Job stress has proven to be a difficult issue for the occupational health community and the labor movement to tackle. Unlike physical or chemical hazards, there is not an obvious tangible hazardous agent. This issue has also been preempted by corporate stress management, health promotion, or employee assistance programs, which explain stress as a purely personal reaction, and often treat the symptoms, not the causes, of job stress. (1) (2) There has been legitimate resistance to this "stress management" model which "blames the victim" and ignores the objective basis of job stress. The occupational stress field of job stress. The occupational stress field also has been plagued by a variety of definitions and difficulties in measurement of stress. In addition, changes in job design or work organization are often inherently more "systems challenging" and require more radical restructuring of workplaces than reducing levels of exposure to toxic substances or ergonomic hazards. This article was prompted, in part, by the need to address such concerns. (3)

A number of specific stressful working conditions, such as repetitive work, assembly-line work, electronic monitoring or surveillance, involuntary overtime, piece-rate work, inflexible hours, arbitrary supervision, and deskilled work, have been studied and recently reviewed. (4) Over the last 14 years, a new model of job stress (figure 1) developed by Robert Karasek (5), has highlighted two key elements of these stressors, and has been supported by a growing body of evidence. Karasek's "job strain" model states that the greatest risk to physical and mental health from stress occurs to workers facing high psychological workload demands or pressures combined with low control or decision latitude in meeting those demands. (6) Job demands are defined by questions such as "working very fast," "working very hard," and not "enough time to get the job done." Job decision latitude is defined as both the ability to use skills on the job and the decision-making authority available to the worker. In some recent studies, this model was expanded to include a third factor - the beneficial effects of workplace social support. (7)

While there are a variety of models of "job stress, the "job strain" model (which is the focus of this paper) emphasizes the inter-action between demands and control in causing stress, and objective constraints on action in the work environment, rather than individual perceptions or "person-environment fit." In addition, other important work related and "social" stressors exist that are less directly connected to the concept of "job strain," but may also have significant health consequences. These include increasing work hours (8,9), and sexual (10,11) or racial harassment or discrimination.

Most studies of the "job strain" model have focused on outcomes such as cardiovascular disease (high blood pressure, heart disease) and psychological distress (anxiety and depression). Since heart disease is the most common cause of death in industrialized countries, and because the "job strain" and heart disease studies have been recently reviewed (12), this article will summarize the evidence that "job strain" is linked to heart disease. We then review the evidence demonstrating that interventions targeted at altering objective working conditions can produce beneficial effects on job demands, decision latitude, and social support. (13) We also review collective bargaining, legislative and other efforts to reduce "job strain" specifically and job stress in general. Together, these studies and programs provide convincing evidence that job design and work organization are risk factors for cardiovascular illness and that these can be modified.

SIGNIFICANCE OF JOB STRESS

The issue of job stress is of more importance to the occupational health community and the labor movement for two compelling reasons:

First, there is the potential for preventing much illness and death. More than 50 million Americans have high blood pressure, and, in 95 percent of cases, the cause is unknown. (14) (This type is called "essential" hypertension, as opposed to cases where there is a recognized cause, such as an adrenal gland disorder or kidney disease. ) High blood pressure is a major cause of heart disease and stroke. (14) While estimates of the proportion of heart disease possibly due to "job strain" vary greatly between studies, Karasek and Theorell (5, p. 167) calculate that up to 23 percent of heart disease could potentially be prevented (over 150,000 deaths prevented per year in the U.S.) if we reduced the level of "job strain" in jobs with the worst strain levels to the average of other occupations. (15) The economic costs of job stress in general (absenteeism, lost productivity) are difficult to estimate but could be as high as several $100 billion (5, p. 167-8).

Second, Karasek's model emphasizes another major negative consequence of work organization-how the assembly­line and the principles of Taylorism, with its focus on reducing workers' skills and influence, can produce passivity, learned helplessness, and lack of participation (at work, in the community, and in politics). The "job strain" model (figure 1) has two components - increasing risk of heart disease following arrow A, but increasing activity, participation, self esteem, motivation to learn, and sense of accomplishment following arrow B. Thus, this model provides a justification and a public health foundation for efforts to achieve workplace democracy. Democracy at work should be promoted as not only just and fair, but also as a method to reduce ill health, and to allow for fuller development of people's emotional, intellectual, and social capabilities.

EVIDENCE LINKING "JOB STRAIN" TO HEART DISEASE

Over the last decade, 14 studies on "job strain" and heart disease including two of all-cause mortality and 20 studies on "job strain" and heart disease risk factors have been published. These have been conducted in Sweden, the U.S., Finland, Denmark, Australia, and Japan, provide strong evidence that "job strain" is a risk factor for heart disease. (12)

Of the 14 heart disease studies, 12 showed clear associations between "job strain" and heart disease. Most of these studies controlled for other (potentially confounding) heart disease risk factors. More importantly, of the eight cohort studies of heart disease and all-cause mortality, seven showed strong positive associations (7, 16­21). (The cohort studies followed people over time and therefore are less likely to be biased.)

In the 20 studies of "job strain" and independent risk factors (for heart disease), the following patterns were seen. Three studies found no association between "job strain" and serum cholesterol. Two studies found a link between "job strain" and smoking (22, 23), while two did not. (One of the negative studies (24), however, did show higher rates of smoking at lower levels of job decision latitude.) Of nine studies of blood pressure measured in a clinical setting, only one found a significant association. However, of the eight studies where an ambulatory (portable) blood pressure monitor was worn during a work day, five showed strong positive associations between "job strain" and blood pressure (2~30), while the other three provided mixed results. (25, 31, 32) Since ambulatory blood pressure is both more reliable (since there is no observer bias and the number of readings is greatly increased) and more valid (since blood pressure is measured during a person's normal daily activities) than casual measures of blood pressure, we feel confident in placing more emphasis on the ambulatory blood pressure results. We conclude that one pathway between "job strain" and heart disease is elevated blood pressure (12), possibly mediated by increases in catecholamines and cortisol (33), increased autonomic nervous system activity (for example, increased heart rate), and/or increased mass of the hearts left ventricle. (34)

CURRENT ISSUES IN RESEARCH ABOUT "JOB STRAIN" AND HEART DISEASE

A variety of issues were raised in these 34 studies, including methodological concerns, and factors which may modify the impact of "job strain" on heart disease (for example, gender, race, and socioeconomic status (SES)). For example, the

studies used various methods and survey questions to measure the concept of "job strain," one limitation of the research. However, Karasek's Job Content Questionnaire (35), which maintains 14 basic questions on task­level demands, skill use, and authority, was frequently used in these studies. The concept of task­level job demands was primarily measured by questions about workload demands. However, it is important to expand the concept to include other stressors such as responsibility for people, role conflict, role ambiguity, and threat of violence or injury. In addition, while questions on task­level control and demands were available in U.S. and Swedish national surveys, questions on control, either individually or collectively (36, 37), over departmental or organization - level policies and decisions (38) were not. Strengths of the "job strain" model are its simplicity and clarity, its prediction of both health and behavioral outcomes, and its emphasis on classifying features of the work environment into the categories of demands or control. However, the model would benefit from the inclusion of various dimensions of demands and control (such as those listed above) used in the more complex Michigan job stress model (39) and the NIOSH Generic Job Stress questionnaire. (40)

In addition, in 10 studies, a technique was used to develop more "objective" measures of job characteristics. National averages of job characteristics for a particular job title were assigned to individuals having that job title, ignoring the fact that job characteristics vary for people even within the same job title. Six of these 10 studies provided positive results. Thus, despite errors in measurement which make it more likely that links between "job strain" and illness will not be found even when they do exist, it is remarkable that consistent positive patterns have been found.

