover, female immigrants have not yet
caught up with the LFPR of native-born women
in any EU Member State.
The positive relationship that education has with
LFPR is weaker for immigrants than for those who
are native-born, especially for female immigrants.
Immigrants who arrived during childhood, however,
tend to have more similar outcomes to those who
are native-born, having gone through schooling
in the host Member State.

immigrants on the size of the future
EU labour force
The number of immigrants admitted into the EU
would strongly alter the overall population size
and thus the labour-force size. As shown in table
3.1, where the volume of immigrants is high (and
there is high integration), there could be around
30% more workers in the labour force by 2060
compared to the baseline scenario, and 60% more
workers compared to the low-volume scenarios.
This high volume of migration would lead to
signicant growth in the labour force size in terms
of absolute numbers.
In addition to migration volume, integration
dynamics have a signicant impact on the EU's
future labour force size. If migration volumes are
high, the stakes of integrating into the labour force
are much higher. For instance, table 3.1 shows
that the dierence between the high volume of
immigrants and high integration (between 300-310
million people) and high immigration and low
Figure 3.1: The migration scenarios, contrasting volumes of migrants with their level
of human capital
Source: CEPA M
Low Volume,
High Human Capital
(Japan)
High Volume,
High Human Capital
(Canada)
High Volume,
Low Human Capital
EU-28 Baseline
(Central Scenario)
Low Volume,
Low Human Capital
High
Low
1.2m every 5 yrs. 10m every 5 yrs.
Volume of immigration flows, applied to the EU
20m every 5 yrs.
Human capital
39 3: Impacts of migration on the EU labour force
Figure 3.2: Labour-force participation rate* in the EU by migration status, age at arrival
and duration of stay, 2010-2015
Source: CEPA M
50%
55%
60%
65%
70%
75%
80%
85%
90%
95%
100%
Natives Duration
of stay<5
10<=Duration
of stay
5<=Duration
of stay<10
Immigrants arriving
during childhood
Labour force participation rate* according to immigrant status,
age at arrival and duration of stay, 2010-2015, European Union
Men Women
*Statistical control for age (30-34) and education (post-secondary education)
Immigrants arriving as adults
integration (264-271 million people) can vary
by around 14%, whereas the dierence between
the low volume of immigrants and high integration
(between 171-172 million) and low immigration
and low integration (168-169 million) is far less
at around 2%.
Because immigrants’ labour-force participation
varies much less according to education than
it does for a Member State’s native-born (or host
population), the eect of the immigrants’ level
of education on the projected labour-force size
is limited. Under both a medium volume
and baseline integration situation, the high-
education assumption (of 229 million people) only
results in 3 million more workers (1.3%) relative
to the medium-education assumption (226 million).
As table 3.1 indicates, a high volume
of immigration will lead to more people
in the labour force than the number currently,
irrespective of their level of education or how
well they integrate into the labour market. However,
not only does this mean that there will be a larger
labour force, but that there would also be a larger
non-working population, partly because immigrants
tend to have lower participation rates than natives,
particularly female immigrants. Immigration also
causes growth of the non-active population
in the long run as immigrants age and leave
the workforce. Most importantly, as discussed later
in this section, it must be noted that the size
of the labour force is neither a good indicator
of economic health nor a way to address
population ageing.
3: Impacts of migration on the EU labour force 40

of immigrants on the educational
composition of the future EU
labour force
In Section 2, it has been projected that, by 2060,
the EU’s labour force will be better educated than
it is now. Table 3.2 shows that the proportion of
workers with post-secondary education could reach
between 40% and 52% in 2060, compared to 35%
in 2015.
Unless migrants are highly prioritised according to
their education level, higher volumes of migration
tend to decrease the proportion of workers with
post-secondary education.
10
As shown in table 3.2,
in cases where migrants’ educational composition
is at the medium level, the EU’s proportion of
post-secondary workers is highest (51%) when the
volume of migrants is low; 48% when the volume
is at its baseline; and 45% when the volume is high.
The declining proportion of post-secondary workers
in the EU – moving from low to high volumes of
immigrants – is because, on average, immigrants
tend to have a lower level of education than those
who are native-born.
When the volume of immigrants is high, their
educational composition would signicantly alter
the future educational composition of the labour
force in the EU as a whole. For example, within
the high-volume scenarios in table 3.2, a strong
selection for highly educated immigrants would
lead to 53% of workers in the EU with post-
secondary education. However, if a high volume
of immigrants with low education entered the EU
labour force, the proportion of the labour force with
post-secondary education would decrease to 41%.
In contrast, where immigrant educational levels are
more moderate and the volume of immigrants is
low, this would have a smaller impact on the overall
education composition of the EU labour force with
post-secondary education.
Furthermore, the integration assumptions (high
and low) inuence the absolute number of
workers with post-secondary education to an
extent, but have a very limited impact on their
Immigration
volume
Educational
composition
of immigrants
Integration assumptions
Low Baseline High
Low
Low 168 170 171
Medium 168 170 171
High 168 171 172
Baseline
Low 212 224 231
Medium 214 226 233
High 217 229 237
High
Low 264 287 300
Medium 265 289 302
High 271 295 310
LABOUR FORCE IN 2015 = 245
Tab le 3.1: Projected labour force size (in millions) of the EU in 2060
Source: CEPA M
41 3: Impacts of migration on the EU labour force
Immigration
volume
Educational
composition
of immigrants
Integration assumptions
Low Baseline High
Low
Low 51% 50% 50%
Medium 51% 51% 51%
High 52% 52% 52%
Baseline
Low 46%45%44%
Medium 48%48%47%
High 52% 53% 53%
High
Low 42%41%40%
Medium 46%45%45%
High 53% 53% 54%
PROPORTION IN 2015 = 35%
Tab le 3.2 : Projected proportion of the labour force with post-secondary education
in the EU in 2060
Source: CEPA M
proportion relative to other education groups in
the overall workforce.

