Journal of Occupational Health Psychology 2004, Vol. 9, No. 1, 28 – 45
Copyright 2004 by the Educational Publishing Foundation 1076-8998/04/$12.00 DOI: 10.1037/1076-89126.96.36.199
Gendered Work Conditions, Health, and Work Outcomes Meg A. Bond, Laura Punnett, Jean L. Pyle, Dianne Cazeca, and Manuela Cooperman University of Massachusetts Lowell This cross-sectional study of nonfaculty university employees examined associations among gendered work conditions (e.g., sexism and discrimination), job demands, and employee job satisfaction and health. Organizational responsiveness and social support were examined as effect modifiers. Comparisons were made by gender and by the male–female ratio in each job category. The relationship of gendered conditions of work to outcomes differed on the basis of respondents’ sex and the job sex ratio. Although the same predictors were hypothesized for job satisfaction, physical health, and psychological distress, there were some differing results. The strongest correlate of job satisfaction was social support; perceived sexism in the workplace also contributed for both men and women. Organizational factors associated with psychological distress differed between female- and male-dominated jobs.
as well as race, age, disability, and sexual orientation. In the workplace, women experience interpersonal sexism in many forms ranging from jokes or comments demeaning women as a group to stereotyping, exclusion of women from a work team or informal work networks, and more direct sabotage or harassment (e.g., Gutek, 2001; Pogrebin & Poole, 1997). Experiences of sex discrimination at work have been associated with more negative relations with coworkers and supervisors, as well as with lower levels of job satisfaction and organizational commitment (Murrell, Olson, & Hanson-Frieze, 1995). Furthermore, women who experience discrimination, negative gender stereotyping, and social isolation tend to experience higher levels of physical and psychological distress (such as high blood pressure, ulcers, tension, anxiety, depression, and sleeplessness) than those who can avoid such experiences (Goldenhar, Swanson, Hurrell, Ruder, & Deddens, 1998; Klonoff, Landrine, & Campbell, 2000). Similarly, sexual harassment has been associated with loss of work motivation and higher levels of distraction that ultimately lead to poor work performance, absenteeism, lateness, and high rates of turnover (Hanisch, 1996; Schneider, Swan, & Fitzgerald, 1997). Considerable research has also found that sexual harassment is associated with negative psychological outcomes such as anxiety, depression, alienation, lower self-esteem, tension, and nervousness and somatic outcomes such as gastrointestinal disturbances, nausea, headaches, and insomnia (Dansky & Kilpatrick, 1997; Goldenhar et al., 1998; Gutek & Done, 2001; Hesson-McInnis & Fitzgerald, 1997). In addition, experiencing sexual harassment at work has been correlated with smoking and alcohol
There is broadening recognition that psychological, social, and structural elements of the work environment can affect individuals’ physical and mental health as well as their work performance and satisfaction (Messing, 2000; Ostlin, 2002). Simultaneously, there is evidence that U.S. workplaces are changing in response to more women entering paid employment at all socioeconomic levels (Blau, Ferber, & Winkler, 2001). Thus we need to understand better how gendered dynamics affect the work environment, how specific physical and psychological job conditions differ by sex, and what impact all of these aspects of the workplace have on the health and satisfaction of workers, both male and female. Past research has documented differential treatment and has shown adverse effects of discrimination, at work and in other settings, on the basis of sex
Meg A. Bond and Manuela Cooperman, Department of Psychology, University of Massachusetts Lowell; Laura Punnett and Dianne Cazeca, Department of Work Environment, University of Massachusetts Lowell; Jean L. Pyle, Department of Regional Economic and Social Development, University of Massachusetts Lowell. All authors are or have been associated with the Center for Women and Work, University of Massachusetts Lowell. Manuela Cooperman is now at the Department of Psychology, Clark University. This research was made possible by grants from the University of Massachusetts Lowell Committee of Federated Centers and Institutes and the University of Massachusetts Lowell Council on Diversity and Pluralism. We thank Lana Tsurikova, Theresa Milewski, Eileen Maloney, and Elizabeth Adler for technical and substantive assistance. Correspondence concerning this article should be addressed to Meg A. Bond, Center for Women and Work and Department of Psychology, University of Massachusetts Lowell, Lowell, MA 01854. E-mail: [email protected]
GENDERED WORK CONDITIONS AND HEALTH
abuse (Richman et al., 1999). Women of color experience the negative effects of both sex- and racerelated discrimination in the workplace (Evans & Herr, 1991; Xu & Leffler, 1996). Hughes and Dodge (1997) surveyed a sample of full-time employed African American women and found that both interpersonal and institutional racism, but particularly interpersonal prejudice, were negatively correlated with job satisfaction. Sexist stereotypes and prejudice shape not only the interpersonal-level dynamics associated with harassment and differential treatment but also the broader organizational and institutional contexts of work (Bond, 2003; Dekker & Barling, 1998; Lott, 1995; Pyle & Bond, 1997) and, at all levels, operate as sources of occupational stress (Korabik, McDonald, & Rosin, 1993). Most past studies have examined the effects of interpersonal discrimination or harassment without considering them in the context of institutional manifestations of differences in the gendered conditions of work, such as occupational segregation by sex, salary inequities, differential levels of job decision latitude, and work schedules that interfere with access to child care. Such factors are also likely direct predictors of worker well-being and job satisfaction, and knowledge of the interrelationships of interpersonal behaviors and institutional practices would greatly enhance interpretation of research findings. To illustrate, Fitzgerald and her associates have found evidence that the job gender context of work, conceptualized as a combination of the sex ratio of the workgroup and the gender traditionality of the work role, is related to the likelihood of experiencing sexual harassment (Fitzgerald, Drasgow, Hulin, Gelfand, & Magley, 1997). This, in turn, was subsequently associated with lower job satisfaction and psychological distress. In addition, perceptions of an organization’s responsiveness to harassment were associated with the occurrence of sexual harassment, implying that potential harassers are more likely to curb their behavior when they anticipate organizational-level repercussions (Fitzgerald et al., 1997; Hesson-McInnis & Fitzgerald, 1997). Research on the gendered patterns in U.S. workplaces reveals considerable de facto occupational sex segregation, both across occupations and within general occupational categories. Women typically work in occupations that are valued less and accorded lower status (Blau et al., 2001; Xu & Leffler, 1996). Within-occupation differences are also generally to women’s disadvantage (e.g., female managers are disproportionately found at lower decision-making levels of their organizations), and women earn less
