Abstract
This chapter examines the diversity of existing family-policy institutions in 15 member states of the European Union and develops a typology of family policies by means of cluster analysis. The countries included are Austria, Belgium, Denmark, Finland, France, Germany, Greece, Ireland, Italy, Luxembourg, the Netherlands, Portugal, Spain, Sweden, and the United Kingdom.
Keywords
These keywords were added by machine and not by the authors. This process is experimental and the keywords may be updated as the learning algorithm improves.
This chapter examines the diversity of existing family-policy institutions in 15 member states of the European Union and develops a typology of family policies by means of cluster analysis.Footnote 1 The countries included are Austria, Belgium, Denmark, Finland, France, Germany, Greece, Ireland, Italy, Luxembourg, the Netherlands, Portugal, Spain, Sweden, and the United Kingdom.
As pointed out before (see Chapter 2), most existing studies include other non-European countries, such as the United States and Australia, but exclude the Southern European countries, such as Greece, Portugal, and Spain (e.g., Gornick et al. 1997; Korpi 2000; Ferrarini 2006: an exception is the work done by Bahle (2008)). The current study contributes to existing research in this field and advances our understanding of family-policy variation in Europe in at least three ways. First, the focus on the 15 “old” member states of the European Union reduces complexity to some extent and keeps some influences constant, e.g., the influences from the supranational level, particularly the European Union. The results thus give an impression of the variability among countries belonging to the same economic and political union. Second, the inclusion of the Southern European countries allows for the investigation of whether these countries form their own, distinct family-policy cluster, as suggested by Flaquer (2000) and Bahle (2008a) as well as by Ferrera (1996) and Karamessini (2008), who discussed the distinctiveness of the “Southern Model” with regard to various fields of welfare-state intervention. Third, by considering up-to-date family-policy indicators from 2008, this study provides a recent picture of the fast-changing family-policy landscape in Europe.
Following the approach suggested by Korpi, the choice of indicators used to classify the 15 countries is restricted to the structure of family-policy institutions and does not include policy outcomes (e.g., female labor-market participation). This restriction allows for understanding institutions as “intervening variables,” which mediate between causal factors and policy outcomes (Korpi 2000: 141).
Along these lines of argumentation, the following chapters (4 to 6) apply the obtained typology as a conceptual framework for the interpretation of variations in public attitudes toward family-policy issues in the EU member states.
This chapter proceeds with some considerations of the typology concept and a brief description of the family-policy typology suggested by Korpi (2000) and Ferrarini (2006). The following sections discuss the indicators used in this study (Section 3.1) as well as data and methods (Section 3.2). Section 3.3 then presents the obtained family-policy typology resulting from the cluster analysis. The last section (3.4) concludes and discusses the potential relationship between the family-policy typology and public attitudes.
According to Ebbinghaus (2006: 52-53), two types of typologies have to be distinguished. With reference to Max Weber, the first kind of typology – the so called “ideal-type” – refers to theoretical models of reality, not reality as such. Typologies of this kind can serve as conceptual instruments for a comparison with and measurement of reality (see also Arts and Gelissen 2002). The family-policy typology developed by Korpi, which is used here as point of departure, is an example of such an ideal-typical typology. The second type of typology – the “real-type” – refers to classifications of empirically observed patterns. This kind of typology is used to reduce empirical complexity and to describe typical patterns found in reality (see also Obinger and Wagschal 1998; Shalev 2007). The typology developed in this chapter is a real-type typology that provides a classification of existing patterns of family-policy institutions in 15 EU member states.
Family policy is a rapidly changing field of welfare-state intervention and includes different kinds of government benefits that relieve or solve family-related problems. The focus in this study is on policies that are meant to contribute to the health and well-being of families with young children before their start in compulsory schooling. What this study does not capture are policy measures supporting the care for the elderly or other dependent family members. However, even this focus does involve a broad range of different policy measures, including direct cash transfers, taxation, in-kind benefits, direct services, and healthcare- and working-time policies (Gornick et al. 1997; Gornick and Heron 2006). Therefore, the selected indicators used here to classify and describe existing family-policy institutions in Europe inevitably represent only a limited variety of the diversity in this policy field. The selection of indicators is based on the ideal-typical family-policy typology provided by Korpi (2000) and Ferrarini (2006) (see also Chapter 2).
