Abstract
The increase in fathers’ involvement in childcare in Western societies has not translated into an equal distribution of childcare between mothers and fathers. While some couples actually succeed in “undoing gender” when the first child arrives, their characteristics that define them are not yet clear. This article provides four different explanations that contribute to understanding how parents share routine care in Spain, using a sample of dual-earner, heterosexual parents from the Spanish Time Use Survey 2010. The results show that fathers are more likely to be “equal sharers” when they earn less than their spouse; do not hold traditionally masculine jobs (e.g., managers or blue-collar workers); have time available and, more important, their spouses are not available from 5 p.m. onward. In line with previous research, this study provides evidence that dismantles the “myth” that higher status employees, and/or those with college education, are more egalitarian in practice.
Introduction
Despite the widespread ideal of “involved fatherhood” in Western societies, sociological studies have systematically shown that couples usually adopt gender traditional patterns of sharing housework and childcare after having their first child (Craig & Mullan, 2011; González, Domínguez-Folgueras, & Baizán, 2010; Sayer, Bianchi, & Robinson, 2004), particularly with regard to the so called “routine tasks” (both in terms of domestic work and childcare 1 ). However, research has also shown that some couples manage to “thwart the forces” that sustain inequality (Deutsch, 1999, p. 5) and become equal sharers. But it is still not known exactly how many these “equal sharers” are, and what specific characteristics have enabled them to achieve this.
This article seeks to further explore to what extent the practices engaged in by Spanish couples 2 who equally share routine childcare duties can be explained by different sociological theoretical perspectives on men’s contribution to unpaid work and care. This is addressed by using a nationally representative time use survey. Four specific explanations will be tested: bargaining power (derived from the mother’s relative resources, such as earnings or education), time availability, status, and the “need” to share childcare (Deutsch, 1999). The latter refers to a situation in which the mother has restricted time availability.
This analysis aims to contribute to the existing empirical evidence and academic debate in several ways. First, it involves a specific study of parents’ relative contribution to childcare. More specifically, it examines the percentage of routine care duties (i.e., physical care and supervision) provided by the mother alone, the father alone, and either parent in the presence of the other, respectively. Measuring this relative contribution is very significant from a gender perspective, since a general trend toward “intensive parenting” may obscure the real nature and impact of the changes in men’s contribution—given that not only fathers but also mothers are increasing the time they spend with their children. A more equal distribution of childcare contributes to gender equality, both by reducing the differences in the amount of effort that mothers and fathers can devote to their jobs and by exposing children to a richer family environment, in which both their father and their mother take care of them. Second, special attention is paid to those work conditions (more specifically, working time and occupations) that could contribute to equality in parenting practices. Spain is a particularly interesting context to analyze how work constrains limit gender equality and work–life balance, given the high prevalence of long work days (Gracia & Kalmijn, 2016), especially among men. There is strong evidence of an increase in fathers’ involvement in childcare in recent years in Spain, but this has been found particularly among men who became unemployed after the Great Recession (Ajenjo Cosp & García Román, 2014; González et al., 2010; Gutiérrez-Domènech, 2010) than among working fathers. Also, as Sullivan pointed out, The use of occupation as a key independent variable in the study of division of labor is a relatively new development, but its significance is that it acts as a mediating connection between the worlds of the public and the private. (Sullivan, 2004, p. 217)
Third, nationally representative data drawn from the Spanish Time Use Survey (STUS) is used. This is a data set that provides information for all members of the household, allowing for a rich analysis of contextual factors that may influence behavior. Time use surveys are the main instrument used to obtain empirical information about unpaid work and care activities, given that they are more likely to avoid perception bias and are more precise than stylized questions over activities performed (i.e., “How often do you . . .?”). Fourth, the focus is on childcare rather than domestic work, as there are relevant theoretical reasons for these two areas to be analyzed separately. Basically, childcare falls into a “category of experience, which, at least in some of its aspects, is perceived as more rewarding and enjoyable” (Sullivan, 2013, p. 74). According to recent qualitative studies (Domínguez-Folgueras, Jurado-Guerrero, & Botía-Morillas, 2017a), an equal division of routine domestic work after first parenthood mainly depends on standards on housework, men’s proactive attitudes toward domestic involvement and schedule flexibility, and on women’s greater participation in paid work; in contrast, the relative resources hypothesis (except for education) does not seem to account for this “undoing gender” process in any significant way. However, it remains unknown whether these explanations are equally valid for the division of childcare duties. The analysis ultimately focuses on a type of activity that is especially sensitive to equality issues in the domestic sphere: routine solo care provided by fathers and mothers, that is, primary, basic or physical care (e.g., feeding, bathing, or simply supervising children) provided while the other parent is not present. The main objective is to reflect fathers’ actual assumption of responsibility for their children’s needs, that is, responding to basic needs (e.g., changing diapers) and taking decisions over how to dress them, feed them, and so on. Along with interaction and availability, responsibility is one of the three dimensions of fathers’ involvement (Lamb, Pleck, & Levine, 1985). It is of great importance, as it is still mothers who overwhelmingly “organize, plan, orchestrate and worry,” despite fathers’ growing participation in domestic life across different cultures and ethnicities (Doucet, 2006, p.11). Embracing an involved fatherhood ideal prior to childbirth does not automatically translate into assuming real responsibility for caregiving when the child is born (e.g., taking initiative and making arrangements) (González, Lapuerta, Martín-García, & Seiz, 2018, p. 34). Even some fathers who have taken paternity leave 3 alone may be “dependent caregivers,” that is, they feel “overanxious” when taking responsibility for solo caregiving, as research has shown in the case of Spain (Meil, Romero-Balsas, & Rogero-García, 2017, p. 118). Routine caregiving greatly accounts for the lagged adaptation of reality to ideals about “new fatherhood” (Sullivan, 2010), as the gap between mothers and fathers is generally wider than in the case of more recreational childcare activities. In sum, this study aims to illuminate the couple-level and workplace-level conditions that may enhance fathers’ involvement in those tasks in which their presence is less significant. Research has not yet found concluding explanations as to why even engaged fathers rarely assume real responsibility for childcare matters (González et al., 2018).