of migration on the labour-force
dependency ratio
Although higher immigration volumes have
a large impact on increasing the labour-force size,
their impact on the LFDR is much smaller. Table
3.3 shows that under the medium-education
composition and baseline integration assumptions,
a high volume of immigrants only manages to
decrease the LFDR by 0.06, compared to the
baseline scenario (1.27 vs. 1.33).
Under all combinations of immigrant volumes,
levels of education and integration assumptions,
the LFDR would be higher than in 2015 (1.08) due
to the unstoppable momentum of an increasing
average age. This is partially because immigrants
make up a relatively small proportion of the
population and therefore have limited ability to
aect the overall population, age structure or lower
participation rates. In any case, as table 3.3 shows,
migrants add to both the overall active and inactive
populations in the long-run.
The integration assumptions, however, have a larger
impact on the LFDR than the actual volumes of
immigrants. Table 3.3 shows that, under the high-
volume and medium-level educational compositions,
the ratio reaches 1.17 with high integration and
climbs to 1.48 if integration is unsuccessful.
Furthermore, low immigration and high integration
yields a better LFDR than high immigration and low
integration (1.39 vs. 1.48). Similarly, moving from


baseline integration rates (1.26 vs. 1.27). In other
words, the impact of immigration on the labour force
depends largely on the immigrants’ ability to access
the labour market rather than on the volume of
immigrants.
Variations in the educational composition of the
volumes of immigrants have a smaller impact
on the LFDR. For most combinations concerning
the volumes of immigration and integration
assumptions, the three educational composition
variants do not signicantly inuence the LFDR.
However, they matter for the productivity-weighted
LFDR (not shown here) which takes account of the
fact that more educated people tend to earn higher
salaries and thus pay more into the system.
These economic aspects go beyond the scope
of this scenario exercise but deserve in-depth
follow-up research.
Conclusions
This section has considered how the dierent
volumes of immigration and their various levels of
integration into the labour force impact the size of
the EU’s future labour force, the proportion of post-
secondary education in the future EU labour force,
and the LFDR by 2060.
A number of key ndings can be asserted. First,
high volumes of immigration would increase
labour force size in the EU, but would have a very
limited impact on the LFDR, which is in line with
the ndings of a recent OECD study (Spielvogel
and Meghnagi 2018). Therefore, a larger labour
force size will not necessarily curb or reduce the
number of non-working people depending on the
labour force, which undoubtedly has repercussions
on economic growth and public spending, as
mentioned in Section 2.
Even though immigrants tend to be younger on
average than those who are native-born, the
positive eect on the labour market is short-lived,
in part because the participation rate of immigrants
is lower than that of native-born, particularly for
female immigrants. In the long run – regardless of
the LFPRs – higher volumes of immigration would
not only increase the working population but also
the population of non-workers too as immigrants
inevitably age, leave the labour force and require
social assistance, as do native-born workers.
Integration dynamics have a signicant impact on
the EU’s future labour force size. Consequently,
to maximise the possible impact of immigration
on lowering the LFDR, any increase in the
Immigration
volume
Educational
composition
of immigrants
Integration assumptions
Low Baseline High
Low
Low 1.44 1.41 1.39
Medium 1.44 1.41 1.39
High 1.43 1.41 1.39
Baseline
Low 1.47 1.34 1.28
Medium 1.46 1.33 1.26
High 1.43 1.30 1.22
High
Low 1.50 1.29 1.20
Medium 1.48 1.27 1.17
High 1.43 1.24 1.12
VALUE IN 2015 = 1.08
Table 3.3: Projected EU labour-force dependency ratio in 2060
Source: CEPA M
3: Impacts of migration on the EU labour force 42
volume of immigration accepted must come
with successful policies to improve immigrants’
access to the labour market, especially for female
immigrants. Otherwise, higher inows combined
with deteriorating participation rates could result in
a situation that is worse than that with medium or
low volumes of immigrants.
Finally, while a careful selection of immigrants
in terms of levels of education would only have
a limited impact on both the size of the labour

greatly for the human capital of the immigrant
population and thus for the productivity and
income levels earned by immigrants as a group.
Also, for other aspects of social, linguistic and
cultural integration that are beyond the scope of
this study, immigrants’ education level tends to
matter greatly because of a steeper learning curve
in adjusting to new conditions.
43 3: Impacts of migration on the EU labour force
4: Impacts of internal EU mobility over time 44
SUMMARY
Movement between EU Member States, or intra-EU mobility, has facilitated population
changes within the EU in recent decades. Pre-existing economic disparities between
Member States have encouraged many citizens to search for work in places other than
their country of origin, likely to the economic benet of the union as a whole, but not
necessarily for all sending Member States.
This section focuses on the potential for change in total population sizes over time,
leaving the topic of the education-selective nature of such emigration for Section
5. If ow patterns from the east and south continue, some Member States will see
signicant reductions in their total population. This raises important implications for
economic development and Cohesion Policy goals.
44
BOX 4: SCENARIOS FEATURED IN SECTION 4
Central scenario:   



-

No intra-EU mobility scenario:
Double intra-EU mobility scenario: This scenario doubles the intra-EU mobility rates used in the

