Sexuality and the Dual-earner Couple, Part II: Beyond the Baby Years.

Two patterns warrant the attention of those who study sexuality in close relationships: the large influx of women into the labor market over the past 3 decades, and the fact that a substantial minority of employees work considerably more than 40 hours per week. According to the U.S. Census Bureau (1998), 8.5% of Americans work 60 or more hours per week and an additional 11.7% work between 49 and 58 hours per week. Given these trends, some sex researchers and therapists have raised concerns about whether sexual expression will be dampened for dual-earner couples, and especially for those who work long hours (e.g., Sarrel & Sarrel, 1984). Despite these concerns, there is remarkably little empirical research addressing these questions. The purpose of the studies reported here is to provide empirical tests of the hypothesis that sexual expression suffers in dual-earner couples. The research is framed by two theoretical perspectives on the relationship between multiple roles and psychosocial functioning; to these we add a third theoretical perspective that insists on the importance of salience of identities.

Theoretical Frameworks

Two broad hypotheses have been advanced regarding the relationship between multiple roles and psychological functioning. One is the scarcity hypothesis or role-strain perspective (Baruch, Biener, & Barnett, 1987; Goode, 1960), which rests on the premise that humans have a fixed amount of time and energy. Adding a role (e.g., motherhood, worker) therefore creates stress and negative psychological consequences for the individual and her or his marital relationship, including the quality of sexual expression. Negative outcomes are especially likely when the demands associated with work and family roles are particularly high, for example, when a mother works full time or more than full time (Rogers, 1996).

The enhancement hypothesis, in contrast, argues that multiple roles enhance psychological and interpersonal functioning (Baruch et al., 1987; Coser, 1991; Marks, 1977; Thoits, 1983). According to this view, multiple roles provide people with additional resources that help them cope with the demands of and reduce stress associated with multiple obligations (Wethington & Kessler, 1989). Epstein (1987), for example, argued that two positive consequences of multiple roles are variety and buffering. Variety refers to the idea that multiple roles expose individuals to diverse situations that help them to take a broad perspective. Buffering occurs when the impact of stresses or dissatisfactions in one role can be buffered by success and satisfaction in another. More generally, Linville (1985, 1987) has argued that considering oneself in terms of a greater number of, and more differentiated, self-aspects buffers one from the effects of stressful life events. Other theorists have noted that employed women benefit from additional opportunities to build a sense of competence and from the additional social support available at work (e.g., Aneshensel, 1986). Employed women have more power within the family because of their earnings, and this enhanced power leads to decreased psychological symptoms such as depression (Rosenfield, 1989). Furthermore, the woman's economic contribution reduces financial strain for the family, improving marital quality (Barnett & Rivers, 1996).

Empirical research does not support the scarcity hypothesis when the outcome variable is psychological distress; in general, employment for women is associated with better mental health (e.g., Glass & Fujimoto, 1994; Wethington & Kessler, 1989), supporting the enhancement hypothesis. An exception occurs when the person works more than 54 hours a week, which is associated with increased psychological distress (Glass & Fujimoto, 1994).

When the outcome variable is marital quality, the preponderance of studies find that work status per se (wife employed or not) is unrelated to marital quality (e.g., Houseknecht & Macke, 1981; Smith, 1985). This finding, rather than disconfirming the scarcity and enhancement hypotheses, may instead suggest that both scarcity and enhancement processes occur and can balance each other.

Current theorizing and research framed by the enhancement hypothesis no longer focuses simply on the number of hours of employment or the number of roles, but rather takes into account the quality of roles (e.g., Barnett, Marshall, Raudenbush, & Brennan, 1993; Baruch & Barnett, 1986; Hyde, Klein, Essex, & Clark, 1995; Klein, Hyde, Essex, & Clark, 1998). According to this view, a person's psychological distress or relationship distress is influenced not so much by whether the person is employed or by the number of hours of employment, but rather by the quality of the work role. Empirical research supports this view (e.g., Bartlett et al., 1993; Greenberger & O'Neill, 1993).

The scarcity and enhancement perspectives take little account of individual differences, specifically individual differences in identities. Identity theory (Stryker & Serpe, 1994; Stryker & Statham, 1985), in contrast, introduces the related constructs of commitment and identity salience. The individual has multiple role identities--for example, wife, mother, and attorney. The construct of identity salience rests on the notion that the individual's identities fall into a hierarchy from most to least salient, based on the individual's readiness to enact a role identity. The salience of a role identity is a function of commitment to that role; commitment reflects the social and personal costs that would result if the person no longer fulfilled the role (Stryker & Serpe, 1994). Insofar as salience indicates the individual's readiness to enact a role, high family salience (compared with work salience) should be associated with more positive relationship outcomes. Inter-role conflicts will be strong only for salient identities; thus for people who are high in both work salience and family salience, the conflict should be strong and should be reflected in negative outcomes for the relationship. Statistically, this leads to a prediction of an interaction between work salience and family salience in predicting relationship outcomes.

Multiple Roles and Sexual Functioning

When the focus shifts specifically to the impact of multiple roles on sexual functioning, the literature is sparse. Sex therapists Sarrel and Sarrel (1984) raised concerns about couples in which one or both of the partners are workaholics, overcommitted to their job and working considerably more than 40 hours a week. Sex therapists have speculated that multiple roles, particularly in dual-earner couples, lead to inhibited sexual desire, or ISD. LoPiccolo and Friedman (1988) identified several causes of ISD. One is anhedonic style, the lack of capacity for play. Clients characterized by this style include very ambitious, task-directed, high achievers--that is, workaholics. Based on her clinical experience, Renshaw (1993) argued that stress and fatigue due to overwork are common causes of sexual dysfunctions. These clinical reports are consistent with the scarcity hypothesis.

Henderson-King and Veroff (1994) interviewed both members of 199 Black and 174 White newlywed couples. They included measures of positive sexual relations (including joy and excitement) and negative sexual relations (including upset). They found that income was negatively related to reports of joy by Black women and positively related to reports of upset by White women. They speculated that, among Whites, higher income couples are dual-earner couples, and that women committed to careers have higher expectations regarding sexual satisfaction or experience higher levels of stress, which negatively affects sexual satisfaction.

Call, Sprecher, and Schwartz (1995) addressed the decline in frequency of marital sexual intercourse over time reported by several previous studies. They speculated that this decline reflects increasing involvement in other roles, such as parenthood or a demanding job. According to Call et al., because wives who work outside the home also do much of the housework, dual-income couples may be especially likely to report a decline in frequency of sex. Analyzing data from the National Survey of Families and Households, involving 6,785 persons who were married and living with a spouse, they found that, contrary to predictions, employment of both spouses was not related to sexual frequency.

In regard to the effect of children--both presence and number--and the addition of the parent role, previous research is ambiguous. Blumstein and Schwartz (1983) asserted that children had a negative effect on couples' sexual functioning but did not report specific data analyses on the question. Greenblatt (1983) found a significant negative correlation, r = -.17, between frequency of intercourse and number of children. However, number of children was correlated with or confounded with length of time in the relationship, r = .44, and the latter is negatively correlated with frequency. Therefore, it is unclear whether the effect was due to number of children or length of time in the relationship. Call et al. (1995) found that marital sex was significantly less frequent when the couple had a baby or young child (vs. no children or older children). In contrast, couples who had a child between the ages of 5 and 18 had a higher frequency of intercourse than other couples. We know of only one study that has provided empirical data on a community sample and that explicitly examined the sexual functioning of dual-earner couples. Hyde, DeLamater, and Hewitt (1998) studied the sexual functioning of a sample of more than 500 married couples as a function of employment variables. They found no significant difference between homemakers and women employed part time, full time, and high full time on measures of frequency of intercourse, sexual satisfaction, and decreased sexual desire. Work-role quality was a stronger predictor of sexual outcomes than hours of work; better work-role quality was associated with more positive sexual functioning. Fatigue was a strong negative predictor of sexual outcomes for women, yet homemakers reported as much fatigue as employed women.

A strength and a limitation of Hyde et al.'s (1998) data set, taken from the Wisconsin Maternity Leave and Health (WMLH) Project, was that it was longitudinal, covering three time points from the middle trimester of pregnancy through 1 year after the birth. Hyde et al. (1998) argued that the timing of data collection around the birth of a child provided a particularly sensitive test of the possible deleterious effects of multiple roles on sexual functioning, since role overload should be prominent at that time. However, one might argue the opposite, namely that the demands of pregnancy and caring for a new baby have such a profound impact on couples' sexual relationships that employment counts for little.