than men in almost every occupation, even those dominated by women (Blau et al., 2001). They face both “glass ceilings” (wherein promotion above certain levels is rare) and “sticky floors” (wherein women are not able to move beyond entry-level jobs), which has the effect of keeping most women in positions with less power and status than men (Pyle & Bond, 1997, 2002). The relationship between the occupational segregation of men and women and workers’ health status has been of considerable interest because “men’s jobs” and “women’s jobs” often have qualitatively and quantitatively different occupational exposures, whether physical workload, psychosocial strain, or even toxic chemicals (Hall, 1992; Messing, 1995, 1997; Punnett & Herbert, 2000; Quinn, Woskie, & Rosenberg, 2000). An intriguing finding in this regard is the report of increased sick leave for all causes, by both men and women, in jobs with high sex segregation—with the most problematic outcomes occurring for women in male-dominated groups (Alexanderson, Leijon, Akerlind, Rydh, & Bjurulk, 1994). Similarly, Evans and Steptoe (2002) found that women in male-dominated jobs had higher anxiety levels than men in the same job, and that men in female-dominated jobs had more extended sick leave than their female counterparts. Further research is needed to determine whether sex segregation acts as a stressor per se or is a determinant of or a marker for sex differences in other exposures, including sexist attitudes and discrimination. The study of gender influences in the workplace can also be enhanced by attention to specific job content. Domains of the psychosocial work environment typically include constructs such as psychological job demands, decision latitude (job control), social support from coworkers and supervisors, and intrinsic and extrinsic rewards. Considerable research has documented the implications of these job features for physical and mental health as well as work outcomes such as absenteeism or quitting the job (Israel, Baker, Goldenhar, Heaney, & Schurman, 1996; Karasek & Theorell, 1990; Kristensen, 1996; Messing, 1997). Job demands (workload and time pressure) have been associated with psychological, musculoskeletal, and nonspecific (psychosomatic) health symptoms in both men and women (e.g., Cheng, Kawachi, Coakley, Schwartz, & Colditz, 2000; de Jonge, Reuvers, Houtman, Bongers, & Kompier, 2000; Elovainio & Sinervo, 1997; Sainfort, 1991; Spector, 1987). The constructive role of social support has been studied extensively; many have reported that work supports enhance workers’ physical
BOND, PUNNETT, PYLE, CAZECA, AND COOPERMAN
health (e.g., de Jonge et al., 2000; Karasek & Theorell, 1990; Niedhammer & Chea, 2003), psychological well-being (e.g., Bourbonnais, Brisson, Moisan, & Vezina, 1996; Paterniti, Niedhammer, Lang, & Consoli, 2002), and job satisfaction (e.g., de Jonge et al., 2000, 2001). There is evidence that work supports can buffer the negative impact of workplace stress on health outcomes, particularly when there is a match between the nature of the stressors and supports (Cohen & Wills, 1985; Frese, 1999). These job characteristics are unevenly distributed in the working population; for instance, jobs in which women predominate generally have lower decision authority and skill utilization, on average, than men’s jobs (Evans & Steptoe, 2002; Josephson et al., 1999; Karasek & Theorell, 1990; Matthews, Hertzman, Ostry, & Power, 1998; Nordander et al., 1999; Vermeulen & Mustard, 2000). Additionally, they may have impacts that vary by sex; for example, women and people with more feminine traits have been found to benefit more than men from work and family supports (e.g., Beehr, Farmer, Glazer, Gudanowski, & Nair, 2003; Perrewe & Carlson, 2002). In sum, past research has documented that interpersonal biases and organizational values about appropriate gender-related behaviors, as well as direct experience of harassment, can affect health and wellbeing. Alongside these findings, physical and psychosocial job demands are another set of variables that both influence worker well-being and vary by sex or the sex ratio of the job. It is also plausible that perceived work supports influence the relationship between gendered conditions of work (such as sexist attitudes and biased treatment) and personal outcomes. In addition, there is evidence that people’s experiences of the work environment often vary by sex, and that workforce segregation leads to contrasting work conditions that have differential effects on workers. However, the interrelationships among gendered attitudes, discriminatory behaviors, job content, and occupational segregation have been largely ignored. We posit that they are not independent of each other and that research on their common health effects should include all of these domains to better understand their consequences, separately and in combination. To address these issues, we developed an integrated conceptual model based on (a) the theoretical foundation that gender is a powerful differentiator of experiences and outcomes in the workplace and (b) individual empirical links within this model that have been established in past research (Bond, Pyle, & Punnett, 2000). We studied selected elements of that
model (see Figure 1) through a survey of all nonfaculty staff members of a midsize university. We hypothesized that gendered conditions of work (interpersonal and institutional sexism, gender harassment and discrimination, and low organizational responsiveness to discrimination) negatively affect the job satisfaction, physical health, and psychological distress of workers. We further hypothesized that perceived support (peer and supervisory) would act as a protective main effect and as an effect modifier, especially for the negative effects of gendered work conditions on job satisfaction, physical health, and psychological distress. Organizational responsiveness to discrimination was hypothesized to act primarily as an effect modifier because, in the absence of direct experience of discrimination, there would be less opportunity to judge responsiveness. Last, we hypothesized that the relationships between the gendered conditions of work and individual work and health outcomes are more evident for women and for people in female-dominated jobs. To thoroughly explore the model, we also examined selected psychological and physical job demands to assess their role in any relationship between gendered working conditions and individual outcomes.
Method Participants The participants in this study were nonfaculty university employees of all departments and administrative units, including the quasi-independent grants and contracts office. The target group included 703 part-time and full-time workers, of whom 51% were female and 95% were White. Completed questionnaires were received from 208 staff, which represented a 30% response rate. Participation varied by job type: Surveys were returned by 39 clerical staff (32% response), 21 (78%) grants and contracts office staff, 68 (30%) administrative and mid-management employees, 33 (17%) employees supported by grants and contracts, 32 (34%) maintenance and trades personnel, and 15 (41%) university police officers. Most respondents (62%) held nonsupervisory positions, 33% were mid-level managers, and 6% were upper-level managers or department heads. Just under half were in gender-integrated jobs; slightly more were in female-dominated jobs (largely clerical) than maledominated positions (maintenance and police; see Table 1). Among those who reported their gender, women were nearly 85% (n ⫽ 44) of the respondents in female-dominated jobs, 59% (n ⫽ 51) of those in gender-integrated jobs, and 13% (n ⫽ 6) of those in male-dominated jobs. The average age of both men and women was 41 years. About 2% of the respondents self-identified as African American, 6% Asian, 1% Hispanic, 88% White, and 5% as “other.” Over one third of all respondents (41%) had been employed at the university in the some position for more
GENDERED WORK CONDITIONS AND HEALTH
Figure 1. Gendered conditions of work, job demands, social support, and work and health outcomes. VDT ⫽ video display terminal.
than 10 years, whereas 35% had been in their current position from 1 to 5 years.
Procedure University employees were contacted by mail. Along with a paper-and-pencil, self-administered, standardized questionnaire, they received a cover letter that explained the goals of the study and the procedures that would be followed to ensure confidentiality. The packet also included
letters of support from campus union representatives, the survey instrument, and a prestamped return envelope. To encourage responses, we gave participants the opportunity to enter a raffle for tickets to university sports events or plays. Ten days after the first mailing and again a few weeks later, reminder postcards that also announced extensions of the survey deadline were sent to all staff. After completing preliminary analyses, we shared the results in separate meetings with the director of human relations and the director of the grants and contracts office,
as well as in an open meeting to which both union representatives and the broader university community were invited. Our goal was both to communicate results and to gain feedback on emerging questions, possible interpretations, and additional analysis that would be of interest to participants or university administrators.
7 0 Row percentages.
