Does Fertility Respond to Work and Family-life
Reconciliation Policies in France?
Olivier Thévenon
CESifo Conference on Fertility and Public Policy
Does Fertility Respond to Work and Family-life Reconciliation
Policies in France?
First draft
Olivier Thévenon♣
1st February 2008
Institut National d’Etudes Démographiques
133, Boulevard Davout
75980 Paris Cedex 20
[email protected]
Abstract :
The current total fertility rate in France has been increasing over this last ten years and
is has reached its highest level since the early 1980s with a rate at 1,98 in 2006. Compared to
European standards, this high level of fertility makes France an outlier, in spite of rather
similar trends in the transition to adulthood, in partnerships, or in attitudes on birth control or
in economic situation (rather low growth, increase in poverty rates). Thus, the French case
challenges some of the hypotheses ventured to explain the current low fertility observed in
European countries. France’s fertility level can be explained by its longstanding family
policy, which has changed in-depth since the 1980s to accommodate with women’s increasing
labour force participation. This policy encompasses a wide range of instruments, based on
different actors and motivations, since this policy is aimed to serve different objectives.
Despite some ambiguities, family policy seems to have created especially positive attitudes
towards 2 or 3 children families in France, and to have bounded the propensity to remain
childless. We argue that a key aspect is the favourable context created for the conciliation
between work and family through a relatively comprehensive and continuous support over the
family life-course. The all set of complementary instruments (financial transfers to large
families, parental leave schemes and provision of childcare support) creates a rather secure
climate for the decision relating to child bearing. It also explains why the decision to have
children or to be in employment is less polarised according to socio-economic status than in
other countries.
♣
I thank Didier Breton, France Prioux, Laurent Toulemon, Pierre Courtioux, Nadine Laïb, Sylvie Le Minez and
Benoît Mirouse for the kind provision of data and figures. Original figures can be found in the references
mentioned at the end of the paper. Usual disclaimer applies.
1
Since about two years, France delights at the prospect that fertility rates are now
approaching the replacement level, with a periodic fertility rate at 1,98 in 2006 and 1.96 in
2007. After a ten-year period of slight increase, fertility rate has thus reached its highest level
since about thirty years (Prioux, 2007). With such a performance, France is situated at the
very top of European and other OECD countries. However, this high level of fertility is
somewhat puzzling, because most of the trends emerging from the 70s in Western countries
are also observed in France. Young people have increased their educational investment,
delayed their entry in the labour market and the time of leaving parental home. Changes are
especially important for women whose labour force participation is still growing.
Young generations also face relatively high rates of unemployment and assume
increasing uncertainty towards their labour force status. Demographic changes are also shaped
by the transformation of values, such as those related to the affirmation of what Giddens and
subsequent authors have named the “individualism of society”, or the increasing concern
about gender equality. Such changes may contribute to explain the increase in non marital
cohabitation, in the use of contraceptive and birth control, or the increase of divorce.
However, one may refer cautiously to these changes in values since some resistance against
gender equality can also be noted, as reflected by the very slight changes in the division of
labour between men and women.
All this factors contribute to explain why younger generations are inclined to delay
childbearing in France like in other Western countries. They are also mentioned as factors that
could lead to a permanent situation of low fertility (Bongaarts, 2001; Kohler, Billari, Ortega,
2006; Morgan and Taylor, 2006; Lutz et al., 2006). However, despite France exhibits most of
the previous mentioned trends, the level of fertility is maintained at a relatively high level of
fertility compared to other countries of continental or southern European countries. As we
detail in the first part of this paper, this situation reflects some specific characteristics of
fertility behaviour: a relatively low proportion of women remaining childless; an increasing
norm towards 2 children and a standardisation of childbirth timing; there are also rather low
differences in family size across social classes. All these factors produces relatively stable
rates of fertility, when they are estimated after control of tempo changes (Toulemon et al.,
2008).
One element explaining this situation is the longstanding family policy implemented
in France, especially since the Second World War. Because of its multi-faceted grounds,
France has a variable and often contradictory position in the classification of Western welfare
systems (Meulders et al., 2002; Thévenon, 2006). The second part of this paper presents
briefly the background the policies supporting families, their relation with the standard of
living of families and how they have been progressively balanced by the issue of work and
family-life conciliation. The support to families and especially to working parents is argued to
be relatively comprehensive, i.e. quite diversified and continuous over the family life-course.
A key difference with some other Continental European countries is that women find
relatively extended support to combine entry into employment and motherhood, i.e. full-time
employment with the birth of the first child (Thévenon, 2006).
However, it is somewhat difficult to assess the effective impact of these policies on
fertility behaviour. One obvious reason is that the family policy package forms a set that can
not be reduced to the sum of its parts, as stated by Héran (2002). Thus, there are no doubt that
the density created by the whole set of family policy instruments contributes to a favourable
environment for fertility and families well-being; nevertheless, at the same time, such
interconnection of instruments makes very difficult to assess separately the impact of each of
them. In spite of such a limit, the third section reviews the studies of the impact of policies on
fertility and related labour market behaviour. We first review cross-country comparisons that
2
conclude to a rather weak effect of direct financial support on fertility but attract attention on
the impact of labour market related policies. The analyses for France also conclude to a rather
weak impact of cash and benefit transfers on fertility, but also stresses the incidence of labour
market anchored policies on female labour supply and the conciliation between work and
family-life. We argue that this rather comprehensive and longstanding policy create a rather
secure climate to enter simultaneously into employment and motherhood. It may explain why
a rather low proportion of women remain childless and the weak impact of the presence of
one child on the probability for women to work at full-time. The extensive support received to
care for children (through parental leave payment, childcare subsidies, the provision of
services, early enrolment in preschool, and the provision of out-of-school care) is also one
reason why women do not forgo having a second or a third child. Thus, it also explains why
the decision to have children or to be in employment is less polarised according to socioeconomic status than in other countries.
I.
Fertility trends and family size
I.1. Long-term or cyclical decrease in Fertility?
The high level of fertility rate in the recent years is now taken as a sign of the
optimism of the French population since this rate is for the second consecutive year at his
highest level since 1982 (Prioux, 2007). France was the first European country to display a
periodic fertility rate at 2 in 2006 (1,98 without the DOM/TOM), after an increase of 2.9% in
comparison to 2005. Although fertility rates still remain a bit lower than over the Atlantic,
France, together with Iceland and Ireland, exhibits one of the highest fertility rates in
Europe. This high level is a consequence of the reversal of trend observed since the mid 90s,
but this up-turn of fertility concerns also other countries like the UK, and the Nordic
European countries where the TFR is now higher than 1.8 child per woman. By contrast, the
fertility rates just started to increase on average in European countries, while such a reversal is
not detectable for all OECD countries.
This rebound of fertility in France makes it also different from other continental
European countries. Two groups can be distinguished among these countries (Prioux, 2007):
on the one hand, Belgium, Luxembourg and the Netherlands show a relatively stable TFR
around the intermediate level of 1.7; on the other hand, Germany, Austria and Switzerland
stand in a situation which is far much closer to other countries with low-fertility, with fertility
rates at 1.3-1.4.
3
Chart 1 : Fertility trends since 1970
4.0
3.5
3.0
2.5
US, IC
France
IR
UK, DK, SW
2.0
EU-15
OECD-27
1.5
GE
SP
JP
PL
2006
2005
2004
2003
2002
2001
2000
1999
1998
1997
1996
1995
1994
1993
1992
1991
1990
1989
1988
1987
1986
1985
1984
1983
1982
1981
1980
1979
1978
1977
1976
1975
1974
1973
1972
1971
1970
1.0
Source: Eurostat; OECD
In France, the (re-)increase of total fertility rate started from the mid-1990s after a
period of relative stability over the 80s and a slight decrease in the early 1990s. This trend
was relatively parallel to the rebound of economic growth and the decrease of unemployment
rate observed at the same time. Some analysis of the period variations of fertility showed that
the number of births depends on future parents’ subjective judgement on their economic
situation at the time of the conception of children and depends also on their expectation at the
time of birth (Bessone et al., 2003). Childbearing is so found to be higher when
unemployment is low and the expectations relating to economic growth are relatively good.
Nevertheless, in spite of this relation between economic situation and the TFR, this latter
continued to rise after 2000, while the growth rate and the associated expectations started to
deteriorate. One explanation of this may be that the longstanding and re-affirmed policy
support to families create a sufficiently confident and stable climate to persuade people not to
forgo nor to postpone childbearing.
In order to draw a picture on fertility, only relying on the evolution of periodic fertility
rate would be misleading since it is very sensitive to change in the timing of childbirth. Thus,
it is now rather widely acknowledged that the usual TFR leads to underestimation of fertility
level compared to those estimated by more sophisticated indexes that take into account the
parity distribution of women and order of births, as well as the interval duration between
births (Rallu and Toulemon, 1994; Sobotka et al., 2005; Bongarts and Feeney, 2006). The
Parity and Duration Total Fertility Rate (PDTFR, also named Period Average Parity) is such
an index which is more consistent to estimate periodic fertility (Toulemon and Mazuy, 2001).
Its value is around 2,0 children per woman in France, which is very close to the cohort
completed fertility rate estimated for the corresponding cohorts (Chart 2).
4
Chart 2: Long term trends of Fertility in France
3,1
2,9
2,7
2,5
2,3
Period total fertility rate
2,1
Cohort completed fertility
PDTFR-PAP (1960-98)
1,9
1,7
1,5
1,3
1,1
1900 1906 1912 1918 1924 1930 1936 1942 1948 1954 1960 1966 1972 1978 1984 1990 1996 2002
1872 1878 1884 1890 1896 1902 1908 1914 1920 1926 1932 1938 1944 1950 1956 1962 1968 1974
year
cohorts
Source: Toulemon, Pailhé, Rossier (2008)
Chart 2 also allows to get a longer perspective on the evolution of fertility rate over the
XXs century. It especially shows the current level of cohort fertility is quite similar to that of
cohorts born around 1900, but there has been strong variations in between: a strong increase
for the cohorts born until the end of the 60s, and a decrease thereafter. One major difference
now compared to the situation one century ago is that infant mortality strongly decreased over
the period, from 9% of children dying in their first year in 1900 to now less than 0.5%
(Toulemon et al., 2008).
