Educational Gradients in Parents' Child-Care Time Across Countries, 1965-2012: Educational Gradients in Parents' Child-Care Time

Abstract

Parental time with children leads to posive child outcomes. Some studies have reported a posive educational gradient: More educated parents devote more time to children than other parents. Furthermore, some research finds that parental child care increased over time. Less certain is whether more educated parents increased their time more than less educated ones did, whether parenting trends for mothers and fathers are the same, and whether observed patterns characterize all Western countries or only some. Hypotheses inspired by theories of social diffusion, class differentiation, and ideologies of child rearing are tested with time-use data for 11 Western countries between 1965 and 2012. For both mothers and fathers, results indicated a widespread educational gradient and an increase in child-care time. In a number of countries, the posive educational gradient increased; nowhere was it dished. Thus, the advantages of intensive parenting continued to accrue to the well-educated elite.

Full text

G M. D S Collegio Carlo Alberto
J T University of California Irvine∗
Educational Gradients in Parents’ Child-Care Time
Across Countries, 1965–2012
Parental time with children leads to positive
child outcomes. Some studies have reported a
positive educational gradient: More educated
parents devote more time to children than other
parents. Furthermore, some research nds
that parental child care increased over time.
Less certain is whether more educated parents
increased their time more than less educated
ones did, whether parenting trends for mothers
and fathers are the same, and whether observed
patterns characterize all Western countries or
only some. Hypotheses inspired by theories of
social diffusion, class differentiation, and ide-
ologies of child rearing are tested with time-use
data for 11 Western countries between 1965
and 2012. For both mothers and fathers, results
indicated a widespread educational gradient
and an increase in child-care time. In a number
of countries, the positive educational gradient
increased; nowhere was it diminished. Thus, the
advantages of intensive parenting continued to
accrue to the well-educated elite.
The time parents devote to child care is
important. Children benet from parental
Collegio Carlo Alberto, Via Real Collegio 30, 10024,
Moncalieri, Turin, Italy (giulia.dottisani@carloalberto.org).
∗Department of Sociology, SSPA 3151, University of
California–Irvine, Irvine, CA 92617 (jktreas@uci.edu).
This article was edited by Jennifer Glass.
Key Words: child care, cross-national, education, parental
investment/involvement, social change, social trends.
interaction, notably in terms of academic
achievement (Bernal & Keane, 2011; Hill,
Waldfogel, Brooks-Gunn, & Han, 2005), cog-
nition (Lugo-Gil & Tamis-LeMonda, 2008;
Tucker-Drob & Harden, 2012), language acqui-
sition (Leibowitz, 1977; Rowe, 2008), and
behavior (Laird, Pettit, Bates, & Dodge, 2003;
Vandell et al., 2010). Parents in several Western
countries have been reported to be spending
increasingly more time in child care (Bianchi,
2000; Gauthier, Smeeding, & Furstenberg,
2004; Sayer, Bianchi, & Robinson, 2004). Para-
doxically, the parents who spend the most time
with children seem to be more educated ones:
those whose time commanded the highest earn-
ings in the labor market (England & Srivastava,
2013; Guryan, Hurst, & Kearney, 2008; Sayer,
Gauthier, & Furstenberg, 2004). Although the
well educated may use higher incomes to free
up time for children, their behavior is con-
sistent with an intensive parenting ideology
that promotes practices that not only benet
children but also can be considered a status
marker differentiating higher from lower social
classes (Lareau, 2003). Although social class is
a multidimensional concept, this article focuses
on the educational facet, which has gured
prominently in the literature on parental child
care. Because the positive educational gradi-
ent dees the logic of economic opportunity
costs, it points up the need to consider educa-
tion not merely as human capital but also as a
pipeline for new ideas about parenting. Moving
beyond prior research to a systematic analysis of
country-to-country differences, this article asks
Journal of Marriage and Family (2016) 1
DOI:10.1111/jomf.12305

2Journal of Marriage and Family
how parents’ education has gured in changes
in the time mothers and fathers spend caring for
children.
Trends over time in the positive educational
gradient in parental child care are not entirely
understood. Higher social classes are typically
the rst to adopt new practices (Rogers, 1962).
Thus, we would expect to nd increases over
time in the educational gradient as the more
educated increase their time with children. But
innovations diffuse eventually to lower social
classes. We would expect less educated peo-
ple ultimately to narrow the education gap in
child-care time by adopting the intensive par-
enting practices of their more educated counter-
parts. If this parenting style consequently loses
its status-conferring association with the elite,
the more educated might retreat from intensive
parenting. Even a negative educational gradient
is a possibility, because the better educated have
the means to outsource child care to paid helpers.
Whether the educational gradient is increas-
ing or decreasing may well depend on when
intensive parenting practices were introduced
into different societies. A historical account of
the rise of intensive mothering in the United
States outlines such developments (Hays, 1996),
but the onset of such parenting has not been
established across countries for mothers and
for fathers.
Drawing on surveys from the Multinational
Time Use Study for 11 Western countries
between 1965 and 2012, this article has three
main objectives. First, we develop and test alter-
native hypotheses regarding parents’ child-care
time that are informed by theories on diffusion
of innovations, class differentiation, and cultural
ideologies of child rearing. Second, to clarify
trends and differentials in parent time spent car-
ing for children, we move beyond piecemeal and
sometimes conicting ndings with a system-
atic, cross-national analysis of a comparatively
large and diverse set of 11 Western countries.
Third, for these countries, we ask whether more
educated mothers and fathers have increased the
time they spend in child care to a greater extent
than their less educated counterparts. That is, we
ask whether the positive educational gradient is
increasing and whether an increase is occurring
everywhere. Because parental time is linked to
many positive child outcomes, any growth in
the gap in parental time in child care has serious
implications for the perpetuation of inequality
across generations.
B
Parenting norms have undergone consider-
able change in response to new scientic
evidence on child development as well as
cultural shifts in relationship ideals for men
and women, parents and children. Already
responsible for more child care than men (Gau-
thier et al., 2004; Sayer, Bianchi, et al., 2004),
women have been called on to do “intensive
mothering,” an evolving maternal ideal that
is “child-centered, expert-guided, emotion-
ally absorbing, labor-intensive and nancially
expensive” (Hays, 1996, p. 54). Articulat-
ing an essentialist and ecological philosophy
of child well-being, the ethos of “natural
motherhood” described for France justies
demanding child-rearing practices, such as
lengthy breast-feeding and use of cloth diapers
(Badinter, 2012). Although new child-rearing
philosophies are directed explicitly or implicitly
at mothers (England & Srivastava, 2013), “new
fatherhood” celebrates the paternal role in child
development and fathers’ active engagement
in children’s lives (Henwood & Procter, 2003;
Marsiglio, Amato, Day, & Lamb, 2000). Orig-
inally described for the United States, new
fatherhood norms have been identied in other
countries as well (Duyvendak & Stavenuiter,
2004; Hook & Wolfe, 2012; Kim, 2014; Svab,
2013). In the American middle class, a cultural
repertoire of child rearing dubbed “concerted
cultivation” encourages both mothers and
fathers to prepare children for success by engag-
ing them in time-consuming verbal interaction,
reasoning, and structured enrichment activities
(Lareau, 2000). Parental engagement is viewed
as necessary not only to protect children in
anonymous urban environments (Bianchi, 2011)
but also to ensure their success in an increas-
ingly competitive economy (Ramey & Ramey,
2009). Parenting is a valued priority. Not only
did Dutch, American, and British fathers and
mothers nd child care more enjoyable than
housework; they also reported holding higher
standards for child care (Poortman & van der
Lippe, 2009; Sullivan, 2013).
More and more, parents’ time use resonates
with contemporary parenting ideologies that
emphasize high parent engagement in children’s
lives. According to studies on individual coun-
tries in North America and Europe (Bianchi
2000; Gauthier et al., 2004), parental time spent
in child care increased over time. Similar results
were reported when countries were pooled—for