32 15 95 37
Percentages do not sum to 100% because of rounding. a
Total Female dominated Clerical personnel Grants/contracts office staff Gender integrated Administrative Employees supported by grants/contracts Male dominated Maintenance/trades personnel University police
68 33 229 191
39 21 124 27
38 34 2 6
7 22 10 17 84 56
% Non-White Column %
Survey respondents Target population Group
Table 1 Distribution of Participants by Sex, Race, Sex Ratio of Job Group, and Type of Work
BOND, PUNNETT, PYLE, CAZECA, AND COOPERMAN
The survey instrument assessed the following: (a) demographic and occupational descriptors, (b) perceptions of gendered elements of the work environment, (c) job demands, (d) perceived supports, and (e) work satisfaction and physical and mental health status. Except where noted, participants were asked to respond on a 5-point scale that ranged from 1 ⫽ strongly disagree to 5 ⫽ strongly agree. Except where noted, summary scores were calculated as the mean of each subscale. Demographics. Sociodemographic and occupational descriptors included age, sex, ethnicity, type of work, work location, part-time/full-time status, work position, and number of years at the university in their current position and in any position. A three-level variable, job sex ratio, was created using information from the university about the actual proportion of males and females in the job category. If a job was held by more than 60% women across the full university population, it was defined as female dominated. If it was staffed by between 40% and 60% women, it was considered gender integrated, and if more than 60% men, male dominated. Gendered conditions of work. The gender-related aspects of the work environment assessed included (a) perceptions of interpersonal and institutional sexism toward women, as well as institutional racism toward non-Whites, (b) personal experience with gender harassment and discrimination, and (c) organizational responsiveness to discrimination. Items used to assess interpersonal sexism, institutional sexism, and institutional racism were adapted from the short form of the Working Environment Scale. The original scale, designed to assess gendered elements of work climate, has good reliability and validity (Stokes, Riger, & Sullivan, 1995).1 Items were added and adapted to refer to race. Owing to the small number of non-White employees, racial discrimination was assessed only at the institutional level by 7 items, whereas gender discrimination was assessed at both the interpersonal (4 items) and institutional (9 items) levels. A sample item to assess interpersonal sexism is “Some men in my department refer to some women as honey, cutie, sweetheart, or other ‘endearing’ terms.” Sample items to assess institutional discrimination are “Promotions are given without regard to gender [race]” and “The people who run this university are serious about treating men [Whites] and women [non-Whites] equally.” Reliabil-
1 Selected items from the Dual Standards, Sexist Attitudes and Comments, and Informal Socializing subscales were adapted to measure interpersonal sexism. Selected items from the Dual Standards, Family/Career, Global, and Sexist Attitudes and Comments subscales were adapted to measure institutional sexism, and selected items from the Dual Standards and Global subscales were adapted to refer to race/ethnicity to measure institutional racism.
GENDERED WORK CONDITIONS AND HEALTH ity of the combined sexism scale was good (13 items, Cronbach’s ␣ ⫽ .86). This was highly correlated with the institutional racism scale (r ⫽ .64, p ⫽ .01); given the relative ethnic homogeneity of the sample, we used only the sexism variable in multivariate analyses. Personal experiences with gender harassment and discrimination were assessed by asking women and men directly, “Over the last year have YOU been treated unfairly or harassed at the university because of your gender?” Participants responded on a 5-point scale ranging from 1 ⫽ never to 5 ⫽ very frequently. This item was later dichotomized into those who responded “never” and those who responded “ever” because very few reported multiple incidents. Organizational responsiveness to discrimination was conceptualized as the organization’s likelihood of responding to gender and racial discrimination at both the interpersonal and institutional levels. Organizational responsiveness was assessed using two items adapted from the Remediation subscale of Stokes et al.’s (1995) Working Environment Scale and then adding two parallel items developed specifically for this study that refer to race, such as “Other employees would support people who raise issues about the unfair treatment of women [non-White] employees in this institution.” The four-item scale of organizational responsiveness to gender and racial discrimination had a Cronbach ␣ value of .88. Job demands. Scales from the Job Content Questionnaire (JCQ; Karasek, 1985) were used to assess psychological and physical work demands. The JCQ has been widely used and has good psychometric properties. The nine-item subscale for psychological work demands (␣ ⫽ .79) was constructed according to the standard JCQ algorithm. Physical work demands were measured with five items. Three additional items developed by the authors were used to assess specific ergonomic job features. These items asked participants to estimate typical daily number of hours spent working at a video display terminal (VDT), ranging from 0 to more than 8 hr in increments of 2 hr, the degree of hand/wrist repetition (on a scale from 1 ⫽ no repetitive activity to 7 ⫽ constant repetitive activity), and the use of hand/wrist force (on a scale from 1 ⫽ no force to 7 ⫽ extremely high force). The final physical work demands scale (six items, ␣ ⫽ .87) combined the five JCQ items with the item on hand/wrist force (after rescaling from 1–7 to 1–5). Social support. Work supports were assessed using the JCQ scales for coworker support (6 items) and supervisor support (5 items). Each of these includes items of both affective and instrumental support. Employees’ sense of community (SOC) was measured using a 12-item scale developed by Lambert and Hopkins (1995), who conceptualized SOC as a sense of belonging, related to both feeling committed to and perceiving commitment from the organization. A principal-components factor analysis of the social support and SOC items revealed that all 23 items loaded most favorably on a common factor accounting for over 41% of the variance, with an average factor loading of 0.64 (range ⫽ 0.43– 0.79). Thus the variables for supervisor and coworker relations and employees’ SOC were summed for a single measure of social support (␣ ⫽ .93). Work and health outcomes. For job satisfaction, we used four items from the Job Dissatisfaction subscale of the JCQ and one item adapted from Hughes and Dodge (1997).
Participants responded from 1 ⫽ absolutely not to 5 ⫽ definitely yes. The five-item summary measure had high reliability (␣ ⫽ .91). Respondents’ physical health was assessed by the Health Satisfaction subscale of the Retirement Descriptive Index (Smith, Kendall, & Hulin, 1969), a 9-item scale on which respondents’ rate their own health status. An additional global item to examine respondents’ self-rated health status was drawn from the Short Form–36 Scale (Ware, Kosinski, & Keller, 1996). The combined physical health status score included 10 items (␣ ⫽ .89). Respondents’ psychological health status was assessed using nine items from the depression and anxiety subscales of the Mental Health Inventory (MHI; Veit & Ware, 1983). Participants responded on a 5-point scale ranging from 1 ⫽ never to 5 ⫽ all of the time. MHI is reliable and has exhibited stability over a 1-year interval (Veit & Ware, 1983). In this data set, the nine-item psychological distress score had a Cronbach alpha of .92.
