Transcript Spacing of children in Switzerland: constancy or change?

Spacing of children
in Switzerland:
constancy or change?
Marion Burkimsher
Affiliated to
University of Lausanne
1969
1970
1971
1972
1973
1974
1975
1976
1977
1978
1979
1980
1981
1982
1983
1984
1985
1986
1987
1988
1989
1990
1991
1992
1993
1994
1995
1996
1997
1998
1999
2000
2001
2002
2003
2004
2005
2006
2007
2008
2009
Age of mother
Change in mean age at birth of each birth order 1969-2007
1st birth
2nd birth
3rd birth
Year
4th birth
5th+ birth
37
36
35
34
33
32
31
30
29
28
27
26
25
Source BEVNAT
Birth spacing - what and why?
• Not difference in mean age of successive birth orders
• Can obtain spacing data from surveys (or census data), but not
birth registration by birth order
• Why these are different…(women who end up having bigger
families start childbearing at a younger age; women who start
later have fewer children)
• With birth spacing data (probability of subsequent birth knowing
duration since last birth) - and female population by parity & age can make good TFR projections by McDonald-Kippen method
1969
1970
1971
1972
1973
1974
1975
1976
1977
1978
1979
1980
1981
1982
1983
1984
1985
1986
1987
1988
1989
1990
1991
1992
1993
1994
1995
1996
1997
1998
1999
2000
2001
2002
2003
2004
2005
2006
2007
2008
2009
Years
Differences in mean age at nth and (n+1)th birth
1st-2nd birth
2nd-3rd birth
3rd-4th birth
3.0
2.8
2.6
2.4
2.2
2.0
1.8
1.6
1.4
1.2
1.0
Source BEVNAT
1969
1970
1971
1972
1973
1974
1975
1976
1977
1978
1979
1980
1981
1982
1983
1984
1985
1986
1987
1988
1989
1990
1991
1992
1993
1994
1995
1996
1997
1998
1999
2000
2001
2002
2003
2004
2005
2006
2007
2008
2009
Standard deviation, years
Change in standard deviation of age at birth
Birth order 1
Birth order 2
Birth order 3
Year
Birth order 4
Birth order 5+
5.4
5.2
5.0
4.8
4.6
4.4
4.2
4.0
3.8
Source BEVNAT
Example of erroneous deduction: spacing in Switzerland
FFS sample data from 1994 for women aged 40-44
ie. cohorts 1950-1954
Mean age at 1st birth 26.9
Mean spacing 1st-2nd child: 3.13 years
Mean spacing 2nd-3rd child: 3.62 years
Mean spacing 3rd-4th child: 3.13 years
Houle & Shkolnikov, 2006 then made the following incorrect
deductions using age at 1st birth as the starting point:
Deduced mean age at 2nd birth 30.0
Deduced mean age at 3rd birth 33.6
Deduced mean age at 4th birth 36.8
Restating…
We cannot deduce anything about birth spacing from differences
in mean age at successive birth orders!
We cannot deduce anything about mean age of successive birth
orders from information on birth spacing!
Why?
Because women who have their first child at a younger age have
more likelihood of having a larger family than women who start at
a later age.
This is easier to understand with an illustration…
2 hypothetical scenarios…
Scenario 1: 4 women, having 1, 2, 3 and 4 children respectively
28
31
34
25
Age
QuickTi me™ and a
decompressor
are needed to see thi s pi ctur e.
QuickTi me™ and a
decompressor
are needed to see thi s pi ctur e.
QuickTi me™ and a
decompressor
are needed to see thi s pi ctur e.
QuickTi me™ and a
decompressor
are needed to see thi s pi ctur e.
Mean age at 1st birth: 25
Mean age at 2nd birth:
28
Mean age at 3rd birth: 31
Mean spacing between all
birth orders: 3 years
Mean age at last birth:
29.5
Scenario 2: 4 women, having 1, 2, 3 and 4 children respectively
28
31
34
25
Age
QuickTi me™ and a
decompressor
are needed to see thi s pi ctur e.
