THE EFFECTS OF FAMILY CHARACTERISTICS AND TIME USE ON TEENAGE GIRLS’ AND BOYS ’ HOUSEHOLD LABOR

Center for Research on Child Wellbeing Working Paper #98-01

Constance T. Gager, Ph.D. Teresa M. Cooney, Ph.D. Kathleen Thiede Call, Ph.D.

THE EFFECTS OF FAMILY CHARACTERISTICS AND TIME USE ON TEENAGE GIRLS’ AND BOYS’ HOUSEHOLD LABOR 1

Constance T. Gager, Ph.D. Office of Population Research 21 Prospect Avenue Princeton University Princeton, NJ 08544 Teresa M. Cooney, Ph.D. Human Development and Family Studies 14B Gwynn Hall University of Missouri Columbia, MO 65211-7700 Kathleen Thiede Call, Ph.D. Institute for Health Services Research University of Minnesota Minneapolis, MN 55455

1

An earlier version of this paper was presented at the annual meetings of the Population Association of America, March, 1997, in Washington, D.C. Support for the Youth Development Study, on which this paper is based, was provided by Grant MH42843 from the National Institute of Mental Health awarded to Jeylan T. Mortimer and Michael D. Finch. We greatly appreciate their willingness to make this data set available to us. We also thank Margaret Mooney Marini for her valuable comments on earlier drafts of this paper. Please direct all correspondence to Constance T. Gager, Office of Population Research, Princeton University, Princeton, NJ 08544

THE EFFECTS OF FAMILY CHARACTERISTICS AND TIME USE ON TEENAGE GIRLS’ AND BOYS’ HOUSEHOLD LABOR

Abstract Studies suggest that children’s housework contributions are minimal. However, much of this research focuses on young or early adolescent children and utilizes data regarding adult tasks, while ignoring chores children more often perform. We address these gaps by analyzing longitudinal time-use data collected from teens on the types of household chores that children are most likely to perform. Specifically, we examine gender inequity in teens’ contributions to household labor and how it changes over high school. Additionally, we explore how teens’ household contributions vary based on family structure characteristics such as maternal employment and parental marital status, and as a function of teens’ involvement in such nonfamily contexts as school and paid work. We find that girls devote more time to household tasks than boys with this gender gap increasing across high school. Teens’ efforts are greater in larger families, especially for girls, and those headed by single or remarried parents. Lastly, high school males spend more time on extra-curricular and leisure activities than girls who work longer hours in both unpaid and paid labor. Thus, we conclude that the leisure gap and the “second shift” documented between wives and husbands (Hochschild, 1989), may originate in the family of orientation.

INTRODUCTION Researchers investigating children’s contributions to household work in the past 10-15 years generally seem to concur with Shelton and John (1996) that children’s participation in household labor “is typically occasional and their time investment small (p. 311).” Most studies of children’s household inputs, however, focus on young children or early adolescents (Bianchi & Robinson, 1997; Crouter, Manke & McHale, 1995; Peters & Haldeman, 1987); the household contribution of older teenagers, who are more capable of sharing the work at home as a result of increased maturity and skills, has received much less attention (for an exception, see Call, Mortimer & Shanahan, 1995). In addition, researchers typically have measured children’s contributions to household work by surveying the time they spend on an array of chores, many of which may actually be more appropriate for adults than children. Tasks that children are more likely to assume in families are often omitted from consideration. These problems are evident, for example, in studies employing data from the National Survey of Families and Households (Blair, 1992a; 1992b; Demo & Acock, 1993), where the household division of labor is assessed with a list that includes “adult” tasks like paying bills and driving, but excludes such “children’s” tasks as babysitting siblings and pet care. Therefore, the household contributions made by offspring may be underestimated. To address these gaps in past research that may have led to underestimates in children’s household inputs, this paper focuses on the household labor of teenagers as they approach adulthood. The analyses are aimed at examining longitudinally, across the high school years, the extent to which teens participate in a wide array of household chores--some of which may be unique to children--as well as identifying the factors that produce variability in their time

investments. Both individual and contextual factors have been linked to variations in children’s household work efforts. Gender is the key individual factor associated with variations in time devoted to housework, for both adults (Blair & Lichter, 1991; Shelton, 1991) and children (Antil et al., 1996; Blair, 1992b; White & Brinkerhoff, 1981). Researchers have found that similar to married couples, household chores among children are also sex-typed, with girls spending more time on a greater number of tasks than boys (Blair, 1992b; McHale et al., 1990; White & Brinkerhoff, 1981). Therefore, a central issue in this analysis is to determine whether a gender gap in household labor exists for teens as well, and whether that gap changes as teens approach adulthood. Concerning contextual factors, research on children’s housework has focused on a number of family structure characteristics, including maternal employment, marital status, and family size, but often in isolation of one another (Benin & Edwards, 1990; Demo & Acock, 1993; Hilton & Haldeman, 1991; Peters & Haldeman, 1987). Although insightful, these studies make it difficult to identify which family structure variables most significantly influence children’s household input; this is especially problematic given that these factors are often highly correlated. Therefore, our analysis improves upon past research by simultaneously examining the role of several key family context factors to determine their relative importance. In addition, we extend past research by recognizing that teens’ household efforts are likely to be influenced by the time demands they experience in social contexts beyond the family, in addition to those resulting from family structure characteristics. Thus, we specifically examine whether time spent doing homework, paid work, volunteer work, and extra-curricular activities is associated with teens’ contributions to household work. Consideration of time 2