This article is organized as follows: after the introduction, the second section provides an overview of the Spanish context. The third section presents the theoretical background and the hypotheses. The fourth section introduces the method and data used, while the fifth one discusses the main results. The concluding remarks are contained in the sixth and final section.
Distribution of Childcare Among Parents in Spain
In 2003, the sex ratio of time devoted to childcare by mothers and fathers in Spain was 3.1, higher than in countries or regions such as the United Kingdom, Denmark, and Flanders (Gracia, 2012). However, some recent studies that analyzed macro-level patterns and institutional factors have shown how social change toward gender equality in unpaid work may be accelerating precisely in those Western countries characterized by more traditional gender regimes, that is, Southern European countries. This is probably due to a more rapid process of social diffusion of less traditional gender behaviors in those countries (Altintas & Sullivan, 2017; Sullivan, Billari, & Altintas, 2014). If this change is confirmed over time, it may have important consequences as women in Southern Europe are especially burdened by the overall amount of work they perform, compared to their counterparts in other Western countries (Burda, Hamermesh, & Weil, 2013; Giménez-Nadal & Sevilla, 2014).
The change is starting to be visible. Today’s first-time fathers seem to reject the model of the “absent father” (González, Domínguez-Folgueras, & Luppi, 2013) and show concerns about the sustainability of their work patterns in terms of future childcare demands (Abril et al., 2015). Those fathers who report that they engage in a greater share of routine childcare (stylized question) tend to earn less than others, have more egalitarian gender attitudes, work in family-friendly companies and be partnered with women who work long hours (Fernández-Cornejo, Escot, Del-Pozo, & Castellanos-Serrano, 2016). However, Spanish fathers who, for example, decide to take parental leave alone (as mothers do) still largely receive a “surprised” reaction from their workplaces (Meil et al., 2017). Mothers continue to be more prone to making use of policies that entail being penalized in their careers (especially through a decrease in their income). Therefore, those who opt for unpaid parental leave or reduced work hours are overwhelmingly women (Jurado-Guerrero, Monferrer, Botía, & Abril, 2018).
The absolute time spent on childcare by Spanish fathers has nonetheless increased considerably in recent years, while their paid work time has remained the same (Domínguez-Folgueras, 2015). Changes in the labor market, with a tremendous growth in male unemployment since the beginning of the Great Recession in 2008, 4 may be in part responsible for this. For the first time in Spain, dual-earner couples outnumbered traditional couples (male breadwinner model) between 2003 and 2010, with a growing trend of female breadwinners too (Ajenjo Cosp & García Román, 2014). Unemployed men seem to have contributed more to childcare than working men (González et al., 2010; Gutiérrez-Domènech, 2010) and have even become “equal sharers” in some cases (Ajenjo Cosp & García Román, 2014).
Despite the ongoing social change, it is still to be established to what extent fathers are becoming “equal sharers.” Quantitative evidence on this is still weak, as the difference between time spent on childcare by men and women has remained almost unchanged over the 2002-2010 period. The gap has even slightly increased due to an increase in the number of households with children, as found by Ajenjo Cosp and García Román (2014). According to them, the gender gap in daily time spent in care-related activities had reduced by around 5 minutes from 2003 to 2010 when looking specifically at dual-earner couples, but again this may be due to demographic change. Overall, the absolute time spent on childcare by both men and women increased during this period (in 1.64 vs. 1.57 weekly hours, respectively, for nonretired/nonstudent men and women aged 21-65 years; Giménez-Nadal & Sevilla, 2014). Moreover, as in other countries, it has been confirmed that the gender gap in parental care time is more remarkable in basic nurturing duties (González et al., 2010; Gutiérrez-Domènech, 2010; Sevilla-Sanz, Giménez-Nadal, & Fernández, 2010), particularly in washing and buying children’s clothes, cooking children’s food, organizing housework and childcare, and feeding children. In contrast, the lowest gap has been found in taking children to park, teaching them and playing with them at home (Fernández-Cornejo et al., 2016). Work conditions can be a particularly important hindrance in terms of sharing the responsibility for children in Spain, given the prevalence of long work days and the so-called split-shift (involving a long lunch break and leaving work late) among Spanish fathers (Gracia & Kalmijn, 2016). Long work hours, especially leaving work late in the evening, are particularly costly in terms of father–child time (Gracia & Kalmijn, 2016; Gutiérrez-Domènech, 2010; Nock & Kingston, 1988). A greater emphasis on family would necessarily imply trading off time at work for time at home, especially in the evening and at weekends, when children are not at school. This is the schedule adaptation that mothers usually do. Schedules and time constraints play a great role in explaining childcare involvement and may obscure or reinforce the effect of values, attitudes, and ideology. In sum, fathers are increasingly engaged in childcare in Spain, but they still face structural resistance to an equally shared division of caregiving duties.
