Time Allocation of Parents and Investments in Sons and Daughters

Abstract

This paper examines the time parents spend with their children using the PSID Child Development Supplement and the American Time Use Survey in order to disentangle whether parent-child gender differences in time use result from parental preferences, especially a preference by fathers for sons, or rather from boy-girl differences in child production functions. We find that married fathers spend significantly more time with sons than with daughters. We also find that same-sex parent-child time in stereotypically gendered activities is pervasive and becomes increasingly evident in the teen years. Single mothers spend more time with daughters relative to sons than married mothers, suggesting that parents prefer time with a same-sex child, rather than that boys need more time inputs than girls. By examining the leisure and housework time of married mothers, we find no evidence of increased bargaining power of mothers of sons that would support a preference by fathers for sons as the sole explanation. This suggests either that mothers also prefer fathers to spend more time with boys, or that children need more time from the same-sex parent.

Full text

Preliminary - Please contact authors before citing
Time Allocation of Parents and Investments in Sons and Daughters*
Shelly Lundberg
Department of Economics,
University of Washington
and IZA, Bonn
lundberg@u.washington.edu
Sabrina Wulff Pabilonia
U.S. Bureau of Labor Statistics
2 Massachusetts Ave., NE Rm. 2180
Washington, DC 20212
Pabilonia.Sabrina@bls.gov
Jennifer Ward-Batts
Claremont McKenna College
500 E 9
th
St
Claremont, CA 91711
jennifer.ward-batts@mckenna.edu
November 2006
* All views expressed in this paper are those of the authors and do not necessarily reflect the
views or policies of the U.S. Bureau of Labor Statistics. Lundberg wishes to thank the Castor
Professorship and NIH/NICHD (R01 HD042785) for financial support. The authors are grateful
to Dori Allard, Laura Argys, Kristin Dale, Anastasiya Osborne, Lucie Schmidt, and Jay Stewart,
for helpful comments and to Simona Blaziene for excellent research assistance.

Time Allocation of Parents and Investments in Sons and Daughters*
Abstract:
This paper examines the time parents spend with their children using the PSID Child
Development Supplement and the American Time Use Survey in order to disentangle whether
parent-child gender differences in time use result from parental preferences, especially a
preference by fathers for sons, or rather from boy-girl differences in child production functions.
We find that married fathers spend significantly more time with sons than with daughters. We
also find that same-sex parent-child time in stereotypically gendered activities is pervasive and
becomes increasingly evident in the teen years. Single mothers spend more time with daughters
relative to sons than married mothers, suggesting that parents prefer time with a same-sex child,
rather than that boys need more time inputs than girls. By examining the leisure and housework
time of married mothers, we find no evidence of increased bargaining power of mothers of
sons that would support a preference by fathers for sons as the sole explanation. This suggests
either that mothers also prefer fathers to spend more time with boys, or that children need more
time from the same-sex parent.
1

I. Introduction
In many developing countries, parental son preference is openly acknowledged and has
measurable consequences for child well-being. With increasing availability of ultrasound
technology, sex-selective abortion has substantially increased the ratio of young boys to young
girls in Korea (Park and Cho, 1995). In addition, Strauss and Thomas (1995) find that girls in
South Asia are disadvantaged in their access to nutritional and health inputs. In most developed
countries, however, there is little apparent gender discrimination in parental investments in
children: boys and girls receive equivalent education, bequests, and transfers (Taubman, 1991).
Since son preference is linked to cultural norms that place a higher value on having sons relative
to daughters, such as a reliance on male children for old age support, it is not surprising that
increasing economic gender equality in industrialized societies should lead to more equal
treatment of sons and daughters. Meadows et al. (2005) survey a wide variety of quality-of-life
indicators for boys and girls in the United States, and find that overall levels of wellbeing are
equivalent.
However, a substantial body of research has documented consistent discrepancies
between the behavior of parents of sons and parents of daughters in wealthy, non-traditional
societies—child gender appears to have a significant effect on outcomes ranging from the marital
status of mothers (Dahl and Moretti, 2004) and paternal happiness (Kohler et al., 2004) to
parents’ political views (Oswald and Powdthavee, 2005). The effect of child gender on the time
allocation of parents is particularly well-documented. Fathers spend more time with sons than
with daughters (Yeung et al., 2001), and more time with their children if at least one of them is a
son (Harris and Morgan, 1991; Mammen, 2005). Several recent studies (e.g., Lundberg and
Rose, 2002; Choi, Joesch, and Lundberg, 2005) have also found that men in industrialized
nations, including the U.S., increase their market work hours more after the birth of a son,
relative to a daughter. What remains unclear is the source of these wide-ranging parental
responses to child gender—whether they result from gender bias in parental preferences (e.g.,
fathers prefer boys) or from boy-girl differences in child production functions (e.g., boys need
fathers, or boys demand more time in general).
1
Different sources of the bias will have different
1
These competing hypotheses, as well as the empirical evidence on child gender effects, are further discussed in
Lundberg (2005b).
2

implications. For example, if fathers prefer boys but mothers do not, then the household
bargaining position of mothers may increase if they have a boy, resulting in her having more
leisure time or less housework time.
In this paper, we use two recent sources of time-use data to investigate differences in
investments of parental time in sons and daughters aged 0 to 17 years in the United States, and
the possible motivations for such differences. The Child Development Supplement of the Panel
Study of Income Dynamics (PSID-CDS) provides time diaries for children aged 0 to 12 years in
1997, including the time spent interacting with, and in the presence of, parents for a detailed set
of activities. With these data, we can examine not only the relative quantities of time mothers
and fathers spend with sons and daughters, but also how this time is spent. The new American
Time-Use Survey (ATUS) collects a daily time diary from one individual (aged 15 or over) in a
representative sample of U.S. households in 2003, 2004, and 2005. These diaries allow us to
document the time allocation of parents in more detail than has previously been possible with
U.S. data, including not only time with children and in market work, but also in leisure and
housework. In addition, the ATUS provides a sample of time diaries for children aged 15 and
over that we can use to supplement what we learn from the PSID diaries for younger children.
Previous studies using PSID-CDS (Yeung et al., 2001) and ATUS (Mammen, 2005) have
found that fathers spend more time with sons than with daughters. Our goal in this paper is two-
fold: 1) to examine how child gender differences in time spent with both mothers and fathers
vary by parental characteristics and across activities and 2) to disentangle the possible
explanations for these differences by investigating parents’ non-childcare time.. The results of
this work may have important implications for the wellbeing of both children and parents. For
example, if fathers spend more time playing sports with their sons than daughters, this may give
boys a developmental advantage. Extra time fathers spend with boys may crowd out father’s
work or leisure time. If men value time with sons and/or if they replace more of the wife’s
childcare time when they have a son, then married mothers may get more leisure time or spend
less time in housework if they have a son relative to a daughter.
From both the child time diaries of PSID-CDS and the adult time diaries of ATUS, we
confirm that married fathers
2
do indeed spend more time with sons than with daughters, and that
2
While we examine the time-use of single mothers, we will not examine the time-use of single fathers due to small
sample sizes.
3

much of this additional time is spent in sports and recreational activities. College-educated
married fathers also spend more time with sons than daughters in educational activities. In one-
child and two-child families, married mothers in the ATUS spend significantly more secondary
time with pre-school aged daughters than sons while the mothers of CDS daughters spend more
direct time with them (shopping and doing housework) than with sons. We also find evidence
that single mothers spend more time with daughters than sons, and that they spend more time in
market work and less time in household work if they have a son rather than a daughter. Using
the ATUS teen diaries, we also find that both married and single mothers spend more time with
teen daughters than teen sons—primarily doing housework, shopping, and social activities.
In the ATUS, fathers in two-son families spend substantially more time in both direct and
indirect childcare time than fathers of two daughters, and this is clearly offset by significantly
less leisure time for these fathers. Unlike previous studies, we do not find evidence of increased
market work by fathers of sons relative to fathers of daughters. Overall, we find that same-sex
parent-child time in stereotypically gendered activities is pervasive and becomes increasingly
evident in the teen years. We find little support for the hypothesis that these investments are
driven purely by father-son preference and that child gender affects the relative bargaining power
of spouses: mothers of sons are not significantly better off than mothers of daughters in terms of
their leisure consumption or household chores.
II. The Effects of Sons and Daughters on Parental Behavior in the U.S.
A. Previous Literature
Recent research by sociologists and economists has found significant, but generally
small, differences between the behavior of parents of sons and parents of daughters in
industrialized countries. Lundberg (2005b) reviews the empirical evidence for child gender
effects on marital and relationship stability, fertility, and the time allocation of parents. Boys are
more likely to live with their father than are girls, both because their parents are more likely to
marry or stay married, and because paternal custody of boys is more common following divorce.
Using U.S. data, Lundberg and Rose (2003) find that an unmarried mother is more likely to
marry the father of her child following the birth of a boy, and Dahl and Moretti (2004) report that
a first-born son has a positive effect (2.6 percent) on the probability that his mother has ever been
4

