Folie 1

1
LectureSeminar
Migration, Integration, and Ethnic Conflicts
Hartmut Esser University of Mannheim Fakulty of
Social Sciences Mannheim Centre for European
Social Research April/May 2006
2
Program
3
1. Migration
2. Integration Basic Concepts and Debates
3. Acculturation The Case of Language
4. Education and Labor Market
5. A General Model of (Intergenerational)
Integration
6. Ethnicity and Ethnic Conflicts
4
Lection 4 Education and Labor Market
1. The Problem Ethnic Stratification
2. Education
3. Labor Market
4. Effects of Bilinguality
5
Lection 4 Education and Labor Market
1. The Problem Ethnic Stratification
2. Education
3. Labor Market
4. Effects of Bilinguality
6
general aim
to identify theoretical mechanisms and empirical
relations between language and structural
integration (education, labor market)
7
Two Examples
8
Language and Education
9
Educational career and Language (GSOEP, Alba,
HandlMüller 1994 odds more than elementary
school, Ref. German children)
1,4
1,2
1
0,8
0,6
0,4
0,2
0
10
Educational career and Language (GSOEP, Alba,
HandlMüller 1994 odds more than elementary
school, Ref. German children)
1,4
1,2
1
0,8
0,6
0,4
0,2
0
11
Educational career and Language (GSOEP, Alba,
HandlMüller 1994 odds more than elementary
school, Ref. German children)
1,4
1,2
1
n.s
0,8
0,6
0,4
0,2
0
12
Language and Labor Market
13
Remember
14
Integration and the Functions of Language
15
three functions
resource
part and precondition of control of valuable
resources (access to interactions, effective
communication, precondition for investments)
symbol
signalling of (latent) traits activation of
stereotypes and emotions reinforcement of
boundaries and identities
medium
(main) instrument to solve problems of
coordination and information flow
16
the three functions of language and the problem
of integration
17
the three functions of language and the problem
of integration
18
the three functions of language and the problem
of integration
19
the three functions of language and the problem
of integration
20
the three functions of language and the problem
of integration
21
the three functions of language and the problem
of integration
22
hence
L2
Bilinguality
and bilinguality
23
Main hypothesis concerning bilinguality
Bilinguality has effects on educational and labor
market success additionally to L2
24
why?
Education

• Cummins Interdependency-Hypothesis

• Zhou Transmission of values and social control

• MouwXie transititional familial
  communication

Labor market

• Globalization and transnationalism

• ethnic economies

• interethnic transactions (middleman)

25
hence
L2 Bilinguality
additional effect of L1 over and above L2
26
The full model
L2 (only) Bilinguality
27
Lection 4 Education and Labor Market
1. The Problem Ethnic Stratification
2. Education
3. Labor Market
4. Effects of Bilinguality
28
Theoretical Mechanisms
29
Determinants of educational success
Educational success
30
Empirical Relations AbilitiesEducational
Success
31
Reading Scores (PISA 2000)
32
Reading Scores PISA Study OLS-Coefficients
boldface significant results
33
Reading Scores PISA Study OLS-Coefficients
34
Reading Scores PISA Study OLS-Coefficients
35
Reading Scores PISA Study OLS-Coefficients
36
Reading Scores PISA Study OLS-Coefficients
37
Reading Scores PISA Study OLS-Coefficients
38
Reading Scores PISA Study OLS-Coefficients
39
Reading Scores PISA Study OLS-Coefficients
40
The explanation of ethnic differences
41
Reading Scores, PISA, (OLS Reference category
Germans)
40
20
0

