July 25
1
Published in Journal of European Social Policy, online First
December 2017
Family Policy in High-income Countries:
Five Decades of Development
Mary Daly and Emanuele Ferragina
Abstract
This article empirically traces trends in family policy in 23 high -income countries since the 1960s. A range of data on income supports for families with children, family-related leave and early childhood education and care are brought together and analysed. The results show that family policy has developed by layering, in terms of both content and time period. A ‘foundational phase’ is characterised by investment in cash and tax allowances for families and employment leave for mothers, while a ‘consolidation phase’ sees states adding to their family policy portfolio, especially through the diversification of family - related leave and augmentation of child-related care services, increasing their overall family policy expenditure and continuing to support fami lies financially but with a preference to direct this through the tax system. There is no inexorable development path though, either within or across countries. A layering development pattern suggests that analysis of family policy over time needs to be oriented to examining both continuity and change and, as the conclusion makes clear, there are many fruitful lines of further research.
2
Introduction
Family policy is the subject of intensive reform and is among the fastest growing areas of
social expenditure within the OECD area. Scholarship, too, is proceeding apace but we still
know too little about the development pattern of family policy and the package of measures
involved at different points in time. Historical comparative analyses are rather scarce, and
family policy tends to be sidelined in the more global analyses of welfare state change. This
article takes up the challenge of exploring developments in family policy since the 1960s,
using a range of evidence and a systematic frame to assess cross-national continuity and
change in the composition of the policy portfolio. The piece seeks to contribute to the
comparative study of family policy by providing a complementary set of information to that
yielded by the more prevalent regime approach andenriching the evidential basis for family
policy analysis.
The overarching question driving the analysis asks: what are the main elements and reforms
in family policy over time in Europe and other highly-developed countries? We adopt a
global picture approach to answering this question, looking for long-range trends across
countries in regard to both the line of development, nature and cumulation of different
components of policy. Our research strategy has two layers. First, using a quantitative
descriptive approach we interrogate the evidence available for 23 OECD countries from the
1960s onwards, focusing on child-related income support to families, child-related leave from
employment and early childhood education and care (hereafter ECEC). Second, we uncover
the patterning underlying these trends and examine how they cohere into an overall set of
temporal developments and policy packages over time. The analysis rests on extensive work
to assemble evidence from different sources and time periods.
The article is organised as follows. The first part introduces the comparative family policy
literature. The methodological approach, indicators and data sources are next outlined. The
third part presents the policy evidence, looking in turn at trends in each. The fourth section
takes an overview of developments, considering the patterning across each policy sub-area
and the periodisation and packaging characterising the content and timing of family policy as
a whole. The conclusion draws out the implications of the empirical investigation and
suggests potential new research directions.
1. Family Policy in a Comparative Perspective
The scholarship on family policy is diverse and rich. The comparative literature can be
summarised as falling into two main streams: research examining the constituent elements of
family policy; work focused on theorising family policy and systematising different types of
approach and signature policy motivations.
The first stream has helped to identify the aims, instruments and political ideologies
associated with family policy (e.g., Fox-Harding, 1996; Saraceno, 2011). It reveals the classic
functions of family policy as relating to: childrearing and the maintenance and care of
children and adults; financial redistribution; combating poverty in families; and fertility
(Wennemo, 1994; Gauthier, 1996). When this type of work has applied a comparative lens,
characterisation of countries’ policy packages or profiles has been a preferred focus (e.g.,
Kamerman and Kahn, 1978; Wennemo, 1994; Gauthier, 1996; Hantrais, 2004; Ferrarini,
3
2006; Thévenon, 2011; Ferragina and Seeleib-Kaiser, 2015). An important insight emerging
from this scholarship is the merit of comparing family policy along a range of dimensions
rather than focusing on one dimension or element of the policy package only.
The second stream of work is more explicitly theoretical. Here too one finds diverse
approaches. One body of work focuses on identifying and explaining qualitative variations in
policy as revealed by historically-embedded comparison (Wennemo, 1994; Kaufmann et al.,
2002; Bahle, 2008). There is also a strong stream of feminist-inspired work. This has made a
significant contribution to the knowledge base, explicating: the state’s role in the organisation
of childcare and other forms of care and its management and distribution (e.g., Bettio and
Plantenga, 2004); the move towards the reconciliation of work and family life (Gornick et al.,
1997; Lewis, 2009); the power relations and inequalities associated with the treatment of
women and men in the welfare state (O’Connor et al., 1999). There is also a body of this
work that theorises family policy development in relation to fertility and child welfare
(Therborn, 1996; Montanari, 2000). An over-arching claim of this comparative strand of
work taken as a whole is that family policy comes in types and that it is possible to identify
signature approaches to family (and gender) policies (Lewis, 1992; Bettio and Plantenga,
2004; Bahle, 2008; Saraceno, 2011). When the types are pinned to countries – largely in an
attempt to engage with the regime approach of Esping-Andersen (1990) – four models are
identified for Western Europe: Nordic, Mediterranean, liberal and, at the centre of the
continent, two different subtypes exemplified by France and Belgium on the one hand and
Germany and Austria on the other.
