Women and Leadership Project

INTERPERSONAL RELATIONS AND GROUP PROCESSES

Sharing of Home Responsibilities Between Professionally Employed Women and Their Husbands

Monica Biernat Camille B. Wortman University of Florida State University of New ~rlc at Stonybrook

A sample off 39 married couples with youDg children and with relatively equal status careers (wives were univen.ity professors or busiDesswomen) were interviewed about work and home life. Consid- erable, traditional inequity in the distribution of child-care tasks and chore responsibility was noted, but women were generally satisfied with their husbands' home involvement. In the academic sample, the longer hours each spouse worked, the more child care the other performed; io the business sample, child-care involvement was largdy determined by the husband's worlc hou.n, income, and education. Overall, women were more self-critical than were men about their perfor- mance in home roles, and women's role performance was rated more hi&hJy by husbands than by thernselve11- Women professionals' continued use of tniditional sex role standards and the imJXJr• tance ofattending to both panners' penpectives in studies of married life are discussed .

The female labor force in the United States is both larger and more demographically diverse today than it was just a few de- cades ago. Not only do women make up nearly 50% of the total Jabor force, but also they are more likely to be in administrative and professional occupations, to be married, and to have chil- dren under the age of6 years than they were in the past (Beller, J982; Hayghe, 1990; U.S. Bureau ofthe Census, 1988; US. De- partment of Labor, 1977). Women who represent the conjunc- tion of these demographic features-married professionals with young children-are growing in number as well: In 1975, about 720,000 married women with childKn less than 6 years old were employed in professional and technical occupations; by 1988 that number had increased by half (US. Bureau of Labor Statistics, 1989; U.S. Department of Labor, 1977). These women, whose Jives reOect the elevation ofcareer to primary status along with marriage and parenthood, are, nonetheless, underrepresented in research on women's work and dual- e11rner couples.

Thiis rcsc:arch projc;ct wu supported by gmnts from the National Science Foundation (DNS•841-7745) and lhe National Institute of Mental Health (MH40255-01), both awarded to C.amillc B. Wortman and Carol-Ann Emmons. We extend warm thanks to our colleaaueson this project-Carol-Ann Emmons, Linda Beth Tiedjc, and Eric Lang - tor their extensive involvement in c,,,nccpluali1.ing, developing, and implementina tbe survey instnunent5. We 11lso thank our int.crvir..-wen; and office staff for their suppon and Chris Crandall, Roxy Silver, and several anonymous reviewers for their helpful commems on an earlier version of this article.

The data for this project were collected while we wereat the Institute for Social Research, University ofMichigan.

Correspondence com.-crning this article should be addre$Scd to Monica Biernat. Department of Psychology, University of Florida, Gainesville, Florida '.3261 1.

It seems. however, that underscandlng how working mothers who choose to remain fully involved in their careers experience and meet the competing demands oftheir multiple roles8hould be considered a research priority. Although there is evidence that paid employment can promote psychoJocica) well-being in married women (Gove & Geerken, 1977; Kessler & McRae, 198 l; Welch & Booth, 1977), it is also clear that the pressuresof managing multiple roles are the greatest, and the psychological benefits ofemployment are the least, under conditionsofh~y family respansibilities, that is, wbeo young children are at home (Gore & Mangione, 1983; Staines, Pleck, Shepard, & O'Connor, 1978; White, Booth, & Edwards, 1986). Further- more, the stresses faced by professionally employed mo1hers are costly not only in terms ofindividuaJ mental health, but also to the welfare ofthe family and to society as a whoJe. Dropping out of the labor force to meet family demands, even for a short time, can negatively atfecta woman'scareer(Rapo.,>Ort &Rapo- port, 1971 ), as well as career opportunities for women in gcm- eral (Fraker, 1984). And cutbacks on career involvement remove women from the labor market at a ti.me when they are DlOllt well-trained and valuable (Behrman, 1982). It seems essential, tben, that what happens in the homes ofmarried women profes- sionals, where joh and family pressures meet, be undentood. This is our prirnary objective in the presentarricle-to examine lhe inner workings of marriages characterized by the rclatively equ.al-st.o.tus careers of the partnen and the presence of YoWl& children in the horne. SpeciftcaJ~ we explore each spouse's perceptions of the distribution of cbild-QIC mponsibilities and then examine potential causes and consequences of the patterns of home respansibility we observe.

A number of studies have shown that the ~t majority of husbands and wives be~e that when the wife is employed, husbands should increase their iDV<>lvement in houschoJd chores (Ferber, 1982). \'et these beliefsdo not translate well into

Journal of P,:no11111ity ■ nd Social l's>dnlntl,.; 19'11, Vol. 60, No. 6, 8-4+-860 l"opyrip,t l!'91 t,y 11.. 1\m,,;.,un Psychc,logical AUOCiatiofl, Inc. 0022-3jl4/91/S3.00

844

845 HOME LIVES OF WOMEN PROFESSIONALS

action, A wife's labor participation may be associated with a decrease in the hours of housework she performs, but does not increase her husband's level of involvement (Berheide, Berk, & Berk, 1976; R. Berk & Berk, 1979; S. Berk, 1985; Bryson, Bry- son, Licht, & Licht, 1976; Condran & Bode, 1982; Geerken & Gove, 1983; Model, 1982; Oakley, 1974; Robinson, 1977, 1980; Slocum & Nye, 1976; Stafford, Backman, & Dibona, 1977; Vanek, 1974). Furthermore, ahhough employed wives may de- crease their total amount of housework time, they nonetheless retain primary home responsibility, thus carrying on "double- duty days" (Firestone & Shelton, 1988; Hochschild, 1989). In short, the redistribution ofinstrumental roles within the home does not seem to match the sharing of these roles outside the home.

One might expect a more egalitarian pattern in marriages in which partners' careers are of equal status and their work re­ sponsibilities are equally demanding. There is evidence that people in professional occupations are more likely than others to express egalitarian sex role beliefs (Acock & Edwards, 1982; Huber & Spitze, 1983; Thornton, Alwin, & Camburn, 1983), but nonetheless, those few studies that have examined dual-car- eer couples as a special case of dual-earner couples have found that wives still perform a disproportionate share of household duties (Berardo, Shehan, & Leslie, 1987; Weingarten, 1978; Yo- ge'\ 1981 ). Almost nothing is known, however, about how care of young children is managed in such marriages. To our knowl- edge, there has not been a single published study focusing on the division of child care and home responsibilities in profes- sionally employed couples with young children. Our data pro- vide this focus and allow us to ask whether these couples have adapted to the recent cultural trend they represent, in which women work at careers rather than stay home during their chil- dren's preschool years. Adaptation is clearly necessary, as the traditional, "male" model of career involvement, in which job success is made possible in part by the wife's assumption of total responsibility for family life, is no longer a viable option for either member of these couples (see Hunt & Hunt, 1982). If the traditional model is maintained, wives cannot benefit from freedom from home responsibilities and consequent single- mindedness toward careers. With a nontraditional model, how- ever, husbands may find themselves competing at work with other men who do have wives in traditional roles.

How, then, are professionally employed couples with young children managing? How are child-care and housework respon- sibilities divided in these homes? Do equal-status careers make for equal division of home duties? This is the first issue we address in this article. The second is this: If home responsibili- ties are not divided equall)o; why is that the case? What are the likely causes and consequences of inequitable division oflabor?

Potential Causes of Inequitable Chore Division

Socialization by Employed Mothers

Families today are making decisions about division of house- hold responsibilities against the backdrop oflifelong socializa- tion in traditional sex role beliefs (Ferber & Birnbaum, 1977). Several studies have indicated that men have more traditional expectations about marriage than do women (Komarovsk),;

1973; Mason & Bumpass, 1975; Mirowsky & Ross, 1987; Os- mond & Martin, 197 5) and that when husbands take on house- hold responsibility, they typically choose tasks that do not threaten their masculine self-images {Weingarten, 1978; Yoge'\ 1981 ). One basis of continued inequity in professional couples may be their conventional sex role beliefs, built up through lifetimes ofsocialization by mothers and fathers who held tradi- tional roles. We expect to find greater equity in division of household responsibility in couples in which either spouse was raised by an employed mother or female guardian.

Education and Income

Power or status differentials between spouses (resulting, for example, from income, education, and occupational prestige) may be at the heart of who does what in the home (Blood & Wolfe, 1960; England & Farkas, 1986; Eriksen, Yancey, & Erik- sen, 1979). Traditionally, husbands have played the more power- ful role in marriages, primarily because they have provided the financial resources. In our sample ofprofessional couples, how- ever, this is obviously not the case, and it is possible to test the hypothesis that spouses (men or women) who earn more money and have attained higher educational levels are least responsible for performing chores and child care. It is also true, however, that college education has a liberalizing effect on attitudes (Newcomb, 1943). Therefore, one might expect highly educated husbands {despite their high status) to be less traditional and to participate to a greater extent in home chores. Of course, re- spondents in the present data set represent a relatively restricted range of education and income levels, thus these hypothesis tests are necessarily conservative.

A subsidiary goal in examining the income variable is to assess its relationship to respondents' self-evaluations of their role performances. Fulfilling the breadwinner role has tradi- tionally been very important to men's identity in that it is neces- sary to being a good husband and father {see Veroff, Douvan, & Kulka, 1981 ). Using the present data, we examine whether in- come has comparable evaluative implications for women whose primary life roles include a demanding career.

Time Spent on the Job

It seems plausible that "who does what" in the home is based, at least in part, on the simple reality of available time. More time spent on job-related activities means less time available at home. A number ofstudies have suggested that the more fathers work, the less involved they are with their children {McHale & Huston, 1984; Robinson, 1977), but little is known about this relationship among mothers. Again, women have traditionally had more time to engage in home chores; in this sample, how- ever, both spouses' careers demand considerable time. We ex- pect to find direct, negative relationships between hours spent at work and involvement in child care and household chores for both spouses.

Psychological Job Importance

Finally, although wives and husbands may have equally de- manding or equal-status careers, it is still possible for the career

846 MONICA BIERNAT AND CAMILLE B. WORTMAN

of one spouse to be more psychologically important than the other's. The spouse with the more significant job is likely to be absolved of home responsibilities. To operationalize relative job importance, we assessed spouses' perceptions of whose career took precedence.

