ESRC Centre for Population Change
l
Briefing 18
l
March 2014
CPC
centre for population change
Do high-income or low-income immigrants
leave a host country faster?
Many people think of international migration as a one-way move, whilst in reality many immigrants only
move overseas temporarily. An important question is: Who leaves and who stays? For example, is it the
successful immigrants who leave whilst the unsuccessful ones, who might be relying on the generosity
of the welfare state in the host country, stay? And if so, who leaves faster: the successful immigrants or
the unsuccessful ones?
Key Points
• Lowest income immigrants tend to return to their home country the fastest, followed by the
high-income group.
• Low-income migrants returning faster can be interpreted as a result of failure; low income tends
to be the result of unemployment or very low earnings.
• High earners leaving is usually due to them successfully meeting their target savings or gaining
their planned skills.
• The concern that host countries have about being burdened by immigrant welfare seekers
is unfounded, as many immigrants leave after they become unemployed or earn no or lowincome.
• The idea that immigrants, particularly from poor countries, move especially to benefit from the
generosity of the welfare state is tenuous as many do not qualify for welfare benefits.
• The concern by less-developed countries about “brain drain” is exaggerated as migration might
actually lead to “brain circulation”.
Introduction
Although international labour migration tends to
be mainly motivated by economic reasons such
as higher wages and better job opportunities,
return migration tends to be more difficult to
understand. One theory is that some migrants
are target savers; they plan to migrate for a
determined period of time to save and then they
will return to their home country. Another theory
is that immigrants prefer their home country and
just migrate for a period of time to build up money
Improving our understanding of the key drivers and implications of population change
www.cpc.ac.uk
Do high-income or low-income immigrants leave a host country faster?
or skills before returning. Others argue that return
be examined separately from the labour market. For
migration is usually unplanned, driven by unexpected
example, unemployment triggers loss of earnings.
events, and is a sign of failure. So, if migrants are
This research addresses the effect of income on
target savers, it might be expected that the more
migration duration in a new way that considers the
income they earn, the faster they will leave a host
changing nature of income experienced by migrants.
country. On the other hand, if immigrants leave very
It takes into account the connection between the
early after their arrival, this might suggest they have
potential reverse causality between the migrant’s
been unsuccessful in finding employment or earning
labour market status and their decision to return to
an income and have had to cut their losses. This then
their home country.
begs the question:AnDo
high-income
or has
low-income
interesting
issue that
been understudied Although
is the relationship
betweenconsiders
how long aall immigrants who
the research
migrant stays in a host country and their income while they are there. Although it is generally
immigrants leave faster?
in between
the Netherlands
from
developing countries
agreed that migration is driven by the differencearrived
in wages
the host and
home
An interesting issuecountries,
that has the
been
understudied
is
the
(LDC)
between
1999-2007,
given
the potential
effect of changes in wages (or income) in the decision to return to a home
variation
between
countries
of
origin,
it
relationship between
how
long
a
migrant
stays
in
a
country is less clear. Migrants would, on the one hand, like to extend their stay overseas as ahas focussed
response
to higher
wages;
other hand, theongain
staying
longer abroad
decreases.
host country and their
income
while
they on
arethethere.
fivefrom
main
countries
of labour
immigration to the
As
a
consequence,
higher
wages
abroad
may
have
a
positive
or
a
negative
effect
on
migration
Although it is generally agreed that migration is driven
Netherlands, namely India, Turkey,
China, South
duration.
by the difference in wages between the host and home
Africa and Morocco. As seen in Table 1, almost 19
countries, the effect of changes in wages (or income)
per cent of recent labour immigrants from developing
Study
in the decision to The
return
to a home country is less
countries came from India and 10 per cent from
Theon
research
uses
administrative
data from the
Netherlands,
all immigrants
whorepresented
clear. Migrants would,
the one
hand,
like to extend
China.
Labourobserving
immigrants
from Turkey
have
entered
the
country
between
1999-2007,
their
motive
for
migration:
whether
for
labour
their stay overseas as a response to higher wages; on
11 per cent and those from Morocco only three per
migration or not, the timing of return and the exact detailed information on their labour
the other hand, themarket
gain from
staying
longer
abroad
as the majority
of immigrants
from
status and income. Given the diversitycent,
of immigrants’
background,
the analysis
is these two
decreases. As a consequence,
higher
wages abroad
countries
tendsince
be family
migrantsofrather
limited to labour
immigrants
from developing
countries,
the behaviour
those than labour
may have a positiveimmigrants
or a negative
effect on
immigrants. Finally, eight per cent of immigrants came
is paramount
formigration
policymakers.
duration.
from South Africa. The distribution of income group
Of course, a migrant’s income is highly connected
to their
experience.
