Parental leave policies and time use for mothers and fathers: a case study of Spain and Sweden

ABSTRACT States play an important role in gender equality through policy structuring. In this case study, the aim is to explore whether changes in parental leave policies over two decades trickle down to changes in gendered time use in two polarised countries: Sweden and Spain, represented by the Basque Country. Sweden represents dual-earner countries with high relative gender equality, whereas Spain represents a south European policy model supporting a breadwinning/homemaker ideal. The results show that changes in the gendered time use among mothers and fathers in both countries are associated with changes in parental leave policies. Changes in policies directed towards increasing gender equality reduce the gender gap in time use among mothers and fathers and seem to increase gender equality within a country. From these results, the conclusion is that parental leave policies that are structured to promote or enable gender equality could reduce the gender time gap in work among mothers and fathers.


Introduction
Throughout the last decades, and even still today, women have often had less access to labour markets; they constitute a higher percentage of part-time work than men and perform most housework duties (Connell, 2009). Gender-normative beliefs regarding care responsibilities and household work are strong barriers to employment opportunities and career possibilities for many European women (Haas, 2003). Family-friendly policies have been argued to be an important factor in the possibility of balancing work and family and reducing gendered work specialisationwomen's focus on housework duties and men's on paid work. A lack of policy support for women to combine work and family often means that women must choose between a career and children. Family-friendly policies structured to promote gender equality might give women greater autonomy and capability to make choices in life and to barging their position as worker, housewife and primary homemaker (Bianchi, Milkie, Sayer, & Robinson, 2000;Brines, 1993Brines, , 1994Coltrane, 2000). The question is whether these types of policies can inflect changes in normative behaviours among both men and women? Can fathers' involvement in parental care duties influence changes in work task specialisation? In 2007, O'Brian, Brandth and Kvande called for more research in the field and, to date, this need remains. Thus, the aim of this study is to explore whether changes in parental leave policies over two decades are visualised in the gendered time use among mothers and fathers. Thus, we will structurally analyse policy change in a broader welfare and social context using examples of two polarised countries with regard to gender equality -Sweden and Spain, represented by the Basque Countryto study behavioural changes in individuals. The purpose of using two polarised countries is that it will reflect two sides of a coin, thereby enabling us to draw a fuller picture of the situation.
Sweden and Spainwelfare state structure and gender context Parental leave policies are integrated into a wider social protection scheme in which welfare state structure and gender context play an important role. Korpi (2000;; see also Korpi et al., 2013), and later Thévenon (2011), attempted to develop welfare classifications that include dimensions of family policies and gender in order to contextualise institutional differences. These family policy models distinguish between policy structures in which women's role in the labour market is strong, and those that support more conservative gender-normative work specialisation. The welfare context in Spain and Sweden is described below.
Spain, represented by the Basque Country, belongs to a welfare model named "the southern model" (Ferrera, 1996), and is characterised by policies that depend on one's position in the labour market. The states included in this model provide weak social protection for people working in the irregular labour market with no job security, and the family constitutes the primary provider of welfare services, including child care. Family-friendly policies in these countries are scarce; social expectations maintain the traditional gendered division of work, with men as breadwinners and women as caretakers, and there is a lack of governmental support (Haas, 2003). In Spain, there has traditionally been a low level of female participation in the labour market, a strong tradition of male breadwinners, and a wide gender gap in the division of unpaid work (González-López, 2001;Mínguez, 2010).
In contrast, Sweden, which belongs to a dualearner model, is characterised by policies that support the equal sharing of work and child care (Esping-Andersen, 1990;Korpi, 2010). Dual-earner countries are often characterised by extensive goals to integrate women into the labour market and to provide comprehensive support systems for working fathers and mothers (Haas, 2003). Thus, policies and laws support women's labour market participation. Consequently, women's labour market participation is high and family policies are generous, with statesupported child care and parental leave for both parents. Sweden has been considered a role model in developing family-friendly policies (Haas, 2003). In the early 1970s, leftist parties in the Nordic countries began initiating policies to increase women's participation in the labour market; these policies included day care for children under school age and parental leave insurance granting parents paid leave to care for infants (Korpi, 2010).
