Industrial Psychology and the Study of Fatigue in Britain

The scientific study of employee well-being can be traced to the founding of the Health of Munition Workers’ Committee (HMWC) in Britain during World War I (Kreis, Reference Kreis1995). As discussed by Kreis, a major activity of the HMWC was to investigate working conditions that would contribute to fatigue and lost productivity. Whereas counterparts in the United States were focused almost entirely on maximizing efficiency, the HMWC researchers were interested in how working conditions that led to fatigue would affect the well-being of employees, for example by increasing anxiety and boredom. At the end of the war, the HMWC was replaced by the Industrial Health Research Board that continued the study of worker well-being.

The leading figure who helped shape early British industrial psychology was Charles Myers, whose writing provided a more balanced treatment of efficiency and well-being than was seen across the Atlantic. His book Mind and Work (Myers, Reference Myers1920) dealt with employee accidents, boredom, fatigue, and mental health. He described research on rest pauses showing that they could increase productivity even though the amount of time spent working was less. Myers’s (Reference Myers1926) industrial psychology textbook included many topics that are prominent in modern occupational health psychology, including accidents, mental health, negative emotions particularly anxiety and boredom, and withholding of output (today counterproductive work behavior). Some of the topics in his text were those a reviewer suggested I remove from mine nearly 70 years later.

Socio-technical Systems

One of the most influential papers in the evolution of thinking about employee well-being was written by Trist and Bamforth (Reference Trist and Bamforth1951). Their paper documented how the social disruption of technological change adversely affected coal miners in Britain. This classic paper talks about how increased isolation and insufficient autonomy were stressful and led to strains of emotional distress and absence. It serves as a foundation for the study of working conditions as stressors, and for socio-technical systems theory that has dominated British organizational research ever since.

Socio-technical systems theory provides a point of view in which the social system and technological system are considered together. The principle of joint optimization suggests that in the ideal workplace, the social and technical systems are designed in a way that best fits with one another. This extends the purpose of human factors from designing technology to fit people to the idea that you have to consider elements of both in designing ideal systems. This allows employees to perform tasks efficiently while reducing strain. A number of established practices such as autonomous work groups and job enrichment can be considered from the point of view of social-technical systems.

The Rise of Stress Research in Scandinavia

Researchers in Scandinavian counties, particularly Norway and Sweden, became early leaders in the study of general and occupational stress (Barling & Griffiths, Reference Barling, Griffiths, Tetrick and Quick2011). This work on stress began in the 1960s and 1970s, underscoring the importance of the workplace (Cooper & Dewe, Reference Cooper and Dewe2004). One particular focus was on the link between occupational stress and cardiovascular disease, particularly the impact of life changes, such as increased responsibility at work (Theorell, Reference Theorell2019). A number of prominent researchers contributed to the foundation upon which occupational stress research would be built.

During the 1980s most of the research on occupational stress was coming from Europe and particularly Scandinavia. A content analysis by Erez (Reference Erez, Triandis, Dunnette and Hough1994) examined the topics in applied psychology articles by country. She found that the majority of Swedish workplace research involved employee health and well-being, whereas only 5% of American articles covered those topics. The number of researchers in the United States who were studying employee health and stress was small, and it would be more than a decade before this topic became mainstream in the United States.

Developments in North America

American industrial psychology developed at the same time as its British counterpart, beginning during World War I. Whereas the British linked employee productivity and employee well-being, Americans focused primarily on productivity. Their study of worker well-being would take decades to fully develop, beginning with the study of job satisfaction and eventually embracing a broader focus on employee health.

Occupational Stress: The Explosion

The study of employee well-being began slowly, with the main interest in Britain and Scandinavia. The topic got off to a slower start in North America, but by the end of the twentieth century, research had exploded. During the 1980s, a handful of researchers in North America and elsewhere became interested in the connection between working conditions and employee health/well-being from the perspective of occupational stress. By the 1990s, the study of this topic took off, and interest has accelerated since. Figure 1.1 shows the number of papers published on occupational stress from 1950 through 2019. I queried the Web of Science (WoS) database using the search term “occupational health” to get a count of papers published each year. I combined yearly counts into decades from the 1950s to the 2010s. As the figure shows, there were very few papers published until the 1980s, with a rapid acceleration after that. Fifty-seven percent of occupational stress papers in the WoS database were published between 2010 and 2019. In 2019 alone, there were as many papers as in all the years up to 1993. Clearly, occupational stress is a major topic of study worldwide.

