Helga Hernes (Reference Hernes1987) started the discussion about how the welfare state influences gender relations with her book Welfare State and Woman Power, where she claimed that the Scandinavian countries were developing ‘women friendly welfare states’ that encourage greater gender equality. The debate on family policy and gender really picked up when Esping-Andersen (Reference Esping-Andersen1990) wrote his famous Three Worlds of Welfare Capitalism. Feminist theorists were quick to point out that the book was based on the notion that the goal should be ‘decommodification’ so that people are not forced to work. However, the women’s movement in many countries had struggled for the right and possibility of women to work, which means they wanted to become commodified. In addition, his scheme did not take into account the unpaid labour that women carry out in the household (e.g., Jenson, Reference Jenson1997; Hervey and Shaw, Reference Hervey and Shaw1998). This criticism sparked a long-going debate on alternative typologies of family policy from a gender perspective. After discussing the main approaches that have emerged, this article continues by briefly discussing attempts to broaden the discussion by bringing in children and by adding an intersectional perspective.
Developing gender approaches
In this section, I will discuss some of the most influential institutional and cultural approaches to analysing family policies.
Institutional approaches
Jane Lewis (Reference Lewis1992) was the first to publish a criticism of Esping-Andersen (Reference Esping-Andersen1990) from a gender perspective and developed a typology that took into account family policy and gender relations by dividing countries into three groups: those with strong, modified, or weak male breadwinner models. The problem with her typology is that the terms ‘modified’ and ‘weak’ male breadwinner models described that they countries were not, rather than what they were (e.g., Sainsbury, Reference Sainsbury1994). Lewis herself (Reference Lewis1997) later admitted that typologies should include the aspect of what policies ought to be, while the breadwinner typology only ranks degrees of negativity.
This led to an upsurge of approaches, many of which did well in capturing one type of policy, but not in creating a full typology. A popular example is the adult-worker model (Lewis and Guillari, Reference Lewis and Giullari2005) in which policies encourage adults of all genders to work. As Daly (Reference Daly2011) notes, the adult-worker model is underspecified, as it does not make clear what variations exist. Furthermore, it does not describe the division of unpaid labor in the home, which is a central theme for feminist analyses. Similarly, Sainsbury (Reference Sainsbury and Sainsbury1999) developed the individual earner-carer, male breadwinner, and separate gender roles models, in which the first named category caught on but not the others. An advantage of the individual earner-carer model is that it indicates that all parents regardless of gender are expected to both work and take care of the family. The problem with her typology is that it is difficult to differentiate between the male breadwinner and separate gender roles models. Others have written about the ‘dual-earner’ model without mentioning caring (e.g., Kangas and Rostgaard, Reference Kangas and Rostgaard2007) which is a step backward from Sainsbury’s idea of the individual earner-carer model.
Evenutally, the familialisation/defamilialisation typology emerged as the leading model (e.g., Lister, Reference Lister, Baldwin and Falkingham1994; McLaughlin and Glendinning, Reference McLaughlin, Glendinning, Hantrais and Glendinning1994; Hantrais, Reference Hantrais2004). Even Esping-Andersen, himself, accepted this typology for family policies as a complement to his commodification/decommodification typology (Esping-Andersen, Reference Esping-Andersen1999, Reference Esping-Andersen2009). It has the advantage over the other typologies in that it allows for a parallel to Esping-Andersen’s scale of degrees commodification and decommodification, thus providing feminists with a clear alternative to Esping-Andersen’s manner of measuring welfare regimes. It also incorporates the ‘ought to’ dimention, because just as decommodification is the goal of social democratic welfare policy according to Esping-Andersen, ‘defamilialisation’ is the goal of feminist family policy according to most of its adherents. The original idea was that while de-commodified workers in Esping-Andersen’s typology gain bargaining power vis-a-vis their prospective employers, defamilialised women also gain power vis-a-vis their male partners (if they are in a heterosexual relationship), because they are no longer dependent on their male partners’ incomes to survive. If women are economically independent, they can also more easily survive without having a male partner or when raising children in a same-sex relationship (c.f. Lister, Reference Lister, Baldwin and Falkingham1994).
