Gender Inequality in the School-to-Work Transition in 29 ... · | 7 EU-LFS ad hoc module 2009...
Transcript of Gender Inequality in the School-to-Work Transition in 29 ... · | 7 EU-LFS ad hoc module 2009...
Melinda Mills
Patrick Präg
3rd European User Conference for EU-LFS
and EU-SILC, Mannheim, March 2013
Gender Inequality in the
School-to-Work Transition
in 29 European Countries
| 2
Gender inequality in the labor market
For instance:
› Female managers
Background and research problem | Data and method | Results | Conclusion
0
10
20
30
40
50
% F
em
ale
man
ag
ers
, 2
01
1
LV IS
HU
FR
PL
LT SI
BG
EE
SE
GB
PT IE
EU
-27
CH
NO FI
SK
RO
NL
ES
DE
BE
DK
AT
CZ
LU IT
MT
GR
CY
Source: Eurostat News Release 37/2013
| 3
Gender inequality in the labor market
For instance:
› Female managers
› Gender pay gap
0
10
20
30
Ge
nd
er
pa
y g
ap
(u
nad
juste
d),
20
11
EE
AT
DE
GR
CZ
SK
GB FI
HU
NL
CH
DK
CY
EU
-27
ES
NO
SE
FR IE LV
BG
MT
PT
RO LT
BE
LU IT PL SI
Source: Eurostat, tsdsc340, date of extraction: 2013-03-15
Background and research problem | Data and method | Results | Conclusion
| 4
Gender inequality in the labor market
For instance:
› Female managers
› Gender pay gap
› Part-time work
Background and research problem | Data and method | Results | Conclusion
0
20
40
60
80
% P
art
-tim
e o
f to
tal em
plo
ym
en
t, 2
01
1
NL
CH
DE
AT
BE
LU
GB
NO
SE IT
EU
-27
FR
DK IE IS
MT
ES FI
EE
CZ
PT
CY
GR
PL SI
HU
LV
LT
SK
RO
BG
Women
Men
Source: Eurostat, lfsa_eppga, date of extraction: 2013-03-15
| 5
› Females outperform males in all EU countries
Gender inequality in educational attainment
100
150
200
250
Fem
ale
te
rtia
ry e
du
ca
tio
n g
rad
ua
tes
pe
r 1
00
me
n, 2
01
0
LV
EE IS PL
LT
HU
SK
RO
SE SI
NO
BG
CZ
PT
CY FI
GR
EU
-27 IT BE
MT
DE
DK
ES
NL
LU
GB
FR IE AT
CH
Source: Eurostat, educ_itertc, date of extraction: 2013-03-15
Background and research problem | Data and method | Results | Conclusion
| 6
Gender inequality in educational attainment
› Female advantage fairly recent phenomenon
20
30
40
50
60
70
% F
em
ale
stu
de
nts
in
te
rtia
ry e
du
catio
n,
19
71
–2
010
1970 1980 1990 2000 2010
EU-27 average (unweighted)
Source: UNESCO Institute of Statistics, date of extraction: 2013-03-14
Background and research problem | Data and method | Results | Conclusion
| 7
EU-LFS ad hoc module 2009
‘Entry of Young People into the Labor Market’
› Young people: 15-34 year-olds
Transition from school to work
› School: leaving formal education for the last time
› Work: first job that lasted for more than three months
› Restricted to transitions in the last five years
29 countries (EU-27 plus Norway and Iceland)
› Switzerland and Germany: questionable data quality
Background and research problem | Data and method | Results | Conclusion
Analytical strategy
› Outcome: First job of more than three months 0/1 › Person-month file to accommodate
time-varying country-level covariates
› Random-effects complementary log-log (cloglog) models (Mills 2011) › Time-constant country variation accounted for
by country dummies
› Analyses conducted in Stata 12.1
| 8
EU-LFS ad hoc module 2009
Background and research problem | Data and method | Results | Conclusion
Individual-level predictors
› Educational attainment › low (ISCED 0-2), medium (ISCED 3-4), high (ISCED 5-6)
› Educational field › Main groups of Andersson and Olsson (1999)
› Workplace-based VET › Yes/no
› Worked during education (> 1 month/year) › Yes/no
| 9
EU-LFS ad hoc module 2009
Background and research problem | Data and method | Results | Conclusion
Country-level time-varying predictors
› GDP per capita (logged) › Quarterly, source: Eurostat
› Unemployment rate › Monthly, source: Eurostat
› Employment protection legislation › For temporary and permanent contracts separately
› Yearly, sources: Venn (2009) and Muravyev (2010)
› Not available for BG, CY, IS, LU, MT, RO, and SI
| 10
EU-LFS ad hoc module 2009
Background and research problem | Data and method | Results | Conclusion
| 11
Sample size N 15-34 year-olds 321,000
After removing …
› Swiss respondents 310,000
› those who left education before 2004 198,000
› those still in education at time of data collection 61,242 of which: non-censored cases 47,131
EU-LFS ad hoc module 2009
Background and research problem | Data and method | Results | Conclusion
| 12
› Gender inequality in educational attainment
Preliminary results
0
5
10
15
20
Perc
en
tage
po
int
diffe
rence
be
twe
en
fem
ale
an
d m
ale
te
rtia
ry e
du
ca
tio
na
l a
tta
inm
ent
CY
DK FI
LV
NO
MT
EE IE PL
BG
PT SI
LT
ES IS SE IT LU
GR
HU
BE
UK
FR
NL
CZ
RO
SK
AT
DE
Source: EU-LFS 2009 AHM, own calculations
Background and research problem | Data and method | Results | Conclusion
| 13
› Gender inequality in educational attainment