married. Fathers have been found to spend more time with, and to be more involved with, sons
than daughters (Harris and Morgan, 1991; Lundberg, McLanahan, and Rose, 2005; Mammen,
2005; Morgan et al., 1988; Yeung et al., 2002). Recent studies using the Panel Study of Income
Dynamics and the German Socioeconomic Panel (Lundberg and Rose, 2002; Choi, Joesch, and
Lundberg, 2005) have also found differential labor supply responses by fathers in the U.S. and
Germany—in general, fathers of boys increase their work hours more than fathers of girls.
In general, the observed parental responses to child gender are consistent with a positive
effect of sons on marital surplus, which could explain both the increase in marital stability and,
as a result, greater specialization within the household. This son-induced marital surplus
premium could arise from two sources: gender bias in parental preferences or boy-girl
differences in child production functions. If fathers, for example, have biased preferences and
receive greater utility from living with sons than daughters, and if maternal custody after
separation is customary, then the birth of a son increases the value of marriage relative to
separation more than does the birth of a daughter. This increase in marital surplus will make the
relationships of parents of sons more stable and, if household resources are distributed between
husbands and wives by bargaining, make mothers of sons better off than mothers of daughters.
Sons will increase the relative bargaining power of their mothers because fathers have a greater
desire to maintain contact with sons than with daughters.
3
Alternatively, marital surplus may be greater for a couple with sons if paternal time is a
more valuable input to the care of boys than of girls. This may be the case if the set of attributes
or skills that parents wish boys and girls to develop differ, or because boys and girls have
different developmental requirements. Morgan et al. (1988) refer to the “expectation that fathers
will teach their sons to play and appreciate sports” and argue that “(r)esearch on child
development supports the notions that fathers have a special role to play in the emotional
development of sons and that marital disruption and the absence of the father are more harmful
for boys than for girls” (p. 112). If sons and daughters require different care, or if paternal time
is believed to be more important to the development of sons, then child gender could affect not
only the total time parents spend on childcare, but also the relative quantities of paternal and
maternal time and the way in which childcare time is disaggregated into specific activities such
3
In the context of a divorce-threat cooperative bargaining model (McElroy and Horney, 1981; Manser and Brown,
1980), father-son preference combined with maternal custody implies that the father of a son will have a relatively
lower threat point in marital bargaining than the father of a daughter.
5

as playing, reading, and physical care. Although the “preference” explanation for child gender
effects on relationship stability and other parental outcomes predicts that mothers of sons will be
better off, the implications of the “constraint” story are not so clear. If mothers value the
“quality” of their sons, then they may be willing to transfer intrahousehold resources to fathers to
induce them to stay in the marriage (Lundberg, 2005b). These alternative explanations imply
that child gender differences in indicators of mothers’ well-being, such as leisure time, may help
us distinguish between preference-driven parent responses to sons and daughters and those that
reflect differences in child production functions.
4
Surveys that restrict time use data to hours of market work do not allow us to test
between these alternative explanations. Lundberg (2005a) finds significant effects of child
gender on the labor supply of both mothers and fathers in the National Longitudinal Survey of
Youth, 1979 (NLSY79), and these effects are opposite at the two ends of the education
spectrum—boys increase specialization among parents with less than a high school education but
reduce specialization among the college-educated. These heterogeneous responses to sons and
daughters across education groups could also be either constraint- or preference-based. As
desired child quality increases, there may be a bias towards same-sex parental inputs in the child
production function. Alternatively, if fathers prefer sons and a son increases his mother’s
bargaining power in the household, highly-educated women may expect a high payoff to
additional work experience and choose continuous employment while low-wage women choose
to increase their leisure consumption. Either parental time diary data or expenditure data that
includes private goods is required to test whether the preference-bargaining story is valid.
The recent availability of the American Time Use Survey (ATUS) data has generated a
flurry of studies on how parents spend their time and how their time allocation relates to gender
of children. Kalenkoski et al. (2005) use both the 2003 ATUS and the U.K. Time-Use Study
data to examine parental childcare and market work time in single-parent, cohabiting, and
married-couple households. They find that single fathers in the U.S. spend more primary care
time with their children on weekdays and less “passive” time with them on weekends compared
4
An alternative explanation of differential parental treatment of sons and daughters based on evolutionary
considerations is provided by a version of the Trivers-Willard hypothesis that natural selection will favor an ability
to adjust child sex ratios to parental condition. This variant argues that reproductive success is increased by
investing more in sons when conditions are good and more in daughters under adverse conditions. The related
hypothesis that higher status parents invest more in sons is discussed by Keller et al. (2001), who find no evidence
to support it using outcome measures from the PSID-CDS, including breast-feeding, interviewer-reported warmth,
and total parental hours with the child.
6

to married and cohabiting fathers. They also find that single mothers in the U.S. spend more
time in market work on weekdays than married or cohabiting mothers
Mammen (2005) uses the 2003 ATUS data to examine the effect of children’s gender on
time fathers spend with their children. She examines the effects of being a boy, of being the
oldest boy in the household, and of having a brother, for both boys and girls, on a child’s receipt
of father’s time. She finds that child gender as well as the gender composition of the sib-ship
affects the time a child spends with his or her father. Being a boy or being the oldest boy in the
household both increase the child’s time with the father. Girls with brothers spend more time
with their father than girls without brothers, due mainly to an increase in time watching TV with
their father.
In this paper, we use 2003 - 2005 ATUS data and 1997 PSID-CDS data to examine the
time use of married parents and single mothers of minor children aged 0-12. We examine the
behavior of mothers and fathers separately, and include not just childcare time, but – in ATUS
data – also leisure, household chores, and market work. We allow child gender differences in
parental time allocation to vary by the age of the youngest child and the education of parents.
This allows us to address parental time resources allocated to boys relative to girls in both
datasets, and in the ATUS, possible effects of child gender on the relative bargaining power and
well-being of mothers and fathers. ATUS also provides time diaries for an older group of
children than the PSID-CDS, and we examine the time that ATUS respondents aged 15 to 17
spend with their parents in various activities.
B. Child Time Diaries from the Panel Study of Income Dynamics Child Development
Supplement
There are two basic methods of documenting the time parents spend with children—
collecting time diaries for adults in which childcare is one possible activity, or collecting time
diaries for children, recording who is with them or engaged in activities with them. In 1997, a
supplement to the Panel Study of Income Dynamics collected data on children’s time use from a
nationally representative sample of families that had at least one child between the ages of 0 and
12. A time diary for a random preassigned weekday and weekend day was completed for up to
two children in each family, preferably by the child’s primary caregiver (usually the mother) and
7

in cooperation with the child when possible (Yeung et al., 2001). The respondents completed a
time grid for the 24 hours between midnight and midnight for each of these days, reporting the
child’s activities, the time an activity began and ended, who was doing the activity with the
child, who was present but not directly involved in the activity, and what else the child was
doing. The time diary method of documenting time use has been found to be more accurate than
other methods such as asking questions about the time devoted to a fixed set of activities (Juster,
1985).
These data have provided a wealth of information about how young American children
spend their time (Hofferth and Sandberg, 2001), and about parental involvement with children.
Yeung et al. (2001) analyzed father’s time with children in intact families, and reported that boys
spend significantly more time with their fathers in play and companionship activities on
weekdays than do girls.
5
In this section, we use a similar sample of 1,754 children who were
living with both biological parents and had completed both a weekday and weekend diary in
order to compare the average times that boys and girls spent per day with each of their parents in
different activities. To the extent that gender affects the likelihood that a child is living with both
parents, these differences may be biased by parent selectivity. For the moment, we assume that
child gender is random even in a sample of intact households, but we will return to this issue
later.
Child activities in which parents are directly involved are divided into six groups—
household work and obtaining goods and services (shopping), personal needs and care,
educational activities, social and organizational activities, sports and active leisure, and passive
leisure. Table 1 reports the mean minutes per day that boys and girls spend in these activities
with mothers and fathers, and also the time spent with parents who were not directly involved in
the child’s activities (secondary time). We use a weighted average of the child’s weekday and
weekend diary time to calculate average daily minutes.
The first panel of Table 1 shows that mothers spend more time directly involved in the
activities of daughters than of sons (about 17 minutes per day) and that most of this extra time
consists of housework and shopping. The gender gap in direct time is partially offset by the
additional secondary time mothers spend with sons, so that the child gender difference in
5
The multivariate analyses of Yeung et al. (2001) control for factors such as parental work hours, which may be
endogenous with respect to child gender.
8

mother’s total time is not significant. Fathers, in contrast, spend about 14 minutes per day more
in direct time with sons than daughters, and also significantly more total time. The second and
third panels report similar results for subsamples of parents with and without a college degree.
More-educated mothers devote more direct time to daughters across a range of activities, such
that none of the individual gender differences are significant, while less-educated mothers
concentrate their extra time with daughters in housework, shopping, and passive leisure.
College-educated fathers spend an additional half hour per week with sons, compared to
daughters, on educational activities and an additional hour and a half on sports and active leisure
activities, and the child gender gap in both direct time and total time with fathers is large and
significant. The extra time spent with sons by less-educated fathers is smaller (about 2 hours per
week in total, rather than 4 hours) and not significantly concentrated in specific activities.
The gender pattern of parent-child activities that emerges from these data is quite clear.
Mothers spend more direct time with daughters than with sons, particularly in stereotypically
female activities, but do not spend significantly more time with them overall. The additional
secondary time that mothers spend with sons is concentrated in the active leisure category (not
reported)—mothers go shopping with their daughters and attend their sons’ sporting events.
Fathers, on the other hand, devote more direct as well as secondary time to sons, and this is
particularly true of college-educated dads. Some of this father-son time is, not surprisingly,
spent on sports, but there is also a significant difference in the weekly time spent by more-
educated fathers on educational activities with sons and daughters.
These simple descriptive results from recent child time diaries suggest greater
investments of time in sons, particularly by more-educated fathers, but parental motivation for
child-gender time differences is not always apparent from the data. Father-son time engaged in
sports or hobbies can yield both parental enjoyment and child skills, though the time that fathers
spend helping sons with homework or reading to them is probably intended as an investment.
When mothers shop and do housework with daughters, this can be regarded as an investment in
domestic skills that may be particularly valuable to daughters, or mothers may prefer interacting
with daughters in some of these activities. There are two possible, but not mutually exclusive,
explanations for the same-sex pattern of direct parent-child activities: mothers may prefer
interacting with daughters and fathers with sons or, if the desired skills or attributes of daughters
are different from those of sons, it may be efficient for the parent most proficient in manly skills
9