Factors That Promote Involved Fatherhood and Shared Childcare
The change to a more involved model of fatherhood started to be seen in the 1980s (Lamb, 2000), although there was little empirical research at the time to show evidence that fathers in dual-earner couples were doing much more than being particularly involved during weekends, at least in the United States (Nock & Kingston, 1988). Overall, men have increased their absolute contribution to childcare and general unpaid work across very different national settings since the 1960s (Altintas & Sullivan, 2017; Hook, 2006). This increase has happened along with an emergence of the ideal of “involved” or “nurturing” fatherhood (Doucet, 2006; Wall & Arnold, 2007) and a growing body of research on fathers’ involvement in childcare (Marsiglio, Amato, Day, & Lamb, 2000), triggered in part by the increased use of time use methodologies (Lamb, 2000). Important theoretical contributions to the field of involved fatherhood have been made by authors such as Pleck and Coltrane in the United States, Lamb in Europe, and Doucet in Canada (Coltrane, 1997; Doucet, 2006; Lamb, 2000; Marsiglio et al., 2000; Pleck, 1977).
The following subsections will present the proposed hypotheses, which are framed within the previous empirical evidence and theory that support each of them. More specifically, this study assesses whether the greater involvement in routine solo care by some fathers in Spain in relative terms (i.e., as a percentage of all routine care provided by both parents) is associated with bargaining power, status, time availability, and/or need, and to what extent the effect of these dimensions is interrelated. Most of the existing quantitative evidence on fathers’ involvement has focused on absolute time rather than on parents’ relative contribution. It is difficult therefore to pinpoint the specific mechanisms that lead to a fairer distribution of childcare among spouses, a relative gap in the literature that this study addresses. In the hypotheses presented below, the term routine solo childcare will be used to refer to the percentage of routine childcare performed by the father alone.
Spouses’ Relative Resources and Bargaining Power
The material or symbolic resources of each spouse may provide them with a greater or lower ability to bargain against undesirable or unprofitable tasks in different ways. The mechanism that explains this is marital power, that is, resources provide potential alternatives to an undesirable/unfair relationship (Demaris & Longmore, 1996). Therefore, one individual’s resources would be ultimately associated with an implicit “threat” of marital breakdown. Depending on the type of resource involved (material or symbolic), this mechanism operates in a different, nuanced way: educational attainment tends to be more directly associated with bargaining “skills” (i.e., persuasion), while occupational status and earnings are usually associated with a more rewarding job (both in economic terms and intellectual or emotional terms). Again, this may be a reason to explicitly or implicitly avoid care or domestic work. In addition, fathers who earn less than their spouses are more likely to take parental leave alone, which can contribute to creating new routines (Meil et al., 2017). The first hypothesis, is as follows:
Socioeconomic Status
Socioeconomic status consists of educational attainment and occupational status, which are considered key determinants for parental involvement (Barbeta-Viñas & Cano, 2017). Educational attainment is widely recognized as being positively associated with more egalitarian gender attitudes (Barbeta-Viñas & Cano, 2017). Ideals over how paid and unpaid work should be divided among men and women are one of the core dimensions of gender ideologies (Grunow, Begall, & Buchler, 2018). As the STUS does not allow gender ideology or attitudes to be measured directly, being college educated could serve as a proxy for an egalitarian gender ideology. Craig and Mullan (2011) found that college educated fathers carried out a higher proportion of routine care (alone or in the presence of the mother) in four Western countries, and, if both parents had a college degree, fathers increased the proportion of childcare (not necessarily routine childcare) provided by them alone. However, the latter was not the case in two countries that share many contextual elements with Spain (Italy and France).
College educated fathers contribute more to childcare and housework (in absolute time) than the rest of fathers in very low-fertility countries (Sullivan et al., 2014). However, other findings on the effect of education on childcare involvement are not so conclusive for the Spanish case (Barbeta-Viñas & Cano, 2017; Fernández-Cornejo et al., 2016; Romero-Balsas, 2015). Again, it cannot be concluded that highly educated fathers share routine childcare in a more egalitarian way, even if their contribution is higher in absolute terms. The following hypothesis will be tested:
With regard to occupational status, this study aims to disentangle if particular occupations are associated with being more or less involved in childcare, and whether or not this effect is mediated by time availability. It not, then the effect could be more clearly associated with how having certain occupations shape differing focuses on work and family, and also with individual identity. In particular, two relevant occupational statuses will be compared: managers and middle-level employees (white-collar workers and technicians).