In seven of the 10 studies where comparisons could be made, the effects of "job strain" were similar for men and women. (However, a higher proportion of U.S. women workers face "job strain." (5, p. 45-6) As the Framingham Heart Study (41) and other research (42) make clear, many women face a dual set of demands from work outside and within the home. "High strain" work (that is, " job strain" may interact with home demands to increase heart disease risk for certain subgroups of workers. In addition, "job strain," as currently measured, may not adequately "capture" other stresses faced by women workers such as salary and promotion inequities or sexual harassment. (10, 11)

Another factor is race or ethnic group. Two of three studies, which reported this information, showed a higher proportion of African­American workers in the "high strain" group (27, 31), suggesting that the increased risk of hypertension faced by African-Americans in the U.S. may be, in part, a result of more "job strain." Since only three studies have been conducted with a predominantly non­Caucasian population, further research is needed to determine to what extent the concept is appropriate and still a risk factor among other racial/ethnic groups.

The relation between "job strain" and SES has also been debated. Many studies statistically control for education (as a measure of social class), and still find that "job strain" increases the risk of heart disease. Conventional social status scales are poorly correlated with "job strain" (5, p. 77), indicating that the "job strain" findings are not simply due to the association between lower SES and heart disease. (43) There are "high strain" white collar clerical jobs in the low­to­middle part of the job status hierarchy, and some blue-collar craft jobs that allow for a high level of skill development and autonomy. Low status and low income occur in "low strain," "passive," and ''high strain" jobs (figure 1).

Only four studies compared social class groups- finding that blue­collar workers, workers with less education, or female clerical workers had a substantially stronger association between "job strain" and heart disease than higher SES groups. Lower SES groups also have higher rates of heart disease and heart disease risk factors (43), and "job strain" may interact with these risk factors. (44) Limited economic resources may play a role. "Job strain" may also interact with chemical and physical health hazards on these jobs (carbon monoxide, solvents, lead, noise, shiftwork), or other psychosocial hazard s such as fear of job loss. (44) Such occupational risk factors may cluster together, and "job strain" may be increased by automation, or by increased work quotas due to budget cuts.

While scientific proof for the "job strain" model is not yet conclusive, preventive action can be undertaken to reduce potential health risks. In the U.S., efforts to reduce occupational stress continue to focus primarily on changing the individual behavior of employees (for example, relaxation techniques, exercise, diet, cognitive/behavioral skills). (45) However, a growing number of programs and interventions are attempting to change various workplace sources of stress. While none of these interventions was specifically designed to reduce "job strain," many focused on changing components of "job strain" (for example, inducing demands, increasing control, enhancing support). In addition, these programs have rarely included objective measures of heart disease risk. However, their lessons provide a valuable guide to future illness prevention and job redesign efforts, and to broader efforts to increase workplace democracy. Efforts to reduce or prevent job strains have been work site based, community­based, industry­wide (in some cases of collective bargaining) or statewide or national in scope, in the case of legislation or regulations. U.S. work site programs have mainly been the result of both social science based organizational reform efforts (known as Action Research) and collective bargaining. These programs are reviewed in the following sections.

SOCIAL SCIENCE­BASED INTERVENTIONS

The most well­developed applied research tradition on bringing about planned change in organizations is the field of Organization Development (OD). OD has its roots in the "human relations" management and social theorists of the 1940s­'50s, who were reacting to the dehumanization, alienation, and bureaucracy characteristic of scientific management (Taylorism). (46, 47) OD practitioners conducted innovative work reform experiments during the 1950~'70s, including early joint labor­management Quality of Work Life (QWL) programs. These focused primarily on social relationships (for example, a sense of belonging, supportive supervision, participation in decision­making) rather than the technical features of production and work organization. In the 1980s, OD practitioners "discovered" the importance of technology, especially European Socio-Technical Systems (STS) theory, which promotes semiautonomous work teams. More importantly, by the 1980s, many OD professionals lost sight of their original stated mission to attempt to serve both employer interests and employee needs and applied their trade primarily on behalf of employers. (48, 49).

Scandinavian work reform experiments in the 1960s-'70s, while influenced by the same human relations research (and also reacting against the dehumanizing effects of scientific management), placed a greater emphasis on technical aspects of production (for example, piece­rate, shiftwork, technology) as well as an understanding that physical illness and injury is an outcome of work organization (50) -- an outcome which has been largely ignored by OD. These different emphases, along with a progressive political climate and a highly unionized work force, led eventually to work environment legislation in the 1970s in Scandinavia and continuing job redesign and work reform efforts today. (5, 51) These experiments, and the emphasis on health as an outcome of work, also laid the foundations for Karasek's model, and much stress research both in Scandinavia and the U.S.

Many OD and QWL efforts have failed, however, because of factors such as lack of support by top management or supervisors, failure to delegate authority, a bureaucratic, authoritarian climate, and rigid job descriptions and personnel practices. (52, 53) Some interventions have led to increased workload or "speedup" (54, 55), work force reductions (46), or were initiated as attempts to avoid unionization (56, 57) or weaken the existing union. (49, 58) However, positive experiences with cooperative programs have also been reported by some unions (59, 60), and the debate continues in the labor movement over the potential value of these programs in specific situations.

Recognizing these limitations, unions and occupational health professionals have much to gain by adopting the valuable set of techniques and processes (intervention research methods) developed by OD, and using them on behalf of workers. One of these methods is known as Action Research (AR). AR involves a partnership between outside experts (usually social scientists) and members of organizations in defining problems, developing intervention tactics, introducing changes that benefit organization members, and measuring outcomes. (38) Issues and changes that this approach typically involves include decision­making structures and processes, task and role demands, information and communication practices, work schedules, and training policies. AR can be classified into "expert-dominated" approaches (also allied "weak" AR), in contrast to "strong" versions where there is relative equality among researchers and organization members in all aspects of the intervention and research process also termed Participatory Action Research (PAR). (61) While few studies have compared these approaches, one review suggests that PAR generates more positive outcomes. (62) Several key examples of "expert" AR and PAR interventions, which focused on improving workers' physical or mental health, are briefly summarized below, followed by a discussion of policy and research issues.

Expert­Dominated Action Research.

In a classic example, Jackson took advantage of a state legislative mandate for more frequent staff meetings in hospitals to measure the effects of participation in decision­making on job stress, job satisfaction, absenteeism and turnover. (63) Units where the intervention was implemented held twice as many staff meetings as in non­intervention units. Workers in participating units reported greater influence, less role conflict and ambiguity, less emotional strain, and greater job satisfaction at three month and six­month follow­up.

In another example, Golembiewski and colleagues worked with 31 "burned out" and overworked Human Resources (HR) staff of a corporation in the midst of rapid growth. (64) Four action planning groups developed recommendations, and the entire staff prioritized them and prepared implementation plans, which were presented to a corporate oversight committee. As a result, an HR career ladder was introduced as well as a change in reporting structure. Effects included a 50 percent reduction in reported ',burnout' that remained low four months after the last intervention, a turnover decline from 37 percent to 17 percent, and a significant increase in reports of "innovativeness."

Participatory Action Research (PAR).

An example of PAR was a six­year study by Israel, Schurman and colleagues in a components parts plant of a major unionized automobile company. (38, 65) With agreement from local union leadership and plant management, and working with union and management representatives, they set up a representative employee committee, primarily comprised of shop-floor employees -- the Stress and Wellness Committee (SWC) -- to implement the project. Using the PAR process of iterative cycles of diagnosis, action­taking and evaluation, the committee identified four primary sources of stress and designed interventions (through subcommittees) for each: lack of participation and influence, hassles with supervisors, lack of information / communication, and "production vs. quality." Interventions included establishment of a pilot cross­functional team in one department to address quality issues, convincing factory management to conduct state of­the­business meetings in each department, and creation of a weekly plant newsletter. Overall, SWC members report high levels of trust in and influence over the committee process. In addition, other employees who were more involved in and knowledgeable about the PAR project reported greater increases in participation, perceived participative climate and co­worker support than others with less exposure. (66)

Another example of PAR in a unionized setting began with a survey by Cahill of "burn­out'' and symptoms of stress among employees of the New Jersey child protection agency. (67) The survey, which found significantly higher levels of "burn­out" than in national samples of social workers, was presented by the employees' union in a legislative hearing. One result of the hearing was the formation of a labor­management stress committee, which identified the agencies existing mainframe computer system as a major source of stress. The system included repetitive de­skilled work for clericals, lack of control of data for administrators and social workers, hard­to­interpret monthly reports, and ergonomically poor work stations. The stress committee recruited a computer programmer to design software jointly with the local employees who would use a new PC­based system. Once the new system was in operation, workers reported significantly higher levels of job satisfaction, decision latitude, skill discretion, control over equipment, a more streamlined information flow between local and central offices, and improved ergonomic conditions.