Results Descriptive Results Women and men reported similar levels of institutional and interpersonal sexism, organizational responsiveness to gender and racial discrimination, psychological demands, psychological distress, and physical health (see Table 2). Women reported significantly higher levels of social support and job satisfaction (Table 2) but more frequent gender harassment and discrimination than men (see Table 3). Men reported higher whole-body physical demands than women; women spent more hours per day at the VDT. People in female-dominated and gender-integrated jobs experienced lower levels of institutional and interpersonal sexism than those in male-dominated jobs (Table 2). People in female-dominated and gender-integrated jobs also reported higher levels of social support and job satisfaction than those in maledominated jobs. In contrast, job segregation by sex was not associated with organizational responsiveness to gender and racial discrimination, psychological demands, psychological distress and physical health, or the frequency of experiencing gender harassment and discrimination (Table 3). Members of male-dominated jobs reported higher physical demands and fewer hours at the VDT than other workers. Non-White employees reported more experiences of gender harassment and discrimination than White employees (Table 3). White employees reported higher levels of psychological demands (p ⫽ .03) but lower levels of psychological distress (p ⫽ .06) than non-White employees. There were no other statisti-
0.72 0.76 1.02 2.61 0.97 0.77 1.19 0.91 0.78
3.31 2.50 3.62
2.58 3.36 2.14 3.87
3.56** 2.44 3.70
2.56 3.37 1.86** 4.27**
1.11 0.89 0.76
0.67 0.82 0.83 2.39
Women (n ⫽ 101)
2.95** 2.55 3.53
2.62 3.30 2.56** 3.01**
1.27 0.98 0.82
0.80 0.71 1.12 2.73
Men (n ⫽ 82)
3.65a** 2.44 3.66
2.43a** 3.17 1.94a** 4.27a**
1.06 0.99 0.75
0.67 0.77 0.85 2.42
Female dominated (n ⫽ 60)
3.42a** 2.51 3.68
2.49a** 3.44 1.76a** 4.65a**
1.12 0.78 0.77
0.67 0.77 0.71 2.25
Gender integrated (n ⫽ 101)
2.64b** 2.56 3.44
2.95b** 3.41 3.23b** 1.74b**
1.29 1.07 0.84
0.79 0.69 1.01 2.44
Male dominated (n ⫽ 47)
Note. All variables on a scale of 1–5, except video display terminal (VDT), which is on a scale of 0 –10. Means with different subscripts differ significantly at p ⬍ .001 in pairwise comparisons, Wilcoxon test. a N in each cell varies slightly because of missing values. ** p ⬍ .01 (sex: Wilcoxon test; sex ratio of job group: Kruskal Wallis test).
Predictive Sexism Psychological demands Physical demands Daily VDT use Moderating Organizational responsiveness Social support Outcome Job satisfaction Psychological distress Physical health
Overall (N ⫽ 208)a
Table 2 Mean Scores for Predictive, Moderating, and Outcome Variables by Sex and Sex Ratio of Job Group
34 BOND, PUNNETT, PYLE, CAZECA, AND COOPERMAN
GENDERED WORK CONDITIONS AND HEALTH
Table 3 Frequency of Experiencing Gender Harassment or Discrimination Ever Group
Overall Sex Women Men Sex ratio of job group Female dominated Gender integrated Male dominated Race White Non-White
57 96 43
15 23 11
26.3 24.0 25.6
a n in each cell varies slightly because of missing values. † p ⬍ .10. * p ⬍ .05.
cally significant differences in gendered work conditions, job demands, social support, and work and health outcomes between Whites and non-Whites, but statistical power was very limited because so few participants (and university employees) were nonWhite (n ⫽ 25). Several of the independent and moderating variables were correlated with each other. Most notably, there were strong associations among low institutional and interpersonal sexism, high perceived organizational responsiveness, and high social support at work (see Table 4). In addition, high physical de-
mands were negatively correlated with social support (Table 4), probably reflecting a gendered dynamic given the association of high support and low physical demands with female-dominated roles (see Table 2). Among the dependent variables, psychological distress (i.e., anxiety and depression) was negatively correlated with both job satisfaction and perceived physical health (Table 4).
Work and Health Outcomes The full matrix of correlation coefficients was examined for bivariate associations and for collinearity among independent variables. Multivariate relationships were analyzed by linear regression. All independent variables or scales (unless they were strongly correlated with each other) were considered in the analysis of each outcome variable, along with selected interaction terms derived from the underlying conceptual model. Some models were fitted after stratification by sex and by sex ratio of the job, to qualitatively consider whether different relationships existed for women and men or among jobs segregated by sex. Retention of covariates in the models was based primarily on whether they changed the point estimate for other variables by at least 15% (Rothman & Greenland, 1998), although in general p values greater than .15 also led to exclusion. Job satisfaction. Job satisfaction was very strongly and positively correlated with social support at work (Table 4); it was also associated with low institutional and interpersonal sexism, organizational
Table 4 Spearman (Nonparametric) Correlation Coefficients Among Predictive, Moderating, and Outcome Variables (N ⫽ 208) Variable Predictive 1. Sexism 2. Gender discrimination 3. Psychological demands 4. Physical demands 5. Daily VDT use Moderating 6. Organizational responsiveness 7. Social support Outcome 8. Job satisfaction 9. Psychological distress 10. Physical health
.21 .16 —
.31 .13 .23 —
⫺.09 .06 .24 ⫺.22 —
⫺.49 ⫺.23 ⫺.24 ⫺.16 .04
⫺.46 ⫺.17 ⫺.23 ⫺.34 .07
⫺.37 ⫺.18 ⫺.22 ⫺.27 .06
.16 .17 .23 .10 .11
⫺.15 ⫺.10 ⫺.16 ⫺.23 ⫺.04
.26 ⫺.57 —
Note. All absolute values of r ⬎ .14 are statistically significant at p ⬍ .05. N in each cell varies slightly due to missing values. VDT ⫽ video display terminal.
BOND, PUNNETT, PYLE, CAZECA, AND COOPERMAN
responsiveness, low psychological demands, and low physical demands. In multivariate analyses, social support was the strongest single correlate of job satisfaction, explaining over one half of the variance (see Table 5). In fact, this association was so strong that no other variable could be added to the model. The magnitudes of the coefficients for social support were very similar for men and women and among the three strata defined by the sex ratio of the job, ranging from B ⫽ 0.98 (95% confidence interval [CI] ⫽ 0.72 to 1.25) to 1.17 (95% CI ⫽ 0.87 to 1.46). Multivariate models with job satisfaction as the dependent variable could be fitted with low perceived institutional and interpersonal sexism in combination with either high organizational responsiveness (Fitted Model I, Table 5) or low psychological demands (Fitted Model II, Table 5).2 In the gender-stratified models, the negative association between sexism and job satisfaction was found only for female- and maledominated jobs (Models I and II, Table 5). A negative association between sexism and job satisfaction was also found for both men and women. Only in the gender-integrated jobs did sexism not contribute to the model. A positive association between organizational responsiveness to discrimination and job satisfaction was found primarily among women and people in female-dominated jobs (Model I, Table 5). In contrast, the negative association with psychological demands was largely confined to men and to people in male-dominated jobs (Model II, Table 5). The difference in coefficient size for men versus women was smaller than the differences among the groups defined by the sex ratio of the job. One additional multivariate model (Model III, Table 5) could be fitted only among women and participants in femaledominated jobs; it showed job satisfaction to be associated with high social support, with no experience of gender discrimination, and with an additional benefit of social support among those who had experienced discrimination. Psychological distress. Psychological distress had only weak correlations with the independent variables, positive correlations with psychological demands, and negative correlations with organizational responsiveness and social support (Table 4). In separate modeling of psychological distress, the univariate associations were all in the expected directions for institutional and interpersonal sexism (⫹), having experienced gender harassment or discrimination (⫹), organizational responsiveness (–), psychological demands (⫹), and social support (–) (see Table 6). In general, these models could explain much less of the variance in psychological distress
than was achieved for job satisfaction. Because of the correlations among the covariates, most could not be retained in the same regression model. However, a multivariate model was fitted that showed higher psychological distress to be associated with both higher psychological demands and lower social support (Model I, Table 6). An interaction term (with marginal statistical significance) showed that among those who had experienced gender discrimination, organizational responsiveness was associated with decreased psychological distress (Model II, Table 6). In the same model, having experienced gender harassment or discrimination also had an association in the expected direction. In stratified Models I and II (Table 6), the magnitude of the coefficient of gender discrimination on psychological distress was greater among women than men, and especially for respondents in femaledominated jobs. Among men and those in maledominated jobs, the magnitude of the coefficient of psychological demands was greater and social support was not a significant correlate of psychological distress. The interaction was observed only in female-dominated jobs. Physical health. The occupational factors measured in this study explained very little variance in respondents’ perceived physical health. Small crude associations, in the predicted directions, were observed with sexism (–), organizational responsiveness (⫹), psychological and physical job demands (–), and social support (⫹) (see Tables 4 and 7). In multivariate analyses, a model could be fitted that included the joint effects of physical demands, psychological demands, and social support (Model I, Table 7). Stratified analyses showed the same directions of associations for each gender, although statistical power was quite limited. Surprisingly, physical demands were not associated with lower ratings of physical health in male-dominated jobs. A series of exploratory analyses also found that, in female-dominated jobs only, experience of gender discrimination was negatively associated with physical health ratings, with neither social support nor physical demands entering into the same model (Model II, Table 7). (text continues on p. 40)
2 The joint distributions of independent variables prevented fitting models with more than two variables at a time, and neither of these models explained as much variability in job satisfaction as did social support alone.
