Burke 2

Work Stress and Women’s Health: Occupational Status Effects1

Interest in work stress, employee satisfaction and health has grown considerably over the past two decades. Much of this work has been conducted under a stress-strain framework (Cooper and Payne, 1988) and accumulating research findings have increased our understanding of this complex phenomenon (Schabracq et al., 1996; Cooper, 1996). Most of the published literature on work stress and health is based on the experiences of men (Offermann and Armitage, 1993). Although this situation is gradually changing, women’s health has received relatively little research attention (Nelson and Burke, 2000a, b). Messing (1997) suggests two reasons for the historical neglect of women’s occupational health issues: women’s jobs are safer than men’s and health problems identified among women workers result from their being unfit for the job or unnecessary complaining. With increasing in numbers of women in the labor force, it is critical that more attention be given to understanding the effects of work stress and women’s health. Women may also have different work stress and health issues than men (Langan-Fox, 1998). For example, Collins et al. (1997) suggest that women may be uniquely affected by work conditions (e.g., exposure to chemicals and reproductive health), disproportionately affected (work and family roles) or differently affected (women’s experience of workplace stress). This research examines the relationship of work stress and women’s health, utilizing a diverse sample of women respondents. This is Ronald J. Burke is Professor of Organizational Behavior, School of Business, York University. His research interested focus on organizational restructuring and downsizing.

Ronald J. Burke

important as previous work has emphasized the work experiences of managerial and professional women (see Burke and McKeen, 1994; Burke, 1996). This sample also permitted a preliminary examination of occupational status effects, linking more traditional work stress research with the fields of medical sociology and epidemiology.

Work stressors women face Davidson and Cooper (1983, 1992), in two books on managerial women and stress, found that managerial women felt isolated at work, exhibited Type A behavior, and experienced greater strain than did men. Extra pressures on managerial women included lack of self-confidence and subtle forms of discrimination. The study confirmed the impression that working women still carry the major burden of home and family problems (Hochschild, 1997). Hochschild (1997) estimates, based on major time-use studies, that women in dual career families work an extra month of 24 hour days each year compared to men. This extra time is spent on what she terms “second shift” work, work outside paid employment such as housework, home management, and childcare. Together, these studies suggest that managerial women may experience more stress than men and that the sources of stress are gender-related; that is, related to the expected and actual roles of women in society, and to the fact that, despite progress, executive women still occupy minority status in organizations. There are some stressors, however, that may be particularly important for working women. These include organizational politics, tokenism, barriers to achievement, overload,

Journal of Business Ethics 37: 91–102, 2002. © 2002 Kluwer Academic Publishers. Printed in the Netherlands.

92

Ronald J. Burke

social-sexual behavior, work/home conflict, and organizational restructuring and downsizing.

Occupational status, social class and health Most of the research on work stress and health has involved managerial and professional women. There is, however, an emerging body of work examining social class and health, sometimes including measures of occupational status and work stress, that covers a greater range of occupations. Social class refers to the underlying structure of industrialized societies in which many social and economic characteristics, such as employment conditions, level of pay, housing quality and prestige tend to vary. The term “socio-economic status” refers to social position and blurs the distinction between two concepts: economic circumstances (income and wealth) and prestige (status). Pearlin (1989) suggested that greater vulnerability to stress exists in social roles reflecting “the unequal distribution of resources, opportunities and self-regard” (p. 245). Inequalities in income, occupation and education, all indicators of social class, are recognized as major determinents of individuals mental health (Veenstra, 2000; Humphries and van Doorslaer, 2000; Diez-Roux et al., 2000). Aston and Lavery (1993) collected data from women in managerial or professional occupations and in clerical occupations. Managerial women reported more intrinsic rewards, and extrusic fewer intrinsic concerns, and higher on selfesteem. However no differences were found in depression, quality of life and symptomatology. Kempen et al. (1999) examined the moderating effect of level of education as an indicator of socio-economic status on the relationships between chronic medical morbidity and six domains of health-related quality of life (physical function, role function, social function, health perceptions, bodily pain and mental health) in a large (N = 5279) community-living elderly sample. Level of education was significantly related to each of the other seven measures. In addition, level of education produced addictive