A final example of PAR to reduce job stress was developed by Lerner and colleagues at the Institute for Labor and Mental Health, and was based outside the workplace. (68) Strategies for raising awareness of the social and workplace sources of stress included: meeting with unions; organizing a conference on job stress where workers told their story to government, public health officials and the media; a "family day" with workshops on stress of family and work life; and Occupational Stress Groups (OSGs). OSGs of 10 workers, led by shop stewards, met for eight to 12 weeks to discuss stress at work, develop social support, discuss the dangers of self-blame for feelings of powerlessness or stress, and to develop strategies for collective action. At follow­up, OSG participants showed significant improvements on virtually all measures of psychological well­being in comparison to controls. Behavioral changes and initiatives taken to improve the workplace were also reported in group interviews.

Other union­sponsored and work site­based initiatives, the OCAW Work and Family Program (69) and the District 65 UAW Stress Project (70), build on the OSG format. Both employ group meetings to raise awareness of stressful working conditions (and their impact on family life) and then develop collective bargaining proposals to improve working conditions.

Discussion.

PAR approaches with strong union involvement have significant advantages over weaker expert-dominated or management­dominated AR programs. Strong union involvement can ensure that the potential dangers of OD are minimized and that interventions genuinely improve the work environment. Unions played important roles in initiating and sustaining structural change in the auto parts factory and in the New Jersey state agency, as well as, of course, in developing the OSG, OCAW and District 65 programs. However, such programs are limited by the low unionization rate in the U.S. The community­based approach used by Lerner can be especially useful in non-union settings (such as COSH group efforts to educate and help organize non­union workers), or where unionized employers refuse to cooperate or commit required support and resources.

PAR is a flexible set of intervention processes and methods, not a pre­packaged canned program. This allows it to be effective in different contexts, with different occupational groups, and with resulting different strategies and tactics. It is also an innovative social research method, which makes it valuable for occupational health research. PAR is an effective tool for the evaluation of change because both quantitative and qualitative data are included, and process, impact, and outcome are assessed (thus requiring multi-disciplinary teams skilled in these techniques). For example, the intervention in the auto parts factory included three administrations of a plant-wide survey (including standardized survey scales), focus group interviews and five surveys of committee members, in-depth interviews of all committee members and plant union and management leaders, and verbatim field notes from committee meetings. Other studies included standardized surveys and objective records such as frequency of staff meetings, absenteeism and turnover. Such multi­method approaches permit "triangulation," that is, cross-validation of and increased confidence in the results. (38, p. 148) Process data enable participants and researchers to assess not just what happened but why it happened (including obstacles to change). Impact data can reveal which organizational or individual factors are affected by the intervention, and through which pathways. For example, in the auto parts factory, regression analysis of survey results indicated that the positive effects of participation were channeled through perceptions of influence. Outcome data can answer questions about health effects.

Another important research issue is the need for longitudinal designs, with adequate time for follow­up. For example, the amount of change reported by the intervention group in Jackson's study increased significantly between the first and second post­tests, suggesting that participation takes time to create effects. In the auto parts factory, 1.5 years was needed to conduct organizational diagnosis and needs assessment prior to engaging in major change strategies.

Thus, PAR to reduce job stress appears to work in two main ways (corresponding to arrows A and B of Karasek's model in figure 1), by: 1. modifying objective stressful conditions in the social and/or technical environment; and 2. the active (individual and collective) learning workers experience in successfully affecting positive change (for example, enhanced perceptions of control and influence, development of skills, positive self­appraisal, strengthened relationships with co-workers).

Genuine PAR allows workers not only to problem­solve but also to, jointly, with researchers, define targets for research and intervention and evaluate change (to be involved in all aspects of the intervention). Workers bang a richness of experience that enhances problem definition and hypothesis development, as well as insights to creating change. (71, 72) For example, workers can specify the concrete manifestations of job demands or low job control in a particular workplace (not captured by standardized scales), necessary for targeting change efforts. Researchers bring a rich knowledge base, methods of questionnaire construction and research design, and other means of improving study validity. While some researchers argue that participant involvement in social research could bias results due to improper wording of questionnaires, or attempts to influence survey response, bias can also result from employees' unwillingness to participate or candidly present their opinions "when involved with conventional research projects, because they associate researchers with management and the existing hierarchical structure." In addition, PAR researchers' use of multiple methods provides limit insights from the participants' "inside" understanding of attitudes, needs, and the social environment. (38, p. 140)

Genuine PAR (as opposed to some QWL programs) increases the skills and activism of those participating in the intervention, although to date there is no evidence that it strengthens union solidarity. However, just as active and assertive union involvement in health and safety training programs strengthens the union's position and credibility in the eyes of its members (73), benefits should be expected when the union is actively involved in improving other issues of concern to workers-job design and psychosocial work environment. (74, 75)

Personal stress management and health promotion was a component in many of these programs (including the District 65, UAW stress program). By discussing personal behavior change within the context of an overall program to improve the work environment, self-blame for behaviors or feelings of stress is avoided, and the union shows it is concerned about the personal welfare of its members. It can also be an organizing tactic to help gain publicity and support for the overall program, as in the auto parts factory study. In general, multiple levels (individual, group, organization, society) need to be targeted for interventions to effectively reduce stress. (76)

Even in successful interventions, many obstacles to change remain, for example, management turnover, lack of management support, pending layoffs and general market conditions in the auto parts factory. In the New Jersey state agency, information and technology managers were initially resistant, perceiving the new technology and software as a threat to their power. Ensuring that they received some credit for the success of the project eventually led to their strong support for the intervention.

PAR can be a valuable technique in traditional occupational health programs. (71, 77) In addition, occupational health professionals and unionists can play a critical role in the next stage of stress research and stress prevention, by 1. adding physical health as an outcome in PAR programs to improve the psychosocial work environment; 2. studying the effect of the physical work environment and fear of injury, on perceived stress and psychological well­being; and 3. studying the possible interaction between physical and psychosocial hazards in the production of heart disease, hypertension, and psychological distress, and other outcomes potentially related to job stress, such as musculoskeletal disorders (78), adverse pregnancy outcomes (79), and "sick building syndrome." (80)

COLLECTIVE BARGAINING APPROACHES

In addition to more recent PAR programs, collective bargaining has been a traditional strategy to increase employee decision latitude (authority, influence, skill), and to regulate demands through contract language on issues such as job security, overtime, seniority, discrimination, technological change, skills training, career ladders, staffing, grievance procedures, and labor­management committees. (81­83) For example, the nurses' shortage during the 198Os in the U.S. has been attributed to factors such as low salary and job stress. Nurses have expressed a strong desire to be treated as professionals, which can be denied through understaffing, lack of autonomy, or an authoritarian work climate. In response, unions have bargained for clinical career ladders for RNs in various specialties, joint physician­nurse committees, greater in­service education (84), and quality patient care and personnel committee. (82)