45 4: Impacts of internal EU mobility over time
IMPACTS
OF INTERNAL
EU MOBILITY
OVER TIME
 Intra-EU mobility and EU Member
State populations
Between 1991 and 2015, the population of the
current EU Member States increased from 476.8
to 508.5 million. As map 4.1 illustrates, in spite of
low fertility, western and southern
14
EU Member
States experienced marked population growth
as they attracted immigrants from within and
outside of the EU. In contrast, Eastern Member
States experienced population decline due to
a combination of very low fertility and emigration.
For example, the Baltic States and Bulgaria lost
between 26% and 16% of their population, intra-
EU mobility being one of the drivers of this decline.
Intra-EU mobility has played an important role
in population change in Member States over
the past decades. In 2017, over 3% of EU citizens
(16.9 million) resided in a country other than their
country of citizenship (Eurostat 2018). Since 2004,
migration among the current Member States
increased following the accession of 13 countries,
although transitional restrictions imposed by some
of the older EU-15
15
Member States have aected
the volume and direction of ows between individual
Member States. In addition, evidence shows that
intra-EU mobility stabilises aer an initial rise
following a country’s accession (de Haas 2018).
Nonetheless, an estimated 1.8% of the population
in the eastern Member States that joined the EU
in 2004 moved to the EU-15 between 2004
and 2009, rising to 4.1% in Bulgaria and Romania
between 2007 and 2009 (Fic et al. 2011). EU-15
populations grew by an estimated 0.4% and 0.3%
due to immigration from new Member States during
the period (Fic et al. 2011).
Separate projections of both international
migration to/from the EU and intra-EU mobility
are necessary to assess the possible impact
45
If internal migration
ows to western
Member States
continue, some in
the south and east
will see signicant
declines in their
populations.
of intra-EU mobility on future population changes
in EU Member States. Such modelling was
not normally done in previous EU population
projections, the only similar projection for the EU
being Bijak et al. (2007).
16
Assessment
of bilateral ows between the Member States
is still particularly challenging due to large
discrepancies in reported in- and outows between
the countries. In spite of signicant eorts
by Eurostat to improve data quality, there are
signicant variations in denitions, data-collection
systems and gaps in information exchange
between the national statistical oces.
17
The data
which were used to compute the mobility rates
for the modelling of bilateral ows between the EU
Member States already include the post-nancial
crisis years when return migration from the eastern
Member States picked up. Thus, return migration
is included in the overall rate although we do not
model it separately. As a consequence, additional
modelling is needed to arrive at integrated,
harmonised bilateral ows between the Member
States. For the purposes of CEPAM, taking into
account the large discrepancies of in-outows,
we estimated intra-EU mobility bilateral ows for
the period 2009-2016 using the method developed
by Raymer et al. (2013).
18
The purpose of this section is to assess the impact
of intra-EU mobility on the EU population. To do
so, three scenarios were developed: the central
scenario, the no intra-EU mobility scenario,
and the double intra-EU mobility scenario.
19
 Intra-EU mobility scenarios
Keeping international migration into the EU
constant
20
and varying only the magnitude
of intra-EU mobility, the central scenario
is juxtaposed against the no intra-EU mobility
and double intra-EU mobility scenarios. The no
intra-EU mobility scenario is used to illustrate
the impact of intra-EU mobility on sending and
receiving countries. We hypothesise that,
Map 4.1: Population change in EU-28 Member States between 1991 and 2015
Source: Eurostat, CEPAM - Created with mapchart.net©
4: Impacts of internal EU mobility over time 4646











non-EU
in the long-term, thanks to greater convergence
toward the living standards found in the most-
developed Member States, greater cohesion,
diminishing regional disparities and stronger EU
integration, it is likely that intra-EU mobility may
slow down from the south and east of the EU
to the west. Therefore, by 2060, it is likely that
the impact of intra-EU mobility would be smaller
than that projected under the central scenario,
although still larger than in the no intra-EU mobility
scenario. In contrast, the double intra-EU mobility
scenario assumes that the deepening gap between
the economically stronger and weaker Member
States can lead to increased intra-EU mobility.
 The impact of intra-EU mobility
on the EU population
Intra-EU mobility has the biggest impact
on the past and projected population change
in the eastern Member States, where it exacerbates
their population decline. The magnitude of this
impact is illustrated in map 4.2 and gure 4.1, which
depict relative changes in population size between
2015 and 2060 according to the dierent scenarios.
For example, the population of Romania would shrink
from 19.9 million in 2015 to 13.8 million in 2060 in
the central scenario (losing 30% of its population).
However, the loss would be much lower in the no
intra-EU mobility scenario, which is 14% of the
2015 population (red bar in gure 4.1).
In the case of the double intra-EU mobility
scenario, Romania could lose about 40% of its
population by 2060 (orange bar in gure 4.1). Since
most migrants are young adults, the direct eect
of higher intra-EU emigration on the population
change of a sending Member State is the loss
of the mainly working-age population. The indirect
eect is fewer new families because the number
of the country’s potential parents is smaller than
would be the case with less-intense emigration.
The relative adverse eect of intra-EU mobility
is more pronounced in less populous countries.
Despite large ows in absolute terms from,
for example, Poland and Spain, gure 4.1 illustrates
that the eect on their population size is actually
less dramatic, although still sizeable.
Of all the Member States showing a projected
declining population, Slovenia is the only one where
a no intra-EU mobility scenario would completely
prevent a population decline (see gure 4.1).
On the contrary, Germany, which is a net gainer
from intra-EU mobility, would face a negative
population change should intra-EU mobility stop.

in age structure, Germany’s population would
decline without arrivals from other Member
States (intra-EU mobility) even with sustained
high international migration.
Among net receiving Member States21 , the impact
of immigrants from other Member States
is sizeable in Austria and the UK, although in most
other Member States the changes in volume of the
intra-EU ows do not make much dierence to the
population growth.
The outow of people from the eastern and
southern Member States contributes to an increase
of population concentration in western Member
States (countries depicted in red in map 4.2).
In 2015, 54% of the EU’s 508 million lived in
western Member States. Depending on the volume
of intra-EU mobility in the future, and in case
of sustained increases in international migration
to/from the EU, this proportion would rise
to 59-61% by 2060 according to the three
scenarios. In absolute terms, the population
of western Member States would increase by 30
million in the no intra-EU mobility scenario (due
to gains from international migration) and by 50
million in the double intra-EU mobility scenario
(due to larger gains from intra-EU mobility).
47 4: Impacts of internal EU mobility over time
47
Map 4.2: Relative population change between 2015 and 2060 (projection) in EU Member States
(a) Central scenario
Source: Eurostat, CEPAM - Created with mapchart.net©
4: Impacts of internal EU mobility over time 4848

Central scenario
Map 4.2: Relative population change between 2015 and 2060 (projection) in EU Member States
(b) No intra-EU mobility scenario
Source: Eurostat, CEPAM - Created with mapchart.net©