Potential Consequences of Inequitable Chore Division

We suggest that a spouse's level of involvement in child care may affect his or her self-ratings of role performance and mari- tal and parenting satisfaction. Specifically, if women gain self- esteem and marital stability by taking primary responsibility for the home and therefore conforming to traditional sex role expectations (Yogev, 1981, 1987), those who do more child care and housework should rate themselves higher in subjective satis- faction with role performance. Evidence does suggest that work- ing mothers are often concerned about failing to fulfill tradi- tional home role obligations (Baruch & Barnett, 1986; Robin- son, Yerby, Fieweger, & Somerick, 1977). However, because employed husbands have traditionally not regarded the running of the home as a source of self-identity, their levels of involve- ment in housework and child care may be unrelated to self- rated role performance and satisfaction. Alternatively, because men have traditionally fulfilled their spouse and parenting roles through work involvement rather than direct care ofchildren or the home, husbands may feel resentful and negative about en- gaging in chores and child care (Crouter, Perry-Jenkins, Hus- ton, & McHale, 1987; Russell & Radin, 1983).

Behind these statements is the notion that women, men, or both may want women to maintain primary responsibility for running the home. A woman may derive a sense ofself-worth as a wife from doing so, and a husband's self-evaluation may bene- fit as well. Although equivocal findings have emerged on the association between wife's occupational achievement and mar- ital satisfaction (Hiller & Philliber, 1982; Philliber & Hiller, 1983; Richardson, 1979; Simpson & England, 1981), there is considerable evidence that men find marriage to an "over-edu- cated" wife to be stressful (Hornung & McCullough, 1981) and that a wife's higher earnings relative to her husband are related to marital dissatisfaction in both spouses, threat to stability ofa marriage, and divorce (Burker & Loughlin, 1980; Cherlin, 1979; OJ,\mico, 1983;Pleck, 1978; Udry,1981). In this study, we exam- ined the association between child-care involvement and per- ceptions of one's performance as a parent and spouse. On the basis of the research described above, we hypothesized that partners in dual-career marriages would report the highest lev- els of self-satisfaction when traditional sex roles were main- tained in the home.

Method

The data reported here are part of a larger longitudinal research project investigating role conflict, stress, and coping among women professionals with young children (for more detail on methodology, see Emmons, Biernat, Tiedje, Lang, & Wortman, 1990; and for reports on other aspects of the studi,; see Marshall & Lang, 1990, on optimism, masteri,; and depression; Reifman, Biernat, & Lang, in press, on stress and social support; Tiedje et al., 1990, on perceptions of role compati- bility; and Wortman, Biernat, & Lang, in press, on occupational stress).

Subjects

Respondents were 139 women professionals with preschool children and their husbands. 1 The women were selected from two broadly de- fined occupational fields: university professor and businesswoman. We chose to study these two groups ofwomen for several reasons. First, academia and business are professional fields that seem to attract sub- stantial numbers of women and demand relatively high levels ofeduca- tion and commitment. Second, the occupational groups differ in the types of job stress commonly experienced (e.g, academics have more flexible work schedules but are less able to leave their work at the office, businesswomen are more likely to work under the pressure of strict deadlines). One ofour goals in the broader study was to compare the differential impact ofthese various stresses on women's home lives. The professors were identified from lists of instructional faculty at the four largest universities in Michigan and the two largest universities in the Chicago area. The businesswomen were selected from companies in geographic proximity to the universities (in a five-county area in Michigan and in Cook County, Jllinois) that had one of four Standard Industrial Classification (SIC) codes. The SIC codes corresponded to four industries that tend to attract female professional employees: ac- counting, advertising, banking, and law.

All respondents were identified through their employers. To be eligi- ble for the studi,; the women had to be working at least 30 hr per week, be married and currently living with their husbands, and have at least one child between the ages of I and 5 years. Every university we con- tacted agreed to cooperate with the study and did so by providing us with lists ofall female instructional facuhy. Telephone screening inter- views with departmental administrators were conducted to determine the eligibility (in terms of marital and parenting status) ofevery woman on each list. A total of 116 eligible female university professors were identified, of which 108 (93%) agreed to participate.

To locate eligible businesswomen, we first identified all the compa- nies that met our target SIC codes and were in the geographic areas described above and that also had at least 50 employees. We stratified our sampling frame into small, medium, and large companies, and then randomly selected from within these strata to obtain an even distribution of participants across the four types of professions (ac- counting, advertising, banking, and law) and three sizes ofcompanies. A total of 104 companies were contacted-72 in Michigan and 32 in Cook County. Sixty-eight of the Michigan companies (94%) and 26 of the Chicago-area companies (81 %) agreed to cooperate.

Most of the companies were reluctant to release the names of their employees, so the women could not be contacted directly. Instead, we relied on the personnel director in each company to identify all profes- sional women in the company who fit the eligibility criteria for the study. In 70% of the companies, the personnel director was willing and able to identify such women (usually there were only a few). After ob- taining the women's permission to be contacted by us, these personnel directors released the eligible women's names. Personnel directors in- dicated that nearly every eligible woman agreed to be contacted. In those cases in which the personnel director preferred not to identify eligible respondents, he or she was asked, instead, to distribute letters to all professional female employees. The letter described the study and contained a reply postcard on which the woman was to indicate her eligibility for and interest in participating in the study. Follow-up letters and postcards were sent until we obtained a high response rate (at least 90%) from each company. Using these procedures, 98 eligible businesswomen were identified; 93 of these agreed to participate. As- suming that the personnel directors did a thorough job ofscreening for eligibility and correcting for potential nonidentification of eligible

1 The entire sample included 200 women, but we are analyzing data only from those women whose husbands also participated (n = 139).

847 HOME LIVES OF WOMEN PROFESSIONALS

women because of unreturned postcards, our response rate was ap- proximately 85.5% of all eligible businesswomen in the 94 participat- ing companies.

The analyses we report in this article are based on those cases in which both the woman and her husband agreed to participate (70% of the total sample). This subsample included 68 university professors and 71 businesswomen. We compared women whose husbands did and did not participate in the study on both demographic characteris- tics and the variables of interest in this article. Out of more than 50 comparisons, the only variable significantly differentiating the two groups was job tenure-women whose husbands refused to participate in the study had been in their current jobs significantly longer (M = 5.4 years) than women whose husbands agreed to participate (M = 4.2 years), 1(198) = 2.54, p < .05.

Procedure

The larger study consisted of four structured interviews with the wife and two self-report surveys from her husband over the course ofa year. The first interviews and surveys took place between January 1986 and February 1987; each couple's participation ended after I year. The data discussed here come primarily from Wave I ofthe study, although several analyses focus on Wave 4 data. For convenience in this article, we refer to the Wave I and Wave 4 interviews (which were approxi- mately I year apart) as Time I and Time 2.

Each woman was interviewed, in person, hy a trained interviewer who came to either the home or the office ofthe respondent (depending on her preference). The interviews at Times I and 2 averaged about 2 hr each (the two intervening interviews were conducted by telephone and were much shorte~. Before the face-to-face interview, each respondent received a questionnaire that was to be completed and returned to the interviewer at the time of the interview. Along with this preinterview questionnaire, each respondent received a spouse questionnaire for her husband to complete. This spouse questionnaire included instructions to the husband to complete the instrument on his own and to mail it directly hack to the research site. The women's face-to-face interviews also included periodic "stops" during which the wife was asked to answer questions in a response booklet on her own. This was done to maximize the amount ofdata collected while minimizing time involve- ment and to make the women feel comfortable about answering rela- tively sensitive questions. We note the exact source of the women's data (interview or questionnaire) as we describe our measures belo\>:

Measures

The interviews covered a wide variety of topics ranging from basic demographics; to stresses in the work, marriage, and parenting do- mains; to coping strategies; and to personality dispositions, among others. Those questions and measures to which we refer specifically in this article are described belO\v.

Child-care tasks. We asked each spouse, "When your (child is/chil- dren are) at home with you and your husband, who . . . ." This stem was followed by eight specific child-care tasks: caring for physical needs, teaching skills and things about the world, getting up during the night, playing interactively, staying home with a sick child, responding to requests, making alternative child-care arrangements, and driving children to day care and other activities. These items were developed by our research staff, but are based on work by Hill and Stafford (1980). Responses ranged from my spouse does much more than me (I) to I do much more than my spouse (5). These questions were part ofthe wife's preinterview questionnaire. Instructions clearly indicated that couples should describe child-care involvement during times when no paid child-care provider was present. Thus, these data only describe child-care involvemem on occasions when parents were present.

Household chores. The wife, but not the husband, was asked in her preinterview questionnaire,

For each household chore listed below, please indicate how much of the time that chore is done hy you, and by your husband, by circling the appropriate number in each column.[2] Circle one number in every column for each chore. Use the following key to select a number: I = never or almost never, 2 = less than half the time, 3 =about halfthe time, 4 = more than halfthe time, 5 =all or most of the time.

The tasks were grocery shopping, cooking, doing laundry, house clean- ing, managing household finances, and making repairs and doing home maintenance. The wife made 12 ratings in all-one for wife's and one for husband's involvement in each of the six chores. As a follow-up to this series of questions, the women were also asked, "Now, please indicate who has responsibility for seeing to it that each of the tasks below get done." Respondents rated responsibility for each ofthe above chores on scales ranging from wife much more than husband (I) to husband much more than wife (5).

To assess women's general satisfaction with their husbands' involve- ment in running the home, the interviewers asked, "Overall, how would you describe the contribution your husband makes toward household chores and child care?" Responses could range from my husband does much too little (1) to my husband does a satisfactory amount (3) to my husband does much /CO much (5). Men were asked about their general satisfaction with their wives' home involvement as well. For them, however, we asked separate questions on wife's contri- bution toward household chores and child care.

Role performance evaluation. Women and men were both asked a series of questions assessing their sense of"how well'' they were per- forming as spouses and parents (for women, these questions were part ofthe interview). Three comparable questions were asked about perfor- mance in each role: (a) "Overall, you feel that you are a very good (spouse/parent)," with responses made on a scale ranging from strongly disagree (I) to strongly agree (5); (b) "You often feel guilty about not being a better (spouse/parent);' using the same response scale; and (c) "At this point in your life, how often, if ever, do you feel as though you're not the kind of (spouse/parent) you would like to be," with re- sponses ranging from never or almost never (I) to always or almost always (5). The former two questions were adapted from the Parenting Stress Index (Ahidin, 1983), and the last question was adapted from Baruch, Barnett, and Rivers (1983).