The purpose of the current studies was to reexamine the questions posed by Hyde et al. (1998) using two data sets. The first was the National Health and Social Life Survey (NHSLS: Laumann, Gagnon, Michael, & Michaels, 1994), a well-sampled survey of the sexual practices of approximately 3,400 Americans. The original publication on this data set did not examine employment in relation to sexuality (Laumann, et al., 1994), nor did a later publication on sexual disorders (Laumann, Paik, & Rosen, 1999). The second was the longitudinal continuation of the WMLH study, with sexuality data collected again when the index child was 4.5 years of age. Both data sets permit the examination of the sexual functioning of married persons as it relates to their and their spouse's employment characteristics at times in the family life course beyond the baby years.

The Current Research

In Study 1 we asked the following questions using the NHSLS data set:

1. Are there differences between married women employed part time, full time, or high full time in a variety of aspects of sexuality, including frequency of intercourse, sexual satisfaction, frequency of orgasm, and sexual problems such as lack of interest in sex and inability to orgasm? The scarcity hypothesis would argue that women in the full-time and high full-time groups should engage in intercourse less frequently and experience less sexual satisfaction. The enhancement hypothesis, in contrast, predicts that a couple's sexual relationship should benefit from the wife's employment. This argument might not extend to high full-time employment, however; any benefits to employment should be evident in part-time and full-time employment and no further benefits should occur with increased hours.

2. Are there differences between married men employed full time or high full time in a variety of aspects of sexuality, including frequency of intercourse, sexual satisfaction, frequency of orgasm, and sexual problems such as anxiety about performance and climaxing too early? Although most research has focused on the effect of women's employment, the scarcity hypothesis and enhancement hypothesis should apply equally well to men.

3. Shifting the focus to couples, is there an interaction between the wife's hours of work and the husband's hours of work for sexual outcomes?

4. Is there an effect of number of children on couples' sexual functioning, and is there an interactive effect with parents' work hours? The scarcity hypothesis argues that the greater the number of children, the greater the negative impact on the parents' sexual relationship. The enhancement hypothesis would argue that the addition of a child, by adding an additional role, should benefit functioning, although additional children might not create any further benefits.

On the basis of prior research, and framed by the scarcity and enhancement hypotheses as well as identity theory, we asked the following questions in Study 2, which used the WMLH data set:

1. Are there differences between homemakers, women employed part time, women employed full time, and women employed high full time in frequency of intercourse or sexual satisfaction? Are there differences in sexual satisfaction among husbands of women in these four employment groups?

2. Are there differences between men who are employed full time and those employed high full time in sexual satisfaction? (Men who were unemployed, were home full time, or were employed less than 32 hrs/wk were not included in the analyses because there were too few of them.)

3. In multiple regression analyses, are sexual outcomes (frequency of intercourse, satisfaction) better predicted by hours of employment or by work-role quality? Prior research (Hyde et al., 1998) indicates that the latter should be the better predictor.

4. Based on identity theory, do work role salience and spouse (marital role) salience predict sexual outcomes?

5. Within a multiple regression approach, several interaction hypotheses were framed. First, is there a quality of workaholism captured by a combination of high work hours and high work salience that is a particularly potent negative predictor of sexual outcomes (i.e., is there an interaction between work hours and work salience)? Focusing on the couple, is there an interaction between the wife's work hours and the husband's work hours indicating that the combination of high work hours for both creates a particularly negative outcome? Again focusing on the couple, is it a particularly potent combination when both husband and wife are high on work salience, creating negative sexual outcomes (i.e., is there an interaction between wife's work salience and husband's work salience)? Based on the postulate from identity theory that interrole conflict is strong only between two roles with salient identities, is there an interaction between work salience and family salience in predicting sexual outcomes, with the high/high combination yielding particularly negative outcomes? Finally, based on a finding by Hyde et al. (1998) in their study of sexuality and dual-earner couples, do the husband's work-role quality and the wife's work-role quality interact in predicting sexual outcomes? Their research indicated that the combination of a husband with low work-role quality and a wife with high work-role quality was associated with negative sexual outcomes.

STUDY 1

Study 1 examined the sexuality of dual-earner couples using the NHSLS data, which represent a well-sampled survey of the sexuality of adult Americans.

Method

Sample. A total of 3,432 persons between the ages of 18 and 59 were interviewed for the NHSLS (Laumann et al., 1994). The sample was obtained by probability sampling of households in the United States, which excluded less than 3% of Americans (e.g., the homeless and persons living in dormitories). The response rate was 79%. For the total sample, 71.4% were White (not Hispanic), 16% were Black (not Hispanic), 9.4% were Hispanic, 1.9% were Asian/Pacific Islander, and 1.2% were Native American/Alaskan. Forty-three percent were men and 57% were women. In regard to education, 14.5% had not graduated from high school, 63% were high school graduates, 15.6% were college graduates, and 6.9% held an advanced degree.

For the analyses reported here, we selected those respondents who were married at the time of the interview. Because of the impact of pregnancy on sexuality and because this study sought to examine sexuality beyond the baby years, women who were currently pregnant were omitted, for a total of 969 women and 775 men.

Procedure. The data were obtained in face-to-face interviews supplemented by brief written questionnaires on particularly sensitive topics (e.g., masturbation).

Measures. Respondents reported the hours they worked last week at all paid jobs. In the publicly available data set, however, employment hours are not available as a continuous variable, but rather as a categorical variable, with categories 1 to 10, 11 to 19, and so on.(1) They also reported, for their spouse, whether the spouse was unemployed, worked pan time or full time (hours were unspecified), was a student, was retired, or was a full-time homemaker. Number of children living in the household, under 18 years of age and related to the respondent, were reported.

Participants indicated dichotomously (yes/no) whether they had experienced each of the following sexual problems for several months or more in the past 12 months: lack of interest in sex, unable to orgasm, climaxed too early, painful during sex, sex not pleasurable even if not painful, anxious about performing, and (for those who reported a problem) avoided sex because of the problem. Men were asked, in addition, about trouble achieving or maintaining an erection. Women were asked about trouble lubricating. All participants were asked whether sex with their spouse made them feel satisfied.

A number of items with continuous response options were asked: frequency of sex in the last year (6 categories, from not at all to 4 or more times a week); how emotionally satisfying was the relationship with the spouse (from 1 = extremely to 5 = not at all); how physically pleasurable was the relationship with the spouse (1 = extremely to 5 = not at all); and frequency of orgasm in the past 12 months (1 = always to 5 = never).

Results

Bivariate correlations among all variables as well as means and standard deviations are shown in Tables 1 and 2, for women and men respectively. Notice that for women, of the 13 sexuality variables only one showed a significant correlation with hours worked last week: frequency of sex. The correlation was actually positive, but small, r = .11, p [is less than] .05. For men, no sexuality variables correlated significantly with hours worked per week.

Table 1. Correlations, Means, and Standard Deviations
for all Variables for Married Women, NHSLS Data Set

