Skip to main content
PMC home page
NIHPA Author Manuscripts logoLink to NIHPA Author Manuscripts
. Author manuscript; available in PMC: 2013 Jun 13.
Published in final edited form as: J Occup Health Psychol. 1996 Jul;1(3):319–329. doi: 10.1037//1076-8998.1.3.319

Perceived Race-Based Discrimination, Employment Status, and Job Stress in a National Sample of Black Women: Implications for Health Outcomes

Vickie M Mays 1, Lerita M Coleman 2, James S Jackson 3
PMCID: PMC3681822  NIHMSID: NIHMS313964  PMID: 9547054

Abstract

Previous research has not systematically examined the relationship of perceived race-based discriminations to labor force participation or job related stresses–problems experienced by Black women. The present study investigated the relative contributions of perceived race-based discriminations and sociodemographic characteristics to employment status and job stress in a national probability sample (the National Survey of Black Americans; J. S. Jackson, 1991) of Black women in the United States. Logit and polychotomous logistic regression analyses revealed that Black women’s current employment status was best explained by sociodemographic measures. In contrast, the combination of perceived discrimination and sociodemographics differentially affects patterns of employment status and perceived job stress in the work environment of Black women. Implications of these findings for the health of African American women are discussed.

As greater numbers of women maintain steady employment and move into nontraditional occupations, more attention is being paid to social and psychological effects resulting from their participation in the labor force. The impact of work on marriage, family, mental health status, life satisfaction, and, in particular, job-related stress has been an important area of study (Crohan, Antonnucci, Adelmann, & Coleman, 1989). Historically, Black1 women have constituted a proportionally large share of the labor force (Evans & Herr, 1991; Malveaux & Wallace, 1987). Much of the research on working Black women, however, has concerned itself with objective labor market indices (e.g., rates of participation, occupation, and salary equity). Little research has examined the psychosocial aspects of perceptions of gender and race discrimination in the labor market experiences of Black women, particularly as they affect entry into the labor force, participation rates (employment status), and job stress in the work environment as a function of perceptions of racial-ethnic discrimination.

Studies examining work and the quality of life consistently document the negative impact of work-related discrimination on women and ethnic group members (Ulbrich, Warheit, & Zimmerman, 1989). Disparities between the occupational mobility and work experience of women and ethnic minorities relative to White men are often cited (Barclay, Fields, & Halpert, 1981; Bowman, 1991; Brown & Ford, 1977; Fernandez, 1981; Ford, 1982, 1983a, 1983b; Moch, 1980; Norman, 1981; Vecchio, 1980; Wallace, 1980). African American women, while coping with the normal pressures of daily work, potentially suffer from social status stressors that are the result of discrimination based on their gender and racial–ethnic status (Bowman, 1991; Cochran & Mays, 1994; Ford, 1983b; Mays, 1985, 1995). It is not known, however, if these various forms of discrimination have a differential impact on their entry, participation rates, or job stresses–problems in the labor market. Job stress has evolved as a major psychosocial risk factor in mental and physical illness (Israel, House, Schuman, Heaney, & Mero, 1989; Kessler, Turner, Blake, & House, 1989; Lennon, 1994; Lennon & Rosenfield, 1992) and is strongly associated with absenteeism, productivity, and employment stability (House, 1981; House & Wells, 1981). Similarly, it is not known whether various types of discrimination are relatively more important than sociodemographic characteristics (e.g., age and education) in explaining the labor market experiences of African American women.

The literature on labor market experience and job stress of Blacks and women indicates that both employment status and presence of job stress in these groups are affected by similar factors. Among those identified are lack of advancement opportunities (Bowman, 1991; Bramwell, 1973; Brown & Ford, 1977; Fernandez, 1981; Fields & Freeman, 1972; Wallace, 1980), lack of challenging work and underutilization of skills (Bowman, 1991; Fernandez, 1981; Ford, 1979; Haw, 1982), lack of control over work environment (Ford, 1982), negative treatment and undesirable work atmosphere (Bowman, 1991; Edwards, 1980; Fernandez, 1981; Ford, 1979), inequities in rewards (Fernandez, 1981; Wallace, 1980), high turnover (Wallace, 1980), and lack of training and personal development (Edwards, 1980; Ford, 1978, 1979, 1982, 1983a; Taylor, 1972). While some components of the occupational histories of Black women might best be explained by basic sociodemographic factors (e.g., education, occupational status, and age), other aspects may be better predicted by their perceptions of discrimination in the labor market.

Becker (1980) examined the relationship between perceived discrimination and labor force participation. He concluded that while perceived discrimination may not directly affect wages, employment level, and stability, these factors may indirectly be linked to Black productivity by work attitudes. In other studies, perceptions of discrimination have been linked to locus of control and labor market experiences. Black men who perceived work outcomes to be contingent on their behavior had more positive feelings about work (Becker & Hills, 1981).

In predominantly White work environments, as a result of their race–ethnicity, Blacks often experience role strain and conflict. In a national study of Black Americans, those who worked in predominantly White work groups perceived race-based discrimination to be a serious problem (Bowman, 1991). When probed, these perceptions included restriction in advancement, underutilization, and bad treatment in the form of less privileges and harder discipline.