Income and China
shows
that labour
labourmarket
migrants
from Morocco
The study
cannot be examined separately from the labour market. For example, unemployment triggers
more often start with low paying jobs, while Indian and
loss of earnings. This research addresses the effect of income on migration duration in a
The research uses administrative data from the
migrants
are over-represented
in highnovel way that takes into account the changingSouth-African
nature of income
experienced
by migrants. It
Netherlands, observing
all
immigrants
who
have
jobs. reverse causality between the
takes into account the connection between paying
the potential
entered the countrymigrant’s
betweenlabour
1999-2007,
their motive
market status
and their decision to return to their home country.
Methodology
for migration: whether for labour migration or not, the
Although
the research
all immigrants
who analysis
arrived in
the Netherlands
from
Duration
is used
for the estimation
of return
timing of return and
the exact
detailed considers
information
developing
countries
(LDC)
between
1999-2007,
given
the
potential
variation
between
migration
for
two
reasons.
First,
duration
analysis
on their labour market status and income. Given the
countries of origin, it has focussed on five main countries of labour immigration to the
focuses on the timing of the return decision and not
diversity of immigrants’
backgrounds,
the Turkey,
analysis
Netherlands,
namely India,
China, South Africa and Morocco. As seen in Table 1,
just
on developing
whether it countries
happened
or not.
duration model
is limited to labour
immigrants
from
developing
almost 19 per cent of recent labour immigrants
from
came
fromAIndia
takes
into
account
such
a
change
in
intensity
to leave.
countries, since theand
behaviour
of those
immigrants
10 per cent
from China.
Labouris immigrants from Turkey represented 11 per cent and
Second,
along
with
the
migration
decisions,
other
those from Morocco were only three per cent as the majority of immigrants from these two
paramount for policymakers.
countries tend be family migrants rather than relevant
labour immigrants.
Finally,
eight
per
cent
of
characteristics of the individuals may also
Of course, a migrant’s
income
is highly
connected
immigrants
came
from South
Africa. The distribution
of income
shows
thatmarket
labour status and
change over
time, group
like the
labour
to their labour market
experience.
Income
cannot
migrants
from Morocco
and China
more oftenthe
start
with
low
paying
jobs,
while
Indian
and
migrant’s income.
South-African migrants are overrepresented in high-paying jobs.
Table 1 – Characteristics of LDC labour immigrants at time of arrival
Table 1: Characteristics of LDC Labour Immigrants at time of arrival
India
Turkey
China
South Africa
Morocco
monthly income (%)
<€1000
20.6
33.1
42.5
21.7
56.5
€1000- €2000 17.5
27.2
35.2
20.6
25.4
€2000- €3000
25.1
24.5
10.7
21.5
9.6
€3000- €4000
17.5
7.1
4.3
13.1
3.5
€4000- €5000
8.9
2.9
2.1
6.6
1.2
€5000- €6000
3.0
1.2
`.5
3.9
0.2
> €6000
7.5
4.1
3.8
12.6
3.7
Share in total LDC
labour immigrants
18.7
10.6
9.9
7.5
2.8
ESRC Centre for
Population Change • Briefing 18
2 Do high-income or low-income immigrants leave a host country faster?
Main findings
Policy implications
This research shows that the over-concern that host
The empirical results show that return migration
countries often have about being burdened by welfare
numbers are U-shaped with respect to income,
seekers is actually unfounded, as many immigrants
implying a higher amount of returns in low- and highin fact leave after they become unemployed or earn
income groups. Indeed, the findings suggest that the
no or low-income. Many immigrants do not qualify
low-income group has the highest number of returns.
automatically for any welfare benefits. Hence, the idea
The evidence suggests that the lowest income
that immigrants, especially from poor countries, move
immigrants tend to return to their country the fastest,
especially to benefit from the generosity of the welfare
followed by the high-income group. The fact that lowMethodology
state is tenuous. Another implication is that when
income
migrants
return
faster
can
be
interpreted
as
Duration analysis is used for the estimation of return migration for two reasons. First,
more
successful
a resultanalysis
of failure.
Low
tends
be the
resultand not just
duration
focuses
onincome
the timing
of thetoreturn
decision
on whether
it immigrants return to their country of
thetoconcern
of unemployment
or very
low takes
earnings.