                             1        2        3        4        5

 1. Lacked interest
    in sex Y/N

 2. Unable to              .47**
    orgasm Y/N

 3. Climaxed too           .16**    .07*
    early Y/N

 4. Painful during         .24**    .19**    .09**
    sex Y/N

 5. Sex not pleasurable    .45**    .40*     .06      .31**
    Y/N

 6. Anxious about          .17**    .19**    .14**    .18**    .18**
    perform Y/N

 7. Trouble lubricating    .25**    .28**   -.01      .29**    .27**
    Y/N

 8. Was sex satisfying    -.12**   -.19**   -.04     -.03     -.16**
    Y/N

 9. Avoided sex            .37**    .18**   -.05      .25**    .38**
    bec prob Y/N

10. Frequency of           .26**    .20**    .04      .04      .12**
    sex 1 = not at all

11. Emotionally           -.27**   -.23**   -.01     -.05     -.29**
    satisfying 1
    = extremely

12. Physically            -.28**   -.28*    -.02     -.08*    -.32**
    pleasing 1
    = extremely

13. Frequency of          -.11**   -.31**    .02     -.08*    -.17**
    orgasm 1 = always

14. Hours worked           .03      .04      .05      .03      .03
    last week

15. Hours worked           .02      .06      .02      .01      .02
    Categories

16. Spouse works          -.07*     .01      .00     -.02     -.01
    1 = part-time

Mean                      1.70     1.78     1.91     1.87     1.79

sd                         .46      .41      .29      .34      .41

N                          958      944      944      943      944

                             6        7        8        9       10

 1. Lacked interest
    in sex Y/N

 2. Unable to
    orgasm Y/N

 3. Climaxed too
    early Y/N

 4. Painful during
    sex Y/N

 5. Sex not pleasurable
    Y/N

 6. Anxious about
    perform Y/N

 7. Trouble lubricating    .20**
    Y/N

 8. Was sex satisfying    -.08*    -.05
    Y/N

 9. Avoided sex            .15**    .05     -.01
    bec prob Y/N

10. Frequency of           .06      .12**   -.15**    .13**
    sex 1 = not at all

11. Emotionally           -.09**   -.07*     .42**   -.14**   -.29**
    satisfying 1
    = extremely

12. Physically            -.09**   -.11**    .42**   -.14**   -.27**
    pleasing 1
    = extremely

13. Frequency of          -.07*    -.06      .23**   -.09*    -.13*
    orgasm 1 = always

14. Hours worked          -.01     -.00      .01      .04      .11
    last week

15. Hours worked          -.03     -.01      .02      .00      .07
    Categories

16. Spouse works           .00     -.07     -.01     -.00     -.04
    1 = part-time

Mean                      1.90     1.79     1.04     1.50     3.97

sd                         .30      .41      .19      .50     1.27

N                          947      941      912      509      905

                           11       12       13      14       15

 1. Lacked interest
    in sex Y/N

 2. Unable to
    orgasm Y/N

 3. Climaxed too
    early Y/N

 4. Painful during
    sex Y/N

 5. Sex not pleasurable
    Y/N

 6. Anxious about
    perform Y/N

 7. Trouble lubricating
    Y/N

 8. Was sex satisfying
    Y/N

 9. Avoided sex
    bec prob Y/N

10. Frequency of
    sex 1 = not at all

11. Emotionally
    satisfying 1
    = extremely

12. Physically             .75**
    pleasing 1
    = extremely

13. Frequency of           .25*     .34**
    orgasm 1 = always

14. Hours worked          -.02     -.04      .03
    last week

15. Hours worked          -.05     -.06     -.01     .86**
    Categories

16. Spouse works          -.07*    -.04     -.05    -.00     -.03
    1 = part-time

Mean                      1.82     1.79     2.04   33.36     1.96   .98

sd                         .87      .79      .92   13.91      .68   .15

N                          917      903      616     614      862

Note. All correlations are Pearsonian. For dichotomous variables,
point-biserial is equivalent to Pearson.
Table 2. Correlations, Means, and Standard Deviations
for all Variables, Married Men, NHSLS Data Set

                                  1        2        3        4

 1. Lacked interest
    in sex Y/N

 2. Unable to                   .39**
    orgasm Y/N

 3. Climaxed too                .16**    .15*
    early Y/N

 4. Painful during              .13**    .04      .10**
    sex Y/N

 5. Sex not pleasurable         .34**    .29*     .21**    .13**
    Y/N

 6. Anxious about               .24**    .29**    .26**    .23**
    perform Y/N

 7. Trouble w/                  .26**    .42**    .16**    .16**
    erection Y/N

 8. Was sex satisfying         -.04     -.17**   -.04     -.13**
    Y/N

 9. Avoided sex                 .36*     .25**   -.30**    .25**
    bec prob Y/N

10. Frequency of                .16**    .11**    .10**    .02
    sex 1 = not at all

11. Emotionally                -.10**   -.11**   -.09*    -.01
    satisfying 1 = extremely

12. Physically                 -.15**   -.18**   -.09*    -.01
    pleasing 1 = extremely

13. Frequency of               -.12**   -.18**    .04     -.02
    orgasm 1 = always

14. Hours worked               -.00      .03     -.02      .03
    last week

15. Hours worked                .06      .08     -.01      .05
    categories

16. Spouse works                .04     -.01      .06      .00
    1 = parttime

Mean                           1.88     1.92     1.71     1.91

sd                              .32      .26      .45      .17

N                               772      770      766      767

                                      5             6        7

 1. Lacked interest
    in sex Y/N

 2. Unable to
    orgasm Y/N

 3. Climaxed too
    early Y/N

 4. Painful during
    sex Y/N

 5. Sex not pleasurable
    Y/N

 6. Anxious about               .27**
    perform Y/N

 7. Trouble w/                  .32**    .38**
    erection Y/N

 8. Was sex satisfying         -.18**   -.06     -.23**
    Y/N

 9. Avoided sex                 .24**    .32**    .31**   -.11*
    bec prob Y/N

10. Frequency of                .06      .06      .04     -.08*
    sex 1 = not at all

11. Emotionally                -.16**   -.15**   -.10*     .16**
    satisfying 1 = extremely

12. Physically                 -.24**   -.15**   -.14**    .26**
    pleasing 1 = extremely

13. Frequency of               -.07     -.03     -.09*     .18**
    orgasm 1 = always

14. Hours worked               -.04      .04     -.01      .00
    last week

15. Hours worked               -.03      .09      .05     -.01
    categories

16. Spouse works               -.02      .00     -.05     -.03
    1 = parttime

Mean                           1.94     1.86     1.91     1.01

sd                              .24      .34      .29      .11

N                               768      769      769      742

                                  8        9       10       11

 1. Lacked interest
    in sex Y/N

 2. Unable to
    orgasm Y/N

 3. Climaxed too
    early Y/N

 4. Painful during
    sex Y/N

 5. Sex not pleasurable
    Y/N

 6. Anxious about
    perform Y/N

 7. Trouble w/
    erection Y/N

 8. Was sex satisfying
    Y/N

 9. Avoided sex
    bec prob Y/N

10. Frequency of                .16**
    sex 1 = not at all

11. Emotionally                -.16**   -.15**
    satisfying 1 = extremely

12. Physically                 -.15**   -.21**    .76**
    pleasing 1 = extremely

13. Frequency of               -.00     -.10      .01      .12**
    orgasm 1 = always

14. Hours worked                .00      .02     -.04     -.04
    last week

15. Hours worked                .09      .02     -.04     -.02
    categories

16. Spouse works                .01      .09     -.03     -.01
    1 = parttime

Mean                           1.75     4.06     1.67     1.62

sd                              .43     1.22      .77      .77

N                               321      724      743      743

                                 12       13       14       15

 1. Lacked interest
    in sex Y/N

 2. Unable to
    orgasm Y/N

 3. Climaxed too
    early Y/N

 4. Painful during
    sex Y/N

 5. Sex not pleasurable
    Y/N

 6. Anxious about
    perform Y/N

 7. Trouble w/
    erection Y/N

 8. Was sex satisfying
    Y/N

 9. Avoided sex
    bec prob Y/N

10. Frequency of
    sex 1 = not at all

11. Emotionally
    satisfying 1 = extremely

12. Physically
    pleasing 1 = extremely

13. Frequency of
    orgasm 1 = always

14. Hours worked               -.05
    last week

15. Hours worked               -.11*     .85**
    categories

16. Spouse works               -.03      .03     -.03
    1 = parttime

Mean                           1.32    43.06     1.37   .79

sd                              .74    13.44      .48   .40

N                               741      640      591   504

Note. All correlations are Pearsonian. For dichotomous variables,
point-biserial is equivalent to Pearson.

Chi-square analyses were conducted to test for an association between hours worked and the dichotomous variables reporting sexual problems (see Tables 3 and 4). To make the analyses as close as possible to those reported in Study 2, hours worked were grouped, for women, as follows: part time (1-29 hrs/wk), full time (30-40 hrs/wk), and high full-time (41-70 or more hrs/wk). (Recall that the variable in the NHSLS data set did not report exact number of hours, but rather categories: 1-10, 11-19, etc. Given these categories, we could only approximate the work categories of Study 2.) None of the chi-square tests were significant. Additionally, a chi-square test was run for the association between work hours and the dichotomous variable, did sex with the spouse make them feel satisfied, [chi square](2) = 2.87, n.s.

Table 3. Reports of Sexual Problems by Married Women as
a Function of Hours Worked per Week, From the
NHSLS Data Set

                                                 Work Hours

                                                                High
Sexual problem                        Part-time   Full-time   full-time

Lacked interest in sex                  32.5%       27.0%       29.8%
Unable to orgasm                        24.7%       23.0%       17.8%
Climaxed too early                       9.1%       10.1%        7.1%
Painful during sex                      13.6%       12.6%       12.3%
Sex not pleasurable                     26.0%       20.1%       23.8%
Anxious about performing                 8.4%        7.2%       10.8%
Trouble lubricating                     24.8%       19.9%       23.8%
Avoided sex because of the problems     50.0%       50.9%       49.2%

Note. None of the chi-square tests of the association between work
hours and sex problems were significant. For all variables except the
last, n = 153-154 for part time, 316-322 for full time, and 129-131
for high full-time. Variations in n reflect missing data. For the
variable "avoided sex," n = 94, 167, and 63 for part-, full-, and high
full-time groups, respectively; n is reduced because the question was
asked only to those who reported a sexual problem.
Table 4. Reports of Sexual Problems by Married Men as a
Function of Their Own and Their Spouses' Hours
Worked per Week, From the NHSLS Data Set

                                        Spouse work status

                                 Full-time
Sexual problem                   homemaker   Part-time   Full-time

Lacked interest in sex
  Full-time men                    11.1%       16.0%       11.3%
  High full-time men                8.5         9.1        11.4
Unable to orgasm(a)
  Full-time men                     4.9         4.0        15.3
  High full-time men                5.2         9.1         0.9
Climaxed too early
  Full-time men                    31.3        34.0        28.2
  High full-time men               37.9        39.4        24.8
Painful during sex(a)
  Full-time men                     1.2         4.0         1.5
  High full-time men                0.0         0.0         0.9
Sex not pleasurable
  Full-time men                     6.2         6.0         3.9
  High full-time men                5.1         3.0         7.0
Anxious about performing
  Full-time men                    12.3        16.0        18.2
  High full-time men                5.2         9.1        10.5
Trouble with erection
  Full-time men                     7.4        10.2        10.8
  High full-time men                5.2         6.1         7.9
Avoided sex because of problem
  Full-time men                    18.2        32.0        26.1
  High full-time men               16.0        20.0        15.4

Note. None of the chi-square tests of the association between men's
and women's work hours per week and men reporting sexual problems
were significant.