effects in multiple regression analyses with most of the other measures. Borooah (1999) examined the relationship between occupational class and health inequality in large samples of men and women in Britain. Respondents were placed into three categories (professional, managerial or technical; skilled manual or non-manual; semi- or unskilled). Those who were skilled had higher illness rates than those who were skilled who, in turn, had higher illness rate than those occupying managerial professional/technical positions. Bartley et al. (1999a) examined the relationship of two different but highly correlated measures of social position and cardiovascular risk factors. Bartley and his colleagues had two measures of social position, one based on employment relations, the other based on general social advantage and lifestyle. Social inequality has several dimensions, so it is important to use distinct measures of each. It is also vital to clearly specify the hypotheses which link specific dimensions of some position and circumstances to health. They found that for men, the two social position measures were related to most of the behavioral, physiological and psychosomatic risk factors for heart disease: risks were higher in those with less favorable employment conditions and lower levels of general social advantage and living standards. Similar patterns were present for women as well. Muntaner et al. (1998) analyzed data from two large U.S. surveys in a study of social class, assets, organizational control and the prevalence of common groups of psychiatric disorders. In one survey of 8098 respondents, they report a negative relationship between financial and physical assets and mood, anxiety, alcohol and drug disorders. The second survey also revealed a negative relationship between financial and physical assets and anxiety, alcohol and drug disorders. The second survey also showed that lower level supervisors presented higher rates of depression and anxiety disorders than higher level managers. Inequalities in assets and organizational control, as well as commonly used measures of social class, were associated with specific psychiatric disorders. Bartley et al. (1999b), in a sample of British

Work Stress and Women’s Health women, examined the relationship of two indicators of social position with health. One indicator used five categories (professional/administrative, routine non-manual, self employed, skilled manual workers, non-skilled manual workers). The second indicator was a rating of social and material advantage and lifestyle. They found a strong relationship between social position and self-assessed health using data from 1984 and 1993. They found a stronger relationship with health for the general social and material advantage scale than for the employment conditions measure in both years as well.

The present study This study builds on some of the features of contemporary work stress and health research and extends this work in potentially useful ways. Most previous work in this area has involved men; the present study focuses on women’s experiences. In addition, the respondents cover a wider socio-economic and occupational spectrum, with much of the previous work concentrating on professional and managerial women. Finally, the range of work stressors has been expanded to include work stressors more likely to be experienced by women working in lower occupational status positions (e.g., features of the physical work environment such as air quality, temperature) and demands of a physical nature (e.g. bending, standing, lifting). A research framework was developed to guide both the selection of variables and data analysis. In this framework, predictors of women’s satisfaction and health are specified in a particular order. The first block of predictors were personal demographic characteristics (e.g., age, level of education). The second block of predictors comprised work situations characteristics (e.g., size of employer, tenure in current job). These two blocks of predictors included measures previously found to be related to satisfaction and health. They served as control variables before considering the relationship of the third block of predictors, i.e., work stressors, with women’s

93

satisfaction and well-being. The important question was whether the work stressors would have significant relationships with satisfaction and well-being, controlling for individual demographic and work setting characteristics. Measures of work-family conflict were then added as a fourth block, to determine whether these would explain significant increments in variance in the work satisfaction and well-being indicators controlling for individual and work setting characteristics and work stressors. The diversity of the sample made the use of such controls critical in examining the job demands and health relationship. Indicators of satisfaction and well-being were representative of most research in this area and included job satisfaction, psychosomatic symptoms and days of illness.