Many clerical workers have joined unions in the last decade, in part due to issues related to job stress: career mobility, pay equity, job security, child care, flextime, parental leave, sexual harassment, having a "voice" through union­management committees, and video display terminal (VDT) work. (85) VDT workers have bargained for better ergonomic conditions, but have also learned that adjustable equipment is not enough. For example, at a New York City newspaper, a union-management committee discovered that job design issues such as control over schedule, regular breaks, work variety, and training were as important as the purchase of new equipment. (86) The National Institute for Occupational Safety and Health (NIOSH) is conducting various studies of the role of psychosocial factors in the development of cumulative trauma disorders (CTDs) among VDT operators. (87)

At least six million U.S. workers were electronically monitored in 1987, with the number expected to grow. (88) As part of a 1992 settlement of a Communications Workers of America (CWA) lawsuit, Northern Telecom agreed to prohibit secret voice, computer, and video monitoring of employees. (89) A CWA­U.S. West contract banned monitoring in 1989 with the help of early results from a study that showed that monitored workers had higher rates not just of psychological distress but also "stiff or sore wrists," "loss of feeling in fingers or wrists" other symptoms of CTDs. (90) Similar studies by Bell Canada and the Communications Workers of Canada led to restrictions on monitoring in 1990. (89) Recently, AT&T agreed to ban secret monitoring of the job performance of workers. (91) A new study at U.S. West by NIOSH showed and stress due to monitoring, fear of job loss, increasing work pressure, and little job decision-making opportunity contributes to injures even when proper equipment is used. (92)

The apparent interaction between psychosocial stress and physical stress and injury and illness needs to be better understood. Monitored workers have reported aspects of "job strain" (greater workload, less job control, unfair work standards, less skill use and variety), and poorer supervisor support. Do such factors lead to fewer breaks, longer work hours, or faster typing? Does increased muscle tension play a role? While some of the 10­fold increase in reported CTDs over the last decade (93) is undoubtedly due to better reporting, these studies suggest that some may be due to work speed­up, de-skilling of jobs into simpler, more repetitive tasks, lack of control, and fear of job loss.

Electronic monitoring is often used to punish, not reward (for example, by publicly displaying results), managers over­rely on it, and an emphasis on quantity not quality is created. (94) However, unions have shown that there are productive alternatives to monitoring. For example, CWA members at an Arizona facility, together with AT&T management, "eliminated individual measurement and remote secret observation. AWT (average work time) was measured only for the whole group. Service observation was performed by small groups of peers by the old­fashioned 'jack­in' method, where the observer sits beside the person being monitored, listens to a few calls and then discusses the results with the employee." As a result, AWT was better than under previous methods of supervision, them were fewer customer complaints, and both the grievance rate and absenteeism were lower. (94)

The loss of the 1981 PATCO strike and the firing of 11,000 unionized workers was a major setback in workers' rights to organize and strike. Some argue that PATCO's biggest failure was that it could not make an effective case for job stress a major strike issue. (95) The job of air traffic controller includes many aspects of "job strain:" 1. high demands (through understaffing, mandatory overtime, few vacations); 2. poor skill utilization (because of poor training methods, outmoded

equipment, few opportunities for promotion); 3. little authority (due to an autocratic system and military style management, where grievants are labeled as troublemakers and not promoted). (95, 96) These conditions persist and, not surprisingly, new controllers have joined a new union and stress remains a major issue.

However, medical proof of the job's hazards has remained elusive. While the major 1975-78 health study of controllers did report prevalence of hypertension 1.5 times that of national samples, and incidence of new cases of hypertension up to four times higher (97), much analysis focused on individual and psychological differences among men in the study. In addition, the Federal Aviation Administration (FAA) emphasized only the individual differences (not the high dissatisfaction with "management policies and practices" noted in the study, (97, p. 6281), and never published the non­technical summary of the study. (2, p. 1301­3) For years, the FAA had ordered researchers conducting their stress studies "not to make recommendations" for corrective action. (2, p. 895) The FAA's technical representative to the study later testified that if the findings of the study (and 28 other FAA studies) had been applied, '1 am absolutely certain" that the 1981 strike "would have been averted." (2, p. 874) Air traffic controllers' experience of stress and desire for equity had been deflected into a debate about the quality of scientific evidence on stress and health. (98)

In 1981, PATCO's collective bargaining demands focused on ways of "escaping" rather than "confronting" job stress: reduced work hours, early retirement, and higher salary demands which did not win public sympathy. Alternative strategies such as improving organizational climate, supervision and communication (99) or more power over the work process, for example, flow control, curbing unregulated pleasure aircraft, disciplining of authoritarian supervisors, or more new hires, were not attempted. (95, p. 187) There were, of course, other reasons why the strike was lost, such as failure to effectively build alliances with other unions (95), poor public relations (100), and, most importantly, an intransigent administration in Washington, DC. However, former PATCO officer Bill Taylor emphasized that "knowing what I know now, I think we should have tried to double our effort to inform the public what the strike was all about, which was bargaining rights, not money." (101 )

A more constructive resolution to a labor-management conflict over working conditions and health was arrived at by a union of toll collectors and a New York City agency. While a specific toxin had not been identified as the cause of illness among 34 bridge toll workers in New York City in 1990, union officials had 'bridled" at the suggestion that the outbreak was due to "stress." (102, 103). The union had attempted for years to improve safety and health conditions for the toll collectors, who have elevated heart disease mortality rates, due, at least in part, to documented excess exposure to carbon monoxide (CO) from automobile exhaust. After the outbreak, union officials demanded permanent air-monitoring equipment and better ventilation. Some union officials acknowledged that while the first cases in the outbreak may have been due to inhalation of toxic vapors (arising from the burning of plastic­coated wire), later cases may have been due to "anxiety." (102) The union and the agency recently bargained a substantial medical surveillance program, whose primary focus is on heart disease risk due to CO exposure. The program will also evaluate the possible role of "job strain" as an independent or interactive risk factor for heart disease.

STATEWIDE AND NATIONAL EFFORTS AND STRATEGIES TO REDUCE STRESS

Workers compensation. Spokespersons for the insurance industry argue that claims for "mental injury" rose sharply during the 1980s, and now account for about 15 percent of all occupational disease claims nationwide (104) -- figures used to justify current efforts to limit claims. However, accurate data is difficult to obtain. In California, for example, one of only six states which considers mental injuries caused by gradual mental or emotional stress to be compensable, and a state with the most liberal law, the rate of mental stress (claims increased 540 percent between 1979-88, according to state data. (105) However, the 9,368 reported cases in 1988 represented only two percent of total disabling work injuries. According to an insurance industry institute in California, many claims are not reported to the state agency, and self­insured public employers have higher rates, suggesting that the number of stress claims is actually four fumes higher. (105) However, even the higher estimate does not support arguments that business "is under siege" (104), but is compatible with growing awareness of the job stress­illness link

The California insurance institute study indicated that stress claimants are more likely to be female and older than other work-disabled employees. Sales and clerical workers filed 40 percent of stress claims. Fewer than 10 percent of the claims followed a specific incident (for example, armed robbery), rather job pressures (69 percent) and harassment (35 percent) were the most common cited reasons for the claim. (105) While it is difficult to generalize from this data, since many factors influence workers' ability or intention to file for compensation, it is compatible with the mode! of "job strain" as cumulative exposure to job pressures and low job control. The law still generally works against the worker since the burden of proof is upon the worker to define a condition and establish work­relatedness. (106)

Recently, employers have pushed for tighter standards for stress claims. A 1990 amendment to the New York State law restricts "mental­mental" claims when stress results from a normal personnel decision (work evaluation, job transfer, demotion) when taken in "good faith" by the employer. Similarly, since 1989, in California, the law requires that workers receive a psychiatric diagnosis of mental injury, and that "actual events" in the workplace were responsible for at least 10 percent of the causation of the injury not simply the worker's perception of stress. (105) It remains to be seen to what extent the new scientific evidence on "job strain" will be used in compensation cases to explain causation for mental injury, hypertension, or heart disease.