(a) Chi-square not computed for this variable because of low frequency
of reporting this problem.

Among employed men, very few worked less than 30 hrs/wk. Therefore, we formed two categories: full time (30-49 hrs/wk) and high full-time (50 or more). To permit a couple-level analysis, results were computed jointly for husband's work status and wife's work status (Table 4). Chi-square tests of association were computed for the association between husbands' and wives' work status and sexual problems. None of these chi-square analyses were significant (i.e., there was no relationship between couples' hours worked per week and the incidence of sexual disorders). Note in particular in Table 4 that what some would view as the most difficult combination, high full-time husbands with full-time employed wives, shows no pattern of increased sexual disorders compared with the other five couple combinations. A parallel couple-level analysis could not be run for reports by wives because so few husbands worked part time.

For married women, a MANOVA was run for differences between the part-time, full-time, and high full-time groups, with four dependent variables: frequency of sex, frequency of orgasm, emotionally satisfying, and physically pleasing. The difference between groups was not significant, multivariate F(8, 1100) = 0.95, n.s. None of the univariate tests was significant. To test for the effects of number of children in the household as well as work status, a 3 (work status: part time, full time, or high full-time) x 4 (number of children: 0, 1, 2, and 3 or more) MANOVA was run with the same four dependent variables. Neither the main effect of work status nor number of children was significant, but the interaction was, multivariate F(24, 2172) = 1.53, p [is less than] .05. Two variables showed significant univariate interactions: how emotionally satisfying the sexual relationship was, F(6, 543) = 1.55, p [is less than] .05, and how physically pleasing the sexual relationship was, F(6, 543) = 1.63, p [is less than] .02. A graph of these interactions indicated no clear pattern, with only small differences among cell means; clearly the large sample size rendered small and inconsistent effects significant.

For married men's responses, a 2 (husband's work status: full time or high full-time) x 3 (wife's work status: homemaker, employed part time, employed full time) MANOVA was run with four dependent variables: frequency of sex, frequency of orgasm, emotionally satisfying, and physically pleasing. The main effect of husband's work status was not significant, multivariate F(4,491) = 2.25, n.s., and neither were any of the univariate tests. Similarly, the main effect of wife's work status was not significant, multivariate F(8,984) = .35, n.s., nor were any of the univariate tests. The interaction between husband's work status and wife's work status was not significant, multivariate F(8,984) = 1.35, n.s.

In order to investigate the impact of children, a 2 (husband's work status) x 3 (wife's work status) x 4 (number of children in household) MANOVA was also run with the same four dependent variables. Consistent with the previous MANOVA, neither the main effect of husband's work status nor the main effect of wife's work status was significant. The main effect for number of children was not significant, nor were any of the two-way or three-way interactions.

Discussion

Consistent with the results of Hyde et al. (1998), women's hours of employment were unrelated to a wide variety of sexuality outcomes. Even women who worked high full-time showed no deficits. Moreover, and again consistent with the findings of Hyde et al. (1998), men's hours of employment were unrelated to sexuality outcomes, even when considering those working high full-time. At the couple level, there was no evidence that high full-time husbands combined with wives employed full time is a particularly dangerous pairing; interactions between husbands' and wives' work status were not significant. Thus, these results provide no support for the scarcity hypothesis, nor do they support concerns about the sexual relationships of dual-earner couples.

STUDY 2

Study 1 demonstrated the absence of differences in sexual expression between women employed part time, full time, or high full-time, and between men employed full time and high full-time. The NHSLS data set, although well sampled, lacks detailed psychological measures such as identity salience of particular roles. We therefore examined more complex hypotheses associated with these psychological variables with the WMLH data set at Time 7, when the children were 4.5 years of age, in other words, beyond the baby years.

Method

Participants. A total of 570 pregnant women and 550 husbands or partners of the women were recruited for participation in the Wisconsin Maternity Leave and Health (WMLH) Project (see Hyde et al., 1995, for details about the sample and its recruiting). Approximately 78% of the sample were from the Milwaukee, Wisconsin, Standard Metropolitan Statistical Area (SMSA) and 22% were from the Madison, Wisconsin, SMSA. For inclusion in the sample, the women had to be over 18 years of age and living with the baby's father, although not necessarily married (the intention was to include fathers equally in the study).

Of the original sample, 299 women had useable sexuality data at Time 7, when the index child was 4.5 years old. They are a subset of the 469 women who completed Time 7 data collection. Of those 469 women, the remainder were not given the sexuality questionnaire because they were now divorced or separated from their partner, they agreed only to a brief telephone interview, or they were viewed by project staff as at high risk for loss from the longitudinal study.

Of this subset of 299 women, 13 were excluded because they were currently pregnant and 25 were excluded because they had given birth in the past year (our goal was to investigate sexuality beyond the baby years).

The following statistics describe the sample of 261 women studied here. The average age at Time 7 was 34.9 years (SD = 4.25); 97% were married to the child's father (for that reason we refer to the participants as husbands and wives). Among the wives, 6 (2%) were Native American, 3 (1%) were Asian American, 6 (2%) were Hispanic, 4 (2%) were African American, and 242 (95%) were Euro-American. Among the husbands, 2 (1%) were Native American, 2 (1%) were Asian American, 5 (2%) were Hispanic, 8 (3%) were African American, 228 (89%) were Euro-American, and 10 (4%) did not report their race/ethnicity. The average age of the husbands was 37.4 years (SD = 9.32). The mean yearly family income before taxes was $67,305 (SD = 34,188) at T7. In 1996, the year in which the T7 data were collected, the mean income for dual-earner couples in the United States was $59,833 (U.S. Census Bureau, 1998); thus, this sample is slightly above the national average.

Procedure. Women were interviewed in their homes by a female interviewer on each of the following occasions: (a) during the second trimester of pregnancy (Time 1), (b) 12 months following the birth (Time 4), and (c) 4.5 years after the birth (Time 7). In addition, women completed mail-out questionnaires at the time of each interview. The interviewer administered a brief sex questionnaire during the interview by giving the questionnaire to the woman and allowing her to fill it out in private; she then sealed it in a privacy envelope and gave it to the interviewer. Husbands were interviewed by phone and completed mail-out questionnaires at each of the same times.

Participants were recruited over a period of 15 months beginning in June, 1990. Of those women who completed the T1 interview, 560 had live births and were eligible to continue in the study. The T4 interview was completed by 95% of the sample. The T7 interview was completed by 82% of the original T1 sample. Respondents were paid for their participation.

Measures. Participants were asked to provide information regarding their age, race/ethnicity, family income before taxes, and education level. Levels of education were assigned numerical values as follows: 1 = less than high school, 2 = vocational school, 3 = high school or equivalent, 4 = technical college, 5 = some college, 6 = college graduate, 7 = college graduate plus some other education, and 8 = graduate degree.

At each interview time, for both wives and husbands, the respondents were asked whether they were currently working for pay or profit and, if so, the average number of hours worked per week.

Wives' and husbands' reports of the quality of their jobs were assessed by scales developed by Barnett and Marshall (1989). The Job-Role Quality Scale measures both positive (rewards) and negative (concerns) aspects of the role. The Rewards scale consists of 23 items, such as "When you think about your job right now, how rewarding is it because of the sense of accomplishment and competence you get from your job?" Each item was rated on a scale from 1 (not at all) to 4 (extremely). Coefficient alpha ranged between .87 and .89 for all scales. The individual's score is the average of their scores on the items, and therefore higher scores reflect more concerns or rewards, respectively. In order to obtain a single measure of work role quality, the mean of the individual's work concern items was subtracted from the mean of their work reward items.

Work salience and spouse salience (salience of one's role as spouse) were measured with three items with parallel structure for each construct. An example for spouse salience is "How important is your marriage or relationship to your sense of yourself?" Participants rated from 1 (not at all important) to 7 (extremely important) the degree to which they were involved in their work/marriage. The mean was calculated for the three items; higher numbers indicate greater salience. For the work salience scale, coefficient alpha was .83 for wives and .80 for husbands. For the spouse salience scale, coefficient alpha was .78 for wives and .85 for husbands.