This article draws from the insights of both sets of work. It also draws from the work that has
sought to operationalise and measure family policy more explicitly. This work tends to divide
into a number of strands also. One approach focuses on the family policy package and
concentrates especially on child-related cash and tax supports to families (e.g., Ferrarini et al
2012; Van Mechelen and Bradshaw, 2012). Another approach argues that services have to be
part of the analysis for they are as important as income supports (Förster and Verbist, 2013). A
further trend in existing work tends to concentrate increasingly on specific sub-areas of family
policy (with employment-related leave especially favoured,e.g., Gornick et al., 1997). We opt
for an integrated approach, which means that we analyse childcare services and leaves
alongside income supports to families for children. In sum, for the purposes of this article,
public support to families with children is conceived as a package that includes: money,
services and time.
We seek to take scholarship forward in at least three ways. First, we look at developments
over time, especially in regard to national family policy effort, thereby adding chronological
depth and empirical detail to existing knowledge. Second, as mentioned above, the article
provides an overview picture of the family policy package since it covers different
dimensions. Third, by systematically tracking policy development and examining the layering
or accretion (Streeck and Thelen, 2005) of policy measures in different periods, we hope to
provide a well-defined, macro-chronological picture that can act as a departure point for the
comparative study of national trends in family policy.
2. Methodological Approach and Data
The underdevelopment of global overarching work on family policy is undoubtedly
associated with limited conceptual and methodological cross-fertilisation between the fields
of comparative welfare state studies and family policy. The vast comparative literature on
4
other social policy domains suggests that the evolution of the components of family policy
over time can be assessed from at least two methodological standpoints: first, through an in-
depth historical analysis of a small number of cases characterised by different or similar
characteristics and outcomes (e.g., Barrington Moore, 1966; Skocpol, 1979); second, by
considering a large number of countries, using a mixture of quantitative data and secondary
analysis of historical and socio-economic elements. A large-N research design can follow a
regime (e.g., Esping-Andersen, 1990) or a trend (e.g., Wilensky, 1975) approach.
There is no agreement in the field on the best approach and the comparative welfare state
literature is characterised by a long-standing debate between scholars who propose
taxonomies and varieties and those who are primarily interested in global, long-term trends
and patterning over time. Over the 1990s and early 2000s, following Esping-Andersen’s
(1990) Three Worlds of Welfare Capitalism as well as Hall and Soskice’s (2001) Varieties of
Capitalism, the identification of models was prominent. Today scholars are refocusing on
large trends, trying to pinpoint elements of continuity and change rather than solely relying
on regime types. This development in the literature reflects the extensive processes of
liberalisation, which are driving welfare state reform in similar directions, thus blurring some
of the distinctions between different welfare regime models (Streeck and Thelen, 2005; Hay
and Wincott, 2012).
We adopt the approach of describing and pinpointing overarching trends in a large-N setting
rather than pursuing an in-depth historical analysis of a few cases or pinning countries to
regimes or types. The approach chosen fits our purpose as our intention is to see how specific
country trends define and fit into overall cross-national patterning. ‘Trend’ is conceptualised
to encompass both temporal developments and the component elements or package of
(family) policy at different periods of time.
The analysis presented rests upon an extensive evidence base of the anatomy and details of
family policy spending and characteristics across time and place. The assembly of this
evidence base was a challenging task, helped of course by the existing evidence but also beset
by extant gaps and lack of coverage for some countries in existing sources. OECD data were
a major resource. Two sets of OECD data were used: Social Expenditure Database (2013),
the OECD Family Database (2015). We supplemented this data with data from two other
sources: the Comparative Family Policy Database (Gauthier, 2011) 1 and the Child Benefit
Dataset. 2
The indicators for each of the three policy domains as well as the data sources are set out in
Table 1.
--------------------------------------------------Table 1-----------------------------------------------------
For child income supports, we rely on the Child Benefit Dataset, which compares the value of
the different types of support – cash benefits and those paid through the tax system -
1 This was compiled under the direction of Anne H. Gauthier and is housed at the Max Planck Institute for
Demographic Research. The evidence covers 22 OECD countries between 1960 and 2010. See:
http://www.demogr.mpg.de/cgi-bin/databases/fampoldb/index.plx. 2 The Child Benefit Dataset, located at SOFI University of Stockholm, covers various forms of child income
support programme, for 18 countries between 1960 and 2010. See http://www.sofi.su.se/spin/data/description-
of-datasets/child-benefit-dataset-cbd.