Procedure Data collection involved two approaches; a survey of nine occupational groups and a series of drop-offs. The survey targeted occupations in which women workers predominated such as health care workers, teachers, childcare workers and sales and advertising managers. These women represented a cross-section of the occupations in which most Ontario women work, as well as some non-traditional occupations, and covered both the private and public sectors as well as unionized and non-unionized workplaces. The survey was sent out to about 13 000 working women using mailing lists from a number of sources. The survey resulted in a response rate of 20.4%, 2564 surveys returned by the closing date. Respondents had two months to complete and return the surveys in pre-addressed, postage paid envelopes. The drop-offs were left in day-care centres and women’s centres across the province where a variety of women could pick-up and drop off surveys at their convenience. Six hundred and seventy surveys were distributed to various centres across Ontario. A total of 121 drop off surveys were returned, a response rate of approximately 18 percent. This approach resulted in a large diverse sample of Ontario working women participating

94

Ronald J. Burke TABLE I Demographic characteristics

Year of birth

N

%

1930–1950 1951–1960 1961–1970 1971–1980

0694 0832 0770 0226

27.5 33.0 30.5 09.0

Current marital Married/partner Separated/Divorced Widowed Never married

1856 0245 0028 0404

72.9 09.6 01.1 15.8

Job title CEO Manager Professional Teacher Health care Childcare Office Sales Non-traditional

0071 0277 0096 0462 0345 0809 0159 0028 0161

03.0 11.5 04.0 19.2 14.3 33.6 06.5 01.2 06.7

Level of education Less than HS Completed HS Some tech/Aat Completed tech/CAAT Some university Completed university Post-grad

0066 0170 0242 0827 0305 0456 0473

02.6 06.7 09.5 32.6 12.0 18.0 18.0

in the study (see Table I). The downside however, is that it was impossible to determine the extent to which those participating constituted a representative sample of those women. In addition, the low response rate (about 18%) suggests that caution be used when interpreting the results.

Racial/Ethnic

N

%

First Nation Black East Asian South Asian Middle Eastern South East Asian European/White

0017 0021 0012 0013 0008 0004 2323

00.7 00.9 00.5 00.5 00.3 00.2 97.9

Number of children 0 1 2 3 4 or more

0582 0398 0839 0308 0128

25.8 17.6 37.2 13.7 05.7

Other dependents Yes No

0186 2227

07.7 92.3

Personal income $10 000 or less $10 000–$29 999 $30 000–$49 999 $50 000–$79 999 $80 000–$100 000 $100 000 or more

0158 0891 0712 0403 0065 0080

06.8 38.6 30.8 17.5 02.5 03.1

Family income Less than $10 000 $10 000–$29 999 $30 000–$49 999 $50 000–$79 999 $80 000–$100 000

0061 0232 0427 00007.6 0334

02.9 11.0 20.3 33.6 15.9

Measures2 Personal demographics Personal demographic characteristics were measured by a number of single items. These included: age, level of education, marital status, number of children, personal income and number of jobs currently had where you work for pay.

Work Stress and Women’s Health Work situation characteristics Work setting characteristics were measured by a number of single items. These included: job tenure, size of employing organization, hours worked per week and

95

3 = About once a week, 5 = Never). Items included: radiation at work (e.g., VDTs, X-rays, microwaves, etc.) and liquid or airborne chemicals such as paints, solvents or gases.

Harassment Work stressors Respondents indicated the extent to which twenty work demands caused them stress at their place of work (α = 0.87). Responses were made on a three-point scale (1 = a major cause, 2 = a minor cause, 3 = not a cause). Items included: unpredictable demands on your time, problems with co-workers, lack of control over what you do, inadequate equipment and problems with supervisors.

Physical demands Respondents indicated how typical ten physical demands were of their job (α = 0.66). Responses were made on a three point scale; 1 = yes, most of the time; 2 = yes, once in a while; 3 = no). Items included: standing for long periods of time, repetitive physical motions; and having limited opportunities to move about.

