journal of family issues 2003 hofferth 185 216

8/12/2019 Journal of Family Issues 2003 Hofferth 185 216

http://slidepdf.com/reader/full/journal-of-family-issues-2003-hofferth-185-216 1/33

http://jfi.sagepub.com/ Journa l of Family Issues

http://jfi.sagepub.com/content/24/2/185The online version of this article can be foun d at:

DOI: 10.1177/0192513X02250087 2003 24: 185Journal of Family Issues

Sandra L. Hofferthace/Ethnic Differences in Father Involvement in Two-Parent Families: Culture, Context, or Economy?

Published by:

http://www.sagepublications.com

can be found at:Journal of Family Issues Additional services and information for

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

http://jfi.sagepub.com/subscriptionsSubscriptions:

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

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

http://jfi.sagepub.com/content/24/2/185.refs.htmlCitations:

What is This? - Mar 1, 2003Version of Record>>

at WESLEYAN UNIV on July 25, 2014 jfi.sagepub.comDownloaded from at WESLEYAN UNIV on July 25, 2014 jfi.sagepub.comDownloaded from

http://jfi.sagepub.com/


http://jfi.sagepub.com/content/24/2/185


http://www.sagepublications.com/

http://jfi.sagepub.com/cgi/alerts


http://jfi.sagepub.com/subscriptions

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


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

http://jfi.sagepub.com/content/24/2/185.refs.html


http://online.sagepub.com/site/sphelp/vorhelp.xhtml


http://jfi.sagepub.com/content/24/2/185.full.pdf











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


http://jfi.sagepub.com/subscriptions


http://www.sagepublications.com/





10.1177/0192513X02250087 ARTICLEJOURNALOFFA MILYISSUES/Mar ch2003Hofferth/ RACE/ETHNICDIFFERENCESINFA THERS

Race/Ethnic Differences in FatherInvolvement in Two-Parent Families

Culture, Context, or Economy?

SANDRA L. HOFFERTHUniversity of Maryland

This article examines the contribution of economic circumstances, neighborhood context,and cultural factors to explainingrace/ethnic differences in fathering in two-parentfamilies.Data come from the1997ChildDevelopment Supplement to thePanelStudyof IncomeDy-namics, a nationally representative sample of children younger thanage 13.Black children’sfathers exhibit less warmth but monitor their children more, Hispanic fathers monitor theirchildren less, and both minority groups exhibit more responsibility for child rearing thanWhite fathers. Economic circumstances contribute to differences in paternal engagementand control, and neighborhood factors contribute to differences in warmth and responsibil-ity. Cultural factors, suchas intergenerational fathering and gender-role attitudes, contributeto explainingdifferences fromWhites in control and responsibility on the partof bothBlacksand Hispanics.

Keywords: fathers; parenting; race; ethnicity

Thelag in academic achievement of minoritycomparedwith nonminoritystudents is a criticalnational problem (Jencks & Phillips,1998).Althoughthe substantial socioeconomic difference between their families is a key

factor in achievement differentials, the gap between minority and Whitestudents is present or even greater among middle-income than low-incomefamilies(Jencks & Phillips,1998). Contributingfactorsotherthan incomemay involve parenting practices. Although at least one study has exam-ined maternal behavior (Phillips, Brooks-Gunn, Duncan, Klebanov, &Crane, 1998), few have examined the paternal side. A father’s emotionalinvestment, attachment, and provision of resources are associated withchildren’s cognitive development and social competence (Lamb, 1997;Pleck, 1997), and suchinvestmentvarieswidely. Differentialachievementamong various groups may result from differential father involvement(McLanahan & Sandefur, 1994). Yet, in spite of the growing body of re-search examining fathersand fathering, researchfocuses primarily on ma-

jority, not minority, families (McAdoo, 1988).

185

JOURNAL OF FAMILY ISSUES, Vol. 24 No. 2, March 2003 185-216DOI: 10.1177/0192513X02250087© 2003 Sage Publications

at WESLEYAN UNIV on July 25, 2014 jfi.sagepub.comDownloaded from







In addition, research often focuses on father involvement by fathersoutside the home rather than by coresidential fathers (Mott, 1990; Seltzer& Bianchi, 1988).Minority children areespecially likelyto grow up with-out a father present (Hernandez, 1993). Although the part played by ab-sent fathers is important, it is also importantto assessfathering behavior intwo-parent families; research has found predivorce behavior to explainmuchof the relationship between divorceand children’s outcomes,partic-ularly for boys (Cherlin et al., 1991). Fathers who are more involved withchildren predivorce may remain more involved postdivorce. Including re-marriage families is important; little research exists on fathering after di-vorce and remarriage.

Fathering patterns may vary by race/ethnicity for several reasons. Dif-

ferent fathering patterns by race/ethnicity may be due to different eco-nomiccircumstances and neighborhood environments. Differentialpater-nal involvement may also result from attitudes and values that areculturalin origin because parenting occurs in a social context (Bronfenbrenner,1979). Given that fathering plays an important part in children’s develop-ment, research is needed to determine whether economic conditions, cul-tural beliefs, or other factors are more important in explaining majority-minority fathering differences. Each has very different implications forpublic policy efforts to increase father involvement.

This article examines factors associated with differential involvementwith fathers among Black, White, and Hispanic children in two-parentfamilies, using data drawn from a new study of children and their parentsconducted in 1997. This article takes advantage of a rich set of parenting

measures compared to those used in past studies, a new set of measures of the time parents spend with individual children, and attitude and behaviordata obtained directly from both mothers and fathers. Although manystudies of fathers and fathering are based on small selective samples, thisstudy is based on a large nationally representative database that can begeneralized to all children living in two-parent families in the UnitedStates.

THEORETICAL PERSPECTIVES

KEY PARENTING PRACTICES

Children learn through interacting with and observing parents(Bandura, 1969), and such learning is affected by the level of interactionor engagement (Lamb, Pleck, Charnov, & Levine, 1985). Although moth-

186 JOURNAL OF FAMILY ISSUES / March 2003








ers’ engagement is generally high, fathers vary in the level of engagementwith children, and research has attempted to quantify this variation by ex-amining the time fathers spend with them. Recent research finds that fa-thers spend about 1.5 hours on a weekday and about 3 hours on a weekendday with children younger than age 13 (Yeung, Sandberg, Davis-Kean, &Hofferth, 2001). Other studies have shown levels of engagement rangingfrom 2 to 2.8 hours per day (Pleck, 1997). Fathers also vary in the degreeof responsibility they assume for their children. Responsibility encom-passesa father’s contribution tohisfamily inhisroleas economicprovider(McAdoo, 1988; Pleck, 1997), and employment and work hours are keyindicators of this dimension. However, responsibility also encompasses adegree of management of the child’s welfare—making sure that the child

is fed, clothed, housed, monitored, managed, examined by physicians,and cared for when needed, a dimension thathas not been examined in thecontext of race/ethnic differences. Few men take primary responsibility,but many share it with their partners (Hofferth, Pleck, Stueve, Bianchi, &Sayer, 2002; Sandberg, 2000).

In addition, mostdevelopmental psychologists argue that the quality of parenting and of the parent-child relationship are crucial to developingcompetent children. According to Baumrind’s typology (1967, 1978,1991), parents are characterized as using one of four disciplinary styles:authoritarian, authoritative, permissive, and uninvolved, based on thecross-classification of demandingness and responsiveness (Maccoby &Martin, 1983). The authoritarian style, a combination of warmth and con-trol, is believed to create the best environment for child development. In

the present study, warmth measuresresponsiveness by providing informa-tion on the emotional content of the interaction between parent and child.Parental monitoring and control, which includes setting rules and enforc-ing them, measures demandingness.

CULTURAL DIFFERENCES IN PARENTING

Because parentingoccurs within and interactswith thatcontext, social-ization practices in race/ethnic minority families may differ from those of White families (Garcia-Collet al., 1996). Ogbu (1981), Kohn (1977), andothers have proposed that child rearing is oriented toward the develop-ment of instrumental competencies that are adaptive for the settings inwhich children are expected to develop and participate. Minority parentsmay teach skills to assist their children to navigate ethnic and racialbarri-ers. As suggestedby previous research (Bartz& Levine, 1978; Baumrind,1968; Garcia-Coll, 1990), Black parents may exhibit more control and

Hofferth / RACE/ETHNIC DIFFERENCES IN FATHERS 187








less warmththan White parents, perhaps as a reaction to a more dangerousand hostile environment. Alternatively, values and attitudes may varyacross different ethnic groups as they have been shown to vary by socio-economic status (Alwin, 2001; Kohn, 1977). Latino families have consis-tently been shown to endorse warm parenting practices and to emphasizecollective forms of obligationover individualism(Delgado-Gaitan, 1994;Fuller, Holloway, & Liang, 1996). Latino fathers may exert less controlover their children in thecontext of sharedextended-kin responsibility forchild rearing (Baca Zinn, 1994). The greater familism of minority groupsmay increase the responsibility of fathers for children.