The relationship between occupational status and involvement in childcare seems complex and has remained relatively unexplored by the literature. High status employees supposedly have more resources to achieve a work–life balance than nonprofessional workers (Gracia & García-Román, 2016) and are presumably particularly concerned about the benefits of spending substantial time in childcare. However, recent empirical evidence challenged the assumption that working-class fathers adopt less egalitarian practices in terms of sharing childcare with their spouses (Abril et al., 2015; Norman, Elliot, & Fagan, 2014). It can therefore be argued that parents with higher occupational status do not necessarily share childcare in a more egalitarian way for at least three reasons. First, higher status couples may be more concerned with the aforementioned ideals of “intensive parenting” (and intensive mothering) and therefore also mothers (and not only fathers) may spend high amounts of time on childcare (Craig & Mullan, 2011). Previous studies have found that the time devoted to care by fathers and by mothers within couples is positively correlated (Romero-Balsas, 2015). Second, as gender equality ideals become mainstream in Western societies, education and class may lose power as a predictor of gender egalitarian attitudes. For example, recent cross-national evidence shows almost residual support for traditional and moderate traditional gender ideologies in eight European countries, including Spain (less than 8% for the population aged 18-45 years;) (Grunow et al., 2018). Third, material constraints and orientation to work (closely related to class and status) may play an important role in determining childcare involvement. Conditions such as being underworked, unable to externalize childcare, or having nonstandard schedules may explain childcare arrangements among spouses to a greater extent than gender attitudes (Deutsch, 1999). With regard to orientation to work, working-class men may be more likely to have an instrumental relationship with their jobs, be somehow “disenchanted” with the labor market, and be more willing to be involved in family activities (Domínguez-Folgueras, Jurado-Guerrero, Botía-Morillas, & Amigot-Leache, 2017b; Seiz et al., 2016). Managers are not only strongly work oriented but also usually integrated in organizational cultures where deviating from traditional masculinity may be more penalized (Abril et al., 2015). Both holding leadership positions (Gasser, 2015) and having high earnings (Fernández-Cornejo et al., 2016) are negatively correlated with time devoted to childcare. In contrast, mid-level white-collar employees (e.g., clerical workers) are more likely to have family-friendly schedules than other employees and seem to be less affected by the penalties associated with breaking the long hours culture (Fernández-Lozano, 2018a, 2018b). For all these reasons, this study proposes that occupational status is not positively related to equal practices at home in a linear way, due to the complex interaction between gender and socioeconomic structures. Therefore, it is hypothesized that
Spouses’ Time Availability
Much of the explanations based around jobs or resources may in fact be related to time availability. Men who work longer hours are more likely to have more resources (i.e., bargaining power) than their spouses, to embrace the traditional “male breadwinner” role, which involves being less family-oriented and more work-oriented. In her classic study on equal sharers in the United States, Deutsch found that time (or lack thereof) was the explicit reason that couples frequently used to explain why the mother was the primary caregiver, even if work time did not tell “the whole story” (Deutsch, 1999, p. 126), as gender shapes decisions about careers that will ultimately affect parenting practices. In fact, she found that parents who shared childcare on a 50-50 basis worked very similar hours.
The extent to which work schedules help explain the time devoted to childcare varies greatly from country to country. Paid work hours are associated with less time devoted to childcare in Australia, the United Kingdom, and the United States, but not in Norway and France (Sayer & Gornick, 2012), and fathers with nonstandard schedules seem to devote more absolute time to childcare in the United States (Wight, Raley, & Bianchi, 2008), but not in Australia (Craig & Powell, 2011).
This study proposes that time availability (i.e., time not spent on paid work) influences the model of fatherhood that Spanish fathers adopt. It is also proposed that in the specific time range when they are available is significant. In particular, three time segments will be analyzed: from 8 a.m. to 5 p.m. from Monday to Friday, after 5 p.m. from Monday to Friday, and on weekends, which will make it possible to compare the effect of standard versus nonstandard work hours. What has been called a “tight time schedule” (Fernández-Lozano, 2018a; Jurado-Guerrero et al., 2018), that is, one which concentrates work time within standard office hours, that is, 8 a.m. to 5 p.m., could be associated not only with more time availability in the afternoon and evening but also with an alternative role to that of the traditional “male breadwinner” who spends long hours at work. Additionally, research has found that fathers with nonstandard schedules seem to be more involved with their children in some countries, like the United States (Wight et al., 2008). It is therefore proposed that:
The last dimension analyzed is need, which refers to the mother’s time availability. If the mother has little time availability, the father is forced to assume more responsibility for childcare. Men tend to be more involved in the household when their spouse is not at home (Presser, 1994). Craig and Mullan (2011) found that the increase in fathers’ proportion of routine care was higher in the case of dual full-time earners than in the case of fathers not employed full-time, which highlights the importance of the mother working full time. In the case of Spain, mothers’ longer work week has been confirmed to be one of the most important factors (along with egalitarian gender attitudes) associated with fathers’ greater share in routine childcare (Fernández-Cornejo et al., 2016). This happened in part because fathers take longer parental leaves when mothers work longer hours. Also, if the mother works longer hours than the father, she is more likely to be more strongly focused on work or attached to her job, which has proved to be associated with more equal patterns of sharing domestic chores (Domínguez-Folgueras et al., 2017a). Again, it would be possible to compare the effect of standard vs. non-standard work hours. The sixth hypothesis proposes the following:
Finally, it is hypothesized that all the theoretical explanations proposed are closely interrelated. For example, mothers who earn more than their spouses are also more likely to have less available time. This study tries to assess to what extent the effect of relative resources and occupational status is mediated by time constraints: time availability and needs. The final hypotheses are as follows:
Table 1 summarizes the hypotheses proposed and the variables associated with them.
Summary of Hypotheses and Variables on Equal Share Parenting.
Note. M-F = Monday–Friday.
