How American Children Spend Their Time

Author(s): Sandra L. Hofferth and John F. Sandberg

Source: Journal of Marriage and Family, Vol. 63, No. 2 (May, 2001), pp. 295-308
Published by: National Council on Family Relations
Stable URL: http://www.jstor.org/stable/3654593 .
Accessed: 12/09/2013 15:22
Your use of the JSTOR archive indicates your acceptance of the Terms & Conditions of Use, available at .
http://www.jstor.org/page/info/about/policies/terms.jsp

.
JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide range of
content in a trusted digital archive. We use information technology and tools to increase productivity and facilitate new forms
of scholarship. For more information about JSTOR, please contact support@jstor.org.

National Council on Family Relations is collaborating with JSTOR to digitize, preserve and extend access to
Journal of Marriage and Family.

http://www.jstor.org

This content downloaded from 201.213.38.62 on Thu, 12 Sep 2013 15:22:24 PM

All use subject to JSTOR Terms and Conditions

SANDRA
L. HOFFERTH
ANDJOHNE SANDBERG
Universityof Michigan

How AmericanChildrenSpendTheirTime

The purpose of this article is to examine how

Americanchildrenunderage 13 spendtheirtime,
sources of variationin time use, and associations
with achievementand behavior.Data comefrom
the 1997 Child DevelopmentSupplementto the
Panel Studyof IncomeDynamics.Theresultssuggest thatparents'characteristicsanddecisionsregarding marriage,family size, and employment
affect the time children spend in educational,
structured,andfamily activities,whichmayaffect
theirschool achievement.Learningactivitiessuch
as readingfor pleasureare associatedwithhigher
achievement,as is structuredtime spentplaying
sports and in social activities.Family time spent
at meals and time spent sleeping are linked to
fewer behavior problems, as measured by the
child's score on the Behavior Problems Index.
The results supportcommonlanguage and myth
about the optimaluse of timefor child development.
We often assume that how childrenspend their
time affects their cognitive and social development. Much of our languagerefers to children's
behaviorin terms of time-whether they spend
too littletime studying,reading,or helpingaround
the house or too much time watchingtelevision
and hangingout with friends.Despite the use of
languagein which time is the acceptedcultural
mediumfor communicatingabout children'sacInstitutefor Social Research, University of Michigan, P.O.
Box 1248, Ann Arbor, MI 48106-1248. (Hofferth@
umich.edu).
Key Words: achievement, behavior problems, children,
time use.

tivities andbehavior,however,time is rarelystudied. This articletakes advantageof nationaldata

collected in 1997 to describe how younger and
olderchildrenunderage 13 spendtheirtime,what
factors are associatedwith these time expenditures, and whethervariationsin time use are associated with children'sachievementand behavior.
BACKGROUND

The developmentof a childfrom infancyto adulthood requiressubstantialparentaland community

investmentsof time, money, and psychosocialor
emotionalcapital(Haveman& Wolfe, 1994). Besides providingopportunitiesfor engagementwith
others, activities provide contexts for learning
(Larson& Verma, 1999). Each context engages
participantsin a set of behaviorsand rules and
resultsin learningskills anda body of knowledge.
Much researchfocuses on the acquisitionof literacy, numeracy,and problem-solvingskills. In
additionto cognitive skills, activitiessuch as play
and conversationprovideopportunitiesfor developing social and emotionalskills. The quantityof
time serves as an estimateof exposureto different
social experiences, with more time leading to
greaterabsorptionof the skills and knowledgeof
that context. Althoughwe may know the demoof children's
graphicandeconomiccharacteristics
families and the communitieswherethey live and
attend school, we rarely know how individual
children spend their time (Larson & Verma).
Thus, children'sown investmentsand exposure
and their links to achievement and emotional
growthare unknown.

Journalof Marriageand Family 63 (May 2001): 295-308

This content downloaded from 201.213.38.62 on Thu, 12 Sep 2013 15:22:24 PM

All use subject to JSTOR Terms and Conditions

295

Journal of Marriage and Family

296
In this article, we addressfour key areas of
children'sactivities:(a) school and day-caretime,
(b) discretionarytime in free play versus organized activities, (c) time in outside-of-school
learningactivities, and (d) time spent in family
activities.
TimeSpentin School Settings
One of the criticalconcernsof parentsis optimizing children'slearningin formalsettings.Educationalresearchhas focusedon such factorsas expendituresper capita, class size, and teacher
qualifications(Hanushek,1989), but less attention
has been paidto how the amountof time children
spendin school may be relatedto achievementon
standardized
tests. Americanchildrenspendmuch
less time in school and studyingthan do Chinese
and Japanese children (Fuligni & Stevenson,
1995; Juster& Stafford,1991). It is hypothesized
thatthis differencemay accountfor the substantial
differences in achievementbetween children in
both countries.Yet time in school-basedsettings
has increasedover the past severaldecadesin the
United States (Hofferth,1996). We examine the
amountof time childrenspent in school and day
care in 1997, how family and individualfactors
are linked to in-school time, and whetherschool
and day-caretime are linked to achievementon
standardizedtests.
Free Play versus OrganizedActivities
Childrendo not learnonly in formalsettings.For
young children,play is theirwork.Besides motor
skills, in play childrendevelopinitiative,self-regulation,and social skills (Larson& Verma,1999).
Recent researchhas pointedout both the importance of physical activity to child development
(Bredekamp& Copple, 1997) and the sedentary
nature of many children's lives (Pellegrini &
Smith, 1998). In this research,playing is a broad
categorythatincludesplayingcards,boardgames,
and puzzles; playing social games such as jump
rope; playing pretendgames; playing with toys;
and unspecifiedindoor and outdoorplay. It also
includesplaying computergames and surfingthe
Internet.A relatedcategory,outdooractivities,includes gardening,boatingand camping,picnicking, pleasure drives, walking, and hiking. The
types of activities in which childrenengage are
likely to be shapedby theircurrentfamilycontext,
including maternalemployment,education,and
family structure.We ask herewhetherthe amount

of time spent in play or otherfree-timeactivities

mattersto children'sachievementand behavior.
Instead of these unstructuredactivities, children today may be spendinga large fractionof
time in highly structuredactivities,such as sports
programs, church-sponsoredactivities, and a
broad category called "visiting." Majorlegislative efforts made in the 1970s to improvegirls'
opportunitiesshould show up today in only a
small gender gap in sports at the elementary
school level. Organizationssuch as LittleLeague
in sports,but
andGymboreepromoteparticipation
their
as
to
contribution
have
little
guidance
parents
to childdevelopment.Churchactivitiesincludeattendingchurchas well as participatingin churchsponsored meetings. Visiting is included as a
structuredactivity because it includes participation in (non-church-sponsored)
youth clubs and
organizations.Such activitiesare expectedto promote children'sachievementand behavior(Task
Force on Youth Development and Community
Programs,1992).
Out-of-SchoolLearningActivities
Readingis an activitythathas been shown to be
strongly linked to children'sscores on standard
verbalachievementtests (Snow,Burns& Griffin,
1998; U.S. Departmentof Education,1999), as
has studying (Keith, Reimers, Fehrmann,Pottebaum,& Aubey, 1986). Watchinga lot of television, in contrast,has been linked to lower cognitive test scores (Timmer,Eccles, & O'Brien,
1985). Several studies show that more television
viewing is associatedwith less time spent in activities such as readingand studying(Koolstra&
Van Der Voort, 1996), which may partiallyexplain its negativeeffect. In the early 1980s (Timmer et al.), childrenspentonly 8 minutesper day
readingfor pleasure,whereas2.3 hourswerespent
watching television. It is importantto explore
what factors affect time spent in these activities
and what differencethese activitiesmaketo children'sachievement.
FamilyActivities
Family activities, including household work,
household conversations,and mealtime provide
importantopportunitiesfor childrento participate
in householdroutines.Small amountsof market
work,such as babysittingandnewspaperdelivery,
can also providetrainingin importantskills and
responsibilitiesat a young age. Althoughit is ar-