For the African American woman faced with the additive effects of both gender and ethnic discrimination (Smith & Stewart, 1983), such a position places them at double jeopardy for experiences of discrimination and job-related stress in me workplace (Gurin & Pruitt, 1978; Mays, 1995; Puryear & Mednick, 1974). These experiences of discrimination and job stress sometimes occur as a function of differential attitudes and expectations that Whites have regarding the skills, authority, and leadership abilities of African American women in the workplace (Adams, 1978; Dumas, 1979; Parker, 1976; Richards & Jaffe, 1972). Specifically identified job stressors of Black women include greater socioemotional demands (Dumas, 1979; Mays, 1995), less structural and informational support for career advancement and upward mobility (Edwards, 1980), and lack of control over their work environment (Ford, 1982). As studies have suggested that women’s career aspirations can be influenced by their beliefs and expectations of opportunity (Evans & Heir, 1991; Smith, 1981; Turner & Turner, 1975), perceptions of racial–ethnic discrimination by African American women could potentially influence their entry into the labor force.

This brief literature review suggests mat employment-related activities (e.g., labor force entry and participation rates, level of productivity, and occupational mobility) may be more than a simple function of structural variables and include perceptions and attitudes about the social system and immediate work environment. Because of the paucity of social psychological literature on Black women and their labor force experiences, many unanswered questions remain. Given the interactive nature of the multiple discrimination Black women face, questions that examine the influence of sociodemographic and perceived discrimination variables on employment status and job problems or job stresses are of particular interest.

While some prior research on labor force participation rates of Black women has included sociodemographic variables (Wallace, 1980), there is not any research, to our knowledge, that has examined the role of perceived discrimination in employment status or job problems-stress. The need for an examination of the relative importance of psychosocial factors in employment is of increasing significance. This is particularly true as more is understood about the negative effects of job problems–stress or worker productivity, efficiency, and absenteeism as well as the costs in related physical and psychological disabilities (Kessler, 1979; Kessler & Cleary, 1980; Kessler & Neighbors, 1986; Mays, 1995; Thoits, 1983; Ulbrich et al., 1989).

In the present study we examined the relationships of sociodemographics and perceived race-based discrimination variables to employment status and job problems–stress within a national probability sample of Black women. Analyses were conducted to explore the contribution of social psychological components of perceived race-based discrimination independently and in combination with sociodemographic characteristics to the variation in employment status and presence of job problems. Two hypotheses were tested within this vein of inquiry. First, are perceptions of race-based discrimination having an impact on the labor force participation of Black women or are sociodemographic variables such as age, education, and the like reasonable predictors of employment status? Second, are job related problems–stressors of African American women better predicted by their sociodemographic characteristics or perceptions of race-based discrimination?

Method

Sample

Data for the analyses were obtained from the National Survey of Black Americans (Jackson, 1991). This data set represents the first full national probability cross-section sample of the entire adult (18 years and older) Black population living in the continental United States (Jackson, 1991). The sample was drawn according to a multistage area probability procedure designed to ensure that every Black household had the same probability of selection. A total of 76 primary areas was selected for interviewing on the basis of the National Census distribution of the Black population. The sites were stratified according to racial composition, and smaller geographic areas (“clusters”) were randomly chosen. Next, professionally trained interviewers listed the habitable households in each cluster. Because correct identification of eligible dwellings was critical, special screening procedures were developed for finding Black households in low-density Black areas. Finally, within each selected Black-occupied household, a single person was randomly chosen from the list of eligible adults for a 2½hr (on average) face-to-face interview that was conducted by a Black interviewer. The survey was conducted over a 7-month period in 1979 and 1980 (Jackson, 1991). These sampling and interviewing procedures resulted in 2,107 completed interviews, representing a response rate of nearly 69%. Of this total, 1,301 of these respondents were female and represent the sample drawn on for the analyses conducted for this study. In the demographic distribution of the female sample, the majority resided in the South (54%), followed by the North (41%) and Western (5%) regions. A full 47% were urban residents; whereas 53% lived in nonurban areas. The respondents ranged in age from 18 to 101 years old, with the median age being 41 years. The women in the sample ranged in their educational level from 44% not completing high school, 33% high school graduates, 15% some college, to 9% college graduates. We classified women’s occupations using the Department of Labor employment categories, and the following results were obtained: 36% of the women were listed as in the service arena, followed by 24% as clerical, 15% as professional, 15% as operatives, 4% as managers, 3% as sales, and 1% as farm laborers.

Instrument

The first step in instrument development was the examination of previous research literature on African Americans to develop a pretest questionnaire that addressed the specific ethnic–racial concerns of African Americans (Jackson, 1991). Work groups, composed of African American students and professionals from a broad cross-section of disciplines, were created. The foci of the various work groups paralleled the scientific issues that emerged from the examination of the research literature and project staff input. Work group topics, which later served as the organizing sections for the construction of the pretest instrument, were (a) identity; (b) consciousness and self-esteem; (c) employment and unemployment; (d) social support including family, religion, coping, help seeking, and mental health resource utilization; and (e) demographics and methodological interviewer–interviewee issues (Jackson, 1991). An advisory board of African American PhDs, also from a broad cross-section of disciplines, was assembled that met before the first pretest to further discuss topic areas and items for inclusion in the pretest instrument.