devoted to non-family activities is especially important in an analysis of teens’ household roles because their social roles expand as they progress through adolescence (Youniss & Smollar, 1985). Finally, in examining variability in teens’ housework contributions, we consider whether differences in these contextual factors account for any gender variation in housework that we document. Just as research on adults has examined how other constraints on available time account for the gender gap in housework among wives and husbands (Coverman, 1985; England & Farkas, 1986), these same issues are likely to be relevant for teens as well. BACKGROUND Time availability theory proposes that the division of househo ld labor within a family results from different constraints on various family members’ time. Typically, the theory has been applied to married couples, where it is argued that husbands devote less time to housework than wives, because of the greater time invested in paid labor by husbands (Coverman, 1985; England & Farkas, 1986). How the time availability of both parents and children affects the household inputs of the younger generation has not been directly studied. In using time availability theory to conceptualize relevant family context factors, we consider which family structure characteristics create added time demands on parents. Time availability theory would suggest that parents who devote more time to paid work outside the home, have more children to care for, or who lack a spouse with whom to share family demands, will have less time available for housework. Consequently, offspring may be expected to pick up the slack. Thus, in this paper, we compare the effects of maternal employment, parents’ marital status, and family size on children’s contributions to household labor. Indeed, although research documents relatively low levels of household work by children 3

in general, the literature on children’s and adolescents’ housework supports the time availability theory. For example, offspring whose mothers are employed full-time appear to assume more responsibility for household tasks than do those whose mothers are not employed (Benin & Edwards, 1990; Peters & Haldeman, 1987; White & Brinkerhoff, 1981). This research consistently shows that having a full-time employed mother increases girls’ time spent on housework (Blair, 1992a,b; White & Brinkerhoff, 1991), while the effect on boys is less consistent. Some studies find no effect on boys’ inputs (Blair, 1992b), while others find mothers’ employment associated with a reduction in boys’ housework hours (Benin & Edwards, 1990). Additionally, while we might expect that living in a single- versus a two-parent household would increase children’s household labor inputs due to the absence of a second adult in the home, the research findings are mixed. Several studies indicate that children’s contributions to the maintenance of the home are greater, both in absolute hours and in the total share of the household workload, when they live with a single parent rather than two parents (Demo & Acock, 1993; Goldscheider & Waite, 1991). Other researchers report no effect of parents’ marital status (Bianchi & Robinson, 1997), while still others find children’s contributions are lower in single-parent families (Hilton & Haldeman, 1991; Peters & Haldeman, 1987). The inconsistencies in these findings may be due to reliance on non-representative convenience samples (Peters & Haldeman, 1987) or samples with low response rates (Benin & Edwards, 1990). Further differences may arise because some data were collected from mothers, (Blair, 1992a,b; White & Brinkerhoff, 1981), while others were collected from children, whose reports are likely to differ. Thus, further research is needed as prior findings may not be generalizable to all children of working mothers or single parents. 4

Studies of stepfamily structures also predict a greater contribution to total household labor by stepchildren, although the effects appear smaller than those for single-parent families (Demo & Acock, 1993; Goldscheider & Waite, 1991). Time constraints may also be responsible here, as parents and stepparents in relatively new remarriages may want to devote more time to their relationship and less to household labor (Ferri, 1984). Family size is also a significant contextual factor, as the number of individuals in a family is likely to affect both the amount of work created in a household, as well as the availability of persons to do it (Brody & Steelman, 1985). Bianchi and Robinson (1997) report a positive relationship between number of siblings in a family and the time a child spends on housework. This finding suggests that the extra work created by added family members is more important than the actual housework input contributed by each additional member. While parental time availability and subsequent parental expectations for help influence what teens do at home, certainly teens’ own experiences outside the family and home are likely to play a role in their level of involvement with housework as well. Research demonstrates that with increased age, the time teens allot to socializing with friends (Larson & Richards, 1994; Youniss & Smollar, 1985), participation in school activities (Kirshnit, Ham & Richards, 1989), homework, and paid employment (Manning, 1990; Mortimer et al., 1994) increases substantially. Thus, although teens may have a heightened capacity to shoulder more household responsibilities as they mature, changes in their time use beyond the family context may constrain the time available for various household tasks. Indeed, one of the few studies to look at the interconnections of family and non-family time use for teens shows that employed teens are significantly less likely than non-working peers to be assigned household chores by their parents (Manning, 1990). Thus, an important omission in research to date is consideration of this 5