This content downloaded from 201.213.38.62 on Thu, 12 Sep 2013 15:22:24 PM

All use subject to JSTOR Terms and Conditions

297

American Children and Time

gued that youth need to be engaged in the lives

of their families and communities(Zill, Collins,
West, & Hausken,1995), researchershave demonstratedthat childrenspend very little time in
any of these activities (Goldscheider& Waite,
1991;Medrich,Roizen,Rubin,& Buckley,1982).
Mealtime,whenparentsandchildrengatheras a
familyto sharetheirdailyexperiences,has beenlittle discussedin the literature
but is likely to be an
importantcorrelateof children'swell-being.Time
spent in meals at home is likely to be associated
with a morestable,organizedfamilylife andtherefore withchildrenhavingfewerbehaviorproblems.
Finally,sleep routinesand sleep time areimportant
componentsof individualand family time and
shouldbe associatedwithchilddevelopment.
FACTORSASSOCIATEDWITH DIFFERENCESIN

TIMEUSE
Children'stime is likely to be affected,first and
foremost,by theirage. Froma developmentalperspective, systematicchangesin children'sactivities-a decline in sleeping, eating, and playing
and an increasein school and otherstructuredactivities-occur as they grow and mature(Robinson & Bianchi, 1997). Althoughless pronounced
among young children,genderdifferencesin activitiesbegin to appearin elementaryschool.Variationin use of time is also likely to be associated
with the employmentstatusof the mother,parental
educationandincome,the numberandage of parents, family size, and race and ethnicity.
As mothersspendmoretime in the workforce,
children'sactivitieschange.Comparedwith children of nonemployedmothers,childrenwith employedmothersspendmoretime in day care.This
includes preschoolprogramsor family childcare
for young childrenand before- and after-school
programsfor school-agechildren.Withmoretime
away, the time women spend doing household
work has declinedsignificantlyover the past several decades(Gershuny& Robinson,1988); children'shouseholdworkmay also decline.Because
the lives of young childrendependon the tightly
scheduledlives of theirworkingparents,we may
see more structuredactivities at the expense of
unstructured
activities.Maternalemploymenttime
may increase at the expense of informalfamily
and personalactivities,such as eating meals and
sleeping, or discretionary activities, such as
church attendanceand visiting. The time spent
readingwith childrenor helping them complete
homeworkmay also decline, althoughresearchin

a Californiasamplefailed to find such an association (Bianchi& Robinson,1997).

Education reflects preferencesfor academic
skills, and higherincome may facilitatespending
on books. We expect that the childrenof better
educated and higher income parents will spend
moretimereadingandstudying.BianchiandRobinson (1997) found that childrenspentmoretime
readingor being readto in householdswith moreeducatedthanless-educatedparents.The sameauthors found an inverse relationshipbetweentelevision viewing time andparentaleducationbutno
significantrelationshipwith income.
Childrenliving with singleparentstendto take
on family responsibilitiesat youngerages thando
children living with two parents (Longfellow,
1979); consequently,we expect them to participate more in household chores. We expect the
time spent in day-careprogramsto increase as
well becausea single parentcannotrely on another parentfor day care, and we expect these children to spend less time playing at home. Television watching,used as an inexpensivebabysitter,
may increase.Childrenof single workingparents
may exhibitlower participationin secularand religious organizationalactivities.Previousresearch
failed to find a relationshipbetween the number
of parentsand 3- to 11-year-oldchildren'sreading, studying,or televisiontime (Bianchi& Robinson, 1997), nor did it find any associationbetween the numberof parentsandthe timechildren
spent in householdwork. But previousresearch
did not takeinto accountthejoint influenceof the
numberandemploymentof parentsas we do here.
Smallfamilysize mayhavepositiveornegative
effects. In smallerfamilies,parentsmay increase
the time spentin educationaland leisureactivities
(Blake, 1989). Havingfewer childrenalso means
that siblings are less availablefor play and help
with homework(Zajonc & Markus, 1975). Research has found that childrenin large families
spendmore time in householdwork thando childrenin smallfamilies(Bianchi& Robinson,1997).
Culturaldifferences reflected in family race
and ethnicity may also affect children'sactivity
time. With their greateremphasis on familism,
Hispanic families may involve their childrenin
householdchores(Taylor,1994);withtheirgreater
concernfor academicsuccess,Asianfamiliesmay
promotemore studying(Kao, Tienda,& Schneider, 1996) thanmajorityfamilies.Black andolder
parent families may encourage involvement in
churchactivities(Taylor& Chatters,1988).

This content downloaded from 201.213.38.62 on Thu, 12 Sep 2013 15:22:24 PM

All use subject to JSTOR Terms and Conditions

Journal of Marriage and Family

298
IMPLICATIONSOF TIME FOR CHILD
DEVELOPMENT

Because daily time must add up to 24 hours,

spendingtime on some activitiesmeansless time
on others.It is importantto know the relativebenefit to children'scognitive and emotionaldevelopment of an additionalhour spent reading,for
example,versusan additionalhourspentwatching
television.Based on the literature,we expecttime
spentreadingand studyingto be linkedto verbal
achievement,and time spent watchingtelevision
to be associatedwith poorerverbalscores. There
is less guidancefrom the literatureon the effects
of time expenditureson children'ssocioemotional
behavior.Familyand social activities,such as going to church,eatingmeals, andvisitingwith others, are expectedto be relatedto fewer behavior
problemsbecause those activities provide social
support.Participationin structuredactivitiessuch
as sportsis also likely to be associatedwith fewer
behaviorproblems,bothbecausechildrenwho are
aggressiveor withdrawnare not likely to participate and because such activitiesare likely to improve children'sself-esteem throughthe developmentof physicaland social skills.
DATA AND METHOD

The ChildDevelopmentSupplementto the Panel

Studyof IncomeDynamics
The studysamplewas drawnfromthe 1997 Child
Development Supplement(CDS) to the Panel
Studyof IncomeDynamics(PSID),a 30-yearlongitudinal survey of a representativesample of
U.S. men, women, children,and the families in
which they reside. In 1997, the PSID addeda refreshersampleof immigrantsto the UnitedStates
so that the samplerepresentsthe U.S. population
in 1997. When weights are used, the PSID has
been found to be representativeof U.S. individuals and their families (Fitzgerald,Gottschalk,&
Moffitt, 1998).
With funding from the National Instituteof
ChildHealthandHumanDevelopment(NICHD),
datawere collectedin 1997 on up to two randomly selected0- to 12-year-oldchildrenof PSID respondents,both from the primarycaregiversand
from the childrenthemselves.The CDS survey
periodbeganin March1997 andendedin December 1997 with a breakfrommid-JunethroughAugust; thus, the study took place only duringthe
school year. Interviews were completed with

2,380 child householdscontaining3,563 children.

