TOWARDS THE UNIVERSAL CARE COURSE MODEL: Care policies and employment patterns in Austria, The Netherlands and Sweden Open Access

Although we witness a strong increase in women's labour market participation, academics as well as policy makers argue for a still greater number to be employed more continuously over the life course. Lewis (2004, 2006; Lewis et al.2008) suggests that there is no simple policy that could represent a solution to the complex problem of managing time and care, and of the question of who is to paying for it. There is thus ample research on the determinants and consequences of women's and men's work and care patterns. Studies about family policies showed the complexities within and between different political measures at the national and the comparable level (Gornick and Meyers 2003; Hantrais 2004; Moss and Deven 2006). While some of the studies focus on a selection of topics – for instance on social services (Anttonen and Sipilä 1996) – others take a much broader approach, including arrangements about leave arrangements and financial provisions (Bettio and Plantenga 2004; Jensen 2008). Going into more depth, it has been shown that countries cluster differently for different policies (Leitner 2003; Szelewa and Plakowski 2008). In other strands of research, the focus lies on the impact policies have on female labour market participation (Jaumotte 2003, 2005; Blossfeld and Hofmeister 2006) or fertility rates (Castles 2003; D'Addio and D'Ercole 2005). The general picture emerging from a survey of previous studies shows that there are at least four influential factors shaping the policies on female employment patterns: childcare services, leave facilities, flexible working-time arrangements and financial allowances (Plantenga and Remery 2005). There is strong empirical evidence that a lack of outside childcare facilities negatively affects female labour supply (Stier and Lewin-Epstein 2001; Esping-Andersen 2002; Uunk et al.2005; Haas et al. 2006).

Contrary to comparative studies which refer to a large number of countries, we provide a description of survey data and policy-relevant indicators for only three countries. This is also to do justice to the peculiarities of national policies (Daly and Rake 2003; Crompton et al.2005; Crompton and Lyonette 2006). Austria, for example, has recently witnessed a growing interest in work-care issues. However, the results of this study have so far mainly been provided to policy makers and have not been made available yet to an international audience. To the present day, only a dwindling number of cross-national studies have included the Austrian context, whereas many studies deal with The Netherlands (Knijn et al.2003; Abrahamson and Wehner 2006) and Sweden (Morgan and Zippel 2003; Blomqvist 2004; Isaksson et al.2006).

The article could be understood as a further development of Fraser's (1994) work-care models, linked with the policy concepts of de Henau et al. (2006). In the following, these theoretical frameworks will be outlined in greater detail. Next, we provide a brief overview of the national context in care policies and mothers' employment patterns. In the concluding part, we argue that the policy framework and work-care models should be adapted in order to more effectively meet the changing requirements according to the age of the children, as outlined in the Universal Care Course model.

Drawing on the classification of the ‘childcare triad’ (de Henau et al. 2006), we analyse how different care policies, namely childcare services, leave facilities and financial allowances as well as flexible working-time arrangements (Plantenga and Remery 2005) create either incentives for (paid) time to care privately or for time to be in paid work (Morgan and Zippel 2003; Lewis 2006a ,b). In a further step, we refer to Nancy Frasers' well-known models (1994) about how to share the paid and unpaid work more equally between the sexes.

Although cross-national differences in labour market outcomes cannot only be ascribed to (care) policies, varying sorts of policies affect female employment. They may either be supportive of enabling parents to have time to care privately or to have time to be in paid work. Following the assumptions of a recent comparison of the old EU-15 countries (de Henau et al. 2006), state interventions on childcare can mainly take three different forms, called ‘childcare triad’. The findings provide three clusters of countries differing in policies which promote either public care through direct supply of public (publicly-funded) childcare or home care by replacing the income and providing job protection for parents taking leave. Finally, policies may favour the private external care option. This implies direct financial support to families to help them purchase care on the market (cash and tax benefits). Unless the authors emphasize that the countries clustered together are very heterogeneous in practice, their main features are as follows: the Nordic European countries are classified as the most effective policies to support female labour market participation by favouring public care institutions. The other two countries investigated here, The Netherlands and Austria adopt the second way of state interventions fostering private home care through extensive leave arrangements. The third type, the private external care is typical for the care situation in the United Kingdom and Spain, but plays a minor role in the here investigated countries.