After topic areas were identified, broad open-ended questions were developed that would allow the investigators to gather input from African American community groups on the meaning and significance of these topic areas to their lives. Focus groups of African Americans homogenous in socioeconomic characteristics were assembled. Each focus group while similar in socioeconomic status varied on a number of status variables such as age, gender, and region of birth in order to obtain a cross-sectional view of the significance and specifics of the topics under review for inclusion in the instrument. Responses from the focus groups were tape recorded, transcribed, and used to write questions for the first version of the pretest instrument. Focus group material allowed questions to be written that utilized language that contained shared meanings of the participants’ experiences (Bowman, 1983; Jackson, 1991; Jackson, Tucker, & Bowman, 1982; Mays & Jackson, 1991; Mays et al., 1992; Word, 1977). This instrument was then pretested on a sample of approximately 200 adults in Detroit, Michigan, and Montgomery, Alabama. Black interviewers were trained by the project staff of the Program for Research on Black Americans and the Institute of Social Research field survey branch. After data collection from the pretest instrument, preliminary analyses accompanied by responses to open-ended questions and interviewer comments were reviewed by the working groups and the advisory board. A second iteration of the instrument was produced and, similar to the first version, was shared with focus groups whose tasks were to comment on the meaningfulness of the language, help refine questions, and ensure its cultural sensitivity and responsiveness to community concerns (Jackson, 1991).

The final instrument used in the face-to-face interviews addressed many important issues of concern to Black Americans. The section on labor market participation was one of the most substantial in the questionnaire. Utilizing information from two pretests, we designed questions not only to gather information regarding groups of working and not working individuals but also to gather information from the discouraged or “hidden unemployed,” who often do not appear in the not working (unemployed) statistics. Questions in the labor market section were often asked of both working and not working respondents with adjustments made for work status. For instance, a working respondent was asked the question, “In the place where you work, do Black people tend to get certain kinds of jobs?” The not working respondent was asked, “In the place where you worked on your last job, did Black people tend to get certain kinds of jobs?”

Measures

Dependent variables

The dependent variables used in this study were employment status and job problems–stress. Every respondent was asked a series of questions to determine her employment status. Working respondents for the analyses included anyone currently employed or laid off but who were expected to return to their jobs in 6 months (n = 753). Respondents who were retired or disabled were removed from the sample to avoid confounding work status with health and age (n = 278).

The measure of job stress was based on a response to the question, “Over the past month or so have you had any job problems?” If the respondent answered yes, then a question followed to determine the degree of distress this problem caused. “How much did that upset you?” With responses varying from not at all to a great deal. A three-level job stress variable (representing no job problems or job-related distress, some job problems and distress, or a great deal of job problems and distress) was employed in the analyses. Contrasts between respondents who were working, laid off less than 6 months ago, and laid off more than 6 months ago on the job stress measure revealed significant differences (p < .01). Therefore, respondents who were laid off were removed from the job stress analyses as a way of controlling for stress as a reaction to job loss.

Predictor variables

Questions were selected from the national data set that addressed perceived race-based discrimination. These questions comprised the 14 items included in a factor analysis (see the Appendix). An iterated principal-components analysis (available through the FACTAN program in OSIRIS, Survey Research Center, 1981) with a varimax rotation was performed on the 14 discrimination items (N = 1,310). The results yielded five factors of which only four received loadings at .40 and above. Two items, “treated badly because of race” and “have more skills and ability for a better job,” did not have significant loadings and were not included in any further analyses. When restricted to four factors, the items accounted for 51% of the total variance. Items loading at .40 or better were used to develop the perceived race-based discrimination indices by summing across items loading on each factor. Factor scores were developed by using Method 1 as described in Comrey and Lee (1992). The four perceived race-based discrimination indices (predictor variables) that emerged from the factor loadings used for the analyses were (a) race discrimination, (b) resource or opportunity barriers, (c) general discrimination-system blame, and (d) general discrimination–past and future orientation. Cronbach’s alpha coefficients (1951) were determined for each index by using the Statistical Package for the Social Sciences (1992) Reliability program.

Factor 1: Resource and opportunity barriers

This index comprised four items that focused on the importance of attributes that keep Black people from obtaining a really good job. These attributes included ability, motivation, training, and family responsibilities (α = .68).

Factor 2: Racial–ethnic discrimination

This index was composed of four items (see the Appendix) related to incidents of racial–ethnic discrimination within the context of employment (α = .63).

Factor 3: General discrimination–system blame

Two items were included in this index of perception of the system (society) as factors responsible for their predicament in life as a Black person (α = .57).

Factor 4: General discrimination–past and future orientation

This two-item index measured a respondent’s sense of past and future racial–ethnic discrimination against Blacks in general (α = .55).

Sociodemographic variables

In addition to the two dependent measures and the predictor variables created from the factor analysis, the analysis involved five sociodemographic variables. These variables were included to determine their independent and joint contribution to employment status and job stress. The variables were age, education, geographic region, urbanicity, and occupational status.

Analytic Method

Both bivariate and multivariate analyses were used in examining the prediction and correlates of employment and job stress in Black women. Because work status was a dichotomous variable (working–not working), its estimated value produced by a model may be interpreted as the probability that an individual woman was employed. Because the dependent variable was categorical, the analytic procedure used was logistic regression (Hosmer & Lemeshow, 1989). A dichotomous regression analysis was performed by using the DREG program in OSIRIS (Hosmer & Lemeshow, 1989). Logistic coefficients were transformed into proportional effects; log-odds was the predicted change in the proportion of cases falling in Category 1 (working) rather than in Category 0 (not working) for a one-unit change in the independent variable. Because the proportional effect varied at different points on the curve, it was necessary to select a single point on the distribution at which to evaluate effect parameters. For the current analyses, the mean of the dependent variable was the point of comparison.

The improvement of goodness of fit of the logistic model can be further evaluated by comparing the log-likelihood statistics. The coefficients were analogous to that of regression coefficients in dummy-variable multiple regression with K-1 categories (Andrews, Morgan, Sonquist, & Klem, 1973).