Job insecurity Job insecurity was measured by three items (α = 0.71). Respondents indicated their agreement with each item on a four-point scale (1 = strongly agree, 4 = strongly disagree. Items included: My job security is good, and My company will downsize in the near future (reversed).

Hazards Respondents indicated how often they had been exposed to three physical health hazards at work over the past month (α = 0.60). Responses were made on a five-point scale (1 = everyday,

Respondents indicated how often each of six incidents had happened to them at work over the past year (α = 0.81). Responses were made on a five-point scale (1 = everyday, 3 = About once a week, 5 = Never). Items included offensive jokes or remarks, being sworn at or yelled at and physical abuse such as being pushed, spit at and so on.

Global work stress Global work stress was measured by two items (α = 0.62). One item assessed how stressful their job was – the things they do; the other item assessed how stressful the work environment is – the people you work with, your supervisor and so on. Responses were made on a four-point scale (1 = very stressful, 4 = not at all stressful).

Psychosomatic symptoms Respondents indicated how often they had each of eight symptoms over the past month (α = 0.74). Responses were made on a five-point scale (1 = every day, 3 = About once a week, 5 = never). Items included stomachaches, back problems, headaches and trouble sleeping.

Global job satisfaction Two areas of global job satisfaction were measured (α = 0.60). One assessed satisfaction with your job – the things you do; the second assessed satisfaction with the environment in which you do your job – the people you work with, your supervisors and so on. Responses were made on a four-point scale (1 = very satisfied, 4 = very dissatisfied).

96

Ronald J. Burke

Work-related psychosomatic symptoms Respondents indicated how often they experienced eight symptoms as a result of work (α = 0.79). Responses were made on a five-point scale (1 = everyday, 3 = About once a week, 5 = Never). Items included stomachaches, back problems, headaches and arthritis/pains in your joints.

Work-family conflict Work-family conflict were assessed by three single items. One item asked respondents to indicate how many hours per week they spent doing things for their household, including cooking, cleaning, grocery shopping, doing laundry and dishes, doing repairs, paying bills, making arrangements and, caring for children. A second item using a four-point scale (1 = often, 4 = never) asked whether family or personal responsibilities/problems at home ever made it difficult for them to do their job. The third item, also using a four-point scale (1 = very easy, 4 = difficult) asked respondents in their job, how easy was it to make arrangements to deal with family or personal responsibilities/problems at home.

Work stressors, work-family conflict and women’s health Hierarchical multiple regression analyses were undertaken in which the indicators of women’s health were separately regressed on the four blocks of predictors. Predictors were entered in a specified order. Table II presents the results of these analyses. Individual measures having independent and significant relationships with particular health outcomes within blocks accounting for significant amounts or increments in explained variance, are also indicated with accompanying βs. The following comments are offered in summary. First, all four blocks of predictors accounted for significant amounts or increments in explained variance on levels of psychosomatic

symptoms reported during the past month. Single women indicated higher levels of psychosomatic symptoms, and those with less formal education reporting more symptoms. Women who had lower personal incomes reported more psychosomatic symptoms; those working more hours per week reported more psychosomatic symptoms. Women reporting more work stressors indicated higher levels of psychosomatic symptoms. Women in more physically demanding jobs also reported more psychosomatic symptoms. And women reporting greater family-work conflict indicated higher levels of psychosomatic symptoms during the past month. Second, two of the four blocks of predictors (personal demographics and work stressors) accounted for significant amounts or increments in explained variance on global job satisfaction. Women reporting more work stressors indicated less job satisfaction. Women reporting greater job insecurity indicated less job satisfaction; those in jobs exposing them to more physical hazards indicated less job satisfaction; and those reporting more harassment in their workplaces indicated less job satisfaction. Third, all four blocks of predictors accounted for significant amounts or increments in explained variance on work-related psychosomatic symptoms. Single women and women working more hours per week reported more work-related psychosomatic symptoms. Women indicating more work stressors, women in jobs having greater physical demands, women reporting more harassment in their workplaces, women in jobs exposing them to more physical hazards, and women reporting greater familywork conflict indicated higher levels of psychosomatic symptoms. Fourth, two of the four blocks of predictors (work situation characteristics and work stressors) accounted for significant increments in explained variance on days of illness during the past year. Women having longer job tenure and women in jobs exposing them to greater physical demands reported more days of illness. A few more general observations are worth noting. First, the four blocks of predictors accounted for more variance on the measures of job satisfaction and psychosomatic symptoms