Although research has examined interactions and control attempts of fathers, little research has examined the time that minority fathers spend

with children. A recent study shows a puzzling picture of Black fathersspending less time in meals but more time talking, reading, and doinghomework with their children than White fathers (Cooksey & Fondell,1996). However, time expenditures to date are based on single-item re-ports (see Yeung et al., 2001, foran exception) rather than diaries andmaybe inaccurate. The greater familistic orientation in many Hispanic com-munities could increase thevalue to fathers of spending time with children(Baca Zinn, 1994;Vega, 1990),but again,little is known about thevalue tofathers of raising children and the time they spend with them.

Differences in intergenerational experiences . Learning theory positsthat motivation, enjoyment of raising children, and appropriate skills arelearned andsustained through a variety ofmechanisms (Lamb etal., 1985;

Pleck, 1997). For example, growing up with a father who was very in-volved couldbe related to exhibiting moreinvolved fathering behaviorsasan adult. Taking parenting classes could be linked to more involvement.Finally, learning parenting from one’s father could be linked to fatheringbehaviors in later life. To the extent that Blacks, Whites, and Hispanicsdiffer in exposure to father involvement while growing up, their own fa-thering behaviors will differ. The smaller proportion of childhood Black children spendliving witha father could contribute to differential involve-ment of Black men with children.

Gender-role attitudes . Differences among Black, White, and Hispanicparents in gender-role and parenting attitudes may explain differences infathering behaviors. Traditional gender-role theory argues that mothers’andfathers’attitudes andvalues will determinehow much time they spendwith their children. Traditional fathers provide instrumental support, in-cluding financial support and discipline, but less emotional support. Con-









sequently, they arelikely to exhibit less warmthwith children than nontra-ditional fathers. From a role perspective, fathers who hold traditionalvalues with regards to marriage and parenthood are likely to be less in-volved, whereas fathers who endorse gender equity are likely to be moreinvolved with their children. To the extent that fathers endorse the impor-tance of fathers in children’s lives, they shouldbe warmerin their relation-ships with their children and take more responsibility for them.

Based on thestrong financial role of theBlackmother,Blackchildren’sfathers are likely to have less traditional attitudes toward marriage andmotherhood than do White children’s fathers. Black fathers may espousemore gender equitable and more individualistic attitudes than White fa-thers as well. Hispanic children’s fathers are likely to have more tradi-

tional attitudes toward marriage and mothering. However, the research onLatinofamilies suggests that machismo is more a stereotype than an accu-rate depiction of the Hispanic male (Baca Zinn, 1994), and maternal em-ployment has increased equity in Mexican families. Less traditional atti-tudes may be associated with more warmth and responsibility forchildren. Different beliefs about the appropriate roles of men and womenmay also alter the parenting of various groups, but again, little is knownabout gender-role attitudes of minority fathers.

ECONOMIC DIFFERENCES AMONG RACE/ETHNIC GROUPS

Differences in parentingamongdifferent race/ethnic groups mayresultfrom such economic differences as lower labor earnings, lesser employ-

ment, and fewer hours of work. Differences in educational levels betweenBlacks, Hispanics, and Whites may help to explain differences inparenting practices as education affects the human capital of parents andtherefore their wage rates. There are several ways that income and work hours could influence parenting practices. First, fathers who contribute tothe family economically may feel that they have made their contributionand that theydo not need to contribute in other ways. The more hours theywork, the less time they have to participate as well. Fathers who do notwork and who do not contribute economically to the family may attemptto make up forthis lack ofresponsibility in thetraditionalsenseby becom-ing involved in other ways. On the other hand, viewed more as resources,earnings couldbe positively related to behaviors suchas parentalmonitor-ing. According to Goode (1982) and others, middle- and upper-incomeparents have monitoredand controlled theirchildren’sbehaviormore thanlower-income parents because they have control over valuable resources,both money and opportunities.









However, notonly are there differences in socioeconomic status vis-à-vis the outside world, but minority and majority families also differ in therelative statusof spouses within themarital relationship. Divisionof labortheory (Becker,1991) suggests that earning more is associated with doingless household work, including caregiving. Bargaining theory attends tothespouse’s earnings as well (Blau,Ferber, & Winkler, 1998).Earnings of husbands and wives are more equal in Black families than in either Whiteor Hispanic families. As a result, we would expect Black fathers to spendmore time caring for children than White fathers. Black husbands havelowerrates of employment andwork hours than White husbands,whereasBlack wives have higher rates of employment and work hours than Whitewives. According to bargaining theory, in families in which their mothers

work more, children should spend more hours with their fathers.

DIFFERENCES IN RESIDENTIALNEIGHBORHOOD CHARACTERISTICS

Previous research suggests that parenting varies by context. BecauseBlacks, Whites, and Hispanics live in very differentneighborhoods, thesecontexts may explain race/ethnic differences in fathering. Social charac-teristics of neighborhoods have been shown to be at least as important asfamily economic characteristics in containing negative youth behavior(Sampson, Raudenbush, & Earls, 1997). We anticipate that the extent of control that fathers exert over their children’s behavior will be a functionof two aspects of neighborhood—the overall quality of the neighborhood

and its racial/ethniccomposition. Parents may parent theirchildrendiffer-ently in an ethnically homogeneous communitythanonein which they area minority of the population (Duncan, Brooks-Gunn, & Klebanov, 1994).In particular, an ethnically homogeneous community may support ethnictraditions that maintain a sense of communitycohesion and pride, leadingto greater familial involvement by fathers and less control. In contrast, fa-thers areexpected to exertmorecontrol inneighborhoods that areless thanideal for raising children.

OTHER EXPLANATIONS FOR RACE/ ETHNIC DIFFERENCES IN FATHERING

Differences between Blacks and Whites or between Hispanics and

Whites may be due to a number of factors other than economic, neighbor-hood, and cultural differences. It is important to control for these possibleconfounding factors in examining race/ethnic differences in parenting.









Child characteristics . The present analysis adjusts for differences be-tween race/ethnic groupsin ageand genderof thechild. Fathers spend lesstime with children as they grow older (Yeung et al., 2001). Because theirhigher fertility levels imply more young children, Black and Hispanic fa-thers may appear more involved if we do not control for child age.

Family structure and composition differences . Adjustment for differ-ences in family types between race/ethnic groups is important. The mostimportant distinction we make is whether the child’s residential father isthe biological or nonbiological parent. The latter tend to be less involvedwith children (McLanahan & Sandefur, 1994). Among two-parent fami-lies, Black children are more likely than White children to live with a

nonbiological father, whereas Hispanic children are similar to Whites inthe proportion living with a biological father.The age of the father and number of children are also controlled. The

age of the parents may influence their interaction and parenting style. Inaddition, controlling for thenumberof children in thefamily is important;with more children, parents are expected to spend less time with each one(Zajonc & Markus, 1975).

DATA AND METHOD

DATA: THE CHILD DEVELOPMENT SUPPLEMENTTO THE PANEL STUDY OF INCOME

The study sample comes from the Panel Study of Income Dynamics(PSID), a nationally representative sampleof U.S.men, women, children,and the families in which theyreside that has been followed for more than30 years. Until 1997, measures were collected annually from interviewswith 1 adult respondent about all family members, but only limited infor-mation was available on children and parental interaction. During thespring and fallof1997, information on up to 2 randomly selected 0- to 12-year-old children of PSID respondents was collected from the primarycaregivers, from other caregivers, and from the children themselves(Hofferth, Davis-Kean, Davis, & Finkelstein, 1999). The Child Develop-ment Supplement (CDS) completed interviews with 2,394 child house-holdsandabout3,600children.The response rate was90% forthosefami-

lies regularly interviewed in the core PSID and 84% for those contactedthefirst time in 1997 foran immigrantrefresher to thesample, with a com-bined response rate for both groups of 88%. When weights are used, as is









done throughout this article, the results have been found to be representa-tive of U.S. individuals and their families (Fitzgerald, Gottschalk, &Moffitt, 1998a). Weights are also applied to adjust for differential non-response across instruments. Case counts represent actual sample sizes.