5
measured in terms of the average production worker’s wage (APW) for a model single-
earner, two-parent and two-child family (children aged 2 and 7 years) from 1960 to 2010.
In regard to the second policy sub-type, three types of family-related leave are covered:
maternity, parental and paternity. This is the most debated element of the family policy
package in terms of what constitutes ‘quality’ and how outcomes are to be evaluated (Keck
and Saraceno, 2013). That said, it is generally agreed that the elements that crucially
distinguish leave quality and orientation reside in the duration, the payment or wage
replacement level and the degree of flexibility around taking the leave – especially the extent
to which it is transferable between parents. These are the focus of analysis here. For
maternity and paternity leave both the total number of weeks and the average replacement
levels are assessed. For paternity leave, data availability allowed access only to the number of
weeks of unpaid and paid leave. However, in order to reveal further detail, we also measure
the number of paid and unpaid weeks of leave (including parental leave, childcare leave and
paternity leave) reserved for exclusive use by the father.
Finally, the evolution of ECEC is measured using two indicators. The first is spending on
ECEC for children up to 5 years of age as a percentage of GDP and the second the percentage
of children below the age of 3 years enrolled in publicly-financed child-care and pre-school
institutions as a percentage of their age cohort. 3
In each case, the global trends are analysed by calculating the average value of each indicator
for the 23 countries at different points in time. The analysis covers the following 23 OECD
countries: Australia, Austria, Belgium, Canada, Denmark, Finland, France, Germany, Greece,
Iceland, Ireland, Italy, Japan, Luxembourg, Netherlands, New Zealand, Norway, Portugal,
Spain, Sweden, Switzerland, United Kingdom (UK), and the United States (USA).
Unfortunately, Eastern European countries and several other OECD member countries (Chile,
Israel, Korea, Mexico and Turkey) had to be omitted because of gaps in the information
available.
The data used has several limitations, reflecting both general data weaknesses in comparative
family policy (Lohmann and Zagel, 2016) and compromises regarding global indicators.
First, because the data are drawn from different sources, the periods compared are not always
exactly the same. To take just one example, it proved impossible to get reliable and
comparable cross-national information on enrolment in ECEC before 1988. The latest dates
for which data are available also vary, which means that parts of the analyses end at different
points, although we have in all cases used the most up-to-date information available at the
time of writing (November 2016). Second, the series for ECEC spending provided by the
OECD changed after 1997. Hence, ECEC spending data are only comparable in two blocks,
between 1980 and 1997 and 1998 and 2013. The limitations of our system of standardisation
and generalisation are acknowledged; unfortunately there are no better alternatives available.
Data standardised using the APW model as well as spending data are not perfect for
evaluating the generosity of policy developments (as recognised in the literature, e.g., Clasen
and Siegel, 2007; Van Mechelen and Bradshaw, 2012). That said, they provide useful
information about trends, but need to be interpreted with caution. Moreover, as the evidence
is global our indicators do not take account of demographic or other background changes and
likely functional pressures.
3 In this case data are gathered from two distinct data sources: Gauthier’s (1999: 958) historical work provides
data for1988 and 1993, while the OECD Family Database presents data for the period 2003-2014.
6
3. The Development of Income Supports, Family-related Leave and ECEC
Child-related income supports to families
We know from existing work that family policy as a field of social policy originated in the
early twentieth century. 4 It is also clear that financial transfers to families were core to the
emergent policy field. The first measures consisted of assistance programmes for destitute or
needy mothers and children, which started to be introduced in Europe and elsewhere from the
first decade of the twentieth century (Gauthier, 1996: 41). Children’s health and welfare was
a driving concern, especially in the context of high child mortality rates and the interpretation
of these as linked to maternal health and mothers’ relative ignorance about the needs and care
of children. There was, though, little in place in terms of organised economic support for
families before 1930. But within a decade of the end of World War II, state-financed,
economic support for families in the form of cash allowances became a nearly generalised
feature of the most advanced market democracies across the world (Montanari, 2000: 307–8).
Before 1960 child income support was mainly provided through employment-based child
supplement to wages (usually paid to fathers), especially in continental European countries,
i.e., Austria, Belgium, France, Italy, Netherlands and Switzerland.
When we pick up the empirical story in 1960, all of the 18 countries for which the evidence is
available were financially supporting families with children (Table 2). Countries varied in
terms of their chosen instrument. Between 1960 and 1980, there was a marked increase in the
use of universal child benefit (accompanied by a smaller increase in the use of child tax
credit) and a steady decline in the prevalence of employment-based child supplement. The
Anglo-Saxon countries (Canada, Ireland, New Zealand and UK) especially favoured the
universal transfer approach as did the Nordic countries (with Denmark as an exception here).