Figure 1.1

Occupational Stress as a Discipline

In many ways, the study of occupational stress can be considered a discipline in its own right because it shares many of the features with recognized disciplines.

• It is interdisciplinary. Occupational stress researchers come from many disciplines, including I-O psychology and other areas of psychology, management, and the health sciences. Papers on occupational stress are published in journals from many different disciplines.

• It has its own journals. Work & Stress was founded in 1981 by Tom Cox at the University of Nottingham in the UK. Although the focus at the time was occupational stress, today it has broadened to incorporate all topics in occupational health psychology. Other journals focus on stress in general, including occupational stress (e.g., International Journal of Stress Management), whereas occupational health psychology journals such as Journal of Occupational Health Psychology and Occupational Health Science are major outlets for occupational stress research.

• It has its own conference. The biannual Work, Stress, and Health conference in North America began with a main focus on occupational stress, although in recent years it has broadened to include other occupational health psychology (OHP) topics.

• There are books on the topic. Many books, including this one, are concerned entirely or mainly with occupational stress. There is even the Handbook of Work Stress, published in 2005 by Sage Publications, and an annual book series, Research in Occupational Stress and Well-Being, published since 2001 by JAI.

Early Occupational Stress Contributions

There are many contributions to the field of occupational stress coming from Europe and the United States. One that linked the early focus on job satisfaction to the broader concern with stress was the book Work and Well-being (Warr & Wall, Reference Warr and Wall1975). Although most of the book reviewed research on job satisfaction, there was a chapter devoted to occupational stress. This book was my introduction to the topic of occupational stress and inspired my lifelong interest in the topic. There are many early contributions to the study of occupational stress, but three deserve mention based on their impact in shaping the study of occupational stress and occupational health:

• Contributions by the Institute for Social Research. The Institute for Social Research at the University of Michigan began a program of research on employee well-being at the end of the 1950s (Cooper & Dewe, Reference Cooper and Dewe2004). This program produced many influential products that helped shape the field of occupational stress and broader occupational health. One of the most significant was Katz and Kahn’s (Reference Katz and Kahn1966) The Social Psychology of Organizations. In this book, they discuss the role stressors of role ambiguity, role conflict, and role overload that came to dominate the study of occupational stress research for quite some time. Another is the Job Demands and Worker Health study (Caplan et al., Reference Caplan, Cobb, French, Van Harrison and Pinneau1975), funded by the National Institute of Occupational Safety and Health (NIOSH). This large-scale study of high stress occupations showed a clear link between stressors and strains.

• Beehr and Newman’s (Reference Beehr and Newman1978) Occupational Stress Model. One of the most influential papers published in the early days of occupational stress research was Beehr and Newman’s (Reference Beehr and Newman1978) review of the newly emerging literature on occupational stress. One of their major contributions was the inclusion of a model that outlines the complex stress process involving environmental characteristics, individual differences, and both short-term and long-term outcomes for employees and organizations.

• The Demand-Control Model. The Demand-Control Model (Karasek, Reference Karasek1979), built on the idea that stressful working conditions (stressors) would lead to poor health and well-being (strains). It suggested that control would buffer the adverse effects of demands and reduce their negative impact. This provided an important stressor-strain framework that would drive much subsequent research and led to the more complex Demand-Control-Support model, which added a buffering effect of social support (Karasek & Theorell, Reference Karasek and Theorell1990). These models drew attention to the prominence of control in the occupational stress process, and although the buffering effect of both control and support have proven to be elusive (de Lange et al., Reference de Lange, Taris, Kompier, Houtman and Bongers2003), there is little doubt that control and support play important roles.

Twenty-First Century Developments

As the study of occupational stress expanded into the twenty-first century, it developed in several directions. Some directions were derived from the general stress literature, whereas others built on work from the occupational domain.

