tipping point in inequality€¦ · meeting with agent-based computational modeling, we can ... cz...
TRANSCRIPT
Tipping point in inequality
The new gender distribution in education is likely to affect reproductive behaviour
The inequality in educational attainment has changed to the advantage of women
First evidence that the new mating squeeze affects educational assortative mating
Tipping point in inequality
Hypotheses about future developments need to consider complex macro-micro interactions
Acknowledgements
An agent-based computational model of European marriage markets
First results suggest that the model approximates actual partner search processes well
Data Sources
Tipping point in inequality
Education specific mating
squeeze
Assortative mating
Probability and stability of unions Fertility
The Reversal of Gender Inequality in Education in Europe: Implications for Reproductive Behaviour
Jan Van Bavel & André Grow
Centre for Sociological Research, KU LeuvenUnit for Family and Population Studies
Contact: [email protected], [email protected]
This project is funded by the European Research Council under the European Union's Seventh Framework Programme (FP/2007-2013) / ERC Grant Agreement no. 312290 for the GENDERBALL project
Until the 1970s, the vast majority of students in higher education were men.
In recent years, more women than men successfully complete higher education.
As a result, there are more highly educated women than men on the marriage market.
Since many decades, educational homogamy has been dominant. If partners differed in their educational attainment, his education was typically higher than her education.
The larger the number of highly educated women became, relatively to the number of highly educated men, the less feasible traditional mating patterns became.
Today, partners still tend to be similarly educated. However, if there is a difference in their educational attainment, her education now tends to be higher than his education.
Focus of this poster
Gender-specific distribution of educational
attainment
Patterns of assortative mating
Constraints that individuals face during
partner searchMating decisions
Macro-level
Micro-level
We first build a theoretical model about mate choice on the micro-level.
We use computational simulations to see whether our theoretical model is able to replicate observed macro-level patterns.
After calibrating the model, we use it as a virtual laboratory to explore potential future scenarios.
More highly educated women than men on the marriage market
His and her relative education will change, with implications for his and her
earning potential in the labour market
Changing educational similarity in couples will affect household dynamics
Changing household dynamics will affect timing and frequency of
childbearing
Death/Reproduction
Central interest
Dating decisions
Marriage decisions
Meeting
With agent-based computational modeling, we can specify theoretically motivated rules for partner search and study their implications in large populations.
Male and female agents prefer partners with similar
educational attainment
α − − − − =
max max
max max max
s ay
w ww
i j j iiij
S s s A ayv
S Y A
Male and female agents prefer partners with high
earning potential
Male agents prefer partners who are in their mid-twenties; female agents prefer partners who are slightly older than themselves
Male and female agents look for partners with high
subjective mate value
Importance of each dimension can differ between male and
female agents
The calibrated model generates patterns of assortative mating that are similar to observed patterns.
The model can capture variation both across countries and across time.
The next step is to extend the simulation period and to explore the implications of potential future scenarios.
• European Social Survey Rounds 1 to 6 Data (2002, ed. 6.3; 2004, ed. 3.3; 2006, ed. 3.4; 2008, ed. 4.2; 2010, ed. 3.1; 2012, ed. 2.0).