A major problem with this alternative is that it is not so clear what defamilialisation means. Taken literally, one could conclude that the goal of policy-making should be to enable parents to give their children to formal childcare as soon as possible, since responsibility for childcare would come away (i.e., de-familialising) from the family. Of course, most proponents of this term do not mean it that way, and they think it is good if fathers share in the parental leave time. Nonetheless, the term creates confusion because literally father leaves are familialising, since the family is taking care of the children. Thus, Daly and Schweiwe (Reference Daly and Schweiwe2010) support the notion that fathers should share the parental leave, but they still consider father leaves to be familialising. Similarly, Leitner (Reference Leitner2003) endorses the Swedish model, but still claims that Sweden has a familialising regime, although it is ‘optionally’ familialising, because according to her, parents can choose whether to go on leave or send their children to daycare. Consequently, not only is there no consensus on whether familisation is good or bad, there is not even any consensus on which countries are defamililised and which are familialised. For example, while Daly and Schweiwe (Reference Daly and Schweiwe2010) and Leitner (Reference Leitner2003) consider Sweden to be familialized, Hantrais (Reference Hantrais2004) considers Sweden to be defamilalised. Or Hantrais (Reference Hantrais2004) considers the UK to be partially defamilialised, while O’Connor et al. (Reference O’Connor, Orloff and Shaver1999) classify the UK as ‘familialist’.
Saxonberg (Reference Saxonberg2013) offers a solution in developing a typology based on the degrees of genderisation and degenderisation. He argues that the primary goal for feminist family policy has been to eliminate gender roles, which was epitomised in Sainsbury’s (1994) book Gendering Welfare States. Policies should be categorised according to whether they promote the elimination of gender roles (‘degenderising’) or whether they support the continuation of gender roles either directly (‘explicitly genderising’) or indirectly (‘implicitly genderising) policies. In the empirical typology that he presented, Saxonberg focused on the types of parental leaves and daycare policies. Parental leaves that were based on the income replacement principle and include some months reserved only for the father are degenderising, while market oriented policies that offer no leave benefits or means-tested ones are implicitly genderising and policies that include long parental leaves that pay lump-sum benefits are explicitly genderising. When it comes to daycare, policies that provide easy access to daycare for children under six are degenderising, while policies that rely on the market are implicitly genderising. Explicitly genderising policies follow the continental conservative model in that they give very little support to daycare for children under three years old, but moderate support to daycare for children three to six years old, although this daycare is often part-time.
Some scholars criticised the degenderisation model for neglecting the intergenerational effect (Lohnmann and Zagel, Reference Lohmann and Zagel2016) or gay rights (Hildebrant, Reference Hildebrandt2018). These authors make the mistake of concentrating on the empirical analysis of Saxonberg’s (Reference Saxonberg2013) article, rather than its theoretical discussion. As the article points out, theoretically, one can apply the degenderisation typology to all kinds of social policies – even those that are not connected to family policies. The indicators used in the article only provided an example of how one could apply the typology. For example, one could apply the typology to healthcare and analyse whether sick leave benefits and benefits for carrying for sick family members either promote the elimination of gender roles or aim to strengthen the gendered division of labor.
Another important aspect of Saxonberg’s (Reference Saxonberg2013) model is that it focused on policies, while the regime-type tradition which Esping-Andersen and most scholars applying the defamilialisation model include the private sector in their typologies. The advantage of concentrating on policies and excluding outcomes is that it makes it possible to study what influence different types of policies have on society when these policies are implemented in different cultural and economic settings. Similar policies can lead to much different outcomes depending on the socioeconomic and cultural backgrounds of particular countries (Saxonberg, and Szelewa, Reference Saxonberg and Szelewa2007; Saxonberg, Reference Saxonberg2014). Thus, rather than only relying on institutional factors such as policies and opportunities offered by the market, cultural sociologists began to investigate the role of culture in family policy.
Cultural approaches
Cultural theorists observe that even if two countries have the exact same types of parental leave, the percentage of leave time taken by fathers might radically differ depending on the differences in cultural norms about the ‘proper’ roles for mothers and fathers (e.g., Pfau-Effinger, Reference Pfau-Effinger2000). Cultural matters at both the macro and micro levels.