From Occupational Stress to Occupational Health

As the study of occupational stress and worker well-being developed, it became clear that it dealt with a very broad and complex set of issues. Researchers who studied these issues, many of whom were not psychologists, became convinced that it was large enough to represent a separate field within psychology, and the field of OHP emerged. Quick (Reference Quick1999) credits Raymond et al. (Reference Raymond, Wood and Patrick1990) as using the term for the first time in print. Their vision was to integrate the training of OHP practitioners and researchers across a variety of fields, including business, medicine, nursing, occupational health, and psychology. The extent to which this integration has been successful is debatable, as most training is focused mainly in psychology, but there is no doubt that the field of OHP has emerged, initially in Europe and then in North America.

OHP is a field concerned with psychological factors in employee health, safety, and well-being. Much of its attention focuses on the psychosocial aspects of the physical and social working environment. Occupational stress is central to the field of OHP, but OHP covers more territory than stress, and its point of view is broader. It connects with fields concerned with occupational health and safety, investigating how psychosocial aspects of the workplace contribute to accidents, injuries, and illness. It takes a public health perspective in being concerned with exposures to both physical and psychological risks in the workplace, and in recognizing the distinction between primary prevention (changing the job), secondary prevention (giving employees tools), and tertiary prevention (treating illness and injury).

The field got its early impetus in the United States from a partnership between the American Psychological Association and NIOSH. Over a period of about six years, they awarded small seed grants to more than a dozen graduate programs in the United States, most of them I-O psychology doctoral programs, to stimulate the training of OHP. Five of these programs received NIOSH training grants: Colorado State University, Ohio University, Portland State University, University of Connecticut, and University of South Florida. The partnership also supported two small conferences at Portland State University and University of South Florida where ideas for developing the field in the United States were discussed. This resulted in the founding of the Society of Occupational Health Psychology in the United States.

OHP can be considered a major subfield within psychology, although it is interdisciplinary and goes beyond psychology. There are several characteristics that define it as an established field.

• OHP Societies. There are societies devoted entirely to OHP. The UK has the European Academy of Occupational Health Psychology (EAOHP); in the United States, there is the Society for Occupational Health Psychology (SOHP).

• OHP Graduate Training. There are more than two dozen universities in both Europe and North America that provide graduate training in OHP. Most are attached to I-O psychology graduate programs, but some are attached to other programs, such as cognitive or social psychology.

• OHP Journals. APA has supported the development of OHP by publishing Journal of Occupational Health Psychology. In 2016, SOHP launched Occupational Health Science to be an interdisciplinary OHP journal. Work & Stress originally devoted to occupational stress evolved into a broader OHP journal as it became affiliated with the EAOHP (Cox & Tisserand, Reference Cox and Tisserand2006).

• OHP Conferences. Europe has an annual OHP conference, and as mentioned earlier, the Work, Stress, and Health conference sponsored initially by APA and NIOSH has broadened its purview to OHP and is now cosponsored by SOHP.

• OHP Books. There are many books dealing with OHP topics, but most notable is the Handbook of Occupational Health Psychology, edited by James Quick and Lois Tetrick, and OHP textbooks Occupational Health Psychology, written by Irvin Schonfeld and Chu-Hsiang (Daisy) Chang (Reference Schonfeld and Chang2017), and Essentials of Occupational Health Psychology, by Chris Cunningham and Kristen Black (Reference Cunningham and Black2021).

Today the OHP field is rapidly growing within Europe, the United States, and many other parts of the world as concern for employee health and well-being becomes increasingly recognized. This is reflected in job ads for psychologists that list OHP as a research focus of interest, and in the personal statements of prospective graduate students who note OHP interests. It can be seen in the programs of I-O psychology conferences (European Association of Work and Organizational Psychology and Society for Industrial and Organizational Psychology) and management conferences (Academy of Management, European Academy of Management, Southern Management Association) as OHP topics are frequent topics of sessions. All of this research attention has taught us much about the connection between the work environment and employee health, but also where more work is needed on how best to manage organizations to maximize employee health and well-being.

Realign Management Thinking

There is a tendency to think of human resources as costs, both the direct costs of salary and benefits and indirect costs of resources provided to employees. This is wrong thinking and should be replaced with the idea of human resource investments that will provide future returns in productivity, quality, and reputation. One of the early insights provided by Myers (Reference Myers1920) was that working fewer hours often resulted in greater total output. Thus, investing in employees by providing rests can lead to a return of even more performance without overly straining the employees. Many other human resource investments can provide impressive returns.

