Transcript Labour mobility and immigration (Borjas ch. + Schröder)

Labour market mobility:
 Geographic mobility
◦ Within a country
◦ International

Job mobility
◦ Intrafirm mobility (change of position)
◦ Interfirm mobility (change of employer)



The basic economic approach to all kinds of
mobility:
The worker moves if and only if the PV of lifetime utility is increased by mobility.

The worker moves if and only of the
(discounted) costs of moving are smaller than
the (discounted) gains.



Costs and gains involve BOTH economic
and social or psychological ones – the place
where one wants to live, relations with
people in a work-place or location.
We do expect economic gains and losses to
increase/decrease the probability of
moving.
We don’t expect each individual to move so
as to maximise (PV of) life-time money
income. (See ex. in Borjas pp. 324-325)


Voluntary (quits) and involuntary (layoffs)
Young people are more likely to do both
◦ Layoffs because of
 ”Last-in-first-out”
 Older workers have acquired more firm-specific human
capital
◦ Quits because
 They are looking for a better match
 They have not yet acquired so much firm-specific human
capital
 There is a selection – those who do not like to move/find
a good match early/are wanted by the employer will stay
long. This gives a negative relation between probability of
separation and tenure. (There are movers and stayers.)

Neither employers nor job-seekers have
perfect information about each other. Job
changes can be part of a processes of ”trialand-error” to get a better match.




10-12 percent of the labour force changed
employer during one year from the mid-60s
to the late 1970s. Slight downward trend.
Mobility decreased at the end of the 1970s
but increased again in the 1980s to about
10%.
Dropped to 6-7% during the 1990s.
There is more mobility when there are more
vacancies.




Between autumn 2007 and autumn 2008, out of
4,4 million employed, 900 000 had changed work
place or employer.
550 000 had changed both work place and
employer.
More young workers than older and more men than
women had changed work place and employer.
http://www.scb.se/statistik/_publikationer/L
E0001_2010K03_TI_02_A05TI1003.pdf



Studies indicate that the wages of jobchangers grow faster.
They also seem to gain a better work
environment
But a study which separates voluntary and
involuntary separation finds that only those
who quit increase their wages, not those who
are laid off.
Tenure in firms of different size
1st semester 2003
Years in firm (tenure)
Mean
Years with present tasks
Median
Mean
Median
Self-employed
12
10
13
10
Employees:
1-9
6
3
6
3
10 - 49
7
3
6
3
50 - 99
8
4
6
3
100 - 249
9
5
7
3
250 - 499
11
8
6
3
500 - 999
11
7
7
3
1000 -
12
8
8
4
All
10
6
7
4
Source: Statistics Sweden, SM UF 39 SM 0302



Increased from  3 to 4 percent in the late
1970s, was relatively stable but fell back to
about 3% in the 1990s.
Is not correlated with vacancies.
Over a long period, the share of mobility
which is internal has increased.

Probability of moving from A to B
◦ increases if economic opportunities improve at B
◦ decreases if economic opportunities increase at A
◦ increases if the costs of moving decrease.

Not only wages matter but also chances of
getting a job.





The larger the distance between states, the
less migration between them.
With larger wage differentials between
states there is more migration.
With increased employment opportunities in
the state of origin, there is less outmigration.
Older workers move less than young.
More educated workers move more than
less educated.
Partly only.
 Experience and skills acquired during the
period of migration can improve prospects
after returning.
 Information is imperfect.




Very high mobility across parishes just after
WWII. (Maximum of 10% in one year.)
Lower (about 7%) in the 1950s and early
60s.
Increase (to 8-9%) in the late 60s and the
first part of the 70s. Mainly from ”forest
counties” (Värmland, Dalarna, the North of
Sweden) to those with the largest cities
(Sthlm, Gbg, Malmö).
Back to 7% in the late 70s and 1980s, then
a gradual increase to 8-8 ½ .





The unemployment rate in a region is an
important determinant of migration in all
studies.
Differences in regional wage rates also have
an impact but the results vary more
between studies.
Nearly half of those who move are 20-34
years old. A large share 20-24.
About equal mobility among women and
men.
Less mobility of dual-earner households.




Cross-section surveys show that those who
have moved earn more.
But panel studies that follow the same
individual find much smaller effects.
Those who moved earned more before
moving too.
For some groups, particularly the younger,
there is an effect on earnings. For dual
earner households, the effect is almost
zero.
A single person (or a single earner family)
moves if that person’s gains from moving are
positive, ΔPV>0
 In a dual earner household, one person’s
gain, ΔPV1, may be positive and the other’s,
ΔPV2’negative.
 The family will move if ΔPV1+ ΔPV2>0
(assuming that members are indifferent to who
earns).