› Gender differences in fields of education
Preliminary results
-30 -20 -10 0 10
Percentage point difference female minus male
Services
Health and welfare
Agriculture
Engineering, manufacturing, and construction
Sciences
Social sciences, business, and law
Humanities and arts
Education
General
EU-27 plus Norway and IcelandSource: EU-LFS 2009 AHM, weighted, own calculations
Background and research problem | Data and method | Results | Conclusion
| 14
› Marked differences across countries
Preliminary results
Background and research problem | Data and method | Results | Conclusion
0
5
10
15
20
Me
dia
n t
ime
to
en
try in
to f
irst
job
(in
mon
ths)
IT
GR
RO
ES
BG
MT
EU
-27
CY
FR
DE
PT
SK
HU SI
SE
PL
NO LV
LU LT IE FI
EE
DK
CZ
BE
AT
UK IS NL
Source: EU-LFS 2009 AHM (authors' calculations)
| 15
› Gender difference in school-to-work transition arises after ~18 months
Preliminary results
Background and research problem | Data and method | Results | Conclusion
0.00
0.25
0.50
0.75
1.00
Pro
po
rtio
n in
fir
st
job
0 12 24 36 48 60 72
Months since leaving education
Men
Women
Note: EU-27 plus Norway and IslandSource: EU-LFS AHM 2009 (weighted), authors' calculations
Kaplan–Meier failure estimates
| 16
› Adjusting for education, however, increases the gender difference
Preliminary results
Background and research problem | Data and method | Results | Conclusion
0.00
0.25
0.50
0.75
1.00
Pro
po
rtio
n in
fir
st
job
0 12 24 36 48 60 72
Months since leaving education
Men
Women
Note: EU-27 plus Norway and IslandSource: EU-LFS AHM 2009 (weighted), authors' calculations
adjusted for education
Kaplan–Meier failure estimates
| 17
› Gender difference across countries
Preliminary results
Background and research problem | Data and method | Results | Conclusion
Women enterfirst job earlier
Nodifference
Men enterfirst job earlier
-5
0
5
10
15
Ge
nd
er
ga
p in
entr
y in
to f
irst
job
(Me
dia
n m
on
ths u
ntil tr
an
sitio
n in
to f
irst
job
)
IT
DE
ES
PT
DK IS
NO
BE
CZ
EE
EU
27
HU IE LU
LV
NL
PL
SE SI
SK
UK
AT
FR LT
RO
BG FI
MT
EL
CY
Source: EU-LFS 2009 AHM (authors' calculations)
| 18
Preliminary results
Background and research problem | Data and method | Results | Conclusion
.05
.1
.15
.2
Pre
dic
ted
pro
ba
bili
ty
Male Female
Didn't work during education
Worked during education
Note: Controlling for age, education, time-constant country variance, and GDP per capitaSource: EU-LFS AHM 2009 (weighted), authors' calculations
› Gendered effect of working during education
| 19
Preliminary results
Background and research problem | Data and method | Results | Conclusion
› Stricter employment protection legislation (EPL) impedes transition
Strictness of EPL: regular contracts
Strictness of EPL: temporary contracts
Unemployment rate
log(GDP per capita)
Cou
ntr
y-l
eve
l p
red
icto
rs
.5 .6 .7 .8 .9 1 1.1
Odds ratio
Notes: Model without BG, CY, IS, LU, MT, RO, and SI. Controlling for sex, age, education,and country dummies. Source: EU-LFS 2009 AHM, own calculations
| 20
Preliminary results
Background and research problem | Data and method | Results | Conclusion
› Effect of EPL largely similar between sexes
.05
.1
.15
.2
Pre
dic
ted
pro
ba
bili
tie
s
0 1 2 3 4
Strictness of EPL: temporary contracts
Male
Female
Notes: Model without BG, CY, IS LU, MT, RO, and SIControlling for gender, age, education, GDP, unemployment, and country dummiesSource: EU-LFS 2009 AHM, own calculations
Predictive Margins of female with 95% CIs
› Selection: Large part of sample remains in education
› Unobserved heterogeneity: Family situation can’t be modeled
› EPL: No distinction between temporary/permanent first contract (yet)
| 21
Limitations
Background and research problem | Data and method | Results | Conclusion
› Country differences much more pronounced than gender differences
› Women clearly benefit from better educational attainment and working during education
› Substantial heterogeneity in the size of gender differences across countries
| 22
Conclusions (so far)
Background and research problem | Data and method | Results | Conclusion
› Andersson, Ronnie, and Anna-Karin Olsson. 1999. Fields of Education and Training Manual. Eurostat.
› Mills, Melinda. 2011. Introducing Survival and Event History Analysis. Sage. doi: 10.4135/9781446268360
› Muravyev, Alexander. 2010. "Evolution of Employment Protection Legislation in the USSR, CIS, and Baltic States, 1985-2009." IZA Discussion Paper 5365.
› Venn, Danielle. 2009. "Legislation, Collective Bargaining, and Enforcement. Updating the OECD Employment Protection Indicators." OECD Social, Employment, and Migration Working Papers 89. doi: 10.1787/223334316804
| 23
References
Background and research problem | Data and method | Results | Conclusion