to spend more time with male offspring. In the recent literature on child gender effects on
parental behavior in developed countries, the difficulty of distinguishing empirically between
“preference” and “boy-girl differences in child production functions” explanations for
sons/daughter differences in time spent with parents has been a continuing theme.
A future draft of this paper will include regression analysis of the PSID-CDS data, as
well as some analysis of time diaries of children in single-mother households.
C. Parent Time Diaries from the American Time Use Survey
In 2003, the ATUS began collecting 24-hour time diaries every day of the year from one
respondent aged 15 and older in each household from the sample of households exiting the
Current Population Survey, except for the day before significant holidays. The overall response
rate in each year (2003-2005) was about 57 percent. Respondents were asked to report their own
activities from 4:00 AM the previous day to 4:00 AM the day of the interview as well as who
they were with and the location of the activity. The length of each activity was not limited.
Respondents were also asked to report secondary childcare time when they were involved in
another primary activity but still had responsibility for their children.
We analyze pooled 2003-2005 data from the ATUS using samples of married parents and
single mothers with own household children under the age of 13 in order to examine the effects
of child gender on parental time use.
6
With child diaries, the comparison of boys and girls is
straightforward, but parental time use may depend upon the gender composition of all children in
the household. There are many possible ways to specify the effects of child gender for families
with both boys and girls (e.g., birth order may matter, and parental behavior may depend upon
the presence of at least one boy or girl, rather than the sex ratio). First, we use samples of
parents with only one child. The first sample includes parents with only one child aged 0 to 12
in their household (1,620 married mothers, 1,462 married fathers, and 924 single mothers). The
second sample includes parents with two children aged 0 to 12 in their household who are of the
same gender (949 married mothers, 850 married fathers, and 293 single mothers).
7
Choosing
samples with only same-sex children simplifies the analysis, but may introduce selection bias if
6
The sample is restricted to exclude parents with additional older children living with them.
7
We exclude individuals who have more than ninety minutes of unaccounted time.
10

the parents’ decision to have additional children (conditional on the gender composition of those
they already have) is correlated with attitudes and preferences that influence their parenting. For
example, it is possible that parents with two girls (i.e. who have to date not had a third child)
may be more (or less) child-oriented than parents with two boys. Angrist and Evans (1998) do
not find any difference between the average propensity to have a third child for parents with two
boys and two girls, but this does not rule out such a correlation of unobservables.
8
Thus, we also
examine households with one or more children under age 13, irrespective of gender, and use the
sex of the eldest child as a more exogenous child gender measure. Restricting the sample to
married parents may also result in selection bias if child gender affects marital stability, but these
effects appear to be small for recent cohorts (Dahl and Moretti, 2004) and will be limited in our
regression analysis that controls for parental characteristics. We also examine a sample of single
mothers. Finding that married mothers spend more time with daughters relative to sons and that
married fathers spend relatively more time with sons would lend support for the hypothesis that
same-sex specialization is desirable in child production functions. On the other hand, if single
mothers spend more time with sons relative to daughters compared to married mothers, this
would lend support for the hypothesis that sons need more parent time.
One focus of our analysis is parental childcare time. “Childcare” is the time that a parent
reports childcare as the parent’s main activity. “Secondary childcare” is the time a parent reports
that they are responsible for a child(ren) under age 13, but this is not their main activity.
Secondary childcare time may be nurturing and may also contribute to skill acquisition (eg. a
parent may report making meals as the primary activity but their children may be helping to
cook), and thus could be important to child development. We also examine parental “time with
children,” which is any time that the child is in the room, whether the parent is directly engaged
in an activity with them or not. We also examine the time mothers and fathers report in market
work, household work, and leisure. Market work time from the time diaries includes work on all
paid jobs. Leisure time includes time spent on socializing, recreation and personal care,
including sleep. A measure of child-free leisure—time without any children under age 18
present—yields patterns that are similar to total leisure and thus are not reported here.
9
8
However, Dahl and Moretti (2004) find a small degree of “son preference” in parity progression.
9
The presence of children was determined from the ATUS “who” file.
11

C. 1. Descriptive Analysis
Table 2 presents the mean time spent by mothers and fathers in childcare and other
activities, including leisure, household work, and market work, by child gender. These times are
minutes per day, based on weighted averages of weekday reports and weekend and holiday
reports (hereafter referred to as weekend), and so are representative of time use during the week.
Significant differences between time spent with sons and time spent with daughters are indicated
in bold. All estimates throughout the paper have been weighted using the ATUS respondent
sample weights.
Panels 1 and 2 of Table 2 report the time use of mothers and fathers with one child in two
age categories — under the age of 6 and from age 6 to 12. For parents with only one child under
age 6, the only significant difference between the time use of married parents with sons versus
daughters is the substantially higher recreational childcare time of married fathers of sons.
Recreational childcare represents activities such as “horsing” around, attending child’s events,
and teaching your child to ride a bike. Single mothers spend a half hour less on household work
if they have sons versus daughters. In families with a single child aged 6-12, married mothers
spend more time in recreational childcare of sons while married fathers spend more overall time
with sons than daughters. Single mothers spend significantly less time with their child if they
have a son rather than a daughter.
Panel 3 of Table 2 reports the average time use of mothers and fathers of two children of
the same sex under age 13. Among two-parent two-child families, the sons’ advantage in main
activity childcare time and total time with fathers is large and significant. About half of the
difference in main activity childcare time for fathers is attributable to greater recreational time
with sons relative to daughters. Fathers also spend significantly more time in secondary
childcare and more total time with sons than with daughters. Fathers of sons also report a sizable
deficit in leisure time, relative to fathers of daughters. Married mothers spend significantly more
time in secondary childcare if they have daughters rather than sons. Single mothers spend
significantly less time in main-activity childcare if they have sons rather than daughters.
In Panel 4, we examine two-parent and single-mother families with one or more children,
and report mean times for those families in which the eldest child is a girl versus a boy. This
approach allows us to avoid the potential bias due to endogeneity in the number and spacing of
children in the other samples. We find that married fathers spend significantly more time in
12

childcare and with their children overall if their eldest child is male rather than female, and this
difference is largely attributable to greater recreational childcare time. They also spend more
secondary time with children if their eldest child is male rather than female. They take slightly
less leisure time if their eldest child is male rather than female. Single mothers spend
significantly more time with their children if the eldest is female rather than male.
C. 2. Parent Time-Use Regressions
Table 3 presents the coefficients on a ‘son’ dummy from OLS time-use equations for
parents with one child between the ages of 0 and 12, and Table 4 reports similar results for the
parents of two same-sex children. The dependent variables are minutes per day that the parent
reports engaging in each activity. Each equation also includes controls for parental age and age
squared (for both parents in two-parent families), highest level of education (high school
graduate, some college, college degree, or advanced degree with less than high school degree the
omitted group), race (black or other race with white being the omitted group), Hispanic ethnicity,
age of youngest child (3-5 or 6-12 with 0-2 being the omitted group), and region of residence.
Each equation also includes controls for temporal factors that could influence time use, such as
the survey year, season, and whether the time diary refers to a weekend day. Sample means are
shown in Appendix Table A2. We also estimate models that include an interaction of the son(s)
dummy and either the age of the youngest child being older than 5 or the responding parent
having a college degree.
10
In one-child families, there are no significant differences in the main-activity childcare
that married mothers provide to sons versus daughters, or in the total amount of time that
mothers spend with boys relative to girls. They do spend a half hour less time in secondary
childcare of a son relative to a daughter if the child is younger than six, but spend a half hour
more time with a son than a daughter if the child is elementary school aged. Fathers, however,
spend significantly more time in recreational childcare for sons than daughters, and the interacted
model indicates that this effect comes from paternal care of a child under age 6. Further, this
10
Zero time spent in a particular activity may be due to a parent’s non-participation in that activity on all days, or in
a parent participating on some days and not others. If all parents spend time in all of the activities we examine at
some point in time, and if the day, season, and year we observe the parent’s time use is random (i.e., is independent
of the characteristics determining frequency of participating in the activity), then OLS will provide consistent
estimates of the determinants of time use. If the zeroes are the result of infrequency of time allocated to a particular
activity, then the tobit model would be misspecified (see Cragg 1971; Blundell and Meghir 1987). To the extent that
some parents never spend time in a particular activity, then OLS estimates may be biased.
13