Legislation and political action. In the U.S., job stressors are not covered by OSHA. There are no health standards for shift work, piecework, machine-pacing, de-skilling, job security, isolated work, or technological change (as in Scandinavia). (107) An innovative campaign, however, is being waged by the Service Employees international Union (SEIU) in Pennsylvania to reduce back injuries and stress caused by inadequate staffing in nursing homes. (OSHA has already cited several nursing homes under the General Duty Clause for insufficient staff to do person transfers.) The campaign is in support of a proposed state law that would compel nursing homes to reveal information about staffing, injuries and profits, and set minimum staffing levels. (108) A recent SEIU national survey of nurses re-emphasized concerns about work load demands, understaffing and stress, and called for OSHA standards for nursing (including staffing), and providing health care workers with a voice in decisions. (109)

On the national level, support by the Clinton administration for the concepts of ''high­skill, high­wage strategies" and "worker participation" (110) to improve the competitiveness of U.S. businesses holds the promise for a new focus on developing healthier work environments and reducing "job strain." However, in order to genuinely promote ''high skill," active and lower "strain" jobs, job training and job design programs need to: 1. go beyond basic job skills, or narrow technical skills, and include "job ladders" or "career paths;" 2. promote computer software that encourages discretion and flexibility ("system knowledge"); 3. make skill training accessible to workers' schedules; and 4. keep skilled jobs in the bargaining unit and therefore increase rather than decrease union strength. (111, 112)

In addition, a variety of current legislative proposals could help increase job control and support, for example, laws that limit electronic monitoring and regulate VDT work. Other proposals could reduce the more general burden of social stress on individuals, such as laws on parental and personal leave, day care and elder care, voluntary overtime and shift work, a limited work week to create jobs, job sharing and part time work (8, 9) Even the OSHA reform bill (through mandated joint committees, improved worker training and enforcement, protection against discrimination, and improved recordkeeping) could spur efforts to identify and reduce psychosocial risk factors, most likely through investigation of hypertension and musculoskeletal disorders. Psychosocial risk factors could be considered for inclusion in the forthcoming ergonomics standard.

The goal of all these interventions and strategies is to produce a healthy workplace -- in which workers are respected, where they have the opportunity to develop their skills and abilities, and where authority is shared, in other words, workplace democracy. Therefore, it is also important to consider legislation that would strengthen workers' collective voice (that is, unions) through banning of permanent replacements for strikers, and, in general, reforming labor law, as well as other means of increasing workers' influence and economic security, such as full employment and opportunities for employee ownership.

CONCLUSIONS AND RECOMMENDATIONS

The "job strain" studies and other research support the idea that social factors play a critical role in the production of common chronic diseases, such as heart disease and hypertension. The intervention studies, and other prevention strategies, indicate that the work environment can be modified to increase employee influence, skills, authority, and support, and to regulate demands. Participatory action research, collective bargaining, and legislation can be effective tools to achieve these goals. Effective PAR requires strong union involvement, while collectively bargained programs can benefit from PAR

methods to involve workers and evaluate change. While the growing evidence linking job stress with illness helps to overcome the notion that psychosocial explanations for disease are not legitimate, vigilance needs to continue against our society's dominant ideology which uses the stress explanation to "blame the victim" indicting those who become ill as well.

We believe that the evidence presented supports the following actions. First, "job strain" assessment instruments should be included in workplace health surveillance and health promotion programs, and in occupational health clinic educational material. Second, unions and their allies need to further increase their emphasis on contract language, education, organizing, and legislation on issues related to their members' job design, work organization, quality of work life, schedule flexibility, and work and family concerns. Third, multidisciplinary teams (including workers, union and company officials, occupational health specialists, epidemiologists, labor and health educators, social psychologists, physicians and nurses), using PAR methods, can design, implement, and evaluate interventions to reduce or prevent exposure to psychosocial and physical health hazards and risk of illness. Fourth, further research is needed on various health outcomes (other than cardiovascular disease) potentially related to "job strain" or stress in general, including psychological disorders (4), musculoskeletal disorders (78), adverse pregnancy outcomes (79), "sick building syndrome" (80), work injuries (113), and immune system functioning (114), and the possible synergistic effects of psychosocial and physical health hazards. Modern workplaces embrace a complex set of risk factors, including psychosocial and physical/chemical.

Research is also needed on the connection between "job strain" and heart disease risk factors such as smoking, alcohol and diet, physiological mechanisms underlying heart disease, the effects of gender, race and social class, and time trends. Similarly, further research is needed on the mechanisms and pathways underlying the effects of participation (for example, perceived influence, skill development, social support) on improvements in satisfaction and self esteem, as well as aspects of intervention strategy associated with genuine organizational change. The Karasek "job strain" model has contributed greatly to the field through its clarity, predictions of health and behavioral outcomes, and emphasis on the concepts of demands, control, and support. It can now benefit from the expansion of the concepts of demands and control, to include measures contained in the Michigan stress model. (39, 40)

We believe that the "high demands + low control + low support" paradigm also provides a useful working model for understanding associations between more general social stress and health. Since hypertension is prevalent in all industrialized societies (both market and state­owned economies), and since blood pressure does not typically rise with age in non­industrial societies (for example, hunter-gatherers and agricultural communities) (115), we need to consider what aspects of industrial society (such as social class differences or "job strain") account for this effect. For example, home and family demands and lack of control may impact on health. (42) Unemployment, with its resulting health effects (116), can be perceived of as an extreme case of loss of control. Even the threat of unemployment can increase competition (demands) and lead to a decreased sense of control among remaining employees. (5, p. 307) The decline in the standard of living since the 1960s and the economic necessity for both parents to work is a major reason for increased work hours (increased demands) in the U.S. (9) Finally, lower SES presents increased cardiovascular and other health risks possibly due to limited influence, resources, and opportunities, as well as a poorer physical environment. (43)

For example, rates of heart disease mortality and all-cause morbidity have risen (primarily for men) in Eastern Europe since the 1960s in contrast to substantial declines in Western Europe, Canada, Japan, and the U.5. (117) This has been attributed by public health officials to 'lifestyle" factors such as smoking, alcohol, and a fatty diet, rather than, for example, environmental pollution. (118) However, the post­World War 2 period was also a period of urbanization, social migration, industrialization based on the principles of Taylorism, and introduction of and adjustment to a political system which allowed citizens limited control both in society and in the workplace. We need to consider the possible effects of these social changes not only on lifestyle behaviors, but on the prevalence of "job strain," or more directly on cardiovascular health.

Just as the elimination of infectious diseases as the major causes of mortality over the last century occurred due to social changes, improvements in sanitation and nutrition, and elimination of slum conditions (and just as the reappearance of diseases such as tuberculosis has resulted from social neglect), chronic diseases are related to the physical and social environments in which people live and wore Our social epidemiological model of illness explicitly recognizes that work reorganization, workplace democracy, and broader societal changes (social and economic democracy) are needed to reduce the risk of cardiovascular disease and improve emotional well-being.

ACKNOWLEDGMENTS

The authors would like to thank Philip Landrigan, David LeGrande, and Dominic Tuminaro for their advice on portions of this article, as well as the suggestions of anonymous reviewers.

REFERENCES

1. For example, in 1984, Dr. Robert Karasek reviewed a film on stress produced by the Federal Aviation Administration and shown to all air traffic controllers. The film stated that stress depends on demands and the individual's coping style-control over the situation was not mentioned. To cope with stress, the film advised controllers to question that real world events are causes of stress, rather interpretations of events are the major cause. The film stated that employee expectations of fair treatment are often emotional in the modern setting, and that workers' misformed expectations of fairness are the source of problems. Finally, exclusively individual solutions (for example, visualizations. of babbling brooks) were recommended. (2)

2. U.S. Congress. House. Committee on Public Works and Transportation, Hearing before the Subcommittee on Investigations and Oversight. Status of the Air Traffic Control System, 98th Congress, May 4, June 7; October 28, 1983; March 20, 21, 22, 27, 28, 29; April 4; June 26, 27, 1984; H. Rept. 98-93 pp. 785-787.