In addition to rating themselves on these questions, husbands also were asked to rate their wives on three of the six items. Specifically, husbands rated their wives on Question a(above) in regard to her perfor- mance as a parent and on Question c in regard to her performance as a spouse and as a mother. The only question on which we asked the wife to rate her husband was Question a (above) as applied to his parenting performance. Some of our analyses refer to these items individually; others are based on the relevant three-item indexes (i.e., evaluation of performance as a parent and as a spouse for each spouse, and an overall evaluation of the wife by the husband).

Satisfaction in life roles. Spouses rated their own sense ofsatisfaction in their jobs, in their marriages, and as parents. Wives completed these measures during the intervie\>: Specifically, respondents were asked to "think about all the pleasures and problems that go into daily life (on the job/with your spouse/as a parent). Taking everything into consider- ation, during the past month, how often have you . . . ." For the women, six items followed; these assessed how often they had felt "happy;• usatisfied," ••frustrated or angry," "insecuret and "unhappy"

2 Women were also asked to indicate how often "someone else" per- formed these tasks. These data are not considered here because we are most concerned with the distribution of chores between spouses.

848 MONICA BIERNAT AND CAMILLE B. WORTMAN

and how often they had "enjoyed" themselves in the relevant life role. For the men, we dropped two of the items: "unhappy" and "enjoyed."' All of the items were answered on scales ranging from never or almost never (I) to always or almost always (5). To create indexes of satisfac- tion, we first created separate averages on the negative and positive items, then subtracted the mean negative rating from the mean positive rating. Satisfaction refers, then, to the difference between positive and negative aspects of a life role. These items are based on Pearlin and Schooler'• (1978) work.

Problems around the home. We asked each spouse (by means of a questionnaire) to tell us how often he or she experienced a set of com- mon "problems" having to do with home life. Those problems were as follows: the house not being as clean as you would like it to be, big projects ( like cleaning out the garage, attic, basement, or closets) not being done, little projects and errands (like taking clothes to the cleaner or getting things repaired) not being done, meals not being well balanced, being too tired to enjoy time with the family, not having time to manage family finances well, and having to worry about coordinat- ing schedules with your spouse so that child-care responsibilities are taken care of. The response scale for each item ran from never(!) toa/1 the time (5 ).

Job involvement. To assess respondents' degree of commitment to or sense ofinvolvement in their jobs, the questionnaires contained two questions developed by Lodahl and Kejner (I 965): "The most impor- tant things that happen to me involve my job" and "I live, eat, and breathe my job." Both wives and husbands responded to these ques- tions on a scale ranging from strongly disagree (I) to strongly agree (5). Our measure of job involvement is based on the average response to these two items.

Balancing roles. Each spouse answered this question about the wife: "Compared to other women in your/her situation, how well would you say you/your wife balance(s) the demands of job, marriage, and fam- ily?" Responses ranged from much worse than other women (I) to much better than other women (5). This question was asked of women during the face-to-face interviews.

Results

Sample Demographics

In Table I , we present the demographic characteristics of the women and men in our sample, stratified by woman's occupa- tion. Given that we chose our women from professional occu- pations, it is not surprising that they were more educated, on average, than their husbands and that their husbands were more likely to be in nonprofessional occupations. Husbands, none- theless, were older and earned higher salaries than did their wives. As for differences between the professions, academic women (and their husbands) were significantly older, more edu- cated, and wealthier than their business sample counterparts. Academic women had also been married longer than business- women, and husbands ofacademics were more likely than hus- bands of businesswomen to be in professional/technical occu- pations and less likely to be in clerical/sales or managerial posi- tions (all ps < .05).

The last two rows in this table are also of interest. Here, we compare the levels of psychological job involvement of women and men from each occupational sample, as well as their aver- age hours at work per week.4 Businesswomen reported spend- ing significantly more time working than any other group, yet academic women were significantly higher in job involvement than their longer working business counterparts (ps < .05). In

general, we characterize our sample as one in which (a) wives and husbands have relatively equal job status (although there clearly is some degree of variability in this regard) and (b) sev- eral demographic differences between the two occupational groups exist.

In subsequent analyses, we report whether and when differ- ential findings appear in our academic and business samples. Because of this analysis plan, the variables that appear in Table I were also examined to test the assumption of equal variance in the samples (academics vs. businesswomen, wives vs. hus- bands, academic wives vs. academic husbands, etcJ. This as- sumption was violated for three variables: (a) variability in edu- cation was higher for businesswomen (SD = 1.26) than for aca- demic women (SD= 0.12), Cochran's C(2, 136) = .99, p < .0001 and for husbands of businesswomen (SD = 1.35) than for hus- bands ofacademics(SD= 0.93), C(2, 136)= .68, p< .0001; (b) variance in income was greater among husbands in general (SD= 29,512)thanamongwives(SD= 21,706), C(2, 275)= .65, p < .000 I; and (c) academic women showed more variability in job hours (SD= 13.95) than did businesswomen (SD= 6.38), C(2, J36) = .83, p < .000 I. We address the implications of these variance differences when relevant for other analyses. The as- sumption ofequal variance in the samples was upheld for every other variable discussed in this article.

Child Care and Housework

Our first aim was to assess how child-care and household chore responsibilities are distributed in these dual-career marriages. We conducted comparisons ofthe wife and husband responses to the eight questions about child-care task involve- ment by means of profile analysis (Morrison, 1967). A profile analysis is roughly comparable to a repeated measures analysis of variance; it performs a comparison of the mean profiles in each stratum and tests the null hypotheses that (a) profiles of the dependent variables are parallel, (b) the profiles are horizon- tal (i.e., there are no variable effects), and (c) the profiles are at the same level (i.e., there are no stratum effects). In this analysis, all three tests were significant, as is demonstrated in Figure I. The profiles were not parallel, F(7, 222) = 31.00, p < .000 I, and there were significant variable effects, F(7, 222) = 2.33, p < .05. The latter effect indicates that husbands and wives endorsed

'This was done, in part, to reduce the length of the husband's ques- tionnaire. The items we chose to drop were those having the lowest inter-item and item-to-total correlations in a pilot study. To test whether the different compositions of the wife and husband satisfac- tion scales were responsible for our results, we also created wife scales, using only the four items that appeared in the husband questionnaires. The patterns ofresults were very comparable regardless of whether the wife's satisfaction indexes contained six or four items.

• We measured hours at work differently for men and women. Hus- bands were asked: "In a typical week, how much time would you say you spend on work or career-related activities? Please include time spent on professional work both on the job and at home and include time spent during the day, in the evening, and on weekends." Wives, on the other hand, completed a complex time diary for the week prior to their face-to-face interviews. This diary included the category "work or career-related activities, including time spent on professional work both on the job and at home."

849 HOME LIVES OF WOMEN PROFESSIONALS

Table I Characteristics ofthe Sample

Attribute

Age College degree (%)' Graduate degree (%) Median income Median family income Race (% White) Years married No. of children(%)

l 2 3 4 5

Husband's occupation(%) Professional/technical Manager/proprietor Self-employed/business Clerical/sales Craftsmen Laborer/service Operative

Weekly work hours Job involvementb

Academic women sample Businesswomen sample

Wives Husbands Wives Husbands

37.53 39.44 32.62 35.10 100 92.5 70.0 62.0 98.5 90.9 35.7 46.4 $37,500 $40,000 $32,500 $37,500

$77,500 $67,500 89.7 89.7 93.0 93.0

9.42 7.39

49.1 54.9 36.8 28.2 12.3 8.5 1.8 4.2

4.2

70.1 31.8 16.4 25.8 4.5 4.5 7.5 22.7 1.5 7.6

4.5 3.0

46.28 49.94 55.85 49.14 2.64 2.53 2.34 2.28

• Some percentages may appear anomalous given sample sizes of 68 (academics) and 71 (businesswomen). Such anomalies are due to missing data values, which are for 0-5 cases in all descriptions. b High numbers indicate greater involvement on a 1-5 scale.

sharing some child-care tasks (e.g., playing interactively) more equally than they did others (e.g., arranging child care). FinaliY, a large effect of gender (husband vs. wife) was apparent, F(l, 228) = 198.24, p< .0001, such that wives were more involved in child care than were their husbands. In fact, on every child-care task wives indicated that they were more involved than their husbands, who also reported being less involved than their wives. Subsequent to the profile analysis, we performed corre- lated t tests on each child-care item separately and found that wives were more involved than their husbands on every task except playing (all ps < .0 I). These data suggest not only that women play the primary role in child care, but also that the spouses agree that this is so. The overall correlation between husband's and wife's ratings of who performed child-care tasks was a robust . 7 5 ( p < .000 I ). Additional analyses performed separately for academics and businesswomen revealed the same patterns of results within each occupational group.

The child-care duties in our list ranged from the unpleasant or mundane (getting up during the night) to the pleasant or fun (playing). It is interesting that the only task on which there was no significant difference between spouses in perceived amount of involvement came from the "fun" category-interactive pla}: The largest differences between the spouses were found, in- stead, on those tasks that are less enjoyable, that is, making alternative child-care arrangements and caring for physical needs. This suggests that husbands' involvement in child care, when it occurs, may not be geared toward the mundane but necessary aspects of parenting; these are shouldered primarily by wives.

Regarding household chores, wives' responses to the six rele- vant questions are displayed in Figure 2A. s A series of corre- lated t tests indicated that wives were significantly more in- volved than their husbands in managing finances and in house- cleaning (both ps < .00 I), and marginally so in cooking ( p < .08). Two other household chores-laundry and repairs-ap- peared to be performed significantly more frequently by hus- bands than by wives (ps < .0001), whereas grocery shopping was a shared chore. These findings remained significant after using the Bonferroni method to correct for number ofcompari- sons made l05/6 comparisons= .008 as the criterion). In sepa- rate analyses for the two occupational groups, the only differen- tial finding was that academic women and their husbands did not significantly differ in their contributions to managing fi- nances (Ms= 1.92 and 1.66, respectively, p > .20).

These data suggest that there is greater equity in division of

' Unfortunately, because our time with the husbands was limited, we were unable to include these household chore items in their question- naires. The child-<:are data do, however, inspire some confidence that the spouses agree on who performs such objective tasks. We did ask the husbands to globally compare "the amount of time that your wife spends on home chores, per week, to the amount of time that you spend," with the added note that "this includes things like home re- pairs and car maintenance too." The largest response category was "husband and wife spend about the same amount of time," with 39.6% of the husbands so indicating this. Thirty-six percent indicated that they spent less time than their wives did, and 24.5% said they spent more time than their wives did.