Based on the policy implications for gendered agency inequality in terms of labor-market participation, Korpi (2000) distinguished welfare states along the two ideal-typical policy dimension of general family support and dual-earner support. The first dimension, general family support describes a situation in which the state supports the traditional family model assuming a gendered division of paid and unpaid work both within families as well as in society more generally. The second dimension, dual-earner support, in contrast, implies that the state supports a dual-earner family model and encourages women’s labor-market participation and a more equal distribution of social care work within the family as well as in society. Based on these two dimensions, Ferrarini (2006) distinguishes four ideal-typical family-policy models. The first model is the “general family-policy model,” which provides a high level of general family support and a low level of dual-earner support. The second model is the “dual-earner family-policy model,” in which dual-earner support is high but general family support is low. The third ideal-typical model is the “market-oriented family-policy model,” which implies that neither general nor dual-earner support is very well developed. In this model, market forces are assumed to shape gender relations and individuals have to find private solutions in order to reconcile work and family responsibilities. The fourth theoretical model, finally, is the “contradictory family-policy model.” This model is characterized by a high level of both general family and dual-earner support.
The main advantages of this typology are at least two-fold. First, the two dimensions explicitly focus on policy implications for agency inequalities in terms of female labor-market participation as well as fathers’ involvement in social care work and thus underline the role public policies play in structuring both gender relations and the opportunity structures for individual actors. Second, the inclusion of different types of family support (i.e., general versus dual-earner support) allows for capturing contradictory or pluralistic policy orientations (Ferrarini 2006). Since countries are never completely clear cut regarding the type of family model they support, capturing this plurality is crucial, especially if the typology is meant to provide a framework for the interpretation or explanation of international variation in social outcomes or public attitudes that are potentially contradictory or pluralistic.
A disadvantage, in contrast, is the fact that existing institutions usually do not unequivocally conform to ideal types. An empirical analysis of actually existing institutions therefore inevitably results in a more heterogeneous picture. Moreover, one must keep in mind that typologies are useful means for reducing the complexity of reality, but this simultaneously implies a danger of neglecting important details of institutional variation. As Korpi (2000: 144) stated, “At best, typologies can give us a bird’s-eye view of the general contours of the landscape, thus facilitating orientation without giving guidance in details.”
3.1 Indicators for the construction of family-policy types
Based on the distinction between general family support and dual-earner support, this study selects up-to-date policy indicators from the year 2008 that capture leave entitlements (i.e., maternity, paternity, and parental leave), childcare services, child allowances, and tax incentives. These indicators are then used in a cluster analysis in order to detect country groups with similar policy profiles. The subsequent paragraphs describe the chosen indicators, data sources, and coding decisions in more detail and specify which indicators represent the general family support dimension and the dual-earner support dimension. Table 3.1 provides a brief summary of the indicators and the last part of this section presents the country-specific values for all indicators.
Leave entitlements. Three types of leave are included: Maternity-, paternity-, and parental leave. Maternity leave is a break from employment related to maternal and infant health and welfare. It is available only to women and usually limited to the period just before and after confinement. Paternity leave is only available to the father and refers to an entitlement to take a short period of leave immediately following the birth of a child. Parental leave and childcare leave are two forms of leave provided for childcare purposes. Parental leave starts after maternity leave, whereas childcare leave can usually be taken immediately after parental leave, thereby creating one continuous period of leave. Both forms are considered together here, although conditions, such as benefits paid, need not be the same. Earnings-related leave benefits have particular potential to increase female labor-market participation and attachment and to encourage a more equal distribution of paid and unpaid work if fathers are also entitled to such benefits (Ferrarini 2006; O’Brian et al. 2007; Kangas and Rostgaard 2007). Footnote 2 Moreover, generous earnings-related parental leave benefits reduce the poverty risk contingent upon childbirth and can ease the childbearing decisions for potential parents (Ferrarini 2006). In contrast, a major concern associated with low paid, flat-rate parental leave benefits is that “mothers with lower levels of education, who have worked in less skilled occupations, are most likely to take these low-paid leaves, which may further marginalize them from the labor market” (OECD 2001: 33).
Consequently, maternity- and paternity leave represent dual-earner support, since maternity leave supports female labor-market attachment and paternity leave encourages fathers’ participation in care work and thus a more equal sharing of paid and unpaid work between men and women. With regard to parental leave, income-related benefits indicate dual-earner support. Flat-rate benefits, as well as time rights without any financial compensation, in contrast, indicate general family support, since these types of parental leave do not stimulate women’s labor-market participation and attachment. Longer and better-remunerated periods of leave indicate a higher level of support. More generous financial-compensation rates provide larger economic resources to the household and therefore increase the attractiveness and actual use of the entitlement.