No intra-EU mobility scenario









non-EU










non-EU
In contrast, only between 15-18% of EU residents
would reside in the eastern Member States
compared to 20% in 2015 (falling from 103 million
to 82-93 million).
 Conclusions
Intra-EU mobility is driven to an important extent
by demand for labour and wage inequality. In spite
of strong economic growth in some eastern Member
States (largely driven by foreign direct investment),
convergence is relatively slow. For example, Slovenia
– the most advanced country in the group in terms
of economic performance – had a GDP per capita
(in purchasing power standards) at only 76%
of the EU-15 average in 2017, although up from
67% in 2000 (Economist Intelligence Unit 2018).
Although it likely beneted the EU as whole,
unusually large emigration from the eastern
Member States has had negative eects
on the sending countries, contributing to lower
economic growth and thus slower convergence
to western EU Member States (IMF 2016). Gaps
in wages and living standards keep driving

implementation of EU cohesion policies, leading
to a loss of home-grown talent and innovation
as well as having consequences for inter-
generational replacement and population ageing
(Atoyan et al. 2016).
Policies targeting economic inequality between
Member States resulting in greater cohesion
and integration can help those countries facing
population decline, a loss of working-age population
and population ageing. But these policies also need
to pay special attention to skills and try to reverse
the education selectivity of emigration by oering
interesting employment opportunities to the highly
skilled and possibly facilitating return migration
for some of the talent that has le.
-50
-40
-30
-20
-10
0
10
20
30
40
AT DE UK BE NL FR IT GR HU ES PL PT EE SI BG HR LV RO LT
Population change 2060/2015 (%)
Central scenario Double intra-EU mobility No intra-EU mobility
49 4: Impacts of internal EU mobility over time
49
Figure 4.1: Relative population change (in %) between 2015 and 2060 (projection) in selected EU
Member States, by scenario
Note: Countries are ranked by the relative importance of the impact of intra-EU mobi lity on their populat ion (double intra-EU
mobility scenario – no intra-EU mobility scenario) from positive to negative.
Source: Eurostat and CEPAM
5: Illustrating the consequences of ‘brain drain’ 50
SUMMARY
The loss of talent to comparatively higher-income countries continuously confronts
some societies within the EU, and many others around the globe. This section
demonstrates at the EU level, what has already been seen in certain Member States
in terms of high emigration. The purpose of this demographic exercise is to illustrate
the relationship between high emigration and demographic variables, rather than
presenting a scenario to be viewed as a likely future for the EU as a whole. It can also
help demonstrate how the currently observed pattern of net migration inows should
not be automatically assumed to continue in the long term.
If the EU stagnated and started to see its highly educated citizens leave – on the scale
already observed in some Southern and Eastern Member States – it would nd itself
with a smaller and less-educated workforce. Such changes would coincide with a more
rapidly ageing population because emigrants tend to be early-career adults.
50
BOX 5: SCENARIOS FEATURED IN SECTION 5
High emigration scenario:
                  
                



Central scenario:  


Zero migration scenario:

51 5: Illustrating the consequences of ‘brain drain’
DEMOGRAPHIC
CONSEQUENCES
OF ‘BRAIN DRAIN’
 The hypothetical scenario of an EU
‘brain drain’
Just a century ago, Europe lost a substantial share
of its population as millions went in search of
a better life overseas. Between 1850 and 1913,
it is estimated that more than 40 million people
emigrated from Europe to the New World (Hatton
and Williamson 1994). Unlike what went before,
the EU has seen far more immigrants than
emigrants in recent decades and migration has
become a driver of population growth. Yet, not only
in the newer Members States but also in some
older ones – particularly Spain, Portugal and Greece
– emigration has exceeded immigration as many
young adults le these countries during the most
recent nancial crisis.



To illustrate this hypothetical – but not impossible
– situation, we developed a high emigration
scenario (see descriptions of scenarios in box 5).
The conditions of this scenario are that the EU
cannot keep pace with economic, scientic and
technological developments taking place in other
parts of the developed world, particularly in East
Asia and North America. Emerging economies also
become attractive and create robust alternatives
for potential immigrants.
As a consequence of this, immigrant ows
into the EU would lessen considerably as other
destinations become comparatively more attractive
or advantageous. For similar reasons, many
EU nationals would also decide to seek career
opportunities in North America, East Asia and other
more economically dynamic regions. This slow-
moving, long-term shi eventually returns Europe
back to its historical patterns of more people
emigrating than immigrating.
51
The loss of talent
to comparatively
higher-income
countries continuously
confronts some
societies within the
EU, and many others
around the globe.

high-emigration scenario
Steep decline in the EU population
This high-emigration scenario leads to a net
migration outow of 9.4 million (nearly the
population of Sweden) that would gradually lessen
to 7.8 million for the period 2055-2060 (greater
than Bulgaria’s population). Should this happen,
at these levels, the EU would start to rapidly
lose its population through a combination of
high emigration and below replacement fertility.
Therefore, as gure 5.1 demonstrates, the size
of the EU population would decline by 31% from
508.5 million in 2015 to 353 million people
in 2060 (gure 5.1, blue line), i.e. 155 million
less than in 2015. Such a rate
of population decline might seem extreme,
however, this is not unseen in the EU. Latvia
and Lithuania have lost 27% and 23% of their
populations, respectively, in a shorter period

conditions of the post-Soviet era and the
. By comparison,
in the zero-migration scenario, the EU population
would gradually reduce by 42 million people
(8% of its population size) by 2060 (zero
migration scenario, gure 5.1, orange line).
Figure 5.1: Population size of EU-28 in 2015-2060, by scenario
Source: CEPA M
300
350
400
450
500
550
2015 2020 2025 2030 2035 2040 2045 2050 2055 2060
Population in millions
Central scenario Zero migration High emigration scenario
Severe population ageing
The high-emigration scenario would also lead
to severe population ageing. Highest immigration
rates are currently observed among young adults
(peaking at 20-29 years old), who are more likely
to immigrate as they are looking for better work
and life opportunities and have a longer time span
to benet from the expected gains of migrating.
Therefore, sizeable and lasting emigration from
the EU would severely deplete the size of
the working-age population. Under the high
emigration scenario, the working-age population
would shrink by half from 306 million in 2015
to 150 million in 2060 (see gure 5.2) and its
share of the total population would drop from 60%
in 2015 to 43%. This is ve percentage points lower
than in the other scenarios (48%) respectively.
Simultaneously, as older people are less likely
to emigrate out of the EU, the share of the
population aged 65+ would more than double from
19% in 2015 to 41% in 2060. The zero migration
and central scenarios lead to dierent overall
population size (see gure 5.1) but a very similar
5: Illustrating the consequences of ‘brain drain’ 5252
Figure 5.2: Size and share of the working-age (dark green) and 65+ population (light green)
in EU-28 in 2015 and 2060, by scenario
Source: CEPA M
proportion of working age and population aged 65+
(gure 5.2). In the central scenario, the size
of the working-age population would decrease
by 17% by 2060 compared to its size in 2015,
whereas in the zero-migration scenario the size
would decrease by 27%.