Note. Tests of statistical significance are computed from standard errors. † p ⬍ .10. * p ⬍ .05. ** p ⬍ .01.
Crude Associations Full data set Full data set Full data set Full data set Full data set Fitted Model I Full data set Women Men Female dominated Gender integrated Male dominated Fitted Model II Full data set Women Men Female dominated Gender integrated Male dominated Fitted Model III Women Female dominated
0.21 0.25 0.12 0.36 0.21 0.04
0.18* 0.21† 0.10 0.32** 0.18 0.04
⫺0.18 ⫺0.17 ⫺0.39 ⫺0.05 ⫺0.04 ⫺0.54
⫺0.12† ⫺0.13 ⫺0.22* ⫺0.04 ⫺0.03 ⫺0.28†
Table 5 Regression Coefficients for Job Satisfaction: Maximum Likelihood Estimates (Unstandardized and Standardized)
Discrimination ⫻ Support GENDERED WORK CONDITIONS AND HEALTH 37
0.04 0.03 0.05 0.05 0.06
0.12 0.19 0.08 0.23 0.09 0.16
0.13 0.19 0.07 0.29 0.09 0.15
204 196 198 206 206
195 94 77 56 96 43
194 94 77 55 96 43
0.85 0.80 0.47 2.65 0.32 0.77
0.22 0.30 0.14 0.87 ⫺0.10 0.06
0.40* 0.42 0.18 1.13** 0.18 0.32
0.11 0.16 0.05 0.38** ⫺0.05 0.03
Note. Tests of statistical significance are computed from standard errors. † p ⬍ .10. * p ⬍ .05. ** p ⬍ .01.
Crude Associations Full data set Full data set Full data set Full data set Full data set Fitted Model I Full data set Women Men Female dominated Gender integrated Male dominated Fitted Model II Full data set Women Men Female dominated Gender integrated Male dominated
Table 6 Regression Coefficients for Psychological Distress: Maximum Likelihood Estimates (Unstandardized and Standardized)
⫺0.22 ⫺0.17 ⫺0.11 ⫺0.58 ⫺0.16 ⫺0.25
⫺0.31† ⫺0.27 ⫺0.13 ⫺0.74* ⫺0.25 ⫺0.31
Discrimination ⫻ Organizational Responsiveness
38 BOND, PUNNETT, PYLE, CAZECA, AND COOPERMAN
0.03 0.01 0.02 0.03 0.04 0.00 0.04
0.07 0.06 0.03 0.08 0.04 0.03
202 193 196 203 204 196 203
202 98 80 58 97 47
⫺0.12† ⫺0.21* ⫺0.03 ⫺0.20 ⫺0.04 ⫺0.22
⫺0.12 ⫺0.19 ⫺0.04 ⫺0.20 ⫺0.04 ⫺0.27 ⫺0.19
Note. Tests of statistical significance are computed from standard errors. VDT ⫽ video display terminal. † p ⬍ .10. * p ⬍ .05. ** p ⬍ .01.
Crude Associations Full data set Full data set Full data set Full data set Full data set Full data set Full data set Fitted Model I Full data set Women Men Female dominated Gender integrated Male dominated Fitted Model II Female dominated
⫺0.10 ⫺0.10 ⫺0.04 ⫺0.10 ⫺0.27 0.10
⫺0.13† ⫺0.09 ⫺0.05 ⫺0.11 ⫺0.23* 0.12
Table 7 Regression Coefficients for Physical Health: Maximum Likelihood Estimates (Unstandardized and Standardized)
Daily VDT use
0.14 0.12 0.21 0.21 0.08 0.17
0.14† 0.12 0.21 0.20 0.07 0.17
GENDERED WORK CONDITIONS AND HEALTH 39
BOND, PUNNETT, PYLE, CAZECA, AND COOPERMAN
Discussion This study of nonfaculty university employees was designed to explore how gendered conditions of work contribute to job satisfaction and to health. The study was undertaken in an exploratory and descriptive way as a preliminary evaluation of our model, given the dearth of literature examining the broad interrelationships among these domains. The key factors of concern were perceived institutional and interpersonal sexism in the workplace and direct experience of gender discrimination. Psychological and physical demands of the job were also considered risk factors. Organizational responsiveness to discrimination and social support in the workplace were included as potential moderating factors. Rather than adjusting for gender, we elected to qualitatively compare these associations between men and women as well as by sex ratio groups, as others have recommended (Mergler, 1995; Mergler, Brabant, Ve´ zina, & Messing, 1987). Sex of the individual and job sex composition are not equivalent, and thus these two sets of stratified analyses are two conceptually different approaches to necessarily overlapping phenomena. Although we had initially hypothesized the same model for job satisfaction, physical health, and psychological distress (see Figure 1), there were different associations with each outcome. We found that the gendered elements of the work context were associated with both job satisfaction and psychological distress, but some covariates differed by sex and some by sex ratio. More specifically, although social support was the strongest single correlate of job satisfaction for all participants in all types of jobs, job satisfaction was also associated with perceived sexism in the workplace, direct experience of gender discrimination, and social support, at least for women and people working in female-dominated jobs. Increased psychological distress was similarly associated with gender discrimination and low organizational responsiveness to discrimination, particularly for those in female-dominated jobs. For men and those in male-dominated jobs, high psychological demands were associated with both decreased job satisfaction and increased psychological distress. This finding contrasts with work by others that indicates that job demands affect mental health among women workers, specifically in clerical and service jobs (Cheng et al., 2000; Elovainio & Sinervo, 1997; Sainfort, 1991; Spector, 1987).