850 MONICA BIERNAT AND CAMILLE B. WORTMAN

5

■ HUSBANDS • 1m1ES

t-~.,o~~~~f,.v,11"'~~~ ,r>1

Child Care Taab

Figure/. Division ofchild-care duties between wives and husbands. (I = spouse does more than me. 3 -= bOlh the same, S ""' l do more)

household chores than in division of cbild-care, even though our household data come solely from the wife's report. We did, however, follow up the chore questions by asking, ..who bas responsibility for seeing to it that each of the tasks gets done?" Wives· mean responses to these items are graphically displayed in Figute 28. It is clear that on every chore except repairs, wives had significantly gm,.ter responsibility than did their husbands (all the means were sigrufi.cantly different from the ..equity" point of3, using a p < .008 criterion). This was true in both the academic and the business samples. Thus, although each spouse may have been the primary performer ofdifferent house- hold chores, the wife c-0ntinue(j to be primarily responsible for

A 19 5

4

3 ••••••••••••••••••

2 2

1111 HUSBANDS ■ W(VE.S

1

~~\\~~o,;,?\~t,o~;\lo~G(c.~"'1 ~

Household Cbore8

Figurt 2. Division ofhousehold chocn (A-) and responsibility for household chores ( B). (A: I =- never, 3 • abouL halftlui time, 5 =- aJJ or most ofJhe time. B: I - wife has more rtSponsibiliry, 3 = both the same, S = husband has more resf}On.ribilityJ

seeing that the chores got done. Only on the traditionally .. mas- cuJjoe"' repair and maintenance task was this pattern reversed.

In termsoftheirsatisfaction with husband's performance, the majority of the women (62.4 %) indicated that their husbands did a satisfactory amount, and 13. 7% actually indicated that they did too much, with the remaining 24% respondin8 that husbands did too little. Academic and businesswomen did not differ in their satisfaction with husbands' performance (I< I). In general, and in both occupatiooal groups, women were more positive about their husbands' contnoutions (M - 2.89) than husbands e:tpeeted they would be (M = 2.67), 1(117) = 3.52, P < .OOl.

Regarding child care. th.e vast m~ority of the husbands be- lieved their wives did a satisfactory amount (18.8%) or too much (18.6%). A somewhat smaller majority indicated that theirwives did a satisfactory amount of (59.4%) or too much household work (19.5%). Husbands of businesswomen were more satisfied with their wives' contribution to household chores (M =- 3.11) than were husbands ofacademics(M = 2.67), 1(136) • 3.08, p < .01, but these husbands did not differ in their satisfaction with wives' contributions to child care.

lo sum, both partners reported oonsiderablesa.tisfaction with their spouses' cbild~are and household chore involvement On every relevant question, the "satisfactory" and "too much" re- sponse options were used by at least 75% of the sample. Corre- lated l tests did, however, indicate somewhat higher satisfaction in the husband with his wife's involvement in child care, t(l 17) = 3.53, p < .OOL, and household chores, l(l 16) • 2.86, p < .Ol , than in the wife with her husband's involvement. ·

R.cle ENaluation

A separate, though related, issue of intenst is how spouses evaluate their own performances in their home roles. Husband

851 HOME LIVES Of' WOMEN PROFESSIONALS

and wife self-assessments of performance in their spouse and parent roles were analyzed with profile analysis. The test for parallelism of the profiles was significant, F(5, 259) = 8.48, p < .000 I , as was the test ofequality ofvariable means F(,5, 259) = 38.72, p < .0001. In other words, the six evaluative questions were endorsed at different levels by women and men. Most important for our purposes was the significant effect of gender -husband versus wife, F(l, 263),., 13.63, p < .00 1. Ali indi- cated in Figure 3, in live of six cases women evaluated them- selves more harshly than did men. That is, women were signin- can1ly more likely than men to indicate that they were "not the kind of" spouses and parents they would lllce to be, that they felt guilty about not being better spous~ and parents, and that they less often felt they were good spouses. Each ofthC$e effects was statistically significant aocording to correlated , tests per- formed after the profile analysis (all at p < .0 I). When separate profile analyses were conducted for academics and business- women, the same pattern ofeffects emerged.

Husbands were also asked to rate their wives on three ofthe above questions. We were able, then, to compare wives' ratings of themselves with husband's ratings of wives. These compari- sons appear in the firs t three bar pairs in figure 4 and suggest that wives were mote critical of themselves than were their husbands. Husbands perceived their wives as betterspouses and parents than their wives perceived themselves, and they were also more positive about their wives' ability to balance roles than were their wives. To statistically test these apparent differ- ences between husbands' and wives' ratings ofth.e wife, we once again pciformcd a profile analysis, as described earlier, fol• lowed by a serie,5 of cortelated t tests. The tests for parallelism of profiles, F(3, 254) = 17.90, and for equality of variable means, F(3, 254) = 22.69, were both significant, as was the key test for spouse differences, F<.I, 256) -= 58.34, all ps < .0001. Thus, although different evaluative qur.stions prompted differ- ent levels ofendorsement from each spouse, the m05t striking finding from this a nd th.e previous profile analysis is that wives

6l HUSBANDS ■ IHVES

Figure 3. ~If-evaluations of role performall<!e by wives and hus- bands. (1 = poor se{f-evalua1ion, S .. good se(fevaJuatum)

li!l HUSB,\NDS ■ WIVES

Figure 4. Evaluations oftbe wife by both spouses. (I .. poor evaJua• tion ofwife, S = good evaluation ofwife)

were generaUy more critical of themselves than husbands were ofthemselves and than husbands were ofwives. The single-item analyses (correlated t tests) supported this finding on every eval- uative question at p < .05 or better. F\Jrtbennorc, this patternof findings was noted in both the academic and business samples.

Jt would be helpful to rule out the possibility that women are simply more negative, critical, or willing to express negative feelingS in general than are men. We asked only one question that allowed us to compare both spouses' ratings of the hus- band: "Overall, you feel that you are [your husband is) a very good parent." Whereas husbands rated themselves quite highly on this question (M = 3.27), wives rated their husbands even higher (M = 3.4), t(l 30) = 2.33, p < .05. Furthermore, oo mea- sures of satisfaction in the marriage, parenting, and job do- mains, we found that wives and husbands were equivalent in the first two and that wives were significantly more satisfied in the jobdomain (Ms ... 1.37 and 0.96, for wives and husbands, l'CSJ)CC· tively), t(l 14) = 2.60, p < .05. These findings were clear in both occupational samples. In sum, it does not seem to be the case that the women were more critical or negative than were the men. Rather. the relatively negative evaluations appeared on items concerning the self.

We did, however, find evidence ofwomen's apparent applica- tion of harsh self-standards in one other domain: their evalua- tions of their homes. Both spouses answered a series of ques- tions about their concerns with various problems around the house; these average ratings are displayed in Figure 5. A profile analysis indicated that the profiles were not parallel, F(6, 226) =13.15; that the variables were not endorsed at equal lev- els, that is, concern was higher for some problems than for others, F(6, 226) -= 54. 19; and that spouses gave significantly different responses, F(l, 231)= 107.07;allps<.0001. Compara- ble results were obtained in the separate profile analyses of academic and business oouples. In genera], wives reported I» ing significantly more bothered by home problems than did husbands; in the simple-effects analyses, this difference was

852 MONICA BIERNAT AND CAMILLE B. WORTMAN

E.. 4 " (J

0 t.> .. .. 0

t- 3

~ ... 0

l!,I H USBANDS ■ WI VES

'fc~" 'fc"' cw"'.1' ,:,t<'t. \,\e&'-" •'.l'"v, pc~ OW, ytS~ ~o""se ~oo G~ cc,ot4' f\1'"

Problems a l Home

Figure S. Levels ofconcern about problems around the home by both SJ)Ouses. (I ., ne.,er bolhered, S • always bolhered')

clear on five ofthe seven problems ( ps < .000 I). Wives' applica- tion of relatively harsh standards in evaluating how things are running at home seems to parallel their evaluations ofself.

Accounting for Palterns ofChild-Care lnvolvemem

The above data indicate to us that our relatively equal-status dual-career couples' marriages are characterized by consider- able inequity, particularly in regard to child care and responsi• bility for home chores. Women are more responsible for run- ning the home than are men and invest more time and effort in child care, but nonetheless evaluate themselves less favorably than do men on role performance. Our next goal was to exam- ine some plausible antecedents of this pattern. In this section we look at three sets ofanalyses, attempting to separately pre- dict (a) wife's child-care involvement, (b) husband's child-care involvement, and (c) the difference between the two (wife's in- volvement relative to her husband's). 6

Child arid work traditionality ofmother. We expected that individuals from less traditional upbringings (e.g., in which mothers worked outside the home) would be more li.kely than those from traditional upbringings to deviate from the pattern ofgreater wife involvement in home chores. Ourdata, however, provide no support for this hypothesis. Wives raised by em- ployed mothers (44% ofthe sample) did not differ on child-care involvement from wives raised by mothers in the traditional role (Ms• 3.54 and3.53, respectively), t(l 18) < l. Thesame held true for husbands raised by employed (54% of the sample) and t.raditionaJ mothers (Ms= 2.57 and 2.67, respectively), t(l 30) < 1. Similar additional analyses were perl'ormed OD the relative child-care index; we also compared cbild-care patterns ofcou- ples in which one or both partners had a mother who was the major family btudwinner with child-care patterns of couples in which this was not true. Once again, patterns of child-care involvement were completely unrelated to these variables. Un- fortunately, we have DO measure ofattitudinal traditionality in

care

our data set. This might prove more relevant to division of re- sponsibilities in the home (see Barucl1 & Barnett, 1986).

Child care and income. We expected that the greater a spouse's earnings, the less likely that spouse is to perform child- care duties. Although husbands have traditionally surpused their wives in earnings, this is not necessarily true in the present sample of highly educated, professionally employed women. lbis fact allows u s to examine whether the impactofinoomeon child-care involveroent is comparable for bothsexes. lo the first 9 rows of values in Table 2 we summarize the relationships between spouse's reported levels of child care and their in- comes. This table includes the relevant correlations for the sam- ple as a whole and for academic and business respondents sepa- rately. The more money husbands earned, the less likely they were to involve themselves in child care. There was also some tendency for wives, particularly businesswomen, to do more child care as husbands' income increased. Io the business sam- ple, earnings of wives were not directly associated with either partner's involvement, but the involvement in child care ofhus- bands of academics did, surprisingly, decrease as the money their wives earned increased. Also surprising was the fact that as their earnings increased, academic women performed more child care relative to their husbands.