Data and Method
The data used in this research have been drawn from the latest STUS (2009-2010), which is an official, nationally representative time use survey carried out by the Spanish National Statistics Institute. The original databases include data for 9,541 households and 25,895 individuals. The STUS is a diary-based survey that logs the activities performed by all the individuals in a household (aged 10 years or older) for a 24-hour period (household activity diary) and the paid work time for a whole week (work diary). A subsample was selected of dual-earner couples (i.e., who completed their work diaries) living with at least one child aged 12 years or younger, for whom there were no missing data in either the independent or dependent variables (DVs). The final sample (n = 655) was relatively small for two main reasons. First, the number of dual-earner couples with children in Spain is relatively small, due to low fertility rates. Second, those couples whose members did not report performing any routine childcare on a given day were excluded (as there was no information for the DV). The unit of analysis were couples, but the hypotheses and analysis focused on fathers’ characteristics because this was considered particularly relevant in terms of achieving equality.
The DVs combine two very important dimensions of fathers’ involvement in childcare from a gender perspective (Craig & Mullan, 2011): “solo care” (understood as the care performed when the other parent is not present) and “routine care.” Routine care includes the activities “physical care” and “supervising children” contained in the STUS. Following the procedure used by Craig and Mullan (2011), three interrelated DVs are used in the models: fathers’ routine solo care (DV1), mothers’ routine solo care (DV2), and routine care performed by either parent (or both simultaneously) in the presence of the other parent (DV3, or “family care”), as described in Figure 1. The three DVs are expressed as percentages of the total routine care performed by parents, so they take values ranging from 0 to 100. While the hypotheses will be tested against DV1 (fathers’ routine solo care), the other two variables are informative of how care provided by fathers, mothers, and family care are interrelated: an increase in fathers’ solo care (DV1) resulting from a particular variable could be directly related to a decrease in mothers’ solo care (DV2) and/or “family care” (DV3). The three regressions within one model can be thus read both vertically and also horizontally (the three values for the constants in the regressions add up to 100 and correspond to the distribution of childcare for the reference category).
Operationalization of the dependent variables (DV1, DV2, DV3).
Also following Craig and Mullan (2011), the distinction between “solo care” and “family care” is based on the place in which both members of the couple report to be, instead of looking only at the “with whom” 5 question reported by one of them, which may introduce more ambiguity. This study also identifies the situations in which the mother is not available to “enter the scene” and assist the father, 6 rather than focusing on those when the task is performed by both parents simultaneously (which does not make so much sense in the case of routine activities such as changing diapers). Table 2 summarizes how “family care” and “solo care” have been operationalized.
Operationalization of “Solo Care” and “Family Care” (Togetherness).
Note. For example, in the case both spouses report being with the other, but none of them specified the place they were in, they were considered as being together.
Includes missing data.
Independent variables have been operationalized as follows. Three variables are associated with bargaining power: economic hypogamy (the mother earns more than the father), educational hypogamy (the mother has a higher level of educational attainment than the father), and occupational hypogamy (the mother is a manager or a professional, while the father is not). As for status, the variable college refers to the father having attained a university degree. With regard to the occupation, the STUS provided a two-digit classification based on the National Classification of Occupations (CNO-94). This has been grouped into six relevant categories:
▪ Managers (CNO-94 Code 1), for example, “Chief Executives, Senior Officials, and Legislators.”
▪ Professionals (CNO-94 Code 2), for example, “Medical Doctors” and “University and Higher Education Teachers.”
▪ Mid-level employees (technicians and clerical workers: CNO-94 Codes 3 and 4), includes employees such as “Medical and Pharmaceutical Technicians” and “Secretaries.”
▪ Services and sales workers, including the armed forces (CNO-94 Codes 0 and 5). It includes “Salespersons,” “Bartenders,” and “Hairdressers,” among others.
▪ Blue-collar workers (CNO-94 Codes 6 to 8: skilled agricultural, forestry and fishery workers; craft and related trades workers, and plant and machine operators and assemblers), for example, “Field Crop and Vegetable Growers,” “House Builders,” and “Miners and Quarriers.”
▪ Elementary occupations (CNO-94 Code 9), such as “Domestic Cleaners and Helpers” and “Garbage and Recycling Collectors.”
The time availability and need dimensions refer to the time spent away from work by the father and the mother, respectively. These variables were derived from the (paid) work diary questionnaire, which showed when the respondents were working for a whole week by 15-minute time lapses. While the DVs were based on time spent doing childcare on a given day, this may have been influenced by weekly schedules (e.g., compensation effects may appear, or on the contrary, long work schedules may be associated with a medium-term pattern of little involvement at home/high involvement at work). A threefold classification of this time availability was used, according to three relevant time lapses (i.e. those associated with standard or nonstandard schedules):
▪ Weekly available hours from 8 a.m. to 5 p.m., from Monday to Friday (more frequently associated with nonstandard schedules).
▪ Weekly available hours after 5 p.m., from Monday to Friday (more frequently associated with standard schedules, and particularly, “tight time schedules”).
▪ Weekly available hours on weekends (associated with standard schedules).
These three variables were introduced as continuous variables (i.e., number of weekly available hours).