wider set of non-family contextual factors in accounting for teens’ participation in household labor. Finally, because previous studies have not addressed contextual factors broadly or simultaneously, it is unclear how these factors may account for any observed gender differences in teens’ household input. For example, although family structure characteristics affect the input of offspring into the household workload, the added work resulting from such family situations as single-parent household structure generally is not divided equally among siblings. Specifically, girls appear to assume more of the extra work than boys in both dual-earner (Benin & Edwards, 1990; Blair, 1992a; 1992b) and single-parent families (Goldscheider & Waite, 1991). Therefore, increasing rates of divorce, and mothers’ labor force participation (DaVanzo & Rahman, 1993) will likely influence the gender gap in teens’ household contributions. In addition, if boys and girls differ in their participation in various roles outside the family, then the gender gap in household contributions is likely to be affected as well. The current analysis allows us to examine the extent to which these two sets of contextual factors account for gender differences in teens’ household labor across late adolescence. Past research and time availability theory suggest a number of hypotheses. We expect that teens’ contributions to housework will be greater when mothers are employed full-time rather than part time or not employed, when teens live in a family structure other than a twoparent, intact family, and when family size is larger. However, it is unclear which of these family context factors is likely to be most important. We also hypothesize that paid and volunteer work as well as school-related activities such as homework and extra-curricular activities, will constrain the time teens have available for housework. Yet, given the lack of consideration of this wider range of contextual factors in children’s time use, we are uncertain 6

how their influence may change as teens age. Finally, beyond these expected findings, we anticipate that gender will still account for a significant portion of the variance in teens’ household labor, as the limited evidence suggests that contextual factors may not influence males and females equally. DATA AND METHODS Sample and Procedures This analysis is based on the first and fourth waves of data from the Youth Development Study (YDS). Initiated in 1988, with a representative random sample (1,010) of all ninth graders enrolled in the St. Paul, Minnesota public schools, the YDS collected data on students’ experiences with family, school, work, and peers in addition to their mental health and achievement. Of these respondents, 93% were retained over the four-year period. Family background information also was collected from parents by mail in the first and fourth years of the study. Parent data were available for 96% of the respondents in 1988, dropping to 79% in 1991 at wave four. This analysis of teens’ housework is restricted to the 825 (82%) of the original wave 1 sample, which is a highly acceptable retention rate for data spanning a 4-year period. This analytic subsample includes only youth who participated in both waves 1 and 4 of the study, and who were residing with a parent at both points (otherwise housework contributions are likely to have a different meaning). Also, respondents with missing data on any of the dependent variables, or any of the categorical independent variables were eliminated. In Table 1, summary scores on the background characteristics and time use indicators for the 825 persons in the analytic sample are compared to the values for the 185 omitted respondents. As in most longitudinal samples, this analytic sample is favorably biased in terms of higher family income 7

and parental education than the group dropped from the analyses. Mothers of those dropped from the sample worked fewer hours than the retained group. Similarly, females and whites were more likely to be retained in the sample than males and minorities. Finally, youth dropped from the longitudinal sample were older and reported more time spent on housework in grade 9 than those who remained living with parents and in the study over time. Yet, despite this attrition, this sample still has the advantage of having been originally randomly selected (in contrast to several past studies), and the biases that exist are known so that we can consider their potential effects in evaluating the findings. -----------------------------Table 1 about here -----------------------------Dependent Variables The dependent variables in the analysis are the teens’ reports of time spent on household work. Respondents were asked how much time they spent in an average week doing each of 12 tasks. Based on past categorization in the literature (Berk, 1985; Huber & Spitze, 1983; Shelton, 1991), and patterns of gender-typing in these data (e.g., which gender reported more hours spent on a particular task), tasks were then categorized as “female,” “male,” and “gender-neutral.” Female-typed tasks included: cooking, setting the table, washing dishes, laundry, cleaning the house, taking care of young children, taking care of elderly family members, and grocery shopping. Yard work (mowing grass, shoveling snow, etc.) and taking out the trash were the only tasks for which males reported more hours than females, and thus are referred to as maletyped. Finally, gender-neutral tasks included: cleaning one’s own room and pet care. Total number of hours spent on these three types of tasks was calculated for both males and females at both waves 1 and 4.

8

Independent Variables Data on maternal employment, family size, and parents' marital status were drawn from the 1988 and 1991 parent surveys. Maternal employment was initially coded as the mother’s reported hours at her main job in a typical week. Non-employed mothers received a value of 0 on this variable. Three dummy variables were then constructed to represent : no formal employment, part-time (1 - 35 hours), and full-time (35 or more hours) work. Family size was calculated by summing the number of biological, step, adopted, and foster children under age 18 in the household, including the teen respondent. We added 1 to this total when teens reported caring for elders living in the home. This variable represents the number of “dependents” in the household. Households were further coded into two categories representing large (3 or more dependents) and small (2 or fewer dependents) families. Rather than include parents in the family size variable, we capture their presence in the home with a marital status variable categorized as: 1) single-parent (based on whom the teen lived with at years 1 and 4); 2) step (if the custodial parent was remarried); or, 3) two-parent household. The time respondents spent on activities other than housework was gathered at both waves of the YDS survey. Four activities were considered: homework, extra-curricular activities, paid work, and volunteer work. (We refer to these four time use variables collectively as non-family activities). For homework, respondents reported the minutes they spent, on average, doing homework on week nights and during weekends. Weeknight minutes were multiplied by four (we assume few do homework on Friday nights) and added to minutes reported for weekend homework, and then transformed into total hourly units. Involvement in extra-curricular activities included the number of hours spent after school per weekday and on weekends participating in activities such as clubs, band, and sports. Volunteer unpaid work was 9