Access to formal child care is an important aspect in relation to parental leave schemes. Where formal child care is available, of reasonable quality, and affordable, women to a greater extent participate and stay in the labour market (Hegewisch & Gornick, 2011). Thus, looking at child care facilities, Spain is argued to be of a familistic welfare regime, meaning that the family is the main caregiver and parents are supported to a greater extent by extended family members, such as grandparents, often particularly grandmothers (Badenes Plá & López López, 2011). Data of childcare enrolment (OECD, 2017) for children aged 0-2 years has shown that in Spain the participation rate in formal childcare was 43.1 per cent in 2004, and has since dropped to 38 per cent (no earlier data available). In Sweden, data show that figures went from 35.9 per cent in 1995 to around 48 per cent. Also, in Sweden, there is a well-developed system were parents are provided monetary compensation when they must stay home with sick children.
All in all, the welfare structure and gender context differ substantially between Sweden and Spain, giving parents different possibilities to combine work and families. But is there a link between policies and gender-equal work?
Parental leave policies and gender equalitywhat is the link and what do we know?
Public policies and institutions may shape the gendered patterns and individuals' experience of, and actions in, everyday life (Connell, 2009;Daly & Rake, 2003). States can affect individuals' opportunities and capabilities in life, either directly through laws or by targeting changes of norms through policies, resource allocation, and entitlements (Brighouse & Olin Wright, 2008). The state sets the frame for inclusion and exclusion in society and in the labour market by defining who is entitled to social benefits and capabilities in the public sphere through laws and policies. For example, in Spain benefits are allocated depending on one's position in the labour market. For those working in the irregular labour market, with no job security, the social protection is weak (Ferrera, 1996). In contrast, benefits in Sweden are universal and individualistic, meaning that everyone is entitled to them but that they focus on those in the greatest need. Through these entitlements people are given opportunities and capabilities, which may result in cross-country differences in attitudes and in the autonomy to make choices in life. For example, universal parental leave policies that give both parents the right to stay home with the child will give both mothers and fathers the opportunity to take parental leave and to participate in the labour market while having children. In contrast, policies expressing a homemaker/breadwinning ideal support a conservative division of gendered work and women's economic dependency on their husbands, leaving women with fewer capabilities and less autonomy to make choices in life (Korpi, Ferrarini, & Englund, 2013). Thus, in light of this argument, parental leave policies can be seen as providing either more or less support for dual-earner families.
What is, then, the result of the lack of gender-equal parental leave policies? Turning to employment trends, Hegewisch and Gornick (2011) concluded in a research overview that leave policies matter for individuals' employment decisions and possibilities. This is represented by the different levels of women in the labour market in Sweden and Spain. For example, in 2010, fulltime equivalent employment for women in Sweden was 48 per cent, whereas the percentage in Spain was 36 per cent (European Institute for Gender Equality [EIGE], 2013). Furthermore, the maternal employment rate among Spanish mothers with at least one child under the age of 14 increased from late 1990 to 2014 (OECD, 2017). Starting at 46.8 per cent, the rate increased to 59.5 per cent in 2010 and has since been stable. In Sweden, the employment rate for mothers with at least one child under 14 years has been stable at around 81 per cent over the last decade (OECD, 2017).
The lack of policies with the aim of facilitating the combination of work and family for women can also lead to a postponed transition to parenthood and reduced fertility rate within a country (Bettio et al., 2013). Europe is facing demographic challenges, and the E.U. has recognised the importance of familyfriendly policies as a means to combat these challenges (European Commission, 2013). Haataja (2005) argued that in countries where there is no support for balancing family and work, fertility rates are decreasing and women's labour market participation is increasing. Only the Nordic countries, with their strong social and family policies, have managed to sustain a fertility rate while increasing women's participation in the labour force (Haataja, 2005). Furthermore, there seems to be a link between fathers' uptake of parental leave and sustained fertility rates in the Nordic countries (Duvander, Lappegard, & Andersson, 2010). In Sweden, fertility rates were down to 1.5 in the mid 1990s, but are now just below 2 (OECD, 2017). For reference, for stable demographic development the fertility rate should be at 2.1. The South European countries have had a fertility rate well below the E.U. average over the last decades (Escobedo & Wall, 2015). Spain has shown the most compelling rate of all southern countries. Fertility rates have dropped markedly since the mid 1970s and have, since the beginning of 1990, been around 1 (OECD, 2017).
