American Sociological Review-2011-Craig-834-61 more often than do fathers (Craig 2006a; Fuligni and...

http://asr.sagepub.com/American Sociological Review

http://asr.sagepub.com/content/76/6/834The online version of this article can be found at:

DOI: 10.1177/0003122411427673

2011 76: 834American Sociological ReviewLyn Craig and Killian Mullan

How Mothers and Fathers Share Childcare : A Cross-National Time-Use Comparison

Published by:

http://www.sagepublications.com

On behalf of:

American Sociological Association

can be found at:American Sociological ReviewAdditional services and information for

http://asr.sagepub.com/cgi/alertsEmail Alerts:

http://asr.sagepub.com/subscriptionsSubscriptions:

http://www.sagepub.com/journalsReprints.navReprints:

http://www.sagepub.com/journalsPermissions.navPermissions:

What is This?

- Dec 1, 2011Version of Record >>

at CAL STATE UNIV SACRAMENTO on December 5, 2011asr.sagepub.comDownloaded from

American Sociological Review76(6) 834 –861© American Sociological Association 2011DOI: 10.1177/0003122411427673http://asr.sagepub.com

As maternal employment increased over the past half-century, scholars expected mothers and fathers would more equally share child-care responsibilities (Bergmann 2005). Move-ment in this direction, however, has been slow and uneven (Bianchi, Robinson, and Milkie 2006). Mothers’ workforce participa-tion has not substantially reduced time they spend looking after their children, and most mothers remain responsible for the bulk of family care (Sayer and Gornick 2011). Work and care are divided more equally in some households than in others, which may reflect parents’ different values, resources, and opportunities (Coltrane 2007). Structural fac-tors are also at play. Gendered division of

labor varies cross-nationally, which some research attributes to policy differences between countries as well as cultural and social norms about gender and family care (Gornick and Meyers 2009; Lewis 2009).

Little research, however, has examined cross-national differences in how childcare is shared between spouses within households,

427673 ASRXXX10.1177/0003122411427673Craig and MullanAmerican Sociological Review

aUniversity of New South Wales

Corresponding Author:Lyn Craig, Social Policy Research Centre, G2 Western Campus, University of New South Wales, Kensington, Sydney, Australia 2052 E-mail: [email protected]

How Mothers and Fathers Share Childcare: A Cross-National Time-Use Comparison

Lyn Craiga and Killian Mullana

AbstractIn most families today, childcare remains divided unequally between fathers and mothers. Scholars argue that persistence of the gendered division of childcare is due to multiple causes, including values about gender and family, disparities in paid work, class, and social context. It is likely that all of these factors interact, but to date researchers have not explored such interactions. To address this gap, we analyze nationally representative time-use data from Australia, Denmark, France, and Italy. These countries have different employment patterns, social and family policies, and cultural attitudes toward parenting and gender equality. Using data from matched married couples, we conduct a cross-national study of mothers’ and fathers’ relative time in childcare, divided along dimensions of task (i.e., routine versus non-routine activities) and co-presence (i.e., caring for children together as a couple versus caring solo). Results show that mothers’ and fathers’ work arrangements and education relate modestly to shares of childcare, and this relationship differs across countries. We find cross-national variation in whether more equal shares result from the behavior of mothers, fathers, or both spouses. Results illustrate the relevance of social context in accentuating or minimizing the impact of individual- and household-level characteristics.

Keywordscomparative analysis, cross-national comparison, fathers’ childcare, gender, mothers’ childcare, shares of childcare, time use


Craig and Mullan 835

and none has looked at the extent to which fathers contribute across the range of activi-ties that make up childcare. Research on individual countries finds that mothers spend more time than fathers caring for children, and that they undertake different types of care. That is, mothers undertake more routine childcare, more physical care, and they care solo more often than do fathers (Craig 2006a; Fuligni and Brooks-Gunn 2004). This prompts the question of whether differences in types of care undertaken by mothers and fathers are consistent cross-nationally, or whether some countries promote more equal participation across the board. Furthermore, to what extent do country and individual and household characteristics interact? Specifically, how do mothers’ and fathers’ work arrangements relate to their shares of childcare, and do these relationships differ across countries? To explore these questions, we conduct a cross-national comparison of mothers’ and fathers’ relative time in childcare. We differentiate between routine and non-routine activities, and between performing care together versus performing care solo.

GENDER DIVISION OF PARENTAL CHILDCAREIncreasing levels of maternal employment over the past five decades have not resulted in more equitable gender distribution of house-work and childcare time (Bianchi and Milkie 2010). Gender specialization in market and non-market work remains ubiquitous, with many women withdrawing from paid work upon becoming mothers, and many house-holds with young children composed of a male breadwinner and a female homemaker. In many other households, mothers work part time and retain primary responsibility for managing the home (Gornick and Meyers 2009). Gender specialization is least pro-nounced when both spouses are employed full time, but even in these households, women generally do most of the housework and childcare (Hook 2010).

A considerable body of literature addresses persistence of the gendered division of labor, much of which focuses on three possible causes: limited time availability, unequal rela-tive resources, and conformity to traditional gender ideology (Brines 1994; Coltrane 2000; Greenstein 2000). These factors generate pre-dictions about how couples choose to divide their responsibilities for market and non-mar-ket work. For example, some scholars suggest that the spouse with higher earning capacity will do the most market work, and the spouse who does the most market work will do the least non-market work (Becker 1981; Lund-berg and Pollack 1996). Additionally, some scholars theorize that couples with conserva-tive gender ideology will likely form male breadwinner households; and in couples with progressive attitudes, mothers and fathers will both work full time (Torr and Short 2004). Evidence in support of each of these explanations is not strong, however, not least because they are inextricably linked (Coltrane 2007). Labor markets yield unequal earnings for men and women, which reflects and influ-ences couples’ decisions on who should com-mit more time to market work. Labor markets may encourage women to specialize in unpaid work and men in paid work. Moreover, mar-kets affect mothers’ and fathers’ relative resources and household bargaining power, reinforcing attitudes about appropriate gender roles (Bittman et al. 2003; Ferree 2010; McDowell and Fang 2007).

These explanations are particularly uncon-vincing in regard to the division of childcare—they are not irrelevant, but neither are they adequate. Employed mothers do not trade-off time in paid work against childcare on a one-for-one basis (Bittman, Craig, and Folbre 2004; Hofferth 2001). Unlike housework, which goes down as paid work hours go up, mothers maintain their childcare time by cut-ting back on their own leisure, personal care, and sleep (Craig 2007b). This suggests that employed mothers are unwilling or unable to delegate their childcare duties, even as they commit more time to market work, thus time


836 American Sociological Review 76(6)

availability effects on maternal childcare are reduced. Similarly, mothers with more resources do not necessarily negotiate out of childcare, which contradicts the premise of bargaining theories (Brines 1994; Lundberg and Pollack 1996). For example, highly edu-cated women, who by extension have greater earning capacity, average more childcare time than do less educated women (Craig 2006b; Folbre 2007).

Indeed, parents with more individual resources may use them to bargain into, rather than out of, childcare (Deding and Lausten 2006). Working parents may feel morally compelled to spend as much time as possible with their children, so the children do not miss out on parental attention (Warner 2005). This compulsion is likely gendered, because beliefs about gender equality and parenting roles do not necessarily converge: women who are willing to claim equality in market work and housework time may not wish to make equivalent claims about childcare (Craig 2007a). In challenging the view that mothers universally welcome more equal divisions of care, research suggests that some mothers actively prevent fathers from doing more with children (Allen and Hawkins 1999). Mothers limit fathers’ care through gate-keeping and hampering their involve-ment. Mothers might do this because they wish to retain control of a domain they feel expert in, or they do not trust fathers to deliver as high a standard of care as they themselves provide (Bianchi and Milkie 2010). There is a complex knot of emotion and practicality at the core of the gendered division in childcare. This speaks to the attachment and responsibility many women feel for their children, and subjective diffi-culty they experience in minimizing time devoted to their care.

Yet, time-use studies show that men have increased their participation in childcare in recent decades (Bianchi et al. 2000; Craig, Mullan, and Blaxland 2010). At the same time, mothers have also increased their aver-age care time, suggesting that in addition to gender equality becoming a more widespread

social goal, ideas of what constitutes proper parenting have also intensified (Hays 1998). Compared with times past, parents are now involved in more aspects of their children’s lives, such as their education and friendships, and are less likely to let them play unsuper-vised. Scholars argue that expectations of mothers, in particular, have intensified, cul-minating in an ideology of intensive mother-ing in which childrearing is “child-centred, expert-guided, emotionally absorbing, labor-intensive and financially expensive” (Hays 1998:54). Heightened expectations of paren-tal involvement may mean that the ideal of dividing care equally between mothers and fathers takes a back seat to ensuring children do not miss out on attention.

Further complicating the picture is that parenting values and behaviors intersect with class and socioeconomic factors, including education. For example, highly educated mothers and fathers devote more time to childcare than do less educated parents, par-ticularly interactive and educational activities thought to foster children’s human capital development (Craig 2006b; Deding and Lausten 2006). This might reflect differences in ideas about inputs that children need from parents to succeed in life, with intensive par-enting most prevalent among the middle classes (Vincent and Ball 2007). Research also finds that highly educated people are more likely to have progressive gender atti-tudes, to be in dual-earner households, and to share housework more equally than are peo-ple with lower levels of educational attain-ment (Baxter, Hewitt, and Western 2005). This implies that parents who have the least free time, highest earning capacity, and most progressive gender attitudes do the most intensive childcare, perhaps because they subscribe more whole-heartedly to current expert parenting advice to “concertedly culti-vate” their children (Lareau 2000).

Although parents’ values, constraints, opportunities, and resources all inform how they divide paid work and intra-household care, relationships among these factors are complicated (Ferree 2010). Ideas and practices



relating to family care and gender equality are subject to multiple, sometimes contradictory, influences, including the broader social con-text within which individuals and families are nested.

Social Influences on the Division of Care

Attitudes about raising children develop within a social context, and understandings about the proper care of children are, to a large extent, collectively produced (Duncan and Edwards 2003). Individual and house-hold ideas about family and gender are medi-ated through cultural norms and values and are reflexively influenced by workplace structures and by economic and family policy (Gornick and Meyers 2009). The picture is complex and intersections multidirectional (Ferree 2010; McDowell and Fang 2007). Teasing out relationships is not straightfor-ward, but comparative research offers a win-dow into how social context may interact with family processes to shape divisions of care.