measured with a question asking about total hours per week spent on volunteer work. Respondents who noted that the time “varied” were asked to report a low and a high number of weekly hours for such work, from which an average was computed. Last, respondents were asked to report the average number of hours they spent in paid employment per week. Family Background Control Variables Other family characteristics that may influence teens’ hourly contributions to household labor are race, and parents’ income, education, and age. Although potentially important, these factors are included as control variables because they are not as clearly linked to time availability theory as the other family context variables we have selected. Race of respondent is coded with three dichotomous dummy variables: Black (1, 0); other minority (1, 0); and, Non-Hispanic Whites (1, 0). The latter variable is used as the omitted comparison category assuming a value of 0 on both the “Black” and “other minority” variables. Parental income, age, and education were coded from the parent surveys. Parental income is the income of the custodial parent if the child lived with a single parent. Parents’ education was the average of the years completed by each parent if the teen lived with both parents. Otherwise, we used the custodial parent’s education level. Finally, parent age is indexed by the age of the oldest parent for those in twoparent households, or the custodial parent for single-parent households. These variables are important to control as they are likely to be correlated with family structure. Data Adjustments Median values were imputed to replace missing data for the ordinal and interval level independent variables. Missing data never represented more than 8% of the cases on any one variable pertaining to the respondents. However, for parent data, especially at Wave 4, missing data were more common (20-30% of the cases on a few variables). Therefore, when possible, 10

wave 1 values were imputed for missing data at wave 4, as wave 1 and wave 4 background characteristics were highly correlated in the data that were available. In addition, severe outliers, those values extending beyond a distance three times the size of the interquartile range (Ott, 1984), were handled by recoding them to the maximum permissible value within the range. Follow-up analyses revealed, however, that this recoding did not alter the results of the analyses. FINDINGS Time Use Across Late Adolescence To address the level of teens’ contributions to household work and how it changes as they move through late adolescence, we compare males’ and females’ time spent on various types of tasks in ninth and twelfth grades using repeated measures ANOVAs. These analyses involve a 2 (gender) X 2 (time) design, with time as the repeated measure. This approach allows for a test of main effects of time and gender (revealing potential age differences across gender and gender differences across age in housework hours), as well as a test of the interaction of gender and time. A significant interaction term would indicate that the gender gap in hours spent on a particular activity changed significantly over the high school years. In Table 2, we present the results of this repeated measures ANOVA procedure. In addition to testing for gender and time effects on the household tasks of interest, we also examine gender and time effects for the four non-family time use activities. This latter comparison provides preliminary evidence of how other activities may assume increased importance across the mid to late adolescent years, potentially competing with teens’ household work investments. -----------------------------Table 2 about here -----------------------------11

Gender differences in time use. The results in Table 2 indicate significant gender differences in the time devoted to both household work and non-family activities across the teen years. Not surprisingly, boys contribute more time than girls to male-typed tasks, whereas girls devote more hours each week than boys to female-typed chores. Sex differences are not apparent on gender-neutral tasks, which account for relatively few hours each week for both sexes. When total housework is considered, we document a significant gender gap, with boys devoting only 87% as much time as girls to housework in ninth grade, and about 68% as much time by grade twelve. Non-family activities also reveal significant gender differences. Girls spend more time than boys on homework and paid employment in both grades nine and twelve. Across high school, however, boys significantly surpass girls in the number of hours spent in extra-curricular activities. The sharpest gender distinctions are captured in ninth grade with girls reporting nearly twice as much time spent in paid employment than boys, and in twelfth grade where girls commit only 60% as much time to extra-curricular pursuits as do boys. Longitudinal changes in time use. Across late adolescence, time invested in household tasks declines significantly. Boys hours decrease by approximately 40% between grades nine and twelve, while girls’ contributions decline by about 25%. We find significant time effects for each of the three types of tasks. Although housework contributions consistently decline over high school, the time devoted to non-family activities shows a mixed pattern of change. Teens report an increase in time spent on homework and paid work between ninth and twelfth grades, although time spent on volunteer work declines and extra-curricular involvement remains relatively stable. The most