The sample used in this article consists of 1,229 children living withtwo parents (stepfather as well as biological father) and for whom infor-mation reported by each parent was available. The sample for analysis of parental time with children consists of 1,172 children because the major-ity, butnot allfamilies, completedtimediaries. Thesample size, includingthosewith complete information on thecontrolvariables, rangesfrom937to 1,077 children, depending on the number of complete responses to theparenting measure and on the ages of the children measured.

MEASUREMENT OF RACE/ETHNICITY

Although ethnic groups exhibit different cultural traditions and prac-tices, groups definedby racial or physicalcharacteristicsmay differ in cul-tural characteristics as well (Taylor, 1994a). For this reason, both Black andHispanic fathers maydifferfrom White fathers in parentingpractices.In this article, race/ethnicity is categorized as non-Hispanic White, non-Hispanic Black, Hispanic, and “other” based on the report of the house-hold head on a set of Census Bureau questions about race and about His-panic origin that were standard up until 1999. The reports were comparedwith the reportof the race/ethnicity of the child; in cases of inconsistency,the head’s report takes precedence. In 1997, Black children represented

about 15% of the U.S. population and Hispanic children another 15%. Asa result of the selection of children living with two parents, in our sample,8% were Black, 11% were Hispanic, 5% were of other races/ethnicities,and 76% were White (see Table 1).

Approximately 62% of Hispanics in the United States are of Mexicandescent. Although a considerable proportion of persons of Mexican de-scent have lived in the United States for generations, recent immigrationincreasedthe Mexican-originpopulation by one third since the1980sand,thus, first- and second-generation groups in the United States (Taylor,1994c). Our sample reflects this reality. Eighty percent of the 98 Latinofamilies in our sample arrived post-1968. Of these, 75% are of Mexicandescent, with most of theremainder from Central America anda few fromSouth America. More than 80%of theimmigrantchildren aresecond gen-eration. Thus,ourHispanic sample describesa primarilyMexicanpopula-tion, in which most families arrived within the past 30 years. The “other”









TABLE 1Background Characteristics of Fathers

in Two-Parent Families by Race/Ethnicity

Total Sample Whites Blacks Hispanics

Variable Mean SD Mean SD Mean SD Mean SD

White 0.76 0.43 1.00 0.00 0.00 0.00 0.00 0.00 0.Black 0.08 0.27 0.00 0.00 1.00 0.00 0.00 0.00 0.Hispanic 0.11 0.32 0.00 0.00 0.00 0.00 1.00 0.00 0.

Other race 0.05 0.22 0.00 0.00 0.00 0.00 0.00 0.00 1.Age of child 5.87 3.72 5.87 3.85 6.38 2.40 5.93 4.68 4.Sex of child (1 = female) 0.52 0.50 0.52 0.52 0.43 0.32 0.52 0.60 0.5Mother’s age 34.77 6.57 34.97 7.01 35.20 4.59 32.92 6.28 35.3Father’s age 36.84 7.12 37.26 7.50 38.10 5.11 33.60 6.56 35.9Nonbiological father 0.07 0.26 0.07 0.26 0.14 0.22 0.07 0.30 0.0Number of children 2.34 1.22 2.19 0.96 2.85 1.35 2.88 1.82 2.6Number of cases 1,229 854 230 98

1 9 3



category includes children of Asian, Native American, and other back-grounds; there are too few to analyze separately.

MEASURES OF PARENTAL INVOLVEMENT AND PARENTING

Time children spend engaged with their fathers . The CDS collected acomplete time diary for 1 weekday and 1 weekend day for each child aged0 to 12 in the family. The data obtained are generally considered superiorto those obtained using standard questions asked in most surveys becauseof internal consistency and reduced social desirability (Hofferth, 1999;Juster & Stafford, 1985). The time diary, which was answered primarilyby the mother or by the mother and the child, asked several questions

about thechild’s flow of activities overa 24-hour periodbeginning atmid-night of the randomly assigned designated day. These questions ask theprimary activity that was going on at that time, when it began and ended,andwhether any other activitywas takingplace.An additional question—”Who was doing the activity with child?” when linked to activity codessuch as “playing”or “beingread to”—provides unbiased details on theex-tent of one-on-one interactions of others with the child. For this analysis,times in which the father (biological father, stepfather, or foster/adoptivefather) was engaged in activities with a child were coded as father en-gaged. Times engaged were summed over all activities for weekdays andweekends for each child. Weekly time was computed by multiplyingweekday time by 5 andweekend daytime by 2. Variables indicate whethertime was spent with that parent and the total number of hours (0 = none ).To test results from previous research, we also examined thetime childrenspend with their fathers in twospecific activities—reading with the fatherand eating meals together.

Responsibility . The 8 responsibility items used here focus on the careof children—bathing children and changing diapers; disciplining chil-dren; choosing children’s activities; buying children’s clothes; drivingchildren to activities; selecting a pediatrician and making appointments;selecting a child care program, preschool, or school; and playing withchildren. 1 Response categories are: 1 ( I do this ), 2 (another household member does this ), 3 ( I share this task ), and 4 ( someone else does thistask ). If the respondent did the task, the response was coded 2, and if therespondent shared it, it was coded 1; otherwise, the task was coded 0.Scores were summed over all items. Overall scale reliability usingCronbach’s alpha was 0.73.









Parental warmth . Parentalwarmthis a 6-itemscaledeveloped by ChildTrends for use in measuring the warmth of the relationship between childandparent. Thequestions, askedofeach parentaboutall children,ask howoften the parent hugged the child, told the child they lovehim or her, spenttime with child, joked or played with child, talked with child, and toldchild they appreciated what he or she did. The response categories rangefrom 1 ( not in the past month ) to 5 ( every day ). A scale was created bysumming the number of behaviors that the parent said they did with thechild in the past month. Scale reliability was 0.77.

Parental monitoring and control . Parental monitoring is measured by aset of9 items asking each parentof children aged 3 and older whether they

have rules setting limits on theirchildren’s activities, their schedules, theirfood, their whereabouts, and their homework and whether they discussthese rules with their children. In contrast to the other scales, this scalemeasures control across all children. 2 The response categories (reversecoded) range from 1 ( never ) t o 5 ( very often ), with 45 the highest possiblescore. The reliability coefficient was 0.73.

MEASUREMENT OF KEY VARIABLES

Economiccharacteristics . Economic variables are measured first by theemployment of mother and father. Employment status is a four-categoryvariable describing the employment of both parents—male breadwinner–female homemaker family, dual-earner family, female breadwinner–malenonemployed family, and two-parent, neither-employed family. WhiteandBlackchildren aremost likelyto live in a dual-earner family;Hispanicchildren (and other races) are most likely to live in a male breadwinner–femalehomemaker family. Black children arealso more likelythan Whitechildren to live in a female breadwinner–unemployed male family. Work hours vary, with Black women working more hours than any of theothers.In contrast, Black men work the fewest hours, resulting in only a 5-hourdifference between Black husbands and wives. Consistent with the smallhourlydifference in time, mothers’and fathers’earnings are themost sim-ilar in Black families, with Black mothers earning $17,000 on average andBlack fathers $20,000. Hispanic mothers earn only $5,400, comparedwith $18,000 for Hispanic fathers. White fathers earn the most, $41,000,with $14,000 for mothers (Table 2).