The response rate was 88%. Poststratification
weights based upon the 1997 CurrentPopulation
Surveywereused to makethe datanationallyrepreflectthe charresentative.Samplecharacteristics
acteristicsof the populationof childrenunderage
13 in the UnitedStatesin 1997.
TimeDiaries
Most surveys obtain estimatesof time expenditures by asking parentsdirectlyhow much time
they spendin certainactivities,such as readingto
theirchild. Althoughsimpleand widely used, this
methodis subjectto social desirabilitybias. Parents reportmoretime spenton desirableactivities
(such as readingto a child) thanon less desirable
ones (Hofferth,1999). Second,the validityof the
answers is poor when activities are infrequent
(Juster & Stafford, 1985; Marini & Shelton,
1993). In contrast, substantial methodological
work has establishedthe validityandreliabilityof
data collected in time-diaryform (Juster& Stafford). For example, estimates of elementary
school age children's time spent in television
viewing range from 13 (Bianchi & Robinson,
1997; Timmeret al., 1985) to 30 hoursper week
(Medrichet al., 1982). Most studiesusing diaries
have estimatedtelevisiontime to average13 to 15
hours per week. Studies obtaininghigher figures
may includetime in which the televisionmay be
on but it is not the primaryactivity.The single
study that comparedparent reports with timelapse video home observationreportedthat 5year-old children watched an average of 14.2
hoursaccordingto diariescomparedwith 13.4 actually observed hours (Anderson,Field, Collins,
Lorch, & Nathan,1985). This supportsour contention that diaries producereasonablyaccurate
estimatesof frequentactivities.
The 1997 CDS collecteda completetime diary
for one weekday and one weekend day for 79%
(2,818) of the 3,563 samplechildrenaged 0 to 12.
Comparisonsbetweenchildrenwho provideda diary and those who did not showed no significant
differences on demographiccharacteristics.The
time diary, which was interviewer-administered
either to the parentor to the parentand child,
askedquestionsaboutthe child'sflow of activities
over a 24-hourperiod beginningat midnightof
the randomlydesignatedday. These questionsask
the primaryactivitythatwas going on at thattime,
when it began and ended, and whetherany other
activity was taking place simultaneously.Chil-

This content downloaded from 201.213.38.62 on Thu, 12 Sep 2013 15:22:24 PM

All use subject to JSTOR Terms and Conditions

299

American Children and Time

dren'sactivitieswere firstassignedto 1 of 10 general activitycategories(e.g., sportsand activeleisure) and then coded into 3-digit subcategories
(e.g., playing soccer). Coding was conductedby
professionalcoders employedby the datacollection organization;the level of reliabilityexceeded
90%.Forthis study,the primaryactivitiesof childrenaged 3 to 11 wereclassifiedintothe 18 major
categoriesused by Timmerand colleagues(1985)
plus day care. Time spent travelingfor the purposes of engagingin a specificactivityis included
in thatcategory.Secondaryactivities,suchas having the television on while doing homework,are
not countedas "watchingtelevision."Thus,some
activitiesthatareoften secondarymay appearunderestimated.Weeklytime is computedby multiplying weekdaytime by 5 and weekendday time
by 2, after removingchildrenwho do not have
both a weekendand weekdaydiary and two who
spent the week visiting. Althoughthe first analyses include all 2,818 childrenunder 12, for the
regressionanalyses we included only the 2,151
childrenaged 3 through12 becausetest scoresare
only available for these children. A few cases
were missing demographicvariables;their omission reducedthe analysis sample for the regressions with controls to 2,123 cases. The actual
numberof cases varies by outcome;not all children were assessed. Furthermore,only children
aged 6 and older were administeredthe Passage
Comprehensionand Calculationtests.
Measurementof the DemographicVariables
The demographicvariablesused to analyze the
impactof family factorson the time of children3
to 12 include age and gender of child; age and
race of head of household;family type and employmentof head and spouse;familyincome;educationof head; and numberof children.Age of
child, age of head, educationof head (years of
completed schooling), income, and number of
children in the family are continuousvariables
with means of 7.5, 38, 13, $51,200, and 2.5, respectively.To make the coefficientscomparable,
income is scaled in tens of thousandsof dollars.
Race and ethnicityis categorical-68% are nonHispanic White, 16% non-HispanicBlack, 11%
Hispanic, 3% Asian, and 2% other races. NonHispanicWhite is the omitted comparisoncategory. Genderis dichotomous,with 0 indicating
male and 1 indicatingfemale; 49% were female
respondents.An interactionbetween gender and
age is includedto capturegender differencesin

activities with age. Finally, to capturethe joint

effect of employmentand family structure,we included six dummyvariables-dual-earnerfamily
(44%), female breadwinner-malenonemployed
family (3%),no breadwinnerfamily (4%), single
employed female-headedfamily (15%), single
nonemployed female-headedfamily (6%), and
single male-headedfamily (3%), with the male
homemakerfamily(25%)the
breadwinner-female
omitted,comparisoncategory.
ChildAssessments
This article examines the association between
children'sactivities and their achievementmeasuredby standardized
tests, controllingfor the deassociated with activity
variables
mographic
choice. Four subtests of the Woodcock-Johnson
Revised Test of Basic Achievementwere used:
letter-wordidentification,a test of children'sability to identify and respondto letters and words;
passage comprehension,a test that measuresvocabularyand comprehensionskills; calculation,a
test of mathematicalcalculationperformance;and
appliedproblems,a test of skill in analyzingand
solving practicalnumericalproblems(Woodcock
& Mather, 1989). Scores are age-standardized
with a mean of 100 and a standarddeviationof
15. In addition,children'ssocioemotionaladjustment is measuredby the BehaviorProblemsIndex. The BehaviorProblemsIndex,developedby
PetersonandZill obtainsparentreportsof the incidence and severity of child behaviorproblems
in a survey setting (Peterson& Zill, 1986). The
30-item scale is divided into two subscales,one
measuring withdrawn or distressed behavior,
called "internalizing,"and the other measuring
aggressive behavior, called "externalizing."
Means for the full scale average40, with a standarddeviationof 8. Reliabilitiesfor the total, internal,and externalscales, as measuredby Cronbach's alpha,are 0.90, 0.81, and 0.86.
RESULTS
Numberof Activities
The resultspresentedfirstare based on the 2,818
childrenbetweenbirthand age 12 whose parents
had completed time diaries for them (or with
them) for 2 days in the previousweek. Children
were reportedto have, on average,22 to 24 activities on a weekdayand24 on a weekend(analyses
not shown).The mean numberof differentactiv-

This content downloaded from 201.213.38.62 on Thu, 12 Sep 2013 15:22:24 PM

All use subject to JSTOR Terms and Conditions

300

Journal of Marriage and Family

TABLE 1. PERCENTAGE OF CHILDREN ENGAGING IN MAJOR ACTIVITIES AND WEEKLY HOURS: MINUTES SPENT BY ALL

BY AGE, 1997
CHILDREN,
0 to 2 Years
(N = 667)

3 to 5 Years
(N = 629)

6 to 8 Years
(N = 639)

9 to 12 Years
(N = 883)

Total
(N = 2,818)

Variable

% Hours: Min

% Hours: Min

% Hours:Min

% Hours: Min

% Hours: Min

Marketwork
Household work
Personal care
Eating
Sleeping
School
Studying
Church
Visiting
Sports
Outdoors
Hobbies
Art activities
Playing
Television
Reading
Household conversations
Other passive leisure
Day care
Total
N/A
Percent of time accounted
for by activities

1
68
99
100
100
15
4
12
48
56
9
2
19
100
71
42
41
53
25

0
81
100
100
100
52
17
24
59
73
19
2
36
98
98
53
38
54
23

2
80
100
100
100
91
55
25
61
76
14
2
23
92
96
43
31
46
13

4
88
100
100
100
90
62
28
69
76
17
4
22
88
94
34
28
52
4

2
80
100
100
100
64
36
23
60
70
15
2
25
94
90
42
34
51
16

0:00
4:40
8:55
11:38
86:01
1:33
0:12
0:27
2:40
2:13
0:25
0:01
0:31
24:55
7:47
1:15
1:11
2:49
9:20
166:36
1:25
99.2%

0:00
6:09
8:32
9:22
76:08
12:05
0:34
1:03
3:21
4:10
0:37
0:05
1:15
17:26
13:49
1:26
0.49
2:37
7:27
166:55
1:05
99.4%

0:06
4:45
7:56
8:08
70:34
32:18
2:09
1:20
3:28
5:17
0:29
0:04
0:44
11:55
12:47
1:09
0:30
1:33
1:42
166:54
1:06

0:16
6:09
7:57
7:25
67:39
33:37
3:40
1:30
3:28
6:26
0:42
0:09
0:56
8:50
13:33
1:15
0:27
2:22
0.26
166:45
1:15