The effects of perceptions of race-based discrimination and sociodemographics factors on job problems–stress were measured by using a more generalized logistic regression method called polychotomous logistic regression analysis. This method is appropriate for dependent variables that are either nominal or ordinal with discrete values such as 1,2,3, and so forth. The independent variables can be either categorical or continuous. The multiple classification analysis program in OSIRIS was used for this analysis. The dependent variable of job problems–stress was assigned ordinal values. The response “no job problems and distress” was assigned a value of 1, the response “some job problems and distress” was assigned a value of 2, and the response “a great deal of job problems and stress” was coded a value of 3.

Results

Hypothesis 1: Work Status

Hypothesis 1, in which the predictive ability of perceptions of race-based discrimination versus sociodemographic factors was examined, found the latter to be better predictors. Bivariate relationships revealed that work status was negatively related to age and positively related to education. Older respondents, in comparison to younger respondents, were less likely to be working (γ = .19, p < .001). Similarly, the higher a person’s education, the more likely they were to be working (γ = −.39, p< .001).

Table 1 presents the results of the logit analysis for perceived race-based discrimination and sociodemographic variables on work status. The base rate of working respondents was 72% (n = 753). The two models used in this analysis were designed to determine if the sociodemographic and perceived race-based discrimination predictors could improve on this base rate. The full model in which work status was predicted by both perceived race-based discrimination and demographic factors was significant, χ2(16, N =753) = 84.57, p < .001. Further examination of this model revealed mat most of this effect was due to the sociodemographic variables.

Table 1.

Logit Analysis Summary of Employment Status With Sociodemographic and Perceived Race-Based Discrimination Predictors

Predictor β β/SE
Intercept −23.49 −4.25
Perceived race-based discrimination
 Race −0.027 −0.71***
 Barriers −0.321 −1.44
 System blame 0.005 −0.91
 Past-Future −0.067 0.04
Sociodemographics
 Education 0.203 4.34****
 Age 0.045 5.62****
Region
 North −0.092 −0.54
 West 0.420 0.96
Urbanicity −2.54 −0.78
Occupation
 Professional 1.42 2.86***
 Managerial 0.311 0.53
 Sales 0.815 1.43
 Clerical 0.627 2.60***
 Crafts −0.008 −0.180
 Operative 0.328 1.33
 Laborer-Farmer 0.090 0.064
Open in a new tab

Note. Model 1: χ2(4, N = 1,310) = 14.33, p < .01, R2 = .070; Model 2: χ2 (16, N = 1,310) = 87.18, p < .001, R2 = .072.

***

p <.01.

****

p <.001.

A hierarchical method of testing the predictors in the logit model was used. When perceived race-based discrimination variables were entered first (Model 1), they significantly, χ2(4, N = 753) = 14.33, p < .01, reduced the predictive error. However, when perceived race-based discrimination variables were entered after sociodemographics, the former did not significantly improve me prediction, χ2(4, N = 753) = 4.99, ns. In contrast, the demographic variables reduced error significantly, regardless of whether they were entered first, χ2(12, N = 753) = 82.20, p < .001 (predictive error = .031), or second, χ2(12, N = 753) = 72.85, p < .001 (predictive error = .027). Age and education were by far the most significant of the demographic predictors (p < .001).

The measure of goodness of fit indicated the proportion of the predictive error, under the null hypothesis that all coefficients were zero, was accounted for by the predictor variables. The model was a better fit for respondents who were very high in their perception of racial–ethnic and past–future discrimination. Within the occupational category, for which the overall chi-square was not significant, the model was a good fit for professionals and service workers relative to the other occupational categories. On the other hand, although age was a significant factor in decreasing predictive error, the model was a bad fit for older women and those low in race discrimination perceptions.

Hypothesis 2: Job Problems–Stress

In contrast to the findings for Hypothesis 1, the addition of perceived race-based discrimination variables significantly improved the predictions of job problems–stress in Black women’s labor force activities. Among the behavior relationships of job problems–stress and demographic factors, significant chi-square relationships were found between job stress and age (γ = −.46, p < .001), education (γ = .18, p < .01), and occupation (γ = −.12, p < .05). The greatest job stress was among the 24–29 age group, whereas those 48 years and older reported very few job problems that caused stress. As the level of education increased, the likelihood of job problems increased. A weak but significant gamma indicated that professionals, crafts, and service occupation employees experienced fewer job problems–stress, whereas other occupations, particularly sales, experienced more job problems with greater stress.

Urban status resulted in slightly more job problems. In addition, regional differences showed that respondents in the South had the lowest and the West the highest job problems–stress (γ = .12, p < .05). Both region and urban status effects may be an artifact of sampling; nonresponse rates were slightly higher in urban areas and nonsouth regions.

The perceived race-based discrimination measures in all instances, except system blame, yielded significant positive relationships with job problems–stress. The greater the perception of race–ethnic discrimination (γ = .37, p < .001), system blame (γ = .31, p < .001), and past–future discrimination (γ = .20, p<.01) the higher the reported job problems–stress. The gamma for barriers showed a weak relationship in the same direction. As the perception of system blame increased, there was a tendency for job problems to increase, though this relationship was not perfectly linear. An alternative interpretation is that job problems encourage perceptions of race–ethnic discrimination, system blame, and past–future discrimination. Direction of causality, however, cannot be determined from the analyses presented.