On thesuch
other
happened
or not. A duration
model
into account
a change inorigin,
intensity
leave. by these less-developed countries
Second,
the migration
other
relevant
characteristics
of the
individuals
about
“brain
drain” is exaggerated as migration might
hand along
high with
earners
leaving decisions,
is usually
due
to them
may
also
change
over
time,
like
the
labour
market
status
and
the
migrant’s
income.
actually lead to “brain circulation”. This suggests that
successfully meeting their target savings or gaining
in today’s globalised world, those successful migrants
their planned skills.
Main findings
are able
to respect
circulateto and are not a lost investment for
The
empirical
show
that returnmigration
migrationdurations,
numbers are U-shaped
with
This
U shaperesults
is found
at different
their
country
of origin.
income,
higher amountthe
of returns
in lowand high-income
groups.
Indeed,
the
thoughimplying
in the aNetherlands,
amount
of returns
findings suggest that the low-income group has the highest number of returns. The evidence
There is no doubt that income is only one factor
decline after five to six years. Interestingly, comparing
suggests that the lowest income immigrants tend to return to their country the fastest,
affecting
returncan
migration,
though it is a key factor for
immigrants
the maingroup.
five origin
countries
(India, migrants
followed
by thefrom
high-income
The fact
that low-income
return faster
be
labour
migrants.
Return
migration
is also determined
China,
Turkey,
South
Africa
and
Morocco)
brings
interpreted as a result of failure. Low income tends to be the result of unemployment or very
low
earnings.
On
the
other
hand
high
earners
leaving
is
usually
due
to
them
successfully
by other factors in the host country as well as in the
up consistent evidence of this U shape relationship
meeting
theirincome
target savings
or gaining
country of origin. Even though labour migrants might
between
and return,
withtheir
theplanned
lowestskills.
income
be predominantly swayed by economic factors, there
group having the highest amount, as seen in Table 2.
This U shape is found at different migration durations, though in the Netherlands, the amount
is no doubt
social, political and personal reasons
is consistent
having
bothInterestingly,
successful comparing
highofThis
returns
decline afterwith
five to
six years.
immigrants
fromthat
the main
also
play
a
role
in
the decision to return to their home
income
migrants
leaving
once
they
have
achieved
five origin countries (India, China, Turkey, South Africa and Morocco) brings up consistent
country.
their savings
whilst
at the income
same time,
evidence
of this or
U skills
shape targets,
relationship
between
and return, with
the lowest income
group
having
the highest
amount, returning
as seen inas
Table
2. This
having
low-income
migrants
a result
of is consistent with having both
reading
successful
high-income
migrants
leaving
once
they
have
achieved Further
their savings
or skills
their limited success.
targets, whilst at the same time, having low-income migrants returning
as
a
result
Bijwaard, G. of
E.,their
and J. Wahba (2014), ‘Do Highlimited success.
Income or Low-Income Immigrants Leave Faster?’
Table 2 – Average stay at the Netherlands (months), simulation results for 10 years,
Journal of Development Economics, 108: 54-68.
Table 2: Average Stay at the Netherlands (months),
main countries of origin
Simulation results for 10 years, main countries of origin
India
Turkey
China
South Africa Morocco
monthly income
29.5
42.8
33.2
50.1
50.5
< €1000
monthly income
47.7
60.8
52.1
70.3
69.9
€2000- €3000
monthly income
31.6
47.2
36.8
54.8
54.8
> €6000
Notes:
monthly
income refers to
time at arrival
Notes:
Monthly
income
refers to time at arrival.
Policy implications
This research shows that the over-concern that host countries often have about being
burdened by welfare seekers is actually unfounded, as many immigrants in fact leave after
they become unemployed or earn no or low-income. Many immigrants do not qualify
automatically for any welfare benefits. Hence, the idea that immigrants, especially from poor
countries, move especially to benefit from the generosity of the welfare state is tenuous.
Another implication is that when more successful immigrants return to their country of
origin, the concern by these less-developed countries about “brain drain” is exaggerated as
migration might actually lead to “brain circulation”. This suggests that in today’s globalised
3 ESRC Centre for Population Change • Briefing 18
Authors
Govert E. Bijwaard
(Netherlands Interdisciplinary Demographic Institute (NIDI))
Jackline Wahba
(University of Southampton, CPC)
Edited by Becki Dey
(University of Southampton, CPC)
ESRC Centre for Population Change
Building 58, Room 2043
Faculty of Social and Human Sciences
University of Southampton
SO17 1BJ
Office: +44(0)2380592579
www.cpc.ac.uk