Limitations Several shortcomings of this research are acknowledged. Because this was a cross-sectional study, it was not possible to determine the temporal direction of the observed associations. Our full conceptual model includes a number of recursive relationships, in which, for example, the job satisfaction of individuals might affect the level of social support that they provide to coworkers. The dynamic nature of these processes over time can only be studied with longitudinal research designs. A related issue is that all information except for the sex ratio of the job was collected from individual self-report. This means that information bias cannot be ruled out, especially in a cross-sectional study, in which the work environment was assessed at the same time as the outcomes and thus may have been perceived differently by those experiencing adverse outcomes. However, the different associations for job satisfaction and for psychological health might imply that adversely affected individuals did not simply describe all aspects of the work environment in a uniformly negative manner. Other potential sources of information bias stem from the sensitive and personal nature of many of the study variables. Perceptions of the work environment may be greatly affected by past and present work experiences, including the ubiquity of highly segregated job markets. For example, the economic need to adapt to the work environment may lead people to accept sexist comments from coworkers or view unresponsive organizations as ordinary and not worth remarking on. Potential selection bias is a serious concern. The response rate was perhaps not unusual for a large mailed survey, but it was relatively low for purposes of generalizing from the sample to the entire target population. We have no way of judging whether, and to what degree, the participants might have differed from the nonparticipants in their perceptions of the institution or in their job satisfaction or health status. The healthy worker effect, or differential selection out of the workplace among exposed and affected workers, may also have been operating in this crosssectional population. We were less successful in predicting physical health than mental health or job satisfaction. Our measure of physical health may have been too broad and thus insensitive to specific effects, if any, of the working conditions under study. Unfortunately, because of space limitations, the survey omitted several aspects of the physical work environment (e.g., indoor air quality) that might have been more strongly
GENDERED WORK CONDITIONS AND HEALTH
associated with physical well-being in this population than the few exposures that were assessed. Future research examining more dimensions of exposures and health outcomes could shed more light on whether the effect of sex segregation involves differential physical exposures in addition to discrimination and other psychosocial factors. This study was also limited by sample size and composition. A larger population with more men and women in “nontraditional” jobs would be desirable for better exploration of how the interaction between sex and sex ratio of the job affects individual outcomes. None of the hypothesized effects of racial segregation or discrimination could be explored because there were so few non-White, non-Anglo participants.
Sex, Sex Ratios, and Gendered Conditions of Work This study indicates that gendered conditions of work (i.e., sexism, experiences of discrimination, and organizational responsiveness to discrimination) are associated with important personal outcomes (job satisfaction and psychological distress). A particularly striking theme, within these findings, was that the relationship between work conditions and outcomes differed by respondents’ sex and even more so by the sex ratio of their jobs. In other words, the results illustrate the importance—when analyzing the effects of sexism and discrimination in the workplace— of considering both differences by sex and the different dynamics that can occur in male-dominated, female-dominated, and gender-integrated jobs (cf. Lefkowitz, 1994; Messing et al., 2003). Each of the three “gendered conditions of work” variables were associated with job satisfaction, but their relationships differed by sex and by sex ratio of the job. To elaborate, although workplace sexism was associated with diminished job satisfaction of both men and women, it was associated with lower job satisfaction only in those positions in which there was a gender imbalance (either female dominated or male dominated) and not when there was more equal representation. This is particularly interesting given that our measure of interpersonal sexism focused on biases against women (i.e., not sexist bias against men). These results suggest that in gender-segregated jobs, both men and women might be negatively affected when they feel that the women around them are devalued. Such findings can, perhaps, be understood in light of past research suggesting that gender
imbalances may increase the salience of gender within a group and thus heighten attention to it. This can happen in the form of sex role spillover whereby gender-based expectations are inappropriately carried over onto the work role (Gutek, 1985, 2001). These findings also extend past work by suggesting that the proportion of men or women in a job may not only affect the prevalence of sexual harassment and sexism (e.g., Fitzgerald et al., 1997), but sex ratios of the job may also shape individuals’ reactions (i.e., lowered satisfaction and psychological distress) to such negative experiences. A better understanding of all of the conditions that accompany gender imbalances in jobs should help forward the quest to clarify the conditions under which sexism and discrimination have adverse personal impacts (see Gutek & Koss, 1993). Although sexism emerged as relevant to job satisfaction in both female- and male-dominated jobs, participants described significantly more of it in male-dominated contexts. It seems possible that the specific role of sexism in lowering satisfaction might depend on which sex is in numerical dominance. The dynamics of sexism are likely to be different in female-dominated positions (wherein women may feel that their entire job role is devalued) compared with male-dominated positions (wherein individual women may be devalued but the work itself is not). It is also critical to take into consideration the broader societal context within which being a female in a largely male-dominated job has a quite different meaning (and more negative work consequences) than being a token male in a female-dominated setting (Yoder, 2001). These dynamics warrant more attention in future research with large enough samples to examine Sex ⫻ Sex Ratio of the Job interactions and further distinguish the effect of an individual’s sex from the job’s sex ratio. Further evidence of the differential role of sex ratios emerged in the relationships among gender discrimination, social support, and job satisfaction for women and people in female-dominated job roles. Experiences of discrimination were an important detractor to job satisfaction beyond the influence of sexist attitudes, yet the associations were moderated by support—which was more common in the femaledominated jobs in this study. One implication here is that gender may put women at risk for some negative events (i.e., experiencing gender discrimination), but the job context (e.g., the association here between the sex ratio of the job and social support) may help to soften the impact of discrimination on job satisfaction. It is plausible that perceiving the organization as
BOND, PUNNETT, PYLE, CAZECA, AND COOPERMAN
responsive to gender concerns can also support positive feelings about one’s job even in those contexts in which studies typically find that discrimination is most widespread, that is, for those in female-dominated positions. The complex interactions between risk and protective qualities that accompany gender and gendered work roles deserve further exploration. Although the associations were somewhat different for psychological distress than for job satisfaction, they raise some parallel issues. In both cases, it appears that the impact of discrimination on individual well-being can be moderated by elements of the work context that vary with the gender composition of the job; that is, support and organizational responsiveness appear to be more relevant to well-being among those groups and in those job contexts in which there is more of it. When discrimination was experienced by people in female-dominated jobs, high levels of organizational responsiveness to the discrimination moderated its association with mental health. Social support was associated with decreased psychological distress for women and for people in female-dominated and gender-integrated job roles, which is consistent with some past research indicating that women and those more socialized into “feminine” approaches tend to benefit more from support than men do (Beehr et al., 2003; Perrewe & Carlson, 2002). Future research on social support in the workplace (and the same could be said for future research on organizational responsiveness to discrimination) might do well to examine social support in terms of both its actual existence in the workplace and one’s experience in trying to access it. The significant interaction effects indicate that both support and perceptions of organizational responsiveness might reflect qualities of the work environment that are activated as needed. This view provides a contextual, dynamic way to understand social support and organizational responsiveness rather than describing them as static qualities of a person’s social or organizational context. It is interesting that psychological demands, while not different for men versus women nor based on the sex ratio of the job, were significant correlates of satisfaction and mental health for men and people in male-dominated jobs but not for those in femaledominated jobs. Additionally, for men and people in male-dominated positions, mental health outcomes are associated with specific job demands rather than with gendered conditions of work or the social support available at work. It is possible that normative gender role socialization might make it more socially acceptable for many men to focus on the psycholog-
ical stresses than on the more relational area of coworker and supervisor supports (see Beehr et al., 2003). It would also be worthwhile for future research to consider whether psychological job demands become more relevant when other conditions, such as social supports from peers and supervisors, are low (i.e., in this study, in male-dominated settings), while organizational responsiveness may be particularly relevant in settings in which a group feels that they are collectively devalued (i.e., in this study, female-dominated jobs). In summary, this research demonstrates that workplace sexism and discrimination may affect individuals’ job satisfaction and psychological well-being. It also suggests that the effect of these gendered conditions of work may be shaped by whether the job is male dominated, female dominated, or gender integrated. Sex and sex ratios of the job may influence where sexism and gender discrimination occur, the frequency with which they occur, as well as the availability of important moderating resources such as peer and supervisory support and organizational responsiveness to discrimination. There are important practical implications for organizations from this research. Gender influences are more than differences between women and men; gender expectations, occupational segregation, and differential treatment of individuals all shape the work contexts and supports available to women and men and thus the experiences of both sexes at work. Examining the context of the job in terms of sex ratios may well play a powerful role in better understanding what has previously been incompletely conceptualized as sex based. The relationships among all of the variables we studied are probably more fluid, complex, and reciprocal than our original model was able to capture. There is clearly a need for researchers in a wide range of fields to incorporate gendered elements of the work environment into models of workplace conditions and their effects.