Table 2 presents the results of the dummy-variable multiple regression models of job problems-stress (n = 527). The sociodemographic model with a R2 of .02 revealed no significant coefficients. The perceived race-based discrimination model revealed that race–ethnic (p < .01) and system blame discriminations (p < .05) had significant coefficients. The overall model was significant, p < .001, with R2 = .04. The R2(.06, p < .01) increased slightly in the combined model; the significant coefficients were barriers (p < .05), race (p < .10), past–future discrimination, and system blame (p < .04). Tests of significance for the increment in squared multiple correlations contributed by variables left out of the reduced model were performed for both the perceived race-based discrimination and sociodemographic model relative to that full model. Tests for the effects of the contribution of perceived race-based discrimination factors were significant, F(5, 509) = 3.68. Unlike the analysis of employment status, the combined prediction model for job problems–stress was enhanced by the addition of perceived race-based discrimination variables.

Table 2.

Polychotomous Logistic Regression Analysis Summaries of Sociodemographics, Perceived Race-Based Discrimination, and Combined Job Stress Models for Working Black Women

Predictor Sociodemographic model coefficient Perceived discrimination model coefficient Combined model coefficient
Perceived race-based discrimination
 Race .113**** .105*
 Barriers .028 .032**
 System blame .048** .050**
 Past-Future .520 .061**
Sociodemographics
 Education .000 .000
 Age −.002 −.001
Region
 North −.012 .025
 West .013 −.001
Occupation
 Professional .052 .016
 Manager .026 .111
 Sales .249* .243*
 Clerical .070 −.035
 Crafts .138 −.125
 Operator −.012 .038
 Laborer-Farmer .104 .197
 Service −.052 .001
Urbanicity .044 .017
R = .02 R = .04**** R = .06**
Open in a new tab

Note. Three of the predictor variables in this analysis are represented by dummy variables. The excluded categories were service workers for occupation, rural for urbanicity, and South for region.

*

p<10.

**

p<.05.

****

p<.001.

Discussion

Results of this study support earlier work by Becker (1980) and Wallace (1980) indicating that work status among women is best explained by sociodemographic factors, especially age and education. Our results add specificity to the particular importance of age and education in African American women’s entry into me labor market. The decision to enter the labor market in the case of Black women seems more controlled by economic need (Wallace, 1980) and less by attitudes. Black women, if they are old enough, not too elderly, and in good health, are very likely to be in the labor force.2 The significance of age as a predictor of Black women’s labor force participation may be a function of the type of jobs Black women typically occupy in the labor market. Black women relative to White women and men are more likely to enter the work force earlier, depart later, and occupy lower status jobs with greater physical demands and less benefits to cover sickness or disability. This results in a higher level of terminations or loss of jobs because of uncovered illness or disabilities.

Perceptions of discrimination, particularly perceptions of barriers to employment or perceived racial problems on the job, however, have some influence on working Black women (Mays, 1995). Our findings, along with those of other studies, suggest that the effects of perceived race-based discrimination on labor force participation are not on entry into the labor market but on level of participation, as this may be influenced by perceptions of job problems–stress (Becker, 1980; Becker & Hills, 1981; Becker & Krzystofiak, 1982; Cain, 1976). Perceptions of discrimination may not influence the decision of whether to enter the labor market for African American women but rather may affect activities toward advancement, skill development, and interracial interpersonal working relationships. Other studies have documented that African American teenagers and African American men, for example, who perceived themselves to be victims of racial–ethnic discrimination in the labor market and who perceived that their job success was not a function of their efforts but was controlled by factors beyond their reach were more likely to be unemployed (Becker, 1980; Becker & Hills, 1981).

Attitudes of perceived race-based discrimination may play a greater role in influencing success and failure once an individual is employed in the labor market. Black women’s perceptions of discrimination in given occupations have been shown to correlate significantly with actual discrimination in occupational availability (Turner & Turner, 1975). If African American women perceive their work environment to be discriminatory with few avenues for change, then the possibility of less effort directed toward high-level job performance may be an outcome of such attitudinal perceptions. As lower job performance would decrease the likelihood of upward mobility and increased salary, the possibility of job dissatisfaction by both employee and employer would increase.

Other research indicates that unemployed Blacks who report discrimination in past work environments and the perception that their efforts can have little effect on discrimination are more likely to continue to be unemployed than Blacks who do not feel victimized by discrimination (Becker & Hills, 1981). Wallace (1980) noted that the discouragement in employment may come in the form of jobs with low wages, high turnover, and limited opportunities for advancement. Other researchers (Becker, 1980; Fernandez, 1981; Haw, 1982) argued that these same attributes of the job environment may lead to job stress. In fact, the findings of this study contribute further support to Becker’s (1980) argument that perceived race-based discrimination may indirectly be linked to labor force participation by job attitudes. The results of this study modify this notion by suggesting that perceived race-based discrimination is directly linked to job stress. These perceptions may be related to intolerable work environments and thus indirectly linked to employment status.

The perception of specific types of discrimination was significantly related to the experience of job problems–stress in our study. These job problems–stresses were greatest in the young and those with higher levels of education. The results of this study indicate that particular subgroups of Black women, depending on their age and levels of education, are at risk for work-related stressors or job problems based on their perceptions of discrimination. Perceptions of discrimination in the labor market may influence not only motivation and effort on the job but motivation to search for a new job when dissatisfied or unemployed.

Little is known of the effects of job stress on Black women’s productivity, psychological well-being, health, or stability in the labor force. Findings from other studies suggest that working conditions that include underutilization of skills and unpleasant working conditions, job instability, and low wages may be a source of job stress. These stresses may be among the factors that account for the increase in the levels of stress and depression among Black women noted in national health survey statistics (Cochran & Mays, 1994; Mays, 1995).