effect appears to be limited to fathers without a college degree. Fathers’ increased time with
sons is not reflected in any significant decreases in time spent on other activities: men with a son
work less on the job, but the effect is not significant. Likewise, married mothers do not spend
more or less time on leisure, market work, or household work if they have sons rather than
daughters.
Single mothers, however, spend about three-quarters of an hour per day more in market
work, and about a half hour less doing housework if they have a son rather than a daughter.
Models with interaction terms indicate that single mothers without a college degree spend less
time doing housework and more time in market work if they have a son rather than a daughter.
Single mothers also spend less overall time with their child if the child is male rather than
female, and this result appears to be concentrated mostly among single mothers with no college
degree.
Table 4 presents regression results for families with two children of the same sex. In
two-parent families with two same-sex children, the dominant effect of two sons versus two
daughters is substantially more father’s main activity childcare, secondary childcare, and time
with children combined with a significant reduction in his leisure time. As seen previously, a
substantial fraction of fathers’ additional childcare time with sons is in recreational childcare.
Fathers spend nearly half an hour more per day on direct care of two sons than two daughters,
and slightly more than that in both secondary care and total time with the children. This is offset
by a reduction in leisure time of about a half hour per day. Models with interaction effects
indicate that these results appear to be concentrated among fathers without a college degree, and
among fathers with a child younger than age six.
If we disaggregate main activity childcare time into different activities (only recreational
time is shown here), the extra time that fathers spend with sons compared to daughters is
concentrated in recreation — a result consistent with our findings from the PSID-CDS.
11
The
patterns of father-son interaction across education groups vary between the two data sets,
however. The child gender effects are largest for ATUS fathers without a college degree while
11
Our ability to compare detailed results across data sets is limited because the PSID-CDS categories are based on
the activities of the child, while ATUS parent-child categories are based on the reported activities of the parent. The
respondents are also likely to differ: fathers will be reporting their own childcare time in the ATUS, but most child
diaries are completed by, or with the assistance of, the primary caregiver—usually the mother. The latter may thus
be subject to proxy respondent bias.
14

in the PSID-CDS college-educated fathers spend relatively more time with sons than do less-
educated fathers.
Married mothers spend significantly more secondary childcare time if they have
daughters rather than sons. The interaction model shows that this is true for mothers whose
youngest is under age 6.
Single mothers with two same-sex children spend significantly less time in main-activity
childcare and less total time with their children if they have sons rather than daughters.
However, this effect is only significant for mothers without a college degree and for mothers
with a child younger than age 6. Single mothers with a college degree also spend more time in
market work if they have sons rather than daughters.
Table 5 presents results from samples of families with one or more children, irrespective
of gender, under the age of 13, but no children aged 13 or older in the household. The “son”
dummy here indicates the gender of the eldest child, and we do not control for number of
children or sibling gender composition, as they may be endogenous with respect to gender of the
first-born child. Our previous result that married fathers spend more time on all types of
childcare time if they have a son rather than a daughter is robust to this specification. Married
mothers also spend more recreational time with their children if the eldest child is male. Single
mothers spend less time with their children if the eldest is a son rather than a daughter, and this
effect is concentrated among single mothers without a college degree.
In summary, we find that fathers spend more time with sons, particularly young sons and
particularly fathers with more than one son, and that pre-teen girls in two-parent families are not
compensated by additional time with their mothers, except in secondary childcare time. Fathers
of two sons adjust for this additional childcare time spent on sons by reducing their leisure, and
there is no significant impact of child gender on fathers’ work hours. There are few significant
effects of child gender on mothers’ time use. Married mothers with two children spend less time
in secondary childcare time if they have sons. This result is also found for mothers of one child
if that child is under age six. Married mothers with two children, the youngest of which is
between six and twelve years old, consume less leisure if they have sons rather than daughters.
Single mothers appear to spend more time with daughters than sons of any age. Single mothers
without a college degree with one child, or with a college degree with two children of the same
sex, spend more time in market work if they have sons rather than daughters. Examining the
15

time of non-custodial fathers would help us to better understand the time allocation of single
mothers, but this is not feasible due to data limitations.
D. Teen Time Diaries from the American Time Use Survey
With the ATUS teen sample, we can examine how children’s time spent with parents
changes as the children grow older. Since the diaries are from the child’s perspective, we can
include time spent with both parents. It is not clear a priori how son-daughter effects should
change as children grow older. Parental preferences for interacting with children of the same sex
may intensify as the children develop skills in gender-stereotypical activities; developmental
motivations for these activities may intensify or weaken. The pooled 2003 - 2005 teen sample
consists of 1,470 boys and girls aged 15-17 who live with two parents – 520 who live with a
single mother. We restricted the age to be less than 18 to avoid possible sample selection issues
with teens’ decisions to leave their parents’ household to either attend college or start their own
household. It is possible that teen boys may leave their home later for college if boys start
kindergarten at a later age on average than girls, as evidenced by Datar (2006). As discussed
earlier, there is also evidence that marriages last longer when the couple has a boy rather than a
girl. However, we find no significant difference in the proportion of boys among teens in the
two-parent family sample and that among all teens in the ATUS. Furthermore, we find that the
proportion of boys among teens in our two-parent family sample is not significantly different
from the proportion of boys in a population with a sex ratio of 1.05, which is the typical sex ratio
at birth.
12
We examine teens’ total time with each parent separately and total time with at least one
parent present. We also examine teen activities with parents in activities similar to those in the
PSID-CDS data.
13
We also examine several subcategories of household work and obtaining
goods and services to see whether teens spend more time with parents in traditional gender-
specific activities. These include shopping, general housework, cooking, and home and vehicle
maintenance activities, such as mowing the yard, home repair, building furniture, and chopping
12
The expected proportion male with a sex ratio of 1.05 is .512. This falls within a 95% confidence interval around
the survey weighted proportion males in our two-parent teen sample, which is .534.
13
This excludes personal care, and the education category includes a less-extensive set of activities than is
appropriate for younger children—i.e. help with homework, but not reading to the child.
16

firewood. From the passive leisure category, we report TV watching separately since this is such
a large component of passive leisure.
D. 1. Descriptive Statistics
Table 6 presents the mean teen boy and girl activity times with married mothers and
fathers separately and total time in the presence of at least one parent. We also report time with
mothers in single-mother families. These times are weighted averages of weekday reports and
weekend reports. Mothers – both married and single – spend significantly more total time with
daughters than sons, primarily in household work, shopping, and in social activities. These
effects are generally larger for single mothers than for married mothers. Single mothers also
spend significantly more time eating and drinking with teen girls. Married fathers spend
significantly more time with sons than daughters overall, and they spend this time doing extra
time doing home or vehicle maintenance and in passive leisure, specifically watching TV.
Overall, daughters spend more time doing activities when at least on of their parents are present
than do boys, with significant differences for shopping, housework, cooking, and social
activities.
D. 2. Teen Time-Use Regressions
Table 7 presents the coefficients on a boy dummy from OLS time-use equations for the
teen sample. The dependent variables are minutes per day that the teen spends doing an activity
in the presence of a parent. Each equation also includes controls for teen’s age and age squared,
mother’s age and age squared, father’s age and age squared, race, Hispanic ethnicity, region of
residence, number of siblings in the household, three of four sibling composition groups (no
siblings is omitted; female sibling(s), male sibling(s), and mixed sex siblings are included), and
temporal factors, such as survey year, season, and whether the time diary was collected for a
weekend day. Sample means are shown in Appendix Table A4.
This multivariate analysis confirms that married mothers spend an extra half-hour per day
with daughters relative to sons, and that fathers spend 14 minutes more with sons than daughters.
Overall teen daughters spend more time with any parent than teen sons. Daughters spend more
time doing housework with their married mother than do sons, and more time on shopping,
cooking, and socializing with their married mother than do sons. Teen sons spend significantly
17

more time with their fathers overall and doing home and vehicle maintenance and on passive
leisure activities, with 8 more minutes per day watching TV with their father than daughters.
Single mothers spend even more time with daughters relative to sons – nearly three-quarters of
an hour more in total.
The overall pattern here is quite clear: mothers and teenage daughters spend time together
both in “women’s work” and in social activities (with shopping perhaps having aspects of all of
these), while men and their sons spend more time doing “men’s work” and watching TV
together. The patterns of time use for teens seem to reflect a gendered division of labor in
households, and preferences (of parents or children or both) for same-sex companionship Our
results for teens do not support the hypothesis that teen boys need more parental time. However,
the needs of boys relative to girls may differ for young children and teens.
III. Conclusion
Using two recent U.S. surveys of parent’s time with children, we find that young sons
spend more time with fathers than do daughters, and that daughters aged 0 to 12 are not
compensated by greater overall time with mothers. In the PSID-CDS, we find that married
mothers spend more time directly interacting with young daughters, mainly by doing housework
and shopping together, but this is offset by more maternal secondary time with boys. However,
mothers in the ATUS do not generally spend more secondary time on boys than girls. Fathers
tend to spend more time with boys both directly engaged in their activities and also present but
not involved, and much of this extra time is devoted to sports and other active leisure pursuits.
College-educated fathers also spend relatively more time with sons on direct educational
activities, such as helping with homework, according to child time diaries. In general, the
father/son and mother/daughter engagement in gender-stereotypical activities is consistent with
parental enjoyment in doing ‘boy’ and ‘girl’ things with a same-sex child or with a desire to train
children in gender-appropriate skills. This pattern is particularly pronounced in the ATUS teen
sample, and may reflect a desire to participate in more gender-stereotypical activities in
general.
14
However, both the finding that young boys receive more total parental time than girls
14
We must recognize that teens in general have more say about what types of activities they engage in general and
with whom they do them then do younger children.
18

and the likelihood that the developmental implications of sports and housework will not be
identical suggests that parents are investing more in pre-teen sons than in pre-teen daughters.
Single mothers appear to allocate more time to daughters than sons. In particular for teen
children, they allocate more time to girls, especially in stereotypically female activities, and this
effect is even stronger than in two-parent families. Single mothers also appear to spend more
time working if they have sons rather than daughters.
A gender discrepancy in parental investment is also suggested by the principal effect of a
male child or children on paternal time use—boys reduce fathers’ leisure time in families with
two same-sex children. If fathers of sons spend more time with boys but reduce hours of market
work (and therefore earnings) one could argue that they are responding to boys’ developmental
needs by substituting time inputs for goods. However, the reduction in father’s leisure indicates
that boys are more expensive than girls, particularly in terms of paternal time. Our hypothesis
that, since fathers prefer sons, the birth of a boy will increase the relative bargaining power of his
mother receives little support from our empirical analysis—mothers of sons are not significantly
better off than mothers of daughters in terms of their leisure consumption or time doing
housework. However, our findings that single mothers do not allocate more time to boys, and in
fact allocate more time to girls, suggest that they do not compensate for the lack of extra father-
time for boys. This strengthens the case that it may be same-sex parent time that is important,
and not just overall parental time. However, this study still does not rule out the possibility that
fathers prefer sons and that mothers may gain bargaining power as a result. Future research,
perhaps using data on household expenditures, could further evaluate whether mothers of sons
may be relatively better off than mothers of daughters in terms of material goods. Improved
bargaining power should result in an increase in whatever mothers value. Mothers may value
time their sons spend with their fathers, perhaps more than her own leisure or housework time.
Thus, if both mothers and father prefer that men spend more time with boys, then the presence of
a son may not manifest itself as an observable increase in personal consumption of the mother.
We find that a gendered activities pattern begins early for boys and their fathers, but is
not apparent at early ages for girls and their mothers in two-parent families, although it is visible
in single-mother families. Our results for teens show that time with the same-sex parent
becomes very important as girls get older. When combined with the result that single mothers do
not seem to compensate for the absence of a father by spending more time with their sons, this
19

suggests one of two things—either that mother's time and father's time are simply not
substitutable in some domains, or that some parent-child time is pure parental consumption and
is driven, effectively, by the child gender preferences of the parents rather than by any child
investment motives.
20