I.2. Later childbirth over the life-cycle
Another important change over the last decades concerns the timing of childbirth. The
evolution of Average age at maternity, as illustrated in Figure 3, accounts for such a change in
the EU. France belongs to the group of countries with age at maternity higher than EU
average.
5
Chart 3: Average age at maternity (1995-2006)
IR
ES
30.50
SW
DK
France
GE, IC
29.50
EU-25,
28.50
PL
27.50
26.50
1995
1996
1997
1998
1999
2000
2001
2002
2003
2004
2005
2006
Source : Eurostat
One may look at the same evolution across cohorts and by parity in order to better
understand it. Chart 4 illustrates such evolution in France over all the century. It shows that
women born between 1940 and 1945 have had their children earliest over their life-cycle, at
age of 26 years on average, compared to 28,5 years for women born in 1920. Chart 4 also
reveals that cross-generational differences concern not much the first birth but rather the
timing of further birth. At age 25, two women out to three had already one child, with an
average at 24 years, i.e. one year earlier than for the generation born in 1920. For the same
cohorts, the average age at the birth of the last child decreased by two years and half and
moved from 31,5 to 28,8 years. This reduction is explained by two factors of about equivalent
importance: successive births occurred over a shorter period, but large families become also
progressively less frequent (Daguet 2002).
6
Chart 4: average age at maternity, by parity
32
31
30
Last Child
29
All Children
28
27
26
First Child
25
24
Cohorts
23
1920
1925
1930
1935
1940
1945
1950
1955
1960
1965
Source: Toulemon, 2003.
Trends are changing for women born after 1945 : they have had their children at later
ages, and the postponement of first birth explain alone the rise in the average age of women at
childbirth. For the women born in 1970, the average age at birth of the first child may be
around 27 years, and the average age at motherhood is expected to be about 30 years
(Toulemon, 2003). Prioux (2007) reported that the average age at childbirth continue to
increase for women born after 1970. Thus, this age is estimated at 29.9 years for women born
in 1973, i.e. one year later than for the generations born in 1966. She also argues that the
increase in the age at childbirth has already started to slightly slow down, and may probably
fade for the followings cohorts.
I.4. Increasing concentration towards families with two children
Changes in the family size have also been important during the last decades. Basically,
the relatively favourable level of fertility of these last years did not compensate the delay of
family formation accumulated by generations born in the beginning of the 60s. The drop of
completed fertility has been rapid for cohorts 1961 to 1964, and their fertility after the age of
30 has not been sufficient to compensate the deficit observed before this age in comparison to
previous cohorts. For cohorts born after 1964, the decrease in completed fertility continue but
at a lower rhythm.
In a long-term perspective, the decrease in average fertility across cohorts conveys a
decrease in the proportion of large families and a squeeze of family size around two children
(Chart 5). The proportion of women with two children has increased by half, from 27% to
40%. In the meantime, the proportion of women with four children or more has been reduced
by 2,5 (from 25% to 10%). For generations born in 1930, the share of mother with large
families (four children or more) was the same than the proportion of women with two
7
children ; twenty years after, this proportion is 4 time lower. By contrast, proportion of
women with either one or three children has remained relatively stable around 20%.
Chart 5: Percentage distribution of women according to their final number of children
100
5 children and over
90
4 children
80
70
3 children
60
50
2 children
40
30
1 child
20
10
Projection
0 children
0
1900
1910
1920
1930
1940
1950
1960
1970
cohort
1980
Source : Toulemon, Pailhé, Rossier (2008)
However, such a figure leads to a likely downward bias. Looking at fertility
behaviours after the arrival of children exhibits more stable patterns. Chart 6 illustrate the
parity progression ratios based on duration-specific fertility rates. The progression to the first
child seems to decline, but this may be mainly due to the postponement of family formation,
as stated in Figure. By contrast, the subsequent parity progression ratios appears to be stable,
indicating that the probability of a progression to the second, third and fourth children has not
changed since 1975 (Chart 6).
Chart 6: Period parity progression ratios
100
Per Cent
90
80
70
a0
a1
a2
a3
a4
60
50
40
30
20
10
0
1960
1970
1980
1990
Source : Toulemon and Mazuy (2001)
8
2000
I.3. Rather low proportion of childless women
One another specific characteristic of the evolution of fertility in France is the
relatively low and stable proportion of childless women, compared to other European
countries (Table 1). Thus, the actual rate of childless women among those with completed
families is about 7% in France in 2002, while it is about 12% in Germany or Austria.
Toulemon (2003) estimates from a larger survey on families that only about 10% of women
born in 1950 remained childless, and the proportion were stable since 1935 and expected to
increase a bit in future years.
Table 1 : Family size among women with completed fertility (age 40-64)
Germany
Austria
Denmark
Italy
Poland
Sweden
Ireland
Spain
Portugal
UnitedKingdom
France
EU-15
EU-25
Actual fertilitya
No child
1 or 2
3 or more
12
64.3
23.7
12.1
63
24.9
11.4
64.7
23.9
11
65.4
23.6
10.2
54
35.8
10.1
60.3
29.6
10.1
29.5
60.4
8.4
55.7
35.9
8.7
54
35.8
7.5
57.5
34.9
6.8
10.6
9.2
55.6
58.1
59.3
37.7
31.3
31.5
Source: Eurobarometer 2002 in EFILWC (2004) for actual fertility;
I.5. Low differences in fertility across level of education and social
status
Differences in family size between social groups are also relatively small and
somewhat decreasing. There are, however, some differences in fertility behaviour according
educational background. One may aim to capture these differences with a distinction between
groups of population which vary to account for cohort-specific age at end of school
enrolment. In this perspective, Robert-Bobée and Mazuy (2005) distinguished groups of
individuals with a short education, i.e with 2 years less than mean age at the end of enrolment
in education of their birth cohort, as opposed to men and women with long education (i.e.
with 2 yeas more than mean age). This enables to separate groups with around 30% of the
cohort on each side, compared with a medium group.
For men as well as women, the mean number of children has declined among all
educational level, and the differences between groups are decreasing (Chart 7). For men,
differences are relatively moderate and even decreasing over cohorts. Nevertheless, men with
a short enrolment in education system remain childless slightly more often, and they also
more often have three or more children. The contrasts are much larger for women than for
men. Among women born in 1950-54, and observed at age of completed fertility, the mean
number of children ranges from 1.8 among women with a long education to 2.4 for women
with short education. Women with a long experience in the educational system remain more
frequently childless (15% of women with higher diploma born between 1945 and 1949
compared to 9% for intermediary experience in the educational system). There are basically
9
two explanations for that situation: they live less frequently in couples, but those living in
partnership have also less children and they experience a longer period without children.
Chart 7: Differences in fertility by experience in education
3.1
Women
2.9
2.7
Men
Short
2.5
Mediu
m
2.3
Long
2.1
1.9
19 4
45
19 49
50
19 54
55
-5
9
9
19
40
-4
4
19
35
-3
9
19
30
-3
4
19
25
-2
9
-2
20
19
19
15
-1
9
4
19
55
-5
9
19
50
-5
4
19
45
-4
9
19
40
-4
4
19
35
-3
9
19
30
-3
4
-2
19
25
-2
20
19
19
15
-1
9
1.7
Source: Toulemon, Pailhé, Rossier (2008)
Differences in educational background have also significant impact on the time
interval between successive births. Thus, people with a longer experience in educational
system have nonetheless a longer period without children, but the interval period between the
first and a second child is also reduced. By contrast, individuals with low educational
enrolment have their first child more rapidly, but the second childbirth occurs significantly
later (Robert-Bobée and Mazuy, 2005). The standard of two children is also more widespread
among the highly educated. On the other hand, the share of families with 3 children has
continued to increase for women with lower education, contrary to what happened for higher
educational level. Only the proportion of very large families, with at least 4 children,
decreased also for women with low education, but in smaller proportion than for women with
higher educational level.
All in all, differences across socioeconomic status are relatively small compared to
other European countries. For instance, France does not exhibit such great polarization as
Britain, where more women decide to remain childless, especially among the most educated
groups (Ekert-Jaffé et al., 2002).
France is also characterised by a growing and relatively high proportion of out of
wedlock births, which illustrates the de-standardisation of pathways to childbearing (Chart 8).
These births even became the majority in 2006 and represent 50.5% of all births (Pla, 2007).
10
Chart 8: Out of wedlock births (% of births)
70
IC
60
50
SW
France
40
DK
UK
30
20
EU-25, IR
ALL
10
0
1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006
Source: Eurostat
I.6. A fertility lower than intended fertility?
One basic argument to advocate the development of family policies is that men and
women have not the number of children they want. The validation of this argument faces,
however, different challenges. The first one is to quantify accurately the “desire” or the
“intention” to have children, since the estimation of a gap considerably varies with different
measurements. The second challenge relates to the interpretation of such a gap, since it may
often refer the constraints that are more or less internalised by individuals. Thus, nonetheless
the declaration on fertility intention may change with the formulation of the question, but also
intentions may strongly vary with institutional and policy context.
Despite important differences in institutional contexts in Europe, Testa (2006)
reported an increase in ideal size from 2001 to 2006. France is one for which ideal size is the
highest ideal size (2.63 compared to the EU-15 average at 2.29 (only Ireland exhibit higher
number at 2.80; Denmark is very near with 2.63), and some countries (like Austria for
example) are characterised by a below-replacement family size ideal. She defines the
“ultimately intended family size” by summing up the number of children already born with
those that people still intend to have in the future, and measure the gaps between actual and
intended family size. France appears at rank 6 out to 25, and gap is estimated at means (0.72;
average at 0.73) (Chart 9). However, the gap between personal ideal family size and tempo
adjusted TFR is lower in many countries, which means that the discrepancy between the
desire and actual behaviour is lower than those illustrated in chat 9 (Lutz, 2007). The gap is
even quasi almost fully closed in countries like Austria, but this is not the case in France.