This study of perceived race-based discrimination, work status, and job stress among Black women has left several unanswered questions and has raised many others. One question not answered in this study is that of the causality of discrimination perceptions. We do not know the extent to which actual experiences of discrimination in the labor market produce perceptions of discrimination and result in job problems versus prelabor market attitudes that create job problems. While it is true that our predictors are not fully independent, future research directed toward determining the etiology of attitudes and the experience of discrimination would be helpful. Regardless of the direction, it is important to develop a more informed perspective on Black women’s job stress because of its relationship to poor mental and physical health, absenteeism, and decreased productivity.

Data from this study clearly indicated that for many Black women if they are not too old or too ill, they participate in the labor market. Of concern is the extent to which participation in the labor market serves as a source of discrimination experiences and consequent job stresses resulting in negative health outcomes for Black women. As growing numbers of studies document the association between race–ethnic-based discrimination and negative health outcomes for African American women (Cochran & Mays, 1994; Krieger, 1990; Mays, 1995), it is important to identify the casual mechanisms of this relationship, perhaps through the use of longitudinal studies. Research that identifies with greater specificity how social status factors such as levels of education, age, and types of occupations modify exposures to health-debilitating employment-related discriminations is needed. Future studies should examine me short- and long-term consequences of such exposures on the overall health and well-being of African American women, especially as participation in the labor market is not a choice but a necessity for most of these women.

Equally as important is the identification of interventions for African American women that prevent or reduce the impact of the psychological distress associated with perceived experiences of employment-related discrimination (Mays, 1995). Research has demonstrated that group-based discussions among Black women of similar employment status on the topic of racism and sexism in the workplace can decrease feelings of job-related stress and increase sources of support (Mays, 1995). Such discussions may even serve a preventive function by providing Black women problem-communicative and problem-solving skills to prevent or lessen race- or gender-based discrimination from having such a detrimental impact on their health and well-being. Learning how to effectively handle powerful others who stereotype and discriminate (Fiske, 1993) may serve a useful, health preventative function. If the goal of decreased morbidity in African American women, as outlined in Healthy People 2000, is to successfully be accomplished, behavioral science must lend its full weight in determining how to eliminate factors, such as race-based discriminations, that are detrimental to the health status of these women in the work environment.

Acknowledgments

Support for this research was from the Center for Minority Group Mental Health, National Institute of Aging, and a National Institute of Mental Health New Investigator Award, received white Vickie M. Mays was at the University of Michigan, Institute of Social Research, Program for Research on Black Americans. Grants from the National Institute of Mental Health and National Institute on Allergy and Infectious Diseases (RO1MH42584; RO1MH44345) are also acknowledged.

We acknowledge with appreciation the assistance of Patricia Gurin, Robert Taylor, Howard Lee, Steve Cardoze, and Laura Klem for statistical consultation and Marilyn Brewer, Barbara Henker, and James House for reading a draft of this article.

Appendix. Factor Analysis of Perceived Race-Based Discrimination Items

Variable Factor 1: Barriers Factor 2: Race Factor 3: System Blame Factor 4: Past-Future
Discrimination 20 years before Civil Rights Movement −.023 .096 .068 .615a
Discrimination 20 years from now .069 −.062 −.014 .510a
Why Blacks don’t do well in life −.006 −.131 −.597a .063
Why Blacks don’t get a good job or education −.000 .114 −.496a −.013
Treated badly because of race −.005 −.325 .090 .048
Have more skills and ability for better job .034 .240 .012 −.025
Not hired because Black −.039 .454a .047 −.100
Family responsibilities kept from getting a good job .438a −.046 .041 .092
Ability kept from getting a good job −.776a −.096 −.071 −.067
Blacks kept from getting a good job −.322 −.397a −.357 .049
Not trying hard enough kept from getting a good job .568a −.011 −.015 .035
Not having education or kept from getting a good job −.686a −.067 .009 −.023
Blacks get certain jobs −.088 .447a −.081 .051
Blacks treated badly at work −.000 −.614a .081 −.008
Percentage of total (variance) 12.188 8.313 5.600 4.827
Percentage of common (variance) 36.055 24.591 16.563 15.279
Open in a new tab
a

Item used to create scale.

Footnotes

1

The terms African American and Black are used interchangeably throughout this article to refer to Americans who share a common ancestral descent from people historically indigenous to sub-Saharan Africa.

2

Women who were not working because of ill health were not counted as unemployed nor included in the analysis. This omission might account for the low levels of unemployment among older Black women.

Contributor Information

Vickie M. Mays, Department of Psychology, University of California, Los Angeles

Lerita M. Coleman, Department of Psychology, University of Colorado

James S. Jackson, Institute for Social Research and School of Public Health, University of Michigan