99.3%

99.3%

0:07
5:29
8:18
9:01
74:33
20:55
1:48
1:07
3:15
4:40
0:34
0:05
0:51
15:16
12:04
1:16
0:43
2:21
4.24
166:47
1:13
99.3%

Note: Total weekly time in hours:minutes = 168:00. Weekly times for individualsderived as the sum of 5 times weekday
time + 2 times weekend day time.

ities, the "varietyof activities,"was 11 to 13 on

both weekdayand weekenddays. These estimates
are similar to estimates obtained in earlier research(Timmeret al., 1985).
Participationand TimeSpentin Activities
Table 1 lists the age of childrenengagingin each
of 19 primaryactivitiesin a week, by age, estimatedfrom their time diaries.Fromthis table, it
is clear that all childrenin the sample sleep, eat,
and engage in or receive personalcare and that
most play and watch television. Few engage in
marketwork or hobbies. Otherwise,the types of
activitiesvary by age of the child. For example,
althoughfew preschoolage childrenstudy,more
thanhalf of school-agechildrendo so. Given that
many activitiesare occasional,we would not expect all childrento engage in most of these on a
daily or weekly basis. Nonetheless,we wantedto
abstractfrom the reportsof a representativesample of children'sweeks to describethe activities
of Americanchildren.To do so, we calculatedthe
average time all children spend in an activity,
which was a functionof the proportionof those

who engagein the activityandthe time thoseparticipatingspendin it. Time estimatesbasedon activities with low frequencies,such as hobbiesand
market work, were likely to be unreliableestimatesof the overallallocationof time and arenot
discussedhere.
Table1 also showsthe averagehoursandminutes childrenunderage 13 spent in 19 activities
in an average week in 1997. The bottom row
shows that 99% of the 24-hour period was accounted for, an indicatorof the comprehensiveness of our coding categories.We first focus on
children's nondiscretionarytime; that is, time
spentin school and day care.
Children's time in school. American children
spent about21 hoursper week in school and another4 hoursand 24 minutesin day care in 1997.
As one would expect, time spentin school varies
dramaticallyby age. Childrenunderage 3 spent
almost no time in school. Childrenage 3 to 5
spent 12 hours in school, whereaschildrenaged
6 to 8 and 9 to 12 spent32 to 33 hoursin school
per week. Theseestimatesareconsistentwith earlier research(Timmeret al., 1985), althoughhigh-

This content downloaded from 201.213.38.62 on Thu, 12 Sep 2013 15:22:24 PM

All use subject to JSTOR Terms and Conditions

American Children and Time

er than studies that conductedinterviewsduring
the summermonths(Robinson& Bianchi, 1997)
or that excluded classroomtime not devoted to
schoolwork(Larson& Verma, 1999). Our estimates included travel time to and from school,
which averages about 2 hours per week (not
shown). Childrenunderage 3 spent more than 9
hoursper week in day care, childrenaged 3 to 5
about 7?1/hours, childrenaged 6 to 8 about 13/4
hours,andchildrenaged 9 to 12 about? hourper
week in day care. These estimateswere averaged
over all children,includingthose who spent no
time in the activity.

301
aged 6 to 8 spent about 2 hours and 9 minutes
studying,about26 minutesper school day. Children9 to 12 spentabout3 hours40 minutesstudying, 44 minutesper school day. In contrast,children spent about 1 hour per week reading for
pleasure, with little variation in reading time
among childrenof differentages. At young ages,
parentsread to children;at older ages, children
read to themselves.
Televisiontime can be consideredlearningor
passive leisure, dependingon what childrenare
watching.Childrenwatched about 12 hours per
week during 1997, representing1 of their free
time. Television viewing increased as children
age, fromabout7? hoursfor very youngchildren
to 13? hoursamong9- to 12-year-olds.Afterchildrenreachage 3, televisionviewingremainedstable at approximately13 hoursper week, slightly
less than 2 hoursper day.

Discretionarytime: children'sfree play versus

structuredactivities.To obtainthe amountof free
time childrenhad in 1997, we subtractednondiscretionarytime-time spentin personalcare,eating, sleeping, and school (includingday care)from 168 hours. That amountedto 51 hours or
30% of children'sweek. About half of this free
Family time. Time in householdwork amounted
time was spentin unstructured
to 5? hoursper week in 1997. A substantialcomplay (15 hours)or
televisionwatching(12 hours).Abouthalf an hour ponentof householdworkwas shopping.Children
was spent in outdooractivities. In contrast,4/4
tendedto accompanytheirparentsin this activity,
hours were spent in sports, 1 hour was spent in
which explains the high amountof time among
church,and3 hourswere spentvisiting.Assuming young children.Time spenteatingmeals amountthat the latter representstructuredactivities,the
ed to about 9 hours per week in 1997, about 1
total adds to only 8? hours per week, less than hour and 15 minutesper day. Childrenaveraged
one fifth of theirfree time.
of an hourin conversationas
aboutthree-quarters
The proportionof children's time spent in
a primaryactivity.
structuredactivities (sports,visiting, church)inThe largestcomponentof children'stime was
creasedas childrenaged, from 5? hours among
sleeping,at 74 hoursand 33 minutesper week in
0- to 2-year-oldsto 11? hours among 9- to 121997, or about 10? hoursper night. Sleepingdeclined from 12 hoursper day for infantsand todyear-olds. Concomitantly,the amount of time
spent playing declinedto less than 9 hours (18% dlersto 10 hoursfor olderelementaryschoolchilof time) amongolderelementaryschool age children.Personalcare occupiedabout8 hours.
drenfromalmost25 hoursamonginfantsandtoddlers. The time per week spent watchingteleviFamily Characteristicsand Children'sTime
sion almostdoubledto 13? hoursfrom734hours.
Thus, the amountof structuredtime was larger For the multivariateanalysis,children3 to 12 in
1997 were subselectedfrom the full sample, to
among older than younger children,althoughit
was still only about 22% of their discretionary which their other characteristicswere similar.In
time.
Table2, we presentthe coefficientsfor the impact
The time childrenspentin otherpassiveleisure of family factorson weekly time spentin school,
amountedto only 2 hoursper week. Besidesgoing
day care, play, housework, reading, studying,
to movies and sports events at which children sports, television, church, visiting, and sleeping
were spectators,passive leisureincludeslistening using Tobitregressionmodels. These models adto music and just sitting around.Age variation just for the fact that not all childrenengagedin
was small for this category.
each activity(Tobin,1958). If this techniquewere
not used, the regressionslope would be biasedby
Out-of-schoollearningtime.Childrenspentabout the inclusion of zero values. The coefficientsreflect both the effect of the independentvariable
1? hoursstudying,but therewas substantialvariation by age. As can be expected,preschool-age on the probabilityof the activityand on the hours
children spent little time "studying." Children spentin the activityby participants(McDonald&

This content downloaded from 201.213.38.62 on Thu, 12 Sep 2013 15:22:24 PM

All use subject to JSTOR Terms and Conditions

302

Journal of Marriage and Family

TABLE 2. PARAMETER ESTIMATES OF TOBIT EQUATIONS FOR THE HOURS PER WEEK CHILDREN AGED 3-12
TYPICAL ACTIVITIES

Variable

School

Intercept

-5.64
(3.28)
3.71*
(0.20)
0.02
(0.05)
3.66*
(1.21)
0.66
(1.43)
2.30
(2.48)
5.41
(3.04)
1.71
(1.00)
1.34
(2.46)
0.56
(2.20)
2.44
(1.37)
-1.07
(1.89)
3.20
(2.24)
3.42
(2.23)
-0.55*
(0.27)
0.04
(0.08)
-0.12
(0.16)
0.74*
(0.37)
17.23
(0.32)
-7,643
1,682
441