References
Angrist, J. and Evans, W. (1998) “Children and Their Parents’ Labor Supply: Evidence from
Exogenous Variation in Family Size.” American Economic Review, 88(3), 450-477.
Blundell, R. and Meghir, C. (1987). “Bivariate Alternatives to the Tobit Model.” Journal of
Econometrics, 34(2), 179-200.
Choi, H.J.; Joesch, J. and Lundberg, S. (2005) “Work and Family: Marriage, Children, Child
Gender and the Work Hours and Earnings of West German Men.” working paper,
University of Washington.
Cragg, J. C. (1971). “Some Statistical Models for Limited Dependent Variables With
Application to the Demand for Durable Goods.” Econometrica, 39(5), 829-844.
Dahl, G. and Moretti, E. (2004) “The Demand for Sons: Evidence from Divorce, Fertility, and
Shotgun Marriage.” NBER Working Paper 10281.
Datar, Ashlesha (2006). “Does Delaying Kindergarten Entrance Give Children a Head Start?’
Economics of Education Review, 25(1), 43-62.
Harris, K. M., and Morgan, S. P. (1991), “Fathers, Sons, and Daughters: Differential Paternal
Involvement in Parenting.” Journal of Marriage and the Family, 53, 531–44.
Hofferth, S. L. and Sandberg, J. F. (2001) “How American Children Spend Their Time.” Journal
of Marriage and the Family, 63(2), 295-308.
Juster, F. T. (1985) “ The Validity and Quality of Time Use Estimates Obtained from Recall
Diaries,” in Juster, F. T. and Stafford, F. P. (eds.) Time, Goods, and Wellbeing, Ann
Arbor, MI: Institute for Social Research, University of Michigan.
Kalenkoski, C.; Ribar, D. and Stratton, L. (2005) “Parental Child Care in Single Parent,
Cohabiting, and Married Couple Families: Time Diary Evidence from the United States
and the United Kingdom.” working paper, Ohio University.
Keller, M. C., Nesse, R. M. and Hofferth, S. (2001) „The Trivers-Willard Hypothesis of Parental
Investment: No Effect in the Contemporary United States.“ Evolution and Human
Behavior 22, 343-360.
Kohler, H. P., Behrman, J. R., and Skytthe, A. (2004) “Partner + Children = Happiness? An
Assessment of the Effect of Fertility and Partnerships on Subjective Well-Being.”
Working paper.
Lundberg, S. (2005a) “The Division of Labor by New Parents: Does Child Gender Matter?”
working paper, University of Washington.
Lundberg, S. (2005b) “Sons, Daughters, and Parental Behavior,” Oxford Review of Economic
Policy, 21(3), 340-356.
Lundberg, S.; McLanahan, S. and Rose, E. (2005) “Child Gender and Father Involvement in
Fragile Families.” working paper, University of Washington.
Lundberg, S., and Rose, E. (2002) “The Effects of Sons and Daughters on Men’s Labor Supply
and Wages.” Review of Economics and Statistics, 84(2), 251-268.
21

22
Lundberg, S. and Rose, E. (2003) “Child Gender and the Transition to Marriage.” Demography,
40(2), 333-350.
Mammen, K. (2005) “Fathers’ Time Investments in Children: Do Sons Get More?” mimeo,
Barnard College.
Manser, M. and Brown, M. (1980) “Marriage and Household Decision Making: A Bargaining
Analysis.” International Economic Review 21(1), 31-44.
McElroy, M. and Horney, M. J. (1981) “Nash Bargained Household Decisions.” International
Economic Review 22(2), 333-50.
Meadows, S. O., Land, K. C., and Lamb, V. L. (2005), ‘Assessing Gilligan Versus Sommers:
Gender-Specific Trends in Child and Youth Well-Being in the United States, 1985-2001’,
Social Indicators Research, 70, 1-52.
Morgan, S. P.; Lye, D. and Condran, G. (1988) “Sons, Daughters and the Risk of Marital
Disruption.” American Journal of Sociology, 94(1), 110-29.
Oswald, A. and Powdthavee, N. (2005) “Daughters and Left-Wing Voting.” working paper
Park, C. B., and Cho, N. H. (1995), ‘Consequences of Son Preference in a Low Fertility Society:
Imbalance of the Sex Ratio at Birth in Korea’, Population and Development Review,
21(1), 59-84.
Strauss, J., and Thomas, D. (1995), ‘Human Resources: Empirical Modeling of Household and
Family Decisions’, in Chenery, H., and Srinivasan, T. N. (eds) Handbook of Development
Economics, Vol. 3A, New York: Elsevier.
Taubman, P. J. (1991), ‘Discrimination within the Family: The Treatment of Daughters and
Sons’, in Hoffman, E. (eds.) Essays on the Economics of Discrimination, 25-42.
Yeung, W. J.; Sandberg, J. F.; David-Kean, P. E. and Hofferth, S. L. (2001) “Children’s Time
with Fathers in Intact Families.” Journal of Marriage and the Family, 63, 136-54

23
Table 1. PSID-Child Development Supplement—Parental Time with Child (Aged 0-12) by Child’s Primary Activity
Panel 1: All Children Mother's time: Minutes per day Father's time: Minutes per day
Means
Difference
Means
Difference
Boys Girls
Boy-Girl
P-value
Boys Girls
Boy-Girl
P-value
Household work and obtaining goods and services 26 36
-10
0.00 16 14
2
0.07
Personal needs and care 72 74 -1 0.58 39 39 0 0.84
Educational and professional training 17 16 1 0.51 6 5 1 0.19
Social and organizational activities 23 24 -1 0.45 17 16 2 0.33
Sports and active leisure 39 41
-2
0.53 34 28
6
0.01
Passive leisure 34 37
-3
0.18 27 24
3
0.09
Total direct time 211 228
-17
0.01 139 125
14
0.00
Total secondary time 183 172
11
0.05 112 105
7
0.10
Total (direct and secondary) time 394 400 -6 0.46 252 230
21
0.00
Number of observations 903 851 903 851
Panel 2: By Mother's Education Mothers without college degree: Minutes per day Mothers with college degree: Minutes per day
Means
Difference
Means
Difference
Boys Girls
Boy-Girl
P-value
Boys Girls
Boy-Girl
P-value
Household work and obtaining goods and services 27 38
-11
0.00 26 31 -5 0.12
Personal needs and care 72 72 0 0.88 72 79 -8 0.15
Educational and professional training 15 15 1 0.57 21 19 2 0.51
Social and organizational activities 22 22 0 0.92 26 31 -5 0.25
Sports and active leisure 39 39 -1 0.81 40 44 -5 0.41
Passive leisure 35 39 -4 0.11 29 28 1 0.75
Total direct time 210 226
-16
0.03 215 234 -19 0.11
Total secondary time 177 170 7 0.27 197 175
22
0.04
Total (direct and secondary) time 388 397 -9 0.35 412 409 3 0.83
Number of observations 685 627 214 218
Panel 3: By Father's Education Fathers without college degree: Minutes per day Fathers with college degree: Minutes per day
Means
Difference
Means
Difference
Boys Girls
Boy-Girl
P-value
Boys Girls
Boy-Girl
P-value
Household work and obtaining goods and services 17 14 3 0.11 15 14 2 0.49
Personal needs and care 38 36 1 0.49 41 43 -2 0.53
Educational and professional training 4 4 0 1.00 10 6
4
0.01
Social and organizational activities 15 14 1 0.49 22 20 3 0.48
Sports and active leisure 31 28 3 0.22 40 27
13
0.00
Passive leisure 30 26 3 0.15 21 19 3 0.34
Total direct time 135 123
12
0.04 150 128
22
0.01
Total secondary time 110 105 5 0.33 118 105
13
0.10
Total (direct and secondary) time 245 228
17
0.04 267 233
34
0.01
Number of observations 648 595 241 247
Note: Average minutes per day based on weighted average of weekday and weekend diaries.