Work Stress and Women’s Health

97

TABLE II Work stressors, work-family conflict and women’s health Women’s health Psychosomatic last month (N = 1063) Personal demographics Marital status (–0.07) Education (0.07) Personal income (0.08) Situational characteristics Hours worked (–0.07) Work stress Work stressors (0.19) Physical demands (0.17) Work-family conflict FWC (0.11) Global job satisfaction (N = 1128) Personal demographics Situational characteristics Work stress Work stressors (–0.45) Job insecurity (–0.11) Hazards (–0.07) Harassment (0.07) Work-family conflict Work-related psychosomatic (N = 973) Personal demographics Marital status (–0.06) Situational characteristics Hours worked (–0.11) Work stress Work stressors (0.29) Physical demands (0.24) Harassment (0.14) Hazards (0.11) Work-family demands FWC (0.06) Days Ill last year (N = 1045) Personal demographics Situational characteristics Job tenure (0.11) Work stress Physical demands (–0.10) Work-family conflict

R

R2

∆R2

P

0.18

0.03

0.03

0.001

0.24

0.06

0.02

0.001

0.46

0.21

0.15

0.001

0.48

0.23

0.02

0.001

0.10 0.13 0.55

0.01 0.02 0.30

0.01 0.01 0.28

0.05 NS 0.001

0.55

0.30

0.00

NS

0.17

0.03

0.03

0.001

0.26

0.07

0.04

0.001

0.66

0.44

0.37

0.001

0.67

0.45

0.00

0.05

0.11 0.16

0.01 0.03

0.01 0.02

NS 0.01

0.22

0.05

0.02

0.001

0.23

0.05

0.00

NS

98

Ronald J. Burke

than on days of illness. Second, three of the four blocks of predictors (personal demographics, work situation characteristics, work stressors) had significant relationships with three or more of the health outcomes. Third, work stressors accounted for considerably more explained variance on the health outcomes than did the other blocks of predictors. The measures of work stressors and physical demands had significant and independent relationships with three of the four health outcomes.

Occupational status and personal demographics Table III presents the correlations between the measure of occupational status and a number of personal demographic characteristics and work situation factors. Almost two-thirds of these correlations were significantly different from zero (p < 0.001). It should be noted that the sample sizes were large (about 2000) and many of the personal and work situation characteristics were themselves significantly correlated. Some of the significant correlations are consistent with the operationalization of the occupational status measure. Thus, women in lower TABLE III Occupational status and personal demographics Personal demographics Visible minority Disability status Level of education Martial status Parent status Hours household Personal income Family income Age Number of jobs Organization size Job tenure Hours worked *** p < 0.001. a Ns range from 1982 to 2381.

Occupational statusa 0.03 0.10*** –0.43*** 0.09*** 0.00 0.08*** –0.52*** –0.40*** 0.17*** 0.11*** 0.02 0.02 –0.22***

status occupations indicated lower levels of formal education lower levels of personal and family income, held a greater number of current jobs and worked fewer hours per week. Other significant correlations further added to our understanding of the occupational status measure. Thus, women in low status occupations were more likely to indicate disability status, were younger, were less likely to be currently married, yet spent more hours per week on household duties and responsibilities. Occupational status had no relationship with visible minority status, parental status, tenure in present job or organizational size.