As a consequence of the assumed ongoing
educational expansion (as discussed in Section
2), the share of the highly educated among the
working-age population increases in all scenarios
because retiring less-educated workers are
replaced by their more-educated and younger
counterparts. Thus, gure 5.3 shows the share
of post-secondary educated increases in all
scenarios. In the central scenario, the educational
composition of the working-age population
is aected by international migration, leading to
a higher share of low educated (5% more than
in the case of the zero-migration scenario) and
a lower share of highly educated (4% less than
in the case of the zero-migration scenario) in 2060.
This is due to the fact that, on average, immigrants
to the EU are less educated compared
to the host population.
In the high-emigration scenario, we double
the emigration rate of the highly educated
compared to the rate in the central scenario.
Highly educated workers thus have a much higher
propensity to emigrate compared to the less-
educated population. Figure 5.3 shows that this
scenario would lead to only 50% of the labour
force having post-secondary education in the
potential workforce, 10% less than would
be the case in the zero-migration scenario.
This scenario illustrates only one theoretically
possible case of an EU that is in decline.
An alternative pattern of migration may also
be considered resulting from a loss of economic
competitiveness with many highly skilled leaving
Europe and, at the same time, large numbers
of low-skilled migrants entering from Africa and
Western Asia. This would result in less population
decline and ageing, but an even stronger decline
in the average skills of the European labour force.
Size
Share
60
48 43 48
19
34 41 34
0
20
40
60
80
100
0
50
100
150
200
250
300
350
20-64 65+
Millions
%
2015 2060
Central
scenario
2060
High-
emigration
scenario
2060
Zero
migration
2015 2060
Central
scenario
2060
High-
emigration
scenario
2060
Zero
migration
53 5: Illustrating the consequences of ‘brain drain’
53
5: Illustrating the consequences of ‘brain drain’ 5454
Figure 5.3: Working-age population of EU-28 by educational attainment in 2015
and in 2060, by scenario
Source: C EPAM
24%
9%
6%
4%
44%
35%
43%
36%
32%
56%
50%
60%
2015
2060 Central scenario
2060 High emigration scenario
2060 Zero migration
lower secondary and below upper secondary post-secondary
 Conclusions
To sum up, this high-emigration scenario can help
illustrate at the EU level what some individual
countries (certain southern and eastern EU Member
States) have already experienced. The stagnant
EU that is hypothesised in the high-emigration
scenario would generate a rapidly decreasing and
ageing population which would be less educated
compared to the population projected under the
central scenario.
Such a scenario illustrates the detrimental eects
that the high emigration of highly skilled workers
can impose on a society. The lessons to be drawn
about the loss of talent are also applicable to other
regions of the world, in particular to developing
countries which are oen le with shortages in
critical professions due to the prospect of higher
pay from abroad.
Within the EU, large emigration of talent can have
a strong impact on the Member States and depress
economic growth, necessitating compensation by
increasing the productivity of the smaller workforce
and tailoring education to be responsive to local
job-market needs.
55 5: Illustrating the consequences of ‘brain drain’
55
PART B:
DEMOGRAPHIC
TRENDS AND
MIGRATION IN
AFRICA AND
WESTERN ASIA
In the same way that EU policies have global implications, the future

around the world. To address this, the report now assesses the demographic
outlook for regions of particular relevance to the EU due to factors
of proximity or migration. A clearer demographic picture of regions outside
the EU helps provide greater context and actionable solutions for
policymakers

resource management and migration policy, among others.
The scenarios detailed in 'Part B' bring to light just how seriously diverging
socio-economic development pathways can alter the course of world

element in accelerating development.
Pressure coming from continuous population growth jeopardises the welfare
and sustainability of many developing countries. This pressure interacts

it harder for people to maintain the ability and desire to build upon
opportunities at home. These considerations and the resulting migration
decisions have important implications for a third country’s accumulation
of human capital and resilience.
6: World population growth trajectories 58
SUMMARY
Many regions of the world have progressed towards late stages of the demographic
transition with both low rates of mortality and fertility, notably East Asia, North
America and Europe. World population may even eventually peak and start to reduce
slightly during the second half of the century, depending on how fast fertility levels
in Africa drop to moderate levels.
This section underscores the urgent need for expanding education in Africa. Education
expansion must keep pace with pressure coming from rapid population growth, as
it holds the key to accelerating the demographic transition and bringing development
successes within reach.
Achieving such goals depends on access to education for girls and young women
in particular, as education and family planning are closely intertwined. Education
broadens horizons and helps fertility enter the realm of conscious choice for women
and men. Evidence from educational sub-populations within countries indicate that
higher living standards and decisions for moderate fertility levels accompany higher
education and the associated wider range of life choices.
58
BOX 6: SCENARIOS FEATURED IN SECTION 6
The Shared Socioeconomic Pathways (SSPs) are global development scenarios for the rest of the century that


SSP1 (rapid social development):  
-


SSP2 (middle-of-the-road):-


SSP3 (stalled social development):



