British and Spanish social policy has been observed to conform to different models in areas such as welfare (Esping-Andersen 1993; 2000), parental leave (Gauthier 2002; Moos and O’Brien 2019; O’Brien and Kowslowski 2016; Escobedo and Wall 2015), childcare (Haas 2003), family (Thevenon 2008) and gender (Fouquet et al. 1999). The two countries also differ in terms of total fertility rate, which is fairly high in the UK and low in Spain, as well as in economic and labour market indicators, such as GDP and unemployment.

Nonetheless, changes in legal provisions on parental leave have followed a similar pattern in the UK (Baird and O’Brien 2015) and Spain (Meil 2013). Recent Spanish and British amendments to parental leave, a tool not only to balance work and family life but also to further gender equality, have much in common, for instance. Similarly, the take-up rate for 2 week paid paternity leave was high in 2010 in both countries, with 74 % of all eligible fathers using the leave in Spain (Romero-Balsas 2012) and 91 % of British fathers taking some time off from work (O’Brien and Kowslowski 2016). Part of maternity leave has also been made ‘transferable’ to fathers in the two countries. These similarities are indicative of a (slow) shift toward gender equality in countries with different gender role traditions and welfare systems. Recent extensions to the duration of paternity leave (to 5 weeks in 2018, 8 weeks in 2019 and, further to a decree-law now in place, to the same 16 weeks as provided for mothers beginning in 2021) in Spain are expediting greater equality in the harmonisation of family and work life. At the same time, legislative measures in favour of gender equality are following a slower course in the UK.

This paper compares the gap between gender attitudes and the distribution of housework and care-giving in the UK and Spain. Basic data drawn from the International Social Survey Programme’s 2002 and 2012 ‘Family and Changing Gender Roles’ module (ISSP 2002; 2012) were applied to assess the relevance of gender values in behavioural changes in two countries with different historical backgrounds and welfare schemes. Whilst prior comparative analyses have addressed the impact of culture on behaviour, restricting the comparison to just two countries put the differences into closer focus. The findings were also compared to similar studies conducted with data from ISSP 2002 (Cunningham 2008; Aboim 2010) to provide further insight into how gender values changed in that 10 year period.

Women’s participation on the labour market has risen in recent decades. Despite the positive correlation between women’s paid work and men’s participation in household chores (OECD 2011), women’s embrace of work outside the home has been greater than men’s within the domestic realm (Bianchi et al. 2000; Holter 2007; Jurado-Guerrero et al. 2012; Meil 1997). Women’s involvement in the labour market has not, then, in itself led to egalitarian sharing of household tasks (Davis and Greenstein 2004). Nor are employment rates anywhere near parity in either the UK or Spain. Further to Eurostat data, 68.4 % of British women between the ages of 20 and 64 and 80 % of the men in that age group were employed in 2012. In Spain, just 54.6 % of the women (and 64.5 % of men) were employed in that year (Eurostat 2019). In addition to labour market structures specific to each country, account must also be taken of the effect of the financial and economic crisis. In Spain it narrowed the gender employment gap (Escobedo and Wall 2015; Flaquer, Moreno and Cano López 2016; Pfau-Effinger 2012), although this was due more to the increase in male rather than female employment (González and Segales, 2014). Moreover, the rise in female activity rates in Spain during the crisis1 may be attributed to the “added-worker effect”, in which women seek (possibly temporary) paid employment in response to their partners’ loss of job or deteriorating working conditions (Addabbo, Rodríguez-Madroño y Gálvez-Muñoz, 2013). In the UK, although male joblessness was higher in the early years of the recession, somewhat later female unemployment climbed as well (Rubery and Rafferty, 2014). The percentage of women with paid jobs also rose during the crisis, although less than in Spain2.

Women’s increasing employment rates in recent decades have lowered support for the male breadwinner model (Cunningham, 2008), ushering in new gender values. Whilst Pfau-Effinger (2015) argued that variations in female employment patterns may be attributable to historical cultural change, public policy may also promote cultural change in some cases. A clear example can be found in the effect of paternity leave policy on fathers’ actual involvement in childcare (Kotsadam and Finseraas 2011; Rege and Solli 2010; Romero-Balsas 2015).