References Alexanderson, K., Leijon, M., Akerlind, I., Rydh, H., & Bjurulk, P. (1994). Epidemiology of sickness and absence in a Swedish county in 1985, 1986, and 1987. Scandinavian Journal of Social Medicine, 22, 27–34. Beehr, T. A., Farmer, S. J., Glazer, S., Gudanowski, D. M., & Nair, V. N. (2003). The enigma of social support and occupational stress: Source congruence and gender role effects. Journal of Occupational Health Psychology 8, 220 –231. Blau, F. D., Ferber, M. A., & Winkler, A. E. (2001). The economics of women, men, and work (4th ed.). Upper Saddle River, NJ: Prentice Hall.
GENDERED WORK CONDITIONS AND HEALTH Bond, M. A. (2003) Prevention of the sexual harassment. In T. Gullotta & M. Bloom (Eds.), Encyclopedia of primary prevention and health promotion (pp. 969 –975). New York: Kluwer. Bond, M. A., Pyle, J. L., & Punnett, L. (2000, November). Discrimination, health, and work outcomes: A model and preliminary data. Paper presented at the meeting of the American Public Health Association, Boston, MA. Bourbonnais, R., Brisson, C., Moisan, J., & Vezina, M. (1996). Job strain and psychological distress in whitecollar workers. Scandinavian Journal of Work, Environment and Health, 22, 139 –145. Cheng, Y., Kawachi, I., Coakley, E. H., Schwartz, J., & Colditz, G. (2000). Association between psychosocial work characteristics and health functioning in American women: Prospective study. British Medical Journal, 320, 1432–1436. Cohen, S., & Wills, T. A. (1985). Stress, social support, and the buffering hypothesis. Psychological Bulletin, 98, 310 –357. Dansky, B., & Kilpatrick, D. (1997). Effects of sexual harassment. In W. O’Donohue (Ed.), Sexual harassment: Theory, research, and treatment (pp. 152–174). Boston: Allyn & Bacon. de Jonge, J., Dormann, C., Janssen, P. P. M., Dollard, M. F., Landeweerd, J. A., & Nijhuis, F. J. N. (2001). Testing reciprocal relationships between job characteristics and psychological well-being: A cross-lagged structural equation model. Journal of Occupational and Organizational Psychology, 74, 29 – 46. de Jonge, J., Reuvers, M. M. E. N., Houtman, I. L. D., Bongers, P. M., & Kompier, M. A. J. (2000). Linear and nonlinear relations between psychosocial job characteristics, subjective outcomes, and sickness absence: Baseline results from SMASH. Journal of Occupational Health Psychology, 5, 256 –268. Dekker, I., & Barling, J. (1998). Personal and organizational predictors of workplace sexual harassment of women by men. Journal of Occupational Health Psychology, 3, 7–18. Elovainio, M., & Sinervo, T. (1997). Psychosocial stressors at work, psychological stress and musculoskeletal symptoms in the care for the elderly. Work & Stress, 11, 351–361. Evans, K. M., & Herr, E. L. (1991). The influence of racism and sexism in the career development of African American women. Journal of Multicultural Counseling and Development, 19, 130 –136. Evans, O., & Steptoe, A. (2002). The contribution of gender-role orientation, work factors and home stressors to psychological well-being and sickness absence in maleand female-dominated occupational groups. Social Science and Medicine, 54, 481– 492. Fitzgerald, L. F., Drasgow, F., Hulin, C. L., Gelfand, M. J., & Magley, V. J. (1997). Antecedents and consequences of sexual harassment in organizations: A test of an integrated model. Journal of Applied Psychology, 82, 578 –589. Frese, M. (1999). Social support as a moderator of the relationship between work stressors and psychological dysfunctioning: A longitudinal study with objective measures. Journal of Occupational Health Psychology, 4, 179 –192. Goldenhar, L., Swanson, N., Hurrell, J., Ruder, A., & Ded-
dens, J. (1998). Stressors and adverse outcomes for female construction workers. Journal of Occupational Health Psychology, 3, 19 –32. Gutek, B. A. (1985). Sex and the workplace: The impact of sexual behavior and harassment on women, men, and organizations. San Francisco: Jossey-Bass. Gutek, B. A. (2001). Women and paid work. Psychology of Women Quarterly, 25, 379 –393. Gutek, B., & Done, R. (2001). Sexual harassment. In R. Unger (Ed.), Handbook of the psychology of women and gender (pp. 367–387). New York: Wiley. Gutek, B. A., & Koss, M. P. (1993). Changed women and changed organizations: Consequences of and coping with sexual harassment. Journal of Vocational Behavior, 42, 28 – 48. Hall, E. M. (1992). Double exposure: The combined impact of the home and work environments on psychosomatic strain in Swedish women and men. International Journal of Health Services, 22, 239 –260. Hanisch, K. A. (1996). An integrated framework for studying the outcomes of sexual harassment: Consequences for individuals and organizations. In M. S. Stockdale (Ed.), Sexual harassment in the workplace: Vol. 5. Perspectives, frontiers, and response strategies (pp. 174 – 198). Thousand Oaks, CA: Sage. Hesson-McInnis, M. S., & Fitzgerald, L. F. (1997). Sexual harassment: A preliminary test of an integrative model. Journal of Applied Social Psychology, 27, 877–901. Hughes, D., & Dodge, M. A. (1997). African American women in the workplace: Relationships between job conditions, racial bias at work, and perceived job quality. American Journal of Community Psychology, 25, 581–599. Israel, B. A., Baker, E. A., Goldenhar, L. M., Heaney, C. A., & Schurman, S. J. (1996). Occupational stress, safety and health: Conceptual framework and principles for effective prevention interventions. Journal of Occupational Health Psychology, 1, 261–286. Josephson, M., Pernold, G., Ahlberg-Hulte´ n, G., Ha¨ renstam, A., Theorell, T., Vingård, E., et al. (1999). Differences in the association between psychosocial work conditions and physical workload in female- and maledominated occupations. American Industrial Hygiene Association Journal, 60, 673– 678. Karasek, R. A. (1985). Job Content Instrument: Questionnaire and user’s guide. Lowell: University of Massachusetts Lowell, Department of Work Environment,. Karasek, R. A., & Theorell, T. (1990). Healthy work: Stress, productivity and the reconstruction of working life. New York: Basic Books. Klonoff, E. A., Landrine, H., & Campbell, R. (2000). Sexist discrimination may account for well-known gender differences in psychiatric symptoms. Psychology of Women Quarterly, 24, 93–99. Korabik, K., McDonald, L. M., & Rosin, H. M. (1993). Stress, coping, and social support among women managers. In B. C. Long & S. E. Kahn (Eds.), Women, work, and coping. A multidisciplinary approach to workplace stress (pp. 133–153). Montreal, Quebec, Canada: McGill-Queen’s University Press. Kristensen, T. S. (1996). Job stress and cardiovascular disease: A theoretical critical review. Journal of Occupational Health Psychology, 1, 246 –260. Lambert, S. J., & Hopkins, K. (1995). Occupational condi-
BOND, PUNNETT, PYLE, CAZECA, AND COOPERMAN