Age of child
Age of head
Black
Hispanic
Asian
Other race
Two earner
Female breadwinner
No breadwinner
Workingfemale head
Nonworking female head
Male head
Gender of child (1 = F)
Gender*age
Income (0000)
Educationof head
Number of children
Scale
Log likelihood
N (noncensored)
N (censored)

Day Care
-0.82
(11.36)
-7.48*
(0.78)
-0.11
(0.19)
0.28
(3.81)
-7.97
(5.75)
- 14.39
(11.44)
-18.55
(12.33)
32.04*
(4.35)
32.11*
(8.55)
-1.66
(12.61)
41.35*
(5.20)
3.46
(7.56)
32.34*
(7.83)
-14.03*
(6.85)
1.66
(1.01)
0.57*
(0.22)
0.17
(0.61)
-5.68*
(1.47)
34.59
(1.77)
-1,714
295
1,828

Play
23.25*
(1.77)
- 1.19*
(0.11)
-0.04
(0.03)
-3.46*
(0.67)
-2.63*
(0.78)
- 3.27*
(1.34)
-3.62*
(1.68)
-2.26*
(0.54)
-2.08
(1.35)
0.79
(1.19)
-3.32*
(0.75)
0.94
(1.02)
-0.03
(1.23)
0.79
(1.18)
-0.28
(0.15)
-0.01
(0.05)
0.09
(0.09)
0.47*
(0.20)
9.53
(0.16)
-7,331
1,903
220

Church
-11.47*
(2.00)
0.14
(0.12)
0.11*
(0.03)
3.29*
(0.74)
-0.09
(0.88)
-0.54
(1.55)
-1.61
(2.01)
-1.01
(0.58)
-1.54
(1.45)
- 5.45*
(1.53)
-5.25*
(0.92)
-3.63*
(1.18)
-2.18
(1.42)
1.15
(1.32)
-0.15
(0.16)
0.04
(0.04)
0.12
(0.09)
0.13
(0.22)
8.23
(0.29)
-2,594
587
1,536

SPENT IN

Sports
0.54
(1.51)
0.70*
(0.09)
-0.01
(0.02)
-3.56*
(0.57)
-4.23*
(0.68)
-4.37*
(1.18)
-4.68*
(1.47)
0.36
(0.46)
-1.97
(1.16)
1.40
(1.04)
0.58
(0.64)
2.53*
(0.87)
-0.66
(1.04)
-0.54
(1.01)
-0.53*
(0.12)
-0.03
(0.04)
0.08
(0.07)
0.20
(0.18)
7.88
(0.15)
-5,983
1,564
559

Note: Standarderrorsin parentheses.

*p < .05.

Moffitt, 1980). The higherthe proportionof childrenwho participatein the activity,the more the
results reflect the hours among participantsand
thus the more similarthe resultsbecome to those
from ordinaryleast squaresregressionsjust on
participants.We discussthe activitiesaccordingto
the fourissues describedearlier-school anddaycare time, unstructuredversus structuredactivities, out-of-schoollearning,and family time.
School or day care. As expected, enrollmentin
day care (but not in school) is linked to the age
of the child andthe employmentstatusof mothers.

As they aged, childrenspentmore time in school

and less in day care.Childrenof employedmothers, regardlessof whetherthe motherwas married
or single, were more likely to be in day care and
spent more time there than children in a male
homemakerfamily,the omitbreadwinner-female
ted category.Childrenfrom higherincome families spentmoretime in day care,butnot in school,
probablybecauseof theirgreaterabilityto afford
privateprograms.Net of income and parentaleducation,Blackchildrenspentmoretime in school.
This may reflecta differencein school scheduling
patternsin schools with minoritypopulationsor

This content downloaded from 201.213.38.62 on Thu, 12 Sep 2013 15:22:24 PM

All use subject to JSTOR Terms and Conditions

303

American Children and Time

TABLE
2. EXTENDED
Visiting

Housework

Eating

Sleep

Reading

Studying

TV

4.26*
(1.60)
-0.11
(0.10)
-0.05
(0.03)
-0.55
(0.61)
- 1.73*
(0.72)
-1.36
(1.25)
3.25*
(1.46)
0.08
(0.49)
1.57
(1.22)
-1.42
(1.12)
-1.46*
(0.69)
-0.57
(0.95)
-0.23
(1.12)
-4.77*
(1.08)
0.67*
(0.13)
-0.00
(0.04)
0.06

1.88
(1.27)
-0.01
(0.08)
0.02
(0.02)
- 1.85*
(0.48)
2.59*
(0.56)
-1.63
(0.96)
-0.03
(1.20)
0.09
(0.39)
-1.45
(0.98)
-0.24
(0.87)
-0.47
(0.54)
-1.45
(0.75)
-0.27
(0.88)
1.15
(0.85)
0.08
(0.10)
-0.05
(0.03)
0.18*

8.49*
(0.62)
-0.24*
(0.04)
0.02*
(0.01)
0.55*
(0.23)
1.02*
(0.27)
0.32
(0.47)
0.36
(0.58)
-0.37
(0.19)
-0.51
(0.47)
- 1.28*
(0.42)
-0.99*
(0.26)
0.21
(0.36)
-0.88*
(0.43)
0.87*
(0.41)
-0.11*
(0.05)
0.03*
(0.02)
0.04

79.37*
(1.47)
- 1.04*
(0.09)
-0.01
(0.02)
- 1.08*
(0.55)
1.02
(0.65)
1.43
(1.12)
0.87
(1.38)
- 1.04*
(0.45)
-0.05
(1.12)
2.73*
(0.99)
-1.53*
(0.62)
-0.07
(0.85)
-1.17
(1.02)
3.95*
(0.98)
-0.41*
(0.12)
-0.04
(0.04)
0.02

-3.11*
(0.90)
-0.35*
(0.06)
0.02
(0.02)
- 1.29*
(0.36)
-0.89*
(0.42)
1.48*
(0.64)
0.33
(0.82)
-0.52*
(0.26)
- 1.55*
(0.72)
- 1.24*
(0.63)
-2.09*
(0.40)
- 1.52*
(0.56)
-0.66
(0.62)
- 1.62*
(0.58)
0.29*
(0.07)
-0.00
(0.02)
0.36*

- 12.92*
(1.31)
1.05*
(0.08)
0.06*
(0.02)
1.85*
(0.46)
3.23*
(0.53)
2.46*
(0.91)
1.40
(1.19)
-0.70
(0.38)
-0.53
(0.94)
-3.13*
(0.88)
-0.52
(0.52)
-1.04
(0.72)
-1.34
(0.84)
0.65
(0.92)
-0.15
(0.11)
0.02
(0.03)
0.18*

18.58*
(1.80)
-0.06
(0.11)
0.04
(0.03)
2.63*
(0.68)
-0.00
(0.80)
4.77*
(1.37)
-2.47
(1.70)
- 1.17*
(0.56)
-0.20
(1.37)
-1.62
(1.23)
0.05
(0.77)
-0.68
(1.05)
-0.86
(1.26)
-0.97
(1.20)
0.17
(0.15)
-0.09*
(0.05)
-0.46*

(0.08)
-0.36
(0.19)
8.15
(0.17)
-5,308
1,332
791

(0.06)
-0.12
(0.15)
6.74
(0.12)
-6,220
1,696
427

(0.03)
0.02
(0.07)
3.36
(0.05)
-5,573
2,118
5

(0.07)
0.04
(0.17)
7.98
(0.12)
-7,407
2,123
0

differentialuse of school-basedafter-schoolprograms.
Free play versusstructuredactivities.Due to maternal responsibilityfor caring for childrenand,
therefore,the increasedtime such childrenspend
out of the home,free play time at homewas lower
when mothers were employed than when they
were not. Consistentwith expectations,as childrengrow older,they spendless time playingand
more time in other activities. Interestingly,the
time spent playing declinedmarginallyfasterby
age for girls thanfor boys. Girls'time in activities
such as readingand visiting increasedfasterthan