Prior to the cluster analysis, effective leave indicators for maternity-, paternity-, and parental leave were calculated, taking into account information on the length, type, and level of remuneration (see Table 3.2). Information on length and payment was mainly taken from the country notes published in the “International Review of Leave Policies and Related Research” by Moss and Korintus (2008). This approach takes into account that a comparison of the sheer length of leave entitlements would say little about the actual impact given the huge differences in payment levels. It can be assumed that a higher level of payment increases the rates of leave take-up (Plantenga et al. 2008; for a similar approach, see Plantenga and Siegel 2004). The effective leave indicators were calculated as the multiple of the length of leave (in weeks) and generosity of wage replacement (as a percentage of wages) divided by 100. In the case of paternity leave, information on length was measured in days instead of weeks. All flat-rate benefits were recalculated as percentages of wages using the national median wage as a reference. The obtained values can be interpreted as the number of weeks (or days in the case of paternity leave) for which leave is provided and remunerated with 100 % of former earnings.
Childcare services. High quality, affordable, and flexible childcare arrangements are among the most important preconditions that enable both parents to enter the labor market and combine paid work and family life (Plantenga et al. 2008). Childcare places for children from birth to three years of age are thereby a key factor for facilitating mothers’ return to the labor market after childbirth. However, facilities for this age group are especially scarce and demand clearly exceeds supply. This study therefore interprets a high availability of childcare services – especially for the youngest age group – as support for the dual-earner family model.
Studies have shown that women are more frequently employed and more likely to work full-time in countries with high service provision compared with women in countries with less or only financial support for families (Ferrarini 2006; Kangas and Rostgaard 2007). In addition, Castles (2003) showed that both public childcare facilities for children under the age of three as well as flexible work arrangements are positively correlated with fertility rates. Early childhood education and care services are additionally important from the perspective of equal opportunity and child well-being, as these institutions can help to reduce the disadvantages of children with special needs or from a lower socio-economic group (OECD 2006).Footnote 3
Despite the unquestioned importance of childcare services for facilitating both parents’ labor-market participation, reliable and internationally harmonized data on available services are scarce (Plantenga et al. 2008; Eurostat 2004). A common approach to overcome the lack of data is to use survey data that provide information on the actual use of childcare services as indicated by the parents, themselves. In this study, data from the EU-SILC provided by Eurostat (2010a) are used. The disadvantage if this indicator is that it includes different types of formal childcare arrangements that do not strictly distinguish between services provided by the (local) government and those provided by other organizations or private persons (e.g., registered childminders). Here, data on the actual hours per week children spend in formal childcare services function as a proxy for the availability of childcare facilities. Using the hours children are cared for institutionally has the advantage of indicating to what extent existing services are sufficient in facilitating mothers’ employment. Two age-groups of children are distinguished. The weekly hours used by children under age three represent the dimension of dual-earner support; those used by children aged three to compulsory school age represent the general family support dimension (for a similar approach, see Ferrarini 2006).
Cash benefits. This study uses child allowances to capture cash benefits. Child allowances are periodical flat-rate payments made to a member of a household with dependent children. These payments help with the costs of raising children and increase families’ net income. However, they are futile with respect to their effects on parents’ labor-force participation and likely to reproduce existing labor-market inequalities between men and women. Child allowances therefore represent the general family support dimension (Korpi 2000; Ferrarini 2006). This type of family support is captured by public expenditure data obtained from Eurostat (2010b). Child allowances were calculated in purchasing power parities (PPPs) per 1,000 children (between 0 and 4 years of age) in order to correct for differences in the age-specific composition of the population in the analyzed countries.
Tax incentives. The tax system is traditionally regarded as a powerful means for vertical redistribution of incomes. Additionally, tax policies may have differentiated effects on the labor supply of men and women and have often decreased female participation in the labor force (Montanari 2000; Smith et al. 2003). If wives’ earnings imply a heavy marginal tax on husbands’ income, the opportunity cost of wives’ employment becomes high for the family and penalizes wives’ employment (Esping-Andersen 1999). Today, most European countries have introduced individual taxation and provide some incentives toward a more equal sharing of paid work between men and women, France and Germany being exceptions (OECD 2010a). Nevertheless, there are persistent differences with regard to the level of taxation effects on households’ income, dependent upon the level of both spouses’ labor-market participation. These differences are used here as a measure for support for the dual-earner model through the tax system.