The sole male breadwinner model declined steadily in the UK and Spain between 2001 and 2013, particularly in the latter country, where it fell from 38.8 % of households in 2001 to 26.9 % in 2013 (Connolly et al. 2016). Since the early twenty-first century, significant differences have existed between the two countries in the proportion of households with no-one employed. Whilst jobless households declined from 6 % to 4.8 % in the UK, the percentage in Spain rose sharply from 4.4 % in 2013 to 10.8 % in 2013 (ibid, 2016). The proportion of households in which males had full-time and females part-time jobs was higher in the UK (30.8% in 2013) than in Spain (10.8% in 2013) (ibid, 2016). As observed by Gregory and Milner (2008), working patterns in the UK, with fathers putting in long hours and mothers holding part-time jobs, constitute an obstacle to the egalitarian assumption of childcare responsibilities.

The welfare regimes in place in the UK and Spain differ. The United Kingdom has an essentially liberal political-economic system, i.e., one in which the market plays a greater role in providing welfare and the benefit system is weaker than in the so-called welfare states. Female employment, albeit frequently part-time, is high. Spain in turn has a rather singular familistic welfare regime with weak market presence, social rights largely linked to employment status and low female employment rates (Esping-Andersen, 2000).

Some studies focus on the gender dimension of such welfare states, identifying different care regimes or social structures for the provision of care and their implications for the gender division of productive and reproductive work. Bettio and Plantenga (2004) measured formal and informal care in a number of European countries using an index based on both the number of adults devoting more than 2 hours a day to child - and other dependent care - and the percentage of households not paying for care. They found that whilst the intensity of informal care was similarly high in the UK and Spain, the gender gap in the provision of informal care differed. Of the 14 countries studied, Spain had the sixth highest percentage of female child-carers (as defined above) whereas the UK, with the fourth lowest, was among the countries with a narrower gender gap. Nonetheless, the share of women carers was above 70 % in both countries.

Drawing from time-use data, Gracia and Esping-Andersen (2015) found that while Spanish fathers engaged more actively in routine care-giving than their British counterparts, in relative terms, Spanish mothers shouldered a heavier routine childcare burden than British mothers. The same authors found fathers’ routine care-giving to be more sensitive in Spain than in the UK to whether the mother had paid work (Gracia and Esping-Andersen 2015). Based on ISSP 2002 data, Aboim (2010) identified three attitudes in 15 EU countries: unequal sharing, the ‘familistic’ unequal and the dual/earner carer model. She classified the United Kingdom, where a dim view was taken of full-time working mothers and less emphasis placed on the dual breadwinner domestic man ideal, under the unequal sharing heading. Spain in turn was found to conform to the familistic unequal pattern, which differs from the above only in that it describes a society more supportive of domestic man ideals. Aboim (2010) also studied the gap between couples’ attitudes and practice around paid and unpaid work. She observed values to be significant but dependent upon welfare tradition and female employment. Crompton and Lyonette (2005), likewise using ISSP 2002 data, contended that structural constraints and factors such as age, education, the presence or absence of children and social class were at least as important as attitudes. Those findings call into question Hakim’s well-known and controversial theory that gender attitudes towards work-family balance are the variables most relevant to couples’ actual practice (Hakim 2000).

Much prior research has been conducted on the factors affecting the allocation of household chores (Coltrane 2010), although from varying perspectives. Inspired by Becker’s (1991) treatise, some authors have argued that the division of housework is the result of a bargaining process in which resources, primarily income and level of schooling, condition the power wielded by each partner (for an overview, see Lundberg and Pollak 1996). Others have shown, however, that the situation is more complex and that greater resources do not always entail less housework. Brines (1994) reported that men with lower economic wherewithal than their partners devoted less time to household chores to compensate for the power asymmetry. Bittman et al. (2013), in turn, detected cases of couples who divided housework along traditional lines even though the woman had a higher income than her partner.