Educational Gradients in Parents’ Child-Care Time 3
example, across Canada, Denmark, the Nether-
lands, Finland, the United States, and the United
Kingdom (Sullivan & Gershuny, 2001) or across
16 Western nations (Gauthier et al., 2004).
Although pooled results did not identify whether
some countries deviated from the overall trend,
there were exceptions to the upward shift in
parenting time. An analysis by Kan, Sullivan,
and Gershuny (2011) raised questions about
how closely individual countries have tracked
the general increase in parents’ child-care
time. For 16 Western countries, they charted
daily minutes of caring labor done by men and
women from the 1960s into the 21st century.
Caring included care of any family member, not
children exclusively. Their graphic presentation
clearly showed declines, not increases, in car-
ing time for the French and Swedes, especially
women, who displayed more country-to-country
variation than men did in their time trajectories.
Consistent with the adoption of new ideologies
that cause shifts in fathers’ time use, Maume
(2011) estimated that 70% of the increase in
U.S. fathers’ time with children was due to
their changing behavior and only 30% to their
changing characteristics. Taken together, these
piecemeal and sometimes conicting results
invited a closer look at country-specic trends
in parents’ investments in child care.
The increase in parental time with chil-
dren was an unexpected development. In the
absence of any decline in fathers’ working
hours (Organisation for Economic Co-operation
and Development, OECD, 2014), the growth
in mothers’ labor-force participation pointed
to competing time demands that should deter
parents from devoting more time to child care
(Craig, Mullan, & Blaxland, 2010; Kan et al.,
2011). Employed mothers did spend less time
caring for children than mothers not working
for pay, but time in child care increased for
American mothers and fathers, working or not,
as well as for European parents (Gauthier et al.,
2004; Sullivan, Billari, & Altintas, 2014). Para-
doxically, women’s wages related positively
to their caregiving (Guryan et al., 2008; Kim-
mel & Connelly, 2007) but negatively to their
housework (Bryan & Sevilla-Sanz, 2011; Gupta,
2006). Denying the logic of time availability
and the opportunity costs of earnings forgone in
child care, this behavior invites cultural expla-
nations for historical increases in child-care
time. Because parents’ schooling was more
consequential for child-care time than earnings
and employment constraints were, England
and Srivastava (2013) argued for closer atten-
tion to child-rearing values, which are closely
associated with parental education.
Parents’ education is positively associated
with child care (England & Saraff, 2010; Lareau,
2003; Sayer, Gauthier, et al., 2004; Sullivan
et al., 2014). For educational, recreational, and
travel-related activities (e.g., museum going,
reading books together) associated with the con-
certed cultivation of the middle class, mother’s
education proved critical in Spain and the United
Kingdom (Gimenez-Nadal & Molina, 2013) as
well as in Italy (Mancini & Pasqua, 2009). This
positive educational gradient in parental child
care is consistent with class-based ideals of
intensive parenting (Gauthier et al., 2004; Sayer,
Gauthier, et al., 2004; Sullivan et al., 2014).
From other perspectives, the positive gradient is
surprising. Parents with more education should
be disposed to do less child care, given their
ability to pay for babysitters, higher employ-
ment rates, more exible jobs, greater opportu-
nity costs for time not spent working for pay,
safer environments, fewer children, and possibly
greater productivity in the time they do devote to
offspring (England & Srivastava, 2013; Lyness,
Gornick, Stone, & Grotto, 2012; Yamauchi &
Leigh, 2011).
In the United States, the more educated
logged no more time in the presence of children,
but they did devote more time to direct child
care (Guryan et al., 2008). Parents with more
schooling are better informed on the norms
and rationale for intensive parenting (Benasich
& Brooks-Gunn, 1996). Conscious of the link
between time investments and positive child
outcomes, they engage in activities that promote
children’s cognitive growth (Hofferth & Sand-
berg, 2001; Lareau, 2000), use child-directed
speech to enhance vocabularies (Rowe, 2008),
and ne-tune their efforts to the child’s devel-
opmental stage (Kalil, Ryan, & Corey, 2012;
Gracia, 2014). Having high standards, more
educated parents nd it hard to locate accept-
able substitutes for their own time with children
(Sayer, Gauthier, et al., 2004).
Whether positive educational gradients hold
for both men and women and across coun-
tries remains to be seen. A pooled analysis on
Australia, Denmark, France, and Italy showed
signicant positive effects of fathers’—but not
mothers’—higher level of education on their
share of routine child care (Craig & Mullan,