To measure sexual attraction, we used items from a scale developed by Barnett and Marshall (1989) to measure rewards individuals perceive from their romantic partners. Within this partner rewards scale is a 3-item subscale of physical attraction. An example item is, "When you think about your relationship right now, how rewarding is it because of your sexual relationship?" Participants rate the degree to which they agree with each statement on a scale from 1 (not at all) to 4 (extremely). The mean of these items was calculated and higher scores indicate greater sexual attraction. Coefficient alpha for wives and husbands at T7 were .84 and .86, respectively.

A pair of questions on the sex questionnaire assessed frequency of intercourse: "In the past month, have you had sexual intercourse?" (Yes, no), and "If yes, about how many times?" The sex questionnaire also included questions about whether the respondent had engaged in the following behaviors in the past month: cunnilingus, fellatio, petting, and masturbation. Response options were yes and no. Another question was, "How satisfied are you with your current sexual relationship with your partner?" This question was rated on a scale from 1 (very dissatisfied) to 5 (very satisfied).

A single item measured loss of sexual interest. Participants responded in a yes/no format to the question, "During the past 2 weeks, was your interest in sex a lot less than usual?" This item was taken from the National Institute of Mental Health Diagnostic Interview Schedule (DIS: Robins, Helzer, Croughan, & Ratcliff, 1981) section on depression completed during the home interview.

Results

Descriptive statistics, including pairwise correlations between all variables, means, and standard deviations, are shown in Table 5.

Table 5. Pairwise Correlations, Means, and Standard
Deviations for Continuous Variables, WMLH Data Set

                        1           2          3         4

 1. Frequency of     1.0
    intercourse
    (T4)(W)

 2. Sexual            .41(**)   1.0
    satisfaction
    (T4)(W)

 3. Sexual            .38(**)    .48(**)   1.0
    attraction
    (T4)(W)

 4. Frequency of      .79(**)    .32(**)    .30(**)   1.0
    intercourse
    (T4)(H)

 5. Sexual            .34(**)    .34(**)    .33(**)    .43(**)
    satisfaction
    (T4)(H)

 6. Sexual            .36(**)    .33(**)    .41(**)    .39(**)
    attraction
    (T4)(H)

 7. Frequency of      .48(**)    .29(**)    .28(**)    .27(**)
    intercourse
    (T7)(W)

 8. Sexual            .18        .35(**)    .34(**)    .12
    satisfaction
    (T7)(W)

 9. Sexual            .27(**)    .34(**)    .60(**)    .23(**)
    attraction
    (T7)(W)

10. Spouse            .05        .30(**)    .48(**)   -.02
    salience
    (T7)(W)

11. Hours worked      .21(**)    .04        .05        .18(*)
    per week
    (T7)(W)

12. Work salience     .05        .04        .16(*)     .01
    (T7)(W)

13. Work role        -.07        .08        .21(**)   -.12
    quality
    (T7)(W)

14. Sexual            .27(**)    .21(**)    .26(**)    .24(**)
    attraction
    (T7)(H)

15. Spouse            .05        .08        .19(**)    .05
    salience
    (T7)(H)

16. Hours worked      .12       -.06        .01        .17(*)
    per week
    (T7)(H)

17. Work salience     .04       -.04        .02        .15(*)
    (T7)(H)

18. Work role        -.04        .02        .01        .11
    quality
    (T7)(H)

Mean                 5.23       3.63       2.75       4.28

Standard deviation   4.14       1.19        .69       4.21

                        5           6          7         8

 1. Frequency of
    intercourse
    (T4)(W)

 2. Sexual
    satisfaction
    (T4)(W)

 3. Sexual
    attraction
    (T4)(W)

 4. Frequency of
    intercourse
    (T4)(H)

 5. Sexual           1.0
    satisfaction
    (T4)(H)

 6. Sexual            .55(**)   1.0
    attraction
    (T4)(H)

 7. Frequency of      .25(**)    .17(*)    1.0
    intercourse
    (T7)(W)

 8. Sexual            .29(**)    .27(**)    .43(**)   1.0
    satisfaction
    (T7)(W)

 9. Sexual            .28(**)    .35(**)    .48(**)    .44(**)
    attraction
    (T7)(W)

10. Spouse            .15        .26(**)    .22(**)    .43(**)
    salience
    (T7)(W)

11. Hours worked      .07        .14        .02       -.12
    per week
    (T7)(W)

12. Work salience     .03        .02        .02        .07
    (T7)(W)

13. Work role        -.00       -.08        .03        .13
    quality
    (T7)(W)

14. Sexual            .32(**)    .48(**)    .31(**)    .25(**)
    attraction
    (T7)(H)

15. Spouse            .23(**)    .34(**)    .16(*)     .22(**)
    salience
    (T7)(H)

16. Hours worked     -.05       -.04        .02       -.06
    per week
    (T7)(H)

17. Work salience    -.05        .15(*)    -.09       -.12
    (T7)(H)

18. Work role         .06        .15(*)     -         -.12
    quality
    (T7)(H)

Mean                 3.4        2.66       5.04       3.61

Standard deviation   1.33        .71       4.01       1.21

                        9          10         11        12

 1. Frequency of
    intercourse
    (T4)(W)

 2. Sexual
    satisfaction
    (T4)(W)

 3. Sexual
    attraction
    (T4)(W)

 4. Frequency of
    intercourse
    (T4)(H)

 5. Sexual
    satisfaction
    (T4)(H)

 6. Sexual
    attraction
    (T4)(H)

 7. Frequency of
    intercourse
    (T7)(W)

 8. Sexual
    satisfaction
    (T7)(W)

 9. Sexual           1.0
    attraction
    (T7)(W)

10. Spouse            .57(**)    1.0
    salience
    (T7)(W)

11. Hours worked     -.02        -.04        1.0
    per week
    (T7)(W)

12. Work salience     .21(**)     .21(**)     .14(*)    1.0
    (T7)(W)

13. Work role         .24(**)     .23(**)    -.14(*)     .45(**)
    quality
    (T7)(W)

14. Sexual            .40(**)     .30(**)     .01        .01
    attraction
    (T7)(H)

15. Spouse            .39(**)     .34(**)    -.05        .09
    salience
    (T7)(H)

16. Hours worked     -.03          -         -.01       -.03
    per week
    (T7)(H)

17. Work salience    -.00         .07         .02        .05
    (T7)(H)

18. Work role        -.00         .06         .00        .04
    quality
    (T7)(H)

Mean                 2.7         5.83       29.87       5.41

Standard deviation   1.0          .67         .94      13.34

                      13      14         15         16

 1. Frequency of
    intercourse
    (T4)(W)

 2. Sexual
    satisfaction
    (T4)(W)

 3. Sexual
    attraction
    (T4)(W)

 4. Frequency of
    intercourse
    (T4)(H)

 5. Sexual
    satisfaction
    (T4)(H)

 6. Sexual
    attraction
    (T4)(H)

 7. Frequency of
    intercourse
    (T7)(W)

 8. Sexual
    satisfaction
    (T7)(W)

 9. Sexual
    attraction
    (T7)(W)

10. Spouse
    salience
    (T7)(W)

11. Hours worked
    per week
    (T7)(W)

12. Work salience
    (T7)(W)

13. Work role        1.0
    quality
    (T7)(W)

14. Sexual           -.03   1.0
    attraction
    (T7)(H)

15. Spouse            .07    .48(**)   1.0
    salience
    (T7)(H)

16. Hours worked     -.05   -.07        .03       1.0
    per week
    (T7)(H)

17. Work salience    -.02    .12        .29(**)    .22(**)
    (T7)(H)

18. Work role         .03    .12        .24(**)    .06
    quality
    (T7)(H)

Mean                 1.46   2.64       5.69       4.04

Standard deviation    1.1    .67        .67       1.06

                       17        18

 1. Frequency of
    intercourse
    (T4)(W)

 2. Sexual
    satisfaction
    (T4)(W)

 3. Sexual
    attraction
    (T4)(W)

 4. Frequency of
    intercourse
    (T4)(H)

 5. Sexual
    satisfaction
    (T4)(H)

 6. Sexual
    attraction
    (T4)(H)

 7. Frequency of
    intercourse
    (T7)(W)

 8. Sexual
    satisfaction
    (T7)(W)

 9. Sexual
    attraction
    (T7)(W)

10. Spouse
    salience
    (T7)(W)

11. Hours worked
    per week
    (T7)(W)

12. Work salience
    (T7)(W)

13. Work role
    quality
    (T7)(W)

14. Sexual
    attraction
    (T7)(H)

15. Spouse
    salience
    (T7)(H)

16. Hours worked
    per week
    (T7)(H)

17. Work salience    1.0
    (T7)(H)

18. Work role         .57(**)   1.0
    quality
    (T7)(H)

Mean                 5.37       1.15

Standard deviation   7.79       1.04

Note. T4 = variable measured when children were 12 months old,
T7 = variable measured when children were 4.5 years old.