The measure uses the difference in net transfers to the government. It compares the average payment of a single-earner couple (in which the husband earns 200 % of the average wage and the wife earns nothing) with the average payment of an equal dual-earner couple (in which both partners earn 100 % of the average worker’s wage). The resulting numbers indicate the tax advantage or disadvantage (in percent) between the two prototypical household models (single-earner couples versus equal dual-earner couples) both of which have two dependent children, ages 6 and 11. Hence, this indicator is in fact a continuous measure ranging from general family support (i.e., negative tax incentives for an equal sharing of paid work between the spouses) to dual-earner support (i.e., positive tax incentives for an equal sharing of paid work between the spouses). Tax incentives were calculated and provided by the OECD (2010a).
The indicators used to capture the dimension of general family support are thus 1) (effective) parental leave entitlements (either unpaid or remunerated flat-rate); 2) childcare services for children between three years of age and compulsory school age; and 3) the level of child allowances. The dimension of dual-earner support is represented by 1) effective maternity-, paternity-, and parental leave (only income-related benefits); 2) childcare services for children under age three; and 3) tax incentives for an equal sharing of paid work between spouses.
Table 3.4 presents the country-specific values for the final family-policy indicators and gives a first impression of variation of family-policy design among the 15 countries included in the analysis. Effective paternity leave varies between as much as two weeks in Spain and Denmark and two days in Luxembourg, the Netherlands, and Greece, and is not at all available in several other countries. Effective maternity leave varies between 18 weeks in Denmark and Ireland and 8 weeks in Sweden and the UK. What is striking here is the fact that both Ireland and the UK provide a very long period of maternity leave (i.e., 42 weeks in Ireland and 52 weeks in the UK). These leaves were introduced instead of parental leave because they were easier for employers to accept (Lewis et al. 2008b). However, these leaves are scarcely compensated financially. The calculation of the effective leave indicators (which took into account both length of leave and payment) results in a low amount of effective maternity leave in both countries. Most other countries provide much shorter periods of maternity leave but remuneration equals mostly 100 % of former earnings (see Table 3.2).
With regard to parental leave, most countries provide either a flat-rate benefit (Austria, Belgium, France, and Luxembourg) or an income-related compensation (Sweden, Denmark, and Italy), while some countries provide solely uncompensated time rights (Greece, Italy, Spain, Portugal, the Netherlands, Ireland, and the UK). Additionally, a few countries provide a mix of different types of leave. Germany provides a 12-month, income-related leave, which can be prolonged without any further financial compensation, while Finland provides an income-related benefit that can be prolonged with a leave that is compensated flat-rate. Great variation is also apparent when looking at the weekly hours young children (under the age of three) spend in formal childcare. On average, children in Austria spend less than 1.5 hours per week in formal childcare, indicating low availability of care services, whereas children in Denmark spend an average of 25 hours per week in childcare services, indicating high availability. Differences in tax incentives for the dual-earner model are also striking. While Germany provides strong negative incentives of -20 %, positive incentives are especially strong in Sweden, with 37 %.
3.2 3.2 Methods
This section illuminates the method of cluster analysis (CA), which is applied in Section 3.3 to classify the 15 countries into different types of family policy. Cluster analysis (CA) generally aims to group cases (here, countries) by simultaneously taking a number of selected characteristics (here, family-policy indicators) into account. CA has already been applied in the comparison of welfare states as a whole (Obinger and Wagschal 1998; Kautto 2002; Powell and Barrientos 2004) as well as in specific policy fields, such as healthcare systems (Wendt 2009; Reibling 2010; Wendt et al. 2011), social-assistance schemes (Gough 2001; Wendt et al. 2011), and family policy (De Henau et al. 2006; Jensen 2008; Wendt et al. 2011). An important advantage of this method is that the number of resulting clusters is not predefined (e.g., by theoretical considerations) but rather remains an empirical issue. This approach thereby allows for capturing not only pluralistic policy orientations but also the whole variety of existing family-policy types.
This study applies the method of hierarchical cluster analysis (HCA) using Ward’s method with the Euclidean dissimilarity measure, which is a standard technique for hierarchical cluster analysis (for an extended introduction to cluster analysis, see Everitt et al. 2001). HCA starts when each country forms a cluster of its own and then gradually joins similar countries to form new clusters until all cases finally come together within one group. Accordingly, the researcher has to determine the number of clusters that best represent the structure of the data. This is done here via a graphical approach, i.e., the dendrogram, presented in the following section. The cases are usually grouped such that both homogeneity within clusters as well as heterogeneity between clusters is maximized. Ideally, countries within a certain cluster should be more similar to each other than to any country of another cluster across all their characteristics.