4Journal of Marriage and Family
2011). Whatever the parent’s gender, higher edu-
cation was associated with more time in basic
child care in Spain but not the United Kingdom
(Gimenez-Nadal & Molina, 2013). As for cog-
nitively enriching educational child care, U.K.
mothers and Spanish fathers showed no associa-
tion with parental education net of the spouse’s
schooling, U.K. fathers a negative association,
and Spanish mothers a positive one. In con-
trast, for Australian mothers and fathers, edu-
cation had no association with the interactive,
talk-based care (e.g., listening, playing games)
that contributed most to children’s cognitive
development (Craig, Powell, & Smyth, 2014).
Studies of changes over time in the educa-
tional gradient also led to inconclusive results.
A study focused on the United States reported
college-educated mothers and fathers increas-
ing their weekly hours of child care more than
less educated parents (Ramey & Ramey, 2009).
In Australia, by contrast, education’s impact
on child-care time declined and ceased to be
signicant (Craig, et al., 2014). In analyses
with pooled U.S. and U.K. data on fathers,
educated men increased their child care more
than other men between 1975 and the early
2000s (Sullivan, 2010). Although there was also
an increase in the paternal educational gra-
dient in most of the seven countries consid-
ered by Gimenez-Nadal and Sevilla (2012), the
results for mothers showed more cross-national
variation. Large-scale, cross-national compar-
isons of trends in the educational gradient in
child care are lacking. For 13 OECD countries,
Sullivan et al. (2014) reported a greater increase
in domestic time for college-educated fathers
than for others, but their analysis did not con-
sider mothers or distinguish child care from core
housework activities.
Indirect evidence on education’s role in
trends comes from studies of particular types
of child care. There is evidence of an increase
in the child-oriented, enrichment activities that
are favored by the middle class (Lareau, 2003).
According to 1988–2002 Italian time-use data,
“quality time” (e.g., helping with homework,
playing) increased compared to “basic care”
(e.g., supervision, physical care; Mancini &
Pasqua, 2009). In Canada, the increase in child
care was mostly due to parent–child interac-
tion time (Gauthier et al., 2004). A long-run
increase in interactive child-care activities was
also reported for the United States (Bianchi,
Robinson, & Milkie, 2006). Some evidence also
linked child-care practices to parenting ideals.
In a small, Midwestern U.S. city, parents in 1924
were less likely than those in 1978 to say they
preferred children’s independence over strict
obedience. In both years, the “business” class
more so than the “working class” favored inde-
pendence, behavior that presumably required
the time-intensive parenting (e.g., verbal inter-
action, reasoning with children) associated with
concerted cultivation (Alwin, 1988). Not known
is whether the overall increase in parental time
in child care reected a diffusion of parenting
practices from the middle to lower classes or
a doubling down on intensive parenting on the
part of the middle class. In other words, we
do not know whether the positive educational
gradient became steeper over time.
Of course, other variables predict parental
time in child care (Monna & Gauthier, 2008).
For example, having more and younger children
is linked to spending more time in child care
(Sayer, Bianchi, et al., 2004). As a result of com-
peting time demands, employment is associated
with less child-care time, as is being a nonpart-
nered, single parent (Bianchi, 2011). Older par-
ents are reported to spend less time in child care
(Sayer, Bianchi, et al., 2004), and mothers do
less on weekends than on weekdays (England &
Srivastava, 2013; Sayer, Bianchi, et al., 2004).
Across countries, key issues on parents’ time
in child care remain unsettled. Several studies
have offered valuable insights, but results have
not necessarily cumulated. In part, this reects
the difculty in drawing comparisons from stud-
ies that have employed different measures of
care, focused on different countries and time
periods, adjusted for different covariates, and
provided information on only fathers or only
mothers. To clarify patterns in parental time
in child care, including trends in the important
educational gradient, systematic analysis over a
large set of countries is needed.
H
Contemporary ideologies of child rearing
that emphasize time-intensive parenting are
consistent not only with an increase in parental
time in child care but also with the positive
educational gradient that shows parents with
more education doing more child care than
other parents. According to the classic theory
on the diffusion of innovations, new practices
are embraced rst by educated elites (Rogers,

Educational Gradients in Parents’ Child-Care Time 5
1962). Stressing mundane practices and cultural
capital acquired in childhood, Bourdieu’s (1984)
theory of class distinctions also observes that
dominant upper classes dissociate themselves
from lower classes by deploying unique social
practices. Thus, the time-intensive parenting
style of the well educated may be a means of
social closure, separating elites from the masses
(Weber, 1922/1978). Diffusion and distinction
processes predict an increase in the positive
educational gradient over time as well-educated
parents double down on their time spent in
child care.
We would not expect the gap in child care
time between more and less educated parents
to grow indenitely. There is a ceiling to the
amount of time well-educated mothers and
fathers can spend with their children. Fur-
thermore, the theory of diffusion predicts that
practices adopted by elites will eventually be
taken up by parents in lower social classes,
portending a narrowing of the class difference in
time invested in children. More educated early
adopters may well retreat from child care if
intensive parenting loses its cachet as a marker
of elite status. Without timing information on the
onset of intensive parenting, it is hard to predict a
tipping point at which a widening education gap
in child care would begin to narrow or a positive
educational gradient turn negative. Nonethe-
less, this possibility cannot be discounted
across nations.
In short, prior research is at odds with eco-
nomic theories emphasizing economic oppor-
tunity costs. Consistent with culture-based,
sociological theories emphasizing new norms of
intensive child rearing, parents have increased
their time in child care in at least some coun-
tries. Furthermore, more educated parents seem
to do more child care than those with less
schooling. We do not know whether the positive
educational gradient in parental child care has
increased, and if so, whether the increase is
driven by developments in a few countries or
many. Although both parents seem to have been
subject to intensive parenting expectations, we
do not know whether mothers and fathers have
responded in the same way. Given the strong evi-
dence linking parental time to a host of positive
child outcomes, a pronounced shift toward more
hours in child care among more educated moth-
ers and fathers has important implications. If the
educational gradient is growing, this stands to
compound the advantages of children growing
up in middle-class families, thus entrenching
a mechanism by which parents sustain social
inequality.
The prior research on changing cultural
norms for parenthood inspires hypotheses
about the allocation of time to child care. First,
more educated parents are more aware of the
advantages of spending time with their chil-
dren (Benasich & Brooks-Gunn, 1996). Thus,
the education-gradient hypothesis (H1) holds
that more educated parents will spend more
time on child care than less educated mothers
and fathers. Second, more educated parents
may intensify their child care and/or their
parenting practices may diffuse to those with
less schooling. Either way, the increasing-care
hypothesis (H2) anticipates that parents in
Western countries will devote more time to
child care over the years. Third, because inten-
sive parenting practices not only contribute to
positive child outcomes but also distinguish
elites from lower social classes, we expect
more educated parents to increase their time
in child care more rapidly than other parents.
This implies a widening educational gradient
(increasing-gradient hypothesis, H3A).Innova-
tions, however, eventually diffuse downward
from higher social classes, losing the status
marker attraction for elites as they become more
widely adopted. In this case, more educated
parents will not step up their intensive parenting
as much as their less educated counterparts
do. The decreasing-gradient hypothesis (H3B)
predicts a decline in the educational gradient
over time.
M
Data come from the Multinational Time Use
Study (MTUS) Heritage Simple Files (Fisher &
Gershuny, 2013). A harmonized collection of
cross-national time-use studies recoded to pro-
vide consistent data across surveys, the MTUS is
the only data set that allows for the comparative
study of historical changes in time allocation.
Therefore, the MTUS is the most suitable data
source for analyzing changes in child care across
a large number of countries and over time. The
analysis focuses on parents, aged 18–65, living
in households with at least one child under 13
years old. Excluding the 2.3% of respondents
with missing values on independent variables
leaves 122,271 observations (68,532 mothers
and 53,739 fathers) nested in 11 countries.