Other relevant variables were included as control variables. The age of the youngest child was also introduced, as a way of controlling for the demands of routine childcare in the family (younger children require more physical care), which is usually assumed by the mother. The number of minutes that the father spent doing domestic tasks was introduced as a continuous variable aimed to capture possible trade-offs between childcare and housework. Alternatively, this variable could be a proxy for gender egalitarian attitudes and beliefs. Although it can be argued that this may not be a very precise indicator for gender ideology (as some men may do less domestic work just because they have less time available), it seems implausible that true believers in gender inequality (Risman, 2017) spend a lot of time doing domestic chores. 7 A dummy variable captured a free or vacation day as reported by the father (32% of the sample diaries corresponded to Saturdays and Sundays) in order to control for more time available on that particular day. The father’s age was also introduced as a continuous variable. Finally, another dummy variable controlled for external (formal or informal) childcare (for children aged 10 years and younger) used by the family, excluding compulsory education. For reasons of space, only significant associations will be discussed for control variables. Multicollinearity between independent and control variables has been discarded by carrying out variance inflation factors for the regressors (estat vif command in Stata 9).
The hierarchical linear modelling statistical technique is used to gradually incorporate the different sets of in DVs. Hierarchical models allow the effect of different sets of variables and the interrelation among them to be isolated, as is the case, for example, when new variables absorb the significant effect of variables in the previous model. In particular, Model 1 aims to assess the effect of bargaining power and status; Model 2 adds needs to Model 1; Model 3 includes bargaining power, status and time availability; and Model 4 contains all variables. As this study specifically focuses on the separate effects of the fathers’ and mothers’ time constraints (Hypotheses 7 and 8), two separate models (Models 2 and 3) incorporates the two sets of variables. All models included the control variables. As the DVs are continuous variables ranging from 0 to 100, and due to its more straightforward interpretation (Craig & Powell, 2011), ordinary least squares regression models are used. Comparative descriptive results from the 2003 STUS and 2010 STUS will be presented before discussing the regression results.
Results
Descriptive Results
Figure 2 shows a comparison of the time that fathers and mothers spent on different forms of childcare in 2003 and 2010, respectively. As can be seen, mothers spent more time doing activities more closely associated with daily routines (such as routine care or picking children up from school), while fathers and mothers spent a similar amount of time doing intellectual stimulation–related activities. On an average day, around 32% of couples shared routine solo care (at least) on a 40-60 basis (see Figure 3).
Daily minutes spent by fathers and mothers in different types of care (STUS 2003-2010).
Distribution of couples according to the father’s contribution to total solo routine childcare on a particular day (STUS 2010).
Table 3 shows the descriptive statistics for the independent and control variables included in the multivariable analyses. Table 4 and Table 5 show the results of the ordinary least squares regression models. These results will be analyzed in the following sections.
Descriptive Statistics for Independent and Control Variables (Weighted; STUS 2010).
Note. STUS = Spanish Time Use Survey; M-F = Monday–Friday; DV = dependent variable.
OLS Results for the Percentage of “Routine Care” (Models 1 and 2; STUS 2010).
Note. OLS = ordinary least squares; STUS = Spanish Time Use Survey; DV = dependent variable; M-F = Monday–Friday; DV = dependent variable. Sample of dual-earner parents living in a couple, with at least one child younger than 13 years, who reported doing some routine care. Controls: age of youngest child, daily minutes of domestic tasks by father, free/vacation day, father’s age, and family receives external childcare.
p < .1. *p < .05. **p < .01. ***p < .001.
OLS Results for the Percentage of “Routine Care” (Models 3 and 4; STUS 2010).
Note. OLS = ordinary least squares; STUS = Spanish Time Use Survey; DV = dependent variable; M-F = Monday–Friday; DV = dependent variable. Sample of dual-earner parents living in a couple with at least one child younger than 13 years who reported doing some routine care. Controls: age of youngest child, daily minutes of domestic tasks by father, free/vacation day, father’s age, and family receives external childcare.
p < .1. *p < .05. **p < .01. ***p < .001.
Bargaining Power, Education, and Jobs
The first hypothesis proposed that relative resources could be associated with more routine childcare performed by fathers, as they may give women the opportunity to strike a bargain out of doing the lion’s share of routine care. According to the data analyzed, this hypothesis (Hypothesis 1) was partially confirmed: only the mother’s greater earnings seemed to be significantly associated with the father’s higher proportion of routine solo care (8-10 percentage points). Important, this is true even after controlling for their time availability, which means that the effect of earnings did not hide the effect of the mother’s longer work days. With regard to status, first it was proposed that college-educated fathers would perform a higher proportion of routine solo care. This was not found to be the case and Hypothesis 2, as to whether these fathers actually had a more egalitarian gender ideology than fathers with a lower educational attainment, could not be confirmed with these data. It was then proposed that fathers in managerial positions would be less involved in routine childcare, even after controlling for their time availability (Hypothesis 3). This hypothesis was confirmed; however, when controlling for all the variables (Model 4), the effect was only significant at p < .1. It cannot be confirmed that fathers in mid-level occupations were the most relatively involved, as those working in services and sales showed higher coefficients, but the differences were not significant. Consequently, Hypothesis 4 could not be confirmed. Fathers’ contribution to routine childcare was found to be higher (in descending order, and net of the effect of time available for mothers and fathers) for services and sales employees, elementary occupations, mid-level employees, professionals, blue-collar workers, and managers (although not all differences were significant at p < .05). Managers and blue-collar workers are those jobs in which traditionally male patterns operate more clearly. It could be concluded, then, that jobs matter in terms of practicing gender equality at home, beyond time constraints. Fathers whose jobs were not closely related to the masculine ideal of power, status, competitiveness, and physical effort were more likely to be equal sharers, as they might experience a lower identity conflict when deviating from traditional masculinity at home.