(0.05)
-0.10
(0.11)
4.15
(0.11)
-3,193
807
1,316

(0.06)
-0.10
(0.14)
5.84
(0.14)
-3,799
1,020
1,103

(0.09)
-0.06
(0.21)
9.78
(0.16)
-7,607
2,043
80

thatfor boys, whereasboys' time increasedfaster

in sports. Net of other factors, Black, Hispanic,
and Asian childrenspent less time playing than
White non-Hispanicchildren. Childrenin large
families spentmoretime playing,mostlikely with
each other.The head's educationand family income were not relatedto children'splay time.
Comparedwith childrenin male breadwinnerfemale homemakerfamilies, childrenin all other
family types spentless time in church.Giventraditional churchteachingswith regardto divorce
and nonmaritalchildbearing,we suspectthatfamilies thatdo not fit prevalentlifestyle andparenting norms were simply less likely to attendreg-

This content downloaded from 201.213.38.62 on Thu, 12 Sep 2013 15:22:24 PM

All use subject to JSTOR Terms and Conditions

304

Journal of Marriage and Family

ularly. Althoughbusy schedules are part of the

story, it is not simply a matterof having a busy
schedule because children in families with no
breadwinneror with a nonemployedfemale head
also spentless time in church.As expected,childrenfromBlackfamiliesandfamilieswithan older head spent more time in church.Churchtime
offset the lower play time of Black children.
Family type is importantto sports participation. Childrenof nonworkingfemale heads spent
more time in sportsthan did childrenliving in a
male breadwinner-femalehomemaker family.
Perhapssuch parentsutilize sportsprogramsas a
source of supervisionfor their children,whereas
childrenof employedfemale heads utilize formal
child-careprograms.As childrenaged, they spent
moretime in sportsactivities.Althoughtherewas
no significantoverall genderdifferencein sports
participationat youngerages, as they aged, girls'
participationdeclinedrelativeto thatof boys, reflectingan increasinggendergap. Despitethe fact
that sportsis an avenuefor skill-buildingandupward mobility,minoritychildrenspent less time
in sportsthan did White children,all else equal.
This may reflect differentialaccess to such programs.Childrenin largefamilies spentmoretime
in sports,perhapsdue to exposurethrougholder
siblings. Neither parentaleducationnor income
was associatedwith participationof childrenin
sportsactivities,net of otherfactors.
The time spent socializingand participatingin
privatelysponsoredclubs and programswas also
linked to family structureand employment.Children from families with a working female head
were less likely to participatein such activities
thanwere childrenin a male breadwinner-female
homemakerfamily, probablybecause they were
too busy. Girls spent significantlyless time at
young ages visiting than boys; however,as they
grow older,girls' time in such activitiesincreases
fasterthanboys' time. Childrenin largerfamilies
spent marginallyless time visiting thandid those
in smallerfamilies,perhapsbecauseof the greater
opportunityfor social activitieswithinthe family.
There were no differencesin visiting by family
income or education.
Family time. In contrastto expectations,no differences in children'shousehold work time by
family structureand maternalemploymentwere
found. Children did not help more or less in
householdchores if their motherwas employed.
Consistentwith greaterfamilism,childrenin Hispanicfamiliesspentmoretime in householdwork

than did childrenin Whitenon-Hispanicfamilies,

whereas children in Black and Asian families
spentless time. For Hispanicchildren,the greater
amount of household work offset the lower
amountof time spentplaying.Thereareno gender
differencesin time spentin householdwork;these
childrenwere still ratheryoung. Finally,in families with a bettereducatedhead,childrendo more
housework.This may be due to greaterexpectations for childrenin such familiesor to moretime
spent shopping.
Although maternalemploymentis associated
with reducedtime childrenspend eating, the income it bringspermitsmoretime eating.Cultural
differencesin eating patternsmay be reflectedin
the findingthatchildrenfromBlack andHispanic
families spentmore time eatingthandid children
fromWhitefamilies.As childrenaged, they spent
less time eatingand moretime in otheractivities.
Childrenliving with olderparentsalso spentmore
time eatingmeals.
Maternal employment was linked to sleep.
Childrenin dual-earnerfamilies and childrenin
families with a working female head spent less
time sleepingthandid childrenin male breadwinner-female homemakerfamilies. If there was no
breadwinner,childrengot more sleep. Sleep deof course.As
pendedon children'scharacteristics,
expected, older childrenslept less than younger
ones. Girls slept longerthanboys, but the difference declinedwith age. Blackchildrenspentmore
time in otheractivities.As a result,they sleptless
thanWhite children.
Out-of-schoollearning. Children in all family
typesreadless thanchildrenin malebreadwinnerfemale homemakerfamilies. The amountof differencein readingtime varied,however.The difference between children's reading time in a
family and a male breadworking-female-headed
winner-femalehomemakerfamily (-2.09) was
larger than the differencebetween a two-parent
dual-earner and a male-breadwinnerfamily
(-0.52). These differencesreflect differencesin
time spent at home and availabilityof a second
parent.Consistentwith the emphasison learning
in the Asian community(Kao et al., 1996) and
offsetting some of the lower time spent playing,
Asian childrenspentsignificantlymoretimereading. In contrast,BlackandHispanicchildrenspent
significantlyless time readingthan White nonHispanic children.Older children spent time in
otheractivities;consequently,readingfor pleasure
declinedas childrenages. Girlsreadless perweek

This content downloaded from 201.213.38.62 on Thu, 12 Sep 2013 15:22:24 PM

All use subject to JSTOR Terms and Conditions

American Children and Time

305

TABLE 3. REGRESSIONS OF ACHIEVEMENT AND BEHAVIOR PROBLEMS ON ACTIVITIES, WITH CONTROLS FOR FAMILY AND
CHILD BACKGROUNDa

Behavior Problems

Woodcock-JohnsonAchievement Test

Activity
Intercept
Marketwork
Personal care
Conversation
Passive leisure
Housework
Eating
Sleeping
In school
Studying
In church
Visiting
Playing sports
Outdoors
Hobbies
Art
Playing
Watchingtelevision
Reading
Day care
R2

Letter-Word
Passage
Applied
ComprehensionComprehension Problems Calculation
82.45*
0.11
-0.11
-0.19
-0.11
-0.02
0.32*
-0.03
0.07
0.14
0.15
0.18*
0.02
-0.16
0.56
0.27
-0.04
-0.04
0.75*
-0.07
0.21
1,726

89.84*
0.23
-0.06
0.47
-0.01
0.13
0.12
-0.04
0.07
0.10
0.21
0.21*
0.08
0.09
0.22
0.26
-0.02
-0.01
0.53*
0.13
0.21
1,218

83.55*
0.43
-0.09
0.06
-0.23*
0.13
0.27*
-0.08
0.15*
0.11
0.10
0.22*
0.20*
0.01
0.05
0.10
0.06
-0.03
0.65*
0.12
0.25
1,719

88.30*
0.64*
-0.07
0.84
0.01
-0.03
0.14
-0.11
0.12
0.00
0.29
0.25*
0.16
0.12
-0.31
-0.21
0.06
-0.01
0.58*
0.10
0.17
1,211

Total
Behavior
Problems

External
Behavior
Problems

Internal
Behavior
Problems

49.48*
-0.07
-0.02
0.17
-0.03
-0.01
-0.19*
-0.04
-0.04
0.03
-0.02
-0.07
-0.07*
-0.00
-0.01
-0.11
-0.00
0.00
-0.02
-0.02
0.07
2,121

28.71*
0.01
0.01
0.09
-0.02
-0.01
-0.14*
-0.01
-0.03
0.03
-0.01
-0.03
-0.04*
0.03
-0.01
-0.07
0.01
0.01
-0.01
-0.01
0.07
2,121