(W) = Wife's report.

(H) = Husband's report

(*) = p [is less than] .05.

(**) = p [is less than] .01.

One concern is that a substantial segment of the original sample did not, for any of several reasons, complete the sexuality questionnaire at T7 and therefore the sample with complete data might be selective in regard to their sexual behaviors. To address this concern, we compared those who completed the T7 sex questionnaire (n = 299) with those who did not (n = 279) on their responses to the sexuality questionnaire at T4. A two-group MANOVA was run, with four dependent variables: husband's report of frequency of sex in the past month, husband's report of satisfaction with the sexual relationship, wife's report of frequency of sex in the past month, and wife's report of satisfaction with the sexual relationship. The MANOVA was not significant, multivariate F(4, 357) = 0.88, and neither were any of the univariate ANOVAs for differences between the two groups. Chi-square tests were used to test for the relationship between completing or not completing the T7 sexuality questionnaire and three dichotomous sexuality variables at T4: whether they had engaged in petting, cunnilingus, or fellatio. According to wives' reports, those who completed the T7 questionnaire were significantly more likely to have engaged in cunnilingus (54%) and fellatio (52%) than those who did not complete the T7 questionnaire (41% and 40%, respectively). Therefore, the subsample seems to be representative of the entire sample on measures of intercourse frequency and sexual satisfaction, but to be slightly more experienced (and perhaps comfortable with) the somewhat more sensitive behaviors of oral-genital sex.

The first hypothesis concerns differences in sexuality as a function of the wife's employment status. Table 6 shows a comparison of the women's four employment groups: homemakers, employed part time (6-31 hrs/wk), employed full time (32-44 hrs/wk), and employed high full time ([is greater than] 44 hrs/wk), for the sexuality outcomes. For the four dichotomous variables (had sexual intercourse, masturbated, engaged in fellatio, engaged in cunnilingus), chi-square analyses were used to test for the association between employment group and yes versus no responses. None of the chi-square tests was significant. For the continuous variables, a 4-group MANOVA was run for differences among the four employment groups, with four dependent variables: frequency of intercourse, satisfaction with sex, her sexual attraction to her husband, and his report of his sexual attraction to her. The overall MANOVA was not significant, multivariate F(12, 585) = 0.74. Neither were any of the univariate F's significant.

Table 6. Comparisons of Wives Who are Homemakers and
Those Employed, and Husbands Employed for Sexuality

                                             Employed

                                      Homemakers     Part time
Sexuality variable                     (n = 40)      (n = 81)

Wives

Had sexual intercourse                   92%           92%
Masturbated                              14%           26%
Engaged in fellatio                      43%           48%
Engaged in cunnilingus                   41%           49%
Frequency of intercourse              5.65 (5.29)   4.53 (3.60)
Satisfaction with sex                 3.63 (1.33)   3.62 (1.29)
Wife's sex attraction to husband      2.82  (.68)   2.70  (.70)
Husband's sex attraction to wife      2.64  (.66)   2.61  (.67)

Husbands

Frequency of intercourse                            4.60 (3.98)
Wife's satisfaction with sexual
 relationship                                       3.61 (1.25)
Husband's sexual attraction to wife                 2.65  (.66)
Wife's sexual attraction to husband                 2.67  (.62)

                                              Employed

                                                    High full
                                      Full time        time
Sexuality variable                     (n = 56)      (n = 23)

Wives

Had sexual intercourse                   99%           86%
Masturbated                              20%           28%
Engaged in fellatio                      48%           52%
Engaged in cunnilingus                   47%           45%
Frequency of intercourse              5.82 (3.49)   4.39 (4.56)
Satisfaction with sex                 3.80 (1.17)   3.17 (1.07)
Wife's sex attraction to husband      2.85  (.66)   2.57  (.52)
Husband's sex attraction to wife      2.72  (.68)   2.57  (.67)

Husbands

Frequency of intercourse              5.51 (4.23)
Wife's satisfaction with sexual
 relationship                         3.63 (1.26)
Husband's sexual attraction to wife   2.62  (.69)
Wife's sexual attraction to husband   2.83  (.68)

Note. None of the tests for significance of differences between
groups were significant. All reports are for the past month. For
continuous variables, means are provided, with standard deviations
in parentheses. Wives: Homemakers, n = 40; Part-time employed,
n = 81; Full-time employed, n = 56; High full-time employed,
n = 23. Husbands: Full-time employed, n = 106; High full-time
employed, n = 93.

One final dichotomous variable was loss of sexual interest. To test for an association between this outcome and women's employment status, a 4 (employment group) x 2 (loss of interest: yes/no) chi square was run. The result showed a marginally significant association, [chi square](3) = 7.01, p [is less than] .10. Inspection of the cell percentages indicated a fairly strong trend that was obscured because of small n's in several cells. Therefore, women's employment categories were collapsed into two categories: homemakers and employed women. The test then indicated a significant association, [chi square](1) = 6.97, p [is less than] .01. Full-time homemakers (21%) were more likely than employed women (8%) to have experienced loss of interest in sex.

To address our second major question, differences between husbands employed full time and high full-time were tested using a 2-group MANOVA with four dependent variables: frequency of intercourse, wife's report of her satisfaction with the sexual relationship, husband's report of his sexual attraction to his wife, and wife's report of her sexual attraction to her husband. The overall multivariate F was not significant, F(4, 194) = 1.48. Neither was any of the univariate tests.

Our third and fourth questions concerned whether, in multiple regression analyses, work hours or work-role quality would be the better predictor of sexual outcomes, and whether work or spouse salience would be important predictors. To address these questions, four hierarchical multiple regressions were computed, one for each of the following outcome variables: couple's frequency of intercourse (average of husband's and wife's reports), wife's sexual satisfaction, wife's sexual attraction to spouse, and husband's sexual attraction to spouse. In each case, three variables were entered first as controls: wife's education, husband's education, and family income. In the next step, wife's hours worked (WHW), wife's work salience (WWK), and wife's spouse salience (WSP) were entered. In the next step, husband's hours worked (HHW), husband's work salience (HWK), and husband's spouse salience (HSP) were entered. Next, wife's work-role quality (WWQ) and husband's work-role quality (HWQ) were entered. Finally, to address the fifth question, each of the following interactions was entered by itself as a final step: HWK x HSP, WWK x WSP, HWK x WWK, HSP x WSP, HWH x HHW, and WWQ x HWQ. All variables were standardized into z scores before the regressions were computed. Note that the interaction HHW x WHW addresses the couple-level analysis by asking whether wife's and husband's work hours interact in predicting sexual outcomes. All interactions were graphed based on values of the predictor variables set at one standard deviation above and below the mean. The regression results are shown in Table 7.

Table 7. Multiple Regression Analyses Predictions,
WMLH Data Set

                                             [Beta]

                                     Frequency       Wife's
                                        of           sexual
Step and Predictor                  Intercourse   satisfaction

Wife

  Wife's hours worked (WHW)           .02          -.08
  Wife's work salience (WWK)          .02          -.00
  Wife's spouse salience (WSP)        .09           .35(**)
    Adjusted [R.sup.2]                .07           .22(***)

Husband

  Husband's hours worked (HHW)        .03           .06
  Husband's work salience (HWK)       .00          -.16
  Husband's spouse salience (HSP)     .25(**)       .26(**)
    Adjusted [R.sup.2]                .11(*)        .28(*)

Work-role quality

  Wife (WWQ)                         -.05           .07
  Husband (HWQ)                      -.26(**)      -.05
    Adjusted [R.sup.2]                .14(*)        .27

HWK x HSP                             .04           .07
WWK x WSP                            -.12          -.16
HWK x WWK                            -.11          -.03
HSP x WSP                             .07           .09
WHW x HHW                             .09          -.03
WWQ x HWQ                            -.07          -.04
[R.sup.2]                             .20           .33
Adjusted [R.sup.2]                    .14           .27

                                                [Beta]

                                    Wife's sexual   Husband's sexual
                                     attraction        attraction
Step and Predictor                   to spouse         to spouse

Wife

  Wife's hours worked (WHW)            .00               .03
  Wife's work salience (WWK)           .03              -.01
  Wife's spouse salience (WSP)         .47(**)           .06
    Adjusted [R.sup.2]                 .28(**)           .06(**)

Husband

  Husband's hours worked (HHW)         .08              -.09
  Husband's work salience (HWK)       -.06              -.01
  Husband's spouse salience (HSP)      .24(**)           .42(**)
    Adjusted [R.sup.2]                 .32(**)           .21(**)

Work-role quality

  Wife (WWQ)                           .08               .02
  Husband (HWQ)                       -.10              -.11
    Adjusted [R.sup.2]                 .32               .21

HWK x HSP                              .07              -.09
WWK x WSP                             -.01               .11
HWK x WWK                             -.02              -.12
HSP x WSP                              .15(*)            .32(**)
WHW x HHW                             -.09              -.08
WWQ x HWQ                              .00              -.25(**)
[R.sup.2]                              .37               .27
Adjusted [R.sup.2]                     .32               .21

Note. Betas and [R.sup.2] are shown for the last step before the
addition of interactions terms. The following variables were
controlled by entry at Step 1: wife's education, husband's
education, and family income. For frequency of intercourse, these
variables were significant, adjusted [R.sup.2] = .07; family
income was negatively associated with frequency. For wife's sexual
satisfaction, adjusted [R.sup.2] = .05; family income was
significantly negatively associated with satisfaction. The control
variables were not significant in the other analyses.