The data also clearly showed that workers at the bottom of the occupational scale (those in elementary occupations and services and sales) performed less “family care” (along with those who had less time available in nonstandard hours). These may be couples who have desynchronized schedules more often, that is, they are more likely to work at different times during the day, either due to externally imposed work conditions or as the result of a conscious decision taken to minimize the need for paid external childcare (see Carriero, Ghysels, & Van Klaveren, 2009, on couples schedule coordination in different countries).
Mothers’ and Fathers’ Time Availability
Models 2 and 3 show the effect of the mothers’ and the fathers’ time availability on an independent basis. Coefficients should be interpreted as the effect on the percentage of solo routine childcare performed by each weekly available hour. As a result, each coefficient should be multiplied by 5 to assess the effect of a Monday to Friday pattern. The data clearly showed that the most significant effect was that associated with the mother’s time availability after 5 p.m. This means that the greatest contribution to equality would be the fact that the mother was at work after 5 p.m., as this increased the father’s relative participation (and decreased the mother’s). Thus, Hypothesis 6 was confirmed. The fathers’ time availability also contributed to equality in a significant way, even net of the effect of the mothers’ time availability, as shown in Models 2 and 4. The fathers’ time availability on weekdays (especially after 5 p.m.) significantly increased their contribution to solo routine childcare. Therefore, Hypothesis 5 was also confirmed. Also, the fathers’ time availability on weekend significantly reduced the mothers’ contribution.
It was also proposed that when controlling for the fathers’ and the mothers’ time availability, some of the other significant associations would lose their effect. In the light of the analyses carried out, these hypotheses cannot be confirmed, as the coefficients only changed slightly. Only the effect for managerial occupations lost predictive power when controlling for both the fathers’ and the mothers’ time availability. This suggests that fathers in managerial positions may be less involved because they have particularly long work days and their spouses do not. However, it must be recalled that the direction of causality cannot be confirmed using cross-sectional data (men who are less likely to engage in childcare may choose more demanding jobs and their spouses may respond to this fact by spending more time on care duties). In contrast, the effect of having a blue-collar occupation or economic hypogamy remained significant and continued to be very similar.
To conclude, it is worth noting that Model 3 had greater predicting power than Model 2, as the R2 and adjusted R2 coefficients suggested. This means that the mother’s time availability was more important than the father’s when it came to explaining how families distributed routine care. Also, the coefficients in Model 4 seem to indicate that both sets of variables (the mother’s and the father’s time availability) reinforced each other’s effect rather than absorbed it. This may be the case because the time availability of each of the spouses proved to be positively correlated (the longer the mother worked, the longer the father worked too).
Other Factors Associated to Fathers’ Involvement in Routine Care
As expected, the lower the age of the youngest child, the more traditional the division of childcare among spouses. The time the father devoted to domestic tasks was clearly associated with a lower percentage of solo care performed by the mother and more care performed in the presence of both parents. This analysis cannot confirm, however, that men who perform more domestic work also spend more time on childcare duties—it could be that these men are at home more often and therefore have commensurately higher levels of “family care.” On the days when fathers did not work, mothers’ solo care was traded for family care, while fathers’ solo care was not affected.
Conclusions
This study has sought to investigate how four different theoretical explanations could help understand the way in which Spanish heterosexual, dual-earner couples distribute routine childcare (e.g., feeding or bathing children), namely, bargaining power, status, time availability, and need, paying special attention to the father’s role. Ultimately, it aimed at identifying the main characteristics of “equal sharers”—those fathers who assume responsibility (i.e., decision taking) for the most demanding childcare tasks, as well as those couples who engage in these tasks on an equal basis. Routine solo care provided by fathers is relevant from a gender perspective not only because it can contribute to alleviate the “extra burden” that mothers have regarding childcare responsibility but also because it promotes alternative forms of masculinity. Co-responsibility in childcare promotes gender equality in the labor market in at least two ways (Fernández-Cornejo et al., 2016). First, women are better able to focus on their career if they are alleviated from the “extra burden” of doing all or most of the unpaid work. Second, women do not lose opportunities of being hired or promoted due to statistical discrimination (i.e., employers’ assumption that women are the main caregivers and are therefore less reliable employees). Also, taking care of children alone promotes an intense emotional bonding between father and child (Meil et al., 2017). From the perspective of a child’s well-being, father–child time has proved to be associated with children’s better cognitive outcomes (Cano, Perales, & Baxter, 2018). Children who are taken care of by both parents would be exposed to more diverse stimuli. Equal sharers are not yet in the majority among families with young children, but they represent a model with positive consequences for men, women, and children.
These analyses have been based on time use data from the STUS. A sample of 655 heterosexual, dual-earner couples were selected who were living with at least one child aged 12 years or younger. The analytical strategy proposed by Craig and Mullan (2011) was followed, whereby total care was divided into three interrelated DVs, presented as three percentages that added up to 100. Therefore, the father’s involvement in childcare was analyzed in relative terms to the mother’s. This research also suggested an innovative way of operationalizing spouses’ “togetherness” using time use data.