18.79*
-0.08
-0.02
0.06
-0.01
0.00
-0.04
-0.03*
-0.02
0.01
-0.03
-0.03
-0.03
-0.03
-0.01
-0.03
-0.01
-0.01
0.01
-0.01
0.08
2,121

aControlsare age and gender of child, race and ethnicity, education and age of head, family structureand employment,
income, and family size.
*p < .05.

than boys, but, with age, their readingtime increasedrelativeto that of boys. Not surprisingly,
childrenof more-educatedheads read more. It is
likely that the parentsreadmore themselves,had
more books aroundthe home, and encouraged
their childrento read.
Consistentwith the previousresultsindicating
that they read less for pleasure, older children
studiedmore than youngerchildren.Black, Hispanic, and Asian childrenspentmoretime studying than did White non-Hispanicchildren,net of
other factors. Factors associated with parental
nonemploymentmay also be linkedto lowerstudy
time becausewe foundthatchildrenstudyless in
familiesin whichtherewere two nonworkingparents. Familyincomewas not relatedto time spent
studying;however,the educationof the head was
related.Childrenof better educatedheads spent
more time studyingthan did childrenof less-educatedheads.Finally,childrenin familieswith an
older head spent more time studying.Olderparents wereprobablymoreknowledgeableaboutthe
importanceof studyingandencouragedtheirchildrento do so.
Children'stelevision viewing was related to

parental employment. Children in dual-earner

families watchedabout 1 hour less televisionper
week than did those with a male breadwinner.
Such children spent less time at home during
whichthey couldwatchtelevision;they werelikely to be in school or preschool.Race differences
were striking.Black and Asian childrenwatched
considerablymoretelevisionthandid White,nonHispanicchildren,22/3 hoursmore among Black,
and almost5 more hoursper week amongAsian
children. Finally, parental education mattered.
Childrenof bettereducatedparentswatcheda half
hourless televisionper week thandid childrenof
less-educatedparents.Childrenof higherincome
parentsalso watchedless television,althoughthe
effect is small.
Associationof Children'sTimeWithCognitive
and BehavioralOutcomes
In Table 3, we show the results of regressing
achievementandbehavioron children'sactivities.
Althoughwe cannotdeterminecausalitybecause
activities and achievementare measuredconcurrently,we foundnumerousactivitiesto be signif-

This content downloaded from 201.213.38.62 on Thu, 12 Sep 2013 15:22:24 PM

All use subject to JSTOR Terms and Conditions

306

Journal of Marriage and Family

icantly associated with scores on achievement

tests and on the BehaviorProblemsIndexand its
subscales(not shown).These associationsareprimarily due to family factors affectinghow children's time is spent. When demographicfactors
that affect how time is spentwere included,only
a few activitiesretaintheirsignificantassociation
with cognitiveachievementand behavior,and we
focus only on those.
Time in learningactivitieswas particularlyimportantto achievement.Besides spendingtime in
school, which was associatedwith higher scores
on the appliedproblemstest, spendingmoretime
readingfor pleasurewas stronglyassociatedwith
higherscores on all achievementtests. Television
and studyingtime were not associatedwithhigher
or lower scores on any tests.
Family time was importantto both achievement and behavior.Spending more time eating
meals was associatedwith a higher score on the
letter-wordand appliedproblemstests. It was also
associatedwith a reductionin total,externalizing,
and,marginally(p < .10), internalizingproblems.
Time spent sleeping is associatedwith fewer internalizingbehaviorproblems.
Active leisure was more productivethan passive leisure. More time spent playing sportswas
associated with a higher score on the applied
problems test and with reduced total behavior
problems,externalizingproblems,and,marginally
(p < .10), internalizingproblems.Timespentvisiting was associatedwith higherscoreson all four
cognitive tests. In contrast,time spent in passive
leisure was associatedwith lower scores on the
appliedproblemstest.
ANDCONCLUSIONS
SUMMARY
In 1997, 55% of an average child's week was
spenteating,sleepingor in personalcare, with an
additional15%spent in school or day care. This
leaves only 30% of children'stime as discretionary, the primaryfocus of our research.Of this 51
hours, free play comprised 15 hours per week
(29%) and television viewing about 12 hoursper
week (24%)of theirfree time. In contrast,structured activities comprised9 hours (18% of free
time) less than theirtime spentplaying.Children
spent little time in educationalactivities: 1 hour
readingand 1 hour and 48 minutesstudying.Art
activities, household work, conversations, and
otherpassive leisurecompletethe week.
One of the objectivesof this researchwas to
identify the factors associatedwith variationin

how childrenspendtime. Of the varietyof activities in which childrenparticipate,we focusedon

four majorcategoriesof activities-school, structured and unstructuredactivities, out-of-school
learning, and family time. In contrastto earlier
research,we found consistentevidence that maternalemploymentaffects children'stime during
the school year.Childrenspendmore time in day
care if they live with an employedmother.As a
result,they spendless time in everythingelse, includingplay, structuredactivitiessuch as church,
family activitiessuch as eating and sleeping,and
learningtime such as reading.They also watch
televisionless.
Parentaleducation,which reflectsknowledge,
preferences,andvalues,is positivelyrelatedto educationalactivities such as readingand studying
and negativelyassociatedwithtelevisionviewing.
Family structureand size affect the abilityof parents to monitortheir childrenand the availability
of playmates.Childrenfrom single female families spendless time in educationalactivitiessuch
as reading,regardlessof the employmentstatusof
the parent.In contrast,they spend more time in
structuredsports. Childrenfrom larger families
spend more time playing and in sports and less
time visiting than childrenfrom smallerfamilies;
such childrenhave a built-inset of playmates.Income is not as importantas commonlybelieved.
Family income was significantlyassociatedonly
with the time childrenwatchtelevision(negatively) and the time spent eating meals and in day
care (positively).
Race and ethnicity affects most activities.
Many of the differences-such as the greater
amountsof time thatAsian childrenspendin educationalactivities at home, that Black children
spendin churchactivities,and thatHispanicchildren spendin family activitiessuch as eatingand
household work-were not surprising.More research on the sources of race and ethnic differences is needed.
Finally,keepingin mind that we cannotdetermine the causal directionof the effect, we asked
whether time was associated with children's
achievementand behavior.Of the three learning
activities-reading, studying, and televisiononly reading was linked to achievement.We
found that childrenwho spent time readingfor
pleasuredid betteron tests of cognitive achievement. Studyingmay result as much from having
difficultyin school as from motivationto excel;
greaterstudyingmay characterizeboth low and
high achievers.Althoughoften negativein direc-

This content downloaded from 201.213.38.62 on Thu, 12 Sep 2013 15:22:24 PM

All use subject to JSTOR Terms and Conditions

307

American Children and Time

tion, more television viewing was never significantly detrimentalto children'sachievement.Not

surprisingly,spendingmore time in school was
linked to higherachievement.
Structuredactivities were linked to both cognitive and emotional development.Those who
playedsportswere betterproblemsolversandhad
fewer emotionalproblems.Perhapsthe cognitive
skills learned on the playing field contributeto
problemsolving,andthe social andphysicalskills
developed contributeto better emotionaladjustment. Alternatively,children lacking cognitive
and social skills may not participatein sportsactivities.Involvementwith othersin visiting,which
includesparticipationin youthorganizations,was
linked to greaterachievementon all tests. Time
spent playing per se was associatedneitherwith
achievementnor with behaviorproblems.Passive
leisure was found to be associatedwith neither
poorer test scores nor behaviorproblems;these
findingssupportthe conclusionthat active forms
of leisure promotechildren'sdevelopmentmore
thanpassive forms.
Time spent in family activities is associated
with fewerproblembehaviors.We foundthatchildren who spent more hours eating meals and
sleeping had lower levels of behaviorproblems
than did those who spent fewer hours eating or
sleeping. During mealtime,childrenand parents
can discuss what happenedduringthe day. This
is not the only time childrenand parentsspend
talking,but childrenspentonly about45 minutes
sitting and talking as the main activity in 1997.
Sleep is also importantto children'swell-being.
Of course, it may be that childrenwho have behaviorproblemsboth sit still and sleep less. This
researchwas conductedat only one pointin time.
Longitudinalresearchis neededto demonstratea
causal relationshipbetween activities and child
achievementand behavior.
Parents often seek informed guidance as to
how to directtheirchildren'sactivities.This article has shown that the amountof time spent at
home eating, sleeping, and reading is linked to
children'sachievementandbehavior.Nonetheless,
schools, day care centers,and before- and afterschool programsexert importantinfluences on
children'slives today as time spent at home declines. Out-of-homesportsparticipationand visiting represent importantaspects of children's
lives bothin the amountof time spentandin their
relationshipto cognitiveand socioemotionalwellbeing. As childrenspendless time at home, however,time spentreading,sleeping,andeatingmay

decline. A balance of out-of-home and homebased activitiesmay be desirable.