Two ”immigration regimes”
 Settlement (US, Canada, Australia)
 Guest worker (Germany, Switzerland)

Sweden doesn’t fit either model.
◦ More restricted labour market immigration than
”settlement” countries
◦ More equal rights and security for immigrants
than”guest worker” countries.




Basic rules (simplified)
Nordic citizens – no restrictions or permit
required to live and work in Sweden.
Citizens of EU/EES – no permits required to
live, work or study for 3 months. After 3
months, have to apply for residence permit
but not for work permit.
Others – work permit must be arranged
before arrival and requires written job offer
in Sweden. If more than 3 months,
residence permit also required. Students
have to be accepted and to have means of
support.


Family re-unification – close relatives
(spouse, child under 18 and others who
lived in the household) can join a person
with permanent right of residence. (A little
less strict for EU/EES citizens). New
marriages – a ”test period” of 2 years.
Refugees – according to UN conventions
asylum seekers have the right to have their
application tried. Reasons for asylum is
well-founded fear of persecution. In
Sweden, residence may also be granted for
other humanitarian reasons on an individual
basis.



Until the 1930s, Sweden had large
emigration. (1 million to the US.)
In 1950s & 60s, labour market immigration
dominated. Swedish employers invited
workers, from different parts of Europe.
Many from Italy, Greece and Yugoslavia but
the largest group was Nordic citiziens
(mostly from Finland).
From the 1980s, most immigrants have
been refugees or family members.
Immigrants from outside Europe increased
from 10% to 50% in 1990. The proportion
decreased in the early 90s and then
increased.




Until the 1970s immigrants had higher
employment rates than natives.
But immigrants were overrepresented in lowskilled, low paid jobs even after a long time in
Sweden and so were there children (Knocke,
2000, Jonsson, 2007)
Since the 1970’s immigrants have had lower
participation rates than natives (Ekberg,
1995, 1999)
The relative earnings of immigrants have
decreased (Aguilar and Gustafsson, 1994)

Proportion born outside Sweden:
(in per cent)
2005
Population aged 15-74
Population aged 25-54
Of those aged 15-74
Employed,
Less than upper secondary
school
Upper secondary
Tertiary education
2011
14,6
16,7
17,3
20,4
12,5
14,8
15,5
13,3
14,2
19,9
15,1
17,7
Average income of immigrants is lower than that of
natives.
There is an employment gap and a wage gap.
Labour market status 2011, age 16-64
Employment
rate
Unemployment
2005
Men born in Sweden
Men born abroad
Not in LF
2011
2005
2011
2005
2011
6,9
6,0
77,8
79,3
16,5
15,6
14,9
16,0
64,8
68,9
23,9
18,0
6,6
5,8
74,2
76,5
20,5
18,8
13,6
15,9
58,7
57,9
32,1
31,2
Women born in Sweden
Women born abroad




But both employment levels, skill level of job
and earnings varies a lot depending on:
Country of origin
Length of time in Sweden
Time of arrival in Sweden (labour market
conditions)
Men born in
Empl. Unempl
.
Women born in
Empl.
Unempl.
Sweden
86.6
3.5
Sweden
83.1
3.7
Born abroad
64.2
10.9
Born abroad
61.3
9.4
Nordic countries 69.6
5.5
Nordic countries
4.2
Western Europe 74.6
4.0
Western Europe 71.3
4.2
Eastern Europe 66.9
8.2
Eastern Europe 68.8
8.5
South Europe 67.3
10.6
South Europe 57.8
11.0
Middle East 57.2
16.0
Middle East 44.0
14.6
Asia 62.0
10.2
Asia 60.7
10.7
Latin America 71.0
Africa 56.7
9.3
14.1
Latin America 66.1
Africa 50.9
8.0
11.0
Women
Men
20 - 64 years
20 - 64 years
Born in Sweden
204,9
262,5
Born abroad
163,0
194,2
Other Nordic countries
208,9
247,7
EU15 except the Nordic
186,5
237,8
Europe except EU15 and
Nordic
168,0
207,2
Africa
114,9
163,8
North America
161,4
206,7
South America
162,6
190,8
Asia
102,5
140,4
Oceania
181,9
229,8
Source: Statistics Sweden
Women
EU/EES
Other Europe
Outside Europe
Sweden
Men
EU/EES
Other Europe
Outside Europe
Sweden
Earnings
% of Sw.
Wage % of Sw.
149 200
109700
85400
158500
94
69
54
20300
18700
17900
20100
101
93
85
194800
143400
115100
221400
88
65
52
24300
21000
20200
24400
100
86
88