Systematic cross-national differences exist in women’s workforce participation and in how households allocate time to paid and unpaid work (Lewis 2009; Orloff 2009). A substantial body of literature argues that par-ticular policy frameworks encourage particu-lar configurations of household labor supply. In Nordic social democracies, for example, the dual full-time earner (DFT) couple is common among families with young chil-dren. In Southern European countries, includ-ing Italy, the male breadwinner (MBW) family is prevalent. Australia has a high pre-ponderance of households in which men work full time and women work part time (1.5-earner) (Craig and Mullan 2010; Cromp-ton 2006; Gornick and Meyers 2003). These patterns suggest that couples’ decisions about how much time mothers and fathers should spend in market work are at least partly struc-tured by social context. To some extent, this is practical; policies such as publicly funded childcare, mandated parental leave, and the

right to request shorter work hours make it easier for both mothers and fathers of small children to work. Furthermore, policies reflect and shape norms about the proper care of children (Hook 2010; Morgan 2005). For example, when free high-quality childcare is universally available, full-time home-based motherhood is less likely regarded as indis-pensable to children’s well-being, as opposed to countries where childcare is unregulated, expensive, and hard to access (Sayer and Gor-nick 2011). Arrangements that couples strike about paid work and care thus reflect and reinforce the social organization of childcare (McDowell and Fang 2007).

However, the relationship is not always compatible or predictable. For example, social ideals of intensive parenting and expectations of maternal workforce participa-tion can co-exist. Also, the social organiza-tion of childcare within countries may be out of step with some parents’ preferences, so not all will actually use the policy provisions available to them. Another question is whether the social organization of care and household employment arrangements it encourages only affect mothers or whether these things also influence fathers’ involve-ment in childcare. We must also consider what sharing childcare actually entails. Child-care encompasses a range of tasks, done under a range of conditions. To be meaning-ful, definitions of shared care should capture more than simply whether parents spend a similar amount of time in childcare: a more nuanced approach is necessary.

Dimensions of Sharing Care

Sharing care can occur on multiple dimen-sions. As Hook (2010) points out in relation to housework, specialization occurs not only between paid and unpaid work but within these categories as well. Gender equality in the amount of housework can mask important differences in tasks (Hook 2010). The same is true of childcare. The detail is important, because gender differences in the type of care provided have implications for gender equity,



which compounds from disparity in overall time spent in care. A much higher proportion of fathers’ than of mothers’ care time is spent in educational and recreational activities than in routine physical and logistical tasks, and much more of fathers’ care time is done when the mother is also present (Craig 2006a). If men’s care consists mainly of recreational activities performed alongside their spouses, it means that childcare is a more intensive activity for women, and that it is more of a family leisure experience for fathers (Shaw 2008). This also means fathers are not substi-tuting for mothers’ time, so mothers are not freed up for other pursuits. Furthermore, some types of childcare activity are more of a barrier to workforce participation than others. Having to do routine tasks at certain times of the day has more of an effect on one’s ability to work full-time hours than do non-routine activities like talking or playing with chil-dren, which fit around parents’ schedules (Craig 2006a).

Even if mothers are in the workforce, being disproportionately responsible for rou-tine care is demanding and constraining. Hav-ing to leave work at a certain time to pick up children from daycare, and then feed, bathe, and put them to bed, not only means a long day, but it also precludes, for example, stay-ing late to finish an assignment or network with colleagues. Like part-time work (Rubery, Smith, and Fagan 1999), mothers’ lack of flexibility in scheduling home duties likely limits career advancement and promotion. Evidence also shows that mothers regard only certain types of father-care as constituting meaningful sharing—participating in routine care activities and caring for children without a spouse present are most associated with higher female perceptions of shared care (Fuligni and Brooks-Gunn 2004).

Involving fathers in childcare and sharing care time more equally between men and women is not the same thing. Fathers can inten-sify their family participation without neces-sarily lightening women’s loads. Conversely, the gender division of childcare could be made more equal by mothers using out-of-home

substitute childcare, without altering fathers’ childcare at all. To move toward gender equality in childcare requires more than sim-ply getting fathers to increase time they allo-cate overall; fathers must increase the routine care they provide and take on a larger share of childcare as the solo caregiver. That is, gender equality in care requires that fathers alter composition of their childcare time and increase routine and solo childcare relative to mothers.

Identifying factors associated with sharing care time across these multiple dimensions is necessary to understand how greater gender equity in childcare could be promoted more broadly. As we have argued, division of child-care is influenced by values about gender and family, household divisions of paid work, social norms, class, and policy frameworks, all of which interact with each other. To explore the effect of these factors and some intersections between them, we conduct a comparative study of mothers’ and fathers’ relative time in childcare, divided along dimensions of task (i.e., routine versus non-routine activities) and co-presence (i.e., car-ing for children together as a couple versus caring solo).

Social Context

We compare mothers’ and fathers’ childcare in countries with different employment pat-terns, social and family policies, and cultural attitudes to parenting and gender equality. Table 1 summarizes select institutional fea-tures of each country. These institutional fea-tures, while not determinative, are an important influence upon couples’ care divi-sion. For example, Denmark has high public expenditure on childcare, high childcare use, and high maternal employment rates, particu-larly for mothers of younger children. These features are fairly typical of Scandinavia, which enjoys a reputation for being woman-friendly (Borchorst 2008). Raising children is viewed as a shared social concern as well as a personal responsibility; policy rhetoric and public discourse promote gender equality in



paid and unpaid work and father involvement in care (Morgan 2003). Differences exist within Scandinavia, but Nordic men are reputed to be particularly hands-on fathers (Hobson 2002). Denmark has state-funded gender-neutral parental leave, and most moth-ers return to full-time employment after their leave period (OECD 2007). This institutional framework fosters the dual-earner family type, and in keeping with its reputation, Denmark has a more equal gender division of overall, total childcare time than do other countries in our study (Craig and Mullan 2010). We do not know, however, whether this means that Danish men also share routine and independent solo care more than fathers elsewhere.

In France, too, state services relieve fami-lies of much of the direct burden of childcare (Morgan 2003). Average working hours are kept low by regulation, which should mini-mize work-family strain (Fagnani and Let-ablier 2004). Universal state-funded early childhood education and care (ECEC) services

are available for children starting at age three. French parenting norms do not privilege par-ent-child time over adult-only leisure, and average time with children is low in world terms (Craig and Mullan 2010; Sayer and Gornick 2011). Also, in contrast to Scandina-via, French family policy is not aimed at promoting gender equality in paid and unpaid work, encouraging fathers’ greater care of children, or challenging the gender order (White 2009; Windebank 2001). Masculine roles remain traditional (Almqvist 2008). We therefore do not expect French men to per-form a large share of routine and sole care of children; movement toward more equal shares of care would be driven by mothers doing less rather than by fathers doing more.

The normative view that young children need constant and sustained parental attention is often prevalent in Anglo countries, espe-cially among the middle classes (Lareau 2000; Vincent and Ball 2007; Wall 2010). In contrast to Denmark and France, work-family reconciliation and raising children are seen

Table 1. Description of the Social Context in Australia, Denmark, Italy, and France

Australia Denmark Italy France

Maternal Employment Rates 2007a (%) Youngest child under 2 years ND 71 50 57 Youngest child 3 to 5 years 48 78 53 73Female Part-Time Employment (%) 39 24 30 23Paid Maternity Leave 2006/07 Weeks of paid (FTE) maternity leave 0 18 16 16 Percent rate of allowanceb NA 100 76.2 100Public Expenditure on Formal Childcare Percent GDP 2005 .18 .85 .15 .36Enrollment Rates For Formal Childcare/

Early Education 2006 (%) Children 0 to 2 years 25 70 29 43 Children 3 to 5 years 60 91 99 100Total Fertility Rates 2006 1.81 1.85 1.35 1.98Gender Wage Gap in Median Earnings of

Full-Time Employees 200717.0 11.0 16.7 12.0

Global Gender Gap Index 2007c .72 .75 .65 .68

Note: ND = no data; NA = not applicable.Source: OECD Family Database; Global Gender Gap Report 2007.a2005 for Australia; 1999 for Denmark.b“Rate of allowance” is defined as the ratio between the full-time equivalent payment and the corresponding entitlement in number of weeks.cGlobal Gender Gap Index covers a range from 0 = complete inequality to 1 = complete equality (for details, see Hausmann, Tyson, and Zahidi 2007).



as largely private issues outside the respon-sibility of the state (Haas 2003). The United States, for example, has no national paid parental leave policy, nor any policies aimed at involving fathers in childcare in intact families (Wisensale 2003). Similarly in Aus-tralia, most formal ECEC is purchased pri-vately (although it is highly regulated and there are some state subsidies). Australia has seen an active debate about whether non-parental care is harmful to children (Baxter et al. 2007). Full-time paid working hours are long, and a high proportion of mothers are employed part time. This means mothers’ and fathers’ average employment time diverges markedly, and most Australian families with young children are 1.5-earner households (in which fathers work full time and mothers work part time). This occurs despite generally progressive gender attitudes, high educational attainment among young women, and legisla-tion enshrining equal opportunity in the workplace (OECD 2007). Intensive, hands-on parenting is normative and care time is high for mothers and fathers (Craig and Mullan 2010). We are interested in exploring whether this translates to more gender equality in rela-tive shares across the range of care tasks.

In Italy, childcare is usually provided by families rather than by the state or private substitute caregivers (Arts and Gelissen 2002). There are regional differences in gen-der role attitudes, but overall, attitudes are traditional, particularly regarding masculinity and men’s family roles (Miller 2004). Fertil-ity rates are low, which some attribute to Ita-ly’s extremely marked gender inequality in paid and unpaid labor (de Laat and Sevilla Sanz 2007). Few childcare places exist for children younger than 3 years, fees are high, and daily hours are short (OECD 2007). Parental leave is provided in accordance with European Union directives, but average maternal employment is low (see Table 1). Most Italian households with children have a single male breadwinner (Miller 2004). We therefore expect that average shares of care will be unequal by gender. Among some young Italians, however, a dual full-time

earner household arrangement is beginning to emerge (Mancini and Pasqua 2011). Because these households are going against a particu-larly strong prevalence for traditional gender roles, we expect their shares of care to differ from the average more markedly than in other countries.