References

  1. Adams EF. A multivariate study of subordinate perceptions and of attitudes toward minority and majority managers. Journal of Applied Psychology. 1978;63:277–288. [Google Scholar]
  2. Andrews F, Morgan JN, Sonquist JA, Klem L. Multiple classification analysis. 2. Ann Arbor, MI: Institute for Social Research; 1973. [Google Scholar]
  3. Barclay LA, Fields MW, Halpert JA. Perceived differences in job isolation and relationship satisfaction in different racial groups. Paper presented at the 24th annual meeting of the Midwest Academy of Management; Chicago. 1981. [Google Scholar]
  4. Becker BE. Perceived discrimination, work attitudes, and labor market experience. Washington, DC: U.S. Department of Labor, Employment and Training Administration; 1980. [Google Scholar]
  5. Becker BE, Hills SM. Youth attitudes and adult labor market activity. Industrial Relations. 1981;20:60–70. [Google Scholar]
  6. Becker BE, Krzystofiak FJ. The influence of labor market discrimination on locus of control. Journal of Vocational Behavior. 1982;21:60–70. [Google Scholar]
  7. Bowman PJ. Significant involvement and functional relevance: Challenge to survey research. Social Work Research. 1983;9:21–27. [Google Scholar]
  8. Bowman PJ. Work life. In: Jackson JS, editor. Life in Black America. Newbury Park, CA: Sage; 1991. [Google Scholar]
  9. Bramwell J. The Black professional today: The multidimensional human. Contact Magazine. 1973;3:22–24. [Google Scholar]
  10. Brown HA, Ford DL. An exploratory analysis of discrimination in the employment of Black MBA graduates. Journal of Applied Psychology. 1977;62:30–36. [Google Scholar]
  11. Cain GG. The challenge of segmented labor market theories to orthodox theory. Journal of Economic Literature. 1976;14:1222–1257. [Google Scholar]
  12. Cochran SD, Mays VM. Depressive distress among homosexually active African American men and women. American Journal of Psychiatry. 1994;151:524–529. doi: 10.1176/ajp.151.4.524. [DOI] [PMC free article] [PubMed] [Google Scholar]
  13. Comrey AL, Lee HB. A first course in factor analysis. 2. Hillsdale, NJ: Erlbaum; 1992. [Google Scholar]
  14. Crohan SE, Antonnucci TC, Adelmann PK, Coleman LM. Job characteristics and well-being at midlife. Psychology of Women Quarterly. 1989;13:223–235. [Google Scholar]
  15. Cronbach LJ. Coefficient alpha and the internal structure of tests. Psychometrika. 1951;31:297–334. [Google Scholar]
  16. Dumas RG. Dilemmas of Black females in leadership. Journal of Personality and Social Systems. 1979;2:3–14. [Google Scholar]
  17. Edwards KL. Unpublished doctoral dissertation. University of Cincinnati; 1980. The influence of management function and perceived environmental support on perceived stress and job satisfaction of Black females in managerial and professional positions in industry. [Google Scholar]
  18. Evans KM, Herr EL. The influence of racism and sexism in the career development of African American women. Journal of Multicultural Counseling and Development. 1991;19:130–135. [Google Scholar]
  19. Fernandez J. Racism and sexism in corporate life: Changing values in American business. Lexington, MA: Lexington Books; 1981. [Google Scholar]
  20. Fields CL, Freeman ES. Black professionals: The gap is not closing. Masters of Business Association. 1972;6:73, 78, 82–84. [Google Scholar]
  21. Fiske S. Controlling other people: The impact of power on stereotyping. American Psychologist. 1993;48:621–628. doi: 10.1037//0003-066x.48.6.621. [DOI] [PubMed] [Google Scholar]
  22. Ford DL. The Black adult and the world of work. In: Gary L, editor. Mental health: A challenge to the Black community. Philadelphia: Dorrance and Co; 1978. [Google Scholar]
  23. Ford DL. On being culturally disadvantaged: Its impact on the development of job expectancies and valences for job rewards among three ethnic groups. Journal of Social and Behavioral Sciences. 1979;23:26–37. [Google Scholar]
  24. Ford DL. Work, job satisfaction, and employee well-being: An exploration study of minority professionals. Journal of Social and Behavioral Sciences. 1982;26:70–73. [Google Scholar]
  25. Ford DL. The high cost of stress on Black women professionals: A comparative analysis. 1983a. Unpublished manuscript. [Google Scholar]
  26. Ford DL. Job-related stress of the minority professional: An exploratory analysis and suggestions for future research. In: Beehr TA, Bhagah RS, editors. Human stress and cognition in organizations: An integrative perspective. New York: Wiley; 1983b. pp. 287–324. [Google Scholar]
  27. Gurin P, Pruitt A. Conference on the educational and occupational needs of Black women (Rep No UD 018 393) Washington, DC: National Institute of Education Equity Group, Women’s Research Division; 1978. (ERIC Document Reproduction Service No ED 156 768) [Google Scholar]
  28. Haw M. Women, work, and stress: A review and agenda for the future. Journal of Health and Social Behavior. 1982;23:132–144. [PubMed] [Google Scholar]
  29. Hosmer DW, Lemeshow S. Applied logistic regression. New York: Wiley; 1989. [Google Scholar]
  30. House JS. Work stress and social support. Reading, MA: Addison-Wesley; 1981. [Google Scholar]
  31. House JS, Wells JA. Occupational stress, social support, and health; Paper presented at the Conference on Reduction Occupational Stress; White Plains, NY. 1981. [Google Scholar]
  32. Israel BA, House JS, Schuman SJ, Heaney CA, Mero RP. The relationship of personal resources, participation, influences, interpersonal relationships, and coping strategies to occupational stress, job strains, and health: A multivariate analysis. Work and Stress. 1989;3:163–194. [Google Scholar]