Most important : The employment gap.
But there is an ethnic wage differential among
the employed too. The difference between
Swedish-born and those born in Northern
and Western Europe is negligable but nonEuropeans have lower wages.
After controlling for a number of factors there
is an adjusted immigrant-native wage gap
(15 % for men, 12% for women according to le
Grand and Szulkin, 2002)
Possible explanations that have been given:
1.
It takes time to settle in a new country and make
a ”match” on the labour market
2.
Human capital differences (education)
3.
”limited transferability of human capital” (lack of
”country-specific human capital”)
4.
Selection effects in migration and return
migration
5.
Statistical or taste discrimination (chapter 9)
6.
Slow and clumsy integration policies
Median yearly earnings
Yrs in
Sw
Abroad
Nordic
Born in
Sweden=231
EU15
exc
Nor
d.
Other
Eu
ro
pe
Africa
North
A.
South.
Am
Asia
Oceani
a
177,2
223,0
214,6
183,5
142,0
182,7
176,6
121,1
208,9
0-2
40,2
128,3
156,2
55,9
20,3
59,7
51,5
18,4
135,2
3-4
70,0
179,9
196,0
95,1
50,2
119,9
106,2
27,5
174,8
5-9
131,3
208,9
210,9
160,4
107,0
171,4
152,1
78,3
214,3
10-19
173,7
215,5
213,1
189,2
164,8
192,5
182,6
143,2
222,5
20-
211,9
227,9
223,7
207,5
198,1
221,9
192,1
173,9
245,3

If we compare the earnings/the
employment rate of immigrants with
different length of time in the country we
confuse
◦ The effect of time
◦ Differences between immigration cohorts
(education, ability)
◦ Differences in labour market conditions at the
time of arrival
Example; Immigrants to the US, Borjas pp 339-41
(wages)
Assume we observe wages at time T
Newly arrived immigrants earn X.
Immigrants arrived at T-10 earn Y
It seems that after 10 years earnings
have increased by Y-X
But if those arrived earlier earned Z at
time T-10 the real change over time is
only Y-Z
Y
Z
X
T-10
T
Women
120%
100%
80%
Series5
Series4
60%
Series3
Series2
40%
Series1
20%
0%
Native
Immigrants
Native men
Immigrant
Native
Immigrant
men
women
women
Men
120%
100%
80%
No info
post-sec 3+ yrs
60%
post-sec <3yrs
Upper sec.
40%
Not secondary
20%
0%
Total
Native
Nordic
EU27
Europé
Africa
N
S
America America
Asia
Oceania
120%
100%
80%
No info
post sec 3+ yrs
60%
post-sec < 3yrs
Upper sec.
40%
Not secondary
20%
0%
Total
Native
Nordic
EU27
Europé
Africa
N
America
S America
Asia
Oceania




But immigrants more often than natives have jobs below their
level of schooling.
For many people the first occupation in Sweden has a lower
status than the home country occupation. Later, upward
mobility in occupational status sets in but not to the home
country level. (Ekberg & Rooth, 2006, small sample of
Ethiopians/Eritreans, Chileans, Iranians, and
Romanians/Hungarians)
Immigrants from non-European countries who arrived before
age 16 also have a higher probability than natives of being
overeducated relative to their jobs. (Katz & Österberg,
forthcoming)
Immigrants, particularly, non-European have lower returns to
education than natives (Nordin, 2011, Katz & Österberg)
Men
120%
100%
Series5
80%
Series4
60%
Series3
40%
Series2
20%
Series1
0%
Finland
Estland
Frankrike
Grekland
Jugoslavien
Ryssland
women
120%
100%
Series5
80%
Series4
60%
Series3
40%
Series2
20%
Series1
0%
Finland
Estland
Frankrike
Grekland
Jugoslavien
Ryssland
Men
120%
100%
Series5
80%
Series4
60%
Series3
40%
Series2
20%
Series1
0%
Turkiet
Irak
Iran
Marocko
Women
120%
100%
Series5
80%
Series4
60%
Series3
40%
Series2
20%
Series1
0%
Turkiet
Irak
Iran
Marocko




Language skills are important but a study which included
grades in Swedish found that children of immigrants still
earned less. Makes better functioning of Swedish as second
language training extremely important.
”Cultural skills” are hard to define and measure and to
distinguish from discrimination. Can partly consists of
contacts.
A large share of job matches are found through informal
channels and this put immigrants and children of immigrants,
at a disadvantage(Behtoui, 2008)
”Accent” can lead to underestimate of competence (Rödin &
Özcan, 2011)




Are people who migrate those with the
highest or lowest skills?
Are the people who return migrate those with
the highest or lowest skills?
The Roy selection model.
Less relevant to refugee immigration.

Workers with skills lower than S
more likely to emigrate from
”Orange” to ”Purple” (dashed)
w

Workers with skills higher than S
more likely to emigrate from
”Orange” to ”Purple”
w
S
Skill level
S
Skill level