59 6: World population growth trajectories
WORLD
POPULATION
GROWTH
TRAJECTORIES

of Africa
The projections based on three scenarios – Shared
Socioeconomic Pathways (SSP)1, SSP2 and SSP3
(explained in box 6) – are illustrated in gure 6.1,
bringing to light the potential for large population
growth in Africa. According to SSP1, the world
population could be around 8.9 billion
by 2060 and Africa as a whole could reach 2 billion,
whereas under SSP3, global population could rise
to 11 billion and Africa’s share could increase
to 3.1 billion (gure 6.1a). It is clear, therefore, that
whichever scenario is projected, Africa’s population
is expected to increase more than that of the
rest of the world. Africa is the EU’s neighbour and
although it is currently acknowledged that most
migration in Africa takes place intra-regionally
(MEDAM, 2018), it cannot be discounted that
immigration ows from Africa to Europe are likely
to continue (European Commission 2018b), and
due to a number of factors (rapid growth in youth
population, increased education, increased wealth,
climate change, conict), more people might want
and be able to come to the EU. It is therefore
important to consider the projected population
growth of Africa as it could be a source of possible
future migration ows into the EU.
These 
be explained by those populations which
are at various stages of the universal process
of demographic transition. In this process, deaths
(in particular child mortality) initially start to
decline while births remain at high levels because
high fertility norms tend to be embedded in most
cultures. Hence, for a period, the combination
of high-birth and low-death rates result in rapid
population growth. It is only when birth rates also
fall that population growth diminishes. African
countries have entered this transition about
59
The education of
girls in Africa –
whether rapidly
expanded or stalled
– will be decisive
in determining the
future of world
population growth.
a century later than Europe and several decades
later than Eastern Asia, which is why population
growth is still very high in Africa and already low
in Europe. Thus, the future of population growth in
Africa will largely depend on the speed of fertility
decline which is closely associated with female
Figure 6.1a: Population projections for 2015-2060
(a) Projected world population (2015-2060)
Source: CEPA M
Figure 6.1b: Population project ions for 2015-2100
(b) Projected population of Africa (2015-2100)
Source: CEPA M
0
2
4
6
8
10
12
2015 2020 2025 2030 2035 2040 2045 2050 2055 2060
SSP1
SSP3
SSP2
Population in billions
0
0.5
1
1.5
2
2.5
3
3.5
4
4.5
5
2015 2020 20 25 2030 2035 2040 2045 2050 20 55 2060 2065 2070 2075 2080 20 85 2090 2095 2100
SSP1
SSP3
SSP2
Population in billions
6: World population growth trajectories 6060
education and contraceptive use. Therefore, this
section will look at how raising education levels in
Africa, especially among girls, will aect birth rates
and consequently population growth.


The series of Demographic and Health Surveys
(DHS) provide high-quality data on fertility for
most developing countries derived from large
numbers of individual interviews that also collect
key background characteristics for women. Of
all the socio-economic variables measured,
female education shows the strongest and most
consistent relationship with fertility. Figure 6.2
shows the dierentials in fertility for six categories
of educational attainment, collected from 58
developing countries with multiple surveys from
some countries. The dotted lines show the curves
traced through the data for each education level
and display an almost perfect ordering with
uneducated women having the highest fertility
rates and women with post-secondary education
having the lowest.
There are also good reasons to believe that the
eect of education on fertility is in fact causal
because rst, it brings fertility within the calculus
of conscious choice (as suggested by A. Coale 1975);
secondly, more-educated women tend to want
fewer children; and thirdly, better-educated women
are more empowered to access contraception
and actually have fewer children despite possible
resistance from husbands or family.
22
Figure 6.1c: Population project ions for 2015-2060
Projected population growth by region (2015-2060)
Source: CEPA M
-40%
-20%
0%
20%
40%
60%
80%
1
00%
1
20%
1
40%
1
60%
1
80%
Europe Africa East Asia South Asia Rest of Asia Latin
America
Caribbean
North
America
Oceania
SSP1 SSP3 SSP2
61 6: World population growth trajectories
61
Figure 6.2: Total fertility rates (TFR) for six different levels of female education
for multiple DHS surveys from 58 developing countries (most recent survey data for each country, with the x-axis giving
the average fertility level and the coloured dots indicating education-specic levels in that country)
Source: CEPA M
0
1
2
3
4
5
6
7
8
9
1 2 3 4 5 6 7 8
Education-specic TFR
Total fertility rate (TFR)
No education
Incomplete primary
Complete primary
Lower secondary
Upper secondary
Post-secondary
6: World population growth trajectories 6262