These findings for1he academic sample are rather counterin- tuitive and led us to speculate that when women earn more mone~ they are able to hire outside help wilh chores, thus lessening their husbands' involvement. Such an explanation seemed unlike!)\ as we explicitly instructed respondents to an- swer the child-care items by referring only to times when the child or children were at home with the parents. Nonetheless, to test this account, we examined three sets of variables having to do with the employment ofoutside help and used them as co- variates in the correlational analyses. Those variables were (a) wives' responses to an interview question about whether they employed anyone on a regular basis to help with chores, (b) wives' average ratings ofhow much of the time "someone else" performed the various household chores descri~ earlier, and (c) wives' indicati-On of the mean number of hours per week their youngest child was in day-care situations. 1

Sixty percent of the academics and 55% of the business- women answered yes to the question ofemploying outside help on a regular basis; the mean level ofsomeone else doing house-

6 Our analyses io this section focus only on the child-care indexes because these data, unlike the household chore data, are based on each spo115e's self-report.

7 This measure asked women to indicate the number of days per week and hours per day each of their children was cared ror in nine different day.care situations. These included day-care centen, nursery schools, schools.. cooperative programs, own home with relative, own home with nonn:lative, other home with !'dative, other home with nonrelative (the modal response), 8Jld "other." We devdoped an index based on the youngest child because (a) most orour respondents had only one child, (b) the youngest child is the most likely to be in need of extensive structured day care arrangements while the parents worlc:, and (c) we had no way of assessing the overlap ln da}«arc hours of women with two or more chiJdrco aod therefore could not develop an appropriate summary measure.

853 HOME LIVES OF WOMEN PROFESSIONALS

Table 2 Correlations Between Aspects ofJobs and Child-Care Involvement

Level of child-care involvement Predictor variable

and sample Wife's Husband's Wife-husband

Wife's income Total -.04 -.03 .02

Academic .08 -.23** .24** Business -.14 .12 -.14

Husband's income Total .13 -.22••· .16*

Academic .12 -.26•· .20• Business .20• -.20• .17

Relative income (wife-husband) Total -.12 .19** -.12

Academic -.01 .07 .02 Business -.21•• -.2s•• -.29••

Wife's education Total -.20•• .15* -.1s••

Academic -.12 .06 -.JO Business -.20• .28•• -.23••

Husband's education Total -.26°" .16* -.20••

Academic -.13 .07 -.08 Business -.29" .25** -.24*•

Wife's job hours Total -.14 .12 -.14

Academic -.19* .Jo•• -.26** Business -.22* -.03 -.09

Husband's job hours Total .42..... -.42••··· .45••···

Academic .26** -.32*.. .32*** Business .56***** -.st••··· _57••···

Relative job hours (wife-husband)

Total -.42•-·· .34***** -.44••··· Academic -.36••· .45••·· -.46**** Business -.59..*** .33*** -.51"'****

Wife's perception of whose career takes precedence'

Total .22•• -.34***** .21••· Academic .02 -.27** .15 Business .33*** -.38*•• .35***

Husband's perception of whose career takes precedence•

Total .14 -.33**** .24** Academic -.01 -.26** .13 Business .23** -.39** .30••·

Note. N = 139. Academic n = 68; business n = 71. • This variable is coded wifes career takes precedence (l) to careers have equal precedence (3) to husbands career takes precedence (5). • p < .09. •• p < .05. • •• p < .01. ·- p < .001. ••••• p < .0001.

hold chores was 2.49 on a 5-point scale, with no difference between academics and businesswomen; and the mean num- ber of hours per week children were in day care was 41.36 hours for academics and 44.55 hours for businesswomen (a nonsig- nificant difference). All of the correlations between income and child-care levels were then recalculated three times, on each occasion partialing out the effects ofone of the employed help variables. The partial correlations were quite comparable to the bivariate correlations (e.g., the largest change was that the corre- lation of.24 in row 2 of Table 2 was reduced to .2 I), indicating

that the association between academic women's child-care in- volvement and income was not due to their ability to hire out- side help. In fact, we repeated the partial correlational analysis strategy for all of the variables in Table 2. In no case was the difference between a bivariate correlation and a partial correla- tion greater than .04.

Does income have implications for spouses' satisfaction with their role performances? In a series ofcorrelational analyses, we found that the greater the wife's earnings relative to her hus- band's, the worse the wife reported feeling about her perfor-

854 MONICA BIERNAT AND CAMILLE B. WORTMAN

mance as a spouse (r = - .20, p < .05) and, among academics, as a parent (r = -.25, p < .05). Furthermore, the greater the in- come disparity favoring the wife, the worse her husband felt about his performance as a spouse (r = - .27) and parent (r = -.25, both ps < .01). The monetary success ofa professionally employed wife thus seems to carry a price: the loss of perceived competence as a spouse in both partners. On the other hand, husbands, by earning more mone); are not only relieved of home duties (see Table 2), but also derive a sense of worth as spouses and parents. In fact, a husband's self-esteem in these roles was directly tied to his income-the more he earned, the better a spouse and parent he perceived himself to be (rs = .25 and .23, respectively, ps < .0 I). These relationships were sup- ported in both the academic and the business samples. Being a good breadwinner, which has long been associated with being a good husband and father, does not yet have the same implica- tions for the comparable female roles.

Child care and respondents' education. Education levels' of the spouses also showed some relationship to child-care distri- bution, as can be seen in rows 10-15 of Table 2. The more educated the wives, the less child care they performed, and the more did their husbands (particularly husbands of business- women). Similarl); with more education, wives performed less child care relative to their husbands. Husbands' higher levels of education were, however, related to lesser wife involvement and greater husband involvement in child care. It seems, then, that higher education among husbands has a liberalizing effect, making these men less traditional rather than less likely to par- ticipate in child care (as a ~status~ hypothesis would suggest). A strong caveat must be placed on these results, however. The lack of effects in the academic relative to the business sample is almost certainly due to the low variability in education in the former group.

Child care and time spent working. In rows 16-24 of Table 2 appear the correlations between work hours and child-care in- volvement. Wife's job hours were modestly related to child-care involvement, such that wives performed less and husbands more child care the more hours the wife worked. However, this pattern was primarily true ofacademics; husbands ofbusiness- women were unaffected by their wives' job hours. This differ- ential finding may be due to the greater variability in academic women's hours than in businesswomen's hours, although the comparable effects of wives' job hours on wives' child care in each sample suggests that this is probably not the case. Hus- bands' hours at work were strongly related, in both samples, to increases in wives' and decreases in husbands' child-care in- volvement. It seems that aspects of the husband's work life-his income and his hours-have a larger impact on home life than the comparable aspects of the wife's professional career.

Child care and perceptions of career precedence. Spouses were in strong agreement on the question of whose career takes precedence (r = .68), and the modal response was that both careers were equally important. We correlated perceptions of career precedence with levels of child-care involvement. As rows 25-30 of Table 2 indicate, when women (particularly businesswomen) believed that their husbands' careers took pre- cedence, they engaged in more child care and their husbands engaged in less. Men's perceptions showed a similar pattern. When husbands believed their own careers took precedence,

they engaged in less child care, but only wives of business- women engaged in more. We interpreted these results some- what conservative)); given the possibly circular relationship be- tween career precedence and child-care involvement: Infer- ences about career precedence may be made primarily from observations of spouses' contributions to child-care duties.

In related analyses unaffected by this possible circularit); we discovered that the perception that husband's career takes prece- dence is related to feeling better about one's performance as a spouse among both wives and husbands (both rs= .18, p < ,05). Analyses within each sample, however, revealed that academic women did not show this pattern (r = .05). In other words, businesswomen and husbands of both types of women profes- sionals were more satisfied that they were "good spouses" when they believed the husband's career took precedence.

Summary. We have uncovered some evidence that career- based status affects who does what in the home. Husband's in- come was associated with reductions in his child-care involve- ment, and higher education levels of both spouses were asso- ciated with nontraditional shifts in child-care responsibilities. Similarly, involvement at home was affected, to some extent, by sheer amount of time spent on the job, as well as by perceptions of career precedence.

To more cohesively present the variables associated with each spouse's child-care involvement, multiple regression models were developed in an effort to identify the best predictors from the list in Table 2. Specificalfy, we used a stepwise forward selection algorithm that allowed variables in the model at p < .05 and removed them at p > .10. We calculated separate mod- els predicting wife's child-care involvement, husband's child- care involvement, and relative involvement (wife's minus hus- band's), first for the sample as a whole, then separately for the academic and business samples. The results, which include the standardized regression coefficients and squared multiple correlation statistics at each step, appear in Table 3.

The most consistent predictor of child-care involvement (in every analysis) was husband's work hours: The more the hus- band worked per week, the less child care he performed and the more his wife did, both in absolute terms and relative to her husband. Academic women's work hours also affected child- care patterns in comparable ways, but this was not true for the businesswomen. Higher education ofthe husbands reduced the child-care involvement of businesswomen, but not ofacademic women. Effects of income also differed, depending on the sam- ple: Husband's income was positively related to the child-care involvement of businesswomen; wife's income was negatively related to the child-care involvement of academic husbands. Finall); a wife's perception that her husband's career took prece- dence resulted in less child care performed by academic hus- bands and more child care performed by business wives.

It seems, then, that the strongest correlates of child-care in- volvement var); depending on the wife's career. In the business sample, child-care involvement was primarily determined by aspects of the husband (his work hours, income, and educa-

• Education is a four-level variable coded as follows: I =not a college graduate, 2 = college graduate, 3 = some graduate training, and 4 - graduate degree.

855 HOME LIVES OF WOMEN PROFESSIONALS

Table 3 Multiple Regression Models Predicting Child-Care Involvement

Standardized Overall R2 at Sample and predictor beta weight F each step

W child-care involvement

Whole sample H job hours H education

.43..*** -.22...