24
Table 2. American Time Use Survey: Parental Minutes per Day, by Activity
Married Mother’s Time Married Father’s Time Single Mother’s Time
Means Means Means
Boy(s) Girl(s)
Difference
Boy – Girl
P-value Boy(s) Girl(s)
Difference
Boy – Girl
P-value Boy(s) Girl(s)
Difference
Boy – Girl
P-value
Panel 1. Mothers and Fathers with 1 Child Aged <6
Childcare 162.4 157.2 5.2 0.64 81.2 71.5 9.7 0.26
122.6 119.2 3.4 0.82
Subcategory: Recreational Childcare 57.9 53.7 4.2 0.53 39.8 24.9
14.9
0.01
37.5 31.6 5.9 0.50
Secondary Childcare 401.9 417.1 -15.2 0.44 267.9 262.1 5.8 0.75
340.2 359.2 -19.0 0.57
Time with Child 430.1 461.3 -31.2 0.14 278.1 281.4 -3.3 0.85
361.9 383.5 -21.6 0.53
Market Work 173.3 156.5 16.8 0.37 321.0 351.3 -30.3 0.12
213.9 176.4 37.6 0.26
Leisure 786.7 796.8 -10.1 0.48 779.1 753.1 26.0 0.14
860.2 870.5 -10.3 0.79
Household Work 169.9 168.3 1.6 0.88 98.5 92.5 6.0 0.51
108.0 139.6
-31.6
0.03
Number of Observations 503 450 471 445
182 197
Panel 2. Mothers and Fathers with 1 Child Aged 6-12
Childcare 67.6 65.9 -0.2 0.82
40.5 35.4 5.1 0.42
59.7 71.9 -12.1 0.16
Subcategory: Recreational Childcare 11.8 5.7
4.4
0.02
13.0 10.0 3.0 0.44
9.6 6.5 3.1 0.32
Secondary Childcare 352.1 334.2 39.5 0.35
281.2 261.4 19.9 0.35
318.1 324.5 -6.3 0.83
Time with Child 276.8 278.7 2.6 0.92
223.7 193.0
30.7
0.09
220.9 289.7
-68.8
0.00
Market Work 217.1 235.4 -18.3 0.45
349.9 366.5 -16.6 0.47
257.1 220.5 36.5 0.21
Leisure 819.0 805.1 13.9 0.46
772.4 787.4 15 0.49
830.0 827.7 2.4 0.93
Household Work 179.1 180.5 -1.4 0.92
96.4 103.3 -6.9 0.62
137.7 158.0 -20.3 0.28
Number of Observations 332 335
271 275
264 281
Panel 3. Mothers and Fathers with 2 Same-Sex Children Aged <13
Childcare 163.1 146.3 16.8 0.11
95.7 59.5
36.2
0.00
121.9 168.9
-47.0
0.03
Subcategory: Recreational Childcare 37.7 29.8 7.9 0.11
32.9 15.4
17.5
0.00
16.9 16.6 0.3 0.97
Secondary Childcare 405.4 445.9
-40.5
0.04
292.0 248.6
43.4
0.05
347.7 354.6 -6.9 0.84
Time with Child 458.6 459.2 -0.6 0.98
302.5 251.1
51.4
0.02
353.1 389.3 -36.2 0.38
Market Work 149.1 147.8 1.3 0.94
331.7 339.0 -7.3 0.71
161.9 171.3 -9.4 0.79
Leisure 779.4 787.2 -7.8 0.61
738.2 782.7
-44.5
0.03
847.6 803.9 43.7 0.27
Household Work 196.8 200.5 -3.7 0.76
103.1 101.4 1.7 0.88
170.9 151.4 19.5 0.36
Number of Observations 501 448
437 413
147 146
Panel 4. Mothers & Fathers with Multiple Children Aged<13, by Gender of Eldest Child
Childcare 152.1 145.7 6.6 0.20 80.3 64.2
16.0
0.00
113.6 119.8 -6.2 0.47
Subcategory: Recreational Childcare 36.6 32.5
3.1
0.08 30.3 21.9
8.4
0.00
19.8 18.9 0.8 0.79
Secondary Childcare 415.5 418.2 1.0 0.76 275.1 261.8
13.3
0.13
353.1 364.3 -11.2 0.79
Time with Child 438.9 441.6 -5.2 0.78 280.4 263.8
16.6
0.04
336.4 364.9
-28.6
0.10
Market Work 162.8 161.6 5.4 0.89 339.5 350.1 -10.6 0.25
196.5 187.7 8.7 0.58
Leisure 779.4 782.5 -5.5 0.66 759.5 762.5
-2.9
0.74
839.9 841.8 -1.9 0.91
Household Work 197.7 198.1 -0.7 0.95 97.5 101.6 -4.2 0.40
140.8 144.2 -3.4 0.69
Number of Observations 2,380 2,183 2,058 1,974
909 867
Note: Average minutes per day based on weighted average of weekday and weekend diaries.
Survey weights used.

25
Table 3. American Time Use Survey: Son Effects on Parental Time Use by Activity in One-Child Families (Child aged 0-12)
Childcare
Recreational
Childcare
Secondary
Childcare
Time with Child Leisure Market Work Household Work
Panel 1. Married Mother's Time: Minutes per day (N=1,620)
Son 7.14 6.66 -8.63 -18.56 -15.37 11.38 1.04
(6.55) (4.00) (13.37) (13.85) (10.74) (14.05) (8.43)
Son 12.27 7.01
-32.69*
-36.59 -20.47 24.04 -2.52
(9.95) (6.43) (19.60) (19.88) (13.15) (17.85) (10.90)
Son*Child Aged 6-12 -12.84 -0.87
60.21**
45.11 12.02 -31.69 8.91
(13.06) (7.10) (26.67) (28.02) (22.38) (29.01) (17.92)
Son 9.74 7.96 -4.06 -21.59 -15.76 13.58 4.98
(8.45) (4.90) (18.63) (19.27) (15.17) (18.80) (11.63)
Son*College Degree -6.45 -3.24 -11.34 7.52 0.97 -5.47 -9.79
(13.34) (8.18) (25.94) (27.63) (20.73) (27.91) (16.28)
Panel 2. Married Father's Time: Minutes per day (N=1,462)
Son 2.25
9.94***
4.13 2.54 5.70 -13.44 -1.80
(5.89) (3.67) (13.68) (12.74) (13.04) (14.49) (7.54)
Son 6.80
14.09***
4.58 -4.12 23.21 -25.34 5.72
(8.34) (5.28) (17.50) (16.82) (16.32) (18.73) (8.92)
Son*Child Aged 6-12 -3.54
-11.50*
-1.24 18.46 -48.51 32.97 -20.83
(10.76) (6.79) (27.44) (24.76) (27.30) (29.67) (16.08)
Son 2.68
8.88**
2.89 -3.31 2.25 2.00 -5.41
(7.46) (4.55) (17.37) (15.74) (17.99) (19.07) (10.73)
Son*College Degree 7.68 2.88 3.36 15.77 9.30 -41.68 9.75
(11.64) (7.25) (28.40) (26.15) (25.68) (29.15) (13.97)
Panel 3. Single Mother's Time: Minutes per day(N=924)
Son -1.60 5.20 -12.14
-44.15**
-14.37
44.52** -25.62**
(7.08) (3.90) (20.61) (20.12) (17.54) (18.81) (10.60)
Son 7.20 6.13 -14.64 -12.83 -15.71 40.77
-34.80***
(12.05) (7.47) (30.63) (33.47) (27.71) (28.23) (14.22)
Son*Child Aged 6-12 -16.23 -1.72 4.62 -57.80 2.49 6.92 16.94
(14.55) (8.07) (41.18) (42.01) (36.31) (38.88) (21.36)
Son -0.73 6.80 -14.57
-51.22**
-25.76
57.66*** -27.92**
(7.94) (4.46) (23.71) (23.01) (20.20) (20.80) (11.64)
Son*College Degree -5.03
-9.30*
14.15 41.13
66.22*
-76.37 13.36
(16.98) (8.79) (41.05) (43.29) (37.48) (48.20) (26.51)
Notes: Standard errors in parentheses. Survey weights used. *** significant at 1%; ** significant at 5%; * significant at 10%.
Each panel contains a separate model for each activity. Recreational Childcare is a subcategory of Childcare.
Control variables include a quadratic in age of respondent and spouse, race and ethnicity (black, other, Hispanic), region, year, and season, age category of youngest child,
education category of respondent and spouse, presence of a grandparent in household, and weekend dummy.