From a cross-national perspective including a range of different countries, care policies may follow in principle one of the three above mentioned paths. However, taking a closer look at a small range of countries, we challenge these assumptions made at an aggregate level. First, the situation in Austria differs from the one in The Netherlands, which have both been classified as ‘home care models’ (de Henau et al. 2006). Second, we argue that the countries may choose either one policy and/or implement a bundle of contradicting policies, as has been shown by various other studies (Leitner 2003). Thirdly, we stress the argument that the cross-national comparative policy analysis misses one important point, since they over-emphasize the role of policies and implied homogenous outcomes without referring to concrete differences in employment patterns, which we aim to do. Therefore, we link these policy models to Nancy Fraser's work-care models.

Since paid work and care patterns differ largely between countries and are gendered all over the world, Fraser's (1994) ‘postindustrial thought experiment’ about how to integrate carers into the labour market has enriched the debate on gender equity. The concepts range from the ‘Universal Breadwinner model’, the ‘Caregiver Parity model’ to the ‘Universal Caregiver model’ and they have been elaborated in many ways leading to new country-typology buildings. As far as we can see, empirical and theoretical studies on women's employment participation can be divided into three main strands with opposite normative assumptions. First, a number of scholars puts a strong emphasis on how to universalize the breadwinner role (Fraser 1994: 51) and how to make work and formal care available and affordable for all. The proponents of this ‘universalizing, androgynous strand’ argue for a de-construction of differences between men and women by equalizing the social, economic and private conditions. An equal labour market participation of both sexes, an adult-worker (Lewis 2004) or parent-worker model (Lewis and Hobson 1997) yield hope for more equality to avoid negative long-term consequences for women who (temporarily) leave the labour market for reasons of caretaking. Women should at least be able to form an autonomous household (Orloff 1993) and become independent from men or from official state benefits.

Second, Fraser defines the Caregiver Parity model emphasizing ‘difference, but equality’. As she puts it: ‘The aim is not to make women's lives the same as men's but, rather, to “make difference costless”. Thus, childbearing, child rearing, and informal domestic labor are to be elevated to parity with formal paid labor’ (Fraser 1994: 55). The main idea is to upgrade the unpaid work by making it pay and to provide social insurance for carers.

Finally, both challenges, either the Universal Breadwinner or the Caregiver Parity model, do send different and contradictory messages about the social value of paid and unpaid work. Hitherto, the most promising vision about how to achieve more gender equity is to overcome the ‘gendered opposition between breadwinning and caregiving’ (Fraser 1994: 61) in the Universal Caregiver model. Here, men and women would both spend as much time in paid labour as in caring for others. A huge amount of recent studies thus deal exactly with the question of how to induce more fathers to care (Knijn et al.2005; Bygren and Ann-Zofie 2006; Plantin 2007) and to do unpaid housework (e.g., by incentives for leave options on a use-it-or-lose-it bas). However, since male work hours do not significantly change after the birth of a child and nearly all fathers in Austria, The Netherlands and Sweden are full-timers irrespective of the age of their children, we focus on mothers' work hours in the following description. Official data may give us a distorted picture, for instance, since Austrian mothers on maternity or parental leave are officially counted as ‘employed’, the overall female employment rate is comparably high (Lewis et al.2008). To enable cross-national comparability, we thus draw on a survey (the European Social Survey 2004).

At an abstract level, the countries to be chosen are good heuristic examples for the three above mentioned normative paths about how to integrate carers into the labour market. In theory, the Universal Breadwinner model (Sweden as an example) necessitates public care, the Caregiver Parity model (Austria) would be best reached by well-paid home care policies, whereas the Universal Caregiver Parity model (The Netherlands) needs a reduction and flexibilization of paid work hours and measures to divide the unpaid work more equally between men and women.