  33. Jackson JS. Methodological approach. In: Jackson JS, editor. Life in Black America. Newbury Park, CA: Sage; 1991. [Google Scholar]
  34. Jackson JS, Tucker MB, Bowman PJ. Conceptual and methodological problems in survey research on Black Americans. In: Liu WT, editor. Methodological problems in minority research. Chicago: Pacific/Asian American Mental Health Research Center; 1982. (Reprinted in Jones, R. L. [Ed.], Advances in Black Psychology [pp. 5–29]Richmond, CA: Cobb & Henry) [Google Scholar]
  35. Kessler RC. Stress, social status and psychological distress. Journal of Health and Social Behavior. 1979;20:259–273. [PubMed] [Google Scholar]
  36. Kessler RC, Cleary PD. Stress, social status and psychological distress. American Sociological Review. 1980;45:463–478. [PubMed] [Google Scholar]
  37. Kessler RC, Neighbors HW. A new perspective on relationships among race, social class, and psychological distress. Journal of Health and Social Behavior. 1986;27:107–115. [PubMed] [Google Scholar]
  38. Kessler RC, Turner J, Blake J, House JS. Unemployment, reemployment and emotional functioning in a community sample. American Sociological Review. 1989;54(4):648–657. [Google Scholar]
  39. Krieger N. Racial and gender discrimination: Risk factors for high blood pressure? Social Science Medicine. 1990;30(12):1273–1281. doi: 10.1016/0277-9536(90)90307-e. [DOI] [PubMed] [Google Scholar]
  40. Lennon MC. Women, work and well-being: The importance of work conditions. Journal of Health and Social Behavior. 1994;35:235–247. [PubMed] [Google Scholar]
  41. Lennon MC, Rosenfield S. Women and mental health: The interaction of job and family conditions. Journal of Health and Social Behavior. 1992;33:316–327. [PubMed] [Google Scholar]
  42. Malveaux J, Wallace P. Minority women in the workplace. In: Koziara KS, Moskow MH, Tanner LD, editors. Working women: Past, present and future. Washington, DC: Bureau of National Affairs; 1987. pp. 265–298. [Google Scholar]
  43. Mays VM. Developmental sequence in Black women’s small groups: An absence of pairing. International Journal of Small Group Research. 1985;2(1):94–99. [Google Scholar]
  44. Mays VM. Black women, work, stress and perceived discrimination: The focus support group model as an intervention for stress reduction. Culture and Mental Health. 1995;1(1):53–65. [PMC free article] [PubMed] [Google Scholar]
  45. Mays VM, Cochran SD, Bellinger, Smith RG, Henley NF, Daniels MG, Tibbits T, Victorianne GD, Osei OK, Birt DK. The language of Black gay men’s sexual behavior: Implications for AIDS risk reduction. Journal of Sex Research. 1992;29(3):425–434. doi: 10.1080/00224499209551657. [DOI] [PMC free article] [PubMed] [Google Scholar]
  46. Mays VM, Jackson JS. AIDS survey methodology with Black Americans. Social Science and Medicine. 1991;33(1):47–54. doi: 10.1016/0277-9536(91)90449-m. [DOI] [PMC free article] [PubMed] [Google Scholar]
  47. Moch MK. Racial differences in job satisfaction: Testing four common explanations. Organizational Behavior and Human Performances. 1980;63:299–306. [Google Scholar]
  48. Norman B. Black Enterprise Magazine. 1981. Jul, Career burnout; pp. 42–48. [Google Scholar]
  49. Parker WS. Black–White differences in leader behavior related to subordinates’ reactions. Journal of Applied Psychology. 1976;61:140–147. [PubMed] [Google Scholar]
  50. Puryear G, Mednick M. Black militancy, affective attachment, and the fear of success in Black college women. Journal of Counseling and Consulting Psychology. 1974;42:263–266. [Google Scholar]
  51. Richards AS, Jaffe CL. Blacks supervising Whites: A study of interracial difficulties in working together in a simulated organization. Journal of Applied Psychology. 1972;36:234–240. [Google Scholar]
  52. Smith A, Stewart AJ. Approaches to studying racism and sexism in Black women’s lives. Journal of Social Issues. 1983;39(3):1–15. [Google Scholar]
  53. Smith E. The career development of young Black females: The forgotten group. Youth & Society. 1981;12:277–312. [Google Scholar]
  54. Statistical Package for the Social Sciences. Statistical Package for the Social Sciences users guide. Chicago: SPSS; 1992. [Google Scholar]
  55. Survey Research Center. OSIRIS IV user’s manual. 7. Ann Arbor, MI: University of Michigan, Institute for Social Research, Computer Support Group; 1981. [Google Scholar]
  56. Taylor SA. The Black executive and the corporation: A difficult fit. Masters of Business Association. 1972;8:91–102. [Google Scholar]
  57. Thoits P. Dimensions of life events that influence psychological distress: An evaluation and synthesis of the literature. In: Kaplan H, editor. Psychosocial stress: Trends in theory and research. New York: Academic Press; 1983. pp. 59–81. [Google Scholar]
  58. Turner BF, Turner CB. Race, sex, and perception of opportunity structure among college students. The Social Science Quarterly. 1975;16:345–360. [Google Scholar]
  59. Ulbrich PM, Warheit GJ, Zimmerman RS. Race, socioeconomic status, and psychological distress: An examination of differential vulnerability. Journal of Health and Social Behavior. 1989;30:131–146. [PubMed] [Google Scholar]
  60. Vecchio RP. Worker alienation as a moderator of the quality–job satisfaction relationship: The case of racial differences. Academy of Management Journal. 1980;23:479–486. [Google Scholar]
  61. Wallace PA. Black women in the labor force. Cambridge, MA: MIT Press; 1980. [Google Scholar]
  62. Word CO. Cross-cultural methods of survey research in Black urban areas. Journal of Black Psychology. 1977;3:72–87. [Google Scholar]

RESOURCES