At the macro level, Pfau-Effinger (Reference Pfau-Effinger2005) concludes that variations in the dominant cultural family models among societies can explain their different development paths of policies towards family and gender in Europe. Kremer (Reference Kremer2007) develops the notion of ‘national ideals of care’. According to her, national family policies reflect the dominating cultural norms about who should care for children. If policies go against these norms, parents might not follow the economic incentives that these policies promote. She distinguishes four ideals that form alternatives to the full-time motherhood ideal: intergenerational care, surrogate mothers, parental sharing, and professional care.
At the micro-level, different groups might behave differently to the same policies, because they have different ‘gendered moral rationalities’. Regardless of what cultural values might dominate national policymaking, norms about child care and motherhood at the individual level still produce potentially different ‘gendered moral rationalities’ (Duncan and Edwards, Reference Duncan and Edwards1999). Even if childcare is free and of high quality, for example, some mothers will still stay at home with their children, because sending their children to day-care violates their moral views as to the ‘proper’ role of the mother (Duncan, Reference Duncan2005). Socio-economic and educational levels influence gendered moral rationalities. Thus, women with higher prestigious jobs are more likely to take shorter leaves than women with less prestigious jobs, and working-class mothers tend to be more in favor of long leaves than are middle-class working mothers (Evertsson and Duvander, Reference Evertsson and Duvander2010; Stefansen and Farstad, Reference Stefansen and Farstad2010).
Hakim (Reference Hakim2000) goes so far as to claim that individual values are so strong that national family policies have little influence. According to her preference theory, women in post-industrial societies have developed different preferences, so she breaks them into three groups: ‘career-oriented’, ‘family-oriented’, and ‘adaptable’ women. The career-oriented will always want to work regardless of policies, while the family-oriented will always want to give priority to having a family over working regardless of policies. Consequently, family policies can only influence ‘adaptable’ women. She also claims that policies cannot influence men, because all men are career-oriented and will not take care of children no matter how generous the parental leaves are.
Hakim’s typology came under heavy criticism. One problem is that she considers the group of adaptable women to be the largest (representing about 60 per cent of all women). Thus, she admits that policies will influence the choices of the majority of women. Another problem is that while she does show survey data concerning women, she just assumes all men are the same without testing her claim empirically. Yet, it is well-known that in some of the Nordic countries, policies have in fact induced fathers to greatly increase their share of the parental leave time. For example, in Sweden last year, fathers took over 31 per cent of the parental leave time (https://tco.se/fakta-och-politik/jamstalldhet/jamstalldhetsindex). Thus, she has been accused of biological essentialism and underplaying the role of policies (e.g., Ginn, et. al., Reference Ginn, Arber, Brannen, Dale, Dex, Elais, Moss, Pahl, Roberts and Rubery1996; Crompton and Lyonette, 2005).
However, conservative scholars are not the only ones who appear to have some sort of essentialism; some feminist scholars also display some amount of essentialism in assuming that very few men will take up parental leave, which means that parental leave becomes a trap for women, because it pressures mothers to leave the labor market for long periods (Morgan and Zippel, Reference Morgan and Zippel2003; Bergmann, Reference Bergmann2008; Lewis, Reference Lewis2009; Sipilä et al., Reference Sipilä, Katja, Tapio, Sipilä, Repo and Rissainen2010; Ferragina, Reference Ferragina2020). Pettit and Hook (Reference Pettit and Hook2009) state this clearly in presenting their ‘inverted U-shape’ curve that shows that in countries with no or short parental leaves and in countries with long parental leaves, women’s labor market participation is adversely influenced. However, in countries with medium-length parental leaves, like the Nordic countries, female labor market participation is higher. Thus, it is common to blindly accept the Nordic model, because the parental leaves are approximately one year or slightly longer (e.g., Javornik, Reference Javornik2014).