12

dramatic change is in the time devoted to paid employment, followed by increases in time spent on homework. These time use findings suggest that the hours teens devote to non-family activities, especially paid work and homework, may indeed influence their contributions to housework over the years associated with high school. Changes in the Gender Gap. The results presented in Table 2 suggest that the gender gap in adolescents’ time devoted to housework changes substantially over high school, whereas nonfamily time use differences for males and females remain relatively stable. Two time X gender interactions are significant. First, we find a greater decline in females’ versus males’ contributions to male tasks over four years. Second, we find a more substantial decrease in males’ than females’ time spent on total chores over the same period. This latter interaction means that the gender gap in total household labor expands over the high school years, as shown in Figure 1. -------------------------------Figure 1 about here -------------------------------Gender, Family Structure and Time Use as Predictors of Teens' Household Labor The ANOVA results above address the role of gender in teens’ involvement in housework and in non-family time use activities that may compete with teens’ household responsibilities. Next, we use multivariate regression models to consider how both non-family and family context variables, jointly influence the time devoted to housework by teens. Panels A and B of Table 3 present the unstandardized regression estimates for teens’ household contributions in grade nine and twelve, respectively. Models estimating contributions to genderneutral tasks are not shown here, as they revealed no gender or family structure effects in preliminary analysis. However, we account for pet care and cleaning one’s own room by 13

including them in total tasks. In Table 3, Models 1A, 1B, and 1C show the estimates for the effect of the gender variable only estimates for each set of tasks--total, female-typed, and male-typed. To Models 1B, 2B, and 3B, we add the contextual variables to determine their relative influence and to assess whether they diminish the importance of gender in predicting teens’ household labor. Finally, we present a third model if the follow-up analyses revealed a significant interaction between type of task, gender, and the family context variables. -------------------------------Table 3 about here -------------------------------First, Models 1A, 1B and 1C, which include only gender, show that gender is a strong and consistent predictor of time input. This finding holds for all three sets of tasks and at both grades nine and twelve, as expected given the bivariate results. In ninth grade, girls average over two hours more per week on household tasks than boys. By twelfth grade, this time difference has nearly doubled. The gender difference appears related to the fact that girls devote more time to a greater number of tasks than do boys, as the chores typed as “female tasks” are the tasks that are more time consuming and repetitive. Although boys spend more time than girls on stereotypically male tasks--nearly two hours per week in ninth grade and just under an hour per week in twelfth grade--few household tasks are predominantly performed by boys and those that tend to be the least repetitive (Blair & Lichter, 1991). In the expanded models (Models 2A, 2B and 2C) for both grades, we see that several family structure variables influence household work by teens, with only minimal changes in the effects of gender. In ninth grade, family size consistently and significantly affects teens’ household contributions. Not surprisingly, teens from large families assume a greater total 14

workload at home than their peers in small families. This finding holds for both male- and female-typed tasks in ninth grade. Tests for interactions, however, reveal that a main effects model did not hold for twelfth grade data for both female-typed and total tasks. At this point, females’ workloads are much more responsive than males’ to family size fluctuations. As shown in Panel B, for total tasks and female tasks, (Models 3A and 3C) compared to their male peers, females in large families spend about 5 more hours on housework per week, and 2 hours more than their female peers in small families. Time spent caring for siblings may explain much of this difference. In ninth grade, caring for younger children accounts for approximately 25% of the time reported on female-typed tasks by both males and females in large families (analysis not shown). By twelfth grade, time spent caring for siblings by males declines substantially to about 14% of their time, while for females it still consumes about 19% of their time. In small families, only about 4% of twelfth grade females’ and males’ female-task time is devoted to child care. The effect of the other family context variables is less consistent across task types and grades. First, parents’ marital status was not uniformly related to household inputs in ninth grade. Although, as predicted, teens with single parents spent more time on female-typed chores than those in intact families, surprisingly teens living with a remarried parent devoted less time to male-typed tasks than those living with their married parents. The influence of parents’ marital status was much greater by twelfth grade. Across tasks, twelfth graders living with single parents devote the most hours to housework--averaging a total of three more hours per week than peers in intact two-parent households. Youth living with a remarried parent and step-parent show the same trend, but significant differences from peers in intact two-parent households in terms of an increased contribution to female tasks. Mothers’ employment status does not predict time spent on household chores, at either 15

grade level. Although, in general, the coefficients for these employment variables are in the expected negative direction--indicating less time spent on housework by teens if mothers are not employed full-time, none of the effects were significant. Teens’ time investments in non-family activities predicted contributions to household work in different ways in ninth and twelfth grades. Unexpectedly, those doing more volunteer work reported significantly more time spent on female tasks in ninth grade. Similarly, ninth graders who spent more time on homework and extra-curricular activities also devote more time to household work, and this is true across all types of chores. However, follow-up analyses reveal gender interactions when considering male and female-type tasks. Added time spent on homework predicts greater input into female-type chores for girls but not boys. In contrast, time investment in extra-curricular activities predicts significantly more time spent on male-type tasks, but only for ninth grade boys (Panel A, 3B and 3C). By twelfth grade, the expansion in teens’ paid work involvement has started to negatively affect their housework contributions. More hours spent working for pay predicts less time spent on female-typed chores and total tasks. Still, in twelfth grade, teens of both sexes who devote more time to homework and volunteer work also are spending significantly more time on housework. At this point, however, extra-curricular activities appear relatively unimportant in determining how much teens help out at home. As a last step in model estimation, we tested whether housework may be the stronger predictor of how teens spend their time than vice versa. Yet, when the regression equations were reversed with non-family time use as the dependent variables and housework as the independent variables, none of the time-use variables are significant. Thus, we are confident that our model specification is correct. 16