RESEARCH FOCUSWe fill a gap in the literature by studying how childcare is shared in households with differ-ent employment configurations across four countries. We identify independent effects of employment configuration and country, but our key interest is in the interaction between them. Most people expect childcare to be shared least equally in MBW households, and most equally in households where both spouses work, even more so when both spouses work full time (as opposed to when one is part time). Of the countries we study, researchers usually expect gender divisions of childcare will be most equal in Denmark and least equal in Italy. These expectations over-lap to some extent, because in Denmark most young families are DFT and in Italy most young families are MBW. However, if cou-ples go against the norm, are household-level work arrangements or country the dominant influence on how they share childcare? Specifically, how do mothers’ and fathers’ work arrangements relate to their shares of childcare and does this relationship differ across counties?

We also advance research by analyzing mothers’ and fathers’ time to see how childcare is actually shared between them. We construct ratio measures of household care from time diaries of matched couples. This is an impor-tant contribution because previous research on unpaid work shares relies either on time diary records of unrelated mothers and fathers (e.g., Hook 2006; Sayer, Bianchi, and Robinson 2004) or on one spouse’s estimation, which may be subject to reporting bias (e.g., Fuwa 2004; Geist 2005). Our measures improve on these approaches because they are based on records of the actual time allocation of men



and women in the same household. Addition-ally, measures focus on childcare, not house-work and childcare combined, and we distinguish between different dimensions of childcare. Specifically, we examine relative participation in routine and non-routine activi-ties, and in caring for children together as a couple or solo. A further original contribution is in identifying whether greater gender equal-ity on these dimensions is driven by mothers, fathers, or both spouses, and whether this dif-fers by household employment configuration, parents’ education, or country.

METHODData

We used data from the Australian Time Use Survey 2006 (AUSTUS; Australia Bureau of Statistics 2008), the Danish Time Use Survey 2002 (DTUS; Danish National Institute for Social Research 2002), the Italian Time Use Survey 2002 to 2003 (ITUS; National Institute of Statistics 2007), and the French Time Use Survey 1999 (FTUS; National Institute of Statistics and Economic Studies 1999). We chose surveys from years as close in time as possible. Each survey contains nationally representative samples of the respective pop-ulations of each country, although sample sizes vary widely. Response rates ranged from a low of 49 percent (Denmark) to a high of 92 percent (Italy). Research finds that non-response presents no significant bias in DTUS (Bonke 2005), and estimates weighted for non-response are broadly comparable with unweighted estimates (Abraham, Maitland, and Bianchi 2006). All surveys collected time-use information using a time-budget diary instrument completed by respondents. This method provides the most reliable and comparable estimates of time use (Robinson and Godbey 1997). In all surveys, multiple members of sampled households were required to provide data, so we can analyze men and women living together as couples and derive individual- and household-level results simultaneously.

Sample

We drew a sample of married couples with at least one child age 0 to 12 years and at least one parent who reported doing childcare in the diary day. Table 2 gives an overview of sample characteristics. Household employ-ment configurations differed systematically across countries, indicating that couples’ choices about work and care were associated with country. The most common household employment configuration in Denmark was the DFT (66 percent); in Australia it was the 1.5-earner (42 percent); and in Italy it was the MBW (41 percent). In each case, these per-centages were much higher than in the other countries. By comparison, France had more diversity in household employment configu-ration, with a particularly large proportion (25 percent) of households in which fathers did not work full time (FNFT).

In regard to education, Italy had the lowest proportion (6 percent) of households in which both parents had a college degree. In Den-mark, France, and Australia, both members of sampled couples had college degrees in 24, 18, and 16 percent of households, respec-tively. Italian households were most likely to have only one child. The possibility that selection into parenthood differs systemati-cally across countries should be considered when interpreting results.

Dependent Variables

We sought to compare parents’ relative time in childcare by dimensions of task (i.e., rou-tine versus non-routine activities) and co-presence (i.e., caring for children together as a couple versus caring solo). To do this, we created a set of eight ratio measures as depen-dent variables, which were designed to cap-ture shares of household total care provided by mothers and fathers, disaggregated by activity and who was present when childcare was performed. We first created a set of measures capturing absolute care time, which we then used to create our dependent ratio measures.



First, we computed total minutes per day mothers and fathers spent performing care as a main or primary activity. We summed these to create total childcare in a household. Second, we disaggregated mothers’ and fathers’ total childcare into two categories by activity type: (1) talk-based care—face-to- face parent-child interaction that includes talking, listening, teaching, helping children learn, reading, telling stories, and playing games, and (2) physical care and accom-panying a child—face-to-face parent-child interaction that includes feeding, bathing, dressing, putting children to sleep, carrying, holding, cuddling, hugging, soothing, trans-portation to school, visits, sports training, music and ballet lessons, parents and teacher nights, time spent waiting and meeting trains or buses, ensuring their safety, and handing

them over to substitute caregivers. Activities categorized as physical care and accom-panying a child include the most laborious care activities and those that need to be done regularly or according to a timetable. We refer to this category as routine care; for the former, we use the terms non-routine care and talk-based care interchangeably.

We therefore have the absolute amount of routine and non-routine care for mothers and fathers, which together sum to total house-hold childcare time. We divided measures of mothers’ and fathers’ routine and non-routine care by total household childcare time to cre-ate four ratio measures (which sum to one) of household care activity. This approach pro-vides a more integrated picture than previous research does, because together the four measures show the ratio of a household’s total

Table 2. Proportions of Individual- and Household-Level Characteristics by Country

Australia (household)

Denmark (household)

Italy (household)

France (household)

N 1,283 582 3,248 1,203Employment Status (%) Male breadwinner (MBW) .26 .19 .41 .27 One-and-a-half earner (1.5) .43 .12 .17 .22 Dual full-time earner (DFT) .20 .65 .36 .26 Father not employed full time (FNFT) .12 .04 .06 .25Degree Status (%) No parent has a degree .63 .48 .83 .63 Both parents have a degree .16 .24 .06 .18 Only father has a degree .08 .10 .05 .09 Only mother has a degree .13 .17 .06 .10 Not known .01 .00 .00 .00Number of Children Age 0 to 12 Years (%) One child .43 .47 .56 .44 Two children .41 .41 .38 .42 More than two children .16 .12 .07 .14Age of Children Youngest child 0 to 4 years .52 .45 .48 .46 Youngest child 5 to 12 years .48 .55 .52 .54 No teenagers in household .80 .70 .75 .69 Teenager in household .20 .30 .25 .31Day Type (%) Weekday .58 .54 .38 .77 Weekend .42 .46 .62 .23

Average Minutes 236 187 174 156



routine care to a household’s total non-routine care and fathers’ share of routine care relative to mothers’. This means our multivariate analyses can be interpreted by reading coef-ficients and models together as well as sepa-rately. In each set of models, the intercept row sums to one, and an increase in coefficients in one model implies decreases in others in the same row.

Then, we again disaggregated mothers’ and fathers’ total primary childcare into two categories, this time by whether care was performed in the company of a spouse, that is, by co-presence. We used information respond-ents provided to survey questions asking “where were you [when doing this activity]?” and “who were you with [when doing this activity]?” We used both sources to create robust measures of the time parents were together and time they were apart while doing childcare. This separates out time fathers looked after children on their own, thus dis-tinguishing taking over from joining in. Spe-cifically, we calculated (1) time childcare was performed in the presence of a spouse and (2) time childcare was performed when the spouse was not present at the same location. This provided four measures covering the amount of time mothers and fathers spent in care away from and together with their spouse (which again makes up total household care). As above, we divided these four measures of time by total household care time to create four ratio measures (in this case regarding co-presence) that sum to one. Again, this offers a more comprehensive picture than previous research does, because the four measures show the ratio of a household’s total childcare performed solo to a household’s total care done together and fathers’ share of solo care relative to mothers’. Recall that each set of multivariate analyses can be interpreted by reading coefficients and models together as well as separately.

To summarize, our dependent variables fall into two groups each containing four ratio measures. The first contains the four ratio measures of mothers’ and fathers’ routine and non-routine care as a proportion of total

household care. The second contains the four ratio measures of mothers’ and fathers’ care solo and in the company of their spouse as a proportion of total household care. Table 3 shows means and standard deviations for all eight ratio measures for all households that reported performing care on the diary day.

Analysis Plan

We begin with a descriptive overview of abso-lute time mothers and fathers spent on childcare in each country. We then estimate a series of OLS regression models to test how routine and non-routine care, and care jointly with a spouse or solo, were shared between spouses within households. Our key independent variables are household employment configuration, country, and interactions between them.

We entered dummy variables for different household employment configurations (i.e., male breadwinner/female homemaker (MBW), father full time/mother part time (1.5-earner), dual full-time earner (DFT), and households in which fathers were not employed full time (FNFT); MBW omitted). FNFT households include dual part-timers, female breadwin-ners, and both spouses not employed. Unfor-tunately, the category was too small to analyze these groups separately, but it yielded a broad indication of whether fathers who were not employed full-time transferred time to child-care. These variables simultaneously capture spouses’ time availability and tap an indirect indicator of their relative contribution to household income. Within single countries, they also give an indirect indicator of parents’ gender ideology. MBW households are viewed as most traditional, and DFT house-holds as most progressive (Hook 2010). We recognize that these variables do not isolate independent effects of time availability, rela-tive resources, and gender ideology, but the earlier discussion suggests these factors are not mutually exclusive and it is impossible to disentangle them (Coltrane 2007). The earlier discussion also leads us to expect that house-hold employment configuration is likely only a partial explanation of shares of care.