QuickTi me™ and a
decompressor
are needed to see thi s pi ctur e.
QuickTi me™ and a
decompressor
are needed to see thi s pi ctur e.
QuickTi me™ and a
decompressor
are needed to see thi s pi ctur e.
Mean age at 1st birth: 29.5
Mean age at 2nd birth: 31
Mean age at 3rd birth: 32.5
Mean age at 4th birth: 34
Mean spacing between all
birth orders: 3 years
Mean age at last birth: 34
Hypothesis
That between 1969 and the present time there has been a
transition from Scenario 1 being the norm to Scenario 2.
But we need birth spacing data…
Using data from SHP
•
Data about cohabiting children had to be combined with data
on (older) children who had moved out of the household
•
Frequency curves of spacing are skewed;
modal gap < median gap < mean gap
•
Likelihood of going on to a 2nd / 3rd child changes over time
•
Solution - survival analysis, life table method
(as only whole year data)
•
Survival analysis gives:
1. change in intensity
2. change in ultimate likelihood
3. median duration of transition (if >50% experience it)
Interesting statistics from SHP sample
Longest gap 1st-2nd for a woman is 18 years
Longest gap 1st-2nd for a man is 32 years
Longest gap 2nd-3rd for a woman is 31 years!
Longest gap 2nd-3rd for a man is 25 years
Range, age at 1st birth, woman - 15-54 years
Range, age at 1st birth, man - 12-59 years
Range, age at 2nd birth, woman - 17-49 years
Range, age at 2nd birth, man - 16-59 years
…as men have no upper bound, only look at women’s rates …
Frequency curve for 2nd child after 1st
and 3rd child after 2nd by duration
Probability of 2nd/3rd birth
0.35
0.30
1960s
1970s
1980s
1990s
1960s
1970s
1980s
1990s
0.25
0.20
0.15
0.10
0.05
0.00
0
1
2
3
4
5
6
7
Number of years after 1st/2nd child
8
9
2nd child
2nd child
2nd child
2nd child
3rd child
3rd child
3rd child
3rd child
Modal gap
1st-2nd &
2nd-3rd is
2 years
Median
gap
% no
2nd
1960s
3.3
20
1970s
3.5
18
1980s
3.3
17
1990s
3.3
18
QuickTime™ and a
decompressor
2000>
3.6
are needed to see this picture.
22
No significant
change in
1st-2nd gap
from 1960s to
present day!
QuickTime™ and a
decompressor
are needed to see this picture.
But significant
changes for
2nd-3rd child
transition!
How does age of mother affect spacing?
…remembering that age at childbearing has been rising over time
QuickTime™ and a
decompressor
are needed to see this picture.
Significantly
less likelihood
of women
having a 2nd
child if over 35
when 1st born
QuickTime™ and a
decompressor
are needed to see this picture.
Very significant
difference in
likelihood of
woman having
3rd child if her
2nd born
before/after 30
Various potential determinants
1st-2nd likelihood
Yes, >2000 slightly lower
Decade of birth of 1st/2nd
child
Age of mother at bi rth of Yes, older mothers lower
1st/2nd
Religious attendance
Yes, regu lar attenders
higher
Education level (3 groups)
Not signifi cant
Gender of 1st child
Linguistic a reas
1st 2 children twins
Mother born in CH
Not signifi cant
Slight
German>French>Italian
Slight, higher if born in
CH
2nd-3rd likelihood
Yes, 1960s higher
intensity
Yes, older mothers lower
Yes, very signif. diffs.
Yes, low ed., higher
intensity
German>French>Italian
Yes, higher if twins
Yes, higher if born in CH
Cannot yet confirm or deny hypothesis
Initial investigations encouraging…
SHP is only up-to-date data source on birth spacing in
Switzerland
More work to be done…
Spacing of children in Switzerland:
constancy or change?
• 1st-2nd child transition > constancy
• 2nd-3rd child transition > change (decline in likelihood)
Thank you!