Build a General Safety Climate

A healthy organization has a climate that protects employees from both physical and psychological risks. This means having a general safety climate that includes elements of safety climate (focus on accidents/injuries) and psychosocial safety climate (focus on mistreatment). There are several steps organizations can take in dealing with such climates (Spector, Reference Spector2019b).

1. 1 Make Climate a Goal. Top management needs to adopt the development of a general safety climate as a strategic goal. This means investing resources into safety and considering it a worthwhile investment. It should be considered okay to take time to be safe, even if it means slowing production a little.

2. 2 Communicate Policies. Policies concerning climate need to be disseminated throughout the organization. This involves messaging from not only the top but also all levels of management. Safety, both physical and psychological, should be a common topic of discussion throughout the organization.

3. 3 Model Safety. Safety policies and practices are for everyone. All managers should follow safety rules (e.g., wearing safety gear) and not assume that telling employees is sufficient without modeling that behavior (Kessler et al., Reference Kessler, Lucianetti, Pindek and Spector2020).

4. 4 Take Corrective Action. Direct supervisors need to monitor employee behavior and take corrective action when necessary. Employees should be recognized for safe behavior and corrected for unsafe actions. This means progressive discipline of starting with gentle reminders and escalating to more punitive measures if unsafe behaviors persist.

Select the Right People

It is important to match people’s capabilities to the demands of the job in order to maximize organizational health. Mismatches will not only result in substandard performance but also create undue stress as individuals struggle to perform job tasks. Effective selection systems focus on hiring the best talent for the specific job by identifying job requirements and then using selection tools to assess job applicants. Those tools can minimize personal biases of decision-makers and achieve greater workforce diversity and a more capable workforce that performs better with less effort and stress.

Develop Your Talent

Training and other developmental activities should be considered investments that can not only improve the performance of employees in their current positions but also provide a talent pool from which to draw for higher level positions. An effective training and development program focuses on three elements (Spector, Reference Spector2021).

1. 1 Training Needs Assessment. Training resources are always limited, so they should be invested wisely so they will have the most impact. Needs assessment means conducting research to determine where knowledge and skill gaps exist for employees, and what future knowledge and skill will be needed. This means first identifying gaps and then prioritizing what and who are to be trained.

2. 2 Training Design. Training needs to incorporate known principles that maximize effectiveness. The common approach to corporate training of having employees attend all-day presentations might be practical, but it is not particularly effective in most cases. Well-designed training needs to incorporate the following principles.

   • - General Principles. This means providing the big picture to put the training in context. A training program on a particular application for individuals new to computing should begin with a brief overview of the computer, its main components, and applications in general. Trainees will better understand the particular application they are being taught if they have a general sense of what a computer is and how it works.

   • - Spaced Training. Learning is retained better when training occurs in shorter sessions spread over time rather than one long session. Further, the longer the interval between training sessions, the longer trainees will remember what was trained (Cepeda et al., Reference Cepeda, Coburn, Rohrer, Wixted, Mozer and Pashler2009).

   • - Feedback. In order to learn effectively, people need feedback. They need to know if what they have learned is correct. With training that teaches knowledge, quizzes can be used to show trainees that they understand the material correctly. Skills training should allow practice sessions that include feedback to indicate if something was done correctly.

   • - Match Practice With the Job. The best training allows practice on tasks that are as close as possible to the job itself. For example, flight training makes use of simulators that include elements that match aircraft on the features being trained. High fidelity simulators have cockpits that mimic the real thing and can simulate the motions of an aircraft in response to the actions taken by the trainee.

3. 3 Training Evaluation. It should not be assumed that training was effective just because it was completed. Training needs to be evaluated to see if employees felt it was worthwhile, if they learned anything (e.g., by use of a quiz at the end of training), if they apply what was learned on the job, and if the training resulted in improvements in employee or organizational outcomes (Kirkpatrick, Reference Kirkpatrick1977). Evaluation means conducting research and collecting data that can be used to determine if training can be improved, or if it is worthwhile to do at all. An effective strategy is to do a pilot test of new training on a small group of employees to determine if the training is effective before rolling it out to all employees who will be trained.