6Journal of Marriage and Family
For one or more days, respondents were asked
to keep a diary recording all their activities in
their own words. In terms of data quality,
time-use diary results are generally preferred
to survey self-reports for unpaid household
labor (Schulz & Grunow, 2012). For diaries
with multiple days, we randomly selected one
day for each respondent. Alternative proce-
dures (e.g., excluding multiday countries, using
robust standard errors clustering days within
respondents) yielded consistent results. Dating
to the 1960s, the harmonized time-use surveys
include a large number of countries. Given the
post hoc MTUS harmonization, comparable
information on two or more time points is not
available for all countries. We chose the 11
countries with comparable data from multi-
ple time points. The diversity of the countries
is indicated by the range of welfare-regime
types (Esping-Andersen, 1999): Canada, United
Kingdom, and United States belong to the lib-
eral welfare regime; Denmark and Norway are
so-called social democracies; France, Germany,
and the Netherlands are part of the conser-
vative group; Italy and Spain are classied
as “southern”; and Slovenia is representative
of formerly socialist countries. Multiple time
points are available for each country: Canada,
1971–1972, 1981, 1986, 1992, 1998–1999; Ger-
many, 1992, 1993, 2001, 2002; Denmark, 1987,
2001; Spain, 1992–1993, 1997, 1998, 2002,
2003, 2008; France, 1965, 1974, 1998–1999;
Italy, 1989, 2002, 2003; the Netherlands, 1975,
1980, 1985, 1990, 1995, 2000, 2005; Norway,
1980–1982, 1990–1991, 2000–2001; Slove-
nia, 1965 (Yugoslavia), 2000–2001; United
Kingdom, 1974–1975, 1983–1985, 1987, 1995,
2000, 2001, 2005; United States, 1965–1966,
1975–1976, 1985, 1994–1995, 1999–2000,
2003, 2004, 2005, 2006, 2007, 2008, 2009,
2010, 2011, 2012. Some other country-years
could not be used because information on
independent variables was lacking.
Variables
Top coded at 12 hours, the harmonized depen-
dent variable for daily minutes in child care
consists of feeding and preparing food for
babies and children, washing, changing,
putting to bed or getting up, unpaid babysitting,
providing medical care, reading to or playing
with children, helping with homework, and
supervising. The variable includes both interac-
tive child care and routine child care, because
only a few countries disaggregate by type of
activity.
To test the education-gradient hypothesis
(H1), a dummy independent variable contrasts
higher education (postsecondary education)
and lower education (incomplete or completed
secondary education), the omitted reference
category. Survey year captures changes in
parental time for the increasing-care hypothesis
(H2). The variable is centered at its median to
facilitate the interpretation of the intercepts. An
interaction between survey year and parent’s
education tests the hypotheses on increasing
and decreasing gradient (H3A and H3B), which
anticipate alternative results for changes over
time in child-care time by education.
Controls include important respondent char-
acteristics: age (mean centered), partnership
status (not living with partner =1; other =0),
number of children younger than age 13 in
the household, age of the youngest child
(5–12 years, with 0–4 as the omitted refer-
ence), and whether the respondent is employed
(employed =0, not employed =1). We also con-
trolled for whether the diary day was a weekend
day (1) or weekday (0). Table 1 reports summary
statistics by country and gender and for the total
sample.
Models
We are interested not only in testing hypotheses
regarding parental time use but also in ascer-
taining whether the results differ from country
to country. Moving beyond the limited general-
izations possible with earlier research, we take
a rigorous multivariate approach with formal
statistical tests for the signicance of differ-
ences over time, between education groups, and
for changes in the educational gradient over
time. We address the cross-national consistency
of results with models that allow effects to
vary by country. Separately for mothers and
fathers, we analyze child-care minutes using
multilevel, random intercepts–random slopes
models for respondents nested in 11 countries.
Multilevel models are the best strategy for
taking into account the nested structure of the
data and for modeling the within-country versus
between-country variation. Following previous
time-use studies (Craig, Powell, & Brown,
2015), we apply a generalized linear model

Educational Gradients in Parents’ Child-Care Time 7
Tab l e 1 . Descriptive Statistics for Mothers and Fathers, by Country and Total
Mothers
CA DE DK ES FR IT NL NO SL UK US Total
Child-care minutes per day 109.29 102.24 77.53 95.33 106.10 74.24 91.66 122.49 80.61 97.85 104.30 98.48
Age (18–65) 32.98 35.63 34.80 36.68 33.27 35.11 33.93 33.46 34.49 33.87 36.09 35.32
Not in a couple 0.20 0.18 0.15 0.16 0.13 0.11 0.08 0.12 0.16 0.24 0.29 0.21
Number of children 1.94 1.81 1.85 1.72 1.70 1.77 2.04 1.94 1.55 2.01 2.04 1.92
Youngest child age 0–4 0.52 0.45 0.56 0.43 0.60 0.35 0.56 0.69 0.62 0.51 0.49 0.48
Youngest child age 5–12 0.48 0.55 0.44 0.57 0.40 0.65 0.44 0.31 0.38 0.49 0.51 0.52
Not employed 0.52 0.34 0.18 0.49 0.46 0.52 0.20 0.31 0.31 0.42 0.35 0.40
Weekend 0.27 0.30 0.44 0.37 0.30 0.65 0.28 0.28 0.43 0.37 0.49 0.44
Lower education 0.52 0.71 0.51 0.75 0.69 0.92 0.82 0.70 0.86 0.78 0.38 0.59
Higher education 0.48 0.29 0.49 0.25 0.31 0.08 0.18 0.30 0.14 0.22 0.62 0.41
Survey year 1988 1996 1994 2002 1989 1995 1988 1992 1991 1994 2004 1999
N5,308 3,680 964 7,511 2,964 9,011 3,501 1,613 1,030 3,122 29,828 68,532
Fathers
CA DE DK ES FR IT NL NO SL UK US Total
Child-care minutes per day 46.46 41.58 39.83 42.43 28.83 30.96 39.27 55.48 32.41 42.74 59.44 46.74
Age (18–65) 35.38 38.26 37.48 38.88 35.48 38.23 36.04 35.83 36.10 36.52 38.18 37.70
Not in a couple 0.05 0.05 0.10 0.09 0.06 0.08 0.03 0.03 0.13 0.07 0.11 0.09
Number of children 1.98 1.85 1.91 1.73 1.69 1.78 2.07 2.02 1.55 2.03 2.05 1.92
Youngest child age 0–4 0.56 0.47 0.54 0.43 0.62 0.35 0.58 0.69 0.59 0.51 0.50 0.48
Youngest child age 5–12 0.44 0.53 0.46 0.57 0.38 0.65 0.42 0.31 0.41 0.49 0.50 0.52
Not employed 0.11 0.07 0.05 0.12 0.10 0.09 0.06 0.06 0.14 0.09 0.10 0.10
Weekend 0.27 0.30 0.42 0.41 0.30 0.65 0.31 0.32 0.45 0.37 0.50 0.45
Lower education 0.49 0.60 0.61 0.76 0.72 0.91 0.72 0.68 0.89 0.74 0.39 0.61
Higher education 0.51 0.40 0.39 0.24 0.28 0.09 0.28 0.32 0.11 0.26 0.61 0.39
Survey year 1988 1995 1994 2001 1989 1995 1989 1992 1992 1993 2004 1998
N3,867 3,168 825 7,356 2,742 8,531 2,101 1,150 844 2,271 20,884 53,739
(GLM) with a log link and the gamma family.
This takes into account the many mostly male
respondents reporting zero time in child care
and ensures that time estimates are not negative,
as might occur with a linear model.
Baseline Model 0 consists of a null random
intercept model with no covariates. Model 1, a
random intercept with xed slope model, adds
not only control variables but also survey year
and education level, which test for the time
trend and the educational gradient, respectively.
Model 2 introduces the interaction between year
and education to determine whether the educa-
tion gradient for child-care time has increased
or decreased over time. To assess whether the
historical changes in child care time by educa-
tion exist for all countries, Model 3 also allows
the interaction of education and year and the
related main terms of year and education to vary
by country.
R
Multilevel results addressed, rst, the hypothe-
ses for the main effects of education and survey
year. This established whether, net of covariates,
mothers and fathers in this set of 11 countries
showed the predicted positive education gradi-
ent (H1) and the anticipated increase over time
(H2) in daily minutes devoted to child care. For-
mal tests assessed whether the educational gradi-
ent increased (H3A) or decreased (H3B ) over the
period and whether this general trend described
all the countries.
Trend and Gradient
Table 2 reports multilevel GLM results for
mothers and fathers separately. In the baseline
Model 0, the random intercept terms for the
average time spent on child care showed that
mothers do more child care than fathers. Models
1 and 2 allowed for random intercepts and xed