3. A number of recent developments suggest a growing awareness of the need to accord stress a major role in the occupational health agenda, including the publication of three essential books on this topic, by Karasek and Theorell (5), Johnson and Johannsson (51), and the International Labor Office. (36) Similarly, recent major conferences have focused on this topic: "Participatory Approaches to Improving Workplace Health," Labor Studies Center, University of Michigan, June 3-5, 1991 (presentation summaries available), and the 1990 and 1992 job stress conferences sponsored by the National institute for Occupational Safety and Health (NIOSH) and the American Psychological Association (APA). (Some presentations available in "Stress and Well­Being at Work: Assessments and Interventions for Occupational Mental Health," Washington, DC APA)

4. Sauter, S. L, Murphy, L R, Hurrell, J.J., Jr. "Prevention of Work­Related Psychological Disorders," American Psychologist 45 (1990):114~1158.

5. Karasek, R, Theorell T. Healthy Work. New York: Basic Books, 1990.

6. The term "stress" is used in this paper to refer to the broad range of psychosocial factors and their insulting mechanisms that affect the worker due to impacts on behavioral, psychological, or physiological outcomes. Within the job stress research community, the term "strain" is typically used to indicate the short­term or intermediate effect of job stress (for example, alterations in the hormonal system of the body), which eventually lead to the development of disease. The terminology of "stress" leading to "strain" and then to disease is actually borrowed from the way these terms are used in engineering. In this paper, we use the term job strain" in a more specific way to refer to the objective workplace causes of "stress" described in the Karasek job strain" model

7. Johnson J. V., Hall E M., Theorell T. "Combined effects of job strain and social isolation on cardiovascular disease morbidity and mortality in a random sample of the Swedish male working population,­ Scandinavian Journal of Work, Environment and Health 15 (1989):271­279.

8. Quinn, M. M., Buiatti, E. "Women Changing the Times," New Solutions 1 (1991):48-56.

9. Brandt, B. "The Problem of Overwork in America Today,. New Solutions 2 (1991 );50-65.

10. Kasinsky, R G. "Sexual Harassment A Health Hazard for Women Workers," New Solutions 2 (1992): 74-83.

11. Spangler, E. "Sexual Harassment: Labor Relations by Other Means," New Solutions 3 (1992):23­30.

12. Schnall, P. L, Landsbergis, P. A. Baker, D. "Job Strain and Cardiovascular Disease," Annual Review of Public Health, 1994 (under review)

13. Schurman, S. J., Israel, B. A., Hugentobler, M. K. "Changing the Work Environment to Reduce Stress: Review of Interventions in the United States and Recommendations for Research and Practice," in preparation.

14. American Heart Association. 1993 Heart and Stroke Facts. Dallas, TX: AHA, 1992.

15. This estimate of population attributable risk was derived from five studies in which multivariate models were used to calculate the association between "job strain" and heart disease. These models controlled for other risk factors (for example, age, education, cigarette smoking, serum cholesterol, and blood pressure) in many cases. To the extent that potential confounding was not adequately controlled for, the 23 percent figure may be an overestimate of the potential benefits of inducing "job strain". However, to the extent that standard heart disease risk factors are in the causal pathway between "job strain" and heart disease, controlling for them provides a conservative estimate of risk-an underestimate of the true complete effect of exposure to "job strain."

16. Alfredsson, L, Spetz, C L, Theorell, T. "Type of Occupation and Near­future Hospitalization for Myocardial Infarction and Some Other Diagnoses," International Journal of Epidemiology 14 (1985):378­388.

17. Falk, A., Hanson, B. S., Isacsson, S­O., Ostergren, P­O. "Job Strain and Mortality in Elderly Men: Social Network, Support, and Influence as Buffers", American Journal of Public Health 82 (1992):1136 1139.

18. Astrand, N. E., Hanson, B. S., Isacson, S. O. "Job Demands, Job Decision Latitude, Job Support, and Social Network Factors as Predictors of Mortality in a Swedish Pulp and Paper Company," British Journal of Industrial Medicine 46 (1989):334­340.

19. Theorell, T., Perski, A., Orth­Gomer, K, Hamsten, A., deFaire, U. "The Effect of Returning to Job Strain on Cardiac Death Risk After a First Myocardial Infarction Before Age 45," International Journal of Cerdiology 30 (1991):61­67.

20. LaCroix, A. Z High Demand/Low Control Work and the Incidence of CHD in the Framingham Cohort. Doctoral dissertation. Chapel Hill, NC: University of North Carolina, 1984.

21. Haan, M. N. "Job Strain and Ischaemic Heart Disease: An Epidemiological Study of Metal Workers," Annals of Clinical Research 20 (1988):143­145.

22. Green, K L, Johnson, J. V. "The Effects of Psychosocial Work Organization on Patterns of Cigarette Smoking Among Male Chemical Plant Employees", American Journal of public Health 80 (1990):1368­1371.

23. Mensch, B. S., Kandel, D. B. "Do Job Conditions Influence the Use of Drugs", Journal of Health and Social Behavior 29 (1988):169-184.

24. Pieper, C., LaCroix, A. Z, Karasek, R A. "The Relation of Psychosocial Dimensions of Work with Coronary Heart Disease Risk Factors: A Meta­analysis of Five United States Data Bases", American Journal of Epidemiology 129 (1989):483­494.

25. Theorell, T., Perski, A., Akerstedt, T., Sigala, F., Ahlberg­Hulten, G., Svensson, J., Eneroth, P. "Changes in Job Strain in Relation to Changes in Physiological States", Scandinavian Journal of Work, Environment and Health 14 (1988):189­196

26. Van Egeren, L F. "The Relationship Between Job Strain and Blood Pressure at Work, at Home, and During Sleep," Psychosomatic Medicine 54 (1992):337­343.

27. Blumenthal, J. A., Siegel, W. C, Phillips, B. G. "Job Strain Affects Blood Pressure in the Laboratory and During Daily Life in Parents with Mild Hypertension," Hypertension (in press).

28. Theorell, T., deFaire, U., Johnson, J., Hall, E. M., Perski, A., Stewart, W. "Job Strain and Ambulatory Blood Pressure Profiles," Scandinavian Journal of Work, Environment and Health 17 (1991):380-385.

29. Schnall, P. L, Schwartz, J. E, Landsbergis, P. A., Warren, K, Pickering, T. G. "The Relationship Between Job Strain, Alcohol and Ambulatory Blood Pressure," Hypertension 19 (1992):488-494.

30. Schnall, P. L, Landsbergis, P. A., Schwartz, J. E., Warren, K., Pickering, T. G. "The Relationship Between Job Strain, Ambulatory Blood Pressure and Hypertension," Paper presented at the Ninth International Symposium on Epidemiology in Occupational Health, Cincinnati, OH, September, 1992.

31. Light, K C, Turner, J. R, Hinderliter, A. L "Job Strain and Ambulatory Work Blood Pressure in Healthy Young Men and Women," Hypertension 20 (1992):214-218.

32. Theorell, T., Knox, S., Svensson, J., Waller, D. "Blood Pressure Variations During a Working Day at Age 28: Effects of Different Types of Work and Blood Pressure Level at Age 18," Journal of Human Stress 11 (1985):36-41.

33. Frankenhauser M, Johansson G. "Stress at Work Psychobiological and Psychosocial Aspects," International Review of Applied Psychology 35 (1986):287­299.