When discussing parental leave policies, it is important to distinguish between parental, maternal, and paternal leave. Typically, paternal and maternal leave is reserved for the father and mother, respectively, and is economically distributed between each of the two. Parental leave indicates neutrality, and is meant for either parent. Subsidies often emanate from the child instead of the parent (see Dearing, 2016or O'Brien, Brandth & Kvande, 2007. Though the literature has stated that all countries in Europe provide maternity leave (cf. Dearing, 2016), this is not the case for Sweden (see below). Instead, parental leave is emphasised, since it is gender neutral and does not convey normative messages about family constellations. In recent years, the concept of "father's quota" has popularly been used for parental leave, which is reserved for the other parent (often the father is the other parent). To withhold the neutrality of wording parental leave, we rather refer to parental leave reserved to each parent.
As a way to reflect the level of gender equality parental leave promotes, Brighouse and Olin Wright (2008) grouped these policies into three broad categories: equality-impending leavesincluding policies that exclusively support the mother as care giver, including unpaid leave; equality-enabling leavespolicies that provide generous paid leave to the family as a unit, which makes it easier for the father, if they like, to take part in the parental leave; and equality-promoting leavepolicies that create incentives for families to share caregiving activities. The last group is often signified by parental leave reserved for each parent, which cannot be transferred between parents. This group of polices is currently often found in the Nordic countries, but can also be found in the Netherlands, Portugal, and Quebec. In addition Ray, Gornick, and Schmitt (2009) compared 21 public parental leave schemes on several dimensions, and consequently identified five policies that are considered the most important for gender equality: (1) generous paid parental leave, (2) nontransferable leave reserved for each parent, (3) universal coverage combined with modest eligibility restrictions, (4) financing structures that pool risk among many employers, and (5) scheduling flexibility in the workplace. The laws regulating parental leave in Sweden and Spain from the early 1970s to date are outlined below.

Sweden
Sweden was the first country in the E.U. to introduce a generous and flexible parental leave system with the aim of equalising the shares of breadwinning and child care responsibilities (Haas, 2003). Parental leave in Sweden was first introduced in 1974, and both men and women had the right to stay at home with their children. In the first year of the programme, parental leave was six months with an allowance of 90 per cent of previous pay (up to a certain amount). In 1975, the leave period was extended to seven months, and in 1978, it increased to nine months. In 1980, three additional months of leave were added, but with a low flat rate. Nine years later, in 1989, parental leave increased to 15 months. In 1995, one month of parental leave was reserved for each parent. Since 2002, two months of leave have been reserved for each parent, and parental leave has expanded to 16 months. Leave can be used until a child turns 8 years old (Statistics Sweden, 2007).
Until 1979, there was no maternal leave before or after a baby was born. However, since 1980, pregnant women whose work could physically endanger their pregnancy have been allowed time off of up to 60 days before delivery (Statistics Sweden, 2012). Since 1980, fathers have had 10 paid paternity days to stay at home with the baby and the mother, from the day that the baby is born until 60 days thereafter.
In conclusion, over the studied period Sweden has progressively developed parental leave from what Brighouse and Olin Wright (2008) named an equality-enabling leave policy to an equality-promoting leave policy, first in 1995 and then in 2002.

Spain
Maternity rights in Spain have not changed greatly in recent decades. In 1980, mothers were entitled to maternity leave of 14 weeks with an allowance of 100 per cent of their previous pay, which could be distributed before or after a baby was born. In 1989, maternity leave was extended to 16 weeks, of which six weeks must be taken following the child's birth. Four of these weeks can be transferred to the father (Ley 3/ 1989). Since 1989, mothers have had the right to use one hour of work for breastfeeding each day (Ley 3/ 1989). In 2007, the law changed, and breastfeeding hours can now be accumulated to provide full days of leave for either parent (Ley Orgánica 3/2007).