12.60***** .18 .23'

W job hours -.23*** .28' Academics

Hjob hours .3 I••• 4.28** .09 W job hours -.28** .17

Business Hjob hours H education

.51***** -.21••

11.58***** .26 _33•

H income .29** .40'

H child-care involvement

Whole sample Hjob hours -.36•··· 9.81••·· .13 H career precedence• -.19** .16

Academics W career precedence• -.32•• 4.79** .IO W job hours .28** .18 Hjob hours -.29** .25 Wincome -.27** .32

Business Hjob hours -.40*** 10.96*** .16

W-H child-care involvement

Whole sample Hjob hours .45***** 9.1s•••• .20 W career precedence• .19** .23 H education -.19•• .26 W job hours -.19** .29

Academics Hjob hours _35••• 6.20••· .12 W job hours -.34** .23'

Business Hjob hours .51***** 9.20••·· .26 W career precedence• .23** .31 H education -.22• .35

Note. W ; wife's; H ; husband's. • R' at this step is significantly greater than R' at the preceding step ( p < .01 ). In models with 3 or more predictors, R 2 at the final step is always significantly greater than R 2 at the first step. • Career prece- dence indicates the relevant spouse's perception of whose career takes precedence, always coded on a scale ranging from wife~ career takes precedence (1) to husband! career takes precedence (5). • p<.09.•• p<.05 . ••• p<.01. •••• p<.001. ·-· p<.0001.

tion), but not by any aspects ofthe wife's career or status. Child- care in the academic sample was affected by job-related attrib- utes ofboth spouses, but the negative association between wife's income and husband's child-care involvement-a result that does not appear to be due to the ability to hire outside help at higher income levels-suggests that a wife's paycheck still does not bring the same privileges as does a husband's. In general, the balance ofchild-care duties may improve when a wife has a career that is higher in status than the husband's and when the husband is highly educated. However, we have also previously noted that couples are happiest about their performances as

spouses when it is perceived that the husband's career takes precedence, and that higher wife income is related to less satis- faction with spousal role performance.

Consequences ofChild-Care Involvement

Thus far, we have noted the inequitable division of responsi- bilities in the home, as well as some apparent contributors to that pattern. Our next question concerns the consequences of child-care involvement for the self-evaluations of each spouse. The more child care the wife reported doing, the better she evaluated her performance as a spouse (r = .19, p < .05) and, surprisingly, the worse her husband evaluated his performance as a spouse (r = - .20, p < .05). The husband's child-care involve- ment was unrelated to his own role-performance evaluation, but the more child care he reported doing, the less well his wife felt about her spousal performance (r = - .22, p < .0 I). Spouse's evaluations of themselves as parents were unrelated to their levels of child-care involvement.

The relationship between each spouse's role performance and child-care involvement was assessed longitudinally by correlating Time 1 child-care involvement with change in per- ceptions of role performance from Time I to Time 2 (child-care involvement was not measured at Time 2). In these analyses, perceptions of parental role performance were unaffected by child-care involvement; however, perceptions ofspousal perfor- mance did change over time depending on Time I child-care levels. Specifically, the more child care the husband reported doing at Time I, the more likely his self-perception of spousal performance was to drop from Time I to Time 2 (r = - .20, p < .05, n = 93). At the same time, a wife's esteem as a spouse increased with higher levels of husband involvement in child care at Time I (r = .18, p < .05, n = 129).

The converse pattern appeared when wife's child-care involve- ment was used as a predictor. That is, wives' perceptions of spousal performance decreased as a result of high child-care involvement at Time I (r = -.18, p < .05, n = 129), whereas husbands' self-perceptions tended to improve, though not signif- icantly so(r= .)2, p < .25, n = 93). In other words, each spouse's child-care involvement at Time 1 predicted a decrease in satis- faction with own spousal performance, but an increase in the satisfaction of the other spouse over time. Both the cross-sec- tional and longitudinal resuhs were comparable in the aca- demic and business samples. As a final note regarding these data, our sample of wives did not appear to receive credit from their husbands for being more involved in child care: Husbands' ratings of their wives were unrelated to wives' reported levels of child-care and housework involvement.

Discussion

One ofour goals in this article was to describe the home lives of married women with young children who are also engaged in professional careers. What makes our sample different from those most frequently studied is the relatively equal economic and professional status of the marriage partners. Despite this equality outside the home, we discovered substantial inequity within the home, particularly in the distribution of child-care tasks and the overall responsibility for seeing that tasks get

856 MONICA BIERNAT AND CAMILLE B. WORTMAN

done. Men did contribute significantly to a number of home chores, and their involvement in child care was quite high in the area of interactive play. But on none of the tasks were husbands the primary care providers. In short, husbands of profes- sionally employed women were performing some, but certainly not an equal share, of the duties that come with the "second shift" of running a home (Hochschild, 1989). It is possible, of course, that husbands would have appeared to be more involved in chores had additional "masculine" tasks been included in the checklist (e.g., yard work, automobile repairs, etc.). In this sense, our findings are somewhat limited by our method. However, many other studies of dual-career couples, each using slightly different measures of home chore involvement, have supported this pattern of inequity as well (Berardo et al., 1987; Weingar- ten, 1978; Yogev, 1981 ). The present results suggest it is also true of professional couples with young children.

In spite of this arrangement, our female respondents were generally satisfied with their husbands' contributions, and fur- thermore, they were relatively critical of their own perfor- mances as wives and mothers. They were also more bothered than husbands by problems in their homes. In fact, women were more self-critical than were men, and women's role perfor- mance was rated more highly by husbands than by the women themselves. It does not seem to be the case that women are more negative than men in general; they report, for example, equal or greater role satisfaction than do men, and they evaluate their husbands quite favorably. We suggest that women's harsh criticism is seen only in domains that have evaluative implica- tions for self. The literature on sex differences has suggested that although men and women do not reliably differ in self-es- teem (see Maccoby & Jacklin, 1974), women are more likely than men to (a) accept negative feedback about their perfor- mance (Eagly & Whitehead, 1972) and (b) make external or unstable attributions for their successes (e.g., "luck") and inter- nal or stable attributions for their failures (e.g., "low ability"; Deaux, 1976; Deaux & Farris, 1977). Comparable mechanisms may be responsible for the lower self-evaluations found in our female respondents.

We also suspect that women professionals invoke harsh stan- dards in judging themselves. When evaluating themselves as wives, homemakers, and mothers, they seem to use nonem- ployed wives and mothers or their own mothers as the stan- dards of comparison. We asked respondents an open-ended question: "What makes you feel as though you're not the kind of wife and/or mother you would like to be?" One woman re- sponded,

My sisters are very traditional, and my husband's family is super- traditional. They have wives who are not like me ... subcon- sciously it bothers me.

Other typical responses included the following:

Sometimes I feel like I should do more "wifely" things; like ifmy husband asks me to sew a button on his shirt it may take me a month to do it; things like cooking or making home more like a haven and more enjoyable. I know these things don't bother my husband, but I'm aware of them. My mother wants me home with my son full-time. I don't want to be. I feel guilt in two directions-for not doing what my mother wants and for not wanting to do it.

If only because of the amount of time spent at work, profes- sionally employed women simply cannot run homes, cook

meals, and raise children in the manner of their traditional counterparts. Nonetheless, the traditional standard seems to be evoked as women judge their performances as spouses and par- ents. When men evaluate themselves and when women evaluate men, the standard used may also be a traditional one. However, because the traditional man's involvement in the home is mini- mal, the husbands in our sample can quite easily surpass this standard (see Biernat, Manis, & Nelson, 1991 ).

We have identified a number of specific contributors to the division of child-care responsibilities between spouses. All of these findings, particularly those involving income and educa- tion, must be interpreted with caution given the rather re- stricted range of these variables in our sample of relatively equal-status marriage partners. Furthermore, differences in the variability of some indicators between academics and business- women may account, in part, for differential results in the two samples. The present data nonetheless indicate that a husband's time at work is the strongest and most consistent associate of child-care involvement of both spouses and of the wife's relative to her husband's. Others have reported that the more a father works, the less involved he is with his children (McHale & Hus- ton, 1984; Robinson, 1977). Our data suggest that husbands' work hours affect wives' child-care involvement as well: Women seem to compensate at home for the child-care time the hus- bands miss by being at work. In sum, husbands who worked many hours did less chi Id care and their wives did more, highly educated husbands did more child care and their wives did less, and spouses' perceptions that the husband's career took prece- dence meant that the husband did less child care and the wife did more. In the business sample, characteristics of the hus- band-his education and work hours-typically had a stronger influence on both spouses' child-care involvement than did comparable characteristics of the wife (see Table 3). In the aca- demic sample, however, women's work hours also contributed to child-care patterns. Findings consistent with this latter pat- tern were recently reported by Barnett and Baruch (1987). In that study, the number of hours middle-class mothers (occupa- tions not reported) worked per week was positively related to husband's involvement in child care. Our data suggest that this effect is specific to women in particular occupations, such as academics. More research is needed to determine when wives' work hours are likely to directly affect husbands' child-care performance. On the basis of exploratory analyses not reported here, the responsible factors do not seem to be women's job involvement or commitment, nor any interactions among the predictors listed in Table 2.

Interestingly, wives and husbands were happiest about their performances as spouses when the husband's career took prece- dence (business sample only) and when he earned more money than his wife (both samples). Such preferences seem to be based on traditional sex role stereotypes concerning the importance of the breadwinning, agentic role to men's sense of self and to perceptions of appropriate masculine behavior (see Eagly & Steffen, 1984; Vero ff et al. 1981 ). These results are also consis- tent with our earlier speculation that self-esteem benefits may accrue to women professionals who follow traditional sex role norms in some spheres of their lives. However, given the exten- sive demands of career, parenthood, and marriage, it would appear to be even more adaptive for women (and men) to revise their sex role ideologies to more accurately reflect the nontradi-

HOME LIVES OF WOMEN PROFESSIONALS 857

tional quality of their lives. This is admittedly a difficult task (see Tavris & Wade, 1984), but one hopes that the reality of changing gender roles will gradually come to be reflected in societal norms and values. That women academics experienced satisfaction when their husbands earned more money, but not when their husbands' careers took precedence, reflects some ambivalence about these traditional sex roles. Having invested more time and expense in formal education, and reporting more psychological involvement in their jobs than businesswo- men, academic women are the more likely group to buck tradi- tional expectations about the job status of women versus men.