11
Chart 9: Mean ultimately intended family size, Women aged 25-39
EU-15
1.32
EU-25
1.34
0.73
0.73
0.97
Austria
0.57
1.05
Spain
0.68
0.75
Italy
1.01
1.48
Germany
0.43
1.28
Portugal
0.78
1.15
Greece
0.94
1.34
Poland
0.78
1.52
Belgium
0.65
1.11
Denmark
1.13
1.51
Sweden
0.76
1.48
Netherlands
0.82
1.64
France
0.72
1.64
United Kingdom
Hungary
1.59
Finland
1.61
0.74
0.83
1.01
1.74
Ireland
0
0.5
1.02
1
1.5
Mean actual family size
2
2.5
Mean additionnally intended family size
Source: Testa (2006)
One reason of this gap may relate to the growing perception that fertility can be a more
or less perfectly controlled and planned behaviour. The legalisation of abortion and the
dissemination contraception over the last forty years has assuredly increased the control on
fertility and it has generated a growing demand for planned parenthood (Régnier-Loilier and
Léridon, 2007). In the late 1960s, 15% of pregnancies resulting in a birth were wanted but
later, 15%. The proportion of planned birth rose from 59% in 1970 to 83% in 1995.
Prospective parents are now expecting not only to choose the number of children and the time
for entering into parenthood, but they also are inclined to choose the interval between births
and the exact period within a year for the births. After the first child, the intervals between
successive births are now more rarely left to chance. For example, closely spaced births are
unusual. While in the mid 1960s, almost one second birth in five occurred in the calendar year
following the first birth, the proportion fell to 8% in the early 1980s and to 6% ten years later
(Prioux, 1994).
Yet control over fertility is by no means perfect, since the share of “unintended”
pregnancies remains very high in France. When assessing the effectiveness of birth control,
the number of abortions must also be taken into account. Then, about 36% of pregnancies can
be estimated as “unintended”. From a survey dedicated to the analysis of fertility intentions,
Toulemon and Rita-Testa (2005) found that the majority of individuals (58.5%) report
themselves as hesitant about their intentions to have or not children. Altogether 6 people in 10
report no firm intention with respect to the children they might have in the next five years.
Thus, intentions are only one factor among many others explaining births. Nevertheless, the
authors noted that the predictive value of intentions varies with the level of education: higher
educated individuals seem to benefit from better conditions to fulfil their intentions, or they
12
3
do anticipate better their behaviour (63% have realised their intentions five years later the
intentions declared in 1998, while 48% for all the population).
To sum-up, France exhibits relatively high fertility at almost replacement rate and a
rather singular increasing trend in Continental Europe since the mid 1990s. Maternity is
clearly postponed over the years but the trend is not as high as in many European countries.
When we consider completed cohort fertility instead of periodic variations which are very
sensitive to tempo variations, the gap between fertility level at the baby boom period and
nowadays is not about one but half a child. The same statement derives from index controlling
for birth parity and age of the last child. Cohorts of women born between 1920 and 1940, i.e.
during the most of baby-boom, have had between 2,5 and 2,6 children on average, while the
cohorts born between 1950 and 1960 had a bit more than 2.1; younger cohorts seems prepared
to reach a level barely lower. Thus, cohort fertility looks very stable around or a little above
2.0 children per woman, while mean age at birth increases.
Compared to other countries, women remain significantly less frequently childless,
and, despite the growing standard of 2 children, the proportion of larger families still remain
higher than in most of European countries. The distribution of family size thus looks very
stable with about 10% of women remaining childless, 20% having 1 child, 40% having two
children, 20% having 3 and 10% having four or more children.
In Spite of this stable figure, one notes the persistence of a gap, at the aggregated
level, between actual fertility behaviour and intentions. This gap partly reflects that fertility
were progressively considered as a planned decision, although individuals are confronted with
many barriers to have clear intentions or to meet these intentions. Clearly, policies are among
those factors allowing the realisation of fertility intentions. They even may contribute to shape
such intentions.
II. From Family Income Protection to Work and family-life
reconciliation policies
It is a widespread knowledge that family policy has a long history in France and that
fertility concern has been traditionally one of its motivation. The support received by families
is thus relatively comprehensive in comparison to other countries. Nevertheless, the addition
of reforms, innovation, and the growing concerns about employment, gender equality,
children poverty, etc. has progressively induced a reshaping of family policies towards the
issue of work and family life reconciliation. The support provided to families remains,
however, quite heterogeneous.
II.1. A relatively high level of investments for families
There are many differences between countries in the extent of which these different
instruments are developed and combined. Looking at the level of public spendings for
families in the OECD countries, France provides a rather generous and diversified support to
families. Chart 10 illustrates how important is to consider tax breaks to compare the overall
support to families and to avoid any biased in the comparison; otherwise, financial support to
families may be underestimated for countries like France (or Germany) where fiscal policy
have important redistributive consequences from childless household to families with children
(Adema and Ladaique, 2005). The spending on childcare and early education services are also
relatively high since they correspond to 1.6% of GDP against only 0.87% on average in the
OECD. Taken family benefits, cash payments, spending on services and fiscal support all
together, France is even placed at the 3rd position among OECD countries in terms public
13
investments for families, as measured the percentage of GDP at 3.8, compared to the OECD
average at 2.4% of GDP.
Chart 10: Public spending on families
Public social spending on family benefits, cash payments, services, and fiscal support, 2003, as a percentage of
GDP
Cash
Services
Tax breaks towards family
OECD-24 (2.4%)
4.5
4.0
3.5
3.0
2.5
2.0
1.5
1.0
Lu
xe
m
D bou
e n rg
m
Fr ark
a
N nce
or
Sw way
H ede
u n
U
ni A n g a
te us ry
d tr
K al
in ia
g
Be dom
lg
i
Ic um
el
A and
us
Fi tria
G nlan
er d
Sl
m
ov
I an
Cz ak R rela y
ec ep nd
h u
N Re blic
ew pu
b
N Zea lic
et la
he n
rl d
P o and
rtu s
Sw Po gal
U itz land
ni er
te la
d nd
St
a
G tes
re
ec
e
Ita
l
y
Ja
Ca pan
na
d
Sp a
a
M in
ex
i
K co
or
Tu ea
rk
ey
0.5
0.0
Public support included here only concerns support that is exclusively for families (e.g. child payments, paid leave and
childcare support, etc.). Spending recorded in other social policy areas such as health and housing support also benefits
families, but not exclusively, and is not included here.
OECD-24 excludes Greece, Hungary, Luxembourg, Poland, Switzerland and Turkey where tax spending data are not
available.
Source: OECD (2007), Babies and Bosses - A Synthesis of Findings for OECD Countries.
This situation is the result of a longstanding historical process which settles a
compromise between various issues and political orientations. This compromise is still now
rooted on one historical double-faced tradition: one the one hand, family policy started with
the protection and the provision of assistance to children of deprived families; on the other
hand, the settlement of a new framework for social protection after the Second World War
made a clear issue of the protection of the income of families with children (and especially of
those with large number of children) (Damon, 2006). The issue of fertility is also a traditional
concern of this support, but its has been partly reshaped by the issue of work and family-life
reconciliation during the last decades. A consequence of that evolution is that the current
compromise that settles family policy is a balance between different objectives, but the large
set of support does not avoid some ambiguity (Thévenon, 2006).
II.2. Income protection and fertility: two early orientations of family
policy
Most of the analysts of the French family policy system have already strengthened its
complexity. Basically, family policy is a package with a mix of no less than about thirty
different measures which aim to help parents at the time of a birth or later to support them for
childrearing, to provide education, or help them in combining work and family life. It is of
course difficult to draw very clear tendencies from this complex system. Nevertheless,
compared to many other OECD countries, the support received by parents in France can be
distinguished by three main characteristics (Adema and Thévenon, 2008). First, most of the
support is provided the State, while the contribution of employers is rather weak. Second, the
types of supports appear to be greatly diversified and comprehensive. Thirdly, this support
covers rather continuously the period of childhood.
14
As already stated, the income transfers occurring through the tax and benefit system
are the first pillar of the support received by families. Cash family benefits are one important
piece since they represent about half of the benefits paid to families (Chart 11). They were
introduced with the new social protection system held in 1945, with a design that will not
much vary until now. Only families with at least two children can receive these fixed amount
benefits which are not mean-tested and increasing with the number of children. Some various
pre- and post births grants complete this set (see appendix 1). Transfers to families also exist
through tax breaks granted by the “quotient familial” mechanism since 1945 tax allowances.
This mechanism is designed to reduce the cost of children and offers tax reductions that are
particularly important for families with at least three or more children. Since tax reductions
are proportional to the basic taxable income, the ‘quotient familial’ is especially beneficial for
the richer large families (see box 1). The support to fertility, especially to increase the
propensity to have a third child, is an explicit motivation of this focus (Thélot and Villac,
1998). The pro-natalist argument still continues to shape State support to families, although
recent surveys showed that it is no longer very well received in public opinion, and that the
notion of children rights is starting to substitute to it (Letablier et al., 2002). Moreover, the
above mentioned changes in fertility and conjugal behaviour (illustrated by the move to the
norm of two children, the increase in the number of lone parents) has created debates on the
adequacy of the design of family benefits and of the quotient familial. Despite some reform
attempts, no sufficient consensus has emerged until now to change fundamentally their
baseline.
Chart 11 : Distribution of spending on families in 2005
3%
0%
6%
9%
10%
11%
8%
53%
Parental leave benefit
Support for employed home
Maternity/paternity leave
Family benefits
Social assistance (API, AAH)
Tax credit for employers
State Investment in childcare and preschool
City spending in childcare st
Source : calcul by the author from data provided in CAS (2007), Rapport sur un service public
pour la petite enfance
In all, at the time of just after the Second World War II, the support is mainly focused
on large families and grounded on a representation of traditional male-breadwinner household
(Hantrais and Letablier, 1996). The 70s are characterised by the increase of the assistance to
families with temporary economic or social difficulties. New targeted benefits are created to
15
help people facing social or family difficulties. These benefits are reserved to specific
population and mean-tested. Lone parents with children under age 3 can so claim for the
Allocation pour Parent Isolé (API) benefit; housing benefit are also created during the 70s
with amounts that depends on family composition. One childcare benefit is also created since
1972 for working mothers. The end of the 70s is certainly a turning point since the concern of
family policies gradually shifted from fertility to the issue of reconciling work and family.