Occupational status and work experiences The correlations between the measure of occupational status and a number of work experiences are presented in the top half of Table IV. All but one of the eight correlations were significantly TABLE IV Occupational status, work experiences and health Occupational statusa Work experiences Job stress Harassment Hostile work environment Job insecurity Physical hazards Physical demands Air quality Temperature

0.20*** 0.11*** 0.11*** 0.16*** 0.19*** 0.34*** 0.00*** 0.13***

Work outcomes Job satisfaction Days of work missed

0.16*** 0.15***

Health outcomes Days of illness Psychosomatic symptoms – work Psychosomatic symptoms

0.14*** 0.13*** 0.09***

*** p < 0.001 a Ns range from 1874 to 2328.

Work Stress and Women’s Health different from zero (p < 0.001). It should be noted that the work experience measures themselves were themselves significantly correlated in the majority of cases. Women working in lower status jobs indicated high levels of job stress, more harassment in their workplaces, a more hostile, harassing work environment, greater job insecurity, greater exposure to physical hazards at work, greater physical demands (e.g. bending, lifting) in their jobs, and less comfortable workplace temperatures. Job status had no relationship with measures of air quality and ventilation.

Occupational status, work satisfactions and health The bottom half of Table IV shows the correlations between the job status measure and both work and health outcome, all five correlations were significantly (p < 0.001). Again, it should be noted that these outcome measures ere themselves significantly correlated. Women in lower status jobs indicated less job satisfaction and higher levels of absenteeism in the preceding year. In addition, they reported more psychosomatic symptoms, more workrelated psychosomatic symptoms and more days of illness in the preceding year.

Occupational status, work satisfaction and health Table V shows the results of regression analyses in which four work satisfaction and health indicators were regressed on three blocks of predictors (individual demographics, work situation characteristics, occupational status). The important question here was whether the measure of occupational status would account for significant increments in explained variance on these four outcome measures, controlling for personal demographic and work situation characteristics. The occupational status measure was found to account for significant increments in explained variance in each analysis. Women in lower occupational status positions indicated poorer general

99

health in the month preceeding the survey, poorer work related health, more days of illness in the preceding year and lower levels of work satisfaction. The measure of occupational status explained relatively small amounts of variance in these outcomes however. This finding was consistent with the relatively low correlations between the occupational status measure and work experiences and health (see Tables III and IV).

Conclusions and implications The empirical study reported here had several objectives. These included: extending the study of work stress and health to women, employing a diverse sample of women respondents beyond the more commonly studied managerial and professional levels, and exploring the effects of occupational status on the work and health relationship.

Work stressors and women’s health This study examined the relationship of several work stressors with measures of satisfaction and well-being in a large sample of employed women. The findings were supportive of a research framework developed to guide the study and were consistent with previous conclusions. That is, work stressors had negative relationships with levels of job satisfaction and positive relationships with self-reported psychosomatic symptoms and days of illness (see Table II). These findings extended our understanding of job demands and women’s health by including considerable numbers of women in low occupational status jobs and introducing work stressors that assessed aspects of the physical environment as well as physical demands present in these jobs. These extensions to the more traditional consideration of work stressors and health produced interesting results. First, work stressors resulting from the physical demands within jobs as well as physical hazards posed to job incumbents emerged as important predictors of women’s satisfaction and health. These work stressors are