The problem with this conclusion is that there are hardly any examples of countries that have long parental leaves of the type that would induce fathers to share more equally in the leave time. There is general agreement that fathers are more likely to take up parental leave if the leaves are generous and based on the income replacement principle and if there are months only reserved for fathers (Saxonberg, Reference Saxonberg2009). Hungary is the only country that has a parental leave based on the income replacement principle that is longer than fourteen months, but it does not have a father quota and the government does not encourage fathers to take the leave. In contrast to the Nordic countries, the decision to introduce the two-year leave based on the income replacement principle was to encourage ethnic Hungarian mothers to have more children than the Roma (Saxonberg, Reference Saxonberg2014; Hašková and Saxonberg, Reference Hašková and Saxonberg2016). Thus, we do not have any examples of relatively long parental leaves that are designed to promote gender equality.
Yet, if we are to take the influence of culture seriously, then there is reason to believe that longer parental leaves in many countries would actually increase gender equality. The reason is that in the Nordic countries there is a strong norm for mothers to breastfeed their babies until they reach the age of nine months. Because of this, if given any amount of freedom of choice in the leave time, most parents will opt to have the mother stay at home for the first nine months as they place the child’s health over the issue of gender equality. Consequently, a study of Sweden and Norway shows that this breastfeeding norm is so strong that in every case in which fathers stayed at home longer than the quota period, the parents extended the entire parental leave time by several months, which they could do by taking less than 100 per cent of the leave benefit for the period (Bergqvist and Saxonberg, Reference Bergqvist and Saxonberg2017). Of course, the general norm for how long mothers should breast-feed babies is likely to differ among countries and regions, but this shows the importance of taking culture into account, and it also shows the need to not dogmatically accept the idea that one-year parental leaves are best for gender equality just because it is the most common period among the Nordic countries.
Another worthwhile finding of Bergqvist and Saxonberg’s (Reference Bergqvist and Saxonberg2017) study is that it shows that the state helps create norms. Even though the parental leaves are similar in Sweden and Norway in terms of their lengths and benefit levels, it was much more common for fathers in Sweden than in Norway to go on leaves that were longer than the quota period. The reason being that in Sweden, half the leave time is officially allocated to each parent, so if the father does not use up half the leave time, he must sign over ‘his’ time to the mother. Consequently, when fathers do not share the leave time equally, they are aware that they are not living up to what society expects of them and in interviews with such fathers, they were on the defensive and felt they had to justify their choice. In Norway, by contrast, the state does not officially give half the time to the father, so the fathers usually only stayed at home for the quota period and noted that they were doing their duty – they did what was expected of them.
Another example of how the state can create norms is the development of the ‘norm of threeness’ that has arisen in post-communist Central Europe. Certain historical-institutional developments going back to the nineteenth century have evolved to create the notion that mothers should work full-time until they have children, then they should stay at home until the children reach the age of three, after which they should work full-time again (Saxonberg, Reference Saxonberg2014; Hašková and Saxonberg, Reference Hašková and Saxonberg2016; Saxonberg and Maříková, Reference Saxonberg and Maříková2023). Both surveys (Saxonberg, Reference Saxonberg2014) and interviews with parents (Saxonberg and Maříková, Reference Saxonberg and Maříková2023) show widespread support for the norm of threeness among the post-communist Central European countries.
Bringing in the child’s perspective
Even though the degenderisation concept allows for a much broader gender-based approach than other typologies, a problem with only focussing on gender is that it leaves out the child’s perspective (Kurowska, Reference Kurowska2018). Generally, conservatives have been most successful in blocking family policy reforms that promote gender equality or the elimination of gender roles when they succeed in turning the dominating discourse of a country to being one of “greedy” mothers placing their best interests above those of the child (e.g., Saxonberg, Reference Saxonberg2014). Thus, the question arises as to whether bringing in the child’s perspective would force us to modify our policy proposals if there is a conflict between the mother’s interests and the child’s.
Luckily, to a large extent, measures supporting gender equality or the elimination of gender roles seem to benefit children as well. Thus, a study comparing various indicators of child wellbeing for different welfare regimes shows that dual earner regimes represent the best practice for promoting children’s health and development (Engster and Olofsdotter Stensöta, Reference Engster and Olofsdotter Stensöta2011). Another study shows that father-child attachments have a positive impact on child outcomes (Moullin et al., Reference Moullin, Waldfogel and Washbrook2014), which gives supports for the necessity of fathers sharing the parental leave time. If children are able to form attachments to both parents rather than just the mother, they display fewer depressive symptoms in adolescence (Kerstis et al., Reference Kerstis, Åslund and Sonnby2018). There is also an association between long paternity leaves and the positive perceptions of father-child closeness and communication among nine-year-old children (Petts et al., Reference Petts, Knoester and Waldfogel2020).