(*) p [is less than] .05,

(**) p [is less than] .01,

(***) p [is less than] .001

When frequency of intercourse was the outcome variable, 14% of the variance (adjusted) was predicted. Neither husbands' nor wives' hours at work were significant predictors. Instead, the husband's spouse salience was a significant positive predictor. Unexpectedly, husband's work-role quality was a negative predictor. That is, lower work-role quality was associated with greater frequency of intercourse.

When wife's sexual satisfaction was the outcome variable, 27% of the variance (adjusted) was accounted for. Consistent with identity theory, both wife's spouse salience and husband's spouse salience were significant positive predictors. One interaction was significant: wife's work salience x wife's spouse salience. This interaction is shown in Figure 1. Higher spouse salience was associated with greater sexual satisfaction for both high and low work salience, but the relationship was stronger for those who were low in work salience.

[ILLUSTRATION OMITTED]

When wife's sexual attraction to her spouse was the outcome variable, 32% of the variance (adjusted) was accounted for. Again, neither husband's nor wife's work hours were significant predictors. Wife's spouse salience was a significant positive predictor, as was husband's spouse salience. There was a significant interaction between husband's spouse salience and wife's spouse salience (see Figure 2). Wives' sexual attraction to the husband was higher when the wife was high in spouse salience, regardless of the husband's level. This relationship was stronger when the husband was also high in spouse salience.

[ILLUSTRATION OMITTED]

When the husband's sexual attraction to his spouse was the outcome variable, 21% of the variance (adjusted) was accounted for. Again, neither husband's nor wife's work hours were significant predictors. Husband's spouse salience was a significant positive predictor, and two interactions were significant: husband's spouse salience x wife's spouse salience and wife's work-role quality x husband's work-role quality. As shown in the top portion of Figure 3, when husbands are low in spouse salience, their sexual attraction to their wives is low, regardless of the level of the wife's spouse salience. When husbands are high in spouse salience, in contrast, higher wife's spouse salience is associated with greater attraction. When both partners are high in spouse salience, husbands' sexual attraction to their wives is greatest. As shown in the bottom portion of Figure 3, when husbands are high in work-role quality and wives are low in work-role quality, husbands' attraction to their wives is greatest.

[ILLUSTRATION OMITTED]

Discussion

These results provide no evidence for the belief that the sexuality of dual-earner couples suffers. Even when husbands or wives were employed high full time, significant sexual deficits did not appear, over a variety of measures. Multiple regression analyses indicated a similar conclusion: Hours worked were not associated with sexual outcomes.

Identity theory predicts that one will enact more salient roles more often, and therefore experience the rewards associated with that role more frequently. Consistent with identity theory, spouse salience was significantly positively associated with sexual outcomes. In different regression analyses, both husbands' and wives' spouse salience were significant. Moreover, husbands' spouse salience and wives' spouse salience interacted in two regression analyses. In both cases, the combination of high husbands' and wives' spouse salience was associated with particularly positive sexual outcomes. These findings are especially impressive because, in several cases, they represent significant predictions from a variable reported by one member of the couple (e.g., the husband's report of his spouse salience) to a variable reported by the other member (e.g., the wife's report of frequency of intercourse). Based on identity theory, we also predicted an interaction between the individual's spouse salience and work salience, with the combination of high scores on both leading to negative outcomes for the couple. The interaction was significant in one of eight tests, but the pattern of the interaction did not reveal a negative sexual outcome for those high in both work and spouse salience.

As predicted, work-role quality was a more important predictor (both as a main effect and as an interaction) than hours worked, although role quality was a less consistent predictor than spouse salience. Unexpectedly, husband's work-role quality was negatively associated with frequency of intercourse. Given that husband's work-role quality was not a significant main effect in the other three regression analyses, and that wife's work-role quality was not significant in any regression analysis, it seems likely that the negative association was a spurious finding. The interactive effects of work role quality were found more consistently across different analyses.

GENERAL DISCUSSION

Using two distinct and large data sets, this research found clear evidence that the sexuality of dual-earner couples does not suffer compared with others. These results are consistent with those of Hyde et al. (1998). Again, the results point to other factors, such as spouse salience, as having the important associations with sexual outcomes. Neither did workaholism--captured by high work hours and high work salience, or an interaction between work hours and work salience--appear to be related to negative sexual outcomes.

Theoretical Implications

Three theoretical models framed the current research: the scarcity hypothesis, the enhancement hypothesis, and identity theory. The scarcity hypothesis received no support in either study. In Study 1, using the NHSLS data set, there were no significant differences between women employed part, full, or high full time, nor were there differences between men employed full or high full time for a variety of sexuality variables. In Study 2, the consistent finding was no difference on a variety of sexuality variables as a function of wives' hours of employment, or of husbands' hours of employment. Even for those couples with one or both partners employed high full time, there was no evidence of sexual deficits.

The enhancement hypothesis in its simple form--that adding the work role should benefit women and their relationships--also was not supported, based on the lack of differences among employment groups. The elaborated version of the enhancement hypothesis, which includes the importance of role quality, did not receive support in multiple regression analyses, probably owing to the inclusion of work salience in the regressions and the positive correlation between work salience and work-role quality. However, simple bivariate correlations indicated a significant positive correlation between wife's work-role quality and sexual attraction to spouse. It would be premature to dismiss the elaborated version of the enhancement hypothesis.

The data in both studies are compatible with a different theoretical approach, in which scarcity and enhancement are both conceptualized as processes that may operate simultaneously. Scarcity processes occur because women and men in dual-earner couples have limits on their time, energy, and attentional capacity, all of which can become overloaded leading to feelings of stress. At the same time, enhancement processes occur because of the benefits of employment (social support, income, opportunities to build a sense of competence). Enhancement processes in turn depend on work-role quality. In a job in which role quality is poor, little enhancement will occur. We conclude, then, that the impact of employment on couples' relationships is best understood within a multiple-process framework that includes scarcity processes, enhancement processes, and role quality.

Identity theory and the importance of identity salience received clear support in Study 2 and could not be tested in Study 1 owing to the absence of an appropriate measure. In Study 2, wife's spouse salience was a significant predictor of her sexual satisfaction and sexual attraction to her spouse. Husband's spouse salience was a significant predictor of frequency of intercourse, his sexual attraction to his spouse, wife's sexual satisfaction, and wife's sexual attraction to her spouse. To the extent that individuals hold "husband" or "wife" to be a salient identity, the relationship--specifically, the sexual aspects of the relationship--appears to benefit. Interestingly, work salience was not a significant negative predictor, as might have been hypothesized from models in which commitment to work and commitment to family are seen as being in competition with each other. Indeed, work salience and spouse salience were significantly positively correlated, for both men (r = .29) and women (r = .21) (see Table 5).

Overall, these findings support a conclusion that the impact of employment on the marital relationship cannot be understood on the basis of a single theory. Rather, a multiple-theory, multiple-process approach is required.

Limitations

It is important to recognize that the designs of these studies were not true experiments. That is, women were not randomly assigned to work part time, full time, or high full time, nor were men randomly assigned to their hours of work. Causal conclusions--that employment influences or does not influence sexuality--are therefore not warranted. For some people, work hours are an economic necessity and not a matter of choice. For others, there is a great deal of self-selection of work hours. Moreover, all the analyses reported here are cross-sectional. Therefore, one cannot sort out whether greater spouse salience leads to better sexual experiences, or better sexual experiences increase spouse salience.

In this research and in most research on dual-earner couples, we say that a woman has taken on an additional role if she is employed. Neglected in this approach is the question of whether the employment role itself consists of one role or multiple roles. A clerk at a grocery store, for example, may have only a single work role, whereas a professor may have several roles: teacher, researcher, advisor to graduate students, chair of the department, and member of the board of a scientific society. Future research should attempt to assess the complexity or multiplicity of roles that constitute the work role.