Household employment configurations differed systematically across countries (see Table 2), confirming that couples’ choices about how to divide paid work intersect with social context. To explore whether these intersections also relate to the gendered divi-sion of childcare, we entered dummy varia-bles for each country and interactions between household employment configuration and country. This allows us to layer our investiga-tion of childcare shares within each type of household within each country, so we can examine whether social context promotes or discourages certain types of household care arrangements. We can establish, for example, whether childcare is shared differently in DFT households in Denmark, where DFT is the statistical norm, compared with DFT households in Australia, where it is unusual. We thus capture some of the complexity that results from the fact that different policy frameworks encourage different degrees of inequality in paid and unpaid work (McCall 2005), and we retain individual- and house-hold-level information. Australia is the omit-ted category because, of the countries we can subject to multivariate analysis, it has the fewest statutory provisions for non-parental care and ECEC and the highest average household care (Craig and Mullan 2010).

Another potential influence on care compo-sition and shares of care is education. Educa-tion likely operates at multiple levels that may intersect, compound, or be at odds (Craig 2006b). Qualifications enhance earning

capacity and thus strengthen personal resources, which may improve relative bar-gaining power. However, education is also a partial marker of class, and couples are likely to have similar levels of education, so there may be no relative advantage to each spouse. Also, higher education is associated with more employment for women and with progressive gender attitudes. Greater education predicts higher childcare time for mothers and fathers, particularly in talk-based activities. Together, these factors suggest that education likely has more effect on the amount and composition of household care provided by both parents in total than on intra-household shares of care. We capture couple and individual education levels by entering dummy variables for differ-ent household combinations of educational status: neither parent has a college degree (yes = 0), both parents have a college degree (yes = 1), only father has a college degree (yes = 1), and only mother has a college degree (yes = 1). We experimented with interactions between these education variables and household employment configuration and found no sig-nificant results. We also experimented with interactions between education and country. Broadly speaking, the effect of education on shares of care was constant across countries, so we did not include these interactions in the models presented here, but instead note the few significant exceptions.

Because we are interested in how parents share care, not in the amount of care they do, we control for absolute total hours households

Table 3. Descriptive Statistics for Mothers’ and Fathers’ Shares of Routine and Non-routine Household Childcare and Childcare Performed Away From or Together With Spouse

Share of Routine Household Care

Share of Non- routine House-

hold Care

Share of House-hold Care

Performed Away from Spouse

Share of House-hold Care Per-

formed Together with Spouse

Total House-hold Care Fathers Mothers Fathers Mothers Fathers Mothers Fathers Mothers

Mean .15 .51 .13 .21 .10 .38 .18 .34 1.00SD .23 .32 .21 .26 .21 .34 .23 .29



spent in childcare. Outcome variables are thus net of time differences in overall amount of care, so the influence of differing national averages and varying use of non-parental childcare is to some extent controlled, and we can focus on relative shares and care compo-sition. We controlled for number of children (1 = 0), because fathers may take a greater share of care and widen their repertoire of care activities when there are more children in the family (Craig and Bittman 2008). We controlled for age of the youngest child (under 5 years = 0, 5 to 12 years = 1), because fathers may take a greater share in care as children grow beyond infancy and toddler-hood (Craig and Sawrikar 2009). We con-trolled for presence of a teenager in the household (no = 0) because teens can assist with care of younger children (Kendig and Bianchi 2008) and thus alter the balance of care between spouses. Because weekends offer different opportunities for sharing care than does the work week (Yeung et al. 2001), we entered a dummy variable for day of the week (weekend day = 1).

We recognize that this study has a number of limitations, including the small number of countries compared, and that there are multi-ple complexities our data do not capture. For example, lack of comparable data on income and wealth means we could not explore class differences on these dimensions. Also, the models implicitly assume that the effect of control variables is constant across countries. To confirm this, we ran our models separately for each country (results available on request). We do not discuss control variables in the text due to space constraint and because the results were as expected. We estimated models for mothers and fathers separately, and because our measures are mutually exclusive we can simultaneously analyze effects of independ-ent variables on mothers’ and fathers’ shares of household care. That is, reading across each set of models together shows whether household adjustments toward more equal shares result from fathers doing more, moth-ers doing less, or the actions of both spouses. To directly test gender gaps in outcome

variables, we also ran interactive models on pooled observations of mothers and fathers. Tables A1 and A2 in the Appendix report results from these models.

Our estimation sample for mothers and fathers is 6,302. This is the total sample of household observations (6,316) minus 14 observations with missing education informa-tion (in Australia and France). Using OLS to model proportions (ratios) of time one engages in childcare is potentially problem-atic because the dependent variable is con-strained to take values between 0 and 1. OLS may predict values outside of this range, so the Fractional Logit (FL) model is an alterna-tive (Buis 2006; Papke and Wooldridge 1996). We estimated OLS and FL models. Results show little substantive difference, so we report OLS findings because they are easier to interpret (results from FL models are avail-able upon request). Also, we examined the sensitivity of our results to a violation of the assumption that observations are independent arising from pooling data across countries by estimating Feasible Generalized Least Squares (FGLS) models specified to allow the error term to differ across countries (Cam-eron and Miller 2010). Substantive results from FGLS and OLS are identical.

RESULTSDescriptive Results

Figure 1 shows the mean daily time house-holds spent in childcare in each country. Average total parental childcare time was highest in Australia and lower in Denmark, followed by Italy and then France. This con-firms previous research and reflects the dif-ferent policy structures and norms regarding raising children in each country (Craig and Mullan 2010). Consistent with a more perva-sive ideology of involved, hands-on parent-ing and sparser work-family policy structures, Australia had the highest average family care. Lower average childcare time in Italy, France, and Denmark is consistent with greater use and social acceptability of ECEC



(especially for children older than 3 years), combined in the latter two countries with higher average maternal workforce participa-tion (see Table 1).

Our main interest here is not in the absolute time differences resulting from each country’s social organization of childcare, but the gen-dered division of that childcare. In all four countries, mothers spent more time performing childcare than did fathers, with fathers averag-ing between 35 percent (Denmark) and 25 per-cent (France) of household care time. In Australia and Italy, fathers averaged about 30 percent of household care. In absolute terms, fathers’ average care was highest in Australia (suggesting greater father involvement) and lowest in France and Italy (suggesting that men adhere to traditional masculine roles in these countries), but the shares of care tell a slightly different story. Relatively speaking, Italy and Australia had similar shares of care. France, where men take on the least proportional load, was the outlier.

Figure 2 shows the mean daily time house-holds with different employment configura-tions spent caring for their children, averaged across countries. Average total parental child-care time was highest in MBW households, followed by 1.5 and FNFT households, and lowest in DFT households. Again, our interest is not in the absolute amount of childcare per-formed, but in the gendered division of that childcare. We see that most variation across households is in mothers’ time, not fathers’.

Even fathers in FNFT households, who might be expected to have the most time available for childcare, differed little from fathers in other household types. This means that due to differ-ences between mothers, shares were most equal in DFT and least equal in MBW house-holds, on average, with the other two house-hold types falling in between. Although variation in mothers’ care time was likely far less than variation in their paid work hours, these descriptive findings are in line with expectations that specialization will be most extreme when mothers are not employed, and least extreme when they are employed full time (Craig 2007b; Hook 2010). However, these findings did not distinguish between dif-ferent types of childcare, or between parents with different characteristics. To decompose composition of care and ascertain how these average findings intersect with individual and couple characteristics, and especially the inter-action between household employment con-figuration and country, we turn now to multivariate analysis.

Multivariate Results

Table 4 shows OLS results for the proportion of routine and non-routine household care performed by mothers and by fathers. Table 5 shows OLS results for the proportion of household care performed by mothers and by fathers away from their spouse and in the presence of their spouse.

0

50

100

150

200

250


Min

utes

per

Day

Fathers' Total Childcare Mothers' Total Childcare

Figure 1. Mean Daily Minutes Spent in Childcare by Fathers and Mothers in Four Countries



In each table, intercepts for mothers and fathers sum across, adding up to 100 percent of household care. We thus have an overview of households’ care composition in two dimensions, activity and co-presence. We also have each spouse’s proportional contribution to the household total of each dimension of care. Coefficients in each set of models should be read together, because an increase in one implies a decrease in others. As an example, suppose mothers’ share of routine household care was 10 percent lower when a house-hold’s youngest child was older than 5 years than when the youngest child was under five. The interpretation here is simply that moth-ers’ share of routine household care is posi-tively associated with the presence of a young child. Our analysis goes further, because implications for gender equity can be explored by comparing results across models. For example, mothers’ 10 percent decrease could be offset by a 10 percent increase in fathers’ share of routine household care. Alternatively, there could be a reallocation by mothers themselves from routine to non-routine care, or some combination of both these explana-tions. Our approach allows us to differentiate between these outcomes and identify which adjustments are made, and by whom.

The reference group in all models is a mother/father in an Australian MBW house-hold in which neither parent has a college

degree, there is one child age 0 to 4 years, no teenage children, and observed on a weekday. Reading across intercepts in the top row of Table 4 shows that mothers and fathers in the reference category performed 63 and 5 per-cent, respectively, of total routine household care. These households devoted 68 percent of their total care to routine activities, with mothers contributing more than 12 times as much as fathers did. Continuing to read across, we see that mothers and fathers in the reference category performed 24 and 8 per-cent, respectively, of total non-routine (talk-based) household care. These households devoted 32 percent of their total care to non-routine activities, with mothers contributing three times as much as fathers. This under-scores a profound gender inequity in total care: the widest gender gaps are in tasks that must be done regularly to a timetable, are less flexible, and, arguably, are less enjoyable than the talking, listening, reading, and play activities that constitute non-routine care.

Reading across intercepts in the top row of Table 5 shows that mothers and fathers in the reference category performed 54 and 2.5 per-cent, respectively, of household total care done solo. In these households, 56.5 percent of total care was done away from one’s spouse. Mothers contributed more than 18 times as much of the household care done solo as did fathers. Continuing to read across,

0

50

100

150

200

250

MaleBreadwinner

1.5-Earner Dual Full Time Father NotEmployed Full

Time

Min

utes

per

Day

Fathers' Total Childcare Mothers' Total Childcare

Figure 2. Mean Daily Minutes Spent in Childcare by Fathers and Mothers in Four Household Types



Table 4. Coefficients and Standard Errors from OLS Regression Models Predicting Fathers’ and Mothers’ Shares of Routine and Non-routine Household Total Childcare

Routine Childcare Non-routine Childcare

Fathers Mothers Fathers Mothers

Intercept .05!!! .63!!! .08!!! .24!!! (.02) (.02) (.01) (.02)One-and-a-half-earner (1.5E) .05!! !.06! .03! !.02 (.02) (.02) (.01) (.02)Dual full-time earner (DFT) .09!!! !.03 .01 !.07!!! (.02) (.03) (.02) (.02)Father not employed full time

(FNFT).06!!