-20

-40
-60

-80

-100

-120

-140
42
Reading Scores, PISA, (OLS Reference category
Germans)
40
20
0

-20

-40
-60

-80

-100

-120

-140
43
Reading Scores, PISA, (OLS Reference category
Germans)
40
20
0

-20

-40
-60

-80

-100

-120

-140
44
Reading Scores, PISA, (OLS Reference category
Germans)
40
20
0

-20

-40
Poles
-60
Russians

-80
GW
Turks

-100

-120

-140
45
Reading Scores, PISA, (OLS Reference category
Germans)
40
20
0

-20

-40
Poles
-60
Russians

-80
GW
Turks

-100

-120

-140
46
Reading Scores, PISA, (OLS Reference category
Germans)
40
20
0

-20

-40
Poles
-60
Russians

-80
GW
Turks

-100

-120

-140
47
Reading Scores, PISA, (OLS Reference category
Germans)
40
20
0

-20

-40
Poles
-60
Russians

-80
GW
Turks

-100

-120

-140
48
Reading Scores, PISA, (OLS Reference category
Germans)
40
20
0

-20

-40
Poles
-60
Russians

-80
GW
Turks

-100

-120

-140
49
Mathematics (PISA 2000)
50
(No Transcript)
51
(No Transcript)
52
(No Transcript)
53
(No Transcript)
54
(No Transcript)
55
(No Transcript)
56
(No Transcript)
57
(No Transcript)
58
(No Transcript)
59
if L2, no effect by L1
60
Reading L2 is the problem of Turkish children
CPH !
61
Transition to Hauptschule/Gymnasium
62
(No Transcript)
63
(No Transcript)
64
(No Transcript)
65
(No Transcript)
66
(No Transcript)
67
(No Transcript)
68
(No Transcript)
69
(No Transcript)
70
(No Transcript)
71
Hauptschule/Gymnasium


350





bivariat2
L1Fam
AgeI
LAge
EdPar
StatPar
Books

300

250

200

Polen

Russen

GA


Türken
150

100

50

0
72
Results Determinants of AbilitiesEducational
Success
Language spoken in family Age at immigration and
learning-age East Migrants Cultural capital
Turks Ethnic penalty Turks (only, if any)
73
Contextual Effects
74
Kristen 2002 Recommendation Gymnasium
and Ethnic Concentration Class
75
Kristen 2002 recommendation Gymnasium
76
Kristen 2002 recommendation Gymnasium
77
Kristen 2002 recommendation Gymnasium
78
Stanat 2002 School Effects PISA 2000
79
Stanat 2005 PISA 2000
80
Stanat 2005 PISA 2000
81
Stanat 2005 PISA 2000
82
Stanat 2005 PISA 2000
83
Stanat 2005 PISA 2000
84
Portes and Hao 2004 Interaction of individual
status and status school
85
high status schools increase differences between
high and low status children
low status schools decrease differences between
high and low status children
ASES
scores
low
middle
high
SES
86
Explanation by the Investment Model
87
general theoretical model (see lecture 1 and 3)
investment (L2-reading, math, grades ) if
U(L2) U(L1)
gt C(L2)/p(L2)

(U(L2) U(L1))p(L2)
C(L2) gt 0
(Motivation Efficient Exposure) Cost
gt 0
(SES ASES)
SDistance gt 0
Mechanisms quality, teachers expectations,
intellectual climate, exposure to L2