The Austrian leave system provides paid leave until the child is three years old, but at a fixed flat-rate payment (436 Euros per month). Six months of this three-year paid leave period have to be taken by the other partner on a use-it-or-lose-it basis. There are incentives to stay attached to the labour market insofar as parents receive the so-called childcare money, while being employed up to a certain earnings limit (14,600 Euros in a year). However, the availability of cash for care for all parents can also be a disincentive becoming established to on the labour market before having a baby. Moreover, there is a mismatch between the three years duration of being entitled to care money and the protection against dismissal ending at the second birthday of the child (Table 1).

To assess the effectiveness of paid leave systems, most comparisons refer either to the ‘effective parental leave level’ (Plantenga and Remery 2005) or to the ‘full-time equivalents’ (OECD 2007) which are calculated by the duration of leave in weeks weighted by the level of payment as percent of last earnings received by the claimant. Although the usefulness of these indicators is debated, they enable us to show that the Austrian paid leave system is favourable in time but not in money. Austria provides only 22 weeks in full-time equivalents, as compared to nearly twice as much granted in Sweden. Hence, the Austrian care system does not really provide a comparably good replacement for income loss and it is ‘transfer-heavy’ (Esping-Andersen 1999; Kreimer and Schiffbänker 2003) mainly with respect to benefits to children, tax reductions and social insurance for dependent carers.

Another well-known characteristic is the lack in childcare facilities. Only 9 percent of children under three years old are estimated to be cared for in formal institutions. The provision with childcare facilities for children between 3 and 6 years is comparably good, about 82 percent of them are in formal day care. However, there is a very limited supply of afternoon care facilities provided for school children aged between 6 and 14 years. Another aspect is that Austrian schools as well as some kindergarten in the countryside run on a part-day assuming that a parent is at home to provide lunch and afternoon care. In sum, public expenditures on formal day care are comparatively low with 0.4 percent of the GDP (Table 2).

Against this institutional background, we investigate the participation of mothers on the labour market. Drawing on the European Survey results from 2004, about 42 percent of mothers with children under the age of 6 are in paid work, whereas more than half of these mothers (52 percent) categorize themselves as ‘homemakers’. When children reach school age the share of employed mothers rises to 65 percent. According to the international convention of a 30-hour threshold for defining part-timers, the share of full-timers (60 percent) predominates among the mothers in paid labour.

We are, of course, well aware that the aggregate descriptive data are to be considered as first hypotheses to be proven by panel data analysis. In fact, we do not know whether mothers with children under three will behave like those who now care for pre-school or school-age children (Table 3).

Although there is no statutory entitlement for paid leave-taking in the Netherlands, a number of collective agreements do provide income replacement (e.g., civil servants receive 75 percent of their former wage) (Plantenga and Remery 2005; Fagan and Hebson 2006). The leave system is characterized by enabling parents to adapt their time to care and to paid work more flexibly. Statutory leave arrangements allow mothers and fathers three months, job protection for each parent. This protected leave period can be taken either full-time or part-time for twice the period (six months) before the child is eight years old. Additionally, all employees have a right to ask for part-time and to rever to full-time.

Childcare facilities are available for about one third of younger children under three as of 2003 (see Table 2). Although we do not get reliable information about how long these children are cared for in these institutions, it is highly probable that this means part-time caring in The Netherlands. Contrary to Austria and Sweden, childcare facilities at company level do play a substantial role in The Netherlands. Here, the provision and financing of childcare facilities have been defined as a combined responsibility of the government, the employer and the employee; each party pays approximately one third of the childcare costs. Although the contribution by the employers is not mandatory, it was estimated that in 2004 approximately 70 percent of all employees had received some contribution from their employer (Plantenga and Remery 2005: 21). Since the reconciliation of work and care is defined as a common tripartite responsibility, public expenditures on day care institutions (0.2 percent of GDP in 2003) are low as compared to the other countries. However, The Netherlands stands out as the only country in which the official policy via the ‘Combination Scenario’ (van Wel and Knijn 2006) creates a social vision about how to share care work between men and women more equally, namely through reduced work hours of men and women. A range of different policy measures support the statutory right to request reduced working hours for employees (Anxo et al.2007a: 11). The Working Hours (adjustment) Act (WAA), applicable companies of more than 10 employees, gives them the right to submit a request of working hours adaptation once in two years. The only precondition on it is having been employed within company for a minimum of one year when a request is submitted. An employer may only reject a request in case of serious business interest. Additionally, the Equal Treatment of Full and Part-timers (WOA) guarantees an equal treatment in the conditions of employment. Apart from these legal regulations, innovative policies are designed to help people in combining various activities (such as education or hobbies) more effectively in different phases of their lives. Employees are granted a statutory right to participate in the life course arrangement. A maximum of 12 percent of the annual wage can be saved for a maximum of about two years of leave. They are then entitled to three years of leave on 70 percent of the last earned income (Plantenga and Remery 2005: 38ff).