(.02)!.07!(.03)

.05!(.02)

!.04(.03)

Denmark .07!! !.02 !.02 !.03 (.02) (.03) (.02) (.03)Italy !.01 .03 .02 !.04! (.01) (.02) (.01) (.02)France !.03 .17!!! !.03 !.11!!! (.02) (.02) (.02) (.02)1.5E x Denmark !.06 .06 !.04 .03 (.04) (.05) (.04) (.04)1.5E x Italy .00 !.00 !.02 .02 (.02) (.03) (.02) (.02)1.5E x France .01 !.03 !.03 .05 (.02) (.03) (.02) (.03)DFT x Denmark !.05 !.00 .01 .04 (.03) (.04) (.03) (.03)DFT x Italy !.01 !.05 .01 .05! (.02) (.03) (.02) (.02)DFT x France .04 !.14!!! !.00 .10!!! (.03) (.04) (.02) (.03)FNFT x Denmark !.02 .02 .01 !.01 (.06) (.08) (.05) (.06)FNFT x Italy .04 .03 !.06! !.01 (.03) (.04) (.03) (.03)FNFT x France !.01 .02 !.06! .05 (.03) (.04) (.03) (.03)Both have degree .06!!! !.06!!! .01 !.01 (.01) (.01) (.01) (.01)Only father has degree .05!!! !.06!!! .01 .00 (.01) (.02) (.01) (.01)Only mother has degree .00 .00 !.01 .00 (.01) (.01) (.01) (.01)Two children .01 .02* !.01! !.02! (.01) (.01) (.01) (.01)Three or more children .01 .06!!! !.03!! !.04!! (.01) (.01) (.01) (.01)Youngest child 5 to 12 years .01 !.10!!! .02!!! .07!!! (.01) (.01) (.01) (.01)Teenager in family .03!!! .00 !.03!!! .00 (.01) (.01) (.01) (.01)Weekend .02!! !.06!!! .05!!! !.01 (.01) (.01) (.01) (.01)Total household care .00!!! !.00!!! .00!!! .00 (.00) (.00) (.00) (.00)R-square (adjusted) .05 .08 .04 .03

Note: Standard errors in parentheses. N = 6,302.!p < .05; !!p < .01; !!!p < .001 (two-tailed tests).



Table 5. Coefficients and Standard Errors from OLS Regression Models Predicting Fathers’ and Mothers’ Shares of Household Total Childcare Performed Away from and Together with Their Spouse

Childcare Away from SpouseChildcare Together with

Spouse

Fathers Mothers Fathers Mothers

Intercept .025 .54!!! .105!!! .33!!! (.01) (.02) (.02) (.02)One-and-a-half-earner (1.5E) .06!!! !.05! .02 !.03 (.01) (.02) (.02) (.02)Dual full-time earner (DFT) .09!!! !.10!!! .01 !.00 (.02) (.03) (.02) (.02)Father not employed full time

(FNFT).09!!!

(.02)!.18!!!(.03)

.02(.02)

.06!(.03)

Denmark .02 !.13!!! .03 .07! (.02) (.04) (.02) (.03)Italy .01 .04 .00 !.05!! (.01) (.02) (.01) (.02)France !.03 .01 !.02 .04! (.02) (.03) (.02) (.02)1.5E x Denmark !.04 .08 !.06 .02 (.04) (.05) (.04) (.05)1.5E x Italy !.01 !.03 !.01 .04 (.02) (.03) (.02) (.02)1.5E x France .01 .00 !.02 .01 (.02) (.03) (.02) (.03)DFT x Denmark !.02 .07 !.02 !.04 (.03) (.04) (.03) (.04)DFT x Italy !.01 !.02 .01 .02 (.02) (.03) (.02) (.03)DFT x France .03 !.05 .01 .01 (.02) (.04) (.03) (.03)FNFT x Denmark !.09 .09 .07 !.08 (.05) (.08) (.05) (.07)FNFT x Italy !.00 .01 !.02 .01 (.03) (.04) (.03) (.03)FNFT x France !.04 .08! !.03 !.01 (.03) (.04) (.03) (.04)Both have degree .03!!! !.05!!! .03!!! !.01 (.01) (.01) (.01) (.01)Only father has degree .02 !.04!! .04!!! !.01 (.01) (.02) (.01) (.01)Only mother has degree !.01 !.01 !.00 .02 (.01) (.01) (.01) (.01)Two children .01! .01 !.02!! !.00 (.01) (.01) (.01) (.01)Three or more children .01 .02 !.03!! !.00 (.01) (.01) (.01) (.01)Youngest child 5 to 12 years .04!!! .02! !.01 !.06!!! (.01) (.01) (.01) (.01)Teenager in family .02!! .01 !.03!!! !.01 (.01) (.01) (.01) (.01)Weekend !.01! !.20!!! .08!!! .13!!! (.01) (.01) (.01) (.01)Total household care .00 !.00!!! .00!!! !.00!! (.00) (.00) (.00) (.00)R-square (adjusted) .05 .12 .07 .06




we see that mothers and fathers in the refer-ence category performed 33 and 10.5 percent, respectively, of household total care that was done with their spouse present. These house-holds performed 43.5 percent of their total care in each others’ company, with mothers three times more likely to be doing that care than fathers. This further illustrates the gen-der inequality in total care and that the widest gender gaps are in care performed solo. Ref-erence category fathers did extremely little household care independently of their spouse.

Independent variables of interest in the model are household employment configura-tion, education, country, and interactions between these factors. We first report results for household employment configuration as a main effect, then country as a main effect, and then interaction terms between household employment configuration and country. We then discuss results relating to education. Finally, to illustrate the main findings sub-stantively, we present graphs showing the total effect of household employment config-uration in each country on fathers’ and moth-ers’ shares of household routine care and care performed solo.

Household employment configura-tion. Scholars expect MBW households to have the least equitable division of care. Con-sistent with this expectation, compared with the reference group, in 1.5-earner households, mothers did 6 percent less, and fathers 5 per-cent more, of routine household care. This makes no significant difference to the propor-tion of household care that is routine, but rather is a transfer between spouses. A corresponding transfer of non-routine care happened between mothers and fathers (although in this case only fathers’ adjustment reached statistical signifi-cance). A similar pattern occurred in households in which fathers did not work full time (FNFT). In these households, more equal shares arose from time recalibrations by both mothers and fathers.

By contrast, in DFT households, fathers did a greater share of routine care, but this was not transferred to mothers. Rather, mothers

in DFT households did a lower proportion of household non-routine care than did mothers in MBW (and indeed 1.5 and FNFT) house-holds. Recall that models controlled for abso-lute time in care, so this means routine care constituted a greater proportion of the lower total amount of care in DFT households (see Figure 2). DFT parents likely concentrated their care into shorter time periods around their employment schedules; therefore, nec-essary routine tasks such as picking children up from daycare, feeding, bathing, and put-ting them to bed took precedence over non-routine interactions. This reinforces that childcare time does not follow paid work time allocations in a linear fashion, because some childcare cannot be outsourced. Even when time is scarce, only some care activities can be dropped. Others must still be done.

The effect of household employment con-figuration on the proportion of childcare moth-ers and fathers each perform solo was consistent with the expectation that when mothers work, there will be a transfer of care responsibility to fathers. Mothers in 1.5-earner, DFT, and FNFT households did increasingly less of the house-hold care solo than did mothers in MBW households. Conversely, fathers in 1.5-earner, DFT, and FNFT households did more of the household care solo than did fathers in MBW households. This shows that in non-MBW households, there was a handover of solo responsibility for childcare from mothers to fathers. Household employment configuration affected gender shares of household childcare, although the magnitude of the differences was not large.

Country. To test whether social context plays a significant independent role in encour-aging division of care, we entered dummy variables for each country. Net of other influ-ences, the effect of country was strongest for French mothers, who, all else being equal, did 17 percent more of the household routine care, and 11 percent less of the household non-routine care, than Australian mothers (the reference group). Because French fathers made no corresponding adjustments, this



means that French households spent a lower proportion of their total care time in non- routine care than was the case elsewhere. We found that Italian mothers also did less non-routine childcare than Australian mothers, without corresponding adjustments by Italian fathers.

Results for Denmark are of particular interest because scholars assume that care is shared more equally in Scandinavian coun-tries than elsewhere. Although we found no significant difference between Australia and Denmark in the proportion of routine and non-routine childcare performed by mothers, in both cases the sign was negative. Together, they summed to Danish mothers doing 5 per-cent less of the household care than Austral-ian mothers. Furthermore, Danish fathers did 7 percent more of the routine household care than did Australian fathers, bringing their predicted contribution to 15 percent of the household total. This gives some credence to the view that Denmark’s normative and social policy framework encourages fathers’ involvement in care, including routine care. Moreover, because the reference category is MBW households, this suggests that in Den-mark, even households thought to be most gender-specialized are slightly more egalitar-ian than elsewhere. This was reinforced by results for the co-presence measures. Refer-ence category Danish mothers did 13 percent less of the childcare solo compared to Aus-tralian (and Italian and French) mothers. They did 7 percent more of the household care together with their spouse compared to refer-ence category mothers.

Household employment configura-tion by country. We now turn to the key question of whether household employment configuration has a different influence on shares of childcare across countries. We found interaction effects to be few in number, so overall, household type had the same relative effect in all countries. Although non-significant findings for household employment configu-ration x country predominate, the models did

show six significant interactions out of a total of 72 tested.

The strongest interaction effects were for French mothers in DFT households, who did 14 percent less of the household routine care, and 10 percent more of the household non-routine care, than did MBW mothers. This largely counteracts the influence of France as a main effect, noted earlier, and means that French DFT mothers enjoyed a substantially more gender-equal balance of childcare than did their compatriots in other household types. The same effect, although much less pronounced, was apparent for Italian DFT mothers. French and Italian mothers with the most equality in paid work hours also enjoyed more intra-household childcare equality than did their compatriots. In both countries, it was mothers themselves who adjusted their child-care composition when they worked longer hours, rather than both spouses readjusting. French and Italian fathers in DFT households did no more of the routine care, or of the care solo, than did their compatriots in other households.