26
Table 4. American Time Use Survey: Son Effects on Parental Time Use by Activity in Two-Child Families (Children aged 0-12)
Childcare
Recreational
Childcare
Secondary
Childcare
Time with Child Leisure Market Work Household Work
Panel 1. Mother's Time: Minutes per day (N=949)
Sons 14.00 6.57
-36.66**
0.95 -6.47 0.65 1.55
(9.28) (4.44) (17.60) (19.48) (14.23) (15.95) (11.66)
Sons 15.58 8.82
-39.23*
-0.95 8.31 -4.78 -3.35
(12.43) (6.01) (22.18) (25.00) (16.73) (19.74) (14.27)
Son*Child Aged 6-12 -5.63 -7.98 9.11 6.76
-52.37*
19.26 17.32
(18.61) (8.13) (36.02) (37.02) (31.28) (33.87) (22.94)
Sons 9.17 -1.80 -30.45 4.19 10.37 -11.87 8.99
(12.07) (5.07) (24.31) (28.37) (20.36) (21.24) (16.49)
Son*College Degree 11.93
20.67**
-15.33 -8.00 -41.58 30.90 -18.38
(17.78) (9.56) (35.00) (37.80) (27.06) (32.18) (21.93)
Panel 2. Father's Time: Minutes per day (N=850)
Sons
27.94*** 14.57*** 44.66** 46.91*** -32.36**
-10.59 2.61
(9.16) (4.11) (19.87) (17.86) (15.73) (18.76) (10.55)
Sons
33.67*** 14.74*** 42.23* 41.49* -38.52**
-12.35 7.54
(12.92) (5.30) (24.42) (23.18) (19.38) (22.84) (13.31)
Son*Child Aged 6-12 -19.28 -0.58 8.19 18.25 20.75 5.91 -16.58
(18.02) (8.17) (41.68) (34.33) (34.49) (39.17) (20.99)
Sons
34.49*** 12.52** 71.24*** 67.85*** -30.30**
-24.00 9.15
(13.56) (5.34) (27.13) (24.93) (22.83) (25.23) (15.10)
Son*College Degree -17.29 5.41
-70.21*
-55.31 -5.43 35.44 -17.27
(19.67) (8.54) (39.31) (35.06) (30.79) (36.86) (21.05)
Panel 3. Single Mother's Time: Minutes per day (N=293)
Son
-34.20**
3.16 -21.02
-46.71*
36.57 -8.62 12.17
(17.01) (5.42) (26.38) (28.81) (27.79) (28.77) (18.34)
Son
-51.27**
1.89 -4.64
-56.34*
55.56 -25.02 24.35
(22.59) (6.60) (31.00) (34.31) (36.54) (34.56) (21.64)
Son*Child Aged 6-12
62.69*
4.66 -60.12 35.36 -69.74 60.21 -44.73
(31.40) (10.25) (50.93) (52.18) (64.35) (58.98) (31.41)
Son
-38.33**
3.65 -17.28
-55.78*
49.84 -26.85 17.17
(19.11) (5.79) (28.59) (31.56) (31.34) (31.68) (20.00)
Son*College Degree
33.51*
-3.92 -30.31 73.68 -107.76
147.95**
-40.60
(33.51) (13.29) (75.20) (60.91) (58.35) (68.46) (45.00)
Notes: Standard errors in parentheses. Survey weights used. Significance levels: *** =1%; ** = 5%; * = 10%.
Each panel contains a separate model for each activity. Recreational Childcare is a subcategory of Childcare.
Control variables include a quadratic in age of respondent and spouse, race and ethnicity (black, other, Hispanic), region, year, and season, age category of youngest child,
education category of respondent and spouse, presence of a grandparent in household, age gap between children and weekend dummy.

27
Table 5. American Time Use Survey: Effect of Eldest Son on Parental Time Use by Activity (Children aged 0-12)
Panel 1. Married Mother's Time: Minutes per day (N=4,563)
Childcare
Recreational
Childcare
Secondary
Childcare
Time with Child Leisure Market Work Household Work
Son 4.51
3.81*
-5.31 -7.27 -4.96 4.60 0.55
(4.54)
(2.24)
(8.36) (8.72) (6.49) (8.20) (5.30)
Son -0.57 3.80 -0.42 -9.05 2.39 5.31 2.08
(5.80) (2.73) (11.52) (11.81) (9.13) (10.85) (7.32)
Son*College Degree 13.36 0.04 12.88 4.70 -19.35 -1.88 -4.03
(9.24) (4.67) (16.29) (17.01) (12.42) (16.47) (10.35)
Panel 2. Married Father's Time: Minutes per day (N=4,032)
Childcare
Recreational
Childcare
Secondary
Childcare
Time with Child Leisure Market Work Household Work
Son
14.74*** 8.05*** 15.17* 15.77**
-1.72 -10.72 -3.36
(4.01) (2.35)
(8.60) (7.99)
(8.15) (8.98) (4.79)
Son
15.14*** 7.10** 23.84** 21.33**
-1.64 -11.55 2.12
(5.43) (3.16) (11.22) (10.59) (11.35) (11.98) (6.63)
Son*College Degree -1.10 2.62 -24.03 -15.42 -0.24 2.31 -3.42
(7.65) (4.65) (17.37) (15.87) (15.28) (17.69) (9.08)
Panel 3. Single Mother's Time: Minutes per day (N=1,776)
Childcare
Recreational
Childcare
Secondary
Childcare
Time with Child Leisure Market Work Household Work
Son -4.69 1.69 -12.82
-30.90**
-7.91 13.41 -4.19
(7.84) (2.85) (15.86) (15.96) (13.94) (14.05) (7.82)
Son -9.23 2.10 -15.84
-37.00**
-8.74 17.13 -2.48
(8.35) (3.12) (17.05) (17.79) (15.62) (15.14) (8.48)
Son*College Degree 34.28 -3.12 22.76 46.00 6.32 -28.06 -12.91
(22.92) (7.74) (40.49) (37.05) (29.27) (38.20) (19.31)
Notes: Standard errors in parentheses. Survey weights used. Significance levels: *** =1%; ** = 5%; * = 10%.
Each panel contains a separate model for each activity. Recreational Childcare is a subcategory of Childcare.
Control variables in the married parents regressions include a quadratic in age of respondent and spouse, race and ethnicity (black, other, Hispanic), region, year, and season,
education category of respondent and spouse, presence of a grandparent in household, and weekend dummy. Control variables in the single mother regression
exclude information on spouse.

28
Table 6. American Time Use Survey: Parental Minutes per Day with Teens (Aged 15-17) by Teen’s Primary Activity
Married Mother's time
Married Father's time
Any Married Parent time
Single Mother's time
Means Difference
Means Difference
Means Difference
Means Difference
Boys Girls
Boy-Girl
P
Boys Girls
Boy-Girl
P
Boys Girls
Boy-Girl
P
Boys Girls
Boy-Girl
P
Total time 98.7 127.8
-29.1
0.00
93.3 79.8
13.5
0.09
132.6 149.4 -16.8 0.11
68.9 111.0
-42.1
0.00
Household work and obtaining
goods and services
9.4 20.8
-11.4
0.00 8.8 8.6 0.2 0.91 15.0 23.9
-8.9
0.00
10.1 18.7
-8.6
0.06
Subcategories:
Shopping 5.8 10.6
-4.8
0.03 2.0 3.4 -1.4 0.17 6.3 11.2
-4.9
0.03
4.4 9.2
-4.8
0.05
Housework 0.6 3.7
-3.1
0.00 1.2 1.4 -0.2 0.78 1.5 3.9
-2.4
0.02
0.9 3.8
-2.9
0.02
Cooking 1.5 3.4
-1.9
0.01 0.7 1.3 -0.6 0.23 1.9 4.2
-2.3
0.01
1.0 3.0
-2
0.05
Home & vehicle
maintenance
0.8 0.8 0.0 0.91 3.7 0.6
3.1
0.00 3.8 1.0
2.8
0.00 0.2 1.5
-1.3
0.08
Education (homework) 2.9 3.7 -0.8 0.64 1.0 1.8 -0.8 0.24 3.4 4.3 -0.9 0.61
0.5 1.1 -0.6 0.28
Social and organizational
activities
14.2 18.7
-4.5
0.09 13.7 14.0 -0.3 0.91 15.9 20.9
-5.0
0.09
7.0 21.0
-14
0.01
Sports and active leisure 2.9 2.8 0.1 0.94 4.6 2.7 1.9 0.24 6.0 3.8 2.2 0.21
5.4 5.4 0 0.99
Passive leisure 29.8 32.5 -2.7 0.56 29.7 20.2
9.5
0.01 42.5 39.4 3.1 0.53
24.4 28.6 -4.2 0.52
Subcategory: TV 26.6 27.5 -0.9 0.82 26.0 17.7
8.3
0.01 37.4 33.1 4.3 0.35
23.2 23.7 -0.5 0.93
Eating & drinking 21.8 23.2 -1.4 0.42 18.4 18.6 -0.2 0.91 23.6 24.9 -1.3 0.49
9.2 12.6
-3.4
0.06
Number of observations 772 698 772 698 772 698
254 266
Notes: Average minutes per day based on weighted average of weekday and weekend diaries.
Survey weights used.

Table 7. American Time Use Survey: Male Effect for Teens’ (Aged 15-17) Minutes per Day with Parents by Teen's Primary
Activity
Married
Mother
Married
Father
Any Married
Parent
Single
Mother
Total time
-30.43***
13.85*
-17.04*
-40.70***
(9.00)
(8.21) (10.32)
(12.51)
Household work and obtaining goods
-11.77***
0.34
-9.06***
-8.67**
and services
(2.82)
(1.98) (3.15)
(4.35)
Subcategories:
Shopping
-5.04**
-1.49
-5.16**
-5.32**
(2.26)
(1.03) (2.28)
(2.28)
Housework
-3.19***
-0.15
-2.41**
-2.70***
(0.91)
(0.73) (1.06)
(1.07)
Cooking
-1.96***
-0.64
-2.40***
-1.68*
(0.68)
(0.49) (0.80)
(1.00)
Home & vehicle maintenance
0.06
3.41***
3.07***
-1.56*
(0.51)
(1.01) (1.02)
(0.91)
Education (homework)
-0.69
-0.78 -0.70
-0.52
(1.52)
(0.69) (1.58)
(0.53)
Social and organizational activities
-4.76*
-0.27
-5.11*
-13.92***
(2.85)
(2.90) (3.06)
(5.30)
Sports and active leisure
0.01
2.01 2.24
0.01
(1.15)
(1.62) (1.77)
(2.48)
Passive leisure
-3.18
9.72***
3.13
-2.67
(4.26)
(3.78) (4.90)
(6.40)
Subcategory: TV
-1.75
8.32**
3.73
0.39
(3.92)
(3.49) (4.46)
(6.16)
Eating and Drinking
-1.57
-0.04 -1.19
-3.64**
(1.81)
(1.63) (1.82)
(1.79)
Number of observations
1,470 1,470 1,470 520
Notes: Only the Male dummy coefficient is shown for each model.
Control variables for teens with married parents include a quadratic in age of child and both parents, race and ethnicity (black, other,
Hispanic), number of siblings, 3 of 4 sibship composition groups, region, year, season, and weekend dummy. Control
variables for teens with a single mother exclude the age of father and include whether another female relative, such as a
grandmother or aunt, lives in the household.
*** significant at 1%; ** significant at 5%; * significant at 10%.
Standard errors in parentheses. Survey weights used.
29