TABLE 2Means on Fathering, Gender-Role Attitudes, Economic Status, and Neighborhood by Race/Ethnicity

All Races White Black Hispanic

Variable Mean SD Mean SD Mean SD Mean SD

Engaged? 0.94 0.25 0.95 0.23 0.86**** 0.23 0.91 0.33 0.9Hours engaged 14.93 10.93 15.35 11.47 12.76*** 7.49 14.16 12.20 13.6Time eating meals 4.79 4.62 4.74 4.67 4.13* 3.29 5.26 6.05 5.4Time reading 0.19 0.72 0.22 0.81 0.12* 0.45 0.10 0.47 0.0Responsibility for child 6.03 2.49 5.59 2.05 6.27**** 1.70 8.03**** 4.27 7.8Warmth 5.02 1.46 5.10 1.43 4.40**** 1.24 4.84* 1.96 4.9Control 31.87 5.60 31.96 5.53 34.08**** 3.70 29.48**** 7.55 32.5Traditional marriage 0.03 1.00 –0.11 0.98 0.24**** 0.59 0.51**** 1.28 0.8Traditional mothering 0.08 1.02 0.07 1.11 –0.32**** 0.51 0.25* 0.88 0.4Equity 0.03 0.99 0.01 0.98 –0.04 0.63 0.27** 1.41 0.Individualism –0.08 0.98 –0.22 0.93 0.12**** 0.66 0.60**** 1.33 0.2Grandfather involved 3.11 0.89 3.06 0.91 3.14 0.61 3.45**** 1.00 3.0No parenting class 0.73 0.44 0.73 0.47 0.83**** 0.24 0.75 0.52 0.6Learned from mother 0.70 0.46 0.72 0.47 0.68 0.30 0.71 0.54 0.5Learned from father 0.58 0.49 0.63 0.51 0.45**** 0.32 0.38**** 0.58 0.4Positive attitude 26.01 2.90 26.19 3.06 25.73** 1.74 25.32*** 3.24 25.3Mother’s work hours 24.98 18.87 25.04 19.05 31.52**** 12.94 24.66 24.35 14.30Father’s work hours 44.35 12.87 44.45 11.95 36.47**** 11.14 47.21** 16.03 48.02Father’s educat ion (years of school) 13.21 3.04 13.88 2.30 13.06**** 1.34 9.22**** 4.97 11.86Mother’s earnings (0000) 1.30 1.67 1.40 1.78 1.71** 1.38 0.54**** 0.94 0.66Father’s earnings (0000) 3.67 3.52 4.12 3.94 2.01**** 1.14 1.79**** 1.24 3.50Male breadwinner, female homemaker 0.34 0.47 0.31 0.48 0.20*** 0.26 0.51**** 0.60 0.58

1 9 6



Dual earners 0.55 0.50 0.61 0.51 0.48**** 0.32 0.33**** 0.56 0.2Female breadwinner, male unemployed 0.03 0.18 0.02 0.15 0.15**** 0.23 0.04 0.23 0.02No breadwinner 0.03 0.17 0.02 0.13 0.06**** 0.15 0.06*** 0.28 0.1Good neighborhood to raise children 0.31 0.46 0.36 0.50 0.19**** 0.25 0.09**** 0.34 0.20Majority Hispanic 0.13 0.33 0.03 0.19 0.15**** 0.23 0.71**** 0.54 0.1Majority Black 0.13 0.34 0.06 0.25 0.78**** 0.27 0.17 0.44 0.1Majority White 0.80 0.40 0.92 0.28 0.44**** 0.32 0.31**** 0.55 0.6Number of cases 1,172 819 213 94

NOTE: Significance tested relative to Whites.* p < .10. ** p < .05. *** p < .01. **** p < .001.

1 9 7



Neighborhood . In the CDS, the respondent defines his own neighbor-hood; it then asks several questions about this neighborhood. The qualityof the neighborhood is measured by how the respondent would rate theneighborhood the family lives in as a place to raise children. Of responsesranging from excellent to poor , 1 = excellent and the rest 0. Neighborhoodracial composition was obtained fromquestionsasking whether none , lessthan half , about half , more than half , and almost all of the residents wereBlack, Hispanic, and White. From the means, it is clear that each group ishighlylikely to live inclose proximityto othersof thesamerace/ethnicity.

Gender-role attitudes . Attitudes toward gender roles are measured by20 standard items included in national surveys since the early 1960s (Ma-

son, Czajka, & Arber, 1976).3

Four response categories range fromstrongly agree to strongly disagree , with items (re)coded so that a highscore indicatesagreementwith theitem.Mean values were substituted fora small number of missing items. Using principal components analysis,these 20 items were factored into four main factors separately forprimaryand secondarycaregiversand thenrotated usingvarimaxrotationinto fourorthogonalfactors. Theserepresentthe following: (a) traditional marriagevalues: “Most of theimportant decisions in thelife of thefamily shouldbemade by the man of the house”; (b) traditional mothering values: “Pre-schoolchildrenare likelyto sufferif theirmotheris employed”; (c)equity:“If a husband and wife both work full-time, they should share householdtasks equally”; and (d) individualism: “Personal happiness is the primarygoal in marriage.” Factor scores assigned to each individual child’s fa-

ther have a mean of 0 and a standard deviation of 1. Scale reliabilities cal-culated using Cronbach’s alpha were 0.75, 0.77, 0.63, and0.47 forfathers’attitudes toward marriage, traditional mothering, gender equity, andindividualism.

Fathering attitudes and skills . Questionson fathering were drawn fromthe“Beinga Father” scale (Pleck, 1997) andfrom the“Role of theFather”questionnaire (Palkowitz, 1984), tapping the belief that the father role isimportant in child development. Items include: “A father should be asheavily involved in thecare ofhischild as themother,” and“Ingeneral, fa-thers andmothers areequally good at meeting their children’s needs.” Theresponses were scored from 1 ( strongly disagree ) to 4 ( strongly agree ).These 8 items were factor analyzed using principal components analysis,but one factor was sufficient to describe these data. Consequently, afterrecoding so that a high score indicates a positive attitude toward fatheringand substituting mean values for individual missing items, the total score









on the 8 items was then obtained by summing. The reliability coefficientwas 0.70 on this scale.

Intergenerational learning . Thesources ofknowledgeof thefather andmother about parenting were also explored, including the extent to whichthe respondent’s own father was involved in raising him or her andwhether they use their own father as a role model in their own parenting.The parent was asked whether he or she had ever taken a parenting class.Parenting classes include childbirth preparation classes; such classes of-ten include a parenting component either before or shortly after birth andmany also have a support function. Finally, questions asked from whomthefathermost learned toparent—theirmother or fatheror someone else.

CONTROL VARIABLES

Variables controlled in the analysis include the age and gender of thechild, numberof children, father’s age, andwhether thefather is a biologi-cal or nonbiological father to the child.

RESULTS

MEAN DIFFERENCES IN PATERNAL INVOLVEMENT

In two-parent families,childrenaverage14.93hoursperweek engaged

with their fathers (see Table 2). This result is consistent with other re-search (Yeung et al., 2001), which shows that children spend, on average,13 hours per week engaged with their (biological) fathers. Black chil-dren’s fathers are significantly less engaged than are White children’s,12.76 hours compared with 15.35 for White children’s fathers ( p < .01), adifference of 2.59 hours per week. There is no significant difference be-tween Hispanic and White children’s fathers’ engagement. Consistentwith previous research, Black children’s fathers spend marginally lesstime eating meals with them; however, in contrast to that research, theyalso spend less time reading to children. This difference may be due todifferences in the methodology used to collect data on time with chil-dren. Because of the lack of significance of meal time and reading timein both bivariate and multivariate models, we do not pursue these mea-sures further.

Consistent with the literature, Black children’s fathers rate themselveslower on warmth than do White fathers, although Hispanic children’s fa-









thers do not differ significantly from White’s. Again, consistent with ex-pectations, Black children’s fathers are more controlling than White chil-dren’s fathers, whereas Hispanic children’s fathers are less controlling.On responsibility for caring for children, both Black and Hispanic chil-dren’s fathers rank higher than White children’s.

Could these differences be dueto differences in attitudes? As expected,Black children’s fathers hold less traditional mothering attitudes com-pared with White fathers, whereas Hispanic and other-race fathers’moth-eringattitudesare more traditional. Although Hispanic fathers and fathersof other races/ethnicities hold more traditional marriage attitudes thanWhite fathers, as expected, Black fathers are unexpectedly also more tra-ditional. Other research has pointed to discrepancies between the mar-

riageexpectationsof Blackmales andfemales (Taylor, 1994b). Comparedwith Whites, and in contrast to the stereotype, Hispanic fathers are thestronger believers in genderequity. Finally, Black, Hispanic, andother fa-thers all hold more individualistic attitudes than Whites.

Overall, only 27% of children’s fathers said they ever had a parentingclass (including childbirth classes). This is lowest for Black fathers, of whom only 17% ever took such a class. The proportion who said theylearned to parent from their mother is high for all groups—7 out of 10. Incontrast, the proportion of fathers who learned from their own father var-ies by race/ethnicity, with all minority fathers less likely than White fa-thers to learn to parent from their own father. Although 63% of White fa-thers learned from their father, 45% of Black, 38% of Hispanic fathers,and 44%of other fathers didso. Fathering attitudesaregenerallypositive,

butBlack, Hispanic, and other-race fathers consistently hold less positiveattitudes than Whites.