Another series of studies has stressed the importance of gender to understanding the distribution of domestic tasks. Several approaches have been adopted to explain that indisputable effect, supported by empirical research. One of the most influential, based on ‘doing gender’, was put forward by West and Zimmerman (1987). This idea questions sex role theories, claiming that gender is not a series of inherent traits but a construct of social interactions in which people constantly engage and build on with daily practice. In that approach gender is viewed as a performative process: every interaction and social situation is gendered, and people tend to act in what they assume to be a gender-appropriate manner (West and Zimmerman 1987: 135), knowing that their acts are going to be judged.

‘Gender appropriate behaviour’, however, what is expected of women and men, is based on cultural values that are undergoing change. For that reason, the ways in which different worldviews and gender ideals affect domestic practice need to be taken into consideration. The possible impact of a given gender value on such practice needs to be evaluated, even where behaviours do not conform to general expectations. As Deutsch (2007) noted, some types of social interaction may involve ‘undoing gender’, i.e., changing traditional gender patterns. Earlier research has found consistencies between beliefs about gender roles and the division of housework. Couples with more egalitarian values tend to divide domestic chores more evenly (Blair and Lichter 1991; Lennon and Rosenfield 1994; Meil 2005).

Gender values and practice do not seem to be related to care-giving in the same way as they are to household tasks, however. Exploring the effects of attitudes towards housework and childcare on the absolute and relative time devoted to these responsibilities in The Netherlands, Poortman and Van der Lippe (2009) found that attitudes were aligned more closely with the division of household chores than with childcare, inferring that the two types of responsibilities are not seen in same light. Egalitarian gender values have also been invoked to explain the use of parental leave by fathers (Romero-Balsas 2012; Haas, Allard and Hwang 2002).

As noted earlier, the UK and Spain have been viewed as conforming to different European public policy models in terms of welfare (Esping-Andersen 1993; 2000), parental leave (Gauthier 2002; Moos and O’Brien 2019; O’Brien and Kowslowski 2016; Escobedo and Wall 2015), childcare (Haas 2003), family (Thevenon 2008) and gender (Bettio and Platenga, 2004; Fouquet et al. 1999). This paper aims to compare conditions in two European countries which, while characterised by traditionally different public policies, are both slowly moving toward the same goal: to involve fathers in childcare and housework. Since sex is a key parameter in the performance of the two activities, the study focuses on the connection between gender values and men’s and women’s actual domestic practice in the two countries.

Risman (2004) proposed that to define gender, it should be conceptualised as a social structure embedded in three intertwined social dimensions: individual, interaction when facing cultural expectations and gender-specific institutional regulations. Deemed as a social structure, gender can be analysed where not only institutions but cultural expectations not perceived as mandatory constitute both external constraints, and an opportunity for change through action.

The research question addressed here was whether those two dimensions, cultural expectations and the impact of institutions, are inducing different outlooks on gender in the two models of welfare state. Most studies note that cultural differences define gender preferences, such as those of fathers around type of care (Lewis and Lamb 2003). Therein lies the interest in comparing gender attitudes and practice in two culturally different contexts - the primary aim of this research. As Risman (2009) noted, gender structures are not static but evolve with each new generation. The dual aim sought here was to determine, firstly, how cultural attitudes toward gender have changed with values over the last 10 years, and secondly, how cultural and institutional factors have impacted domestic practice by identifying the effect of the gender value model on care as practised in the two countries. The idea was to explore whether the doing gender theory is taking root in more traditional groups and whether more egalitarian groups should be interpreted from the undoing gender perspective (Risman, 2009).