DISCUSSION AND CONCLUSIONS A central aim of this paper has been to reevaluate the common conclusion that children’s household labor inputs are minimal (see Shelton & John, 1996). This conclusion may be more accurate for young children than for teens, whose contributions have received less attention, and whose greater maturity renders them more capable of sharing housework. Our findings reveal that teens do spend more time on household work than reported in previous research on younger adolescents and teens (Benin & Edwards, 1990; White & Brinkerhoff, 1981) or on younger children (Blair 1992a, b; Bianchi & Robinson, 1997). On average, teen boys in our sample reported spending nearly 15 hours a week on chores in the ninth grade, while girls of the same age devoted 17 hours to chores. Although teens’ time on housework decreases between ninth and twelfth grades, their inputs were still substantial in twelfth grade, with boys and girls spending 9 and 13 hours, respectively. The inclusion of the chores typically performed by children, such as babysitting siblings or taking care of pets, likely explains our higher housework participation. As expected, babysitting siblings accounted for a significant proportion of the time devoted to helping out at home, especially among ninth graders, whose siblings are typically younger, and among those from larger families. Clearly, failing to collect data on time spent caring for siblings is a serious omission in research on children’s housework contributions. Given high rates of maternal employment and single-parenting--family features that heighten the demand for sibling care (Benin & Edwards, 1990)-- it is increasingly important to correct this omission in future studies. We concede that our higher estimates may also be somewhat biased by the data collection method of direct questioning about individual tasks compared to data collected from the daily time-diary approach (Benin & Edwards, 1990; Marini & Shelton, 1993). However, self-report 17

bias is a potential threat with this latter method as well. The strength of our data is the reliance on teenagers’ own estimates of how much time they spend on each task per week, while previous researchers have based estimates on parents’ reports of children’s household time (White & Brinkerhoff, 1981, Blair, 1992a,b). It is not apparent why teens’ estimates would be any less reliable than those of their parents, especially given their maturity and the fact that parent’s may not be home to monitor teen’s housework time. Last, our higher time estimates may also be due to more accurately capturing the range of housework tasks teens do, such as babysitting siblings -- a task that consumes a considerable share of ninth graders’ time. Our analyses do coincide with previous findings, by consistently showing that gender is a key determinant of the assignment of tasks and of the absolute hours spent on household labor. Much of the observed gender gap appears to be attributable to the sex-typing of teens’ household labor. Because many more household tasks are typically female-typed (i.e. females spend more time on these than males do), and female tasks are performed daily or several times daily, girls spend significantly more time on a greater number of tasks than boys do (see also Blair, 1992b; Call, 1996). Furthermore, the analyses show that although both girls’ and boys’ participation in household labor decreases as they age, the gender gap in housework actually increases over the high school years. Specifically, boys reduce their hours spent on housework more than girls, over the high school years. The gender gap widens, despite the fact that both sexes substantially increase their paid employment hours. This finding that teens’ paid work participation reduces their time spent on housework, especially in grade twelve, provides the best support for hypothesis that teens’ increased time on activities outside the home interferes with their housework contributions. This finding also validates Manning’s (1990) evidence that parents are less likely to assign household 18

tasks to employed teens than non-employed teens. Thus, we do find support for the time availability hypothesis that teens’ expanding roles outside the family are associated with reductions in household inputs. The time devoted to other activities also reveals interesting gender contrasts for teens, although the findings refute the time availability hypothesis. First, girls devote more time than boys to all the activities we examined, except extra-curricular activities, which often include leisure type activities such as sports, music, or clubs (FIFCFS, 1998). Thus, it appears that the leisure gap and the “second shift,” which has been documented between wives and husbands (Hochschild, 1989) originates in one’s own family. Second, the finding that involvement in extra-curricular activities (in ninth grade), homework, and volunteer work are associated with increased time spent on total household chores, contradicts the time availability hypothesis. We speculate that teens who spend more time doing non-family activities may be more motivated overall, which may affect household labor inputs. In the case of homework, it is also possible that teens can easily combine their housework and homework; chores such as babysitting younger siblings, which account for a large portion of teens’ household inputs, are easily compatible with academic work. Beyond documenting the gender gap in teens’ time-use as well as the effect of this timeuse on household inputs, we also show that teens’ housework time varies by key family characteristics. The time spent on housework is greatest and most consistent over the high school years in larger families and in single-parent families. Specifically, teens do more housework when their mother’s time is constrained (as a result of having a larger family or being a single parent). It is important to note that our bias analyses revealed more sample attrition of teens from single-parent families and of teens who performed more housework. Therefore, our 19