This analysis shows considerable race/ethnic differences in father in-volvement with their children. Some, such as the greater control and lesswarmth expressed by Black fathers, were expected. Others, such as thegreater responsibility shown by Black and Hispanic children’s fathers,were not. Some of these differences in parenting maybe explained by dif-ferences in the economic situations and living arrangements of these dif-ferent groups. Others could be explained by attitudinal differences. Somedifferences in attitudes, suchas theespousalof moretraditionalmotheringamong Hispanics and less traditional mothering attitudes among Blacks,were expected. Unexpected were traditional marriage attitudes amongBlacks, high levels ofequityamongHispanic fathers andhigh levels of in-dividualism among all three minority groups.









HIERARCHICAL REGRESSIONS

To examine the extent to which race/ethnic differences in fathering re-main after controlling for family and neighborhood factors and gender-role or fathering attitudes, we use a set of hierarchical models (seeTable 3). In thefirststep,measures of parentingareregressed on race vari-ables only. In the second step, our control variables—age and gender of the child, father’s age, nonbiological father, and number of children—areadded. The third step adds economic characteristics of the family—eachparent’s earnings, education of father, employment status, and hours of mother and father. Measures of neighborhood are added in step 4, andmeasures of fathering attitudes are added in step 5. Finally, fathering atti-tudes are dropped and gender-role attitudes substituted in step 6, becauseone of the gender-role attitudes (equity) is strongly associated with posi-tive fathering attitudes. We test the significance of the change in the effectof being in a particular racial or ethnic group on parenting due to the addi-tion of each set of variables (Clogg, Petkova, & Haritou, 1995) (see Ta-ble 4).

CHILDREN’S TIME ENGAGED WITH FATHERS

Table 3, panel A, column 1, shows that the coefficient for children’stime engaged with their father is negative, reflecting a 2.4-hour lowerweekly father engagement time forBlack compared with White children,but this effect is only marginally statistically significant ( p < .10). Eventhough the Black-White difference in engagement is relatively small, weexamine what factors explain this important component of parenting.

With controls for children’s age and gender, nonbiological father, fa-ther’s age, andfamily size (see Table 3, column2), theeffect of Black racedeclines significantly and is no longer associated with children’s timewith their father. Black families in our sample tend to be older, to haveslightly older children, and to have a larger number of children. Fathersspend less time with children in these circumstances. Black children aretwice as likely as White children to be living with a nonbiological father,and children receive less paternal attention in such families. Controllingfor factors that reduce time with children lowers the negative associationof Black race/ethnicity and time with the father. We do not examine fur-ther changes in the coefficient on “Black” because it is not significantly

associated with paternal engagement in later models.


(text continues on p. 207)








TABLE 3OLS Regressions of Parenting Behavior on Race/Ethnicity and Controls

Panel A: Engagement Panel B: Warmth

Variable 1 2 3 4 5 6 1 2 3 4

White (omitted)Black –2.45* –1.31 –2.02 –2.00 –2.14 –1.88 –0.86**** –0.71**** –0.65**** –0.87**** –0.9Hispanic –0.56 –0.50 0.14 1.12 0.68 1.24 0.00 0.01 0.18 0.06 –0Other race –2.54* –2.85* –1.87 –1.87 –1.72 –1.52 –0.10 –0.19 –0.07 –0.08 –0.Age of chi ld –0.59**** –0.58**** –0.57**** –0.52**** –0.55**** –0.14**** –0.14**** –0.14**** –0.1Gender of child 0.15 0.21 0.26 0.15 0.20 –0.05 –0.05 –0.05 –0Age of father –0.11* –0.11** –0.12** –0.11** –0.11* 0.00 –0.01 –0.01 0

Nonbiological father –5.59**** –5.69**** –5.77**** –5.36**** –5.34**** –0.75**** –0.68**** –0.69**** –0.5Number of children –0.76**** –0.57* –0.55* –0.59** –0.64** –0.05 –0.02 –0.02 –0.Father’s education 0.19 0.22 0.11* 0.12 0.01 0.01 –0Mother’s hours 0.04* 0.05* 0.05** 0.04 0.00 0.00 0Father’s hours –0.06** –0.07** –0.07 –0.06** –0.01 –0.01 –0Male breadwinner (omitted)Two-earner family –0.40 –0.49 –0.45 –0.37 0.02 0.04 0Female breadwinner –3.83* –4.18** –4.06* –4.00* –0.18 –0.15 –0No breadwinner –1.10 –1.11 –0.55 –0.30 –0.38 –0.36 –0Mother’s earnings 0.17 0.21 0.16 0.20 0.06* 0.06* 0Father’s earnings –0.25** –0.25** –0.22** –0.25** 0.03** 0.03** 0Good neighborhood –1.59** –1.53** –1.71** 0.14 0Majority Hispanic –1.78 –1.80 –1.91 0.21 Majority Black –0.27 –0.12 –0.24 0.34* Grandfather involved 0.68*

2 0 2



No parenting class –0.48 –Learned from mother 0.42 Learned from father –0.22 Positive attitude 0.36*** Traditional marriage –0.27 Traditional mothering –0.15 Equity 0.69** Individualism –0.51 Intercept 15.50 25.14 24.63 25.18 15.16 26.20 5.10 6.09 6.22 6.10 3

R2 .01 .09 .11 .12 .13 .13 .02 .17 .18 .18 Number of cases 1,027 1,027 1,027 1,027 1,027 1,027 1,077 1,077 1,077 1,077 1,0

Panel C: Control Panel D: Responsibility

1 2 3 4 5 6 1 2 3 4

White (omitted)Black 2.39**** 1.97*** 2.21*** 1.87** 1.77** 2.24** 0.65** 0.60** 0.45 0.07 –0.Hispanic –2.39**** –3.06**** –2.21*** –3.09**** –3.56**** –2.53*** 2.68**** 2.43**** 2.34**** 1.72**** 1.4Other race 0.92 0.52 1.13 1.02 1.16 1.53* 2.12**** 2.00**** 1.66**** 1.55**** 1.Age of child –0.02 –0.01 –0.01 0.04 0.01 –0.01 0.00 0.00 0Gender of child 0.42 0.44 0.43 0.29 0.28 –0.21 –0.17 –0.15 –0Age of father –0.03 –0.07** –0.07** –0.06* –0.05* –0.04*** –0.04*** –0.04*** –0.Nonbiological father –0.23 0.24 0.21 0.59 0.92 –0.67** –0.95*** –0.98*** –0.Number of children 0.56**** 0.63**** 0.65**** 0.59**** 0.44*** 0.11* 0.12* 0.14** 0.Father’s education 0.11 0.11 0.01 –0.04 –0.01 0.00 –0Mother’s hours –0.02 –0.02 –0.02 –0.03** –0.01 –0.01 –0Father’s hours –0.02 –0.02 –0.01 –0.01 –0.01** –0.01** –02

0 3



TABLE 3 (continued)

Panel C: Control Panel D: Responsibility

1 2 3 4 5 6 1 2 3 4

Male breadwinner (omitted)Two-earner family –0.52 –0.44 –0.35 –0.04 –0.41** –0.36* –0Female breadwinner 0.48 0.62 0.62 1.23 –1.55**** –1.53**** –1No breadwinner –4.24**** –4.16**** –3.44*** –2.81*** 1.84**** 1.89**** 2.Mother’s earnings 0.09 0.10 0.03 0.12 0.25**** 0.27**** 0Father’s earnings 0.10* 0.10* 0.12** 0.10* –0.05** –0.05** –0Good neighborhood 0.40 0.50 0.13 –0.25* –0.23 –0Majority Hispanic 1.35 1.34* 0.92 0.85*** 0.87*** 0

Majority Black 0.49 0.63 0.77 0.45 0.49* Grandfather involved 0.53** 0.37****No parenting class –0.73* –Learned from mother 0.80* Learned from father –1.09** –Positive attitude 0.41**** Traditional marriage –0.13 Traditional mothering 0.08 Equity 1.16**** Individualism –1.03**** Intercept 32.04 31.92 32.60 32.09 20.92 33.46 5.54 6.83 7.82 7.49 3

R2 .03 .05 .08 .08 .14 .15 .13 .14 .20 .21 Number of cases 937 937 937 937 937 937 1,077 1,077 1,077 1,077 1,0

NOTE: OLS = ordinary least squares.* p < .10. ** p < .05. *** p < .01. **** p < .001.