8Journal of Marriage and Family
Tab l e 2 . Multilevel Generalized Linear Models Predicting Child-Care Minutes Daily for Mothers and Fathers: 11 Countries,
1965–2012 (Unstandardized Gamma Coefcients)
Model 0 Model 1 Model 2 Model 3
Mothers Fathers Mothers Fathers Mothers Fathers Mothers Fathers
Intercept 4.570*** 3.734*** 4.782*** 4.044*** 4.774*** 4.036*** 4.768*** 3.976***
(0.047) (0.102) (0.038) (0.072) (0.038) (0.073) 0.060 (0.125)
Year (median centered) 0.014*** 0.028*** 0.013*** 0.027*** 0.021** 0.028**
(0.001) (0.001) (0.001) (0.001) (0.007) (0.011)
Higher education 0.199*** 0.345*** 0.207*** 0.354*** 0.191*** 0.368***
(0.010) (0.017) (0.010) (0.017) (0.035) (0.030)
Higher education ×Year 0.005*** 0.003*0.001 0.002
(0.001) (0.002) (0.001) (0.003)
Variance of the random components
Residual 1.190 2.603 1.237 2.965 1.239 2.963 1.236 2.974
Country 0.025 0.113 0.014 0.051 0.014 0.052 0.037 0.163
Year (median centered) 0.001 0.001
Higher education 0.010 0.007
Higher education ×Year 0.001 0.001
AIC 749,065 471,051 739,010 464,934 738,988 464,931 738,481 464,630
Note. Standard errors in parentheses.AIC =Akaike information criterion. Models 0–2 are random intercept models. Model
3 allows random intercepts and random slopes. Models 1–3 control for age, partnership status, number of children, age of
youngest child, employment status, and day of week. Nmothers =68,532; Nfathers =53,739.
*p<0.05. **p<0.01. *** p<0.001 (two-tailed tests).
slopes. In both models, net of covariates, the
positive and signicant coefcients for survey
year (p<.001) conrmed that daily minutes in
child care increased over time for both fathers
and mothers across the 11 countries (H2). To
get a sense of the magnitude of the increases
in caring time, we calculated predicted time in
child care from Model 1 setting at their sample
means the other covariates. While the mean time
the average mother in the 11 countries spent
daily on child care in 1965 was calculated to
be about 54 minutes, it increased to a predicted
104 minutes by 2012. For fathers, the estimates
increased from a scant 16 minutes daily in 1965
to 59 minutes in 2012.
Now, consider the educational gradient.
Adjusting for covariates, both models showed
that more educated mothers (p<.001) and
fathers (p<.001) spent more time on child
care than parents with less education (H1).
The results were consistent with a positive
educational gradient. Comparing Models 0 and
1, note that the variance of the country-level
intercept was reduced when education and time
were included, underscoring the importance
of these variables for mothers’ and fathers’
child-care time. Setting the covariates to their
overall sample means, we estimated that more
educated mothers devoted 18 more minutes to
child care than did mothers with less schooling.
More educated fathers spent 16 more minutes in
child care than other fathers.
In Model 2, the interaction between survey
year and education tests whether the gradient
was increasing (H3A) or decreasing (H3B ), that
is, whether the better educated increased their
child-care minutes more or less than other par-
ents over time. The positive and signicant
interaction between higher education and year
(p<.001 for mothers, p<.05 for fathers) was
consistent with an increase in the education gra-
dient (H3A). Over the decades, more educated
parents in the pooled data adopted intensive
time-use patterns that further differentiated them
from parents with less education. At the begin-
ning of the period, the expected time in child
care was the same for mothers with more and
less schooling, that is, about 50 minutes per
day. By 2012, however, more educated moth-
ers spent an estimated 123 minutes daily on
child care and less educated mothers, 94. The
growing educational gradient among fathers was
even more striking. Reporting 18 minutes a day
in 1965, more educated fathers did only three

Educational Gradients in Parents’ Child-Care Time 9
F . M’ P C-C M D  E  Y   C.
Note. The predictions and condence intervals are calculated from Model 3. The predicted values are adjusted by setting
age, partnership status, number of children, age of youngest child, employment status, and day of week at the overall sample
means.
more minutes of child care than men with less
schooling. By 2012, they were estimated to have
done 74 minutes daily, or 24 more minutes than
their less educated counterparts. Regardless of
schooling, parents spent more time in child care,
but the increase was greater for the more edu-
cated.
Country-Specic Results for Education
To address whether the overall increase in the
education gradient in child care that we observe
in the pooled analysis held for all countries,
Model 3 allowed the intercept and the slopes
for (higher) education, time, and their interac-
tion to vary. Survey year (p<.01) and educa-
tion (p<.001) were statistically signicant for
both mothers and fathers, again indicating both
an increase in parental time (H2) and a positive
educational gradient (H1). When countries were
allowed to differ in slopes, however, the interac-
tion term no longer showed a signicant increase
in the educational gradient over time for either
parent. Apparently, the parent education gap in
child-care time did not widen in all 11 countries.
To better illustrate this nding, we calcu-
lated for each country the predicted minutes
in child care by education and survey year on
the basis of the estimates in Model 3, which
allowed year, education, and their interaction
to vary. In calculating the predicted values, all
other covariates were set to the overall sample
means. In Figures 1 and 2, for mothers and
fathers, respectively, the plotted results show
the changes over time by education, holding all
other covariates at their sample means. The dots
mark the survey years observed. Extrapolating
over all possible time points between 1965 and
2012, we t solid lines that show more educated
parents and dashed lines that show parents with
less education. For each country, any overlap
in the 95% condence bounds for survey year
indicates that educational differences were not
statistically signicant.
In most countries, the overlapping condence
bounds were not consistent with a statistically
signicant educational gradient for mothers at
the .05 level (H1), if only because the disaggre-
gated Ns were small. In no country and no year,
however, were more educated mothers seen
to do less child-care than their less educated
counterparts. Most of the lines in Figure 1
trend upward, consistent with an increase in
mothers’ child-care time over the period (H2).
Furthermore, the increase in child care was
not a new phenomenon, dating to the earliest
surveys (e.g., 1965–1966 in the United States).
France, however, was an anomaly. Contrary
to the increasing-care hypothesis (H2), French
mothers’ child-care time fell visibly over the
years.