In the other countries the early child income supports tended to be directed through the tax
system (Germany and USA for example) or were a continuation of employment-based
subsidies (Austria, Belgium, Italy and the Netherlands). It was the latter countries that offered
the most generous financial support to families at the first time point in our comparison.
Tracing the pattern over the next 50 years, there were two notable trends: the level of support
to families with children became more generous over time and the modality of support moved
in the direction of child tax credits. By 2010 the average replacement rate for all child-income
supports for the sampled countries stood at 12.1% of the APW (up from 7.2% in 1960). The
composition of the most and least generous country groupings changed quite substantially
though. Some of the continental European countries remained relatively generous (e.g.,
Austria, Belgium and Italy) but it was the Anglo-Saxon countries (especially Ireland, New
Zealand and the UK) and Germany that significantly increased their level of support to
families with children over the period. Countries where the value of the support was
significantly below average by 2010 included France, Japan and Norway (Table 2).
-----------------------------------------------------Table 2--------------------------------------------------
In a second trend, one sees especially a favouring of the tax system, in what Ferrarini et al.
(2012) refer to as a ‘fiscalisation of child support’. This reflects some movement away from
4 Although the origins date earlier if one includes education, health and measures against destitution and the
establishment and regulation of institutions to take care of children.
7
universalism in child cash allowances. Universalism was a strong early thrust as countries
sought to put in place generalised systems of financial support for families with children and
moved over time to supporting all children rather than the second and/or subsequent children
(which was a feature of early cash allowances in many countries). There was even then,
though, considerable movement within countries over time, representing varying pressures
and contestation around cash support for families with children and the incentives contained
in different measures, especially with the changed economic environment for families and the
rediscovery of poverty. More recently, tax credits have been favoured, especially in the
Anglo-Saxon countries but also in some continental European countries (e.g., Austria,
Belgium, Germany and Italy). This is associated with a move to reduce the value of the
universal child benefits, with only Denmark, Ireland and Japan showing a counter trend.
Family-related leave
Of the different types of leave, maternity leave is the oldest and in some countries was part of
the foundational architecture of the welfare state. First introduced in the last decades of the
nineteenth century, all developed countries, except for Canada and the USA, had adopted
some form of maternity leave by the eve of the Second World War (Gauthier, 1996: 50). Four
weeks of leave, usually paid but at a low rate, was the early norm. The objective was to ring-
fence a period of protected leave for mothers around the time of childbirth. Policy evolution
between 1960 and 1980 focused on extending the duration and making the payment more
generous, again keeping the focus on the health and protection of the mother (Table 3). Taken
as a whole, this 20-year period saw an average increase of three weeks in the duration of paid
leave – from 9.5 to 12.5 weeks. Canada, Finland and Portugal are notable for strong growth
in terms of leave duration. Sweden and Norway saw significant moves in the opposite
direction, prefiguring the shifting momentum that would occur later in other countries from
maternity to parental leave. This was also a period when the leave was made more generous
in terms of remuneration. For example, in parallel with the increased number of weeks of
paid maternity leave, average replacement rates also substantially increased between 1960
and 1980 – from an average of 42.9% to 54.3% of the APW. Denmark, Finland and Portugal
stand out as countries that significantly increased the generosity of their maternity leave in
this period (however the former two were far more generous than the latter).
---------------------------------------------------------Table 3----------------------------------------------
From the 1980s on, growth in maternity leave became much more muted and exceptional.
While most countries engaged in some policy reform, the 30 year period between 1980 and
2010 registered only an absolute cross-national average increase of two weeks in leave length
and the average replacement rate increased slightly – from 54.3% to 58.6% of the APW in
1980 and 2010 respectively (Table 3). Ireland and the UK along with Switzerland and
Portugal were notable for prioritising maternity leave for development during this period.
From the 1980s it was parental leave (and to a lesser extent paternity leave) that was in the
spotlight. As Table 3 shows, between 1980 and 2010 the average duration of paid parental
leave increased from 18 to 54.2 weeks and the average replacement rate rocketed from 11.3%
to 33.2% of the APW. These are very strong movements, perhaps reflecting strong pressures.
In this period, gender equality especially was interpreted in terms of granting social rights
around child raising and providing greater choice for women around employment and family
life, and to a lesser extent rights for men in their role of father (Daly 2010).
8
But there was considerable national and regional variation also. In regard to parental leave for
example, countries differed in whether they prioritised duration per se or the number of
weeks of paid leave. Australia, Germany, Greece, Ireland, Italy, the Netherlands, Portugal,
Spain and the UK tended to prioritise the unpaid route, whereas Finland, France, Japan,
Luxembourg and Sweden significantly increased both (albeit at different time periods) (Table
3). Notable among the remaining countries are Switzerland and USA which today still have
no universal parental leave – either paid or unpaid.