2 0 4



TABLE 4Changes in Race/Ethnic Coefficients Across Models From Table 3 (With Significance Tests)

Background Earnings Neighborhood Fathering 1-2 2-3 3-4 4-5

EngagementBlack

d –1.137 0.711 –0.028 0.141 S (d )a 0.148 0.281 0.997 0.020 t 7.68 2.53 ns 7.04

Hispanic b

d — — — — S (d ) — — — —t — — — —

WarmthBlack

d –0.154 –0.600 0.221 0.041 S (d ) 0.036 0.048 0.122 0.033 t 4.28 1.23 1.81 1.24

Hispanic b

d — — — — S (d ) — — — —t — — — — 2

0 5



TABLE 4 (continued)

Background Earnings Neighborhood Fathering 1-2 2-3 3-4 4-5

ControlBlack

d –0.426 0.242 –0.336 –0.107 S (d ) 0.141 0.165 0.519 0.130 t 3.02 1.47 ns ns

Hispanicd 0.668 –0.849 0.880 0.476 S (d ) 0.205 0.338 0.578 0.062

t 3.25 2.51 1.52 7.67 ResponsibilityBlack

d –0.056 –0.151 –0.377 –0.162 S (d ) 0.043 0.135 0.202 0.051 t 1.30 1.12 1.86 3.18

Hispanicd –0.252 –0.089 –0.613 –0.314 S (d ) 0.071 0.115 0.199 0.022 t 3.55 ns 3.11 14.27

a. S (d ) = square root [(se 2)2 – (se 1)2 (Root MSE 2 /Root MSE 1)].b. Not calculated because of nonsignificance of the coefficient.

2 0 6



Across all children, the more hours the father works and the more hemakes, and the better the neighborhood (perhaps suggesting wealth), thelower his time with his children. Similarly, the more hours the motherworks, the more hours the father spends. These findings are consistentwith division of labor and bargaining perspectives. Positive attitudes to-wardfatheringand grandfather involvementare associated with increasedpaternal time with children as are more equitable sex-role attitudes.

PARENTAL WARMTH

Consistent with the unadjusted means, Black children’s fathers reportsignificantly fewer warm behaviors with them than White children’s fa-

thers (see Table 3, panel B, column 1). This finding is consistent with pre-vious research and suggestive of more traditional fathering behavior inBlack than White families.This resultholds adjustingforallvariables, al-though it declines or increases slightly at each step. About one fourth of the variance is explained in the full model.

Theeffect of being Black declines significantly with control for familystructureandotherbackground factors (see Table 3, column2).This is be-cause Black children’s fathers aremore likely to be stepfathersor cohabit-ing, and such fathers tend to be less warm with their stepchildren. Con-trolling for these differences in family structure reduces the differential inwarmth between White and Black fathers.

Controls for economicstatus reduce the Black-White warmth differen-tial, but the reduction is not statistically significant. Black children’s fa-

thers’and mothers’wages are lower, and lower wages are associated withreporting fewer warm behaviors. Why higher wages are linked to warmbehaviors may reflect differential parenting values by class, as suggestedby Kohn (1977), with middle-class fathers more likely to provide a sup-portive, warm environment than lower-class fathers. We shall see later onthat they also exertmore control, thus suggesting the dominance of an au-thoritative modelof parenting (warmth with control) among middle-classfathers.

Controlling forneighborhood increasesthe Black-White differentialinwarmthsubstantially, and the increase is significant at p < .10. Black chil-dren tend to live in Black neighborhoods, in which locations, thedata sug-gest, fathers tend to be warmer. Notadjusting for living in these neighbor-hoods masks the fact that Black fathers tend to show less warmth thanWhite fathers.

Adding intergenerational fathering and fatheringattitudes (seeTable3,column 5) increases the effect of being Black on warmth slightly, but not









significantly, even though there are significant relationships between fa-thering attitudes and warm behaviors. Black children’s fathers have a lesspositive attitude toward fathering than White children’s, and having a lesspositive fathering attitude is also associated with fewer warm behaviors.Although having a grandfather involved in raising one’s own father is as-sociated with warm behaviors, Black children’s fathers do notdiffer fromthose of Whites on grandfather involvement.

Addinggender-role attitudesis associated with no changein theBlack-White difference in warmth (see Table 3, column 6 vs. 4), which remainssignificant. Black fathers espouse more individualistic attitudes, whichare associated with less warmth. Believing in gender equity is associatedwith greater warmth, but on this attitude,Blacks andWhites do notdiffer.

PARENTAL MONITORING AND CONTROL

Race/ethnic differences in control areshown to be highlysignificant inTable 3, with 2.4pointsgreater control by Black fathers and2.4points lesscontrol by Hispanic fathers compared with White fathers (panel C, col-umn 1). The Black-White difference in control/monitoring declines sig-nificantly when we adjust for background characteristics of the child andfamily. Blacks have more 6- to 8-year-olds than Whites, and parents con-trol such children more. Blacks also have more children, and having morechildren increases the amount of control exerted. The Hispanic-Whitedif-ference, in contrast, increases significantly. Hispanic children are morelikely to live with a biological father, which is associated with more con-

trol. Adjusting for differences in living arrangements increases the nega-tive relationship between Hispanic origin and monitoring.

Controlling for economic status results in increasing the Black-Whitedifference in monitoring by a small and nonsignificant amount while de-creasing the Hispanic-White monitoring difference by a large and signifi-cant amount. Black families are more likely to have no breadwinner, andthat is associated with less monitoring. Black fathers also have lowerwages, which are associated with less monitoring. Adjusting for thesevariables increasesthe Black-White difference over what itwasin thepre-vious step. Economic differences also explain some of the Hispanic-Whitemonitoringdifference. Hispanic men earn lessmoney, whichis associatedwith lower monitoring. When this is taken into account, the monitoringdifference is reduced.

Our data show that Black and Hispanic families report living in lessdesirable areas for raising children than White families. Under these cir-cumstances, we would expect parents to monitor and control children’s









behavior more, as Black fathers do. However, controlling for neighbor-hood reduces only slightly and insignificantly the Black-White differencein monitoring because the neighborhood rating is not linked to monitor-ing. The Hispanic-Whitemonitoring difference increaseswhen neighbor-hood is controlled, although the increase is not statistically significant.Hispanics live in majority Hispanic neighborhoods; monitoring is morecommonin such neighborhoods. Hispanicsmonitor less giventhe types of neighborhoods they live in.

Adding attitudes does not significantly change the effect of Black raceon monitoring althoughit significantlyincreasesthe effectof Hispanic or-igin. Hispanic fathers are less likely to have learned parenting from theirfather than White fathers. Not having learned from their father is associ-

ated with more monitoring. In addition,Hispanic fathersaremore likelytohave had an involved father. Having had an involved father is associatedwith more monitoring. Adjusting for these two factors, Hispanic fathersmonitor even less than White fathers.

Finally, adding gender-role attitudes significantly increases the strengthof Black race/ethnicity andsignificantly reduces theeffect ofHispanic or-igin on monitoring. Hispanic and Black children’s fathers are more likelyto express individualistic attitudesthan White children’s, and greater indi-vidualism is associated with lesser monitoring and control. When thesedifferences are adjusted, the effect of being Hispanic on monitoring andcontrol is reduced and the effect of being Black is increased. Hispanicsalso express more gender-equitable attitudes, and gender-equitable atti-tudes are associated with greater monitoring.

PARENTAL RESPONSIBILITY

Consistent with mean differences, both Hispanic and Black children’sfathers report themselves significantly more likely than White children’sfathers to take responsibility for their care (see Table 3, panel D, col-umn 1). However, although the coefficient for Hispanic origin remainssignificant over all models, the coefficient for Black declines to insignifi-cance as variables are added. The sizes of the coefficients in column 1 areconsistent with the sizes of effects in Table 1; the effect of Hispanic originis considerably larger than the effect of Black race.

In Table 3, column2, controls added forbackground characteristics re-duce both effects, significantly for Hispanics but not for Blacks. Black children aremuch more likely to live with a nonbiological father, which isassociated withfathers taking lessresponsibility. Hispanicand Blackchil-dren tend to have more siblings, which is associated with greater father re-









sponsibility. Controlling for these variables reduces the effect of Black race and Hispanic ethnicity.