education determines the future

Until the 1970s, fertility in sub-Saharan Africa
remained very high with an average TFR of
around seven births per woman (Bongaarts and
Casterline 2013), and dierences between regions
and countries were relatively modest. Over the
past quarter century, however, signicant fertility
declines have occurred, particularly in Eastern and
Southern Africa, while fertility remains at pre-
demographic transition levels in some countries in
Western and Central Africa.
As a result, dierences among African countries
have grown large over time. In 2015, the TFR in
the African region ranged from 6.8 births per
woman in Niger to 1.5 in Mauritius.
23
Except
for a few countries in Southern Africa, current
fertility rates in sub-Saharan Africa are among
the highest in the world, making it much harder to
reach development goals.
African countries are latecomers in the process
of demographic transition compared to Asia
or Latin America
24
, and some could remain
trapped in low-education and high-fertility rates
longer than others. As argued by Bongaarts
and Casterline (2013), high fertility levels in
Africa are also a consequence of parents having
higher desired family size and thus having less
motivation to adopt birth control.
25
In this context,
there is a controversial discussion about African
exceptionalism
26
to be explained by cultural
factors which also manifests itself via objections
from husbands or other family members about
contraception and concerns about the moral
and social acceptability of family planning.
Whatever the specic cultural embeddedness of
reproductive decisions, as mentioned above, the
empowerment of women through education can
63 6: World population growth trajectories
63
strongly contribute to overcoming such resistance
and enable them to eectively pursue their own
interests, which tends to result in lower fertility.
To see the eect educational attainment can have
on fertility rates, we will now look at the scenarios.
Figure 6.1 shows that if there was little further
progress in education, the associated higher
fertility rates in African countries would generate
a world population size 15%
27
larger (for the SSP3
scenario) than under the SSP2 scenario, whilst
the rapid expansion of education associated with
a more rapid fertility decline (SSP1) would generate
a population 10%
28
lower than that in the SSP2
scenario. Although dierences in population
projections reect assumptions about all
demographic components, these dierences
in future of population growth are to a large extent
due to dierent future education trajectories.
The population pyramids (in gure 6.3) display
the distribution by age, sex and level of education
of the African population in 2060, under the rapid
development scenario (SSP1), the medium (SSP2)
and the stalled development scenario (SSP3),
illustrating the potential impact of investments
in social development, in particular female
education, on future population dynamics.
29
As shown by the SSP3 population pyramid, Africa
is in danger of becoming less educated over time
if education expansion cannot keep pace with
population growth.
For
development
to succeed,
the expansion
of education
must keep pace
with rapid
population
growth
Figure 6.3: Scenario SSP1 - Rapid Development Scenario - Age and education pyramids for Africa in 2060,
according to three contrast ing scenarios; the colours indicate the level of educat ional attainment
Source: CEPA M
200 000 150 000 100 000 50 000 0 50 000 100 000 150 000 200 000
0-4
5-9
10-14
15-19
20-24
25-29
30-34
35-39
40-44
45-49
50-54
55-59
60-64
65-69
70-74
75-79
80-84
85-89
90-94
95-99
100+
Scenario SSP1
African countries, 2060
Under 15 No education Incomplete primary Completed primary
Lower secondary Upper secondary Post-secondary
Men Women
450
2020
Central scenario
Age in years
Population in thousands
200 000 150 000 100 00 0 50 000 0 50 000 10 0 000 150 000 200,000
0-4
5-9
10-14
15-19
20-24
25-29
30-34
35-39
40-44
45-49
50-54
55-59
60-64
65-69
70-74
75-79
80-84
85-89
90-94
95-99
100+ Men Women
Under 15 No education Incomplete primary Completed primary
Lower secondary Upper secondary Post-secondary
Age in years
Population in thousands
6: World population growth trajectories 6464
Figure 6.3: Scenario SSP2 - Central Scenario - Age and education pyramids for Africa in 2060, according to three
contrasting scenarios; the colours indicate the level of educational attainment
Source: CEPA M
Figure 6.3: Scenario SSP3 - Stalled Development Scenario - Age and education pyramids for Africa in 2060,
according to three contrast ing scenarios; the colours indicate the level of educat ional attainment
Source: CEPA M
 Conclusions
This section looked at Africa's future population
growth under three global development scenarios
following the SSP narratives. It discussed how
female education can help reduce fertility rates
and illustrated how alternative education and
social development scenarios aect population
growth as well as the make-up of a population’s
future age and educational attainment. While
population policies tend to be a controversial
topic, a clear policy focus on female education
is much less controversial.
In terms of the consequences of population growth,
there can be little doubt that rapid population
growth makes it much more dicult to expand
essential services such as education and health,
which in turn are essential for social and economic
development. Improvements in reproductive
health and female education also work together
synergistically in reducing the desired family
size and making it easier to access eective
contraception, resulting in lower fertility.
Looking at the scenarios, it becomes apparent that
a policy focus on education has the double eect
of simultaneously reducing fertility and increasing
skills and thus economic productivity. This could
create a virtuous circle of women’s empowerment,
better health, lower fertility and economic growth,
putting human capital investment at the centre
of global development strategies.
65 6: World population growth trajectories
65
Men Women
200 000 150 000 100 000 50 000 0 5 0 000 100 000 150 000 200 000
0-4
5-9
10-14
15-19
20-24
25-29
30-34
35-39
40-44
45-49
50-54
55-59
60-64
65-69
70-74
75-79
80-84
85-89
90-94
95-99
100+
Under 15 No education Incomplete primary Completed primary
Lower secondary Upper secondary Post-secondary
Age in years
Population in thousands
7: Reasons for migration in Africa and Western Asia 66
SUMMARY
Africa’s population will increase by a factor of two to three, or possibly even more over
the coming decades in the event of stalled socio-economic development. In addition,
population growth in Western Asia will also be signicant. For example, the population
of Afghanistan is likely to triple and that of Pakistan to almost double. Such massive
projected population increases raise questions about future job opportunities, economic
growth, as well as the peace and stability necessary to facilitate such opportunities.
Furthermore, the already unavoidable climate change is likely to further complicate
these challenges.
The insecurity and conict arising from worsening economic and environmental
conditions act as key push factors for potential migrants. These factors are associated
with a perceived lack of opportunity to thrive in one’s own country, reected in youth
unemployment and a lack of suitable jobs. Pursuing policies that enhance local
employment opportunities and foster stability and security are critical to building
opportunities in the countries of origin.
66
67 7: Reasons for migration in Africa and Western Asia
REASONS
FOR MIGRATION
IN AFRICA AND
WESTERN ASIA
 
as push-factors
This section looks at the factors confronting Africa
and Western Asia, which are critical in creating
possible push factors for outmigration: conict,
unemployment and instability. Such political and
economic factors are much more volatile and
harder to forecast than the demographic and
human capital trends discussed in the previous
section which are backed by greater inertia
and predictability. Thus, this section focuses on
recent developments and current socio-economic
conditions in Africa and Western Asia that may give
rise to future migration. It is beyond the scope of
CEPAM’s research to be able to make projections on
alternative unemployment rates, conict scenarios
or other political developments over the coming
decades that may cause future migration ows to
the EU1. Climate change, however, is another more
slowly evolving and more predictable change, which
will be explicitly discussed in Section 8. Here we will
only consider it as a possible trigger of conict.
Conicts can cause multiple negative consequences
for human well-being, which can be important
reasons for people deciding to leave their home
country. These consequences range from possible
premature death and injury to psychological
trauma, destruction of housing and loss of
livelihoods. There is abundant evidence that such
experiences or fear of such experiences can be
a major push factor for outmigration.
In parts of Western Asia and Africa recent escalation
in conicts, terrorist attacks and increasing forms
of organised violence have forced people to move
both within their countries and abroad in search of
security and peace. The intensity of the conict and
the location of the ghting can explain an essential
part of the variation in ows of asylum applications
and refugees, both inside and outside the EU.
67
Insecurity, conict
and population
growth interplay
to degrade life
prospects, pushing
migrants away
from building
a future locally.
Results suggest that people ee terror and war,
as well as violence and insecurity emerging from
non-conict-aected areas and areas perpetrated
by dierent criminal groups (Conte and Migali 2018).
Figure 7.1 presents a recent estimate of global ows
of asylum seekers by world regions for the ve-year
periods 2006-2010 and 2011-2015 (Abel 2018).
The arrows indicate the direction of ows, while
the thickness of the arrows shows the volume
of ows. Indeed, this comprehensive estimate
of global bilateral ows conrms the view that
Western Asia and Africa have been the most
important regions of origin for refugees while
Western Europe is the most important destination
region. Interestingly, the pattern changes somewhat
between the two periods: in 2011-2015, refugee
movements from Eastern to Southern Africa
dominate the picture and are much larger than