When it comes to daycare, studies have also concluded that high-quality institutions improve the child’s well-being. The longest study of children so far has been that undertaken by the Swedish psychologist Bengt-Erik Andersson, who studied children from the age of six months until they reached the age of twenty five years. His studies concluded that among thirteen-year-old children, the younger a child begins daycare, the better the school results will be for that child and the better the child will do in terms of social competence (Reference Andersson1992, Reference Andersson2005). In a British longitudinal study, the researchers found that children who attended daycare (where the carers had education training) had less behavioural problems than other children (Deater-Deckard et al., Reference Deater-Deckard, Pinkerton and Scarr1996: 940). A Canadian study concludes that children who attend daycare are less aggressive than children who stay at home with their mothers (Borge et al., Reference Borge, Rutter, Côté and Tremblay2004).
The fact that research indicates that policies which promote the elimination of gender roles can also help children, does not mean we should simply ignore children. For example, as noted above, sometimes conflicts to arise, such as the issue of how long the mother should breastfeed. If parents believe that mothers should breastfeed for the first nine-months and if there is also a norm against having the mother pump out her milk and put them in bottles, then fathers will only share the leave time equally if the parental leave period is extended to eighteen months. Moreover, if one accepts the argument that parental leaves that are longer than one year are not detrimental for gender equality if properly formulated (based on the income replacement principle and long non-transferable periods for each parent), then the question arises as to what the optimal time is for a child to attend daycare. So far, few studies have looked into this issue, as most have simply compared children who attend daycare at a certain age with those who did not. Thus, more psychological research is necessary and social policy experts should work together with psychologists on these issues.
Since infants cannot make choices about who cares for them, the question arises as to how we can frame their perspective. The Swedish discourse, for example, tends to frame it in terms of the child’s right – such as the right to have both parents (Saxonberg, Reference Saxonberg2009). However, this would not solve such issues as the optimal time for children to begin attending daycare. A promising solution comes from the capabilities approach. This framework concentrates on two levels: functionings and capabilities. Functionings have intrinsic value and can include basic forms of wellbeing such as housing, while capabilities are the functionings which are available and readily assessable to an individual (Kurowska, Reference Kurowska2018; Hobson et al., Reference Hobson, Fahlén and Takás2011). Therefore, policies should aim at improving the capabilities of people. Applied to the work-life balance of adults, Kurowska (Reference Kurowska2020: 406) proposes that capability not only to having enough time, but also ‘enough physical and psychical energy to engage in non-work activities while maintaining a given level of paid and un-paid work responsibilities’. When it comes to bringing in the child’s perspective, she interprets capabilities to mean basic children’s rights, such as the right to nourishment ‘independently of parental provision’ (Kurowska, Reference Kurowska2018: 34). Even though this approach represents a step forward, Kurowska (Reference Kurowska2018) is a bit confining in limiting children’s capabilities to the right of nourishment. To fully take into account the children’s perspective, more capabilities are necessary, such as the right to economic security, the right to a good education (including pre-school), and the right to have extended contact with both parents (including parental leaves) assuming the child has two living parents and did not come about through artificial insemination.
Intersectionality
Even though feminist research has long taken into account intersectionality in noting that women from different classes as well as religious and ethnic backgrounds might have different interests, the intersectionality approach as taken a backseat to much of the research on family policies, as researchers have focused more on gender equality, the labor market, or demographics. However, it is obvious, for example, that if daycare is mostly private, then middle-class women will be more likely to afford it than working class women. Or if a country has a parental leave of twelve to thirteen months based on the income replacement principle, but a norm that mothers should breastfeed the first nine months, then fathers will only share equally in the parental leave if it is at least eighteen months long. Consequently, Berqvist and Saxonberg (Reference Bergqvist and Saxonberg2017) found that in Sweden almost all the fathers who stayed at home longer than the quota period came from moderately high-income, middle-class families, where they could afford to extend the parental leave period by taking out less than 100 per cent of the benefits per month. (For example, if parents have the right to 80 per cent of their income for thirteen months, they can stay at home a total of twenty-six months and received 40 per cent of their income for that period). With the rise of rightwing populism movements, welfare chauvinist ideas are also spreading according to which certain groups such as immigrants and ethnic minorities do not ‘deserve’ welfare benefits, including family benefits (e.g., de Koster et al., Reference de Koster, Achterberg and van Der Waal2013; Rathgeb and Busemeyer, Reference Rathgeb and Busemeyer2022).