NOTE

Fundingfor this researchwas providedby Grant

U01HD37563 from the National Institute of Child
Healthand HumanDevelopmentand by the Centerfor
the Ethnographyof EverydayLife, an Alfred P. Sloan
Center for the Study of WorkingFamilies.Thanksto
DianaBlackfordfor researchassistance.
REFERENCES

D., Field,D., Collins,P.,Lorch,E., & NaAnderson,

time
than,J. (1985).Estimatesof youngchildren's
A methodological
of parwithtelevision:
comparison

ent reportswith time-lapsevideo home observation.

Child Development, 56, 1345-1357.

Bianchi,S., & Robinson,J. (1997).Whatdidyou do

use of time,familycomposition,
today?Children's
and the acquisitionof social capital.Journalof Mar-

riage and the Family, 59, 332-244.

Blake, J. (1989). Family size and achievement. Berke-

ley: Universityof CaliforniaPress.

Bredekamp,S., & Copple,C. (1997). Developmentally
appropriate practice in early childhood programs.

Washington,DC: NationalAssociationfor the Educationof YoungChildren.

P.,& Moffitt,R. (1998).An

J., Gottschalk,
Fitzgerald,
analysisof sampleattritionin paneldata:The Michigan Panel Study of Income Dynamics.Journal of

Human Resources, 33, 251-299.

H.W.(1995).Timeuseand
Fuligni,A. J.,& Stevenson,
mathematicsachievementamongAmerican,Chinese,
and Japanesehigh school students.Child Development, 66, 830-842.

Gershuny,J., & Robinson,J. P. (1988). Historical

changesin the householddivision of labor.Demog-

raphy, 25, 537-552.

Goldscheider,E K., & Waite,L. J. (1991).New families,

no families. Berkeley:Universityof CaliforniaPress.
Hanushek,E. (1989). The impactof differentialexpenditures on school performance.Educational Researcher, 18 (4),45-55.

Haveman,R., & Wolfe, B. (1994). Succeedinggenerations: On the effects of investments in children. New

York:Russell Sage Foundation.

S. (1996).Childcarein theUnitedStatestoHofferth,
day. The Future of Children, 6 (2),41-61.
Hofferth, S. (1999, May). Family reading to young children: social desirability and cultural biases in re-

porting. Paper presentedat the National Research

CouncilWorkshopon Measurementof andResearch
on Time Use, Committee on National Statistics,

DC.
Washington,

Juster, E, & Stafford, E P. (1985). Time, goods, and

well-being.Ann Arbor,MI: Institutefor Social Research.

Juster,E T, & Stafford,E (1991). The allocationof

time: Empiricalfindings,behavioralmodels,and

problemsof measurement.Journalof EconomicLiterature,29, 471-522.

Kao, G., Tienda,M., & Schneider,B. (1996). Racialand

In A. M.
in academicperformance.
ethnicvariation

This content downloaded from 201.213.38.62 on Thu, 12 Sep 2013 15:22:24 PM

All use subject to JSTOR Terms and Conditions

308

Journal of Marriage and Family

Pallas (Ed.), Research in Sociology of Education and

Robinson,J. P., & Bianchi, S. (1997). The children's

Socialization (pp. 263-297). Greenwich,CT: JAI

Press.
Keith,T., Reimers,T., Fehrmann,P., Pottebaum,S., &
Aubey, L. (1986). Parentalinvolvement,homework,
and TV time: Direct and indirect effects on high

DC: NationalAcademyPress.
TaskForceon YouthDevelopmentandCommunityPro-

Koolstra,C. M., & Van Der Voort,T. H. (1996). Longitudinaleffects of television on children'sleisuretime reading:A test of threeexplanatorymodels.Hu-

poration.
Taylor,R. J., & Chatters,L. M. (1988). Churchmembers
as a sourceof informalsocial support.Reviewof Re-

Larson,R., & Verma,S. (1999). How childrenand adolescents spend time across the world:Work,play,
and developmentalopportunities.PsychologicalBul-

Taylor, R. L. (1994). Minority families and social

school achievement. Journal of Educational Psychology, 78, 373-380.

man Communication Research, 23, 4-35.

letin, 125, 701-736.

Longfellow,C. (1979). Divorce in context:Its impact

on children.In G. Levinger& O. Moles (Eds.), Divorce and separation: context, causes, and conse-

quences(pp. 287-306). New York:Basic Books.

Marini, M. M., & Shelton, B. A. (1993). Measuring
household work: Recent experience in the United
States. Social Science Research, 22, 361-382.

McDonald,J., & Moffitt,R. (1980). The uses of Tobit

analysis. Review of Economics and Statistics, 62,

318-321.
Medrich, E., Roizen, J., Rubin, V., & Buckley, S.

(1982). The serious business of growing up: A study

of children's lives outside school. Berkeley: Univer-

sity of CaliforniaPress.
Pellegrini,A., & Smith,P. K. (1998). Physicalactivity
play: The natureand functionof a neglectedaspect
of play. Child Development, 69, 577-874.

Peterson,J. L., & Zill, N. (1986). Maritaldisruption,

parent-childrelationships,andbehavioralproblemsin
children. Journal of Marriage and the Family, 48,

295-307.

hours. American Demographics, 20, 22-24.

Snow, C., Burns,M. S., & Griffin,P.(1998). Preventing

reading difficulties in young children. Washington,

grams. (1992). A matter of time: Risk and opportunity

in the nonschool hours. New York: Carnegie Cor-

ligious Research, 30, 193-203.

change. In R. L. Taylor (Ed.), Minority families in the

United States: A multicultural perspective (pp. 204-

248). EnglewoodCliffs, NJ: PrenticeHall.

Timmer,S. G., Eccles, J., & O'Brien,K. (1985). How
childrenuse time. In E Juster& E P. Stafford(Eds.),
Time, goods, and well-being (pp. 353-382). Ann Ar-

bor,MI: Institutefor Social Research.

Tobin,J. (1958). Estimationof relationshipsfor limited
dependentvariables.Econometrica,26, 24-36.
U.S. Departmentof Education.(1999). Startearly,finish strong: How to help every child become a reader

[AmericaReads Challenge].Washington,DC: Author.

Woodcock,R., & Mather,N. (1989). W-J-RTests of
Achievement: Examiner's manual. In: R. W. Wood-

cock & M. B. Johnson (Eds.), Woodcock-Johnson

Psycho-EducationalBattery-Revised. Allen, TX:
DLM TeachingResources.
Zajonc,R. B., & Markus,G. B. (1975). Birthorderand
intellectual development. Psychological Review, 82,

74-88.
Zill, N., Collins, M., West, J., & Hausken,E. (1995).
Approaching kindergarten: A look at preschoolers in

the UnitedStates (NCES 95-280). Washington,DC:

U.S. Departmentof Education.

This content downloaded from 201.213.38.62 on Thu, 12 Sep 2013 15:22:24 PM

All use subject to JSTOR Terms and Conditions