Box 1: The quotient familial
An original mechanism, which is implemented only in France, is the “family ratio” (“quotient familial”)
integrated to the income tax schedule. This device operates as follows :
- let R be the taxable income of a household;
- one calculates a number N of units depending on the family size : one unit for each adult, 0,5 unit for each
child up to parity 2, 1 unit for each child from the third child or for the first child of a one-parent family (in the
general case);
- the “family ratio” Q is equal to R divided by N;
- the marginal income tax t is an increasing function of Q, and not of R;
- the tax T owed by the household is equal to : t(Q)*R-v(Q)*N, where v(Q) is determined to ensure that there
is no discontinuity in the function T(R) when N remains constant.
Hence the fiscal incentive to families is more than a simple lump sum tax reduction per child, as it would be if t
and v depended on R and not on Q, but increases with the income R. Actually, an extra child induces a lump sum
tax reduction (0,5*v(Q) or v(Q) according to the rank of this new child), plus a diminution of t(Q) since Q
decreases. However, “family ratio” is not a tax credit: households with low income who are exempted from
paying the income tax do not benefit at all from it, and households paying a little amount of income tax receive
only a small tax reduction. Hence, because of its counter-redistributive effects, there is a ceiling for the amount
of the tax reduction due to the “family ratio”: 2 200 euros a year for a first or second child, 4 400 euros a year for
a child of parity 3 and over or for the first child of a one-parent family. However, the total “cost” – in fact, losses
for the Government budget – of the “family ratio” may be estimated, using microsimulation models9, to 0.7 % of
GDP in 2004, namely one third of the total amount of cash benefits for families. This confirms the crucial role
“family ratio” plays in reducing the financial burden of families.
The rationale for such a mechanism is the 13th article of the Declaration of Human and Citizen Rights of 1789,
which is part of the current French Constitution. This article tells that the tax owed by households must depend
not simply on income, but on “contributory capacity”. This notion has been interpreted for a long time as
imposing on the tax schedule to take into account the family size, and the French Constitutional Court has
continuously recalled to the Parliament that it could not create a new tax or change an existing one without
taking into account and limiting the impact on household income according the family size.
Source: Caussat (2006)
II.3. A growing but ambiguous focus the work and family-life
reconciliation
However, the support to working mothers is developed in an ambivalent context tied
between two traditions: the ‘familialistic’ tradition which support the focus on traditional and
large family, and the ‘feminist’ movement which claims the individualisation of social rights
and the participation of women in the labour market (Commaille et al., 2002; Revillard,
2006). The issue of work and family-life conciliation is growing during the 70s and 80s. They
add motives for the development of childcare services to those related to the education and
development of children. Nevertheless, the support to working women is developed with
ambiguity, notably because family policy has remained thought and conducted separately
from policies for womens’ rights and gender equality (Strobel, 2004).
This ambiguity is increased during the 80s and 90s because family policy evolves in a
context with strong budgetary restrictions, and because it is more directly subordinated to
employment concerns. Some authors even argued that Family policy has been gradually
16
shaped as an instrument of employment policies on two fronts: unemployment and the
development of domestic and personal services (Martin et al., 1997; Fagnani, 2003). On the
one hand, the Allocation Parentale d’Education is settled to encourage working mothers to
leave temporarily the labour market for childrearing (and thus to reduced the job queue).
Thus, mothers with 3 children can receive, under conditions on prior activity a fixed grant if
they stop working or switch to part-time work until their youngest child reach his third
anniversary. This benefit was extended to mothers with 2 children from 1994. The 2004
reform of the Prestation d’Accueil du Jeune Enfant (PAJE) extend the possibility for parents
to interrupt activity from the birth of the first child for a period of six months. In that case,
they receive the Complément de Libre Choix d’Activité (CLCA), monthly amount of 522€ . In
year 2005, the amount spent for the parental leave benefit (APE and CLCA) is, at 9% of the
family support, approximately the same than to total amount spent in childcare and preschool
structures by the State and municipalities (Chart 11).
On the other hand, the has been involved in the development of individualised and
flexible home care services is supported with tax reduction for household who employ homebased family carers in order to both limit the cost of childcare and create employment.
Subsidised childminders, allowances and tax reduction for household paying childcare costs
were also introduced in the early 80s. Since 1980, an allowance is paid to families with
children under 3 (extended to children under 6) who are cared for by a home-based registered
childminder. In 1986, tax allowance was introduced for families who employ nannies at
home, i.e. richer families than those relying on the crèche or on a registered childminder.
Parents are eligible to deduct from taxable income half of the childcare cost. The net
reduction of tax covers about 70% of the cost (Fagnani, 1998). The support granted through
this mechanism is sizeable since it represents about 10% of the amount spent for families.
II.4. A policy that benefit rather to richer families…
This brief historical presentation shows how transfers are multi-oriented. Some recent
papers attempt to evaluate how they benefit to families with children compared to childless
household. Globally, they state that these transfers are larger for families with a youngest
child under age 3, and they are also more important for large and richer families. Over all
families, the average amount of transfers induced by the presence of children is U-shaped: the
presence of children increases the amount received by low earnings families though social and
housing assistance; while the mechanism of the “quotient familial” is clearly beneficial for the
richer households (Bechtel et al., 2005; Chart 12). Compared to other OECD countries, the
ratio of transfers received by the lowest income families compared to the richest is relatively
low and about half of the OECD average (Adema and Thévenon, 2008).
17
Chart 12: Gain in disposable income due to the presence of children
Average amount of transfers per child (by brackets of initial income)
4 500
Families with children
(youngest <3 years old)
All families
Families without children under age 3
4 000
3 500
Euros
3 000
2 500
2 000
1 500
1 000
500
<5
05
-1
0
10
-1
5
15
-2
0
20
-2
5
25
-3
0
30
-4
0
40
-5
0
50
-7
0
>7
0
<5
05
-1
0
10
-1
5
15
-2
0
20
-2
5
25
-3
0
30
-4
0
40
-5
0
50
-7
0
>7
0
0
>7
0
10
-1
5
15
-2
0
20
-2
5
25
-3
0
30
-4
0
40
-5
0
50
-7
0
-1
05
<5
0
income before transfers (brackets of initial income)
Family benefits
Tax breaks
Social assistance and housing benefit
Total
Source: Bechtel et al. (2005) ; Simulation of the legislation in year 2004 with income on
2001.
II.5. … and to households with children under 3 years old and
larger families
Globally, the transfers received by families help at reducing the gap of income
between households with and without children, and thus contribute to reduce income
inequalities. They also contribute to reduce poverty rates from 21% before transfers for
families with children to 7.2% after those transfers (Courtioux et al., 2005). Restricting to
households with a young child under age 3, the support due to children is increasing with
primary income, because of the reduction of tax due to childcare costs. Concerning the
progressiveness of support with the number of children, the transfers are more important for
large families in France at all income level, while it is rather proportional to the number of
children in a majority of European countries (Math and Meilland, 2007).
Chart 13 shows also the gain obtained from the support to pay childcare cost (CMG)
increases with initial income. For couples with one child, the benefit of childcare support is
clearly higher than those received through the home care allowance (CLCA) for most of
households whatever their position in the distribution of income. The figure changes for
couples with two (or three) children since we can see that households in the 4 first deciles of
income gain far much from the home care allowance than from the addition of subsidies and
tax reduction received to compensate childcare costs.
18
Chart 13 : Income gain due to the presence of children within household
Average amount per child
Couples with children below 3 yeas old
2 000
Couples
with 1 child
1 800
1 600
Couples
with 3 children
Couples
with 2 children
Child care support
(CCA, CMG)
1 400
1 200
Euros
CMG (Cotisations
1 000
800
600
CLCA2
CMG (Complément)
400
200
CLCA1
0
<10 10- 15- 20- 25- 30- 40- 50- >70
15 20 25 30 40 50 70
<10 10- 15- 20- 25- 30- 40- 50- >70
15 20 25 30 40 50 70
<10 10- 15- 20- 25- 30- 40- 50- >70
15 20 25 30 40 50 70
Income before transfers (by brackets)
Childcare support
CMG (complement)
CMG (cotisations)
CLCA1
CLCA2
Source: Bechtel et al. (2005) ; Simulation of the legislation in year 2004 with income on
2001.
II.6. A comprehensive provision of childcare services for the
conciliation between work and family life.
Without considering the tax breaks due to childcare costs, the effective tax rate of a
second worker in the household is clearly below the OECD average, but the incentive to get a
second worker rather than an increase of the first earners’ labour supply is especially low,
given the ‘quotient conjugal’ mechanism. The net cost paid by two earner parents with two
children using full-time care is, in percentage of household net income, about the OECD
average (Barber and Immervoll, 2005). Nevertheless, tax breaks contribute to reduce the
childcare costs and differences in cost between a place in collective crèche and the
employment of a registered childminder (Fagnani, 2003).
This support encourages the conciliation between work and family life, and completes
the relatively comprehensive provision of childcare services which is given by the system of
public Crèches, and also mainly by the early enrolment of children into preschool (Ecoles
maternelles). As already stated, the investments in early childhood services are comparatively
high in France, and represent almost one tenth of the total mount spent for families (Chart 10).
The crèches cover the major part of the day (from 7-8am to 6-7 pm) and are available
immediately after the end of maternity leave, i.e. from the age of two or three months. This
makes an important difference with, for example, the situation in Nordic European countries
where collective public services are only available when the child is one year old.
The Ecoles Maternelles are preschool institutions created in 1881, and large
investments were made since the 1970s to increase their total amount of places. Public state is
19
encouraged to enlarge preschool capacity since one place is relatively cheap compared to the
cost of other child care solutions (Adema and Thévenon, 2008). These preschools are free of
charge for parents (except for the provision of lunch) and often available for children from
age 2. They provide care support to working parents and are relatively compatible with fulltime work since children are basically cared from 8.30-9 am to 11h30-12, then from 1.30 to
4.30 pm. Most of the preschools provide lunch and out-of-school care after school hours (until
6-6.30 pm).
A recent survey (conducted in 2002) on childcare estimated that although the largest
proportion (61%) of children under age 3 are cared mainly by their parents during the
weekdays, 21% are cared by subsidised child-minders, 10% are cared in Crèches. 7% are
cared mainly by their grand-parents and only 1% by nannies at home (Blanpain, 2006). For
older preschool children, about 37% of children of aged 2 are enrolled in preschool but 97%
from age 3 and universal coverage at age 4. combined with other child care solution since
only 14% of children of age 2 attend school on a full-time basis, 21% on a part-time basis.
Nevertheless, most of children can attend out-of school care from 4.30 to 6-6.30 pm.