Against the background to enable work hours reduction, we look at employment outcomes for mothers. Not surprisingly, mothers in The Netherlands overall exhibit the lowest employment participation as compared to the other countries. About one third (36 percent) of mothers with children up to five years are employed, rising to about 51 percent of mothers when they have children between 6 and 15 years. A substantial proportion of these mothers (42 percent) still sees their main activity as performing housework and caring. There is no surprise about the fact that part-time work predominates among women in The Netherlands, where about two thirds of employed mothers work fewer than 30 hours a week. Contrary to other countries, they do not substantially change work hours when children grow older.

Swedish parental leave schemes offer a generous combination of time and financial support to parents. There are high earnings-related payments for most of the leave: one year at 80 percent of time earnings up to a ceiling and about two months with a low flat-rate benefit. In Sweden, the taking of parental leave can be divided flexibly between the parents in the first year; one parent may use two months more. As a result the total job protected period entails a maximum of 14 months for one partner. Swedish parents are protected against dismissal while being on leave for more than 16 months when both parents take some leave. Parents may also reduce working hours to up to 75 percent of normal work hours until the child's first year of school, when it is eight years old, with a right to return to full-time.

In Sweden childcare is framed as a social right insofar as parents of children between 1 and 12 years have a guarantee of a childcare place (Plantenga and Remery 2005: 12ff). Public childcare is available to pre-school children (1–5 years) either on a full-time or part-time basis. School children (6–12 years of age) are entitled to care after school-hours e.g., in leisure time centres. In 2003, 41 percent of the infants up to three years are in formal day care centres and pre-school aged children are nearly exclusively (90 percent) cared for in public institutions. Although these standards, availability and affordability are high by international comparison, there are hints at the fact that the former collectivization of care undergoes a process of change towards the privatization, which means that children of highly educated persons are over-represented within the non-public sector (Blomqvist 2004). Nevertheless, the good provision of public care institutions in Sweden is reflected in the stronger public expenditure, which is more than three times higher in Sweden (1.3 percent of GDP in 2003) than in Austria (0.4 percent) and in the Netherlands (0.2 percent).

Corresponding to the good provision with public care institutions, we witness a strong labour market attachment of Swedish mothers. The great majority of Swedish mothers with children under 6 years (74 percent) is in paid labour. It is interesting to see that only 4 percent of mothers with small children describe their main activity as ‘doing housework or looking after children’, but about 10 percent of these mothers see themselves as being unemployed. The overwhelming majority of Swedish mothers with school-age children (81 percent) are employed or see themselves as being unemployed (6 percent). Being a housewife is of little importance to Swedish mothers with school-aged children (1 percent). The share of full-time employed is as high among mothers with young children as among those with school children.

Studying cross-national differences in policies leads us to a new differentiation between paid and unpaid home care models. Paid home care is the prevailing model for the child's first 16 months in Sweden and up to the age of three in Austria, although at a much lower level. The Netherlands fall far behind the two other countries when we look at the amount of time and money granted for care by the state, a situation which has been named unpaid home care here. Nevertheless, we have to keep in mind that in the Netherlands, employers' contributions for parents might in fact be at least as efficient as the official provisions in the other two countries. Hence, it might be misleading to only study state policies without looking at compensation from the employers' side, as it has been demonstrated in a recent cross-national, comparative study (Anxo et al.2007b).