analysis may underestimate the effect of single-parenthood on teens’ household labor contributions. Thus, although much debate has ensued over the negative consequences of divorce for children (McLanahan & Sandefur, 1994), our findings suggest that growing up in a single-parent family may promote greater involvement in the family work effort by both boys and girls. In turn, this increased participation in household responsibilities may result in greater equity when teens enter their own intimate relationships in adulthood (Gager, 1998; Goldscheider & Waite, 1991). We also found twelfth grade teens in remarried families devoted more time to both female and total household tasks, although ninth grade teens contributed less time to male tasks. These findings confirmed our hypothesis : we expected teens in remarried families to spend more hours on housework given their likelihood of previously having spent time in a single-parent family--a family structure associated with higher housework input by teens (Goldscheider & Waite, 1991). In contrast, the finding that ninth grade teens in stepfamilies spend less time on male tasks than their peers was unexpected. This may result because outdoor chores, like yard work, are often performed with a father/stepfather. Since teens are more likely to have problematic relationships with stepfathers than their biological fathers, they may avoid these chores to elude spending time with their stepfather. Further, Hetherington and Jodl (1994) argue that during early adolescence, stepfathers typically are less controlling than are biological fathers, and may not press stepchildren to spend time on these male chores. The results for maternal employment do not support the time availability theory. In general, we found that teens’ housework input did not vary substantially based on mothers’ employment status. However, as Hochschild (1989) has argued, it is highly likely that mothers who work full-time lower their standards to reduce time spent on housework, thereby reducing 20

the burden on their children. These mixed findings suggest that time availability alone cannot fully explain why girls perform more housework than boys, and that further study of these linkages is warranted. In sum, our findings provide partial support for time availability theory by showing that some family structure characteristics and teens’ own time use do play a significant role in the amount of housework teenager perform, but within a gendered context. Similar to the past, when married women comprised a reserve army of paid labor in times of war, teens may constitute a reserve army of household labor that can be recruited in times of family need, such as in singleparent families and in larger families. However, our findings show that when we consider all household tasks, teenage girls are more likely than boys to pick up the slack when the need arises. Thus, despite major changes in the roles of men and women, we find evidence that teens continue to be socialized into gender-specific roles through the sex-typing of household assignments, and that these role assignments become more pronounced as teens approach adulthood.

21

References Antil, J. K., Goodnow, J., Russell, G., & Cotton, S. (1996). The influence of parents and family context on children’s involvement in household tasks. Sex Roles, 43, 215-236. Benin, M. H., & Edwards, D. A. (1990). Adolescent chores: The difference between dual- and single-earner families. Journal of Marriage and the Family, 52, 361-373. Berk, S. F. (1985). The gender factory: The apportionment of work in American households. New York: Plenum. Bianchi, S., & Robinson, J. (1997). What did you do today?: Children’s use of time, family composition and the acquisition of social capital. Journal of Marriage and the Family, 59, 332-344. Blair, S. L., & Lichter, D. T. (1991). Measuring the division of household labor: Gender segregation of housework among American couples. Journal of Family Issues, 12, 91113. Blair, S. L. (1993). Employment, family, and perceptions of marital quality among husbands and wives. Journal of Marriage and the Family, 14, 189-212. Blair, S. L. (1992a). Children’s participation in household labor: Socialization versus the need for household labor. Journal of Youth and Adolescence, 21, 241-58. Blair, S. L. (1992b). The sex-typing of children’s household labor: Parental influence on daughters’ and sons’ housework. Youth and Society, 24, 178-203. Brody, C. J., & Steelman, L. (1985). Sibling structure and parental sex-typing of children’s household tasks. Journal of Marriage and the Family, 47, 265-273. Call, K. T. (1996). The implications of helpfulness for possible selves. Pp. 63-96 in J. T. Mortimer & M. Finch (Eds.), Adolescents, work and family. Newbury Park, CA: Sage 22

Publications. Call, K., Mortimer, J. T., & Shanahan, M. (1995). Helpfulness and the development of competence in adolescence. Child Development, 66, 129-138. Coverman, S. (1985). Explaining husbands’ participation in domestic labor. Sociological Quarterly, 26, 81-97. Crouter, A. C., Manke, B. A., & McHale, S. A. (1995). The family context of gender intensification in early adolescence. Child Development, 66, 317-329. DaVanzo, J., & Rahman, O. (1993). American families: Trends and correlates. Population Index, 59, 350-86. Demo, D., & Acock, A. C. (1993). Family diversity and the division of domestic labor. Family Relations, 42, 323-31. England, P., & Farkas, G. (1986). Households, employment and gender: A social economic and demographic view. New York: Aldine. Federal Interagency Forum on Child and Family Statistics. (1998). America’s children: Key national indicators of well-being. Federal Interagency Forum on Child and Family Statistics, Washington DC : U.S. Government Printing Office. Ferri, E. (1984). Stepchildren: A national study. Windsor, Berkshire: Nfer-Nelson. Gager, C. T. (1998). The role of valued outcomes, justifications, and comparison referents in perceptions of fairness among dual-earner couples. Journal of Family Issues,19, 622648. Goldscheider, F. K., & Waite, L. J. (1991). New families, no families: The transformation of the American home. Berkley CA: University of California Press. Hetherington, E.M., & Jodl, K.M. (1994). Stepfamilies as settings for child development. Pp. 23