However, about 2 to 6 years old children are cared only by their parents after school day, and
only 17% use out-of school care.
The contribution of the school system to child care stretches also beyond preschool.
Thus, school days are relatively long in France compared to other European countries, and
out-of-school care is often provided (Ananian and Bauer, 2007). In all, 56% of children in
primary and secondary school have lunch at school, and 21% resort on after school care. This
use is more frequent when both parents work, especially women work on Wednesdays.
However, about one women in five care for their children on Wednesday even when they
work all the day, and almost 27% of school aged children are cared by their grand parents or
other family member.
III. The impact of policies on fertility… and work and family
–life conciliation
Basically, the reason why Family-friendly policies are expected to impact fertility is
because they can reduce the cost of children and can improve the conciliation between work
and family life. These policies refer to a large range of institutional instruments including tax
credits, cash benefits, and other financial transfers that aim to limit the potential loss of wellbeing that the presence of children can produce within a household; other benefits or transfers
provide specific assistance for deprived families, or support for housing; some other devices
are rather aimed at helping parents to balance work and family life. Then, benefits and tax
allowances can be designed to help families to meet their childcare cost. Childcare services
and parental leave provisions are also of prime importance to give parents the opportunity to
combine employment with family and care responsibilities, and thus they impact the
opportunity costs of motherhood. Some other tax transfers or subsidies related to labour
market status impact both the direct and the opportunity costs of children. In this case,
policies can create opposite incentives on childbearing.
III.1. Some difficulties to measure policy impact
Cross-country analyses suggest that total fertility rates are higher in OECD countries
with wider childcare availability, lower direct cost of children, higher part-time availability
20
and longer leaves. The precise evaluation of the impact of each of these policy variables faces
several difficulties:
• First, the range of policy instruments that can influence fertility is broad, and
this large range makes necessary to define conventionally the scope and borders of such
policies in order to enable cross-country comparisons. Furthermore, given this wide range of
instruments, its is difficult to capture the potentially complementary nature of instruments, i.e.
the fact that one instrument would have an effective impact only with its combination with
another one or if the associated support is higher than a given threshold. In particular, the
effectiveness of policy instruments may depends on the existence of a complementary and
continuous support over the period of childhood, to deal with the different aspects of
childrearing and the evolution of needs over the life-cycle.
• Second, there are also obvious temporal lags in the adjustments of reproductive
behaviour to policy changes, and such lags make difficult to assign the impact of specific
instruments or specific policy change.
• A third issue is that both fertility and labour market behaviours are decided
simultaneously and this may induce some endogeneity within explanatory variables such as
mothers’ labour supply. This issue is, however, now relatively well-treated when longitudinal
data are available
• A fourth difficulty lies in the estimation of the real cost of children, which
varies with household economic and demographic situations. Thus, most of cross-country
comparison do not account for such inter-group heterogeneity and focus on the average
values. However, policy impact may vary across socio-demographic groups, such as ethnicity,
socioeconomic status, and also according to the form of the relationships between parents
(Sleebos, 2003). Gauthier (2007) stresses, for example, that policy have differential impact by
birth order.
• A related issue is to circumscribe the scope of policy measures since they can
have an impact on the behaviour of subgroups of population who are not directly targeted by
these measures. As wee will see, parental leave payment paid in France to parents with two
children may have had an important impact on behaviour of larger families. Moreover, very
active policies, even targeted to specific groups of population, may act as a signal of a support
to families that can affect a wide proportion of families, by contagion effect.
• This possibility leads to a sixth difficulty: the potential endogeneity of policies.
The development of family policies can surely induce, but can also be generated by it and be a
mark. Long term series finds evidence that childcare policy has been developed, for example
in Sweden, in parallel to the expansion of female labour participation, but was not a prerequisite (OECD, 2005). In that case, childbearing would not be consequences of policies but
the causal relation may be the reverse. Policy may thus contribute to the changes in the
family-related values and in fertility intentions.
• Last but not least, the evaluation of policy should consider a broader impact
than the (immediate) response to financial incentives. Thus one effect of institutions is to
reduce uncertainty and to enable people to make long term commitments (as childbearing is).
There are, however, several conditions that policy background must meet to have such effect.
A longstanding policy is in particular required in order to keep trust and to convince people
that no very profound changes will occur in the future. A continuous and diversified policy is
also needed to guarantee a long-term support and to create ‘systemic’ coherence. This is
certainly a key aspect to explain France performance, but this aspect goes beyond financial
incentives.
21
III.2. Some evidence from comparative studies: from the impact of
financial incentives to the assessment of work and family-life
reconciliation policies
Given these obstacles, it is an intricate piece to assess the exact impact of family
policies on fertility. One approach to such assessment is given by cross-country comparisons
which investigate the effect of policies on fertility rates in industrialised countries. Most of
these studies consider the aggregated level of TFR, and merge the data concerning all
countries, even if they allow country-specific effects. They used either cross-section or timeseries or longitudinal data.
In a survey published in 2003, Sleebos asserted that findings are often inconclusive or
contradictory, partly because of methodological differences. The review of studies suggested,
however, a positive but weak relation between reproductive behaviour and family policy. The
same dubitative statement is reiterated by Gauthier (2007). One obvious reason for the lack of
evidence is that policies’ influence can be detectable only in the long-term. However, both
authors point out that the more obvious and robust results concern the impact of cash benefit
and transfers, while the impact of work-related policies are more contradictory: the duration
and payment of maternity leave are generally found to have a rather weak impact, while some
studies conclude to a larger impact of childcare availability and/or affordability. Flexible
working hours and the availability of part-time work are also pointed as important factors
influencing fertility. Thus, a key issue for fertility decision appears to be the possibility to
combine family formation with labour force participation. For that reason, attention should be
paid to the policies that aim at reconciling work and family-life.
Among the most significant studies assessing the impact of cash transfers, Blanchet
and Ekert-Jaffé (1994) investigate the effect of family benefits on the TFR of 11 industrialised
countries over the period 1970-1983. They construct a family policy index that account the
global generosity of financial support through cash benefits, tax relief, parental leave
compensation, and found a rather weak impact of such a package. They found that the French
family policy could lead to an extra 0.17 child per women as compared to some other
developed countries like the UK. Such an impact is not insignificant, but shows that that no
major fertility upswing may be expected from family policies.
Gauthier and Hatzius (1997) modelled the dynamic relation between fertility rates and
policies for 22 OECD countries over the period 1970-1990. They found that neither the
duration nor the benefits provided by maternity appeared to be significantly related to fertility.
By contrast, direct cash benefits are found to have positive and significant effect, although it
is small. They also considered the differences in policy impact according to birth parity and
found greater effect of benefits for the first child. One of their conclusion is that targeting
benefits at the third child, as is done in France, is unlikely to increase fertility. Overall, the
magnitude of the effect is however small since their estimation is that a 25% increase in
family allowances would result in an increase of fertility of about 0.07 children per women on
average. However, they also observed differences in the responsiveness across groups of
countries. Basically, a greater impact is found in Scandinavian countries and is interpreted as
the effect of the co-variation of in-kind support in this countries. Thus the authors (indirectly)
stress the importance institutional complementarity although they could not quantify it
directly.
Comparing changes in cross-section data, Castles (2003) argue that the provision of
child-care facilities for children aged 0-3 which is crucial to early labour force re-entry may
have been the main factor contributing to the reversal of the relationship between fertility
22
rates and female labour market participation in OECD countries. Micro-evaluation evidence
are, however, more mitigated. While a positive impact of reduced childcare cost and increased
childcare availability on fertility is estimated by Di Prete et al. (2003) and Del Boca et al.
(2007), no statistically significant impact of childcare characteristics is reported for example
by Ronsen (2004) for Norway and Finland, Hank and Kreyenfed (2003) for Germany and by
Andersson et al. (2004) for Sweden.
More recent studies confirm the impact of standard family policy, and provides also
evidence on the effect labour market and work organisation. D’Addio and Mira d’Ercole
(2005) analyse cross-country differences in total fertility rates in 1999 for 19 OECD
countries. Their investigation is based on models that allow for dynamic effects, potential
heterogeneity between countries, and endogeneity of some of the explanatory variables. They
find evidence that transfers to families with children, as well as the provision of services to
working parents to better cope with their care responsibilities have positive impact on
childbearing. The impact is, however, relatively weak: an increase in the total length of
parental leave by one week, in the percentage of wage replaced and transfers to families
would on average an increase in total fertility rate of respectively 0.3%, 0,9% and 1%. They
also estimate that female labour market participation, the share of women in part-time work,
and the ratio of female to male hourly earnings have all a positive impact on fertility.
Del Boca et al. (2007) also modelled the role of childcare arrangement, parental leave,
family allowances and labour market flexibility, but adopted an individual-based approach by
and on womens’ joint decision towards fertility and labour supply. Their results based on the
ECHP show that a non-negligible portion of the differences in participation and fertility rates
for women of 6 European countries (including France) can be attributed to the characteristics
of this institutions, but the impact vary with educational levels. Childcare availability and
optional leave have larger impacts on the fertility and participation decisions at lower
educational levels. Parameter significance on fertility is, however, weak. By contrast, labour
market policies, such as part-time opportunities have larger impact on outcomes of women
with higher education. In all cases, the impact is more significant and larger on labour supply
than on fertility.
When assessing the effectiveness of policies in raising fertility, an important issue is
also whether these policies impact the “completed” fertility rates or only the timing of births.
In the first case, policies would have long-term effect, while they would be temporary in the
second case. Studies that tried to distinguish between the two, using age and parity specific
fertility rates, tend to conclude that impacts are more significant on the timing of fertility
rather than on the total number of children (Sleebos, 2003). Andersson (2001) suggests, for
example, that the introduction of a “speed premium” in the parental-leave system of Sweden
accelerated childbearing decisions, by reducing the spacing between the first and second
births.
III.3. Some evidence for France at the crossroads of the literature on
fertility and female labour market behaviour.