Although decisions regarding employment are multi-faceted (e.g., the economic necessity for paid work, or the female tendency to avoid leave in order to fulfil their career commitment), differences in care policies explain, at least up to a certain extent, why we found vast cross-national differences in the present study. Different elements to policies correspond to country-specific work-care models. The Austrian care-policy structure shows a tendency towards representing the Caregiver Parity model, since private care for children is paid for a comparatively long period, although only at the level of replacement for lower incomes. From a normative perspective, The Netherlands have been chosen as an example for the Universal Caregiver model, because they not only support flexible hours and a reduction of working hours, but also a scenario combining the two policies (Combination Scenario). Employees profit from the right to work short hours in order to have more time to care for their children. When compared to Austria and The Netherlands, Sweden represents the country which comes closest to the Universal Breadwinner model. It has established a social democratic, universal welfare state and is well-known for the state's efforts to back women's continuous full-time labour market participation.

In Austria, paid home care policies lead to the standardized pattern of mothers staying at home full-time, or at least part-time, during the first two or three years. In Sweden and The Netherlands, by contrast, flexible options regarding how to integrate parents more continuously into the labour market are provided throughout different caring phases. While in The Netherlands, mothers either stay connected to the labour market by working short hours, or leave it completely, Sweden provides the best incentives to stay at home while being paid for the first months, with a right to work reduced hours and to return to working full-time afterwards. In contrast to the other countries, there is no option or need for accepting a long-term homemaker role in Sweden. Owing to the number of public caring facilities, mothers can be employed without interruption and are able to constantly increase their work hours over the caring period. Overall, Swedish care policies result in the collectivization of a uniform pattern for all parents: in the child's first 12 to 16 months, the Caregiver Parity model predominates, followed by the reduction in work hours and the return to full-time work, the so-called Universal Breadwinner model. When compared to Sweden, we find a greater diversification among mothers in Austria and The Netherlands, where a lack of public caring facilities preponderates during the pre-school and school age care phases.

Finally, we investigated which country-specific aspects contribute to enhancing the opportunities for greater gender equality. As argued previously (Haas 2002), the extension of formal rights to take caring leave, including non-transferable parts, did not noticeably induce men to go on childcaring leave. In all three countries studied here, fathers did not change their work hours over the course of the caring period. In Austria, fathers even increased paid work hours compared to their male counterparts who did not have children. Swedish and Dutch policy approaches differ with respect to the incentives for fathers to share the caring workload more equally. In Sweden, parents receive adequate payment when going on leave. Also the so-called ‘daddys months’ (paternity leave) provide an incentive for getting familiar with childcare and the responsibilities involved. In The Netherlands, the Combination Scenario includes fathers who also reduce work hours in order to gain time to care for their children. In practice, however, fathers' contributions to care are rather insignificant. In principle, we agree that Sweden ‘has moved closest to the gender equity model’ (Gottfried and Reese 2003), maintained by a public commitment to social care provision, and that it represents a case of the Universal Breadwinner model. However, we neither find the Caregiver Parity model in Austria, nor the Universal Caregiver model in The Netherlands. Both countries indeed follow a Caregiver model without parity since carers do not enjoy financial independence from the male partner when staying at home caring, which would be an essential part of the Caregiver Parity model. As long as only women reduce work hours, and as hitherto are not able to make ends meet, we witness that, more than ten years after Fraser formulated her crucial thoughts, we are still in the phase of the ‘the family wage’, and have not surpassed it yet.

The empirical study shows that the provision of public care institutions and paid leave arrangements differ not only cross-nationally, but also across the different stages of childcare. For example, (paid) leave arrangements refer to a small number of years directly after birth, whereas caring facilities are more likely to be favoured when children reach (pre-)school age. Similarly, employment patterns of mothers strongly vary over the age of the children, also being more noticeable in some countries. Drawing on these insights, we argue for adopting a dynamic perspective on policies and on the organization of paid work and childcare.