When it comes to paternity leave, as Table 4 shows, by 2015 the average duration of paid
leave across the 23 countries was only 0.9 weeks. Looking at the patterns over time the
forerunner countries are clear: Belgium, Luxembourg, Norway, Spain and Sweden. A
sizeable number of countries in 2015 had no (paid or unpaid) paternity leave: Australia,
Austria, Canada, Germany, Ireland, Switzerland and USA. Overall though, there is a move to
increase the duration of paternity leave, to make a part of it paid, and introduce either bonuses
or penalties for fathers (Moss, 2012; Keck and Saraceno, 2013).
--------------------------------------------------Table 4 -----------------------------------------------------
But paternity leave does not reflect fully social policy’s framing of the respective
entitlements and roles of fatherhood and motherhood. One also has to consider the
(re)distribution of leave between parents. This is a source of considerable innovation, with
countries especially reserving a period of leave for the father within the overall leave period.
In terms of number of paid weeks reserved for the father, the average level went from 0.2
weeks in 1990 to 9.4 weeks in 2015 (Table 4). The thrust of reform undertaken was in the
direction of incentivising fathers’ engagement with leave (through ‘use it or lose it’
conditions or the introduction of bonuses if fathers become more involved in taking parental
leave or parents share the leave more equally). The only exceptions are Canada and
Switzerland which have no leave reserved for the father and Ireland and the USA which fail
to guarantee a period of paid leave for the father.
ECEC
The data available on ECEC allowed us to measure both spending (for children aged under 5
years) and enrolment rates (for children aged under 3 years in formal childcare and pre-
school institutions). Data shortages mean that the time period examined in both cases is
shorter than for the other two policy domains, starting in 1980 for the expenditure data and
1988 for enrolment rates.
As Table 5 shows, both expenditure and enrolment have been on an upward trajectory. In the
case of spending, the growth between 1980 and 1997 was relatively modest but from the late
1990s the increase was strong and generally progressive with investment in the sector
growing strongly. In terms of national and regional patterns, the Nordic countries were
forerunners but not to an equal degree. Denmark and Sweden stood out as strong spenders on
ECEC even as far back as 1980. They remain in the vanguard on the latest evidence for 2013,
spending around 1.5% of their GDP on ECEC but they are exceeded by Iceland. Other
relatively heavy ECEC spenders in 2013 included Finland, France, New Zealand and
Norway.
----------------------------------------------------Table 5-------------------------------------------------
9
Enrolment rates for children under 3 years also show strong growth and tend to confirm a
similar picture regarding the periodisation of growth. For example, average enrolment rates
for this cohort increased from 12.2% to 24.3% between 1988 and 2003 and reached 40.9% in
2014 (Table 5). To pinpoint some developments at country level, Belgium, France, Germany,
Ireland, Luxembourg, the Netherlands, Norway, Portugal, Spain and the UK are notable for
the heavy increase in the number of enrolments in the youngest cohort registered over the
period as a whole. Some of these countries are later innovators though – especially France,
Germany, Luxembourg, Portugal and Spain where the growth tends to be concentrated in the
most recent period. As of 2014 the countries with the highest rates of enrolment were
Denmark (with 65.2% of the youngest cohort enrolled in ECEC, in excess of all other
countries and of the average of 40.9%), Iceland (59.7%), the Netherlands (55.9), Norway
(54.7%), France (51.9%), Portugal (47.9%), Sweden (46.9%) and New Zealand (41.9%).
4. Overall Patterning in Family Policy Development
How is one to make sense of these different trends and the chronological pattern over time?
Three global trends are striking. First, the long-term spending trend in family policy and the
three sub-areas considered is generally expansionary. Second, ECEC stands apart from the
other two sub-areas in terms of the strength of its growth, especially since the early 2000s.
Indeed, spending on ECEC is rapidly converging with that on child-related income support.
Third, of the three domains, income supports show the most variation in expenditure levels
over time, suggesting that, while they are among the oldest family policy measures, the
consensus on them and their relative place as a legitimate element of family policy is neither
fixed nor unwavering. As noted, there has been a marked shift to tax credits and a decrease in
the commitment to universalism as a means of supporting families with children. This pattern
appears to have become established in the early 2000s (Förster and Verbist, 2013: 15).
Looking at other indicators apart from spending and also taking into account existing
analyses of family policy (e.g., Wennemo, 1994; Gauthier, 1996; Lohmann and Zagel, 2016),
it seems possible to identify two distinct patterns of family policy development in terms of
components of the policy package and also general movements among countries in particular
directions at particular periods of time (Table 6).