34. Schnall, P. L, Pieper, C, Schwartz, J. E., Karasek, R A., Schlussel, Y., Devereux, R B., Ganau, A., Alderman, M., Warren, K, Pickering, T. G. "The Relationship Between Job Strain, Workplace Diastolic Blood Pressure, and Left Ventricular Mass Index: Results of a Case-control Study," Journal of the American Medical Association 263 (1990):1929­1935.

35. Karasek, R A., Gordon, G., Pietrokovsky, C, Frese, M., Pieper, C, Schwartz, J., Fry, L, Schirer, D. Job Content Instrument Questionnaire and User s Guide. Los Angeles, CA University of Southern California, 1985.

36. DiMartino, V. (Ed.) Conditions of Work Digest: Preventing Stress at Work. Geneva: International Labor Office, 1992.

37. Johnson, J. V. "Collective Control: Strategies for Survival in the Workplace," International Journal of Health Services 19 (1989):469-480.

38. lsrael, B. A., Schurman, S. J., House, J. S. "Action Research on Occupational Stress: Involving Workers as Researchers," International of Health Services 19 (1989):135-155.

39. Caplan, R D., Cobb, S., French, J. R P., Jr., Van Harrison, R, Pinneau, S R, Jr. Job Demands and Work Health Cincinnati, OH: National Institute for Occupational Safety and Health, 1975 (Publication No. 75-168).

40. Hurrell, J. J., McLaney, M. A. "Exposure to Job Stress - A New Psychometric Instrument,. Scandinavian Journal of Work, Environment and Health 14 (suppl 1) (1988):27­28.

41. Haynes, S. G., Feinleib, M. "Women, Work and Coronary Heart Disease: Perspective Findings from the Framingham Heart Study," American Journal of Public Health 70 (1980):133-141.

42. Hall, E. M. "Double Exposure: The Combined Impact of the Home and Work Environments on Psychosomatic Strain in Swedish Men and Women", International Journal of Health 22 (1992):239­260.

43. Marmot, M., Theorell, T. "Social Class and Cardiovascular Disease, International Journal of Health Services 18 (1988):659-674.

44. Siegrist, J., Peter, R, Junge, A., Cremer, P., Seidel, D. "Low Status Control, High Effort at Work and Ischemic Heart Disease: Prospective Evidence from Blue collar Men," Social Science and Medicine 31 (1990):1127­1134.

45. Ivancevich, ]. M., Matteson, M. T., Freedman, S. M., Phillips, ]. S. "Worksite Stress Management Interventions," American Psychologist 45 (1990):252­261.

46. This section is based on a more detailed review in (13).

47. Examples are classic texts from this period: Lewin, K Field Theory in Social Science New York: Harper, 1951; Walker, C R, Guest, R H. Man on the Assembly Line. Cambridge: Harvard University Press, 1952; MacGregor, D. The Human Side of Enterprise. New York: McGraw­Hill, 1960; Kornhauser, A. Mental Health of the Industrial Worker. New York: Wesley, 1965.

48. For recent critical reviews, see the series Woodman, R, Pasmore, W. (Eds.) Research in Organization Change and Development. Greenwich, CT: JAI Press, 1987.

49. Parker, M. Inside the Circle. Boston: South End Press, 1985.

50. Gardell, B., Gustavsen, B. "Work Environment Research and Social Change: Current Developments in Scandinavian," Journal of Occupational Behavior 1 (1980):3-17.

51. Johnson, J. V., Johannsson, G. The Psychosocial Work Environment. Amityville, NY: Baywood, 1991.

52. Hanlon, M.D. "Reducing Hospital Costs Through Employee Involvement Strategies," National Productivity Review 5 (1986):2231.

53. Kopelman, R E. "Job Design and Productivity: A Review of the Evidence," National Productivity Review 4 (1985):237­255.

54. Gordon, M. E., Burt, R.E. "A History of Industrial Psychology's Relationship with American Unions: Lessons from the Past and Directions for the Future," International Review of Applied Psychology 30 (1981):137­156

55. Buller, P. F., Bell, C. H. "Effects of Team Building and Goal Setting on Productivity: A Field Experiment," Academy of Management Journal 29 (1986):305-328.

56. Grenier, G. J. "Twisting Quality Circles to Bust Unions," AFL­CIO News, May 14, 1983.

57. Jones, L. M., Bowers, D. G., Fuller, S. M. Rcport of Findings 1984: Task Force on Management ond Employse Relationships. Washington, DC: Federal Aviation Admiristration, November 7, 1984.

58. "APWU Settles USPS Company Union Charges," AFL­CIO News, May 11, 1992.

59. "Partnenhip Sweetens Work at Nabisco Bakeries," AFL­CIO Ncws, April 29, 1991.

60. "Cooperative Successes Typified at Ohio Plant," AFL-CIO News, June 22, 1992.

61. Whyte, W. F. Participatory Action Research. Newbury Park, CA: Sage, 1990.

62. Karasek, R. "Stress Prevention Through Work Reorganization: A Sununary of 19 lnternational Case Studies," in DiMartino, V. (ed.) Conditions of Work Digest, Preventing Stress at Work. Geneva lntenutional Labor Office, 1992: 23-41.

63. Jackson, S. E. "Participation in Decision Making as a Strategy for Reducing Job Related Strain," Journal of Applied Psychology 68 (1983): 3-19.

64. Golembiewski, R, Hilles, R, Daly, R. "Some Effects of Multiple OD Intervendons on Burnout and Worksite Features," Journal of Applied Behavioral Science 23 (1987): 295-314.

65. Israel, B. A., Schunnan, S. J., Hugentobler, M. K., House, J. "A Participatory Action Research Approach to Reducing Occupational Stress in the United States," in DiMartino, V. (Ed.) Conditions of Work Digest: Preventing Strcss at Work. Geneva: International Labor Office, 1992: 152­163.

66. Heaney, C., Israel, B., Schurrnan, S., House, J., Baker, B., Hugentobler, M. "Evaluation of a Participatory Action Research Approach to Reducing Stress." Paper presented at the APA/NIOSH Conference on Occupational Stress, Washington, DC, November, 1992.

67. Cahill, J. "Computers and Stress Reduction in Social Service Workers in New Jersey," in DiMartino, V. (Ed.) Conditions of Work Digest: Preventing Stress at Work. Geneva: International Labor Office, 1992: 197­203.

68. Lerner, M. Occupational Stress Groups and the Psychodynamics of the World of Work. Oakland, CA: Institute for Labor

and Mental Health, 1985.

69. May, L."A Union Programme to Reduce Work and Family Stress Factors in Unskilled and Semi­Skilled Workers on the East Coast of the United States," in DiMartino, V. (Ed.) Conditions of Work Digest: Preventing Stress at Work. Geneva: International Labor Office, 1992: 164­171.

70. Landsbergis, P. A., Silverman, B., Barrett, C., Schnall, P. I. "Union Stress Committees and Stress Reduction in Blue and White Collar Workers," in DiMartino, V. (Ed.) Conditions of Work Digest: Preventing Stress at Work. Geneva: International Labor Office, 1992:144-151.

71. Mergler, D. "Worker Participation in Occupational Health Research: Theory and Practice," International Journal of Health Services 17 (1987):151­167.

72. Thus, it is analogous to the role proposed for workers in making industrial hygiene more effective, that is, involving workers in inspections, assessment of hazards and health symptoms; problem­solving; and changing work practices and organization as well as technical improvements (Senn, E. "PIaying Industrial Hygiene To Win," New Solutions (1991):72­81).

73. Hugentobler, M. K, Robins, T. G., Schurman, S. J. "How Unions Can Improve the Outcomes of Joint Health and Safety Training Programs`" Labor Studies Journal 15 (1990):16-38.

74. AFL-CIO Committee on the Evolution of Work 77. The Changing Situation of Workers and Their Unions. Washington, D.C. AFL-CIO, 1985.

75. Cahill, J. "Economic and Non­Economic Sources of Job Satisfaction in Higher Education," Paper Presented at the American Association for Higher Education, Washington, D.C., March, 1993.