We also found that fathers in French and Italian FNFT households did not do more of the non-routine household care (a main effect we found for this household type). This sug-gests that descriptive results showing little difference between fathers in FNFT and other households may be driven by France and Italy. French and Italian fathers may stick to traditional masculine shares of childcare even when they have less time constraints and lower relative resources. Masculine childcare patterns seem entrenched in these countries.

Education. As a main effect, fathers’ higher education predicted a transfer of house-hold routine care and solo care from the mother to the father (see Tables 4 and 5). This is the case when both spouses have, or just the father has, a college degree. Similarly, when only the father has a college degree, mothers did less of the household childcare solo. We saw no change in shares when only the mother has a



college degree, suggesting education’s effect operates through men rather than women. Fathers’ education is associated with more egalitarian gender attitudes and is a partial marker of family class. Our findings are there-fore consistent with interpretations that in households with higher socioeconomic status and when fathers have more liberal attitudes, fathers are willing to participate more in hands-on “concerted cultivation” of their children.

In additional models (not shown but avail-able on request), we entered interaction terms to see if the influence of education differed across countries. We found that in France and Italy, fathers in households in which both spouses have a degree did less care solo than did equivalent fathers in Australia (and Den-mark). These fathers are thus exceptions to the main effect findings noted earlier. This may mean that education is not associated with egalitarian gender attitudes in France and Italy to the same degree as elsewhere. We also found that Italian mothers with a degree did a higher proportion of the childcare solo than did equivalent mothers in other countries, further underlining the strong gender division of child-care in Italy. We ran models interacting house-hold employment configuration and education, but they added little additional explanatory power over the model specification presented.

Finally, we present graphs summarizing the main multivariate results to illustrate what they mean substantively. These graphs show the shares of childcare of reference-category mothers and fathers in different household employment categories in each country. To create them, we computed fitted values from Tables 4 and 5 as follows. For Australia, the reference country, we added the intercept and main effect for household employ ment configuration. For the other countries, we added the intercept and main effect for household employment configura-tion together with the main country effect and the interaction effect between country and household employment configuration. Figure 3 shows fathers’ and mothers’ shares of rou-tine childcare in each household configuration

in each country, and Figure 4 summarizes effects for fathers’ and mothers’ shares of childcare performed solo in each household configuration in each country.

The figures show the gender differences discussed earlier, and that effects of country and household employment configuration are more marginal than gender. The figures also illustrate the extent of these marginal differ-ences. In particular, they show the degree to which French mothers’ share of routine child-care exceeded that of other mothers (Figure 3, right-hand panel), and that the intersection between country and being in a DFT house-hold was stronger for French mothers than for mothers elsewhere (Figures 3 and 4, right-hand panels). The figures also illustrate the modest extent to which Danish fathers’ share of routine care exceeded that of other fathers (Figure 3, left-hand panel), and the compara-tively substantial extent to which Danish mothers’ share of childcare performed solo was less than other mothers (Figure 4, right-hand panel).

CONCLUSIONSMultiple factors influence parents’ decisions about dividing responsibility for childcare. Spouses’ lives are linked together within families, which in turn are situated within particular social contexts. Our aim in this article was to explore intersecting influences on the gendered division of childcare, mea-sured on key dimensions of task (routine and non-routine) and co-presence (care together with or away from a spouse). Our purpose was to offer a more detailed exploration of the gendered division of childcare than was pre-viously available by breaking childcare down into its constituent parts.

We should emphasize again that, across the board, gender was the strongest influence on the composition and share of childcare. As was foregrounded in the fully interactive models in the Appendix, and shown clearly in Figures 3 and 4, all other influences were minor in comparison. The gender difference



was stark. Mothers did a significantly higher proportion of total household care than did fathers on all measures. This held true across all household types and across all countries. Even in the most egalitarian household type (DFT) in the most egalitarian country (Denmark), mothers did much more of the care than fathers.

Given the amount of research cataloguing persistence of the gendered division of unpaid work, this is perhaps unsurprising. But child-care is different from other forms of unpaid work; it is an important part of family rela-tionships, essential to good social functioning and a deeply felt moral imperative (Folbre 2007; Lewis 2009). Mothers could arrange more equal intra-household shares by cutting back on their own care giving, but they may be reluctant to do so for fear their children will miss out. Perhaps one reason some mothers do not yield childcare time is because fathers do only certain types of childcare activity. We found that gender differences were widest in forms of care that research suggests mothers would regard as meaningful sharing: routine

care and care performed solo (Fuligni and Brooks-Gunn 2004).

Meaningful equality in childcare requires that fathers do more than join in non-routine care activities alongside mothers. It requires fathers to do more of the routine childcare and a larger share of childcare as the solo caregiver. We sought to identify which fac-tors were associated with movement toward more gender equality on these dimensions of care, and whether variation resulted from mothers’ or from fathers’ behavior, or from a combination of both. We were particularly interested in interplay between household employment configuration and country. The former indirectly taps parents’ values, con-straints, and resources, while the latter taps the social context within which individual parents, and the households they form, are nested. We found that each mattered, but not uniformly and not equally to each gender.

For example, households in which women were employed generally had slightly more equal shares on the activity and co-presence dimensions because fathers did more routine

0

0.1

0.2

0.3

0.4

0.5

0.6

0.7

0.8

MaleBreadwinner

1.5-Earner DualFull Time

Father NotEmployedFull Time

MaleBreadwinner


Father NotEmployedFull Time

Pare

nts'

Rou

!ne

Care

(% o

f Hou

seho

ldTo

tal C

are)

Fathers Mothers


Figure 3. Mothers’ and Fathers’ Average Shares of Routine Household Childcare in Households with Different Employment Configurations in Four CountriesNote: Calculated from results in Table 4, columns 1 and 2. Shares in Australia calculated by summing the intercept and the main effect of household employment configuration; shares in other countries calculated by summing the intercept and the main effect of household employment configuration, the main effect of country, and the interaction between household employment configuration and country. Reference category is a father/mother in households where neither parent has a college degree, there is one child age 0 to 4 years, no teenage children, and observed on a weekday.



and solo childcare. This suggests that house-hold-level characteristics influence divisions of childcare independent of country norms, and that fathers adjusted their care time in response to their wives’ employment. Moth-ers’ shares on each dimension similarly adjusted with employment status but less con-sistently. Most notably, mothers’ relative childcare composition in DFT households was similar to that of MBW mothers. This sug-gests that mothers’ role responsibilities are the same in both types of households, despite their different levels of work commitment.

However, this pattern was not universal. For example, French mothers did more of the routine childcare than other mothers. Coun-terbalancing this fact was that French DFT mothers had more equal shares of routine and non-routine care than their compatriots who worked fewer hours. Italy had a similar pat-tern, although less pronounced. In these coun-tries, household employment configuration made more of a difference to mothers’ shares of care than it did elsewhere. In particular, adopting more masculine work patterns meant more equal relative care composition for French mothers, bringing them closer to the average in other countries. Importantly,

however, readjustment occurred for mothers’ own care composition, with no significant corresponding adjustment among fathers. Fathers in DFT households showed no addi-tional movement toward more equality in France and Italy compared to the other coun-tries. In these instances, household-level work arrangements influenced mothers’ care com-position, while fathers’ care composition remained static. Similarly, education was the individual characteristic that most predicted higher relative shares for fathers across the range of childcare dimensions. This, however, was also country-dependent, as our supple-mentary analyses showed that in France and Italy education did not predict higher relative shares for fathers (supplementary analyses available upon request). The finding that higher education is associated with more rela-tive father-care applied only to Australia and Denmark. The behavior of French and Italian fathers with higher education remained more traditionally masculine.

Results suggest that cultural attitudes about masculinity and fatherhood may out-weigh individual- and household-level char-acteristics. This is further demonstrated by our finding that Danish men did slightly more

0

0.1

0.2

0.3

0.4

0.5

0.6

MaleBreadwinner


Father Not EmployedFull Time

MaleBreadwinner


Father Not EmployedFull Time

Pare

nts'

Sol

o Ca

re (%

of

Hou

seho

ld T

otal

Car

e)


Fathers Mothers

Figure 4. Mothers’ and Fathers’ Average Shares of Household Childcare Performed Away from Spouse in Households with Different Employment Configurations in Four CountriesNote: Calculated from results in Table 5, columns 1 and 2. Shares in Australia calculated by summing the intercept and the main effect of household employment configuration; shares in other countries calculated by summing the intercept and the main effect of household employment configuration, the main effect of country, and the interaction between household employment configuration and country. Reference category is a father/mother in households where neither parent has a college degree, there is one child age 0 to 4 years, no teenage children, and observed on a weekday.



routine care than fathers elsewhere, and Dan-ish women did less childcare solo than moth-ers elsewhere. These findings were robust across all household types, including MBW, and confirm findings by Hook (2006) in rela-tion to domestic tasks generally. Results imply that the Nordic discourse of father involvement and sharing is pervasive in Den-mark, even extending to households in which mothers are currently out of the workforce. Given the policy framework, many of these mothers are likely on maternity leave.

This article examined childcare in cross-national perspective in more detail than pre-vious research. Results confirm persistence of the gendered division of childcare, and the limited degree to which theories of time availability and relative resources explain shares of care. The results also illustrate the relevance of policy, culture, and social norms

in accentuating or minimizing the impact of individual- and household-level characteris-tics. Mothers’ and fathers’ education and work arrangements related to their shares of childcare, and this relationship differed slightly across countries. We also found cross-national variation in whether more equal shares result from the behavior of mothers, fathers, or both spouses. Changes to mascu-line and feminine roles, and to parenting norms for each gender, do not occur in lock-step. Our results suggest disjunctions are more pronounced in some countries than in others. Policies intended to encourage more father involvement or equal shares of child-care must be sensitive to social context, or they may not have the effects intended. Future research should build on this inquiry as more countries extend their national time-use sur-veys to collect information from both spouses.