In 1980, fathers were entitled to two days of paternity leave (Ley 8/1980). In 2007, paternity leave was extended to 13 days (15 days for public employees) with 100 per cent of their pay (Ley Orgánica 3/2007). In 2009, paternity leave was further increased to four weeks with full compensation. This was intended to come into effect in 2011 (Ley 9/2009); however, implementation was delayed, and it came into effect on 1 January 2017.
In the early 1980s, leave without compensation was introduced, allowing one parent to stay at home with their children until the children are three years old (Ley 8/1980). Since 1995, parents with children up to age six have been allowed to reduce their workdays, with a reduction in pay that is not subsidised (Real Decreto Legislativo 1/1995). By 2007, this benefit was extended to parents with children up to age eight. During a child's first year, a parent's return to the same job is protected, and after the first year, job protection is restricted to jobs within the same area (Escobedo & Meil, 2013).
In contrast to Sweden, the regions in Spain are autonomous and can create their own laws and benefits. The parental leave polices presented above cover all of Spain; however, some regions have augmented these policies. The data in this study are collected from the Spanish region of the Basque Country. In 2000, the Basque Country and four other regions implemented a flat-rate benefit given to either parent for three years after a baby is born (Escobedo & Meil, 2013;González-López, 2001;Lapuerta, Baizán, & González, 2011).
According to Ray et al. (2009), this benefit is important because it stimulates and enables increased gender equality. Thus, in 2000 the Basque Country moved from equality-impending leave towards equalityenabling leave, introducing a flat-rate benefit to either parent. However, this flat rate is relatively low; thus, full equality has not been achieved. Nevertheless, Lapuerta et al. (2011) showed that the economic incentives for parental leave implemented in the Basque Country has stimulated the use of parental leave and increased take-up rate, indicating its importance.
Time usecould gendered time use relate to welfare state structure and policy development?
Time use and division of work between men and women is coloured by normative behaviours. These behaviours are socially produced and reproduced throughout life, and strengthen adults' gender expectations. Dividing work becomes an act of "doing gender" (West & Zimmerman, 1987), and work task specialisation is thus based on our ideas of femininity and masculinity. With respect to this, work allocation springs from the fact that men and women affirm and reproduce gender by sticking to their gendered work tasks (Connell, 2009;Hochschild & Machung, 2003). Thus, work (both paid and unpaid) is an arena in which gender order is emphasised, and is therefore of interest to study with regard to gender equality. In addition, as argued by Brighouse and Olin Wright (2008), division of household work is "a significant, systematic determinant of broader patterns of continuing gender inequality" (pp. 364-5). Thus, the time use of men and women reflects both gender relations and prevailing social norms (Sullivan, Coltrane, McAnnally, & Altintas, 2009).
What, then, do we know of gendered time use? In general, research has confirmed that work is gendered and that work task specialisation is based on gendernormative values (see Hagqvist, 2016). Time use and the division of work seem to differ across countries and relate to macro-level gender equality and gender norms. Generally, the division of work is less equal in countries that are more aligned with the breadwinning norm than in countries with a dual-earner welfare structure (Fuwa, 2004;Fuwa & Cohen, 2007;Nordenmark, 2004). For instance, for women, 8.5 per cent of the variation in time spent on paid work is determined by country of residence, and 6 per cent of time is spent on unpaid work (Van Der Lippe, De Ruijter, De Ruijter, & Raub, 2011). In addition, about 12 per cent of the variation in men's housework involvement is explained by the country in which they live (Thébaud, 2010).
The length and structure of parental leave policies in a country seems important for gendered time use and division of work. In a cross-country comparison, results indicated that in countries with long parental leave periods, the division of work seems more unequal than in countries with shorter parental leave (Estes, 2011;Hook, 2006Hook, , 2010Schober, 2011). In addition, well-paid leave and paternity leave (including leave reserved for each parent) provide circumstances for more equally divided work (Dearing, 2016). Fathers who live in countries that offer paternity leave assume a greater share of the unpaid work (Hook, 2006), and for child care activity (Sullivan et al., 2009).