55-79 in A. Booth & J. Dunn, (Eds.) Stepfamilies: Who benefits, who does not? Hillsdale NJ: Lawrence Earlbaum and Associates. Hilton, J. M., & Haldeman, V. A. (1991). Gender differences in the performance of household tasks by adults and children in single-parent and two-parent, two-earner families. Journal of Family Issues, 12, 114-130. Hochschild, A. (1989). The second shift. New York: Basic Books. Huber, J., & Spitze, G. (1983). Sex stratification: Children, housework and jobs. New York: Academic Press. Kirshnit, C., Ham, M., & Richards, M. (1989). The sporting life: Athletic activities during early adolescence. Journal of Youth and Adolescence, 18, 601-615. Larson, R., & Richards, M. H. (1994). Divergent realities: The emotional lives of mothers, fathers and adolescents. New York: Basic Books. Manning, W. (1990). Parenting employed teenagers. Youth and Society, 22, 184-200. Marini, M. M. and & Shelton, B. A. 1993. “Measuring household work: Recent experience in the United States.” Social Science Research, 22, 361-382. McHale, S. M., Bartko, W. T., Crouter, A.C., & Perry-Jenkins, M. (1990). Children’s housework and psychosocial functioning: The mediating effects of parents’ sex-role behaviors and attitudes. Child Development, 61, 1413-1426. McLanahan, S. and Sandefur, G. Growing up with a single parent: What hurts what helps. Cambridge: Harvard University Press. Mortimer, J. T., Finch, M. D., Dennehey, K., Lee, C., & Beebe, T. (1994). Work experience in adolescence. Journal of Vocational Educational Research, 19, 39-70. Ott, L. (1984). An introduction to statistical methods and data analysis (2nd Edition). Boston: 24

Duxbury Press. Peters, J. M., & Haldeman, V. A. (1987). Time used for household work: A study of school-age children from single-parent, two-parent, one-earner, and two-earner families. Journal of Family Issues, 8, 212-225. Pina, D. L., & Bengston, V. L. (1993). The division of household labor and wives’ happiness: Ideology, employment and perceptions of support. Journal of Marriage and the Family, 55, 901-912. Shelton, B. A., & John, D. (1996). The division of household labor. Annual Review of Sociology, 22, 299-322. Shelton, B. A. (1991). Women, men and time. Westport, CT: Greenwood Press. Timmer S., Eccles, J., & O’Brien, K. (1985). How children use time. Pp. 353-383 in T. Juster & F. Stafford, (Eds.) Time, goods and well-being. Ann Arbor MI: University of Michigan Institute for Social Research. West, C., & Fernstemaker, S. (1993). Power, inequality, and the accomplishment of gender. Gender & Society, 9, 8-17 White, L. K., & Brinkerhoff, D. (1981). The sexual division of labor. Social Forces, 60, 17081. Youniss, J. & Smollar, J. (1985). Adolescents’ relations with mother, father, and friends. Chicago: University of Chicago Press.

25

Table 1. Means Values (and standard deviations) for Analytic Sub-sample and Cases Eliminated from the Analyses. Variable Analytic Sample (n=825) Omitted Cases (n=185) a Significance Percent of Sample Male 44.70% 61.62% X2 = 17.29*** Percent of Sample Black 8.85% 16.23% X2 = 7.86** Other Minority Status 13.70% 30.52% X2 = 26.86** Average Age of Respondent

14.64 (.65)

14.93 (.82)

t = 4.02***

Mother’s Work Hours

28.83 (18.37)

24.18 (17.93)

t = 3.12**

Oldest Parent’s Age

42.30 (6.30)

41.56 (7.18)

t = 1.29NS

Parents’ Average Education 1 = < high school to 8 = professional or Ph.D.

3.20 (1.64)

2.83 (1.47)

t = 2.82**

Family Income 1 = < $5000/year to 13 = $100,000+/year

6.29 ( 2.20)

5.90 (2.10)

t = 2.17*

Percent in Families w/ 3 or More Dependents

33.30 %

33.51%

X2 = .00 NS

Percent of Sample - Parent Single Remarried Intact Marriage

25.70 % 15.39 % 58.91 %

40.74% 16.05% 43.21%

X2 = 15.15*** X2 = .04 NS X2 = 13.51***

Table 1 continued Variable

Analytic Sample (n=825)

Omitted Cases (n=185) a

Significance

Total Task Hours - Grade 9 Malesb Femalesc

14.41 (11.60) 17.25 (11.55)

18.06 (12.06) 19.79 (12.20)

t = 2.93** t = 1.99*

Paid Work Hours - Grade 9 Males Females

3.58 (6.65) 6.89 (8.45)

5.56 (10.08) 7.87 (9.46)

t = 1.97* t = .82 NS

Homework Hours - Grade 9 Males Females

5.92 (5.22) 7.56 (4.99)

5.75 (5.59) 7.74 (5.68)

t = .31 NS t = .28 NS

Volunteers Hours - Grade 9 Males Females

.57 (1.37) .65 (1.45)

.41 (1.07) .47 (1.26)

t = 1.24 NS t = 1.01 NS

Extra-curricular Activities Hours - Grade 9 Males Females

9.09 (9.73) 6.46 (7.19)

7.47 (10.26) 6.76 (8.28)

t = 1.54 NS t = .32 NS

* p