10 Journal of Marriage and Family
F . F’ P C-C M D  E  Y   C.
Note. The predictions and condence intervals are calculated from Model 3. The predicted values are adjusted by setting
age, partnership status, number of children, age of youngest child, employment status, and day of week at the overall sample
means.
As for the trends in the gradient, Figure 1
shows that more educated mothers in at least
some countries, notably Slovenia, but also the
United States, Italy, and Norway, increased their
child-care time more than other mothers did
(H3A). Even in France, the educational gradient
seems to have widened over time, because better
educated mothers cut back on their child care
less than their less educated counterparts did.
In Slovenia, the positive gradient increased not
only because better educated mothers stepped
up their child-care time, but also because over
three decades their intensive mothering prac-
tices failed to diffuse to mothers with less
schooling. In other countries (e.g., Denmark),
any differences by education were small, and
mothers with more and less schooling tracked
one another closely over time in increasing their
child care. No support, however, was seen in any
of the 11 countries for the decreasing-gradient
hypotheses (H3B).
Figure 2 reported the predicted values for
fathers in 11 countries. Less educated fathers
were never seen to devote more time to child care
than more educated fathers did (H1), but educa-
tion differences were seldom statistically signif-
icant, probably because of small sample sizes.
A notable exception was France, where a sta-
tistically signicant positive education gradient
emerged in the late 1990s, when more edu-
cated fathers took on more child care than other
fathers did. There was evidence of increases for
men in all countries (H2), even France, where
women decreased their child care. As for trends
in fathers’ educational gradient, some modest
increase seemed to characterize all countries
(H3A). There is no evidence of the theoretical
mechanisms that could lead to a decline in the
education gradient (H3B). Not only do no coun-
tries show a slowing increase in more educated
fathers’ child care; several countries (France,
Norway, Slovenia) give no indication that lower
education groups were beginning to catch up.
In these three countries, the intensive fathering
behavior of more educated fathers had yet to dif-
fuse to less educated men.
D
Both xed- and random-effects models for
11 Western countries conrmed that moth-
ers and fathers increased the time they spent
caring for their children, as anticipated by the
increasing-care hypothesis (H2). France, where
mothers decreased their time in child care, stood
out as an anomaly, but one that was also reported
by Kan et al. (2011), who used a broader mea-
sure of “caring for family members.” In addition

Educational Gradients in Parents’ Child-Care Time 11
to the upward time trend, there was evidence
supporting the hypothesis of a positive edu-
cational gradient (H1) in both random- and
xed-effects models. In general, more educated
mothers and fathers devoted more minutes to
child care each day than less educated ones
did. The gradient was substantively important
in countries such as the United States, where
women with postsecondary schooling spent a
half hour more each day caring for their children
than did other mothers.
Less certain is whether ndings were con-
sistent with the increasing-gradient hypothesis
(H3A). Fixed-effect models found the interaction
between education and survey year to be sta-
tistically signicant; however, random-effects
models (allowing each country’s historical
trends to differ for parents with high education
and low education) did not nd a signicant
education-by-year interaction. If the statisti-
cal results were inconclusive (for reasons to
be discussed), visual inspection of the data
certainly pointed to a trend toward a steeper
gradient for mothers in most countries and for
fathers everywhere. Nowhere had the positive
educational gradient in child care dimin-
ished, as the decreasing-gradient hypothesis
(H3B) suggested. Increasing or stable gradi-
ents argued that the advantages of intensive
parenting continued to accrue to the more
educated elite.
In the unsettled cross-national literature on
parental time use, this study’s most important
contribution may well be showing that the shift
to more intensive parenting could not be traced
only to mothers or only to fathers. With the
notable exception of French mothers, both par-
ents across countries showed similar behavior
in increasing the time spent caring for children.
In several countries (France, Norway, Slovenia),
the child-care time of fathers with lower edu-
cation was unchanged, which suggests that new
fathering ideals had not yet diffused to less
educated men.
Given a comparatively large number of coun-
tries with comparable data, this formal and
systematic analysis has offered greater clarity
than previously existed on cross-national issues
of parental child-care time. Nonetheless, the
existing data fell short of what was needed
to detect statistically signicant trends and
differentials at the country level. For mothers
and fathers, the 95% condence bounds often
overlapped for the two education groups. The
samples were typically too small to support
statistical tests in education-by-gender analy-
ses for mothers and fathers of young children
within country-years. This was illustrated by the
wide condence intervals for the annual U.S.
surveys in the 21st century. Mothers’ condence
intervals also tended to be large in the early
survey years, when relatively few women had
postsecondary educations. Nonetheless, sensi-
tivity tests contrasting incomplete secondary
education against completed secondary or post-
secondary education did not change the results.
Unfortunately, because we had relatively few
time points and cases, modeling any nonlinear
effect of trends in the educational gradient
proved problematic. Conicting trends reported
here for different countries also compromised
efforts to generalize. There was, for example,
the contrarian decline in mothers’ child care
in France, as well as the fact that some less
educated parents (Slovenian mothers and Slove-
nian, French, and Norwegian fathers) did not
conform to the general increase in time with
children. These incongruities worked against
nding statistical signicance for trends. Our
countries reected the expedient of comparable
surveys with multiple observations of harmo-
nized data on key variables. A different set of
countries might have yielded different results.
Trends and differentials might well be clearer
for time spent on the cognitively enriching child
care that cultural theories describe as appealing
to the middle class. Unfortunately, this detail
was not collected in all countries and was not
available in the harmonized data. Similarly,
only a few countries collected data needed to
evaluate how a partner’s education related to the
respondent’s child care.
Across various country contexts, the perva-
sive educational gradient in parental child care,
coupled with the increase in that gradient over
time, provided a compelling explanation for
the historical increases in parental child care
that have been widely reported. Mothers’ and
fathers’ child-care time rose, in part, because
those parents with the highest levels of school-
ing lavished more and more time on their chil-
dren. This is not surprising. Not only are more
educated parents more aware of the develop-
mental payoffs to parental time with children;
their intensive parenting practices conrm their
privileged social status by differentiating them
from parents in lower social classes. Of course,