76. Israel, B., Schurman, S. "Social Support, Control and the Stress Process," in Glanz, K, Lewis, F., Rimer, B. (Eds.) Health Behavior and Health Education: Theory, Research snd Practice. San Francisco: Jossey­Bass, 1990:187­215.

77. "Moving Towards Worker-Oriented Participatory Research," Panel at the American Public Health Association, San Francisco, CA, October, 1993.

78. Bongers, P. M., deWinter, R R, Kompier, M. A. J., Hildebrandt, V. H. "Psychosocial Factors at Work and Musculoskeletal Disease: A Review of the Literature," Scandinavian Journal of Work Environment and HeaIth (in press).

79. Brett, K., Strogatz, D., Savitz, D. "Occupational Stress and Low Birth Weight Delivery." Presented at the 24th Annual Meeting of the Society of Epidemiological Research, Buffalo, NY, June 1991. American Journal of Epidemiology 134 (1991): 722­723.

80. National Institute for Occupational Safety and Health. Indoor Air Quality and Work Environment Study. Cincinnati, OH: NIOSH, 1991 (HETA 88-364-2102).

81. Shostak, A. B. "Union Efforts to Relieve Blue­Collar Stress," in C.L. Cooper, and M. J. Smith (Eds.) Job Stress snd Blue-Collar Work. New York: Wiley, 1985: 195-205.

82. Service Employees International Union. Stress: Contract Provisions. Washington, DC: SEIU, 1983.

83. Unions have also developed valuable worker education materials on job stress that use the Small Group Activity Method and the principles of empowerment education. Sources include the Communications Workers of America, District 1; District 65, UAW; and the Labor Institute, in New York City; 9 to 5, National Association of Working Women, Cleveland, OH; and the Workers Health and Safety Centre, Kingston, Ontario.

84. Landsbergis, P. "Occupational Stress Among Nurses: New Developments in Theory and Prevention," in J.H. Humphrey (Ed.) Human Strcss: Currcnt Selected Research (Vol. 3). New York: AMC Press, 1989: 173-195.

85. AFL­CIO News, January 26, 1985; May 23,1988; July 8, 1989; February 1, 1993; Labor Notes, December 1985; October, 1989.

86. Dena Bunis, personal communication, February, 1993.

87. Bernard, B. "Psychosocial Factors for Musculoskeletal Disorders," Paper to be Presented at the American Public Health Association, San Francisco, CA, October, 1993.

88. U.S. Congress, Office of Technology Assessment, "The Electronic Supervisor: New Technology, New Tensions," OTA-CIT­333. Washington, DC: U.S. Government Printing Office, 1987.

89. "Northern Telecom Ends Worker Monitoring," VDT Ncws, March/April 1992.

90. Smith, M. J., Carayon, P., Sanders, K. J., Lim, S­Y., LeGrande, D. "Employee Stress and Health Complaints in Jobs With and Without Electronic Performance Monitoring," Applicd Ergonomics 23 (1992):17­27.

91. Ramirez, A. "A.T.&T. and Unions Praise New Pact," New York Times, July 3,1992.

92. National Institute for Occupational Safety and Health. Health Hazard Evaluation Report. Cincinnati, OH: NIOSH, July 1942 (HETA 89­299­2230).

93. Bureau of Labor Statistics. "Survey of Occupational Injuries and Illnesses in 1991." Washington, DC: US. Department of Labor, 1992

94. "The Electronic Supervisor," op. cit., p. 39, 46-48, 57; Massachusetts Coalition on New Offlce Technology, "Electronic Monitoring in the Workplace: Supervision or Surveillance?" Boston, 1989.

95. Shostak, A. B., Skocik, D. The Air Controllers' Controversy. New York: Human Sciences Press, 1986.

96. Landsbergis, P. "Is Air Traffic Control a Stressful Occupation?" Labor Studies Journal 11 (1986) :117­134.

97. Rose, R. M., Jenkins, C. D., Hurst, M. W. Air Traffic Controller Health Change Study. Washington, DC: Department of

Transportation, 1978.

98. Tesh, S. "The Politics of Stress: The Case of Air Traffic Control," International Journal of Health Services 14 (1984):569-587.

99. Bowers, D. G. "What Would Make 11,500 People Quit Their Jobs," Organizational Dynamics (1983):5-19.

100. Hund, R W., Kriesky, J. K. "The Rise and Demise of PATCO Reconstructed," Industrial and Labor Relations Review 40 (1986):115-121.

101. "Five Years After PATCO," Frequent Flyer Magazine, August, 1986:54-55, 62.

102. Golden, T. "Was Illness at Bridges in the Minds of Workers," New York Times (March 12, 1990).

103. Cases of similar physical symptoms among a group of workers without an identifiable pathogen (known as "epidemic psychogenic illness," EPI) have been extensively analyzed by NIOSH and others. (For excample: Colligan, M., Pennebaker, J., Murphy, L. (Eds.) Mass Psychogenic Illncss: A Social Psychological Analysis. Hillsdale, NJ: Lawrence Erlbaum, 1982.) Work organization in workplaces where EPI has occurred has been chracacterizd as repetitive or boring work with rigid pacing and high job pressures, strict rules, and a lack of communication and social interaction. While EPI can therefore be regarded as a desperate reaction to excessive "job strain," another explanation is that low­level chemical exposure and stress may have synergistic effects.

104. DeCarlo. D. T., Gruenfeld, D. H. Stress in the American Workplace. Fort Washington, PA: LRP Publications, 1989: 11.

105. California Workers' Compensation Institute, "Mental Stress Claims in California Workers' Compensation -- Incidence, Costs and Trends," San Francisco: California Workers' Compensation Institute, 199O.

106. Elisburg, D. "Workers' Compensation Stress Claims: Employee Issues," Paper Presented at the APA/NIOSH Conference on Occupational Stress, Washington, DC, November, 1992.

107. Deutsch, S. "Work Environment Reform and Industrial Democracy," Sociology of Work ond Occupations 8 (1981):180-194.

108. Service Employees lnternational Union. The High Cost of Short Staffing. Washington, DC: SElU, 1992.

109. Service Employees International Union. The National Nurse Survey. Washington, DC SEIU, 1993.

110. Byrne, M. "Reich, Council Explore Ways to Get Moving," AFL-CIO News, March, 1,1993.

111. Richardson, C "Training to Improve Working Conditions, and Build Union Strength," Paper Presented at the International Metalworkers' Federation Conference, Helsinki, Finland, September, 1991.

112. Richardson, C. "Technology, Training and Work Organization," New Solutions 3 (1993): 5-6.

113. Johnson, J. L. "Work Injury and Stress," Paper Presented at the APA/NlOSH Conference on Occupational Stress, Washington, DC, November, 1992.

114. Henningsen, G. M., Hurrell, J. J., Baker, F., Douglas, C., MacKenzie, B. A., Robertson, S. K, Phipps, F. C. "Measurement of Salivary Immunoglobulin A as an Immunologic Biomarker of Job Stress," Scandinavian Journal of Work, Environment and Health 18 (suppl 2) (1992):133-136

115. Schnall, P. L., Kern, R. "Hypertension in American Society: An Introduction to Historical Materialist Epidemiology," in P. Conrad and R. Kern (eds.) The Sociology of Health and Illness. New York St. Martin's Press, 1981: 97­122.

116. Brenner, H. Economy, Society and Health, Washington, DC: Economic Policy Institute, 1992.

117. Uemura, K, Pisa, Z. "Trends in Cardiovascular Disease Mortality in Industrialized Countries Since 1950," World Health Statistics Quarterly 41 (1988):155-178.

118. Levenstein, C. "Occupational Health in Eastern Europe During Political and Economic Transition," Paper Presented at the American Public Health Association, Washington, DC, November, 1992.