Leadership

One of the key elements in a healthy work organization is leadership. Perhaps the two most important functions of leaders, recognized as far back as the Ohio State Leadership Studies (Fleishman & Harris, Reference Fleishman and Harris1962), are providing structure and providing support. Structure concerns the organization of work and the coordination of effort across employees and functions. It helps clarify to employees what their roles are and where to put their efforts. Support means paying attention to employees and providing assistance to them in dealing with the demands of the job. Leaders vary in the extent to which they provide each function, but they are both vital to good leadership.

Work needs to be structured, and the costs of a passive leader who does not perform the structuring function can be substantial in terms of both employee poor performance and stress. Effective structure can minimize many stressors, including role ambiguity, role conflict, and work overload, because it clarifies expectations and balances workload to be most efficient.

Support comes in two broad forms. Instrumental support provides assistance in getting jobs done. This includes showing employees how to perform tasks and taking on some tasks when employees are overloaded. Emotional support helps employees cope with the stress of the job and, in some cases, stress off the job. Emotionally supportive leaders consider the impact of decisions on employees and build trusting relationships. Their support can help buffer the negative impact of stressful jobs and is particularly vital in industries where jobs are inherently stressful, such as first responders and healthcare. Paying attention to employee health, safety, and well-being is important for providing a general safety climate as it makes clear to employees that their health, safety, and well-being are priorities.

Although it is widely acknowledged that support is a vital leadership function, there has been surprisingly little attention paid to how it can best be accomplished. A new line of research has shown that well-intentioned but poorly delivered support can add to employee stress rather than making it better. Gray et al. (Reference Gray, Spector, Lacey, Young, Jacobsen and Taylor2019) identified ten ways in which workplace support by supervisors and others can be counterproductive. For example, support might be unwanted, delivered in a way that is insulting, or not provide what the employee needs. Gray et al.’s research suggests that managers should consider the following.

• Communicate with Employees. Be sure that you understand what support is needed and that the support is wanted. Sometimes an employee needs to struggle to learn how to perform a task and to build confidence. Unless consequences of failure are severe, which can be the case in healthcare, allow employees to figure it out themselves if they wish.

• Explain Rather Than Do. If an employee does not know how to do a task, it is better to teach that employee how to do it than to step in and do it for him or her. In the long run, it can be more effective to explain how to do something and sit with an employee while they try, than to do it and have him or her watch. Learning requires doing and feedback.

• Do Not Be Critical. Often offers of support can be interpreted as criticism. After all, if you are offering to help an employee with a problem, it implies that the employee is not capable of doing the job. Be tactful in approaching employees and do not assume help is needed. An approach like, “How is it going today?” is better than “You look like you could use some help”.

• Support, Don’t Dismiss. When employees are dealing with difficult events, either on or off the job, it can be tempting to try to make them feel better by looking on the bright side, for example telling them that it happened to you once and everything came out all right. This approach can often make someone feel worse by discounting their feelings. Sometimes it is best to say you are sorry they are having the problem, and just be willing to listen.

Provide Rewards for Contributions

Rewards for performance are important because they provide feedback that employee efforts are recognized and worthwhile to management. One of the major stress theories, the Effort-Reward Imbalance Theory, recognizes that unrewarded effort can be stressful (Siegrist, Reference Siegrist1996). People whose efforts are unrewarded can feel exploited and unfairly treated.

There are many ways to reward employees, not all of which are monetary. Pay-for-performance systems can be effective if administered in a fair and transparent way, but systems in which most pay is linked to output (e.g., commission sales) can be stressful. Nonmonetary rewards can be as effective in terms of feedback and motivation. It is not a stretch to suggest that employees will work for praise, although obviously they will not work for praise alone. Providing praise and showing appreciation can enhance employee motivation because it provides feedback that their performance is good, and it signals that the employee is having a positive impact.

Praising employees and showing appreciation does not always come naturally to supervisors. Spector (Reference Spector2019a) offers five tips for using these sorts of rewards.

• Show appreciation for both effort and results. Results ultimately come from effort, so you want to encourage employees to remain motivated even when they have not yet accomplished their goals. Rewarding effort can be used to motivate employees by acknowledging improved performance over time.

• Appreciation can be shown with a simple “thank you”.