The analysis showed that “equal sharers” had a number of features. First, irrespective of how much time the mother worked, it was found that her income was greater than the father’s. This evidence pointed to the importance of economic resources in terms of the bargaining power between spouses, although alternative explanations may be plausible (e.g., that men with lower income are more family-oriented). Second, the father worked in a mixed or female-dominated occupation: in services and sales, as a clerical employee or a technician, or, to a lesser extent, as a professional. Managers and blue-collar workers were less involved in childcare in relative terms. Causation may run in both directions in this case, however. On the one hand, work cultures may create the conditions for a more traditional male behavior. On the other hand, occupational self-selection may also be at play here (i.e., less family-oriented men choose typically male occupations). Third, the father is available for their children, especially in nonstandard work hours (i.e., after 5 p.m.), which is consistent with previous research (Gracia & Kalmijn, 2016; Gutiérrez-Domènech, 2010; Nock & Kingston, 1988). This evidence is important in policy terms, as it points to the fact that the schedules of school/childcare centers and work schedules are not synchronized, and so far, it has been mothers who have had to adapt to this misalignment. The signaling effect to which mothers are subject (when they request a schedule adaptation to be able to pick up their children from school) can be avoided if work schedules are generally adapted in all organizations where working 9 a.m. to 5 p.m. is feasible. Last, and very important, the mother was not available after 5 p.m. due to work commitments. This variable has a direct impact on both the father’s and the mother’s proportion of routine solo care. This is a very significant finding, as it points to the fact that men may become more involved out of necessity. At the same time this involvement may foster women’s careers, as mothers can devote more time to their jobs (again, reverse causation cannot be ruled out). At the interactional level, mothers who work long hours, especially in nonstandard hours, are probably required to explicitly or implicitly defend the importance of their work to themselves or to the family, and to engage in complex intracouple negotiations.
In line with previous research, this study confirmed that the father’s relative involvement in childcare was associated with his time availability, but even more as to whether the mother was not at home in particularly demanding moments (the mother’s time availability had more predicting power than the father’s). This research has also contributed to a better understanding of the relationship between family practices and socioeconomic status. Three main conclusions have emerged here. First, this study has presented evidence against the “myth” that gender equality goes hand in hand with socioeconomic status or educational attainment. College-educated fathers are not particularly involved in routine solo care in comparison with their spouses. Second, fathers in services and sales occupations perform a greater percentage of routine care alone, while those who are managers tend to engage in childcare less, even when accounting for their time availability. This means that jobs are important predictors of how couples distribute childcare, irrespective of time constraints. Jobs are important sources of personal identity, and it seems that the traditional male identity (i.e., high orientation to work, breadwinner role, high authority, high autonomy, etc.) may be in conflict with adopting a nurturing role at home. In sum, this research suggests that the “male breadwinner model” may act as a “package” of behaviors (high earnings, long work days, traditionally male occupations), which would be associated with the lowest relative involvement in childcare. The behavior of more egalitarian fathers is in opposition with this model in several dimensions. Finally, this study has also provided evidence that lower status jobs (e.g., sales and services or elementary occupations), as well as nonstandard schedules, seem to be particularly costly in terms of “family care” as defined in this research, probably due to desynchronized schedules.
With regard to the limitations of this study and the prospects for future research, it is worth mentioning that distributing childcare equally is an abstract notion that can take very different forms, as it does not inform specifically about the amount (and quality) of childcare. In this sense, parents who conform to the norm of intensive parenting could be described as egalitarian, as would those who receive significant external support with childcare and spend very little (but equal) time with their children. In addition, this research has focused on solo care, leaving “family care” out of the analysis. It must also be acknowledged that time spent on childcare, even on routine childcare, does not translate automatically into autonomous decision taking or responsibility for childcare matters, as qualitative studies have shown (González et al., 2018). Future research could explore further other possible gendered uses of time beyond work schedules (e.g., time devoted to leisure activities by fathers and mothers) and if this is one of the possible explanations as to why Spanish mothers are overburdened in terms of total—both paid and unpaid—work. The most problematic issue with time use data is probably endogeneity. The causal relationship between different time allocations (e.g., home vs. work) cannot be confirmed, as, for example, more family-oriented men may choose to work shorter hours. What can be confirmed, however, is that time is a limited resource, which requires individuals to make both sacrifices and choices.
In this sense, it is relevant to emphasize the importance of time devoted to paid work in terms of gender equality, both in the public and the private spheres. Recent research has found that men’s average work hours are important to understand men’s relative contribution to household chores, even when controlling for the individual’s work hours (Dotti Sani, 2014). This suggests that when family-friendly schedules are normalized among men and women across the board, the whole of society receives a message in favor of a more gender-balanced distribution of unpaid work and care. For this change to take place in all countries, fathers need to push forward social change at the organizational level by demanding to be socially recognized as caregiving fathers, while mothers leave space at home for them to “pick up the slack.”
Footnotes
Acknowledgements
I have to thank Killian Mullan for providing me with very enriching insights on how to understand spouses’ “togetherness” in relation to childcare and to operationalize it using time use data, pointing out the differences among perceptions of “where” individuals are and those of “with whom” they are. I would also like to thank Teresa Jurado-Guerrero for very useful feedback on this article.
Declaration of Conflicting Interests
The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.
Funding
The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: This work was supported by the research project CSO2014-58754-P and the grant BES-2015-07427 to hire doctoral candidates, funded by the Spanish Ministry of Economy.