Since the indicators described in Section 3.1 were measured on different scales, they have to be standardized prior to performing a cluster analysis (Milligan and Cooper 1988). Standardization of the input variables is necessary since the dissimilarity measure used in the analysis (Euclidean distance) is sensitive to differences in the magnitudes of the variables. Standardization furthermore serves the idea of equal weighting of all input variables.
Several approaches to the standardization of variables exist. The present study applies standardization through division by the range of the variable. In a simulation study, Milligan and Cooper (1988) have shown that this approach offers the best recovery of the underlying cluster structure across a variety of conditions (e.g., error conditions, separation distances, clustering methods, and coverage levels) and outperforms the traditional z-score transformation. The following calculation was used to obtain standardized values:
Whenever possible, each variable’s natural minimum and maximum value was used. As there was no natural benchmark, values were set at a level that was below or above the actual minimum or maximum value within the sample of EU countries. The final values range between 0 (low support) and 1 (high support) (for a similar approach, see De Henau et al. 2006; Plantenga et al. 2009).
3.3 Family-policy types: Results from the cluster analysis
The dendrogram below (Figure 3.1) graphically depicts the result of the hierarchical cluster analysis (HCA). This two-dimensional diagram illustrates the fusions made at each stage of the analysis. The y-axis indicates the distance or dissimilarity between two countries or groups of countries; the longer the vertical lines, the more dissimilar countries are to each other or to a group of countries merged at an earlier stage. The dendrogram suggests that four clusters best represent the structure of the data. Table 3.4 summarizes the clusters’ characteristics, which are elaborated subsequently. It is important to note that it is never easy to distinctly classify all countries. Therefore, major differences between countries assigned to the same cluster are pointed out here as well.
Cluster 1, on the far left in Figure 3.1, consists of Austria and Luxembourg. This group is characterized by a high level of general family support and rather low dual-earner support and is labeled the general family support cluster. Both countries provide a high level of child allowances (which are especially high in Luxembourg) as well as extensive parental leave with monetary compensation paid as a flat-rate benefit (which is especially generous in Austria). Neither measure stimulates mothers’ labor-market attachment. The tax system also does little to stimulate an equal sharing of paid work between spouses (the incentive is about 12 %) and paternity leave is nonexistent in Austria and quite short in Luxembourg (two days). Moreover, childcare services are hardly available for the youngest age-group (1 hour in Austria, 7 hours in Luxembourg). The availability for older children lies at a medium position with 17 hours per week.
Cluster 2, also on the left of Figure 3.1, comprises the UK, Ireland, the Netherlands, and the Southern European countries of Greece, Italy, and Portugal. Cluster 2 contains the largest group of countries, all of which provide relatively little support for families in both dimensions, though Italy and Portugal do perform slightly better compared with the remaining countries in this group. This cluster is labeled the low support cluster. In this group, child allowances are at the lower end of the scale (the only exception is Ireland) and only Italy provides earnings-related compensation during parental leave, albeit at a very low level. No other countries provide any financial compensation at all. They have a very short period of unpaid parental leave (about 13 days), and paternity leave is either short or not at all available (as in Italy and Ireland). Compensating for the low support for families, strong tax incentives for the dual-earner model exist in several countries (highest in Ireland with 32 % and lowest in Portugal with 12 %), and childcare coverage is moderate, at least for somewhat older children. There are some notable differences among the countries regarding childcare services. For the youngest children, the availability centers around 4 hours per week in half of the countries and is twice as high in Italy and the Netherlands (8 hours) and tripled in Portugal (12 hours). The availability of services for children from three to compulsory school age centers around 18 hours in most countries and is again clearly higher in Italy (30 hours) and Portugal (26 hours). Overall, public support for families is low in this cluster, and balancing work and family life is tough and remains mainly a private responsibility. There is, however, some support for working parents in terms of taxation and childcare services (the latter especially in Italy and Portugal).