12 Journal of Marriage and Family
ideas and behavior that take root among the priv-
ileged are known to diffuse eventually to the
less advantaged. In most countries, less educated
parents also increased their time in child care,
albeit not as much as the more educated did.
Although the cultural theories align with this
parental behavior, this is not to deny that a lack of
resources may constrain less educated parents’
ability to act on intensive parenting norms. Any
downward diffusion of intensive parenting has
yet to deter more educated parents’ efforts. They
show no sign of hitting a ceiling on child-care
time or of abandoning practices that might have
been losing their cachet as a marker of higher
social class. Greater time investments in child
care by parents with less schooling may even
encourage the most educated parents to escalate
their own involvement in order to maintain a rel-
ative advantage for their children.
This article builds on neglected cultural
explanations for the increase in parents’ time
in child care. Our focus on social diffusion,
class differentiation, and class-based ideolo-
gies of child rearing suggests hypotheses that
prove largely consistent across countries with
observed parental behavior. This approach
brings a distinct class dimension to parental
time. We know that children whose parents have
less education are disadvantaged in numerous
ways. This research highlights another disad-
vantage for children from lower classes, namely,
the more limited time their parents spend caring
for them. At least in some countries, trends
over time in the educational gradient appear to
exacerbate the intergenerational transmission
of inequality.
The growth in parental time in child care
shows that most parents, whatever their educa-
tion, are prioritizing children in the allocation of
time. Children benet from parental time, and
parents undoubtedly nd children rewarding.
Although parents’ intensive time may be good
for children, we know less about the conse-
quences for adults of the high expectations for
time-intensive parenting. Trends in child care
indicate that mothers are being called on to
reconcile enhanced norms of parenting with
increased labor-force participation. Without
particularly supportive public programs and
employer policies, these conicting demands
on mothers seem to bode poorly for gender
equality, workforce attachment, and occupa-
tional advancement (Abendroth, Huffman, &
Treas, 2014). At the same time, fathers must
accommodate the demands of the “new father”
norm, gender-egalitarian expectations for shared
housework, and continuing breadwinner roles.
Both the causes and consequences of intensive
parenting trends merit further attention.
R
Abendroth, A., Huffman, M., & Treas, J. (2014). The
parity penalty in life course perspective: Moth-
erhood and occupational status in 13 European
countries. American Sociological Review,79,
993–1014. doi:10.1177/000312414545986
Alwin, D. (1988). Changes in qualities valued in
children in the United States, 1964 to 1984. Social
Science Research,18, 195–236.
Badinter, E. (2012). The conict: How modern moth-
erhood undermines the status of women.New
York, NY: Metropolitan Books.
Benasich, A. A., & Brooks-Gunn, J. (1996). Mater-
nal attitudes and knowledge of child-rearing:
Associations with family and child outcomes.
Child Development,67, 1186–1205. doi:10.1111/
j.1467-8624.1996.tb01790.x
Bernal, R., & Keane, M. P. (2011). Child care choices
and children’s cognitive achievement: The case of
single mothers. Journal of Labor Economics,29,
459–512. doi:10.1086/659343
Bianchi, S. M. (2000). Maternal employment and
time with children: Dramatic change or sur-
prising continuity? Demography,37, 401–414.
doi:10.1353/dem.2000.0001
Bianchi, S. M. (2011). Family change and time
allocation in American families. Annals of the
American Academy of Political and Social Science,
638, 21–44. doi:10.1177/0002716211413731
Bianchi, S. M., Robinson, J., & Milkie, M. (2006).
Changing rhythms of American family life. New
York, NY: Russell Sage Foundation.
Bourdieu, P. (1984). Distinction: A social critique
of the judgment of taste (Richard Nice, Trans.).
Cambridge, MA: Harvard University Press.
Bryan, M. L., & Sevilla-Sanz, A. (2011). Does house-
work lower wages? Evidence for Britain. Oxford
Economic Papers,63, 187–210. doi:10.1093/
oep/gpq011
Craig, L., & Mullan, K. (2011). How mothers and
fathers share child care: A cross-national time-use
comparison. American Sociological Review,76,
834–861. doi:10.1177/0003122411427673
Craig, L., Mullan, K., & Blaxland, M. (2010).
Parenthood, policy and work–family time in
Australia 1992–2006. Work, Employment & Soci-
ety,24, 27–45. doi:10.1177/0950017009353778
Craig, L., Powell, A., & Brown, J. E. (2015).
Co-resident parents and young people aged
15–34: Who does what housework? Social Indi-
cators Research,121, 569–588. doi:10.1007/
s11205-014-0643-5

Educational Gradients in Parents’ Child-Care Time 13
Craig, L., Powell, A., & Smyth, C. (2014). Towards
intensive parenting? Changes in the composi-
tion and determinants of mothers’ and fathers’
time with children, 1992–2006. British Journal
of Sociology,65, 555–579. doi:10.1111/1468-
4446.12035
Duyvendak, J. W., & Stavenuiter, M. (Eds.).
(2004). Working fathers, caring men. The Hague,
the Netherlands: Ministry of Social Affairs.
England, P., & Saraff, A. (2010). Educational differ-
ences in parents’ time spent in child care: Culture,
incentives, or income constraints? Paper presented
at the annual meeting of the Population Associa-
tion of America, Dallas, TX.
England, P., & Srivastava, A. (2013). Educational
differences in US parents’ time spent in child
care: The role of culture and cross-spouse inu-
ence. Social Science Research,42, 971–988.
doi:10.1016/j.ssresearch.2013.03.003
Esping-Andersen, G. (1999). Social foundations of
postindustrial economies. Oxford, UK: Oxford
University Press.
Fisher, K., & Gershuny, J. (2013). Multinational time
use study: Users’ guide and documentation (Tech-
nical report). Oxford, UK: University of Oxford.
Gauthier, A., Smeeding, T. M., & Furstenberg, F. F.
(2004). Are parents investing less time in children?
Trends in selected industrialized countries. Pop-
ulation and Development Review,30, 647–671.
doi:10.1111/j.1728-4457.2004.00036.x
Gimenez-Nadal, J. I., & Molina, J. A. (2013). Parents’
education as a determinant of educational child
care time. Journal of Population Economics,26,
719–749. doi:10.1007/s00148-012-0443-7
Gimenez-Nadal, J. I., & Sevilla, A. (2012). Trends
in time allocation: A cross-country analysis.
European Economics Review,56, 1338–1359.
doi:10.1016/j.euroecorev.2012.02.011
Gracia, P. (2014). Fathers’ child care involvement and
children’s age in Spain: A time use study on dif-
ferences by education and mothers’ employment.
European Sociological Review,30, 137–150.
doi:10.1093/esr/jcu037
Gupta, S. (2006). Her money, her time: Women’s
earnings and their housework hours. Social
Science Research,35, 975–999. doi:10.1016/
j.ssresearch.2005.07.003
Guryan, J., Hurst, E., & Kearney, M. (2008). Parental
education and parental time with children. Journal
of Economic Perspectives,22, 23–46. doi:10.1257/
jep.22.3.23
Hays, S. (1996). The cultural contradictions of moth-
erhood. New Haven, CT: Yale University Press.
Henwood, K., & Procter, J. (2003). The “good
father”: Reading men’s accounts of paternal
involvement during the transition to rst-time
fatherhood. British Journal of Social Psychology,
42, 337–355. doi:10.1348/014466603322438198
Hill, J., Waldfogel, J., Brooks-Gunn, J., & Han, W.
(2005). Maternal employment and child develop-
ment: A fresh look using newer methods. Devel-
opmental Psychology,41, 833–850. doi:10.1037/
0012-1649.41.6.833
Hofferth, S. L., & Sandberg, J. (2001). How Ameri-
can children spend their time. Journal of Marriage
and Family,63, 295–308. doi:10.1111/j.1741-
3737.2001.00295.x
Hook, J. L., & Wolfe, C. M. (2012). New fathers?
Residential fathers’ time with children in four
countries. Journal of Family Issues,33, 415–450.
doi:10.1177/0192513X11425779
Kalil, A., Ryan, R., & Corey, M. R. (2012). Diverging
destinies: Maternal education and the development
gradient in time with children. Demography,49,
1361–1383. doi:10.1007/s13524-012-0129-5
Kan, M. Y., Sullivan, O., & Gershuny, J. (2011).
Gender convergence in domestic work: Discerning
the effects of interactional and institutional barri-
ers from large-scale data. Sociology,45, 234–251.
doi:10.1177/0038038510394014
Kim, A. J. (2014). Gender boot camp for Korean
immigrant patriarchs: Father school and the new
father conversion process. Sociological Perspec-
tives,57, 321–342. doi:10.1177/073112141452
3734
Kimmel, J., & Connelly, R. (2007). Mothers’ time
choices: Caregiving, leisure, home production,
and paid work. Journal of Human Resources,42,
643–681. doi:10.3368/jhr.XLII.3.643
Laird, R. D., Pettit, G. S., Bates, J. E., & Dodge, K.
A. (2003). Parents’ monitoring knowledge and
adolescents’ delinquent behavior: Evidence of
correlated developmental changes and reciprocal
inuences. Child Development,74, 752–768.
doi:10.1111/1467-8624.00566
Lareau, A. (2000). Social class and the daily lives of
children: A study from the United States. Child-
hood,7, 155–171. doi:10.1177/090756820000700
2003
Lareau, A. (2003). Unequal childhoods: Class,
race, and family life. Berkeley, CA: University of
California Press.
Leibowitz, A. (1977). Parental inputs and children’s
achievement. Journal of Human Resources,12,
242–251. doi:10.2307/145387
Lugo-Gil, J., & Tamis-LeMonda, C. S. (2008).
Family resources and parenting quality: Links
to children’s cognitive development across the
rst 3 years. Child Development,79, 1065–1085.
doi:10.1111/j.1467-8624.2008.01176.x
Lyness, K. S., Gornick, J. C., Stone, P., & Grotto, A.
R. (2012). It’s all about control: Worker control
over schedule and hours in cross-national context.
American Sociological Review,77, 1023–1049.
doi:10.1177/0003122412465331