In this context, there is only a limited number of papers investigating the impact of
family-oriented policies on fertility in France, and which can be use to interpret previous
mentioned specificities in fertility trends. Contrary to the previous mentioned comparative
studies, most of the French focused analyses are based on micro-behaviour and individual
data. Some of these studies have focused on the impact of financial support on fertility. Other
and more recent studies considered the effect of specific work-related policy measures on
fertility and its combination with female labour market participation. In particular, many
23
studies focused on the incidence of the paid parental leave scheme (the APE) on female
labour market behaviour, but only one on childbearing. Many studies also analyse the impact
of childcare characteristics (cost and availability) on female labour supply, but none
investigated, to our knowledge, their specific impact on fertility. These papers, nevertheless,
shed light on the relatively comprehensive support that help many women to combine a fulltime labour force participation with the birth of their first children. Policies that secure the
conciliation between labour market participation and the entry into first motherhood appears
indeed of first importance. There are, however, no studies investigating very specifically the
role of flexible working hours or part-time work on fertility.
III.4. A rather weak impact of financial incentives, but larger for
higher parity births
Ekert (1986) is among the first studies to estimate the impact of financial (non work
related) transfers to families on fertility. Ekert evaluated the incidence of direct cash benefits
(including the “universal” allocations familiales, the mean-tested Complement familial, and
also housing benefits) on fertility in the ending 70s and observed a significant but weak
impact since these benefits are estimated to contribute to fertility rates for about 0.2 children
per woman. The author also expects that a full payment of the child-related cost would
increase fertility by about only 0.3 children per woman. As stated previously, Blanchet and
Ekert (1994) found a rather similar impact from a cross-country comparison.
Ekert et al. (2002) also suggest that family policy is an important factor that reduces
the differences (and polarisation) in fertility behaviour between socio-professional categories,
especially compared to countries like the UK. Nonetheless almost all women enter in
motherhood in France, but the propensity to get a second child is only very weakly dependent
of social categories. The polarisation of fertility behaviour across social groups is far more
important and start from the first child in the UK, where the support to families and to
working mothers is far less developed.
A more recent piece by Landais (2003) scrutinised the impact of the ‘quotient familial’
on the fertility at parity 3. For this, he evaluated the impact of a change that occurred in 1981
in the tax legislation which considered after this date a third child as a full adult while they
previously contributed as half an adult in the calculus of the tax allowance (see box). He also
concluded to a positive but very weak: a 1% variation of the incentive relative to the
households with three children produces a relative increase of their proportion of less than
0.05%. He observed also that the sensitivity of fertility behaviour to fiscal incentives increases
with income (and thus with the magnitude on the incentive). There is, however, a long delay
(between 5 and 10 years) before that the effect are fully disseminated and can be evaluated.
III.5. The 1994 parental leave reform: a limited (positive) impact on
fertility…
Breton and Prioux (2005) also investigated the role of family policies on the transition
from two to three children. They concludes that the measures targeted on the third child had
visible impact on the timing of birth. They may also had contributed to the stability or even
the slight increase in the probability of having a third child, especially the Allocation Parental
d’Education introduced in 1985. This probability is now decreasing, but the targeting of
family support has also diminished, and second birth are now more evidently supported. They
observed the rather same cyclical variations over the period 1970-2000 of the probability of
having a second and third child, but with higher magnitude in the later case when the targeted
support to the 3 children families were increased. However, they could not quantify this
24
impact without any prediction of what would have happened without these policy changes.
Stating the relative similar cyclical trends of the probability of having a child at parity 2 and
3, they also suggest that policy may have had an impact beyond their target (families with 3
children), i.e. on fertility behaviour at lower parity, by effect of contagion due to the secure
climate that it creates and the large campaigns of communication that were associated with
the reforms.
Chart 14: Policy support for a second and third child and associated fertility trends
per
1000
800
Cohort PPR
A
1
800
From first to
second child
a1
Period PPR
700
APE 2nd
Reduction in 3rd
child benefits
600
700
600
Cohort PPR
A
From second to
500
third child
500
a2
Period PPR
400
300
200
1970
1966
400
300
"Giscard" APE 3rd chil
measures (progressive
extension)
1975
197
1980
1976
1985
198
1
1990
1986
Source : Breton and Prioux (2005)
1995
199
1
200
2000 Year
Cohort
In order to determine, Laroque and Salanié (2004; 2005) investigated the role of
financial incentives through tax and benefit transfers received by families. used a microsimulation model with endogenous female labour force participation and fertility response to
financial incentives. The results of these two papers are, however, rather different and
sometimes contradictory. Basically, Laroque and Salanié (2005) stressed two main
differences. The first one is that they used in the first one a fully myopic model, so that the
duration of benefits were not well taken into account. The second one is that they included
fewer variables in the fertility equation in the first generation of papers. Thus, the first results
were marred with omitted variables bias and the 2005 publication provides, according to the
authors, more reliable estimates. For that reason, we consider here only the results of this
latest publication which provides an ex ante assessment of the impact of the financial transfers
received by families. Here again, the impact of financial support is found significant but
relatively weak. In all, the cost elasticity of the “demand” of children is estimated at 0.2: a
25% reduction of the cost of children would increase fertility by only 5%. However, the
responsiveness of fertility behaviour to vary with birth parity: higher is the rank of birth, more
sensitive is fertility behaviour to financial variations. Thus, the estimated for example that the
increased incentives to mothers of a second child to exit out of the labour market that
25
occurred in through the extension of the APE would have increased the number of births by
3.7%, namely an increase of 10.9% of the births at parity 2, reduced by a decrease of the
number of births at parity 3 by 2.4% since the birth of a third child became less attractive 1.
Whatever the robustness of their estimation is, one may be cautious to use these results
to assess the effectiveness of financial incentives on fertility behaviour. Some restrictions on
the sample design2 and on the fertility equation can be criticised. Moreover, ex ante
simulations are certainly not sufficient to assess the impact of policies, and must be
complemented by ex post studies based on actual behaviours.
In this perspective, Piketty (2005) also aimed at quantifying the incidence of the
parental leave compensation (APE) on both fertility and womens’ labour supply ex post
evaluation with both Labour Force Surveys and the 1999 Survey on the Family. The extension
of the grant for the birth of a second child in 1994 serves as a natural experiment. Here, the
challenge is to control for possible windfall effects, i.e. for the fact that some women would
have had a child or leave the labour market without change in the legislation. Regarding
fertility, the effects are hard to quantify, but Piketty argued that the extension of the APE
would have not explained more than 20-30% of the increase of the total number of birth
observed from 1994 to 2001 (at maximum 10% of the third rank births and between 10 and
20% of the births of rank 2)
III.6. … But a larger (negative) impact on female labour market
participation
The effects on female labour force participation are more spectacular, still according to
Piketty’s results. Indeed, the payment of the APE to mothers with two children (with the
youngest under 3) would have induced between 100.000 and 150.000 withdrawals from the
labour market. Thus, between 50 and 70% out of the 220.000 recipients of the APE at full rate
would have not interrupted their activity at the second birth if they have not been covered by
the benefit.
The author also pointed out cumulative effects since having received the APE for a
second birth seems to have led about 50.000 mothers with 3 children to leave the labour
market after the third birth. Thus, the APE reform would have implied between in total
150.000 and 200.000 interruptions of female labour market interruption. Moschion (2007)
women have anticipated their labour market withdrawal: whereas before 1994, mothers
moved away from the labour market when they had a third child; since 1994, the first
withdrawals occur more frequently for the second birth. She concludes that this effect may
prove the role of financial incentives.
Finally, Piketty’s results showed also such that mothers with two children do not seem
having suffered from particular difficulties to return to the labour market since their
probability of working increased over the period. A similar statement is made by Thévenon
(2007) who observed that the probability of being active (at full-time) increased of about the
same proportion than for childless or mother of 1 child over the period 1996-2005, after a lag
which may be explained by the extension of the APE.
1
Note, however, that this result is contradictory to their previous one since they estimated a strong influence of
financial incentive on fertility at parity one, but none for two-children or larger families (Laroque and Salanié,
2004).
2
In particular, the sample leaves aside households which contain an adult retiree, or a self-employed person or a
civil servant ; it also focused on women who have left school for more than two years, so as to avoid dealing
with the schooling versus work decision.
26
III.7. A rather large participation in full-time employment of women
before the birth of a second child.
This trend illustrates the relatively specific situation of France in terms of female
employment and family-life conciliation. Compared to other OECD and European countries,
the employment rate of women of age 15 to 64 years is just above the OECD average at
58,6%. Looking at the correlation between the period fertility and female employment rates in
2005, one can see that compared to other countries with same rate of female employment, the
fertility rate is relatively high (especially compared to Continental, Southern or Eastern
European countries, Korea or Japan). Of course, the correlation can be looked at the other
way round, with the less positive interpretation that for a given fertility rate, female
employment rate of French women is quite lower than in Nordic European and Anglo-Saxon
countries (US, Australia, New-Zealand and the UK). However, one difference with this last
group of countries is that full-time employment remains a wide normative working time basis
in France, although part-time work continued to significantly increase over the 90s
(Thévenon, 2007). In France, like in Denmark, Canada, Finland, Portugal or Sweden, the
most frequent situation is clearly when both parents works at full-time (Aliaga, 2005; Adema
and Thévenon, 2008). Compared to other Continental European countries, a rather specific
aspect is how labour market behaviour of French women adjusts to each stage of family
formation (Thévenon, 2006). The incidence of the first birth is very weaker on the exits from
the labour market, and weaker that the birth of higher parity (Thévenon, 1999). Furthermore,
full-time employment rate is more sensitive to the number of children, and especially to the
presence of a third child, than to the age of these children. Part-time work concerns first
mothers with two or three children, and implies relatively long working hours compared to the
practices in other European countries (Thévenon, 2007; Chaupain-Guillot et al., 2007).
III.8. Childcare costs have more impact on the use of paid care than
on female labour supply.
In this context, it is important to consider policies that may affect primarily female
labour supply, whenever it is not possible to consider simultaneously their potential induced
impact on fertility3. In that case, most of studies limit their investigation on the effect of
institutional variations on the use of policy measures and female labour supply. There is only
a very limited number of studies in France and they mostly investigate the impact of child
care cost on the demand of formal paid care and female labour supply. Choné et al. (2004)
found that globally child care cost impact the use of paid care but has only very low incidence
on female labour market participation. These costs, together with female potential wages,
have more clear and significant effect on working hours. A result of their estimation is also
that a suppression of the APE at parity 2 would have a significant impact on female labour
participation (+11% among women with 2 children and the youngest under 3) and on the use
of formal paid care (+4%). The global impact on employment rate would however be
moderate with an increase of about 4% due the relatively small share of population concerned
by the APE. By contrast, the reduction of tax associated to childcare cost would have only
little impact.