Additionally, studies have shown that women who return to full-time work reduce the amount of time spent on unpaid work to a greater extent compared to those returning to part-time work (Schober, 2011;Thomas & Hildingsson, 2009). Furthermore, fathers whose partners return to full-time work tend to increase the amount of time spent on housework (Schober, 2011). Thus, parental leave policies seem important in relation to time use for parents, not only during a period of parental leave but from a whole-life perspective.

Method
Initially, time use data from the Multinational Time Use Study database (M.T.U.S.) will be explored. M.T.U.S. standardises national time use data based on a nationally representative sample of adults of working age (Gershuny et al., 2012;Gershuny, 2000; for more information see http://www.timeuse.org/mtus). Swedish data was collected from Statistics Sweden, and Spanish data from the Basque Statistics Office (Eurostat), which covers the Basque Country, a region in north-eastern Spain. The data is cross-sectional and, for both countries, extracted from three points in similar years. For the Basque Country, time point 1 was 1992-1993, 2 was 2002-2003and 3 was 2008for Sweden, time point 1 was 1990-1991, 2 was 2000and 3 was 2010(Gershuny et al., 2012. The data included information on how the respondents spent their time according to 150 different activity codes within a 24-hour period. The data was reported in minutes per day, but is presented here as hours per day. The activity codes are standardised within M.T.U.S., and hence similar for the Basque Country and Sweden for all three time points. For this study, data from weekdays was used to better reflect a normal work day. The samples included fathers and mothers from 18 to 65 years old, who were married or cohabiting, and who had one or more children under the age of 19 living at home. The sample size for time point 1 was 148 (58% mothers) for the Basque Country and 1,481 (52% mothers) for Sweden. In time point 2, the sample size was 1,361 (43% mothers) for Sweden and 2,710 (55% mothers) for the Basque Country. Finally, the sample for time point 3 included 936 (56% mothers) for Sweden and 284 (56% mothers) for the Basque Country.

Variables
For this study, three different time use variables were used. Each variable comprises a summary of the time use activity codes present in the time use data. The three variables and the included activities were: (1) Paid work time, including time spent on paid work while employed or self-employed. (2) Unpaid work time, including tasks often referred to as routine household work, such as cooking, washing dishes, making beds, doing laundry, maintaining clothes, cleaning, shopping, budgeting and organising finances, and maintaining possessions.
(3) Child care work, including caring for children, such as feeding and food preparation, washing and changing babies, interacting, helping with homework, supervising, and attending school meetings.

Statistical analysis
Drawing on time-use data, we seek to identify changes in gender relations among parents as a result of a transition in parental leave policies. There are two main analytical strategies to identify changes: absolute change over time and mothers' change in relation to the father (Björk Eydal & Rostgaard, 2016). Thus, as a first step the absolute change is studied, initially using a descriptive analysis that presents the mean daily time use for fathers and mothers in Sweden and the Basque Country for all three time points. This data is presented as diagrams. Subsequently, absolute change among mothers and fathers for each time period and for each time use activity was first tested for significant changes using an ordinary least squares (O.L.S.) regression (the first time period is imputed as a constant). Second, relative changes were tested. Using O.L.S. regression, mothers were compared with fathers for the three time points. In Tables 1 and 2, the B value is presented with a 95 per cent confidence interval.

Results
Starting with absolute change in time use among Basque and Swedish mothers and fathers, Figures 1 through 3 present the mean hours per weekday spent on paid work, unpaid work, and child care, respectively, by time point. The descriptive data on paid work indicates that Basque mothers reported an increase in their mean hours per weekday from time points 1 to 2, but a more noticeable increase from 2 to 3. The mean working hours per weekday for Basque mothers was two hours at time point 1 and nearly four hours at time point 3. Swedish mothers made no changes in the mean time spent in paid work from time points 1 to 2, and only marginal changes from 2 to 3. Swedish fathers decreased their time spent in paid work from time points 1 to 3 by nearly one hour. The time differences for Basque fathers were marginal throughout the studied period ( Figure 1). Over all, the gender time gap in paid work seems to narrow over time in both The Basque Country and in Sweden. Figure 2 shows that between time points 1 and 3, Basque and Swedish fathers marginally increased their time spent on unpaid work, whereas mothers exhibited a considerable decrease in this type of work. Basque mothers marginally decreased their time in unpaid work between time points 1 and 2; however, between 2 and 3, the reduction in time was considerable. Swedish mothers successively decreased the time spent on unpaid work between time points 1 and 3. As with paid work, the gender time gap in unpaid work seems to narrow over time, though mainly due to mothers' reduction in time.