---------------------------------------------------Table 6---------------------------------------------
A ‘foundational phase’ – the beginning of which clearly pre-dates most data presented here
given that it started around the post-war period and continued through to the 1970s – was
characterised by investment in a two-component family policy package: cash allowances for
children on a more or less universal basis and leave for maternity. The cross-national
momentum in this period was to increase the average value of child-related cash allowances
and extend the generosity of maternity leave. The former was motivated by a sense of
solidarity with families, and a recognition that families merited support with the costs of
child-raising, especially to avoid low income and poverty; the latter was inspired by a wish to
protect the health of the mother and child and has roots also in workers (and especially
women’s) rights. Fifteen of the 18 countries for which data is available saw an increase in
the average value of cash allowances between 1960 and 1980 and a trend is observed for 12
10
countries towards increasing the generosity of maternity leave (true for both duration and
average replacement rate) in the same period. There was hardly any increase in the average
replacement value of maternity allowance after 1980.
Once countries complete this foundational period, they tend to expand their family policy
package, both in terms of constituent elements and spending. We might call this ‘a
consolidation phase’ and locate it generally in the 1980s and after. It is marked by greater
diversity in the family policy package and also greater investment. In terms of constituent
elements, the core developments centre on some shift in financial support from the income
support system to the tax system, a growth of parental leave, an increased period of leave
reserved for the father, and more extensive provision of ECEC. The latter shows a very
strong pattern; the enrolment in formal childcare and pre-school institutions for children aged
between 0 and 2 years went up continuously in each country (Table 5). Taken as a whole,
this period is defined by policy that continues to be concerned with support of the family as
an institution of economic maintenance but also intra-familial equity and sharing and more
instrumental concerns like engagement with the labour market.
Thinking about what periodisation means, it is not that the old measures are dropped or
eliminated but rather that newer elements are added and/or prioritised. This means that there
are strong elements of continuity in family policy over time, and that the original measures
are still firmly in place (although of course there have been changes in substance and
intention which cannot be picked up by global spending and other data). We suggest that
family policy can be characterised as developing through a process of layering. Drawing on
work that has been mainly applied to other aspects of the welfare state, Streeck and Thelen
(2005) have emphasised that change does not only occur in abrupt or exogenous ways. Rather
than large changes in response to big shocks, they conceptualise how incremental policy
change produces transformative results. This seems also to be the case for family policy.
5. Conclusion
The evolution of family policy and the patterning for countries individually and globally are
far more complex than any simple periodisation or regime approach can countenance.
Therefore, the exercise undertaken here tried to strike a balance between tracking family
policy development since the 1960s – using the best available evidence – and drawing
attention to the combination and layering of policy components as well as the timing of
reform.
On the basis of the picture drawn by the 23 countries considered, family policy seems to
develop through a process of layering. Cash support to families with a universal orientation
and maternity leave constitute the backbone of family policy in what we call the foundational
phase between 1960 and 1980; and in a consolidation phase newer instruments, such as
parental and paternity leave and more widely available ECEC and greater recourse to the tax
system to support families, are added to reshape existing protection, change the incentives or
disincentives and respond to new societal needs and orientations. The original elements of the
policy package are not dismantled or replaced, but rather progressively accompanied by new
policy instruments which sometimes overshadow them. Although a chronological pattern can
be observed, we are not suggesting that this is fixed either in terms of timing or country
11
adherence. There are no countries that perfectly follow the patterns identified (although the
Nordic countries seem to be at the forefront of change) and it may well be that latecomer
countries (and indeed countries new to family policy) accelerate their development towards
consolidating or putting in place a more diversified portfolio of family policy. Hence, the
interpretation of family policy change should not be taken to imply an inexorable temporal
development path. Continuity and change need to be analysed together in frameworks that are
open to both.
How can the research initiated by this analysis be taken forward?
One way is to investigate whether the patterns observed for the 23 countries studied obtain
for other countries and regions (for example, Eastern Europe, Asia, Latin America and
Africa). Another fruitful line of research, using the macro evidence provided by our analyses,
is to undertake detailed country studies to see how they compare to the general or global
pattern set out here. There is also a need for research – especially country specific studies – to
probe the significance of changes in the distributive profile of family policy, the move away
from horizontal redistribution and towards vertical redistribution (as signalled by the
deceasing value of the universal cash supports to families for example and increased use of
the tax system as a conduit of such support). There is another line of further analysis also:
predicting and testing for future change. Here, we can hazard some suggestions. ECEC
service provision is growing more rapidly in countries which already had the highest level of
spending on it – that is, the Nordic countries. Most other countries are following this general
trend but at a much slower pace. Given this, one may hypothesise that other countries will
continue to follow the Nordic pattern and in a few years ECEC spending will overtake family
allowances as the main component of family policy. The recent evolution of family policy
spending and the evidence for the country level seem to suggest so. But it is important to
stress that this prediction assumes that OECD countries will continue to follow the trend
observed here.