Applications

How might the results of these studies be applied in couple counseling situations? And how can one resolve the discrepancy between reports by sex therapists and marital therapists that some clients seeking sex therapy are having difficulties because of overwork or workaholism, with the results of these studies, which indicate no effects of hours of employment on sexual outcomes? It seems likely that current popular culture construes dual-earner couples as having difficulties and, in particular, not having enough time for sex. This belief then becomes a convenient attribution. That is, if a couple seek therapy because of a sex problem, overwork is a convenient, nonthreatening, external attribution which can be used to mask more serious problems such as lack of couple communication or negative cognitions about sexuality held by one or both partners. Unfortunately, this attribution may keep a couple from addressing the real issues.

The findings regarding individual differences in identity salience may also lend themselves to applications in couple counseling. The findings suggest that more important than work hours is the salience of the individuals' identity as a spouse. Thus, the issue of commitment to the spouse role warrants attention in couple counseling.

(1) For the NHSLS, homemaker was not a response option for self. Therefore, we were unable to distinguish between homemakers and others working 0 hours per week, such as students and the unemployed. As a result, analyses for NHSLS women do not include homemakers as a category. Men reported whether their wife was a homemaker, permitting comparison of that category in the couple analyses of men's reports.

REFERENCES

Aneshensel, C. S. (1986). Marital and employment role-strain, social support, and depression among adult women. In S. E. Hobfoll (Ed.), Stress, social support, and women (pp. 99-114). New York: Hemisphere.

Barnett, R. C., & Marshall, N. L. (1989). Preliminary manual for the role quality scales. Wellesley, MA: Wellesley Center for Research on Women.

Barnett, R. C., Marshall, N. L., Raudenbush, S. W., & Brennan, R. T. (1993). Gender and the relationship between job experiences and psychological distress: A study of dual-earner couples. Journal of Personality and Social Psychology, 64, 794-806.

Barnett, R. C. & Rivers, C. (1996). She works, he works. New York: HarperSanFrancisco.

Baruch, G. K., & Barnett, R. (1986). Role quality, multiple role involvement, and psychological well-being in midlife women. Journal of Personality and Social Psychology, 51, 578-585.

Baruch, G. K., Biener, L., & Barnett, R. C. (1987). Women and gender in research on work and family stress. American Psychologist, 42, 130-136.

Blumstein, P., & Schwartz, P. (1983). American couples. New York: William Morrow.

Call, V., Sprecher, S., & Schwartz, P. (1995). The incidence and frequency of marital sex in a national sample. Journal of Marriage and the Family, 57, 639-652.

Coser, R. L. (1991). In defense of modernity: Role complexity and individual autonomy. Stanford, CA: Stanford University Press.

Epstein, C. F. (1987). Multiple demands and multiple roles: The conditions of successful management. In F. J. Crosby (Ed.), Spouse, parent, worker: On gender and multiple roles (pp. 23-35). New Haven, CT: Yale University Press.

Glass, J., & Fujimoto, T. (1994). Housework, paid work, and depression among husbands and wives. Journal of Health and Social Behavior, 35, 179-191.

Goode, W. J. (1960). A theory of role strain. American Sociological Review, 25, 483-496.

Greenberger, E., & O'Neil, R. (1993). Spouse, parent, worker: Role commitments and role-related experiences in the construction of adults' well-being. Developmental Psychology, 29, 181-197.

Greenblatt, C. S. (1983). The salience of sexuality in the early years of marriage. Journal of Marriage and the Family, 45, 289-300.

Henderson-King, D. H., & Veroff, J. (1994). Sexual satisfaction and marital well-being in the first years of marriage. Journal of Social and Personal Relationships, 11, 509-534.

Houseknecht, S. K., & Macke, A. S. (1981). Combining marriage and career: The marital adjustment of professional women. Journal of Marriage and the Family, 43, 651-661.

Hyde, J. S., DeLamater, J. D., & Hewitt, E. C. (1998). Sexuality and the dual-earner couple: Multiple roles and sexual functioning. Journal of Family Psychology, 12, 354-368.

Hyde, J. S., Klein, M. H., Essex, M. J., & Clark, R. (1995). Maternity leave and women's mental health. Psychology of Women Quarterly, 19, 257-285.

Klein, M. H., Hyde, J. S., Essex, M. J., & Clark, R. (1998). Maternity leave, role quality, work involvement, and mental health one year after delivery. Psychology of Women Quarterly, 22, 239-266.

Laumann, E. O., Gagnon, J. H., Michael, R. T., & Michaels, S. (1994). The social organization of sexuality: Sexual practices in the United States. Chicago: University of Chicago Press.

Laumann, E. O., Paik, A., & Rosen, R. C. (1999). Sexual dysfunction in the United States: Prevalence and predictors. Journal of the American Medical Association, 281, 537-544.

Linville, P. W. (1985), Self-complexity and affective extremity: Don't put all of your eggs in one cognitive basket. Social Cognition, 3, 94-120.

Linville, P. W. (1987). Self-complexity as a cognitive buffer against stress-related illness and depression. Journal of Personality and Social Psychology, 52, 663-676.

LoPiccolo, J., & Friedman, J. M. (1988). Broad spectrum treatment and low sexual desire: Integration of cognitive, behavioral, and systemic therapy. In S. R. Leiblum & R. C. Rosen (Eds.), Sexual desire disorders (pp. 107-144). New York: Guilford.

Marks, S. R. (1977). Multiple roles and role strain: Some notes on human energy, time and commitment. American Sociological Review, 39, 567-578.

Renshaw, D. (1993, November). Impotence, 1994. Paper presented at the annual meeting of the Society for the Scientific Study of Sexuality, Chicago, IL.

Robins, L. N., Heizer, J. E., Croughan, J., & Ratcliff, K. S. (1981). National Institute of Mental Health Diagnostic Interview Schedule: Its history, characteristics, and validity. Archives of General Psychiatry, 38, 381-389.

Rogers, S. J. (1996). Mothers' work hours and marital quality: Variations by family structure and family size. Journal of Marriage and the Family, 58, 606-617.

Rosenfield, S. (1989). The effects of women's employment: Personal control and sex differences in mental health. Journal of Health and Social Behavior, 30, 77-91.

Sarrel, L., & Sarrel, P. (1984). Sexual turning points: The seven stages of adult sexuality. New York: Macmillan.

Simon, R. W., & Marcussen, K. (1999). Marital transitions, marital beliefs, and mental health. Journal of Health & Social Behavior, 40, 111-125.

Smith, D. S. (1985). Wife employment and marital adjustment: A cumulation of results. Family Relations, 34, 483-490.

Stryker, S., & Serpe, R. T. (1994). Identity salience and psychological centrality: Equivalent, overlapping, or complementary concepts? Social Psychology Quarterly, 57, 16-35.

Stryker, S., & Statham, A. (11985). Symbolic interaction and role theory. In G. Lindzey & E. Aronson (Eds.), The handbook of social psychology, Vol. 1 (3rd ed., pp. 311-378). New York: Random House.

Thoits, P. A. (1983). Multiple identities and psychological well-being: A reformulation and test of the social isolation hypothesis. American Sociological Review, 48, 147-187.

U.S. Bureau of the Census. (1992). Statistical abstract of the United States: 1992. (112th ed.) Washington, DC: Author.

U.S. Bureau of the Census. (1998, Sept. 30). Statistical abstract of the United States: 1998. Washington, DC: Author.

Wethington, E., & Kessler, R. C. (1989). Employment, parental responsibility, and psychological distress: A longitudinal study of married women. Journal of Family Issues, 10, 527-546.

Manuscript accepted September 8, 2000

Janet Shibley Hyde, John D. DeLamater, and Amanda M. Durik University of Wisconsin--Madison

This research was supported by National Institute of Mental Health Grant MH44340 to Janet Hyde and Marilyn Essex. Developmental phases of this work were funded by the University of Wisconsin Graduate School and the Wisconsin Psychiatric Research Institute. Special thanks are extended to Jeff Armstrong, Nancy Smider, and other members of the WMLH staff. Rosalind Barnett and Joseph Pleck were helpful consultants to the project in its early stages. We are especially grateful to the families who participated, giving so much time to this project.

We are also grateful to Ed Laumann, John Gagnon, Robert Michael, and Stuart Michaels for archiving the NHSLS data set so that it is widely available to other investigators for analysis.

Address correspondence to Janet Shibley Hyde, Department of Psychology, University of Wisconsin-Madison, 1202 W. Johnson Street, Madison, WI 53706; e-mail: [e-mail].

Editor Note: The editorial processing of this manuscript was managed by Scott Christopher, Ph.D.; as usual, the reviews were blind.