100

Ronald J. Burke TABLE V Occupational status, work satisfaction and health

General health – last month (1533) Ind. demographics Marital (–0.07) Work situation Hours worked (–0.12) Number of jobs (–0.07) Job tenure (–0.06) Occupational status (–0.08) Work satisfaction (N = 1665) Ind. demographics Marital (0.08) Work situation Job tenure (0.07) Occupational status (0.15) Work related health (N = 1362) Ind. demographics Marital (–0.07) Work situation Hours worked (–0.19) Job tenure (–0.12) Number of Jobs (–0.05) Occupational status (–0.13) Days Ill (N = 1540) Ind. demographic Work situation Job tenure (0.09) Occupational status (0.13)

rarely included in job demands research. Second, these findings highlighted the need to devote more research attention to women (and men) in lower occupational status jobs. There seems to have been a trend away from these occupations toward more examination of managerial and professional jobs. A combination of factors seemed to converge on women in these lower occupational status positions. These included personal demographics (lower education, lower income), work stressors

R

R2

∆R2

P

0.12

0.02

0.02

0.001

0.18

0.03

0.02

0.001

0.19

0.04

0.00

0.010

0.11

0.01

0.01

0.010

0.14

0.02

0.01

0.010

0.19

0.03

0.01

0.001

0.13

0.02

0.02

0.010

0.24

0.06

0.04

0.001

0.26

0.07

0.01

0.001

0.09 0.14

0.01 0.02

0.01 0.01

0.050 0.010

0.17

0.03

0.01

0.001

(harassment, physical demands, physical hazards) and well-being (low job satisfaction, more psychosomatic symptoms and more days of illness).

Occupational status and women’s health The present study also examined the influences of occupational status on women’s work experiences, work satisfactions and health. Consistent, though moderate effects, of occupational status

Work Stress and Women’s Health were found (see Tables III and IV). First, there was support for the occupational status measure used in the research. That is, relationships of particular personal demographics likely to be associated with occupational status (i.e., education, income) were evident. In addition, as expected, women in lower occupational status jobs indicated less satisfaction at work and poorer emotional and physical health. One potential explanation for these findings lies in the work experiences reported by women in low status jobs. These women reported more negative work experiences, ranging from heightened job insecurity and work stress to a more hostile and hazardous workplace. Thus there seemed to be a constellation of factors operating in the work lives and family lives of women in lower status jobs that posed potential health risks to them. A large body of research evidence exists in the fields of medical sociology and epidemiology highlighting the role played by occupational status and social class in explaining levels of satisfaction and health (Pearlin, 1989; Borooah, 1999). Researchers and policy makers interested in gender, work stress and health need to expand their samples to include women in jobs having lower occupational status to fully understand the nature and magnitude of women’s health concerns and the focus at potential remedial actions. We need more research on work and health among women. More attention must also be paid to women who work in jobs other than managerial and professional ones. We need to include measures of work conditions, how the work is experienced and measures of a variety of wellbeing and health indicators in this research as well. This is important if we are to understand the broader relationships of social class and health. There is also a need to be aware of and link our research in work and well-being to an increasing and well conducted body of scholarship produced by others interested in health (medical sociologists, epidemologists).

101

Notes 1

Preparation of this manuscript was supported in part by the School of Business, York University. The research was funded by the Ontario Workplace Health and Safety Agency and the Ontario Disease Panel. I thank the Centre for Health Studies, York University and the Institute for Social Research, York University, for making the data available. Graeme Macdermid assisted with data analysis. Sandra Osti prepared the manuscript. 2 All measures were developed specifically for the study.