Cultural expectations (Risman 2004) were explored on the grounds of the answers to the questions on gender values in connection with paid work and childcare contained in the ISSP (2012) surveys conducted in Spain and the UK. Groups were defined according to their attitudes toward gender equality, from the nearest to the farthest from that ideal. As Risman (2004) noted, gender is a structure and men and women are distributed across a variety of structural positions. Consequently, multivariate models were deployed to control for family and class variables in gender attitudes. A second aim was to ascertain how the groups most culturally favourable to equality in different welfare models behave in practice, i.e., how adherence to more or less egalitarian gender principles affects the time devoted to domestic chores. Doing and undoing gender were explored on the grounds of the time spent by men and women on domestic tasks. Less time spent by women and more by men in such tasks would be a proxy for undoing gender and vice-versa. The aims pursued are set out in the following hypotheses.

The primary hypothesis is that in progressive families with more egalitarian values men devote more and women less time to domestic work than in families with more traditional attitudes. The idea was to verify whether, as Risman (2009) contended, people with more egalitarian gender values engage in undoing gender and those with more traditional values in doing gender. Further to Hakim’s (2000) postulate that preferences constitute the key variable to explaining the allocation of childcare tasks, the contention here is that progressive women spend less time and progressive men more on reproductive work. That hypothesis is generally supported by earlier research and some authors have shown that the partners need to share egalitarian values for housework to be divided equitably (see Coltrane 2010).

H1a: ‘Progressive men devote more time to domestic activities than traditionalists.’

H1b: ‘Progressive women devote less time to domestic activities than traditionalists.’

In response to Aboim’s (2010) findings showing that welfare and employment patterns have a bearing on the impact of gender values and family structure on actual time devoted to housework and childcare, the argument here is that changes in attitudes around the dual earner/dual carer model and developments towards a more egalitarian gender ideal in 2012 should be expected to have established an intermediate level of masculine hegemony in both countries. Moreover, the decline in that decade in acceptance of the male breadwinner model (Connolly et al., 2016) should increase the likelihood that gender values would play a similar role in the UK and Spain in the division of household tasks.

H2: ‘Gender values around domestic chores are moving in the same direction and carry the same weight in the UK and Spain.’

Pursuant to the foregoing, the study first compared gender values in the two countries in 2002 and 2012. Inasmuch as Spain has a stronger familistic tradition than the UK, the shift toward an egalitarian model was expected to entail greater change there. The effect of the economic crisis on reducing the employment gap in Spain (Escobedo and Wall 2015; Flaquer, Moreno and Cano López 2016; Pfau-Effinger 2012) was expected to have spilled over into gender values in the 10 years studied.

H3: ‘The UK and Spain are gradually converging on more equitable gender patterns.’

The research hypotheses were tested against quantitative data drawn from the British and Spanish subsamples of the International Social Survey Programme (ISSP) module, ‘Family and Changing Gender Roles’ for the years 2002 and 2012. The change in gender values in that 10 year period was determined on the grounds of gender value variables.

The first step consisted in applying a descriptive method to ISSP 2002 and 2012 data to determine the trends in egalitarian values present in British and Spanish society in that time period. Separate cluster analyses were subsequently conducted for each country to group respondents by gender values in 2012, with a view to identifying national differences in attitudes toward gender. The possible answers to the questions used to create the clusters were: Strongly agree, Agree, Neither agree nor disagree (UK only), Disagree, and Strongly Disagree. The absence in the Spanish subsample of the option ‘neither disagree nor disagree’ was another reason for running separate cluster analyses for the UK and Spain.

Gender variable values were divided into the three topic groups listed below and the responses were recoded, with 1 as the most traditional and 5 (4 in Spain) as the most egalitarian answer.

After creating gender value groups in the two countries, the divide between values and domestic practice was analysed along sex lines. Multivariate ordinary least-squares regression was used to determine whether egalitarian gender values impacted actual domestic behaviour. Household activity was divided into time spent on housework and on care-giving. The dependent quantitative variables were: 1) On average, how many hours a week do you personally spend on household work, not including childcare and leisure time activities?; and 2) On average, how many hours a week do you spend looking after family members?.

Given that sex is a key parameter in assessing change in domestic practice, separate models were developed for men and women in the Spanish and British subsamples. The explicative variable was the cluster group, for which the nominal categories were: traditionalist, intermediate and progressive. Relevant socio-economic variables were included to control for their possible effect.