Table A1. Coefficients and Standard Errors from Pooled OLS Models of Parents’ Shares of Routine and Non-routine Household Total Childcare, Including Three-Way Interactions between Gender, Household Employment Configuration, and Country


Main

EffectsGender

InteractionsMain

EffectsGender

Interactions

Intercept .53!!! .22!!! (.02) (.01) Male !.39!!! !.13!!! (.02) (.02) One-and-a-half-earner (1.5E) !.05!! .10!!! !.01 .03 (.02) (.03) (.02) (.02)Dual full-time earner (DFT) !.03 .13!!! !.06!! .06! (.02) (.03) (.02) (.03)Father not employed full time

(FNFT)!.08!!(.03)

.15!!!(.04)

!.04(.02)

.09!!(.03)

Denmark !.02 .10! !.04 .02 (.03) (.04) (.03) (.04)Italy .03 !.04 !.03 .04 (.02) (.02) (.01) (.02)France .18!!! !.22!!! !.10!!! .06* (.02) (.03) (.02) (.03)

APPENDIX

(continued)




Main

EffectsGender

InteractionsMain

EffectsGender

Interactions

1.5E x Denmark .04 !.07 .04 !.09 (.05) (.07) (.04) (.06)1.5E x Italy !.01 .02 .01 !.01 (.02) (.03) (.02) (.03)1.5E x France !.03 .05 .04 !.07 (.03) (.04) (.03) (.04)DFT x Denmark .01 !.06 .04 !.04 (.04) (.05) (.03) (.05)DFT x Italy !.04 .03 .04 !.01 (.03) (.04) (.02) (.03)DFT x France !.13!!! .17!!! .09!!! !.08! (.03) (.04) (.03) (.04)FNFT x Denmark .04 !.08 !.01 .02 (.07) (.10) (.06) (.08)FNFT x Italy .04 !.01 !.01 !.05 (.03) (.05) (.03) (.04)FNFT x France .02 !.03 .05 !.11! (.03) (.05) (.03) (.04)Both have degree .00 !.00 (.01) (.01) Only father has degree !.01 .01 (.01) (.01) Only mother has degree .00 !.00 (.01) (.01) Two children .01!! !.01!! (.01) (.00) Three or more children .03!!! !.03!!! (.01) (.01) Youngest child 5 to 12 years !.04!!! .04!!! (.01) (.00) Teenager in family .02! !.02!! (.01) (.01) Weekend !.02!!! .02!!! (.01) (.00) Total household care !.00!! .00!!! (.00) (.00) R-square (adjusted) .32 .05


Table A1. (continued)



Table A2. Coefficients and Standard Errors from Pooled OLS Models of Parents’ Shares of Household Total Childcare Performed Away from and Together with Their Spouse, Includ-ing Three-Way Interactions between Gender, Household Employment Configuration, and Country


Spouse

Main

EffectsGender

InteractionsMain

EffectsGender

Interactions

Intercept .47!!! .29!!! (.02) (.02) Male !.37!!! !.15!!! (.02) (.02) One-and-a-half-earner (1.5E) !.04! .09!!! !.02 .04 (.02) (.03) (.02) (.02)Dual full-time earner (DFT) !.10!!! .19!!! .00 .00 (.02) (.03) (.02) (.03)Father not employed full time

(FNFT)!.19!!!(.03)

.29!!!(.04)

.06!!(.02)

!.05(.04)

Denmark !.13!!! .17!!! .07!! !.04 (.03) (.04) (.03) (.04)Italy .03 !.01 !.03 .01 (.02) (.02) (.02) (.02)France .03 !.09!! .05!! !.08!! (.02) (.03) (.02) (.03)1.5E x Denmark .07 !.09 .01 !.07 (.05) (.06) (.04) (.06)1.5E x Italy !.04 .04 .03 !.04 (.02) (.03) (.02) (.03)1.5E x France !.00 .02 .01 !.03 (.03) (.04) (.03) (.04)DFT x Denmark .08! !.11! !.03 .00 (.04) (.05) (.03) (.05)DFT x Italy !.02 .01 .01 .00 (.03) (.04) (.02) (.03)DFT x France !.05 .08 .01 .00 (.03) (.04) (.03) (.04)FNFT x Denmark .10 !.20! !.07 .14 (.07) (.10) (.06) (.09)FNFT x Italy .02 !.03 .00 !.02 (.03) (.05) (.03) (.04)FNFT x France .09! !.13!! !.02 !.01 (.03) (.05) (.03) (.05)Both have degree !.01 .01 (.01) (.01) Only father has degree !.01 .01 (.01) (.01) Only mother has degree !.01 .01 (.01) (.01)

(continued)




Spouse

Main

EffectsGender

InteractionsMain

EffectsGender

Interactions

Two children .01 !.01! (.01) (.01) Three or more children .02 !.02! (.01) (.01) Youngest child 5 to 12 years .03!!! !.03!!! (.01) (.01) Teenager in family .02!! !.02!! (.01) (.01) Weekend !.11!!! .11!!! (.01) (.00) Total household care !.00!!! .00!!! (.00) (.00) R-square (adjusted) .24 .14


Table A2. (continued)

FundingThis research was supported by Australian Research Council DP 1093311.

AcknowledgmentsWe would like to thank Ciara Smyth, Bruce Bradbury, Natasha Cortis, Shelia Shaver, and three anonymous reviewers for their helpful advice and comments on drafts of this article.

ReferencesAbraham, Katherine, Aaron Maitland, and Suzanne Bian-

chi. 2006. Non-response in the American Time Use Survey: Who Is Missing From the Data and How Much Does it Matter? NBER Technical Working Paper 328, Cambridge, MA: National Bureau of Eco-nomic Research.

Allen, Sarah M. and Alan J. Hawkins. 1999. “Maternal Gatekeeping: Mothers’ Beliefs and Behaviors That Inhibit Greater Father Involvement in Family Work.” Journal of Marriage and Family 61:199–212.

Almqvist, Anna-Lena. 2008. “Why Most Swedish Fathers and Few French Fathers Use Paid Parental Leave: An Exploratory Qualitative Study of Parents.” Fathering: A Journal of Theory, Research, and Prac-tice about Men as Fathers 6:192–200.

Arts, Wil and John Gelissen. 2002. “Three Worlds of Wel-fare Capitalism or More? A State-of-the-Art Report.”

Journal of European Social Policy 12:137–58.Australia Bureau of Statistics. 2008. Time Use Survey

2006, Cat. No. 4152.0.55.001. Canberra, Australia: Australia Bureau of Statistics.

Baxter, Janeen, Belinda Hewitt, and Mark Western. 2005. “Post-familial Families and the Domestic Division of Labour.” Journal of Comparative Family Studies 36:583–600.

Baxter, Jennifer, Matthew Gray, Michael Alexander, Lyn-dall Strazdins, and Michael Bittman. 2007. “Mothers and Fathers with Young Children: Paid Employ-ment, Caring and Wellbeing.” Department of Fami-lies, Community Services and Indigenous Affairs, Canberra, Australia: FaHCSIA.

Becker, Gary. 1981. A Treatise on the Family. Cambridge, MA: Harvard University Press.

Bergmann, Barbara. 2005. The Economic Emergence of Women, 2nd ed. New York: Palgrave MacMillan.

Bianchi, Suzanne and Melissa Milkie. 2010. “Work and Family Research in the First Decade of the 21st Cen-tury.” Journal of Marriage and Family 72:705–725.

Bianchi, Suzanne, Melissa Milkie, Liana Sayer, and John Robinson. 2000. “Is Anyone Doing the Housework? Trends in the Gender Division of Household Labor.” Social Forces 79:191–228.

Bianchi, Suzanne, John Robinson, and Melissa Milkie. 2006. Changing Rhythms of American Family Life. New York: Russell Sage.

Bittman, Michael, Lyn Craig, and Nancy Folbre. 2004. “Packaging Care: What Happens When Parents Uti-lize Non-parental Child Care.” Pp. 133–51 in Family



Time: The Social Organization of Care, edited by N. Folbre and M. Bittman. London: Routledge.

Bittman, Michael, Paula England, Liana Sayer, Nancy Folbre, and George Matheson. 2003. “When Does Gender Trump Money? Bargaining and Time in Household Work.” American Journal of Sociology 109:186–214.

Bonke, Jens. 2005. “Paid Work and Unpaid Work: Diary Information versus Questionnaire Information.” Social Indicators Research 70:349–68.

Borchorst, Anette. 2008. “Woman-Friendly Policies and State Feminism: Theorizing Scandinavian Gender Equality.” Feminist Theory 9:207–224.

Brines, Julie. 1994. “Economic Dependency, Gender and the Division of Labour at Home.” American Journal of Sociology 100:652–88.

Buis, Martin. 2006. “Proportions As Dependent Vari-able.” Presented at the 12th UK Stata Users Group Meeting, London, England.

Cameron, Colin A. and Douglas L. Miller. 2010. “Robust Inference with Clustered Data.” Pp. 1–25 in Hand-book of Empirical Economics and Finance, edited by A. Ullah and D. E. Giles. Boca Raton, FL: Chapman & Hall/CRC.

Coltrane, Scott. 2000. “Research on Household Labor: Modelling and Measuring the Social Embeddedness of Routine Family Work.” Journal of Marriage and the Family 62:1208–33.

Coltrane, Scott. 2007. “Fatherhood, Gender and Work-Family Policies.” Pp. 385–410 in Real Utopias, edited by E. O. Wright. Madison, WI: The Havens Center, University of Wisconsin-Madison.

Craig, Lyn. 2006a. “Does Father Care Mean Fathers Share? A Comparison of How Mothers and Fathers in Intact Families Spend Time with Children.” Gender & Society 20:259–81.

Craig, Lyn. 2006b. “Parental Education, Time in Work and Time with Children: An Australian Time-Diary Analysis.” British Journal of Sociology 57:553–75.

Craig, Lyn. 2007a. Contemporary Motherhood: The Impact of Children on Adult Time. Aldershot, UK: Ashgate Publishing.

Craig, Lyn. 2007b. “How Employed Mothers in Austra-lia Find Time for Both Market Work and Childcare.” Journal of Family and Economic Issues 28:69–87.