In ‘search for new policy principles’ (Lewis and Giullari 2005), we refer to Fraser's Universal Caregiver model as the most promising vision for a way to achieve more gender equality. We aim to push her thoughts further by adopting a dynamic perspective. Furthermore, the elaboration we propose is more strongly based on empirical evidence in policies and employment. To address these differences, we have named it the Universal Care Course model (UCC). This model is to be understood as an illustrative model forpreliminary steps towards the utopian aim of achieving Fraser's Universal Caregiver model. In Fraser's vision, both women and men should spend as much time in paid as in unpaid labour, the latter consisting of childcare, housework and civil work. Since an overall reduction of working hours is far from being evident, transitional policies are deemed necessary to change the situation. Due to path dependencies (Esping-Andersen 2002), the complexities within policies (Leitner 2003) and discrepancies in cultures of childcare (Kremer 2007), different countries need different policy measures and principles for distinct care phases. The core question is: which country-specific measures enable us to go one step further into the direction of universalizing childcare? Universal care involves equal sharing of caretaking times between men and women. Consequently, we develop a gender-sensitive theoretical framework for the organization of paid work and care, varying over different care phases: first, the infancy phase comprises the period of care for children up to three years; in the second phase, children reach pre-school age (3–6); and finally there is the phase of school-aged children (older than six years).

A substantial time away from work is adverse to income and career prospects and the entitlement for social security benefits and pensions. In the literature, different assumptions about the long-term effects of leave-taking can be found (Lutz 2004; Lamplmayer 2004; Lalive and Zweimüller 2005). Generally, a break of no longer than several months to one and a half years, with a high earnings-related benefit, does no substantial harm, especially if there is a possibility of returning to the previous employer (e.g., Ziefle 2004). To avoid negative consequences and to de-gender the responsibility for childcare, each parent could thus be entitled to job protection and paid leave for one and a half years at 80 percent of the former salary, or at a fixed rate for parents without former labour market experience. In order to encourage and enable men to take childcare leave, it should not be transferable to the other partner. Parents could be allowed to change periods of leave-taking flexibly, according to their own as well as their employers' needs. Single parents could be entitled to a longer period of time to care, but the shorter the leave of the single parent, the better for their labour market integration and overall social position. The crucial point here is that our vision of a Universal Care Course model differs from Fraser's Universal Caregiver model in the following points: Instead of a continuous labour market integration with short hours, we argue for the option of a well-paid leave during the infancy phase. This is to avoid short-term consequences of labour market absence and to encourage men to take paid leave as well. However, the dilemma faced here is that policies in the infancy phase should address both high care transfers for a short break from work and high quality childcare services to enable labour market participation, if necessary or welcomed by parents.

When children reach the pre-school age, non-parental childcare is increasingly welcomed by parents and is also favoured by care policies (the provision of caring facilities for pre-schoolers is much better than for infants). Obviously, the reduction in work hours seems to be the favourable solution for this care phase if it is affordable and available. Nevertheless, since part time jobs are mainly held down by women who have caring obligations, gender differences are constantly being perpetuated, especially from a life course perspective. It has been widely documented that, compared to the male-dominated, full-time jobs, part-time work entails a large number of negative consequences, touching upon payment, career perspectives, social and employment security, and pensions, just to name a few (Fagan 2003; Dackweiler 2004; Lutz 2004; Klee 2005; Fagan and Hebson 2006). However, part-time work may potentially entail positive aspects when it is limited to a certain period of (care) time or when it represents an alternative to unemployment. In line with Fraser's Universal Caregiver model, we argue that the general reduction of work hours could be a desirable norm to meet the care requirements of this phase. In theory, men and women could share paid work and care work more equally on a part-time work/childcare basis.

For this stage of childcare, a (moderate) full-time employment (between 30 and 40 hours) would be a desirable norm. Furthermore, parents need flexibility in organizing their overtime, taking sick leave, and managing the holidays of their children. This suggestion differs from Fraser's model approach in which reduced work hours enable parents to sustain their living, being well-paid and insured. Since in many cases and countries‘ earning a double income is necessary, men and women may opt for longer paid work hours. As it has been repeatedly shown (e.g., Stier and Lewin-Epstein 2001; de Henau et al. 2006), this model makes good provision with high quality care facilities necessary.

Based on the cross-national comparison, we will now propose improvements on policies in order to universalize care and paid work. Since different national contexts call for different policy measures, we will hint at country-specific directions for changes in a gender-sensitive way.