The results for each country are discussed in this section under three headings: change in gender attitudes, cluster analysis and multivariate analysis.

This study analysed the gap between gender attitudes and actual domestic practice by sex. It first determined whether the UK and Spain are converging in terms of egalitarian values. As reported by Meil (2013) for Spain and Baird and O’Brien (2015) for the UK, whilst legislative change is underway in both countries, progress is slow.

The effect or otherwise of public policy on parenthood and the division of household tasks is controversial. Some authors deem the public realm to have limited impact on private practice within couples and families (Gershuny and Sullivan 2003; Gregory and Milner 2008). Others contend that institutional and structural factors influence childcare significantly (O’Brien, Brandth and Kvande 2007; Brandth and Kvande 2002). The two countries reviewed here, both post-industrial societies, have moved toward more egalitarian values. The route followed by each, however, was observed to be guided by their respective welfare traditions, as noted by Aboim (2010). Ideal childcare was associated more firmly with non-working mothers in Spain than in the UK, due perhaps to the traditional familism that characterises care practice in Mediterranean countries (Bettio and Platenga, 2004; Moreno, 2007). At the same time, however, support for the dual earner model was greater in Spain, denoting a certain cultural ambivalence around the traditional family ideal.

Gender values affected actual behaviour similarly in the two countries, perhaps because neither is a strong welfare state. Differences would be expected to be wider between strong and weak welfare states, with egalitarian practice more widespread in the former and gender values more likely to affect behaviour in the latter. Risman (2004) contended on the contrary, that inequality may stem from unwitting social expectations through which behaviour is affected less by gender values and more by other structural elements such as public policy.

Further comparative research would be needed to test that hypothesis, however. Egalitarian care practices might be extended more widely and the impact of gender values on such practice mitigated by policies able to further effective equality. Future research might address the hypothesis that the rise in the share of parental leaves reserved to males, in conjunction with a generous replacement income, is reducing the dependence of egalitarian practice on gender values.

Cluster analysis was conducted on data from the ‘Family and Changing Gender Roles’ module of the 2012 International Social Survey Programme (ISSP) to classify British and Spanish respondents by gender attitude to determine whether professing more egalitarian values had a positive impact on equitable sharing of household chores. Multivariate analysis based on ordinary least-squares regression was then performed, using the explicative variables resulting from cluster analysis.

Our primary hypothesis that gender values impact actual domestic practice was partly accepted. Progressive men were observed to spend more and progressive women less time doing housework than their traditionalist counterparts. Gender values were not found to affect the amount of time devoted by either sex to care-giving, however. Further to Aboim (2010) and Hakim (2000), then, gender attitudes impact unpaid work, but only when the chores do not involve care-giving, as reported by Poortman and Van der Lippe (2009). Structural conditions were observed to be at least as important as gender values in domestic practice, as contended by Crompton and Lyonette (2005). Despite Aboim’s (2010) findings, different welfare traditions and employment patterns did not alter the impact of gender values and structural conditions on actual domestic behaviour, at least where, while different, the traditions at issue were not particularly strong.

The limitations of this study should be addressed in subsequent research. Inasmuch as the ISSP does not furnish panel data, the direction of the relationship between egalitarian values and egalitarian behaviour cannot be absolutely determined. Do the former lead to the latter, or the other way around? That is a pertinent issue because, according to social psychology literature, attitudes and actual behaviour are reciprocally reinforcing (Ajzen and Fishbein 1980). Gender beliefs, as the cognitive component of attitudes, influence behaviour. Nonetheless, as Ruddick (1995) contended, routine practice also shapes the way people think. As Deutsch (2007) observed, social interaction may be a site for reproducing gender patterns, but also a place for resisting that tendency. The result is a capacity for gender transgression (on the individual identity scale) and gender transformation (on the social scale). Panel data are imperative to conclusively determining the effect or otherwise of domestic practice on gender values in the UK and Spain.