88
Statistical model
L2 c b1SES b2ASES
b3(SESASES) b4SDist
d(L2)/d(SES) b1
b3ASES
expected partial effect of ASES on SES positive
Substantial interpretation effect of SES
increases with ASES!
89
The graphical illustration
90
U(L2)-U(L1)
C/p
Costs
SES has stronger effects in classes with high
ASES
ASES
low
high
91
In principle the same effect as for EC on
language acquisition (see lecture 3)
92
Remember (Lecture 2)
Substitute Education for SES and Concentration
for ASES
(SES ASES)
SDistance gt 0
93
Statistical model (for education and ethnic
concentration)
L2 c b1Educ b2Conc
b3(EducConc) b4SDist
d(L2)/d(Educ) b1 b3Conc
expected partial effect of Conc on Educ negative
Substantial interpretation effect of SES
decreases with Conc!
94
Summary statistical models
for SES and ASES
L2 c b1SES b2ASES
b3(SESASES) b4SDist
d(L2)/d(SES) b1
b3ASES
for education and ethnic concentration
L2 c b1Educ b2Conc
b3(EducConc) b4SDist
d(L2)/d(Educ) b1 b3Conc
95
Summary substantial interpretation
the same
mechanism, statistical effect, interpretation
good conditions increase individual advantages!
96
Education has smaller effects in classes with
high EC
U(L2)-U(L1)
C/p
Costs
Education has stronger effects in classes with
low EC
Education increases
Ethnic concentration
high
low
97
SES has smaller effects in classes with low ASES
U(L2)-U(L1)
C/p
Costs
SES has stronger effects in classes with high
ASES
SES increases
ASES
low
high
98
Summary substantial interpretation
but remember also the other view on the same
relation (lecture 3)
the disadvantaged suffer more from bad conditions!
look at the other partial effect!
99
Statistical model
L2 c b1Educ b2Conc
b3(EducConc) b4SDist
d(L2)/d(Conc) b2
b3Educ
expected statistical interaction effect positive
100
U(L2)-U(L1)
C/p
Costs
Education
Ethnic concentration increases
high
low
101
U(L2)-U(L1)
C/p
Costs
Education has stronger effects in classes with
low EC
Education increases
Ethnic concentration
high
low
102
Summary Contextual effects on educational success
ASES/ECS have (strong) effects
and a remarkable interaction effect
disadvantaged children suffer more from negative
conditions (low ASES, high ECS)
disadvantaged children gain more from
improvements of negative conditions (higher ASES,
lower ECS)
advantaged children gain more from positive
conditions (high ASES, low ECS)
advantaged children lose their advantages from
more negative conditions (lower ASES, higher ECS)
103
The political problem
104
Our results suggest that if a young person from
disadvantaged socioeconomic background has
parents with low social status, and belongs to a
minority race, then diversity would be an
advantage. The student would benefit from the
resources that the more advantaged students would
bring to the social context of the school. On the
other hand, our findings indicate that if a young
person comes from a relatively privileged
background, then diversity could be a
disadvantage, at least in terms of achievement.
(Caldas und Bankston (1996 276)
105
Lection 4 Education and Labor Market
1. The Problem Ethnic Stratification
2. Education
3. Labor Market
4. Effects of Bilinguality
106
Theoretical Mechanisms
107
Productivity/ Wage (w)
Variations by Fit Q-Value Relevance Form
Human- Capital
108
Productivity/ wages (w)
Discount demand/supply Discrimination Transaction
scosts/ intervening opportunities ethnic networks
Abilities in L
109
Abilities im L
110
(No Transcript)
111
Determinants Labor Market Success
discounts
interv. opportunities ethnic networks
(stat.)discrimination transaction costs
employment wages
L2
Bilinguality
112
Empirical Relations
113
Income
114