Cluster 3 includes Belgium, Finland, and France. This group, labeled the pluralistic support cluster, is quite heterogeneous and characterized by both general family support as well as dual-earner support. All three countries provide generous paternity leave (ranging from 9 days in Belgium to 13 days in Finland), and France and Finland also provide generous parental leave, though this leave is compensated flat-rate in France. In Belgium, parental leave benefits are also remunerated flat-rate, albeit humbly compared with the other two countries. A Finnish particularity is the combination of a parental leave entitlement, which is an income-related benefit with the provision of a flat-rate home-care allowance after the end of parental leave. After the end of parental leave, Finnish parents can choose between municipal daycare and this home-care allowance, which allows parents to stay at home to care for their child and receive a flat-rate benefit until the child is three years old (Kröger et al. 2003). This is in line with earlier studies indicating that all three countries advocate parental choice, supporting working mothers as well as emphasizing parents’ right to provide care (Mahon 2002; Kremer 2007). This plurality in policy incentives is also mirrored in a medium level of child allowances (which is lowest in Finland) and the availability of childcare services. Availability of services for the youngest age group is at a medium level (ranging from 9 hours in Finland to 14 hours in Belgium) while at a high level for older children (ranging from 26 hours in Finland to 32 in Belgium). The level of tax incentives, in turn, considerably varies among the three countries. France provides a negative incentive of -3 %, which promotes the traditional family model; Belgium provides a moderate positive tax incentive of 9 %, whereas Finland provides a strong tax incentive for the dual-earner model of the family (with 27 %). Within this Cluster, Finland clearly offers the strongest support for the dual-earner model of the family (especially in terms of income-related parental leave and tax incentives) and France the strongest support for the traditional model of the family (especially through a generous flat-rate remunerated parental leave and negative tax incentives).
Cluster 4, on the right of the figure, is made up of the Nordic countries of Denmark and Sweden. These two countries are characterized by a high level of dual-earner support combined with a moderate level of general family support, which is mirrored in the comparably low level of child allowances provided and a focus on income-related parental leave benefits. This cluster is labeled the dual-earner support cluster. Denmark and Sweden stand out with considerable incentives for fathers’ involvement in care in terms of paternity leave (14 days in Denmark, 8 days in Sweden) and high childcare coverage rates for both age groups. Availability of childcare services for the youngest children is especially high in Denmark, with 25 hours (compared with 14 hours in Sweden). Moreover, both countries offer generous earnings-related parental leave. Sweden additionally offers some flat-rate and unpaid periods of leave to prolong the well-compensated parental leave period. Consequently, mothers’ labor-market involvement and reconciliation of work and family life, as well as fathers’ involvement in care work, are strongly supported publicly. High tax incentives (especially in Sweden) for the dual-earner model reveal this public support.
Two countries, namely Spain and Germany, are considered outliers that cannot be classified unambiguously. Figure 3.1 shows that Spain is fitted into the third cluster (with France, Belgium, and Finland), while Germany is grouped with the fourth cluster (with Denmark and Sweden), both at a very late stage of the clustering process. This classification is due to very specific characteristics of the two countries.
Spain’s closeness to the pluralistic support cluster is due to comparably high values in both policy dimensions. However, a closer look at the countries’ family-policy institutions reveals that existing support remains rudimentary in comparison with France, Belgium, and Finland. On the one hand, Spain provides a comparably high level of childcare services (i.e., 10 hours for the youngest children, 28 hours for older children), generous paternity-leave entitlements (15 days), and a long period of parental leave (150 weeks). On the other hand, Spain provides no financial compensation at all during the parental leave period. This fact reduces the likelihood of take-up and leaves families with considerable difficulties concerning income maintenance upon childbirth as well as during the following, most care-intensive period. Moreover, these entitlements do not stimulate female labor-market attachment. In line with these arguments, Spain is not subsumed under cluster 3 but rather assigned to cluster 2 consisting of the remaining Southern European countries as well as the UK, Ireland, and the Netherlands. Within this group, Spain remains a clear outlier with regard to paternity- and parental leave and resembles Portugal and Italy regarding the higher availability of childcare services compared with the other cluster members.
The main reason for Germany’s proximity to the Nordic countries in Figure 3.1 is the strong dual-earner incentive in terms of income-related parental leave and a moderate availability of childcare services. Germany’s parental leave regulations were reformed in 2007 in order to increase dual-earner support, and the formerly means-tested and flat-rate benefit was replaced by an income-related benefit (Wersig 2007; Henninger et al. 2008). Simultaneously, however, strong incentives supporting a traditional division of paid- and unpaid work between men and women persist, such as an additional period of unpaid parental leave (which can be used to prolong the paid period of parental leave) and strong negative tax incentives (-21 %). Germany’s tax system actually penalizes an equal sharing of paid work between spouses, promoting the traditional family model with a main wage earner. Moreover, paternity leave is nonexistent in Germany and child allowances are generous. The availability of childcare services lies at a medium position of 13 hours for the youngest children and 24 hours for older ones. However, Germany is a special case with regard to childcare services. Due to cultural and institutional differences originating from the division into East and West Germany (i.e., between the former German Democratic Republic and the Federal Republic of Germany), huge differences in the availability and use of childcare services persist between both parts of the country to the present day (Rosenfeld et al. 2004). The indicator used here includes information from both parts of the country, thereby underestimating usage in the East while overestimating usage in the West.