References Aston, J. and J. Lavery: 1993, ‘The Health of Women in Paid Employment: Effects of Quality of Work Role, Social Support and Cynicism and Psychological and Physical Well-being’, Women and Health 20, 1–25. Bartley, M., A. Sacker, D. Firth and R. Fitzpatrick: 1999a, ‘Understanding Social Variation in Cardiovascular Risk Factors in Women and Men: The Advantage of Theoretically Based Measures’, Social Science & Medicine 49, 831–846. Bartley, M., A. Sacker, D. Firth and R. Fitzpatrick: 1999b, ‘Social Position, Social Roles and Women’s Health in England: Changing Relationships 1984–1993’, Social Science & Medicine 48, 99–116. Borooah, V. K.: 1999, ‘Occupational Class and the Probability of Long-term Limiting Illness’, Social Science & Medicine 49, 253–266. Burke, R. J.: 1996, ‘Work Experiences, Stress and Health Among Managerial and Professional Women’, in M. J. Schabracq, J. A. M. Winnubst and C. L. Cooper (eds.), Handbook of Work and Health Psychology (Wiley, New York), pp. 205–230. Burke, R. J. and C. A. McKeen: 1994, ‘Work and Career Experiences and Emotional Well-being of Managerial and Professional Women’, Stress Medicine 10, 65–79. Collins, B. S., R. B. Hollander, D. M. Koffman, R. Reeve and S. Seidler: 1997, ‘Women, Work and Health: Issues and Implications for Worksite Health Promotion’, Women & Health 25, 3–18. Cooper, C. L.: 1996, Handbook of Stress, Medicine and Health (CRC Press, Boca Raton, FL). Cooper, C. L. and R. Payne: 1988, Causes, Coping and Consequences of Stress at Work (John Wiley, New York).

102

Ronald J. Burke

Davidson, M. J. and C. L. Cooper: 1983, Stress and the Woman Manager (St. Martin’s Press, New York). Davidson, M. J. and C. L. Cooper: 1992, Shattering the Glass Ceiling (Paul Chapman, London). Diez-Roux, A. V., B. G. Link and M. E. Northridge: 2000, ‘A Multi-level Analysis of Income Equality and Cardiovascular Disease Risk Factors’, Social Science & Medicine 50, 673–688. Hochschild, A. R.: 1997, The Time Bind (Metropolitan Books, New York). Humphries, K. H. and E. van Doorslaer: 2000, ‘Income-related Health Inequality in Canada’, Social Science & Medicine 50, 663–672. Kempen, G. I. J. M., E. I. Brilman, A. V. Ranchor and J. Ormel: 1999, ‘Morbidity and Quality of Life and the Moderating Effects of Level of Education in the Elderly’, Social Science & Medicine 49, 143–150. Langan-Fox, J.: 1998, ‘Women’s Careers and Occupational Stress’, in C. L. Cooper and I. T. Robertson (eds.), International Review of Industrial and Organizational Psychology (John Wiley, New York), pp. 273–304. Messing, K.: 1997, ‘Women’s Occupational Health: A Critical Review and Discussion of Current Issues’, Women & Health 25, 39–55. Muntaner, C., W. W. Eaton, C. Diala, R. C. Kessler and P. D. Sorlie: 1998, ‘Social Class, Assets, Organizational Control and the Prevalence of Common Groups of Psychiatric Disorders’, Social Science & Medicine 47, 2043–2054.

Nelson, D. L. and R. J. Burke: 2000a, ‘Women, Work Stress and Health’, in M. J. Davidson and R. J. Burke (eds.), Women in Management (Sage Publications, Newbury Park, CA), pp. 177–191. Nelson, D. L. and R. J. Burke: 2000b, ‘Women Executives: Individual and Organizational Health Issues’, Academy of Management Executive 14, 107–121. Offermann, L. R. and M. A. Armitage: 1993, ‘Stress and the Woman Manager: Sources, Health Outcomes, and Interventions’, in E. Fagenson (ed.), Women in Management: Trends, Issues, and Challenges in Managerial Diversity (Sage Publications, Newbury Park, CA). Pearlin, L. I.: 1989, ‘The Sociological Study of Stress’, Journal of Health and Social Behavior 30, 241–256. Schabracq, M. J., J. A. M. Winnubst and C. L. Cooper: 1996, Handbook of Work and Health Psychology (John Wiley, New York). Veenstra, G.: 2000, ‘Social Capital, SES and Health: An Individual-level Analysis’, Social Science & Medicine 50, 619–630.

York University, Faculty of Administrative Studies, 4700 Keele Street, M3J 1P3, North York, Ontario, Canada