Chiswick and Miller 1995 270, Table 5, Columns
2, 4 and 6
115
Chiswick and Miller 1995 270, Table 5, Columns
2, 4 and 6
116
Chiswick and Miller 1995 270, Table 5, Columns
2, 4 and 6
117
Chiswick and Miller 1995 270, Table 5, Columns
2, 4 and 6
118
Chiswick and Miller 1995 270, Table 5, Columns
2, 4 and 6
119
Chiswick and Miller 1995 270, Table 5, Columns
2, 4 and 6
120
Chiswick and Miller 1995 270, Table 5, Columns
2, 4 and 6
121
Communicative relevance
122
Language effects (income) Relevance
Berman, Lang and Siniver 2000 15ff. Figures 1a
and 1b
Programmers Computer Technicians
Change in Income
Construction workers Gas station attendants
Change in Income
Change in Hebrew
123
Language effects (income) Relevance
Berman, Lang and Siniver 2000 15ff. Figures 1a
and 1b
Programmers Computer Technicians
Change in Income
Construction workers Gas station attendants
Change in Income
Change in Hebrew
124
Q-Value
125
Language effects (income) Q-Value
Chiswick and Repetto 2001 222, Table 10.9,
Column 4
126
Language effects (income) Q-Value
Chiswick and Repetto 2001 222, Table 10.9,
Column 4
127
Language effects (income) Q-Value
Chiswick and Repetto 2001 222, Table 10.9,
Column 4
128
Language effects (income) Q-Value
Chiswick and Repetto 2001 222, Table 10.9,
Column 4
129
Form Literacy
130
Speaking, Writing and Earnings (Chiswick und
Repetto 2001, 220, bv)
Earnings
14
12
10
8
6
4
2
0
131
Speaking, Writing and Earnings (Chiswick und
Repetto 2001, 220, mv)
132
Speaking, Writing and Earnings, Chiswick und
Repetto 2001, 220
133
Speaking, Writing and Earnings, Chiswick und
Repetto 2001, 220
134
Discrimination
135
Accents and Earnings, Dávila, Bohara und Saenz
1993 909
136
Accents and Earnings, Dávila, Bohara und Saenz
1993 909
137
Accents and Earnings, Dávila, Bohara und Saenz
1993 912
138
Accents and Earnings, Dávila, Bohara und Saenz
1993 912
139
Accents and Earnings, Dávila, Bohara und Saenz
1993 912
140
Accents and Earnings, Dávila, Bohara und Saenz
1993 912
141
(Ethnic) Group Size and Earnings
142
Group Size and Earnings, Pendacur und Pendacur
2002 172
143
Group Size and Earnings, Pendacur und Pendacur
2002 172
144
Group Size and Earnings, Pendakur und Pendakur
2002 172
145
Group Size and Earnings, Pendakur und Pendakur
2002 170
ethno-linguistic Group
0
linguistic Group
ethnic Group
Earnings Differential
decrease of ethnic penalties
0 0.5 1 1.5
2 2.5 3
3.5 4
Group Size in 100 000
146
Group Size and Earnings, Pendakur und Pendakur
2002 170
ethno-linguistic group
0
linguistic Group
ethnic Group
Earnings Differential
decrease of ethnic penalties
0 0.5 1 1.5
2 2.5 3
3.5 4
Group Size in 100 000
147
Ethnic Networks and Occupational Placement
148
Occupational Placement and Ethnic Networks
Kalter 2005b 36, Table 4, Logit blue-collar
worker/white-collar worker
149
Occupational Placement and Ethnic Networks
Kalter 2005b 36, Table 4, Logit blue-collar
worker/white-collar worker
150
Occupational Placement and Ethnic Networks
Kalter 2005b 36, Table 4, Logit blue-collar
worker/white-collar worker
151
Occupational Placement and Ethnic Networks
Kalter 2005b 36, Table 4, Logit blue-collar
worker/white-collar worker
152
Occupational Placement and Ethnic Networks
Kalter 2005b 36, Table 4, Logit blue-collar
worker/white-collar worker
153
Occupational Placement and Ethnic Networks
Kalter 2005b 36, Table 4, Logit blue-collar
worker/white-collar worker
154
Lection 4 Education and Labor Market
1. The Problem Ethnic Stratification
2. Education
3. Labor Market
4. Effects of Bilinguality
155
Main hypothesis concerning bilinguality
Bilinguality has effects on educational and labor
market success additionally to L2
156
why?
Education