In Figure 3, the descriptive data show that from time points 1 to 2, all groups but Swedish mothers increased their time spent on child care. Basque mothers considerably increased their time spent on child care between time points 1 and 2; however, after time point 2 the changes stagnated. Swedish mothers reduced their time with children between time points 1 and 2; however, between 2 and 3, the changes were marginal. All in all, more time seems to have been spent on care work, and men seem to have taken more of this time over the studied period.
In Table 1   Considering paid work first, Basque mothers showed significant increases between time points 1 and 2; however, the difference between 1 and 3 was larger, indicating that a more considerable change occurred between time points 2 and 3. Basque fathers had a marginal change in the amount of time spent on paid work over the study period. Swedish mothers had a small, insignificant time difference in paid work between time points 1 and 2, but a significant total increase in time spent on paid work between 1 and 3. Table 1 shows that fathers from Sweden significantly reduced their time in paid work during the study period, primarily between time point 1 and 2.
With respect to unpaid work, the results show that both Basque and Swedish fathers had a marginal but significant positive change in the number of hours spent on housework from time points 1 to 3. The data shows that the time spent on unpaid work for mothers from both Sweden and the Basque Country decreased significantly and substantially. Basque mothers show a small change between 1 and 2 and a considerable change between 1 and 3, indicating that most of the change occurred between time points 2 and 3. Swedish mothers had a significant successive reduction in the amount of time spent on unpaid work from time points 1 to 3.
The third column in Table 1 shows that all but Swedish mothers showed a significant increase in the amount of time spent with children.
Turning to relative changes, Table 2 shows mothers' time use in relation to fathers' for each time point for the respective countries. The data shows significant time differences between the amounts that mothers and fathers spent on all time use activities (p > 0.001), confirming the gendered behaviours in time use. For both countries and across time points, fathers spent more time on paid work than mothers, while mothers spent more time on unpaid and care work than fathers.
The differences between Basque fathers and mothers indicate a considerable change between time points 2 and 3, narrowing the gender time gap for paid work. Similar patterns can be observed when examining data on unpaid work for Basque mothers and fathers. There is a significant difference between Basque fathers and mothers regarding the amount of time spent on child care work; the difference seems to remain across time points. From Table 1, we draw the conclusion that this is due to the fact that both mothers and fathers have increased their time spent on care work over timefathers somewhat more than mothers in the Basque Country.
The relative differences in the amount of time spent on paid work for Swedish fathers and mothers appear to be successively moving towards greater equality, and similar patterns are seen for unpaid work and child care work. Regarding change in gender time gap in unpaid work, Tables 1 and 2 indicate that though fathers' time is increasing, it is primarily mothers time in unpaid work that are decreasing.
In line with the progressive changes in parental leave policies, Sweden seems to have successively narrowed the gender time gap in paid work, unpaid work, and care work for mothers and fathers. Respectively for the Basque Country, between the first and second time point the Basque Country belonged to the equalityimpending leave model, which is reflected in the time use data. The gender time gap was rather large in all three variables, and there was a lack of movement towards each other. In 2000, The Basque country introduced a policy change moving from the equality-impending leave model towards equality-enabling leave. In the time use data this step is identifiable: between time points 2 and 3 the gender time gap in paid, unpaid, and care work narrowed among Basque mothers and fathers.

Discussion
This study has explored whether changes in parental leave policies over two decades are visualised in the gendered time use in two polarised countries with regard to gender equality: Sweden and Spain, represented by the Basque Country. Time use data support a transition towards greater gender equality regarding work in both countries over the two studied decades. In addition, changes towards greater equality in parental leave policies can be detected in the gendered time use behaviours among individuals. Thus, this study has contributed to previous knowledge by studying patterns of change on an individual level related to changes in policy development.