Controlling for economic status significantly ( p < .10) reduces theBlack coefficient, and it is no longer significant. This is because Black families are characterized by lower fathers’ earnings and work hours,higher mothers’ earnings, and being more likely to have no breadwinner,which areallassociated with fathers taking greater responsibility for chil-dren. The Hispanic coefficient is not significantly reduced but retains asignificant association with paternal responsibility. Mothers’ wages arelower, leading to less responsibility taken by fathers, but fathers’ lowwages are associated with fathers taking more responsibility. These re-sults are consistent with a division of labor or bargaining explanation for

differential involvement.Controlling for neighborhood (see Table 3, column 4) significantly re-duces the effect of Hispanic origin. This is because Hispanic children arehighly likely to live in Hispanic neighborhoods, and fathers in majorityHispanic areas aremore likelyto take responsibility forchildrenthan fam-iliesin otherneighborhoods.Adjustingforneighborhood reduces the His-panic effect.

Controlling for attitudes further reduces the effect of Hispanic ethnic-ity. Although they are less likely to have learned to parent from their fa-ther, Hispanic fathers are more likely to have had a father involved withthem, and that is associated with greater responsibility forchildren. In ad-dition, having a positive attitude toward parenting is associated withgreater responsibility. Controlling for these Hispanic-White differences

reduces the effect of Hispanic ethnicity on responsibility, although it isstill statistically significant.

Finally, adjusting for gender roles significantly reduces the effect of Hispanic origin on parental responsibility over column 4. Hispanics ex-hibit more gender-equitable values, which areassociated with fathers tak-ing more responsibility for their children.

DISCUSSION AND CONCLUSION

This article has focused on Black-White and Hispanic-White differ-ences in fathering behavior in two-parent families. Several differenceswere identified. Black children’s fathers exhibit less engagement and lesswarmth but exert more control and take more responsibility than Whitechildren’s fathers, thus exhibiting a more authoritarian model of parenting.Hispanic children’s fathers exert lesscontrol but also takemore









responsibility than White children’s fathers, although they do not differfrom White fathers in warmth. Theirparentingstyle appears more permis-sive. The results suggest that parenting attitudes are a major factor in racedifferences in control and responsibility, whereas economic factorsstrongly affect engagement and responsibility. Differences in warmth aredue to background differences, neighborhood, and family structure.

Demographic and economic factors—particularly family size, rela-tionship to father, family structure, and employment—explain mostof thefact that Black fathers appear to spend less time engaged with their chil-dren than do White fathers. In particular, Black fathers have larger fami-lies, are less likely to be the biological father, and are more likely to be un-employed living with a female breadwinner, all of which, by themselves,

would result in less time spent with children. When these factors are con-trolled, Black-White differences in engagement are no longer statisticallysignificant.

Differences in warmth between Black and White fathers are also par-tially explained by the fact that fewer Black fathers are biological fathers,and nonbiological fathers are less warm. Neighborhood characteristics,however, appear to mask larger Black-White differences in warmth. Liv-ing in a good neighborhoodis associated with less time engaged with chil-dren. Black familiestend to live in Black neighborhoods and, accordingtothese data, fathers tend to be warmer in such neighborhoods. This fact ap-pears to mask their lower overall levels of warmth. Perhaps living in a ra-cially homogeneous community is less stressful than living in a heteroge-neous one (controlling for desirability), and this contributes to greater

warmth in the former. Alternatively, Black fathers with interpersonalstyles that are less warm may select heterogeneous neighborhoods. Moreresearch is needed on the interaction between ethnicityand neighborhoodin influencing parenting styles.

Differences in degree of control and monitoring between Blacks andWhites are partially explained by family size. Black families tend to belarger, and fathers in large families exert more control than those insmaller families. This explainswhy Black children’s fathers arehigheroncontrol than White children’s fathers. Hispanic children’s fathers monitorand control their children less than White children’s. This difference doesnotdisappearbut declines as economic conditions areadded. Hispanic fa-thers’incomes arelower, andlowerincomes areassociated with less mon-itoring, thus explaining some of their lessermonitoring andcontrol. The ef-fects of fathers’income areconsistentwith a resource ratherthan division-of-labor model.









Finally, Black, Hispanic, and other minority fathers all show more re-sponsibility for their children than do White fathers. Black-White differ-ences in responsibility are partiallyexplained by neighborhood factors. Inbad neighborhoods, which is where Black families are more likely to re-side, fathers take more responsibility for their children. Differences inneighborhood composition and in culture between Hispanics and Whitesalso explain much of the variation in responsibility between the twogroupsof fathers. Fathers take much more responsibility fortheirchildrenin Hispanic neighborhoods, and (in analyses not shown) this effect isgreater for Hispanic fathers. Hispanic fathers also live in less desirableneighborhoods, and living in such neighborhoods is associated with moreresponsibility.

Although traditional economic and demographic factors are linked tothe attention fathers pay to their children, this study has found evidencethat attitudes, values, and motivational factors—what we call “culturalfactors”—are closely tied to the father’s engagement with children. Fa-thers with more positiveattitudes toward parenting, whose own fatherwasinvolved with him as a child, who have more gender-equitable responses,and who had taken a parenting class spend more time with the child, exer-cise more control, and take more responsibility for their children. Fa-thering attitudes and behavior explain significant variation in control forHispanics and in responsibility for both Blacks and Hispanics. Gender-role attitudes explain some of the difference in control between Black ver-susWhite andHispanicversus White fathers andsome of thedifferenceinresponsibility for Hispanic fathers compared with White fathers. For ex-

ample, although they have more gender-equitable attitudes, which are as-sociated with more control, Hispanic fathers are also highly individualis-tic in attitudes toward marriage and have less positive fathering attitudes,both of which are associated with less control. Adjusting for these factorsexplains some of the lesser monitoring of Hispanic fathers. The fact thatHispanic fathers are more individualistic is inconsistent with the expectedgreater familism of Hispanic fathers. However, if responsibility is sharedby a variety of extended kin, individual control may not be necessary.

Differences in responsibility between Hispanic children’s fathers (andthose of other races) and White children’s remain strong even with con-trols forother factors. Hispanic fathers report their own fathers were moreinvolved in rearing them, and that increases their involvement with theirownchildren. In addition, more genderequitableattitudes, which areheldby Hispanic men,are associatedwith moreresponsibility towardchildren.The potential strength of minority families regarding fathering should beexamined more thoroughly.









Although it examined only a small setof possible parenting behaviors,this article showed that, of the behaviors we studied, minority fathering intwo-parent families varies in only a matterof degreefrom that of majorityfamilies.Black fathers tend tobe more authoritarian, with less warmthandgreater control, whereas Hispanic fathers are more permissive, with lesscontrol than White fathers. The stereotype of a macho, noninvolved mi-norityfather is not typically true; Black and Hispanic,primarilyMexican,males are as involved as are majority fathers, and they share greater re-sponsibility for child rearing with their partners. Unfortunately, our studydoes not describe the fathering experiences of all Black, White, and His-panic children. Although it included a variety of relationships, our studyof children in coresidential two-parent families represents only about half

of Black, 75% of Hispanic, and 80% of White children in the UnitedStates. Nonresidential fathers are typically less involved. Research thatexamines the involvement of non-coresidential fathers in the livesof chil-dren is an important next step for fully evaluating the strengths and weak-nesses of majority and minority families for the United States of the 21stcentury.

NOTES

1. These measures are self-reports; theygive the respondent’sperception. Althoughsub- jective perception is not the same as what would be measured were time diaries of parentalactivities collected, this scale has been used in a variety of previous studies, such as the Na-tional Survey of Families andHouseholds.The crucialfactor is that there is no reasonto ex-pect minority-majority differences in reporting on this scale.

2. It is unfortunate that this measure is not child-specific because control may vary fromchild to child. However, lacking any alternative, we take this as a measure of the overall de-gree of control that a father exerts over his children.

3. A complete list of the items in the attitude scales used in this article is available fromthe author.

REFERENCES

Alwin, D. F. (2001). Parental values, beliefs, and behavior: A review and promulga for re-search into the new century. In T. Owens & S. Hofferth (Eds.), Children at the millen-nium: Where did we come from, where are we going? New York: Elsevier Science.