• Cummins Interdependency-Hypothesis

• Zhou Transmission of values and social control

• MouwXie transititional familial
  communication

Labor market

• Globalization and transnationalism

• ethnic economies

• interethnic transactions (middleman)

157
Bilinguality and Educational Success
158
Reading
159
PortesRumbaut
(reading, mathematics and grades similar)
Model 1 bivariate Model 2 gender and
nationality Model 2 family and migration
biography Model 3 L1-environment (family,
peers, School) Model 5 L2-abilities parents
160
Language adaptation has the expected effects,
with fluent bilinguals performing better than
average when other variables are controlled and
limited bilinguals doing much worse. These
results indicate the value of selective
preservation of linguistic ties to families and
coethnic communities. (Portes und Rumbaut 2001
243)
161
a re-analysis
162
Bilinguality and reading scores CILS, OLS, bold
signifikant
163
Bilinguality and reading scores CILS, OLS, bold
signifikant
164
Bilinguality and reading scores CILS, OLS, bold
signifikant
165
Bilinguality and reading scores CILS, OLS, bold
signifikant
166
Bilinguality and reading scores CILS, OLS, bold
signifikant
167
Mathematics
168
Bilinguality and mathematics (scores) CILS, OLS,
bold signifikant
169
Bilinguality and mathematics (scores) CILS, OLS,
bold signifikant
170
Bilinguality and mathematics (scores) CILS, OLS,
bold signifikant
171
Bilinguality and mathematics (scores) CILS, OLS,
bold signifikant
172
Bilinguality and mathematics (scores) CILS, OLS,
bold signifikant
173
Bilinguality and mathematics (scores) CILS, OLS,
bold signifikant
174
but why find PortesRumbaut effects of
bilinguality?
175
Procedure PortesRumbaut
(reading, mathematics and grades similar)
176
???
177
Bilinguality and school performance Ethnic
differences (Hispanics and Asians in USA CILS)
178
Background
Very different successes between hispanic and
asian migrants children bilinguality?
Effetcs of different group size and
ethno-linguistic conzentrations bilinguality?
179
Reading OLS CILS
Hispanics bv
10
0
-10
-20
-30
-40
-50
-60
180
Reading OLS CILS
Hispanics bv
10
0
-10
-20
-30
-40
-50
-60
181
Mathematics OLS CILS
10
0
-10
-20
-30
182
hence
the successful asian groups have not only no
advantage of bilinguality, but even suffer from
it.
at the best
Penalties by L1 become smaller, the larger the
ethno-linguistic group is (eg with hispanics),
but they disappear with very large groups only.
remember
183
Group Size and Earnings, Pendakur und Pendakur
2002 170
ethno-linguistic group
0
linguistic Group
ethnic Group
Earnings Differential
decrease of ethnic penalties
0 0.5 1 1.5
2 2.5 3
3.5 4
Group Size in 100 000
184
Effects of bilingual education?
185
Determinants of educational success bilingual
education?
(institutional) discriminiation
preschool-exposure
Educational success
186
Nach der Analyse des Forschungsstandes besteht
ein straffer Zusammenhang zwischen dem Erwerb
einer Zweitsprache und dem der jeweiligen
Erstsprache. Hierzu liegt eine große Zahl an
Studien vor, die von der frühen Kindheit bis hin
zum Erwachsenenalter reichen . Die
schulische Förderung der Erstsprache wirkt sich
im schlechtesten Fall neutral aus, hat also
keine negativen Folgen für den Spracherwerb .
In der Regel jedoch wirkt sie sich positiv aus
. (Gogolin, Neumann und Roth (2003 45/6
Hervorhebungen nicht im Original).
187
Greenes Meta-Analysis (1998)
188
Greenes Meta-Analysis (1998)
Effect sizes (Hedges g) of the eleven studies in
Greenes meta-analysis (1998 8, Table 3)
0,9
0,8
average-effect 0.21 (2.46)
0,7
0,6
0,5
0,4
0,3
0,2
0,1
0
-0,1
1
2
3
4
5
6
7
8
9
10
11
189
Greenes Meta-Analysis (1998)
0,9
0,8
0,7
0,6
0,5
0,4
0,3
0,2
0,1
0
-0,1
1
2
3
4
5
6
7
8
9
10
11
190
Greenes Meta-Analysis (1998)
0,9
0,8
0,7
average-effect 0.12 (1.78)
0,6
0,5
0,4
0,3
0,2
0,1
0
-0,1
1
2
3
4
5
6
7
8
9
10
11
191
Greenes Meta-Analysis (1998)
0,9
0,8
0,7
0,6
0,5
0,4
0,3
0,2
0,1
0
-0,1
1
2
3
4
5
6
7
8
9
10
11
192
Greenes Meta-Analysis (1998)
0,9
0,8
and without study 6 0.07 (1.35)
0,7
0,6
0,5
0,4
0,3
0,2
0,1
0
-0,1
1
2
3
4
5
6
7
8
9
10
11
193
Reconstruction by RossellKuder (2005)
194
Reconstruction by RossellKuder (2005)
195
Conclusion RossellKuder 2005
At this point, we can say that our reanalyses of
both Greene (1998) and Slavin and Cheung (2004,
2005) do not support the conclusions they draw
regarding the superiority of bilingual education
over a mainstream classroom. (Rossell und Kuder
2005 69)
196
Conclusion Gogolin (2005)
There is no evidence of negative effects on the
acquisition of a second language . The question
as to whether bilingual children profit from
bilingual education can therefore be answered
with yes, as there were no negative effects on
L2-development, but unambiguous positive effects
on bilingual development. What remains unclear
however, is the question of whether they would
benefit equally (or perhaps even more) from other
measures, if only L2 were concerned. Answering
this question would require comparative
experimental settings and research, which has
been lacking up to now. (Gogolin 2005 142)
197
Bilinguality and Labor Market Performance
198
Theory Carliner
199
Seven Assumptions
1. There are A(n) speakers of a language A and
B(n) of a language B
2. There are three kinds of jobs