As scholars such as Williams (Reference Williams, Irving, Fenger and Hudson2015, Reference Williams, Daly, Pfau-Effinger, Gilbert and Besharov2023) point out, intersectionality is not only limited to the cross-section of gender, race/ethnicity, and class, but other factors, such as the children’s perspective and migration are important. She notes the importance of migration and the use of immigrant women for childcaring and notes the transnational aspect of family policy.
Even though much research has been conducted about families in wealthy countries hiring nannies from poorer countries to take care of their children, very little has been written about the opposite case: when people from poorer countries hire nannies from relatively wealthy countries. Souralová (Reference Souralová2020, Reference Souralová2021) provides a notable exception in her studies of Vietnamese families living in the Czech Republic, who hire Czech nannies, in which she integrates the perspectives of the parents, children, and nannies. She notes (2021) that the Czech parental leave is for up to four years, but the Vietnamese parents usually own small businesses and cannot afford to have a parent stay at home for such a long period; furthermore, they are used to the Vietnamese system in which there is only a six-month maternity leave but easy access to daycare. Because of the lack of daycare for children under three, they feel forced to turn to retired Czech women, who become their nannies. When the children reach kindergarten, the parents cannot pick up the children because of work. Even when the children are old-enough to attend Czech kindergartens, they have the problem that they are only open to around 5:00 pm, which is too early, as many parents work until 8:00 pm. Since schools also close much earlier than the parents finish work, they need nannies even after the child begins attending school. Their work is so demanding that they do not even think they can afford to go on vacations with their children. Thus, Souralová (Reference Souralová2020) terms this ‘doing everything for the children while not being with them’. So, this shows how the conservative Czech model of family policy makes it difficult for work-oriented parents to balance work and family life. It also shows the international power relations in that people from poorer countries feel forced to move to wealthier countries to make ends meet. The power relations also manifest themselves in that the nannies coming from the richer countries raise the children in accordance with the cultural norms of the wealthier country. On the one hand, it is economically advantageous as the children get social and cultural capital, but on the other hand, they lose the culture of their parents and their mother country (Souralová, Reference Souralová2014). An ethnic-racial aspect also arises in that the children have darker skin and are seen as coming from an ‘inferior’ culture while trying to integrate into the ‘superior’ culture.
Studies of family policy from an intersectional perspective are now beginning to emphasise more same-sex couples and those who do not have a binary gender identity (such as those identifying as queer). However, these studies have been mostly sociological analyses of power relations rather than investigations of family policies (for a review, see Few-Demo and Allen, Reference Few-Demo and Allen2020). Another important issue for family policy concerns the work-life conflicts of women with physical disabilities, which is beginning to be addressed in research on intersectionality, but not so much in research on actual family policies (Ryan and Briggs, Reference Ryan and Briggs2019). Similarly, Marra (Reference Marra2020) argues for the need to use an intersectional and reflexively emergent time use analysis for developing better family policies, but she does not actually develop policy proposals based on this or create some kind of policy typology based on this approach.
Conclusion
This brief review of the state-of-the-art points to certain recommended directions for future research. This includes bringing in a more child-centred perspective, using more intersectional analysis and paying greater attention to same-sex relations and people of non-binary gender identities. When taking a more intersectional approach it would also be interesting to look into the variety of private caregivers. Although there have been many studies about the role of migration and nannies, in which parents from wealthy countries employ immigrant nannies from poorer countries, very little has been written about the case in which immigrant families hire natives to be their nannies because the immigrant parents are both working more than full-time at their family businesses. Perhaps further research could even uncover other types of caring arrangements and relationships concerning immigrants.
Open access