Craig, Lyn and Michael Bittman. 2008. “The Effect of Children on Adults’ Time-Use: An Analysis of the Incremental Time Costs of Children in Australia.” Feminist Economics 14:57–85.

Craig, Lyn and Killian Mullan. 2010. “Parenthood, Gen-der and Work-Family Time in USA, Australia, Italy, France and Denmark.” Journal of Marriage and Family 72:1344–61.

Craig, Lyn, Killian Mullan, and Megan Blaxland. 2010. “Parenthood, Policy and Work-Family Time in Aus-tralia 1992–2006.” Work, Employment and Society 24:1–19.

Craig, Lyn and Pooja Sawrikar. 2009. “Work and Family: How Does the (Gender) Balance Change as Children Grow?” Gender, Work and Organisation 16:684–709.

Crompton, Rosemary. 2006. Employment and the Fam-ily: The Reconfiguration of Work and Family Life in Contemporary Societies. Cambridge, UK: Cambridge University Press.

Danish National Institute for Social Research (SFI). 2002. Danish Time Use Survey 2002. Copenhagen: Danish National Institute for Social Research.

Deding, Mette and Mette Lausten. 2006. “Choosing between His Time and Her Time: Paid and Unpaid Work of Danish Couples.” electronic International Journal of Time Use Research (eIJTUR) 3:28–48.

de Laat, Joost and Almudena Sevilla Sanz. 2007. “Working Women, Men’s Home Time and Lowest-Low Fertility in Europe.” University of Oxford, Department of Eco-nomics, Economic Series Working Papers No. 308.

Duncan, Simon and Rosalind Edwards. 2003. “State Welfare Regimes, Mothers’ Agencies and Gendered Moral Rationalities.” Pp. 1–23 in Thoughts on Fam-ily, Gender, Generation and Class: A Festschrift to Ulla Bjornberg, edited by K. Kollind and A. Peterson. Goteberg, Sweden: Sociology Institute, Goteberg University.

Fagnani, Jeanne and Marie-Thérèse Letablier. 2004. “Work and Family Life Balance: The Impact of the 35-Hour Laws in France.” Work, Employment & Soci-ety 18:551–72.

Ferree, Myra Marx. 2010. “Filling the Glass: Gender Perspectives on Families.” Journal of Marriage and Family 72:420–39.

Folbre, Nancy. 2007. Valuing Children: Rethinking the Economics of the Family. Boston: Harvard Univer-sity Press.

Fuligni, Allison Sidle and Jeanne Brooks-Gunn. 2004. “Measuring Mother and Father Shared Caregiv-ing: An Analysis Using the Panel Study of Income Dynamics Child Development Supplement.” Pp. 341–85 in Conceptualising and Measuring Father Involvement, edited by R. D. Day. Mahwah, NJ: Law-rence Erlbaum Associates.

Fuwa, Makiko. 2004. “Macro-Level Gender Inequality and the Division of Household Labor in 22 Coun-tries.” American Sociological Review 69:751–67.

Geist, Claudia. 2005. “The Welfare State and the Home: Regime Differences in the Domestic Division of Labour.” European Sociological Review 21:23–41.

Gornick, Janet and Marcia Meyers. 2003. Families that Work: Policies for Reconciling Parenthood and Employment. New York: Russell Sage.

Gornick, Janet and Marcia Meyers. 2009. Gender Equal-ity: Transforming Family Divisions of Labor. Vol. 4, Real Utopias Project. London: Verso.

Greenstein, Theodore. 2000. “Economic Dependence, Gender and the Division of Labour in the Home.” Journal of Marriage and the Family 62:322–35.



Haas, Linda. 2003. “Parental Leave and Gender Equality: Lessons from the European Union.” Review of Policy Research 20:89–116.

Hausmann, Ricardo, Laura Tyson, and Saadia Zahidi. 2007. The Global Gender Gap Report. Geneva: World Economic Forum.

Hays, Sharon. 1998. The Cultural Contradictions of Motherhood. New Haven, CT: Yale University Press.

Hobson, Barbara. 2002. “Making Men into Fathers: Men, Masculinities and the Social Politics of Fatherhood.” Cambridge, UK: Cambridge University Press.

Hofferth, Sandra. 2001. “Women’s Employment and Care of Children in the United States.” Pp. 151–74 in Women’s Employment in a Comparative Perspec-tive, edited by T. Van der Lippe and L. Van Dijk. New York: Aldine de Gruyter.

Hook, Jennifer. 2006. “Care in Context: Men’s Unpaid Work in 20 Countries, 1965–2003.” American Socio-logical Review 71:639–60.

Hook, Jennifer. 2010. “Gender Inequality in the Welfare State: Sex Segregation in Housework, 1965–2003.” American Journal of Sociology 115:1480–1523.

Kendig, Sarah and Suzanne Bianchi. 2008. “Single, Cohabitating, and Married Mothers’ Time with Children.” Journal of Marriage and Family 70: 1228–40.

Lareau, Annette. 2000. “Social Class and the Daily Lives of Children: A Study From the United States.” Child-hood 7:155–71.

Lewis, Jane. 2009. Work-Family Balance, Gender and Policy. Cheltenham, UK: Edward Elgar.

Lundberg, Shelly and Robert A. Pollack. 1996. “Bargain-ing and Distribution in Marriage.” Journal of Eco-nomic Perspectives 10:139–58.

Mancini, Anna Laura and Silvia Pasqua. 2011. “Asym-metries and Interdependencies in Time Use Between Italian Parents.” ZEW – Centre for European Eco-nomic Research, Discussion Paper No. 11-005.

McCall, Leslie. 2005. “The Complexity of Intersectional-ity.” Signs: Journal of Women in Culture and Society 30:1771–1800.

McDowell, Teresa and Shi-Ruei Sherry Fang. 2007. “Feminist-Informed Critical Multiculturalism: Con-siderations for Family Research.” Journal of Family Issues 28:549–66.

Miller, Pavla. 2004. “Demography and Gender Regimes: The Case of Italians and Ethnic Traditions.” Journal of Population Research 21:199–222.

Morgan, Kimberly. 2003. “The Politics of Mothers’ Employment: France in Comparative Perspective.” World Politics 55:259–89.

Morgan, Kimberly. 2005. “The ‘Production’ of Child Care: How Labor Markets Shape Social Policy and Vice Versa.” Social Politics 12:243–63.

National Institute of Statistics (ISTAT). 2007. L’uso del tempo: Indagine multiscopo sulle famiglie “Uso del

tempo” Anni 2002–2003. Rome: National Institute of Statistics.

National Institute of Statistics and Economic Studies (INSEE). 1999. Emploi de temp 1998–1999. Paris: National Institute of Statistics and Economic Studies.

OECD. 2007. “Babies and Bosses – Reconciling Work and Family Life: A Synthesis of Findings for OECD Countries.” Paris: Organisation for Economic Co-operation and Development.

Orloff, Ann. 2009. “Gendering the Comparative Analysis of Welfare States: An Unfinished Agenda.” Sociologi-cal Theory 27:317–43.

Papke, Leslie and Jeffrey Wooldridge. 1996. “Economet-ric Methods for Fractional Response Variables with an Application to 401(K) Plan Participation Rates.” Journal of Applied Econometrics 11:619–32.

Robinson, John P. and Geoffrey Godbey. 1997. Time for Life: The Surprising Ways Americans Use Their Time. University Park: Pennsylvania State University Press.

Rubery, Jill, Mark Smith, and Collette Fagan. 1999. Women’s Employment in Europe: Trends and Pros-pects. London: Routledge.

Sayer, Liana, Suzanne Bianchi, and John Robinson. 2004. “Are Parents Investing Less in Children? Trends in Mothers and Fathers Time with Children.” American Journal of Sociology 110:1–43.

Sayer, Liana and Janet Gornick. 2011. “Cross-National Variation in the Influence of Employment Hours on Child Care Time.” European Sociological Review. doi: 10.1093/esr/jcr008.

Shaw, Susan. 2008. “Family Leisure and Changing Ideol-ogies of Parenthood.” Sociology Compass 2:688–703.

Torr, Berna Miller and Susan E. Short. 2004. “Second Births and the Second Shift: A Research Note on Gender Equity and Fertility.” Population and Devel-opment Review 30:109–131.

Vincent, Carol and Stephen J. Ball. 2007. “‘Making Up’ the Middle-Class Child: Families, Activities and Class Dispositions.” Sociology 41:1061–1077.

Wall, Glenda. 2010. “Mothers’ Experiences with Inten-sive Parenting and Brain Development Discourse.” Women’s Studies International Forum 33:253–63.

Warner, Judith 2005. Perfect Madness: Motherhood in the Age of Anxiety. New York: Riverhead.

White, Linda. 2009. “Explaining Differences in Child Care Policy Development in France and the USA: Norms, Frames, Programmatic Ideas.” International Political Science Review 30:385–405.

Windebank, Jan. 2001. “Dual-Earner Couples in Britain and France: Gender Divisions of Domestic Labour and Parenting Work in Different Welfare States.” Work, Employment and Society 15:269–90.

Wisensale, Steven. 2003. “Two Steps Forward, One Step Back: The Family and Medical Leave Act as Retrenchment Policy.” Review of Policy Research 20:135–52.



Yeung, Jean, John Sandberg, Pamela Davis-Kean, and Sandra Hofferth. 2001. “Children’s Time with Fathers in Intact Families.” Journal of Marriage and the Family 63:136–54.

Lyn Craig is an Australian Research Council Queen Elizabeth II Fellow at the Social Policy Research Centre at the University of New South Wales, Sydney. Her research interests include parenthood and time equity, work-family balance, and the intra-household

effects of work-care policy structures. She is author of Contemporary Motherhood: The Impact of Children on Adult Time (Ashgate 2007).

Killian Mullan is a Research Fellow at the Social Policy Research Centre in the University of New South Wales, Sydney. He pursues a number of interests in the broad field of time-use research including work-family balance and the time use of young people. He is also interested in methodological issues relating to time-use research.


American Sociological Review-2011-Craig-834-61 more often than do fathers (Craig 2006a; Fuligni and...

Documents

Transcript of American Sociological Review-2011-Craig-834-61 more often than do fathers (Craig 2006a; Fuligni and...