In contrast to Sweden and Denmark, Germany cannot be considered to be a country that provides clear incentives for the dual-earner model of the family. Instead, Germany’s policy profile mainly promotes the traditional model of the family, which suggests that is should instead be assigned to cluster 1, which consists of Austria and Luxembourg. Within this cluster, Germany remains an outlier due to the generous and income-related parental-leave benefit and the higher availability of childcare services (these policies are in line with the dual-earner support dimension) on the one hand and the exceptionally strong negative tax incentive and additional unpaid parental leave entitlements (these policies are in line with the general family support dimension) on the other hand.
The family-policy typology resulting from the cluster analysis suggests four family-policy clusters and two outliers (i.e., Germany and Spain). Taking into account the specific policy profiles of these two cases and the inherent incentives and constraints for families’ work- and care arrangements, the following cluster assignments are suggested: Germany’s policy profile indicates compliance to cluster 1, whereas Spain’s policy profile suggests proximity to cluster 2. The resulting clusters are comparably small (except cluster 2) but seem to represent a meaningful division. Eventually, the following 4 clusters are distinguished:
-
1.
The general family support cluster with Austria, Luxembourg, and Germany. This cluster provides a high level of general family support and rather low dual-earner support. It is characterized by a high level of child allowances, extensive parental leave (mostly compensated flat-rate), limited or non-existing paternity leave, and limited availability of childcare services for the youngest age group.
-
2.
The low support cluster with the UK, Ireland, the Netherlands, Greece, Italy, Portugal, and Spain. This cluster provides relatively little support for families in both dimensions. The level of child allowances is low, parental leave is short and/or scarcely compensated financially, and the provision of childcare services is moderate.
-
3.
The pluralistic support cluster with France, Belgium, and Finland. This group is quite heterogeneous and combines both general family support as well as dual-earner support. It provides generous parental and paternity leave (though parental leave is partly compensated flat-rate) and a moderate to high level of childcare services.
-
4.
The dual-earner support cluster with Denmark and Sweden. This group is characterized by a high level of dual-earner support combined with a moderate level of general family support. These countries provide generous
earnings-related parental- and paternity leave entitlements, a moderate level of child allowances, and well-established childcare facilities.
This cluster solution represents a rather flexible typology, and the classification of countries is subject to change relative to the inclusion of additional countries or indicators in the analysis or if the characteristics of family policies change as a result of policy reform. Table 3.4 summarizes the clusters’ characteristics regarding all policy indicators included. For the two clusters containing the outliers (clusters 1 and 2), the table presents separate values for both outliers as well as the mean values for the respective cluster both with and without Germany and Spain.
3.4 Conclusion
This chapter developed a family-policy typology in order to classify and describe empirically observed family-policy patterns in 15 countries belonging to the European Union. The typology-approach applied here is a useful means for reducing the complexity of reality; however, it necessarily implies the danger of neglecting important details of institutional variation (Korpi 2000: 144). Accordingly, the described family-policy typology provides a rough picture of family-policy diversity in Europe.
The following chapter uses the described typology of existing family-policy systems as a framework for the analysis and interpretation of differences in public attitudes toward family-policy issues. Several attitudinal patterns are expected to emerge. The dual-earner cluster represents a family-policy context that is highly supportive of the dual-earner model of the family. Such a context should nourish a societal consensus in favor of this family model. As a result, demand for government responsibility as well as the level of satisfaction with existing policies is expected to be high within this cluster. Skepticism toward the consequences of benefits and services, in contrast, is expected to be low.
A tradition of low state intervention in the sphere of the family, in contrast, could result in a public climate in which any type of government intervention is rejected. Skepticism toward the consequences of benefits and services is thereby expected to be high. Alternatively, it can be argued that a low level of support for families results in a climate of strong public demand for government responsibility, indicating disagreement with the status quo. Due to increasing female labor-market aspirations and participation rates, it is to be expected that support is especially strong for those policy measures that increase the compatibility of work and family life, such as parental leave and childcare services. Satisfaction with existing policies should then be low as should be the level of skepticism. The empirical analyses in the following chapters reveal, which of the described scenarios actually occur.
The general family support cluster and the pluralistic support cluster, finally, are likely to range in-between the dual-earner support and the low-support cluster with regard to all three attitudinal dimensions.
Compared with earlier research, the four family-policy clusters are only partly consistent with alternative classifications. It is critical to note, though, that the comparability of different classifications is limited because different studies include different countries and indicators and also focus on different points in time. Table 3.5 provides a brief summary of the country groupings suggested by Ferrarini (2006) and Bahle (2008a) as well as the current study. The main differences between these three classifications of family policies are discussed below.