3
Because of the too large number of structural determinants that would have to be considered and also lack of
data, it is often impossible to consider this impact simultaneously on fertility and labour supply, taken as joint
decisions.
27
III.9. A probable (but uncertain) effect of early preschool enrolment
of children on female labour supply…at full-time.
Two very recent studies investigate also the role of early preschool on female labour
supply. Results are, however, rather uncertain. De Curraize (2005) analyses the incidence of
early school enrolment (at age 2) of the duration of mother’s labour market participation
interruption and finds that it is impossible to conclude to a positive effect. Moschion (2007)
finds that supplying mothers of two years old children with developed child care modifies the
effect of fertility on mothers’ labour supply and may have a positive impact the combination
between work and family life. In particular, she finds that having more than two children has
a negative effect on the participation and hours worked in departments with relatively rates of
enrolment of children at age 2. Conversely these effects are positive in departments where the
schooling is high. The estimations are, however, non significant due to lack in precision, and
thus must be interpreted cautiously. Nevertheless, comparative studies suggest that the entry
of children into the preschool at age 3 is a turning point since female employment rate at fulltime is then significantly increasing, this increase being higher than those observed at a later
period of childhood (Thévenon, 2007).
Conclusion: towards a broader approach of policy impact on
fertility regulation?
As stated in the first part of the paper, France exhibits relatively high periodic fertility
rates compared to other European countries, and singularly increasing since the mid-1990s.
This rates even appear rather stable after controlling for tempo effects. Thus, the increase of
the TFR seems to rather reflect some catch-up effect from women who postponed
childbearing. In a more long-term perspective, the decrease in (completed) fertility refers to
the growing standardisation of fertility behaviour towards two children. The proportion of
women remaining childless is also lower than in many other European countries.
Family policies do contribute to facilitate the entry into motherhood. On the one hand,
financial transfers through cash benefits and fiscal allowances reduce the ‘cost’ of children
but their direct effect on fertility are estimated to be weak. On the other hand, substantial
support is also provided to families with children under the age of 3 or 6 years to alleviate the
cost of childcare, and the literature suggests a large responsiveness of female labour market
and its coordination with family formation to these financial support. However, the financial
aspect is certainly only one part of the story to account for the ‘stylised facts’ reviewed in the
paper. In particular, one reason why childlessness is less frequent in France is certainly that
the birth of a first child is more compatible with the insertion into employment, and even in
full-time employment which does not prevent from future career development. One argument
in this direction is that the “stability in employment” appears to be a growing precondition
before having a first child (Toulemon and Leridon, 1999; Méron and Widmer, 2002). Thus,
empirical evidence shows that women who experienced a period of unemployment have
relatively postponed the birth of the first child compared to both inactive and employed
women, and this is more prevalent for less qualified women and younger cohorts (Méron and
Widmer, 2002). Among the individuals who intend to have a child, the experience of
unemployment also reduce the probability to realise the intention within an interval of five
years (Toulemon and Testa, 2005).
Then, the more frequent (or more rapid) to first motherhood in France may be
explained by the quite highly comprehensive support received by women to combine
motherhood with a full-time labour market participation. This support is indeed, as stated in
28
the second section of the paper, relatively diversified in terms of financial support, provision
of services and is also continuous over the period of childhood (Adema and Thévenon, 2008).
Such a comprehensive support is likely to adjust to family needs and to secure the transitions
to motherhood and employment. It is also one reason why women do not forgo having a
second or a third child, and also explain why the decision to have children or to be in
employment is less polarised according to socio-economic status than in other countries
(Thévenon, 2004).
This is, however, how much policies contribute to establish such a secure climate is
difficult to capture and to quantify. In order to really understand the processes involved, one
may need to investigate further if and how labour market flexibility (i.e. flexitime or parttime, but also the increasing mobility between jobs observed during the lifetime career) adds
or reduce uncertainty that leads couples to postpone births. There is, to our knowledge, only
one study with very preliminary results on the impact of working conditions on fertility
intentions. Cette et al. (2005) suggested that the organisation of working life influences the
intended number of children, more significantly for higher socio-economic status. Higher are
household income, weaker is, however, the impact of working conditions. The lack of time
and of predictable hours would influence the timing of childbearing.
One may also need to understand further the role of institutions as “filter” that can
make the impact of an uncertain environment over the life-course very different between
different groups of populations (Blossfeld et al., 2004). This is indeed an important issue
since we can observe that highly educated individuals realise their fertility intention more
frequently, whatever their intention is to have or not a child (Toulemon and Testa, 2005).
Then a challenge is to understand how institutions shape the link between some macrosocioeconomic processes and the micro-individual life-course. Included in this perspective,
one issue is to capture in which extent does the effective impact of family policy institutions
on fertility depend on their complementarity over the family life-course. Another issue is to
characterise how such complementarity or, by contrast, some lack of coherence is likely to
create path-dependencies. Clearly, the fact that fertility behaviour has not change in France as
much as in the neighbouring countries is an indication of the existence of such dependency.
However, even such a sophisticated framework would be misleading to understand the
situation of fertility in France without taking into account the relative positive attitude towards
family in France, as stated by Testa and Grilli (2006) in their European comparison of
fertility preferences. As underlined by Toulemon et al. (2008), 95% of French people want at
least one child, which is very similar to the high proportion in other European countries
except in Continental Europe. There are, however, regional disparities since the ‘preference’
for children is more widespread in Southern regions, which is interpreted by Toulemon and al.
(2008) as an indication that French fertility ideals are linked to its cultural past (Southern
regions being more representative of that past). By contrast, the ideal number of children
among people who affirmed their willingness to have a child is extremely high in France, with
the highest score at 2.34 just behind Ireland, and less heterogeneous between regions. As
argued by the authors, this positive attitude towards large families may be partly nurtured by
the multi-facetted French family policy.
References
Adema W., Ladaique M (2005), “Net Social Expenditure: More comprehensive Measures of
social support”, OECD Social, Employment and Migrations Working Papers, 29, 79p.
Adema W., Thévenon O. (2008), “Les politiques de conciliation du travail et de la vie
familiale en France au regard des pays de l’OCDE. A partir de la synthèse des études « Bébés
et employeurs »”, Recherches et Prévisions, à paraître.
29
Aliaga C. (2005), « Conciliation entre vie professionnelle et vie familiale : des écarts entre les
femmes et les hommes », Eurostat, Statistiques en bref, 4/2005.
Ananian S., Bauer D. (2007), « Le temps périscolaire », Etudes et Résultats, 611, DREES.
Andersson G. (2001), “The impact of labour-force participation on childbearing behaviour:
pro-cyclical fertility in Sweden during the 1980s and the 1990s”, European Journal of
Population, 16, pp. 293-313.
Barber D., Immervoll H., (2005), « Can Parents afford to work ? Childcare costs, tax-benefit
policies and work incentives », Social, Employment and Migration Working Papers, OECD,
Paris.
Bechtel J., Caussat L., Courtioux P., Laib N., Le Minez S. et Mirouse B. (2005), « La
politique familiale française : coûts et bénéficiaires », Complément J, in Godet M., Sullerot
E., La Famille, une affaire publique, CAE, rapport n° 57, La documentation française, p. 409437.
Blanchet D., Ekert-Jaffé O. (1994), « The Demographic Impact of Family Benefits : Evidence
from a Micro-Model and from Macro-Data », in Ermisch J., Ogawa N. (eds), The Family, the
Market and the State in Ageing Societies, Clarendon Press, Oxford, pp. 79-104.
Bessone A.J., Brilhault G., Guérin S. (2003), “Analyse conjoncturelle de la fécondité“,
Mémoire de démographie, ENSAE.
Blossfeld H.P, Klijzing E., Mills M., Kurz K. (2004), Globalization, Uncertainty and Youth in
Society, Routledge, London.
Bongarts J. (2001), “Fertility and reproductive preferences in post-transitional societies”,
Population and Development Review, 27 (Supplement: Global fertility transition), pp. 260281.
Bongarts J., Feeney G. (2006), “The quantum and tempo of life-cycle events”, Vienna
Yearbook of Population Research, pp. 115-151, Vienna.
Breton D., Prioux F. (2005), « Two or Three Children ? Influence of family policy and
socioeconomic factors”, Population (English edition), 60(4), pp. 415-445.
Castles F. (2003), « The world turned upside down : below replacement fertility, changing
preferences and family-friendly public policy in 21 OECD countries » Journal of European
Social Policy, 13, pp. 209-227.
Caussat L. (2006), « Fertility trends and family policy in France : do they match ? »,
International Forum on Low Fertility and Ageing Society, Seoul, Korea, 13th-14th September.
Cette G., Dromel N., Méda D. (2005), « Conciliation entre vies professionnelle et familiale et
renoncements à l’enfant », Revue de l’OFCE, 92, pp. 263-312.
Chaupain-Guillot S., Guillot O., Jankeliowitch-Laval E. (2007), “Choix d’activité des mères
et garde des jeunes enfants: une comparaison européenne”, Recherches et Prévisions, 90, pp.
41-54.
Commaille J., Strobel P., Villac M. (2002), La Politique de la Famille, Repères, La
Découverte.
Curraize Y. (2005), « L’extension de la scolarisation en maternelle : une expérience naturelle
pour comprendre l’offre de travail des mères de jeunes enfants », mimeo.
D’Addio-Dervaux A., Mira d’Ercole M. (2005), “Fertility trends and the impact of policies”,
Social, Employment and Migration Working Papers, OECD, Paris.
30
Di Prete T., Engelhardt H., Morgan P., Pacalova H. (2002), „Do cross-national differences in
the costs of children influence fertility behavior?“, Discussion Paper 355, DIW, Berlin.
Courtioux P., Laïb N., Le Minez S. et Mirouse B. (2005), « L’incidence du système de
prélèvements et de transferts sociaux sur le niveau de vie des familles en 2004 : une approche
par microsimulation », Études et Résultats, n° 408, juin, 12 p.
Daguet F. (2002), « La fécondité en France au cours du XXème siècle », INSEE Première,
873, décembre.