tions and workers’ sense of community: Variations by gender and race. American Journal of Community Psychology, 23, 151–179. Lefkowitz, (1994). Sex-related differences in job attitudes and dispositional variables: Now you see them, . . . . Academy of Management Journal, 37, 323–336. Lott, B. (1995). Distancing from women: Interpersonal sexist discrimination. In B. Lott & D. Maluso (Eds.), The social psychology of interpersonal discrimination (pp. 13– 49). New York: Guilford Press. Matthews, S., Hertzman, C., Ostry, A., & Power, C. (1998). Gender, work roles and psychosocial work characteristics as determinants of health. Social Science and Medicine, 46, 1417–1424. Mergler, D. (1995). Adjusting for gender differences in occupational health studies. In K. Messing, B. Neis, & L. Dumais (Eds.), Invisible: Issues in women’s occupational health/La sante´ des travailleuses (pp. 236 –251). Charlottetown, Prince Edward Island, Canada: Gynergy Books. Mergler, D., Brabant, C., Ve´ zina, N., & Messing, K. (1987). The weaker sex? Men in women’s working conditions report similar health symptoms. Journal of Occupational Medicine, 29, 417– 421. Messing, K. (1995). Chicken or egg: Biological differences and the sexual division of labor. In K. Messing, B. Neis, & L. Dumais (Eds.), Invisible: Issues in women’s occupational health/La sante´ des travailleuses (pp. 177–201). Charlottetown, Prince Edward Island, Canada: Gynergy Books. Messing, K. (1997). Women’s occupational health: A critical review and discussion of current issues. Women and Health, 25, 39 – 68. Messing, K. (2000). Multiple roles and complex exposures: Hard-to-pin-down risks for working women. In M. B. Goldman & M. Hatch (Eds.), Women and health (pp. 455– 462). San Diego, CA: Academic Press. Messing, K., Punnett, L., Bond, M. A., Alexanderson, K., Pyle, J., Zahm, S., et al. (2003). Be the fairest of them all: Challenges and recommendations in the treatment of gender in occupational health research. American Journal of Industrial Medicine 43, 618 – 629. Murrell, A. J., Olson, J. E., & Hanson-Frieze, I. (1995). Sexual harassment and gender discrimination: A longitudinal study of women managers. Journal of Social Issues, 51, 139 –149. Niedhammer, I., & Chea, M. (2003). Psychosocial factors at work and self reported health: Comparative results of cross sectional and prospective analyses of the French GAZEL cohort. Occupational and Environmental Medicine, 60, 509 –515. Nordander, C., Ohlsson, K., Balogh, I., Rylander, L., Pålsson, B., & Skerfving, S. (1999). Fish processing work: The impact of two sex dependent exposure profiles on musculoskeletal health. Occupational and Environmental Medicine, 56, 256 –264. Ostlin, P. (2002). Gender inequalities in health: The significance of work. In S. P. Wamala & J. Lynch (Eds.), Gender and social inequities in health (pp. 43– 65). Lund, Sweden: Studentlitteratur. Paterniti, S., Niedhammer, I., Lang, T., & Consoli, S. M. (2002). Psychosocial factors at work, personality traits and depressive symptoms: Longitudinal results from the
GAZEL study. British Journal of Psychiatry, 181, 111–117. Perrewe, P. L., & Carlson, D. S. (2002). Do men and women benefit from social support equally? Results from a field examination within the work and family context. In D. L. Nelson & R. J. Burke (Eds.), Gender, work stress, and health (pp. 101–114). Washington, DC: American Psychological Association. Pogrebin, M. R., & Poole, E. D. (1997). The sexualized work environment: A look at women jail officers. Prison Journal, 77, 41–57. Punnett, L., & Herbert R. (2000). Work-related musculoskeletal disorders: is there a gender differential, and if so, what does it mean? In M. B. Goldman & M. Hatch (Eds.), Women and health (pp. 474 – 492). San Diego, CA: Academic Press. Pyle, J. L., & Bond, M. A. (1997). Workforce diversity: Emerging interdisciplinary challenges. New Solutions, 7(2), 41⫺57. Pyle, J., & Bond, M. A. (2002). Gender and ethnic divisions in the US labor force. In W. Lazonick (Ed.), The international encyclopedia of business and management: Handbook of economics. London: International Thomson Business Press. Retrieved from http://www.iebm.com Quinn, M. M., Woskie, S. R., & Rosenberg, B. J. (2000). Women and work. In B. S. Levy & D. H. Wegman (Eds.), Occupational health: Recognizing and preventing work-related disease and injury (4th ed., pp. 655– 678). Philadelphia: Lippincott, Williams, & Wilkins. Richman, J. A., Rospenda, K. M., Nawyn, S. J., Flaherty, J. A., Fendrich, M., Drum, M. L. et al. (1999). Sexual harassment and generalized workplace abuse among university employees: Prevalence and mental health correlates. American Journal of Public Health, 89, 358 –363. Rothman, K. J., & Greenland S. (1998). Modern epidemiology (2nd ed.). Philadelphia: Lippincott, Williams, & Wilkins. Sainfort, P. (1991). Stress, job control and other job elements: A study of office workers. International Journal of Industrial Ergonomics, 7, 11–23. Schneider, K. T., Swan, S., & Fitzgerald, L. F. (1997). Job-related and psychological effects of sexual harassment in the workplace: Empirical evidence from two organizations. Journal of Applied Psychology, 8, 223–233. Smith, P. C., Kendall, L., & Hulin, C. L. (1969). The measurement of satisfaction in work and retirement. Chicago: Rand McNally. Spector, P. E. (1987) Interactive effects of perceived control and job stressors on affective reactions and health outcomes for clerical workers. Work & Stress 1, 155–162. Stokes, J., Riger, S., & Sullivan, M. (1995). Measuring perceptions of the working environment for women in corporate settings. Psychology of Women Quarterly, 19, 533–549. Veit, C. T., & Ware, J. H., Jr. (1983). The structure of psychological distress and well-being in general populations. Journal of Consulting and Clinical Psychology, 51, 730 –742. Vermeulen, M., & Mustard, C. (2000). Gender differences in job strain, social support at work, and psychological distress. Journal of Occupational Health Psychology, 5, 428 – 440. Ware, J. E., Kosinski, M., & Keller, S. D. (1996). A 12-Item
GENDERED WORK CONDITIONS AND HEALTH Short-Form Health Survey: Construction of scales and preliminary tests of reliability and validity. Medical Care, 34, 220 –233. Xu, W., & Leffler, A. (1996). Gender and race effects on occupational prestige, segregation, and earnings. In E. Ngan-Ling Chow, D. Wilkinson, & M. Baca Zinn (Eds.), Race, class, and gender: Common bonds, different voices (pp. 107–124). Thousand Oaks, CA: Sage. Yoder, J. (2001, August). Context matters: Understanding
women and work. Presidential address presented at the 109th Annual Convention of the American Psychological Association, San Francisco, CA.
Received May 3, 2002 Revision received August 15, 2003 Accepted August 20, 2003 y