Baca Zinn,M. (1994).Adaptationand continuity in Mexican-originfamilies. In R. L. Taylor

(Ed.), Minority families in the United States: A multicultural perspective (pp. 64-81).Englewood Cliffs, NJ: Prentice Hall.









Bandura, A. (1969). Principles of behavior modification . New York: Holt, Rinehart &Winston.

Bartz, K., & Levine, B. (1978). Child rearing by Black parents: A description and compari-son to Anglo and Chicano parents. Journal of Marriage and the Family , 40 , 709-720.

Baumrind, D. (1968). An exploratory study of socialization effects on Black children:Black-White comparisons. Child Development , 43 , 261-267.

Baumrind, D. (1978). Parental disciplinary patterns and social competence in children.Youth and Society , 9 , 239-276.

Baumrind, D. (1991). The influence of parenting style on adolescent competence and sub-stance abuse. Journal of Early Adolescence , 11 , 56-94.

Becker, G. S. (1991). A treatise on the family (Rev. ed.). Cambridge, MA: Harvard Univer-sity Press.

Blau, F., Ferber, M.,& Winkler, A. (1998). Theeconomics of women,men, andwork . UpperSaddle River, NJ: Prentice Hall.

Bronfenbrenner, U. (1979). The ecology of human development . Cambridge, MA: HarvardUniversity Press.

Cherlin, A., Furstenberg, F. F., Jr., Chase-Lansdale, P. L., Kiernan, K.E.,Robins, P. K., Mor -rison, D. R., et al . (1991). Longitudinal studies of the effects of divorce on children inGreat Britain and the United States. Science , 252 , 1386-1389.

Clogg, C. C., Petkova, E., & Haritou, A. (1995). Statistical methods for comparing regres-sion coefficients between models. American Journal of Sociology , 100 (5), 1261-1293.

Cooksey, E. C., & Fondell, M. (1996). Spending time with his kids: Effects of family struc-ture on fathers’ and children’s lives. Journal of Marriage and the Family , 58 , 693-707.

Delgado-Gaitan, C. (1994). Socializingyoung children in Mexican-Americanfamilies.In P.Greenfield& R. Cocking (Eds.), Cross-cultural roots of minority childdevelopment: Anintergenerational perspective (pp. 55-86). Hillsdale, NJ: Lawrence Erlbaum.

Duncan, G. J., Brooks-Gunn, J., & Klebanov, P. K. (1994). Economic deprivation and earlychildhood development. Child Development , 65 (2), 296-318.

Fitzgerald, J., Gottschalk, P., & Moffitt, R. (1998a). An analysis of sample att rition in paneldata: The Michigan Panel Study of Income Dynamics. Journal of Human Resources ,

33(2), 251-299.Fuller, B., Holloway, S., & Liang, X. (1996). Family selection of child-care centers: The in-

fluence of household support, ethnicity, and parental practices. Child Development , 67 ,3320-3337.

Garcia-Coll, C. (1990). Developmental outcome of minority infants: A process-orientedlook into our beginnings. Child Development , 61 , 270-289.

Garcia-Coll, C., Lamberty, G., Jenkins, R., McAdoo, H. P., Crnic, K., Wasik, B. H., et al.(1996). An integrative model for the study of developmental competencies in minoritychildren. Child Development , 67 , 1891-1914.

Goode, W. J. (1982). The family . New York: Free Press.Hernandez, D. J. (1993). America’s children: Resources from family, government and the

economy . New York: Russell Sage.Hofferth,S. L. (1999, May27-28). Family readingto youngchildren: Social desirability and

cultural biases in reporting . Paper presented at Workshop on Measurement of and Re-search on Time Use, Committee on National Statistics. Washington, DC: National Re-

search Council.Hofferth, S. L., Davis-Kean, P., Davis, J., & Finkelstein, J. (1999). 1997 user guide: TheChild Development Supplement to the Panel Study of Income Dynamics . Ann Arbor:University of Michigan, Institute for Social Research.









Hofferth, S. L., Pleck, J., Stueve, J., Bianchi, S., & Sayer, L. (2002). The demography of fa-thers: What fathers do. In K. Tamis-Lemonda & N. Cabrera (Eds.), Handbook of father involvement . Mahwah, NJ: Lawrence Erlbaum.

Jencks, C., & Phillips, M. (1998). The Black-White test score gap . Washington, DC:Brookings Institution.

Juster, F., & Stafford, F. P. (1985). Time, goods, and well-being . Ann Arbor: University of Michigan, Institute for Social Research.

Kohn, M. L. (1977). Class and conformity . Chicago: University of Chicago Press.Lamb,M. E. (1997). Nonparental childcare:Context, quality, correlates,and consequences.

In I. Sigel & K. Renninger (Eds.), Handbook of child psychology (5th ed., pp. 73-134).New York: John Wiley.

Lamb, M. E., Pleck, J. H., Charnov, E. L., & Levine, J. A. (1985). Paternal behavior in hu-mans. American Zoologist , 25 , 883-894.

Maccoby, E.,& Martin, J. (1983). Socialization in thecontext of thefamily: Parent-childin-teraction.In E. M.Hetherington (Ed.), Handbookof child psychology (Vol. 4, pp.1-101).New York: John Wiley.

Mason,K. O.,Czajka, J.L., & Arber, S.(1976).Changes in U.S. women’s sex-role attitudes,1964-1974. American Sociological Review , 41 , 573-596.

McAdoo, J. L. (1988). The roles of Black fathers in the socialization of Black children . InH. P. McAdoo (Ed.), Black families (pp. 257-269). Newbury Park, CA: Sage.

McLanahan, S., & Sandefur, G. (1994). Growing up with a single parent: What hurts, what helps . Cambridge, MA: Harvard University.

Mott, F. L. (1990). When is a father really gone? Paternal child contact in father-absenthomes. Demography , 27 , 499-517.

Ogbu, J. (1981).Originsof humancompetence: A cultural-ecological perspective. ChildDe-velopment , 52 , 413-429.

Palkowitz, R. (1984). Parental attitudes and fathers’ interactions with their 5-month-old in-fants. Developmental Psychology , 20 , 1054-1060.

Phillips, M., Brooks-Gunn, J., Duncan, G., Klebanov, P., & Crane, J. (1998). Family back-ground, parentingpractices, andthe Black-White testscoregap. In C. Jencks& M. Phil-

lips (Eds.), The Black-White test score gap (pp. 103-145). Washington, DC: BrookingsInstitution.

Pleck,J. H. (1997).Paternal involvement: Levels, sources,and consequences. InM. E. Lamb(Ed.), Therole ofthe fatherin childdevelopment (pp.66-103).NewYork: JohnWiley.

Sampson, R. J., Raudenbush, S. W., & Earls, F. (1997, August 15). Neighborhoods and vio-lent crime: A multilevel study of collective efficacy. Science , 277 , 918-924.

Sandberg, J. F. (2000, March). Modeling multi-dimensionality of involvement in two-parent families: A l atent class analysis of paternal responsibility . Paper presented at AnnualMeeting of the Population Association of America, Los Angeles, CA.

Seltzer, J.,& Bianchi, S. (1988).Children’s contact withabsent parents. Journalof Marriageand the Family , 50 , 663-677.

Taylor, R. L. (1994a). Minority families in America: An introduction. In R. L. Taylor (Ed.), Minority famili es in the United States: A multicultural perspect ive (pp. 1-16).Englewood Cliffs, NJ: Prentice Hall.

Taylor, R. L. (1994b). Black American families. In R. L. Taylor (Ed.), Minority families in

the United States: A multicultural perspective (pp. 19-46). Englewood Cliffs, NJ:Prentice Hall.









Taylor, R. L. (1994c). Minority families and socia l change. In R. L. Taylor (Ed.), Minority families in the United States: A multicultural perspective (pp. 204-248). EnglewoodCliffs, NJ: Prentice Hall.

Vega,W.A. (1990).Hispanic families. Journalof Marriage and the Family , 52 , 1015-1024.Yeung, W. J.,Sandberg, J.,Davis-Kean, P. E.,& Hofferth,S. L. (2001). Children’s timewith

fathers in intact families. Journal of Marriage and the Family , 63 (1), 136-154.Zajonc,R. B.,& Markus, G.B. (1975). Birth orderand intellectual development. Psycholog-

ical Review , 82 , 74-88.


journal of family issues 2003 hofferth 185 216

Documents