A second way forward is to integrate family policy analysis into the studies of the welfare
state. Comparative scholars have tended to focus their research on other policy areas, such as
pensions and labour market policies (Streeck and Thelen, 2005), to all intents and purposes
excluding family policy (although see Gingrich and Häussermann, 2015). What are they
missing? They risk overlooking the role of family policy in a broader policy context. Among
the unasked questions are the extent to which and how closely family policy is used as a lever
for managing austerity, labour market desiderata, or housing need, or population health?
There seems to be evidence that family policy is becoming more instrumental, ever more
tightly tied to developments in the economy and the labour market and much less an
expression of society’s solidarity with families and children (Mätzke and Ostner, 2010).
Others see significance in the trend away from universalism in child supports and towards
both fiscalisation and a changed logic of redistribution towards the worst-off families
(Ferrarini et al., 2012). Without family policy, welfare state analysis misses also key aspects
of social politics; those embedded in questions about the extent to which changing policy
reflects different types of struggle for equality – on the basis of gender, age and/or sexual
orientation for example, but also among socio-economic classes and generations. Furthermore, analysing family policy brings in actors and concerns that are too little
considered in the mainstream welfare state literature, e.g., concerns around children, long-
term care, fertility.
12
A further fruitful line of analysis to consider is the significance of the differences in timing
between family policy and more general welfare state development. For not alone is family
policy characterised by a progressive evolution of spending growth but its period of
expansion has generally coincided with a phase of welfare state decline. That is, family
policy has experienced strong growth since the late 1970s/early 1980s – a time when it is
generally agreed that the expansionary phase of the welfare state was more or less over
(Pierson, 2001). Hence, if family policy is excluded then comparative scholarship risks an
inaccurate characterisation of the development of the welfare state. Finally, one very
interesting aspect of family policy – and a feature that might well set it apart from other
social policy domains – is its close relationship to prevailing beliefs and values. Given the
sensitivities and fine balances involved, family-related policy developments have to be
theorised as a compromise between general societal value systems and economic and other
functional pressures. The study of family policy, then, heightens the necessity to bring values
and contestations around ‘private life’ and the appropriate role and compass of the state into
the comparative welfare state field more broadly. We suggest that pursuing large research
questions such as these will enrich the study of both family policy and welfare state
development. In this endeavour, there is a place for both large-N studies but also a small-N
research approach.
But there are challenges that have to be acknowledged, not least weaknesses in data and
therefore analytic capacity. Some such weaknesses have limited the scope and depth of the
analysis undertaken here, leading to the omission of 12 OECD member countries. The three
main data sources used – the OECD Family Database, the Child Benefit Dataset and the
Comparative Family Policy Database – are major resources but they cover different periods
and have some different emphases. Longer and more detailed historical data series on all
domains of family policy are needed. In addition, better qualitative indicators across the
range of family policy measures would greatly help, e.g., on the flexibility of leave among
fathers and mothers, opening times and flexibility of service availability for ECEC. In sum,
attention needs to be given to the infrastructure for family policy analysis also.
13
Bibliography
Bahle, T. (2008) ‘Family policy patterns in the enlarged EU’, in J. Alber, T. Fahey and C.
Saraceno (eds) Handbook of quality of life in the enlarged European Union, London:
Routledge, pp. 100-25.
Barrington Moore, J. (1966) Social origins of dictatorship and democracy. Lord and peasant
in the making of the modern world, Boston: Beacon Press.
Bettio, F. and Plantenga, J. (2004) ‘Comparing care regimes in Europe’, Feminist Economics,
10, 1, 85-113.
CBD (2015) Child benefit dataset.
URL:http://www.sofi.su.se/spin/data/description-of-datasets/child-benefit-dataset-cbd.
Clasen, J. and Siegel, N. (eds) (2007) Investigating welfare state change, Cheltenham:
Edward Elgar, pp. 133–65.
Daly, M. (2010) ‘Shifts in family policy in the UK under New Labour’, Journal of European
Social Policy, 20, 5, 433-43.
Esping-Andersen, G. (1990) The three worlds of welfare capitalism, Cambridge: Polity
Ferragina, E. and Seeleib-Kaiser, M. (2015) ‘Determinants of a silent (r)evolution:
Understanding the expansion of family policy in rich OECD countries’, Social
Politics, 22, 1-37.
Ferrarini, T. (2006) Families, states and labour markets, Cheltenham: Edward Elgar.
Ferrarini, T., Nelson, K. and Höög. H. (2012) The fiscalization of child benefits in OECD
Countries, GINI Discussion Paper 49, Amsterdam: AIOS.