to Western Europe. The regional denitions used
in this analysis do not directly allow for identifying
the ows into the EU, which mainly comprise
Western and Northern Europe (only receiving
countries of refugees), as well as parts of Eastern
and Southern Europe (both receiving and sending
countries).
This recent study (Abel et al. 2018) conducted
in the context of CEPAM also tried to link conict-
induced migration to several triggering factors,
including spells of drought caused by climate
change. Climate contributed to the outbreak of
conicts in Western Asia from 2010-2012, where
political uprising occurred in countries including
Tunisia, Libya and Yemen, and Syria. In Syria,
particularly, long-running droughts and water
shortages caused by climate change resulted
in repeated crop failures, with rural families
eventually moving to urban areas. This in turn led
to overcrowding, unemployment and political unrest,
and then civil war. Similar patterns were also found
in sub-Saharan Africa during the same period.
However, it is hard to draw generalisations from this
trend because the ndings of this study suggest that
the impact of climate on conict and consequently
on asylum-seeking ows is limited to specic time
periods and contexts.
Figure 7.1: Asylum seeking ows by world region, 2006-2010 (left) and 2011-2015 (right)
Source: Abel 2018
7: Reasons for migration in Africa and Western Asia 6868
 Youth unemployment as a driver
of outmigration
Sub-Saharan Africa (SSA) is currently the youngest
region in the world and has the largest potential
increase in domestic labour supply. Under the SSP2
scenario (Lutz et al. 2018), in SSA, the share of the
working-age population is projected to increase
over the coming decades in all subregions, with the
percentage change ranging from 4.7% in Southern
Africa to 31% in Central Africa between 2015
and 2060. Educational attainment in SSA is also
expected to increase over time in all subregions,
although it would still remain lower than
educational attainment in the rest of the world. For
the North Africa sub-region, by 2060, 77% of the
population would have a post-secondary or higher
education and those with no formal education
would be less than 5%. For the Middle East, 64%
of the population would have a post-secondary
education or higher, against 6% with no formal
education (Lutz et al. 2018).
In recent years, most countries in the SSA region
have experienced high and sustained economic
growth, although starting from a very low level. Six
of the ten fastest-growing economies in the world
are from this region (Samans and Zahidi 2017) and,
according to the World Bank’s latest estimates, GDP
growth is expected to strengthen to 3.1% in 2018
and 3.5% in 2019 (World Bank 2018a).
When studying the relationship between economic
growth and outmigration, however, it is important
to note that it is not linear. Under extremely poor
conditions, people cannot aord to move to another
country. Although South Sudan is currently one of
the poorest countries with the most insecure living
conditions, hardly any migrants from South Sudan
make it to Europe. They simply cannot aord the
cost of travel which may include excessive fees
for trackers.
More generally, there is a well-established pattern
whereby, in the early phases of development
and economic growth, outmigration rates tend
to increase as some people get the means
and a desire to actually leave the country. As
development continues, however, outmigration
rates typically fall again, unless conict and huge
unemployment bring additional hardship. It thus
seems to be to less the objective dierence in
wages between countries of origin and destination
that drives migration, rather the perceived
opportunities for a better future in one’s own
country. If these opportunities for the future look
very bleak, it can be a reason to leave; if they
obviously look better than the current conditions
and people see a brighter future for themselves
and their families, they may choose to stay.
Despite the good news about economic growth in
many SSA countries, the labour market is generally
characterised as being largely informal
31
. Work
When
deciding whether
to migrate,
the objective
differences
in wages between
countries are less
inuential than
the outlook on
opportunities
at home
69 7: Reasons for migration in Africa and Western Asia
69
remains widespread in the informal sector, as
major supply-and-demand challenges hamper the
creation of formal jobs. On the labour supply side,
inadequate supply of human capital
32
is perceived
as a major obstacle to business expansion by
51% of rms in SSA, surveyed by the World Bank
Enterprise Survey (World Bank 2018b).
While employment in the formal sector has
increased in recent decades (especially in small
enterprises), this increase has not kept pace
with the population growth, leading to a growing
mismatch between supply and demand. The
growing number of young people entering the
labour force every year, combined with the limited
transformation of most economies into high-
productivity and non-agricultural jobs, could also
consolidate the informal sector as the main source
of jobs for young people, especially the least skilled
and least educated, in the near future.
Figure 7.2 shows the recent trend in youth
unemployment for selected countries and regions
in Africa and Western Asia. Actually, North Africa
has among the highest youth unemployment in
the world with over 30% in Tunisia and Egypt
in 2017. With continued population growth, the
serious challenge of massive youth unemployment
could further increase in the future. Figure 7.2
also shows that for the youth unemployment
rate SSA seems to fare better than North Africa
and Western Asia. However, this is also linked to
the way unemployment data are collected and
estimated. Low unemployment rates in SSA can
mask high poverty, and countries with higher levels
of economic development also tend to register
higher unemployment rates, as observed in South
Africa for example.
Unemployment also relates to the size of dierent
age groups in relation to others. There is a huge
body of literature on the so-called ‘youth bulge’
33
in Africa and Western Asia which refers to social
and political risks resulting from large numbers
of young adults with high aspirations, but low
economic opportunities and little possibility for
Figure 7. 2: Youth unemployment (% of total labour force aged 15-24)
Source: World Bank 2018
Data from database: World Development Indicators
7: Reasons for migration in Africa and Western Asia 7070
2006 2008 2010 2012 2014 2016 2018
Sub-Saharan Africa Middle East & North Africa Egypt, Arab Rep. Syrian Arab Republic Botswana
Gabon Tunisia South Africa Libya
0 %
10 %