In Austria, necessary measures for this first childcare phase are only partly fulfilled as long as the replacement for income loss is low. Furthermore, there is a substantial lack of places in childcare facilities. Sweden provides wage replacement rates that allow a certain extent of financial independence. However, it could be extended in time and expanded further to include the fathers. In The Netherlands, we witness a dependency of the carers on the male breadwinner, the employer and on the collective agreements guaranteeing paid leave and care places. We therefore propose a general paid leave arrangement. Paid home care measures should be associated with public care for those who choose to stay in or return to the labour market (Table 4).

In the second, pre-school care phase, parents in Austria and in The Netherlands need public care policies. The guarantee of an affordable and high-quality care place would enable a smooth transition to the labour market for them. In the vision of the UCC model, both parents could work on reduced hours, which is necessary to meet the caring responsibilities in exceptional cases (illness, holidays) and to avoid time strains and conflicts. As in Sweden, both parents should be legally entitled to reduce their work hours until the child is 8 years old, with a right to return to full-time employment.

Since full-time employment (30–40 hours but not more) would be a desirable norm for the third care phase, policies in Austria and The Netherlands should address the substantial lack in caring facilities. Parents need to be guaranteed full-time care in schools, leisure centres, and during school holidays. Once again, the Swedish pattern is the most promising in this respect.

This paper started out by arguing that Austria, The Netherlands and Sweden are heuristic examples of different models of distributing paid work and childcare between the sexes. To show the cross-national variations, we presented different child care policies and juxtaposed them with mother's employment rates. It emerged that different childcare regimes lead to different employment patterns, which does not come as a surprise. However, at the end of the descriptive analysis, we came up with a new model called The Universal Care Course. This is based on the theoretical framework of Fraser's models and it takes into account the empirical finding that care policies, as well as labour market patterns, vary substantially over the age of the children. Furthermore, we reflected on the desirable norms for different care phases: the infancy phase for small children, the pre-school phase, and the school-age phase. Drawing on the two principle policy options of having either public or paid home care, we have shown possible directions and paths for the exemplary countries. Cross-national differences would call for different policy options to address. The underlying premise here was to discuss how more equal sharing of paid work and care work between the sexes could be achieved. The aim of these broadly outlined policy measures was to show the directions policies may follow, with a focus on recognizing that care policies always address different care course phases. The proposed UCC model is to be understood as a starting point for discussion. However, it would need an expansion of the theoretical framing and the formulation of more precise policy measures. Key challenges are to show how to avoid negative effects of long-term breaks for more than one child, for the care of disabled or for elderly care. It is vital to find the ‘ideal mix’ of the different types of care, professional and private, and to heighten the value of care.

Last but not least, we follow Fraser's recent perspective in taking a more ‘trans-nationalizing stance’ (Fraser 2007) at power relations. Some women are able to afford domestic and care help, while others have to leave their children at home without supervision in order to look after middle- and upper-class children or elderly persons in the first world. We argue that the universalization and professionalization of childcare could be an alternative to the feminization and the devaluation of (child)care which we actually witness to a varying degree; in some countries (e.g., in Austria and The Netherlands) more than in others (in Sweden).

We are grateful for research support under the Sixth Framework Programme of the European Commission (WORKCARE-project, CIT5-028361, http://www.abdn.ac.uk/socsci/research/nec/workcare/) that have made the scope and depth of this research possible. We would also like to thank the anonymous referees for their helpful comments and suggestions.

Barbara Haas is Reader in Sociology and Social Research at the Vienna University of Economics and Business (WU, Austria). She specializes in cross-national comparative research on gender, work, employment, work-life balance and welfare. Her special methodological focus is put on how to combine quantitative and qualitative research methods. She is currently working on several projects dealing with work and care from different cross-national perspectives.

Margit Hartel is reader in empirical social research and statistics at the University of Vienna. Her research topics are work life balance, gendered employment patterns and promotion of women at the workplace. As a member of the department for Quality Management and Program Delivery at the University of Economics and Business she designs a panel investigation to show courses of studies and career entries for students of economics.