Sweden has, during the studied period, progressively moved towards equality-promoting leave with parental leave reserved for each parent. By the end of the measured period, Sweden had, according to Dearing (2016), one of Europe's most gender-equal parental leave policies. Meanwhile, Swedish fathers and mothers have made successive moves towards more gender-equal time use over the three measured time points. Parental leave for each parent was first implemented in Sweden in 1990, and provides an example of policies that could change the customary involvement of fathers in housework and child care (Brighouse & Olin Wright, 2008). After the introduction of leave reserved for each parent, Swedish parental leave policies now meet all five of the practices that Ray et al. (2009) considered important for gender equality: generous paid parental leave, non-transferable quotas of leave for each parent, universal coverage combined with modest eligibility restrictions, financing structures that pool risk among many employers, and scheduling flexibility in the workplace. Swedish statistics from this period additionally show an increase in fathers' uptake of parental leave days and a reduction in leave among mothers (Statistics Sweden, 2012). The results indicate that perhaps the narrowing gender time gap in unpaid work is foremost a result of mothers reducing their time. Though equalising time use, the norm of mothers as the main homemaker is still strong in Sweden (Hagqvist, 2016). Despite small changes with regard to housework, the results indicate that an equality-promoting leave model seem to trickle down to individuals' gendered behaviours in dividing work. To further promote equality, perhaps a more aggressive leave structure, in line with the Icelandic 3-3-3 system (Björk Eydal & Rostgaard, 2016), should be implemented in Sweden. In January 2016, the Swedish government introduced 90 days of parental leave reserved for each parent, though we have yet to see whether this will have any effects on time use.
Turning to the Basque Country, the progression on parental leave policy development has been more moderate over the studied period. From time points 1 to 2, the Basque Country met none of the criteria for practices that Ray et al. (2009) considered important for gender equality. In addition, the Basque country clearly belonged to an equalityimpending leave model (Brighouse & Olin Wright, 2008), mainly focusing on mothers as care givers. In 2000, politicians in the Basque Country made changes to their parental leave policies that allowed parents, both mothers and fathers, to stay at home with their children with a flat-rate benefit. Thus, the Basque Country moved towards, though without entering, the equality-enabling leave model. In this study, the Basque Country showed only small changes in gendered time use between time points 1 and 2. Since time point 2, however, Basque mothers and fathers have made considerable strides towards equal time use. During the same period of time the amount of births covered by leave entitlement increased, indicating that more parents are financially covered when on leave (Escobedo, Meil, & Lapuerta, 2016). In 1995, 31 per cent of births were covered by maternity leave benefits, and this rose to 68 per cent in 2009. In 2007, 35.2 per cent of fathers were coved by leave benefit, which reached 22.4 per cent in 2014 (Escobedo, Meil, & Lapuerta, 2016). Thus, the results support that the equality-enabling leave model better supports equal sharing and greater gender equality in relation to the equality-impending leave model.
In sum, the findings from this study support the idea that policy changes in Sweden and Spain, represented by the Basque Country, could have trickled down to the behaviours of fathers and mothers with regard to gendered time use. In addition, the structure of parental leave policies is important for gender equality. The two studied countries are contextually different with regard to equality and the following structuring of policies in general. Even so, this study shows that countries, no matters how gender equal initially, might benefit from parental leave that focuses on promoting gender equality. Thus, governments seem to be able to steer towards greater equality by formulising equality-promoting leave policies.
As noted earlier, policies giving both parents the right to stay at home with children, and the benefits that allow them to do so, give both the mother and father incentives and opportunities to combine work and family. Consequently, these policies can contribute to shifts in attitudes and actions among parents toward more gender equality with regard to work, and gendered task specialisation within a country. Duvander and Johansson (2012) showed that incentives for fathers to take on greater responsibility in the home can influence individuals' attitudes and actions when becoming a parent. The introduction of leave reserved for each parent in Sweden, and flat-rate benefits and the possibility for both parents to stay at home in the Basque Country, and the association this has with gendered time use, is in line with this reasoning.