A similar explanation may account for one other difference we noted between academic and businesswomen: how wife's income affected husband's child-care involvement (see Table 2). In the business sample, wife's income was unrelated to hus- band's participation in child care; in the academic sample, this relationship was unexpectedly negative. Why is it that when academic women earned more money, their husbands reported doing less child care? As the partial correlational analyses sug- gested that this finding was not due to the ability to hire outside help when wives earn more money, we can only offer the specu- lation that academic women are more ambivalent about tradi- tional sex role norms than are businesswomen. Academic women do not want their husbands' careers to take precedence over their own, but they are dissatisfied with their perfor- mances as spouses and parents when they earn more money than their husbands. One way to compensate for negative feel- ings resulting from the income disparity is to "absolve" the husband of his child-care responsibilities. Because business- women's incomes have no implications for child-care patterns, such compensation is unnecessary.

This notion of ambivalence about sex roles is somewhat analo- gous to the concept of racial ambivalence (Gaertner & Dovidio, 1986; Katz, Wackenhut, & Hass, 1986), which suggests that White Americans concurrently possess egalitarian values and negative perceptions of Blacks. The sex role analogy is that egalitarian beliefs coexist with norms learned early in life about the "appropriate" roles for women and men. In this study, we did not measure our respondents' sex role ideologies. However, on the basis of academic women's higher education levels and higher involvement in their jobs and our data suggesting that child-care patterns in the business sample are primarily deter- mined by the husband's and not the wife's life situation, we suspect that the women in our academic sample possess less traditional sex role ideologies than those in our business sam- ple. These nontraditional values, as already indicated, may con- flict with beliefs about the appropriate male breadwinner role, creating a state of ambivalence in which greater wife attain- ments must be tempered by conciliations to the husband. It would be valuable to explicitly question respondents about these issues in future studies.

When we examined the effects of child-care involvement on respondents' perceptions of their role performances, results were similar in both the academic and the business samples. It appears, then, that the main difference between the groups is in the factors that contribute to child-care patterns. Consequences of child-care involvement did differ, however, in the short and long terms and for husbands and wives. The more child care the wife performed, the more positive was her concurrent satisfac- tion with her spousal performance. She also felt less satisfied

with herself the more child care her husband performed. This is consistent with the hypothesis that professionally employed women may gain esteem as wives through involvement in tradi- tional home activities and with the evidence that mothers' life satisfaction may suffer with increases in husbands' participa- tion in parenting (Baruch & Barnett, 1986). Husbands' levels of child-care involvement did not directly affect their role perfor- mance perceptions, but husbands did experience some dissatis- faction concurrent with wives' greater involvement in child care. This may indicate that modern fathers' sense of responsi- bility as child-care providers is increasing, in contrast to earlier findings that dual-earner husbands are happiest with their marriages when they perceive that their wives are doing more than their fair share of family work (Yogev & Brett, 1985).

Our longitudinal results were somewhat different. Women became more satisfied with their performances as wives follow- ing greater husband involvement in child care at Time l. This finding differs from previous research reporting no decreases in wives' tensions and conflicts with higher paternal involve- ment in family work (Baruch & Barnett, 1986; Pleck, 1983). Our data suggest that wives may indeed benefit from husband involvement in child care, but that these benefits only surface with time. The longitudinal analyses also indicate that one's own level of child-care involvement predicts a decrease in one's self-rating on the spousal performance variable over time. The decrement in satisfaction among husbands is consistent with recent longitudinal findings by Crouter et al. (1987) that hus- bands in dual-earner marriages experience reductions in mari- tal happiness with high levels of child-care involvement (in part because they feel resentful about being coerced into these du- ties by their wives). The decrement among wives may reflect fatigue from performing as primary child-care provider while engaging in a demanding career, or resentment that coercion is a necessary step to prompt husband participation.

Before we make too much of these findings, however, we must note their limitations. First, it is curious that although levels of child-care involvement predicted satisfaction with spousal performance, they did not predict the more relevant evaluations of parental performance. Why this might be is un- clear. A case can readily be made for child-care involvement implicating one's spousal role, but such involvement should cer- tainly also implicate one's parenting role. Second, the longitu- dinal findings are somewhat limited in that we did not measure (and therefore could not control for) spouses' child-care involve- ment at Time 2. The analysis strategy used is still legitimate (essentially a partial correlation), but a more convincing argu- ment about causal direction could have been made with panel data. At minimum, these results should be interpreted with caution, although they can inform future research on the short- and long-term implications of extensive child-care involve- ment.

One additional issue concerning our child-care measure de- serves mention. Respondents were instructed to indicate their involvement in the various child-care tasks by considering only the times "when your (child is/children are) at home with you and your spouse." In other words, care of the child by a nonpar- ent is not considered in these estimates. Availability of such care did not moderate the relationships between child-care in- volvement and various job characteristics, but paid help may nonetheless play an important role in regard to the conse-

858 MONICA BIERNAT AND CAMILLE B. WORTMAN

quences of child-care involvement for spouses. In this study, respondents' children were in day care for an average of 43 hr per week, but the range was 9-93 hr. We suspect that the family with 9 hr of day care is quite different from the family with 93 hr, even if, on our parental child-care measures, both families report the same relative division of duties between spouses. In the former case, inequity leaves the wife with substantial respon- sibility and possibly negative effects on role performance satis- faction; in the latter case, it may not. To address this possibility, we regressed measures of role performance satisfaction on the interaction between hours per week of day care and the relative child-care involvement of the spouses. Results were nonsignifi- cant in all cases. Nonetheless, future work on the consequences of child-care and home chore involvement should incorporate complete measures of extraparental help, particularly when the financial resources of the sample allow for relatively heavy use of this category.

Conclusion

Our data argue for the importance of assessing the opinions and perspectives of both partners in any investigation of marriage: Findings based on one spouse's report are necessarily incomplete. We were able, for example, to document areas of agreement between spouses (e.g, on "who does what child care?"), as well as areas in which differences in perception were the norm (e.g., "how often is the wife not the kind of parent she should be?").9 An even broader net could be cast in the study of family relations by including an assessment of the children of these couples. Our data already indicate that events that may be good for one spouse are bad for the other. The more hours a husband works, for example, the less child-care he performs, but the wife must pick up the slack. Adding children to the equation would both complicate and complete the picture of these processes of family life.

We chose to study the home lives of married women profes- sionals with young children because their numbers are steadily growing and because their lives include a number of significant, simultaneous stressors. We have documented that equal status outside the home seems not to translate into equal sharing of responsibilities within the home and that women, more than men, are critical of their performances as spouses and parents. One could argue that our findings from a sample with relatively high socioeconomic status may not generalize to the larger pop- ulation of dual-earner couples. However, that inequity contin- ues to exist even in couples in which wives have demanding, nontraditional careers suggests that the processes by which home care and child care are decided is similar in these groups. Furthermore, we submit that nearly all women are socialized to follow standards set by traditional women and that both women and men are struggling with the social changes prompted by women's growing involvement in the labor force. In this sense, the patterns and implications of child-care involve-

• These examples suggest that differences in perception may be most likely to appearon measures involving subjective attitudes rather than on observations of objective behavior (performance of child-care duties).

ment we have discussed are relevant to the broader study of working couples.

References

Abidin, R.R. (1983). Parenting Stress lndex(PSI). Charlottesville: Uni- versity of Virginia Press.

Acock, A., & Edwards, J. (I 982). Egalitarian sex-role attitudes and fe- male income. Journal of Marriage and the Family. 44, 581-589.

Barnett, R. C., & Baruch, G. K. (I 987). Determinantsoffathers' partici- pation in family work. Journal of Marriage and the Family. 49, 29-40.

Baruch, G. K., & Barnett, R. C. (1986). Consequencesoffathers' partic- ipation in family work: Parents' role strain and well-being. Journal of Personality and Social Psychology, 51, 983-992.

Baruch, G., Barnett, R., & Rivers, C. (1983). Lifeprints: New patterns of love and work for todays women. New York: New American Library.

Behrman, D. L. (1982). Family and/or career plans of first time mothers. Ann Arbor, MI: UMI Research Press.

Beller, A. (1982). Occupational segregation by sex: Determinants and changes. Journal of Human Resources, 17, 371-392.

Berardo, D. H., Shehan, C. L., & Leslie, G. R. (1987). A residue of tradition: Jobs, careers, and spouses' time in housework. Journal of Marriage and the Family, 49, 381-390.

Berheide, C., Berk, S., & Berk, R. (1976). Household work in the sub- urbs: The job and its participants. Pacific Sociological Re>iew, l 9, 491-518.

Berk, R., & Berk, S. (1979). Labor and leisure at home. Beverly Hills, CA: Sage.

Berk, S. (1985). The gender foctory:Theapponionmentofworkin Ameri- can households. New York: Plenum Press.

Biernat, M., Manis, M., & Nelson, T. E. (1991). Stereotypes and stan- dards of judgment. Journal of Personality and Social Psychology, 60, 485-499.

Blood, 0., & Wolfe, D. M. (1960). Husbands and wives. New York: Free Press.

Bryson, R, Bryson, J., Licht, M, & Licht, B. (1976). The professional pair: Husband and wife psychologists. American Psychologist, 3 I, 10--16.

Burker, J., & Loughlin, J. (I 980, August). Female labor force panicipa- tion and divorce. Paper presented at the annual meeting of the Ameri- can Sociological Association, New York.

Cherlin, A. (1979). Worklife and marital dissolution. In G. Levinger & 0. C. Moles (EdsJ, Divorce and separation (pp. 151-166). New York: Basic Books.

Condran, J., & Bode, J. (1982). Rashomon, working wives, and family division of labor: Middletown, 1980. Journal of Marriage and the Family, 44, 421-426.

Crouter, A. C., Perry-Jenkins, M., Huston, T. L., & McHale, S. M. (1987). Processes underlying father involvement in dual-earner and single-earner families. Developmental Psychology, 23, 431-330.

Dl>.mico, R. (1983). Status maintenance or status competition? Wife's relative wages as a determinant of labor supply and marital instabil- ity. Social Forces. 61, 1186-1205.

Deaux, K. (1976). Sex: A perspective on the attribution process. In J. Harvey, W J. Ickes, & R. F. Kidd (Eds~. New directions in a11ribution research (Vol. I, pp. 335-352). Hillsdale, NJ: Erlbaum.

Deaux, K., & Farris, E. (1977). Attributing causes for one's own perfor- mance: The effects of sex, norms, and outcome. Journal of Research in Personality. 11, 59-72.

Eagly, A. H., & Steffen, V. J. (1984). Gender stereotypes stem from the distribution of women and men into social roles. Journal of Personal- ity and Social Psychology, 46, 735-754.