• with A as requirement
• with B as requirement
• with bilinguality as requirement

3. Actors prefer jobs in their respective L1, but
would invest in the other language if that would
pay
4. Supply in language abilities varies positively
with incentives for language acqusition
5. Actors are indifferent for jobs in their own
language and bilingual jobs
6. There is a higher demand for one of the two
languages (eg A), due to the structure of trade
relations or the higher (Q-)value of that language
7. Therefore exists a higher supply in the other
language (eg B) than is demanded
200
Implications
1. Income in A-jobs is higher than in B-jobs
(higher demand for A than for B)
2. No difference in income for A-jobs and
bilingual jobs (due to the indifference towards
own-language and bilingual jobs) bilinguality
doesnt pay for A-speakers
3. All bilingual speakers are B-speakers, because
(only) for those exist incentives to acquire the
other language (A)
4. Among the monolinguals A-speakers have the
highest income, followed by bilingual B-speakers
the lowest income have those with neither A nor B
In general bilinguality doesnt pay over and
above the respective (national or
regional) dominant language
201
Effects of Bilinguality Empirical Results
202
Income and bilinguality Fry and Lowell 2003
OLS, controlled for age,
city/country, region)
203
Income and bilinguality Fry and Lowell 2003
OLS, controlled for age,
city/country, region)
204
Income and bilinguality Fry and Lowell 2003
OLS, controlled for age,
city/country, region)
205
Income and bilinguality Fry and Lowell 2003
OLS, controlled for age,
city/country, region)
206
Income and bilinguality Fry and Lowell 2003
OLS, controlled for age,
city/country, region)
207
Income and bilinguality Fry and Lowell 2003
OLS, controlled for age,
city/country, region)
208
Income and bilinguality Fry and Lowell 2003
OLS, controlled for age,
city/country, region)
209
Income and bilinguality Fry and Lowell 2003
OLS, controlled for age,
city/country, region)
210
Income and bilinguality Fry and Lowell 2003
OLS, controlled for age,
city/country, region)
211
Income and bilinguality Fry and Lowell 2003
OLS, controlled for age,
city/country, region)
212
Income and bilinguality Fry and Lowell 2003
OLS, controlled for age,
city/country, region)
213
Income and bilinguality Fry and Lowell 2003
OLS, controlled for age,
city/country, region)
214
Occupational Placement and L1 Kalter 2005b 36,
Table 4, Logit blue-collar
worker/white-collar worker
215
Occupational Placement and L1 Kalter 2005b 36,
Table 4, Logit blue-collar
worker/white-collar worker
216
Occupational Placement and L1 Kalter 2005b 36,
Table 4, Logit blue-collar
worker/white-collar worker
no effect of L1!
217
Canada
218
Official Languages (PendakurPendakur 2002)
219
Canada Bilinguality und income (official
languages) Pendakur and Pendakur 2002 157, Table
2, logit (controlled for standard variables)
too few cases
220
Canada Bilinguality und income (official
languages) Pendakur and Pendakur 2002 157, Table
2, logit (controlled for standard variables)
too few cases
221
Canada Bilinguality und income (official
languages) Pendakur and Pendakur 2002 157, Table
2, logit (controlled for standard variables)
Carliners B-speakers (french)
too few cases
222
Non-Official Languages (PendakurPendakur 2002)
223
(No Transcript)
224
(No Transcript)
225
(No Transcript)
226
(No Transcript)
227
penalties for allophones, especially in
Montreal, even with command of one or both of the
official languages
228
and also
229
(No Transcript)
230
(No Transcript)
231
(No Transcript)
232
(No Transcript)
233
(No Transcript)
234
there are additional penalties for speakers
with even two allophone languages
235
Given official language knowledge, men who know
two nonofficial languages earn less than men who
know just one nonofficial language. (Pendakur
und Pendakur 2002 159 Hervorhebungen nicht im
Original).
236
All in all Carliner