It could be argued that these changes in time use were a natural move towards equality or were induced by other changes in society. An unpublished manuscript by one of the authors (Hagqvist, 2016) supports the fact that societies are moving towards a more gender equal division of housework. However, changes over recent decades are relatively small and seem primarily to be a result of women entering the labour market (Bianchi et al., 2000;Hagqvist, 2016). In European societies, men as a group is the norm and masculine features are valued higher than female (Hirdman, 1990;Sulivan, Coltrane, McAnnally & Altintas, 2009). The political and public strides towards gender equality are foremost based on the idea that women should increase their time in paid labour; thus, equality is based on a male norm (Hagqvist, 2016). Thus, to impose stronger gender egalitarianism, economic or gendered normative constraints related to work need to be removed. For instance, the flat-rate benefit introduced in the Basque Country, though small, increased mothers' economic independence, and hence their autonomy, when at home with children, and hence their position within the family and society (Bianchi et al., 2000). When economically dependent on the breadwinner, mothers tend to perform most of the unpaid work in exchange for economic security (Brines, 1993(Brines, , 1994. Brines (1994) stated that the dependency relationship between the breadwinner and the dependent is contractual within a marriage. The equalising effect on work division in relation to women's higher income is stronger in countries that lack gender-equal labour laws than in countries with more gender-equal laws (Fuwa & Cohen, 2007), indicating that a flat-rate benefit could be important in the bargaining regarding unpaid work for mothers in the Basque Country. In addition, a flat-rate benefit might make it easier for male breadwinners to take time off work to stay at home with children (Brighouse & Olin Wright, 2008). Furthermore, paid leave signals a norm that care work is important and valued in society, which could affect norms of gender equality within a country (Brighouse & Olin Wright, 2008). Indeed, in Sweden the introduction of leave reserved for each parent signals the importance of sharing work and the eligibility of fathers to stay at home. Additionally, it signals that fathers are capable as caregivers. Employers' previous lack of willingness to give fathers parental leave (Haas, Allard, & Hwang, 2002) may decrease if there is a policy supporting, and a general acceptance of, fathers staying at home. Indeed, Thomas and Hildingsson (2009) showed that fathers who use their parental leave take greater responsibility for the housework and for children.
The norm of good motherhood includes dimensions of being irreplaceable as a mother and taking responsibility for the child and family well-being, and demands women's time at home rather than at paid work (Elvin-Nowak & Thomsson, 2001;Fox, 2001). This is true also for Sweden (Elvin-Nowak & Thomsson, 2001). Meanwhile, the concept of fatherhood has often been related to the concepts of breadwinning (Björk Eydal & Rostgaard, 2016;Connell, 2009), which is also possibly reflected in the greater change in housework of Swedish mothers than fathers. It is our belief that to fully acknowledge gender inequality, and in line with changing time use, we need to address and discuss norms of good motherhood and fatherhood, and what it means to be a mother or father (see also Björk Eydal & Rostgaard, 2016).

Methodological considerations
Unlike the data for Sweden, which covers the entire country, the available data for Spain covers only the Basque Country. However, Spain, unlike Sweden, has autonomous regions, which constitute their own laws and policies. In relation to other regions in Spain, the Basque Country is a progressive region with regard to gender equality, and one of few regions in Spain to have introduced a flat-rate benefit. The conclusion is that although the Basque Country cannot represent all of Spain, it is a relevant and important case to study and does represent the southern European family policy model, similar to Spain in general.
Economic recessions could affect gender-segregated work differences (Escobedo & Wall, 2015). Because recessions are often followed by increased unemployment, such periods affect access to labour markets. As unemployment increases, it will affect the average time spent on paid work, which could in turn impact the time spent on unpaid work. Since 1990, there have been two major global economic recessions (Bettio et al., 2013). The second recession occurred in 2008, which is the time at which the data was collected for the Basque Country, and therefore should not have affected the Basque results. In Sweden, the 2008 recession did not markedly affect the labour market. However, the economic crisis in Europe in early 1990 could have affected the labour markets in Sweden and Spain. The data was collected at the end of the recession; hence, if the results were affected then the effect of the recession on the time use data should be small. Nevertheless, the recession should be considered when examining the results. Health-2011) under Grant 278173 (SOPHIE project) and The Swedish Research Council under Grant 421-2009-2153

Disclosure statement
No potential conflict of interest was reported by the authors.