Eagly, A. H., & Whitehead, G. I. (1972). Effects of choice on receptivity

HOME LIVES OF WOMEN PROFESSIONALS 859

to favorable and unfavorable evaluations ofoneself. Journal of Person- ality and Social Psychology, 22. 223-230.

Emmons, C. A., Biernat, M" Tiedje, L. B., Lang, E. L., & Wortman, C. B. (1990). Stress, support, and coping among women profes- sionals with preschool children. In J. Eckenrode & S. Gore (Eds.), Stress between work and family (pp. 61-93). New York: Plenum Press.

England, P., & Farkas, G. (1986). Households, employment, and gender: A social, economic, and demographic view. Chicago: Aldine.

Eriksen, J. A., Yancey, W. L., & Eriksen, E. P. (1979). The division of family roles. Journal of Marriage and the Family, 41, 301-303.

Ferber, M. (1982). Low-market participation of young married women: Cause and effects. Journal of Marriage and the Family, 44, 457-468.

Ferber, M.A., & Birnbaum, B. G. (1977). The "new home economics": Retrospects and prospects. Journal of Consumer Research, 4, 19-28.

Firestone, J., & Shelton, B. A. (1988). An estimation of the effects of women's work on available leisure time. Journal of Family Issues, 9, 478-495.

Fraker, S. (1984, April 16). Why women aren1 getting to the top. For- tune, pp. 4-0-45.

Gaertner, S. L., & Dovidio, J. F. (1986). The aversive form of racism. In J. F. Dovidio & S. L. Gaertner (Edsj, Prejudice, discrimination. and racism (pp. 61-89). San Diego, CA: Academic Press.

Geerken, M., & Gove, W.R. (1983). At home and at work: The fami/ys a/location of labor. Beverly Hills, CA: Sage.

Gore, S., & Mangione, T. W. (1983). Social roles, sex roles, and psycho- logical distress: Additive and interactive models of sex differences. Journal of Health and Social Behavior, 24, 300-312.

Gove, W. R., & Geerken, M. R. (1977). The effect of children and em- ployment on the mental health of married men and women. Social Forces, 56, 66-76.

Hayghe, H. V. (I 990). Family members in the work force. Monthly La- bor Review 113, 14-19.

Hill, C. R., & Stalford, F. P. (1980). Parental care of children: Time diary estimates of quantity, predictability, and variety. Journal of Human Resources, 15, 219-239.

Hiller, D. V., & Philliber, W. W. (1982). Predicting marital and career success among dual worker couples. Journal of Marriage and the Family, 44. 347-355.

Hochschild, A. (1989). The second shift: Uvrking parents and the revolu- tion aJ home. New York: Viking Press.

Hornung, A., & McCullough, B. C. (1981 ). Status relationships in dual- employment marriages: Consequences for psychological well-being. Journal of Marriage and the Family, 43, 125-141.

Huber, J., & Spitze, G. (1983). Sex-stratificaJion: Children, housework, and jobs. San Diego, CA: Academic Press.

Hunt, J., & Hunt, L. (1982). Dual-career families: Vanguard of the future or residue of the past? In J. Aldous (Ed.), Two paychecks: Life in dual-earner families (pp. 41-60). Beverly Hills, CA: Sage.

Katz, I., Wackenhut, J., & Hass, R. G. (1986). Racial ambivalence, value duality, and behavior. In J. F. Dovidio & S. L. Gaertner (Edsj, Preju- dice, discrimination, and racism (pp. 35-59). San Diego, CA: Aca- demic Press.

Kessler, R. C., & McRae, J. A. (1981). Trends in the relationship be- tween sex and psychological distress: 1957 -1976. American Sociolog- ical Review, 46, 443-452.

Komarovsky, M. (1973). Cultural contradictions and sex roles: The masculine case. American Journal of Sociology, 78, 873-884.

Lodahl, T. M., & Kejner, M. (1965). The definition and measurement of job involvement. Journal of Applied Psychology, 49, 24-33.

Maccoby, E. E., & Jacklin, C. N. (1974). The psychology of sex differ- ences. Stanford, CA: Stanford University Press,

Marshall, G. N., & Lang, E. L. (1990). Optimism, self-mastery, and

depression among women professionals. Journal of Personality and Social Psychology, 59, 132-139.

Mason, K. 0., & Bumpass, L. L. (1975). U.S. women's sex-role ideology, 1970. American Journal of Sociology, 80, 1212-1219.

McHale, S. M., & Huston, T. L. (1984). Men and women as parents: Sex role orientations, employment, and parental roles with infants. Child Development, 55, 1349-1361.

Mirowsky, J" & Ross, C. E. (1987). Beliefin innate sex roles: Sex stratifi- cation vs. interpersonal influence in marriage. Journal of Marriage and the Family, 49, 527-540.

Model, S. (1982). Housework by husbands: Determinants and implica- tions. In J. Aldous (EdJ, Two paychecks: Life in dual-earner families (pp. 193-206). Beverly Hills, CA: Sage.

Morrison, D. F. (1967). Multivariate statistical methods. New York: McGraw-Hill.

Newcomb, T. M. (1943). Personality and social change. New York: Dry- den Press.

Oakley, A. (1974). The sociologyofhousework. London: Martin Robert- son.

Osmond, M. W., & Martin, P. Y. (1975). Sex and sexism: A comparison of male and female sex-role attitudes. Journal of Marriage and the Family, 37, 744-753.

Pearlin, L. I., & Schooler, C. (1978). The structure of coping. Journal of Health and Social Behavior, 19, 2-21.

Philliber, W., & Hiller, D. (1983). Relative occupational attainments of spouses and later changes in marriage and wife's work experience. Journal of Marriage and Lhe Family, 45, 161-170.

Pleck, J. H. (1978). The work-family role system. Social Problems, 24, 417-427.

Pleck, J. H. (1983). Husbands' paid work and family roles: Current research issues. In H. Lopata & J. Pleck (Eds.), Research in the inter- weave of social roles: Iv/. 3. Families and jobs (pp. 251-333). Green- wich, CT: JAi Press.

Rapoport, R., & Rapoport, R. (1971 ). Dual-career families. New York: Penguin Books.

Reifman, A., Biernat, M., & Lang, E. (in press). Stress, social support, and health in a sample of married professional women with small children. Psychology ofWomen Quarterly.

Richardson, J. G. (1979). Wife occupational superiority and marital troubles: An examination of the hypothesis. Journal of Marriage and Lhe Family, 41, 63-72.

Robinson, J. (1977). How Americans use time: A social-psychological analysis. New York: Praeger.

Robinson, J. (1980). Household changes in how people spend their time. In A. Michel (Ed.), Family issues of employed women in Europe and America (pp. 148-158). Leiden, The Netherlands: Brill.

Robinson, J.P., Yerby, J., Fieweger, M., & Somerick, N. (1977). Sex role differences in time use. Sex Roles, 3, 443-459.

Russell, G., & Radin, N. (1983). Increased paternal participation: The father's perspective. In M. Lamb & A. Sagi (Edst Fatherhood and family policy (pp. 139-167). Hillsdale, NJ: Erlbaum.

Simpson, I. H., & England, P. (1981 ). Conjugal work roles and marital solidarity. Journal of Family Issues, 2, 180-204.

Slocum, W., & Nye, F. I. (1976). Provider and housekeeper roles. In E I. Nye (Edj, Role structure and analysis of the family (pp. 1-30). Bev- erly Hills, CA: Sage.

Stafford, R., Backman, E., & Dibona, P. (1977). The division of labor among cohabitating and married couples. Journal of Marriage and the Family, 39, 43-57.

Staines, G. L., Pleck, J. H., Shepard, L. J., & O'Connor, P. (1978). Wives' employment status and marital adjustment: Yet another look. PsY- chology ofWomen Quarterly, 3, 90-120.

Tavris, C., & Wade, C. (1984). The Jonges/ war.· Sex differences in per- spective. San Diego, CA: Harcourt Brace Jovanovich.

860 MONICA BIERNAT AND CAMILLE B. WORTMAN

Thornton, A., Alwin, D. F., & Camburn, D. (1983). Causes and conse- quences of sex-role attitudes and attitude change. American Sociologi- cal Review, 48, 211-227.

Tiedje, L. B., Wortman, C. 8., Downey, G., Emmons, C. A., Biernat, M., & Lang, E. (1990). Multiple role women: Inter-role compatibility perceptions, role satisfaction, and mental health. Journal of Marriage and the Family. 52, 63-72.

UdrY, J. R. (1981 ). Marital alternatives and marital disruption. Journal of Marriage and the Family. 43, 889-897.

U.S. Bureau of the Census. (1988). Household and family characteris- tics: March 1987. Current Population Reporls, Series P.20, No. 424.

U.S. Bureau of Labor Statistics. (1989). Special labor force reporls (Re- port Nos. 13, 130, 134; Bulletins 2163, 2307). Washington, DC: Gov- ernment Printing Office.

U.S. Department of Labor (I 977). Working mothers and their children. Washington, DC: Employment Standards Administration, Women's Bureau.

Vanek, J. (1974). Time spent in housework. Scientific American, 231, 116-120.

Vero ff, J., Douvan, E., & Kulka, R. A. (1981 ). The inner American: A self-portrait from 1957 to 1976. New York: Basic Books.

Weingarten, K. (1978). The employment patterns of professional cou-

pies and their distribution of involvement in the familY, Psychology ofWomen Quarterly. 3, 43-52.

Welch, S., & Booth, A. (1977). The effect of employment on the health of married women and children. Sex Roles, 3. 385-397.

White, L. K., Booth, A., & Edwards, J. N. (1986). Children and marital happiness-Why the negative correlation? Jaumal of Family Issues, 7, 131-147.

Wortman, C. B., Biernat, M., & Lang, E. (in press). Coping with role overload. In M. Frankenhaueser (Ed), Women, work, and health. Chicago: University of Chicago Press.

Yogev, S. (1981 ). Do professional women have egalitarian marital rela- tionships? Journal of Marriage and the Family. 43, 865-871.

Yogev, S. (1987). Marital satisfaction and sex role perceptions among dual-earner couples. Journal of Social and Personal Relationships, 4, 35-45.

Yogev, S., & Brett, J. (1985). Perceptions of the division of housework and child care and marital satisfaction. Jaurna/ of Marriage and the Family. 47, 609-618.

Received April 9, 1990 Revision received December 11, 1990

Accepted December 16, 1990 ■