Appendix Table 1. PSID - Child Development Supplement - Summary Statistics
Mean
(std. dev.)
Mother
Age 34.66
(6.17)
White 76.14
Black 6.62
Other race 17.25
With college degree 27.94
Without college degree 72.06
Father
Age 36.85
(6.80)
White 75.81
Black 6.41
Other race 17.78
With college degree 34.38
Without college degree 65.62
Number of observations (children) 1,754
-excluding observations with missing mother’s education 1,744
-excluding observations with missing father’s education 1,731
30

Appendix Table A2. American Time Use Survey: Means for Parent Time Diary Samples
(One or Two Children Aged 0-12)
Married Mothers
Married Fathers
Single Mothers
1 Child 2 Children 1 Child 2 Children 1 Child 2 Children
Age
33.98
(11.07)
33.82
(8.42)
35.85
(11.12)
35.96
(9.28)
31.89
(13.49)
29.94
(9.82)
Spouse’s age
36.80
(11.72)
36.28
(8.85)
34.03
(10.50)
34.05
(9.54)
Black
0.08 0.06 0.09 0.07 .33 .43
Other race
0.07 0.07 0.09 0.06 .05 .05
Hispanic
0.16 0.19 0.16 0.19 .16 .19
Son(s) dummy
0.55 0.53 0.53 0.49 .47 .51
(youngest) child aged 3-5
0.18 0.28 0.20 0.31 .23 .27
(youngest) child aged 6-12
0.40 0.29 0.36 0.29 .53 .27
Age gap between children
3.15
(2.65)
3.02
(2.54)
Interactions with son(s)
Aged 6-12 * son(s)
0.21 0.15 0.20 0.14 .27 .13
College degree * son(s)
0.19 0.23 0.19 0.19 .09 .06
Respondent Education
High School grad
0.28 0.25 0.30 0.27 .34 .43
Some College
0.25 0.24 0.24 0.24 .35 .29
College Degree
0.25 0.28 0.23 0.25 .12 .09
Advanced Degree
0.14 0.13 0.13 0.12 .05 .03
Spouse education
High School grad
0.26 0.23 0.24 0.24
Some College
0.26 0.26 0.23 0.26
College Degree
0.22 0.25 0.28 0.25
Advanced Degree
0.14 0.13 0.11 0.13
Grandparent in HH
0.02 0.03 0.03 0.02 .21 .25
Weekend
0.28 0.31 0.30 0.32 .31 .36
Midwest
0.20 0.21 0.22 0.25 .23 .22
South
0.37 0.34 0.34 0.32 .39 .44
West
0.25 0.26 0.25 0.24 .18 .16
Spring
0.25 0.26 0.25 0.23 .27 .29
Summer
0.23 0.25 0.23 0.25 .28 .23
Fall
0.28 0.22 0.27 0.26 .20 .22
Year
2004.02 2004.00 2004.03 2003.98 2004.01 2004.08
Number of observations
1,620 949 1,462 850 924 293
Note: Standard deviations in parentheses. Survey weights used.
31

32
Appendix Table A3. American Time Use Survey: Means for Parent Time Diary Samples (Multiple
Children Aged 0-12)
Married Mothers Married Fathers Single Mothers
Age
33.73
(9.58)
35.77
(9.90)
30.21
(12.20)
Spouse’s age
36.36
(10.04)
33.88
(9.33)
Black .07 .08 .36
Other race .06 .07 .06
Hispanic .19 .20 .19
Eldest son dummy .52 .50 .51
College degree * eldest son .20 .18 .07
(youngest) child aged 3-5 .26 .27 .27
(youngest) child aged 6-12 .31 .29 .38
Respondent Education
High School grad .26 .28 .38
Some College .26 .24 .30
College Degree .25 .23 .09
Advanced Degree .13 .13 .04
Spouse education
High School grad .25 .23
Some College .26 .24
College Degree .22 .27
Advanced Degree .13 .11
Grandparent in HH .02 .02 .25
Weekend .30 .30 .32
Midwest .23 .24 .23
South .34 .33 .41
West .26 .25 .18
Spring .25 .24 .29
Summer .24 .24 .26
Fall .26 .26 .21
Year 2004.01 2004.00 2004.05
Number of observations 4,563 4,032 1,776
Note: Standard deviations in parentheses. Survey weights used.

33
Appendix Table A4. American Time Use Survey: Means for Teen Time Diary Samples (Children
Aged 15-17)
Living with
Married
Parents
Living with
Single
Mother
Age
16.05
(1.02)
16.13
(1.02)
Male 0.55 0.50
Mother's age
43.91
(7.16)
42.43
(7.51)
Father's age
46.24
(8.10)
Black 0.10 .33
Other race 0.06 .07
Hispanic 0.20 .17
Number of siblings in HH
1.45
(1.47)
1.37
(1.80)
Male siblings .31 .26
Female siblings .29 .24
Mixed-gender siblings .23 .24
Female relative .07
Weekend 0.31 .30
Midwest 0.26 .25
South 0.30 .33
West 0.27 .23
Spring 0.25 .24
Summer 0.26 .23
Fall 0.26 .25
Year 2004.02 2003.99
Number of observations 1,470 520
Note: Standard deviations in parentheses. Survey weights used.

34
Data Appendix: Variables from the American Time Use Study
Notation: TX refers to tier X code in ATUS activity lexicon.
Parent Time Diary Samples
Variable Definition
Childcare Main activity care – sum of educational, primary, and recreational child care
Recreational
Childcare
T1=3 & T2=1 & (T3=3 or 4 or 5 or 10): Playing with child(ren), e.g., playing
basketball, teaching to ride a bike, “horsing around”, building models, other crafts,
attending child’s events (e.g., game, recital, play), taking trick-or-treating, etc.
Educational
Childcare
T1=3 & ((T2=1 & T3=2) or T2=2): Reading to or doing homework with
child(ren), meetings with teachers, PTA meetings, home schooling, etc.
Primary care T1=3 & [T2=1 & (T3=1,6, 7, 8, 9, 11, 12, or 99)] or (T2=3) or (T1=17 & T2=3 &
T3=1): Physical care or everyday care, watching after, talking with, organizing
activities for, waiting for, picking up/dropping off, bandaging knee, taking child to
doctor, etc.
Secondary
Childcare
Any non-childcare activity while a child(ren) under age 13 is(are) in your care
(child need not be in same room)
Time with Child Time in any activity when child(ren) is(are) in the room
Leisure (T1=1, 12, 13, 14, or 15) or [T1=2 & T2=9 & (T3=3 or T3=4)] or [T1=6 & T2=1
& T3=2] or [T1=8 & T2=5] or [T1=16 & T2=1 & (T3=1 or T3=2)]: e.g., personal
care & related services (haircut, pedicure), socializing, reading, going to movies,
sports, exercise, hobbies, religious activities, volunteer work, personal email or
letters, classes for personal interest, phone calls with friends and family, etc.
Home production T1=2, 7, 8, 9, 10, or 16 excluding [T1=2 & T2=9 & (T3=3 or T3=4)]
& [T1=8 & T2=5] & [T1=16 & T2=1 & (T3=1 or T3=2)]: Housework; food prep;
care of pets; cooking; home, appliance, & vehicle maintenance & repair;
household and financial management tasks; shopping; obtaining services for the
household; using government services; civic obligations; household related phone
calls, etc.
Market work T1 = 5: Work and job-related activities, including job search, income generating
activities, entertaining clients, job-specific training, etc.
Notes: Times with children refer to own household children.
Leisure, home production, and market work may overlap with secondary care.

35
Teen Time Diary Sample
Father time Total time (any activity) with father present
Mother time Total time (any activity) with mother present
Any parent time Total time (any activity) with one or both parents present
Household work and obtaining
goods & services
T1=2 or T2=7 or (T1=8 excluding T2=5) or T1=9: Household
activities, shopping, and obtaining professional and household
services
Shopping T1 = 7
Housework T1 = 2 & T2=1
Cooking T1 = 2 & T2=2
Home and vehicle maintenance T1 = 2 & (T2=3 or T2=4): interior/exterior home maintenance,
vehicle repair, yard work,
Education T1=6 & T2=3: Homework
Social and organizational activities [T1=12 & (T2=1,2, or 4 or (T2=5 & T3=1, 2 or 4)] or [T1=13 &
(T2=2 or (T2=4 & T3=2))] or T1=14: general socializing,
attending or hosting social events, attending arts/sport/recreation
events, religious and spiritual activities
Sports and active leisure [T1=3 & T2=1 & (T3=3,4, or 5)] or [T1=12 & T2=3 & (T2=7, 9,
10, or 11)] or [T1=13 & T2=1] or [T1=13 & T2=3 & T3=1] or
[T1=13 & T2=4 & T3=1]: playing sports, arts & crafts, hobbies
Passive leisure [T1=12 &T2=3 excluding T3 = 7, 9-11] or (T1=12 & T2=5 &
T3=3): TV, listening to music, reading, conversations, relaxing,
doing nothing
TV time T1= 12 & T2=3 & T3=3: watching TV, not religious
Eating & drinking T1= 11
Note: Times are in each activity when mother, father, or either parent is present.