14 Journal of Marriage and Family
Mancini, A. L., & Pasqua S. (2009). Asymmetries
and interdependencies in time use between Ital-
ian spouses (Working Paper No. 12/2009). Torino,
Italy: Center for Household Income Labor and
Demographic Economics.
Marsiglio, W., Amato, P., Day, R. D., & Lamb,
M. E. (2000). Scholarship on fatherhood in the
1990s and beyond. Journal of Marriage and
Family,62, 1173–1191. doi:10.1111/j.1741-3737.
2000.01173.x
Maume, D. J. (2011). Reconsidering the temporal
increase in fathers’ time with children. Journal
of Family and Economic Issues,32, 411–423.
doi:10.1007/s10834-010-9227-y
Monna, B., & Gauthier, A. (2008). Review of the
literature on the social and economic determi-
nants of parental time, Journal of Family and Eco-
nomic Issues,49, 634–653. doi:10.1007/s10834-
008-9121-z
Organisation for Economic Co-operation and Devel-
opment. (2014). OECD Family Database. Paris:
Author. Retrieved from http://www.oecd.org/
social/family/database.htm
Poortman, A. R., & Van der Lippe, T. (2009).
Attitudes toward housework and child care and
the gendered division of labor. Journal of Mar-
riage and Family,71, 526–541. doi:10.1111/
j.1741-3737.2009.00617.x
Ramey, G., & Ramey, V. A. (2009). The rug rat
race (Brookings Papers on Economic Activity No.
41). Washington, DC: Economic Studies Program,
Brookings Institution.
Rogers, E. M. (1962). Diffusion of innovations.New
York, NY: Free Press.
Rowe, M. (2008). Child-directed speech: Relation
to socioeconomic status, knowledge of child
development and child vocabulary skill. Journal
of Child Language,35, 185–205. doi:10.1017/
S0305000907008343
Sayer, L., Bianchi, S., & Robinson, J. (2004). Are par-
ents investing less in children? American Journal
of Sociology,110, 1–43. doi:10.1086/386270
Sayer, L., Gauthier, A., & Furstenberg, F. J. (2004).
Educational differences in parents’ time with chil-
dren: Cross-national variations. Journal of Mar-
riage and Family,66, 1152–1169. doi:10.1111/
j.0022-2445.2004.00084.x
Schulz, F., & Grunow, D. (2012). Comparing diary
and survey estimates on time use. European Soci-
ological Review,28, 622–632. doi:10.1093/esr/
jcr030
Sullivan, O. (2010). Changing differences by
educational attainment in fathers’ domestic
labour and child care. Sociology,44, 716–733.
doi:10.1177/0038038510369351
Sullivan, O. (2013). What do we learn about gen-
der by analyzing housework separately from child
care? Some considerations from time use evidence,
Journal of Family Theory and Review,5, 72–84.
doi:10.1111/jftr.12007
Sullivan, O., Billari, F., & Altintas, E. (2014).
Fathers’ changing contributions to child care and
domestic work in very low fertility countries: The
effect of education. Journal of Family Issues,35,
1048–1065. doi:10.1177/0192513X14522241
Sullivan, O., & Gershuny, J. (2001). Cross-national
changes in time use: Some sociological (hi)stories
re-examined. British Journal of Sociology,52,
331–347. doi:10.1080/00071310120045015
Svab, A. (2013). “I only have to ask him and he
does it …” Active fatherhood and (perceptions of)
division of family labour in Slovenia. Journal of
Comparative Family Studies,44, 57–78.
Tucker-Drob, E. M., & Harden, K. P. (2012).
Early childhood cognitive development and
parental cognitive stimulation: Evidence for
reciprocal gene–environment transactions. Devel-
opmental Science,15, 250–259. doi:10.1111/j.
1467-7687.2011.01121.x
Vandell, D. L., Belsky, J., Burchinal, M., Steinberg,
L., Vandergrift, N., & NICHD Early Child Care
Research Network. (2010). Do effects of early
child care extend to age 15 years? Results from
the NICHD study of early child care and youth
development. Child Development,81, 737–756.
doi:10.1111/j.1467-8624.2007.01021.x
Weber, M. (1978). Economy and society: An outline
of interpretive sociology. Berkeley, CA: University
of California Press. (Originally published 1922)
Yamauchi, C., & Leigh, A. (2011). Which chil-
dren benet from non-parental care? Economics
of Education Review,30, 1468–1490. doi:10.1016/
j.econedurev.2011.07.012