A first discrepancy with earlier typologies is the classification of Finland. Both Bahle (2008a) and Ferrarini (2006) classified Finland in the same group as the other two Nordic countries of Denmark and Sweden, whereas this study assigns Finland to the pluralistic support cluster (together with France and Belgium). Consistent with this distinction, Bahle also classified France and Belgium in a separate group. In the ideal-typical family-policy typology developed by Korpi (2000) and Ferrarini (2006), this group would coincide with the so-called contradictory family-policy model. However, this model was not identified in their empirical analyses and no country was assigned to this type. Both France and Belgium were instead allocated to the general family support type. The current typology assigns Germany, Luxembourg, and Austria to the general family support cluster; these countries also form their own group in Bahle’s typology (together with other counties not included here).
An important question to consider was whether the Southern European countries form their own, distinct family-policy cluster, as suggested by Flaquer (2000) and Bahle (2008a). In contrast to their suggestion, this study reveals one low-support cluster, which includes the Southern European countries as well as the UK, Ireland, and the Netherlands. The common feature of this heterogeneous group of countries is a low level of government support in both dimensions and the absence of clear policy incentives for a certain family model. However, it is important to note that these countries differ with respect to important contextual features such as labor-market characteristics and female employment participation.
In comparison with earlier typologies, the low support cluster roughly corresponds with the market-oriented family policy model in Ferrarini’s typology. It is critical to note, however, that Ferrarini did not include the Southern European countries in his study (except for Italy). Based on data from the year 2004, Bahle suggests a distinction between the Southern European countries of Portugal, Italy, Greece, and Spain on the one hand and the UK, Ireland, and the Netherlands on the other hand. He concludes that the Southern European countries provide even less support for families compared with the remaining three countries. An important peculiarity of the Southern European countries is furthermore the traditionally strong role of the Church, the late democratization (except in Italy), and the strong reliance on family solidarity or “familism” (Bahle 2008a: 104; see also Karamessini 2008).
It has been pointed out earlier in this study, that these historical, cultural, and political dimensions (e.g., family systems, religious traditions, the role of the Church, and the relationship between the Church and the state) have played a key role in the development of distinct family-policy systems and that they are mirrored in the institutional design of existing family-policy arrangements (Bahle 1995; Gauthier 1996; Pfau-Effinger 1996, 2005b; Kremer 2007; Strohmeier 2002).
This study assumes that the family-policy setup, as one element of the welfare arrangement (Pfau-Effinger 2005a), impacts public opinion toward family policy. The following chapter therefore investigates the link between the family-policy setup and public opinion. However, all other elements of the welfare arrangement play a role as well and differences in the degree of problem pressure (e.g., low fertility and child poverty) or labor-market characteristics and opportunities (e.g., the availability of part-time employment) are important factors that can be expected to modify public attitudes. Chapter Five elaborates on these issues and develops an adjusted version of the family-policy typology. The consideration of the broader social context suggests that the Southern European countries indeed split off into their own cluster and that Finland should be grouped with the Nordic countries of Denmark and Sweden.
Notes
- 1.
The family-policy indicators used in this chapter were collected during the research project “Attitudes Toward Welfare State Institutions: New Perspectives for the Comparative Welfare State Analysis,” which was funded by the German Research Foundation (DFG) between 2006 and 2009. I already analyzed and published these indicators in the first part of the chapter “Family policy – one for all?” (Wendt et al. 2011).
- 2.
For a brief overview of divergent opinions about the effect of leave on women’s labor-market attachment, see Kangas and Rostgaard (2007).
- 3.
For concerns expressed over too early usage of professional childcare, please refer to Belsky (2001) and Brooks-Gunn et al. (2002).
Author information
Authors and Affiliations
Rights and permissions
Copyright information
© 2014 Springer Fachmedien Wiesbaden
About this chapter
Cite this chapter
Mischke, M. (2014). Family policies in Europe – a cluster analysis. In: Public Attitudes towards Family Policies in Europe. Springer VS, Wiesbaden. https://doi.org/10.1007/978-3-658-03577-8_3
Download citation
DOI: https://doi.org/10.1007/978-3-658-03577-8_3
Publisher Name: Springer VS, Wiesbaden
Print ISBN: 978-3-658-03576-1
Online ISBN: 978-3-658-03577-8
eBook Packages: Humanities, Social Sciences and LawSocial Sciences (R0)