• Reconstruction of populations by age, sex, and level of educational attainment for 120 countries for 1970-2000 by the International Institute for Applied Systems Analysis/Vienna Institute of Demography
AT
BE
BG
CH
CR
CZ
DE
DK
EE
ES
FI
FR
GR
HU
IC
IE
IT
LI
LU
LV
NL
NO
PL
PT
RO
SE
SI
UK
Share of women in tertiary education
30
40
50
60
70
1971
1981
1991
2001
CH
NL
BE
BG
Year
Perc
ent
ATAT
AT
BE
BE
BE
BG
BG
BG
CHCH
CH
CZ
CZ
CZ
DEDE
DE
DK
DK
DK
EE
EE
EE
ESES
ES FI
FI
FI
FR
FR
FR
GB
GB
GB
GR
GR
GR
HR
HR
HR
HU
HU
HU
IE
IEIE
IT
IT
ITLT
LT
LT
LU
LULU
LV
LV
LV
NL
NL
NL
NO
NO
NO
PL
PL
PL
PT
PT
PT
RORORO
SE
SESE
SI
SI
SI
SK
SK
SK
−2
−1
ATATAT
BEBEBE
BGBG
BG
CHCHCH
CZCZCZ
DEDEDEDK
DKDK
EEEE
EE
ES
ESESFI
FIFI
FR
FRFR
GBGB
GBGRGR
GR
HRHRHR
HUHUHU
IEIE
IE ITITIT
LTLTLT
LU
LULULV
LVLV
NLNL
NLNONO
NOPLPLPL
PT
PT
PT
RORO
RO
SE
SESESISISI
SKSKSK
Men = Women
−0.5 0.0 0.5
AT
AT
AT
BE
BEBE
BGBG
BG
CH
CH
CH
CZ
CZ
CZ
DE
DE
DE
DKDKDK
EEEEEE
ES
ES
ESFI
FI
FI
FRFR
FR
GBGBGB
GR
GR
GR
HR
HR
HR
HU
HU
HU
IEIE
IE
IT
IT
IT
LTLT
LT
LU
LU
LU
LV
LVLV
NL
NLNL
NO
NONO
PLPL
PL
PT
PT
PT
RORO
ROSESE
SE
SI
SI
SI
SK
SKSK
Men < Women
−0.5 0.0 0.5 −0.5 0.0 0.5
log
(pe
rce
nt
of
co
up
les)
Percent of couples
Educational attainment W = MMen < Women Men = Women Men > Women
=
=
=
=
=
=
=
=
=
=
=
==
=
=
=
=
=
=
=
=
=
=
=
=
=
=
=
W
W
W
W
W
W
W
W
W
W
W
WW
W
W
W
W
W
W
W
W
W
W
W
W
W
W
W
M
M
M
M
M
M
M
M
M
M
M
MM
M
M
M
M
M
M
M
M
M
M
M
M
M
M
M
=
=
=
=
=
=
=
=
=
=
=
==
=
=
=
=
=
=
=
=
=
=
=
=
=
=
=
W
W
W
W
W
W
W
W
W
W
W
WW
W
W
W
W
W
W
W
W
W
W
W
W
W
W
W
M
M
M
M
M
M
M
M
M
M
M
MM
M
M
M
M
M
M
M
M
M
M
M
M
M
M
M
=
=
=
=
=
=
=
=
=
=
=
==
=
=
=
=
=
=
=
=
=
=
=
=
=
=
=
W
W
W
W
W
W
W
W
W
W
W
WW
W
W
W
W
W
W
W
W
W
W
W
W
W
W
W
M
M
M
M
M
M
M
M
M
M
M
MM
M
M
M
M
M
M
M
M
M
M
M
M
M
M
M
ROEELTESNLDEGRGBPTCHFRSEIE
LUFI
BENOLVSI
DKIT
HUPLATHRBGCZSK
0 25 50 75 100 0 25 50 75 100 0 25 50 75 100
Co
un
try
Cohort 1950 - 1959 Cohort 1960 - 1969 Cohort 1970 - 1979
Men > Women
log(ratio men age 27-36 to women age 25-34 with degree in tertiary education)
Cohort
BE DE
FI FR
NL PT
0
25
50
75
100
0
25
50
75
100
0
25
50
75
100
(194
0,19
50]
(195
0,19
60]
(196
0,19
70]
(194
0,19
50]
(195
0,19
60]
(196
0,19
70]
Pe
rce
nt
of
co
up
les
observed mean of 50 simulation runs
Men > Women
BE DE
FI FR
NL PT
(194
0,19
50]
(195
0,19
60]
(196
0,19
70]
(194
0,19
50]
(195
0,19
60]
(196
0,19
70]
Men = Women
BE DE
FI FR
NL PT
(194
0,19
50]
(195
0,19
60]
(196
0,19
70]
(194
0,19
50]
(195
0,19
60]
(196
0,19
70]
Men < Women
mean of simulation runs +-/1 standard deviation