Damon J. (2006), Les Politiques Familiales, Que-sais-je ?, 3776, Presses Universitaires de
France.
Del Boca D., Pasqua S., Pronzato Ch. (2007), “Motherhood and Work: Labor Market and
Social Policies”, Mimeo.
Ekert O. (1986), « Effets et limites des aides financière aux familles: une expérience et un
modèle », Population, 41/2, 327-348.
Ekert-Jaffé O., Joshi H., Lynch K., Mougin R., Rendall M., Shapiro D. (2002), “Fertility,
timing of births and socio-economic status in France and Britain: social policies and
occupational polarisation”, Population (English edition), 57(3), pp. 475-507.
Fagnani J. (2003), “Supporting Working Parents in France: Is Family Policy at a Turning
Point?”, Journal of Population and Social Security (Population), Supplement to volume 1, pp.
307-332.
Fagnani J. (1998), Recent changes in family policy in France : political trade-offs and
economic constraints, in E. Drew, R. Emerek (eds), Women, Work and the Family in Europe,
Londres, New-York, Routledge, pp. 58-65.
Gauthier A. (2007), “The impact of family policies on fertility in industrialised countries: a
review of the literature”, Population Research Policy Review, 26(3), pp. 323-346.
Gauthier A., Hatzius J. (1997), „Family benefits and fertility: an econometric analysis“
Population Studies, 51, pp. 295-306.
Hank K, Kreyenfeld M. (2002), “A multilevel analysis of child care and the transition to
motherhood in West Germany”, Discussion Paper 290, DIW, Berlin.
Hantrais L., Letablier M.Th. (1996), Families and Family Policies in Europe, Longman,
Londres/New York, 200p.
Héran F. (2002), “Situation démographique de la France dans le contexte européen”, in
Aglietta M., Blanchet D., Héran F. (eds.), Démographie et Economie, Conseil d’Analyse
Economique.
Kohler H.P., Billari F., Ortega J. (2006), “Low Fertility in Europe: Causes, Implications and
Policy Options." In F. R. Harris (Ed.), The Baby Bust: Who will do the Work? Who Will Pay
the Taxes? Lanham, MD: Rowman & Littlefield Publishers, 48-109.
Landais C. (2003), « Le quotient familial a-t-il stimulé la natalité française », Economie
Publique, 13, pp. 3-31.
Laroque G., Salanié B. (2004), « Fertility and Financial Incentives in France », CESifo
Economic Studies, 50(3), pp. 423-450.
Laroque G., Salanié B. (2005), « Does fertility respond to financial incentives ? », Document
de travail, CREST.
31
Letablier M.Th., Pennec S., Büttner O. (2003), “Opinions, attitudes et aspirations des familles
vis-à-vis de la politique familiale en France”, Rapport de recherché du Centre d’Etudes de
l’Emploi, 9, janvier.
Letablier M.Th. (2003), “Fertility and Family Policies in France”, Journal of Population and
Social Security (Population), Supplement to volume 1, pp. 245-261.
Lutz W. (2007), “Adaptation versus mitigation policies on demographic change in Europe”,
Vienna Yearbook of Population Research, pp. 19-27, Vienna.
Lutz W., Skirbeckk V., Testa M.R. (2006) “The low-fertility trap hypothesis: forces that may
lead to further postponement and fewer births in Europe”, Vienna Yearbook of Population
Research, pp. 167-192, Vienna.
Martin C., Math A., Renaudat E. (1998), « Caring for the very young children and dependent
elderly people in France : towards a commodification of social care ? » in Lewis J. (ed.),
Gender, Social care and Welfare State Restructuring in Europe, Ashgate, Aldershot, pp. 139174.
Math A., Meilland Ch. (2007), “Comparaison du soutien financier apporté aux familles dans
douze pays européens”, Recherches et Prévisions, 90, pp. 55-68.
Méron M., Widmer I. (2002), « Les femmes au chômage retardent l’arrivée du premier
enfant », Population, 57(2), pp.327-358.
Meulders D., O’Dorchai S. (2002), « Welfare State comparisons and motherhood », MOCHO, Report
for the European Commission.
Morgan Ph., Taylor M. (2006), “Low fertility at turn of the twenty-first century”, Annual
Review of Sociology, 32, pp. 375-399.
Moschion J. (2007), « Reconciling Work and Family Life : The Effect of French Family
Policies », Document de Travail du CES, 2007.73, Université Paris 1.
OECD (2005), Babies and Bosses : reconcyling work and family-life. Canada, Finland,
Sweden and United-Kingdom, Paris.
Pla A. (2007), « Bilan démographique 2007. Des naissances toujours très nombreuses »,
INSEE Première, n° 1170, janvier.
Prioux F. (1994), “Du premier au deuxième enfant: maîtrise de la contraception, normes
sociales et choix de l’intervalle », in AIDELF (ed.) Les modes de régulation de la
reproduction humaine. Incidences sur la fécondité et la santé, PUF, pp. 85-96.
Prioux F. (2007), « L’évolution démographique récente en France : la fécondité à son plus
haut niveau depuis plus de trente ans », Population, 62 (3), pp. 489-532.
Piketty Th. (2005), « L'impact de l'allocation parentale d'éducation sur l'activité féminine et la
fécondité en France, 1982-2002 », dans Lefèvre C. (Ed.): Histoires de familles, histoires
familiales, Les Cahiers de l'INED, n°156, p. 79-109.
Rallu J.L., Toulemon L. (1994), « Period fertility measures. The construction of different
indices and their application to France”, Population (English selection), 6, pp. 59-130.
Régnier-Loilier A., Léridon H. (2007), « Four decades of legalized contraception in France:
an unfinished revolution?”, Population and Societies, 439., 8p.
Revillard A. (2006), « Work/Family Policy in France : From State Familialism to State
Feminism ?”, International Journal of Law, Policy and the Family, 20(2), pp. 133-150.
32
Robert-Bobée I., Mazuy M. (2005), « Calendriers de constitution des familles et âge de fin
d’études », in Lefèvre C., Filhon A. (eds.), Histoires de familles, histoires familiales. Les
résultats de l’Enquête Famille de 1999, Cahiers de l’INED, pp. 175-200.
Ronsen M. (2004), “Fertility and family policy in Norway – A reflection on trends and
possible connections”, Demographic Research, 10(10), Max Planck Institute for
Demographic Research.
Sobotka T., Winkler-Dworak M., Testa M.R., Lutz W., Philipov D., Engelhart H., Gisser R.
(2005), “Monthly estimates of the quantum of fertility: towards a fertility monitoring system
in Austria, Vienna Yearbook of Population Research, pp. 109-141.
Strobel P. (2004), « L’État et les transformations de la famille », Cahiers Français Famille(s)
et politiques familiales, pp. 57-64.
Testa M.R. (2006), “Childbearing preference and family issues in Europe”, Report for the
Eurobarometer.
Testa M.R., Grilli L. (2006), « The influence of childbearing regional contexts on ideal family
size in Europe”, Population (English edition), 61(1-2), pp. 99-127.
Thélot C., Villac M . (1998), Politique familiale. Bilan et Perspectives, La Documentation
Française.
Thévenon O. (1999), "La durée du travail féminin en Europe : entre flexibilité et conformité.
Une comparaison des relations emploi/famille en Allemagne de l’Ouest, Espagne, France, aux
Pays-Bas et Royaume-Uni ", Recherches et Prévisions, 56, pp 47-66.
Thévenon O. (2004), « Les enjeux pour l’emploi féminin de la stratégie européenne pour
l’emploi », Revue de l’OFCE, n° 90, p. 379-418.
Thévenon O. (2006), « Régimes d’Etat Social et convention familiale : une analyse des
régulations emploi-famille », Economies et Sociétés, série Socio-Economie du Travail, 27, 6,
pp.1137-1171.
Thévenon O. (2007), « Labour force participation of women with children: disparities and
developments in Europe since the 1990s », IZA/Bertelsman Workshop “Balancing work and
of
November,
family
life”,
23-24th
http://www.iza.org/index_html?mainframe=http%3A//www.iza.org/conference_files/BaFaLi_
2007/viewProgram%3Fconf_id%3D1424&topSelect=events&subSelect=conferences
Toulemon L. (2003), La fécondité en France depuis 25 ans, Rapport pour le Haut Conseil de
la Population et de la Famille, 77p., http://www.ladocumentationfrancaise.fr/rapportspublics/034000116/index.shtml.
Toulemon L., Léridon H. (1999), « La famille idéale : combien d’enfants, à quel âge ? »
INSEE Première, 652.
Toulemon L., Mazuy M. (2001), “Les naissances sont retardées mais la fécondité est stable”,
Population, 56(4), pp. 611-644.
Toulemon L., Pailhé A., Rossier C. (2008), « France: high and stable fertility”, in Frejka T.,
Sobotka T., Hoem J., Toulemon L. (eds.) Childbearing trends and policies in Europe,
Demographic Research, vol 18.
Toulemon L., Testa M.R. (2005), « Fertility intentions and actual fertility: a complex
relationship”, Populations and Societies, 415, September, 4p.
33
Annex 1: The main cash family benefits in France
Source: Caussat (2006)
This appendix provides a detailed description of the main family benefits in France, after the
2004 reform. The given amounts are monthly in euros. The average monthly net wage for a
full-time worker in the private sector was in 2004 around 1850 euros.
First, there are 9 legal family benefits defined by the Social Security code, of which 4 are
linked to birth or to the first years of the life of a child – three of them being included within
the “prestation d’accueil du jeune enfant”, 2 are devoted to lone-parent families or orphans,
and the remaining 3 may be provided to any family. The following table provides complete
information on these benefits, ranked by decreasing total amount.
34
Apart from legal family benefits, families benefit also from:
- wage compensations (“indemnités journalières de maternité”) for mothers during the
maternity leave, e.g. 16 weeks for a first or second child, and 26 weeks for a child of parity 3
and over; the daily allowances are of the same amount than for the father leave; the total cost
of those compensations was 2 465 millions euros in 2005;
- grants paid by the Ministry for Education (“bourses d’éducation”) to families including
children going to elementary and secondary school, under means test, for a total amount of 2
393 millions euros in 2005;
- some employers, especially the Government and the main public utilities, pay
supplementary benefits (“supplement familial de traitement”) to their employees, for a total
amount of 2 361 millions euros in 2005.
35