Förster, M.F. and Verbist, G. (2013) Money or Kindergarten? Distributive effects of cash
versus in-kind family transfers for young children, OECD Social, Employment and
Migration Working Papers 135, Paris: OECD.
Fox Harding, L. (1996) Family, state and social policy, Basingstoke: Macmillan,
Gauthier, A.H. (1996) The state and the family: A comparative analysis of family policies in
industrialized countries, Oxford: Clarendon Press.
Gauthier, A.H. (1999) ‘The sources and methods of comparative family policy research’,
Comparative Social Research, 18: 31-56.
Gauthier, A.H. (2011) Comparative family policy satabase.
URL: http://www.demogr.mpg.de/cgi-bin/databases/fampoldb/index.plx
Gingrich, J. and Häussermann, S. (2015) ‘The decline of the working-class vote, the
reconfiguration of the welfare support coalition and consequences for the welfare
state’, Journal of European Social Policy, 25, 1, 50-75.
Gornick, J.C., Meyers, M.K. and Ross, K.E. (1997) ‘Supporting the employment of mothers:
Policy variation across fourteen welfare states’, Journal of European Social Policy, 7,
1, 45-70.
Hall, P.A. and Soskice, D. (eds) (2001) Varieties of capitalism: The institutional foundations
of comparative advantage, Oxford: Oxford University Press.
Hantrais, L. (2004) Family policy matters, Bristol: Policy Press.
Hay, C. and Wincott, D. (2012) The political economy of European welfare capitalism,
Basingstoke: Palgrave.
Kamerman, S. and Kahn, A.J. (eds) (1978) Family policy: Government and families in
fourteen countries, New York: Columbia University Press.
Kaufmann, F-X., Kuijsten, A., Schulze, H-J. and Strohmeier, K.P. (eds) (2002) Family life
and family policies in Europe, vol 2 Problems and issues in comparative perspective,
Oxford: Oxford University Press.
14
Keck, W. and Saraceno, C. (2013) ‘The impact of different social-policy frameworks on
social inequalities among women in Europe: The labour-market participation of
mothers’, Social Politics, 20, 3, 297-328.
Lewis, J. (1992) ‘Gender and the development of welfare regimes’, Journal of European
Social Policy, 2, 3, 159-73.
Lewis, J. (2009) Work-family balance, gender and policy, Cheltenham: Edward Elgar.
Lohmann, H. and Zagel, H. (2016) ‘Family policy in comparative perspective: The concepts
and measurement of familization and defamilization’, Journal of European Social
Policy, 26, 1, 48-65.
Mätzke, M. and Ostner, I. (2010) ‘Introduction: Change and continuity in recent family
policies’, Journal of European Social Policy, 20, 5, 387-98.
Montanari, I. (2000) ‘From family wage to marriage subsidy and child benefits: Controversy
and consensus in the development of family policy’, Journal of European Social
Policy, 10, 4, 307–33.
Moss, P. (2012) ‘Parental leaves and early childhood education and care: From mapping the
terrain to exploring the environment’, Children and Youth Services Review, 34, 3,
523-31.
Pierson, P. (2001) The new politics of the welfare state, Oxford: Oxford University Press.
O’Connor, J.S., Orloff, A.S. and Shaver, S. (1999) States, markets, families: Gender,
liberalism and social policy in Australia, Canada, Great Britain and the United
States, Cambridge: Cambridge University Press.
OECD (2013) Family database.
URL: http://www.oecd.org/els/family/database.htm
OECD (2011) Social expenditure database (SOCX).
URL: http://www.oecd.org/social/expenditure.htm
Saraceno, C. (2011) Family policies, concepts, goals and instruments, Torino: Collegio Carlo
Alberto, Carlo Alberto Notebooks, No 230.
Skocpol, T. (1979) States and social revolutions. A comparative analysis of France, Russia
and China, Cambridge, Mass: Cambridge University Press.
Streeck, W. and Thelen, K. (eds) (2005) Beyond continuity: Institutional change in advanced
political economies, Oxford: Oxford University Press.
Therborn, G. (1996) ‘Child politics: Dimensions and perspectives’, Childhood, 3, 1, 29-44.
Thévenon, O. (2011) ‘Family policies in OECD countries: A comparative analysis’,
Population and Development Review, 37, 1, 57-87.
Van Mechelen, N. and Bradshaw, J. (2012) Child poverty as a government priority Child
benefit packages for working families, 1992-2009, GINI Discussion Paper 50,
Amsterdam: AIOS.
Wennemo, I. (1994) Sharing the costs of children Studies on the development of family
support in the OECD, Stockholm: Swedish Institute for Social Research.
Wilensky, H.L. (1975) The welfare state and equality, Berkeley: University of California
Press.
15