• Be clear about what you are rewarding. Thank employees for specific efforts or accomplishments (e.g., a new person best for monthly sales), rather than providing vague statements not tied to particular actions.

• Use rewards to develop people. You can use the principle of successive approximations to reward employees for getting closer and closer to the level of performance you need. Reward an employee for initial efforts, and as performance improves over time, give occasional praise for doing better.

• Rewards need to be fair. Set your own standards and process for providing rewards and then apply them equally to all employees. Praising one employee and ignoring others can create bad feelings and conflict within a group. You need to acknowledge your stars, but average performers also need appreciation and recognition to maintain efforts to improve.

Impacts in Terms of Economic Costs

Financial impact is often included while assessing impacts of organizational stressors at the organizational or national level. Criteria such as productivity loss, sickness absenteeism, presenteeism, medical claim, or health care cost have been used to estimate direct and indirect costs in past cost-of-illness research (e.g., Hassard et al., Reference Hassard, Teoh, Visockaite, Dewe and Cox2018a, Reference Hassard, Teoh, Visockaite, Dewe and Cox2018b, Reference Hassard, Teoh and Cox2019). It should be emphasized that direct and indirect costs merely reflect a myopic snapshot of much broader impacts. Intangible costs to societies as a whole – such as physical and psychological sufferings, reduced quality of individual and family lives, and loss of lives (Goh et al., Reference Goh, Pfefer and Zenios2015, Reference Goh, Pfeffer and Zenios2016) attributed to occupational stressors, and violation of right to work in just and favorable conditions and right to have adequate living for the health and well-being of workers and their family (United Nations, 1948) – are often not adequately estimated. Thus, keep in mind that estimated economic costs of organizational stressors are likely grossly underestimated.

Productivity losses, manifested as sickness absenteeism, turnover, and presenteeism, etc., are one of the key outcomes while investigating adverse impacts of occupational stressors at the macro level. For example, based on 1513 participants of a Swiss workforce two-wave survey, Brunner et al. (Reference Brunner, Igic, Keller and Wieser2019) found associations between occupational stressors and productivity losses due to sickness absenteeism and presenteeism. Their results revealed that occupational stressors accounts for 23.8% of the total health-related production losses. The losses correspond to 195 Swiss francs per worker each month, or 3.2% of the average monthly earnings in Switzerland.

Kline and Lewis (Reference Kline and Lewis2019) focused on one type of occupational stressor, bullying and harassment, and estimated its related cost from staff in the National Health Service (NHS), which is responsible for health care delivery in England. The total estimated cost attributed to bullying and harassment is £2.281 billion annually, of which sickness absenteeism costs £483.6 million, turnover costs £231.9 million, productivity loss costs £575.7 million, presenteeism costs £604.4 million, and compensation and litigation cost £83.5 million.

To inform the Australian Human Rights Commission’s national inquiry into sexual harassment in Australian workplaces, Deloitte Access Economics (2019) estimated cost of workplace sexual harassment in 2018 to be $3.5 billion, of which productivity losses account for $2.6 billion. Among those $2.6 billion, 70% of the cost is lost by organizations, 23% of the cost is tax revenue loss, and 7% of the cost accounts for income loss of employees.

Discrimination at work is a major occupational stressor (Mays et al., Reference Mays, Coleman and Jackson1996) experienced by disadvantaged social groups. In 2019, a total of 72,675 charges of workplace discrimination were received by the Equal Employment Opportunity Commission in US, leading to $486 million in monetary relief for victims of discrimination (US Equal Employment Opportunity Commission, 2020a). The economic cost of workplace discrimination to the UK economy is estimated at £127 billion every year, of which £123 billion is due to gender discrimination; £2.6 billion to discrimination against ethnic minorities; and £2 billion to discrimination against sexual orientation (Centre for Economic and Business Research, 2018).

Depression has been shown to be related to productivity loss, sickness absenteeism, and presenteeism (Bonde, Reference Bonde2008). To estimate the economic cost of work-related depression, Cocker et al. (Reference Cocker, Sanderson and LaMontagne2017) first estimated 13.2% (men) and 17.2% (women) of depression cases were attributed to job strain (i.e., high job demands and low job control) based on Australian population data. After that, they conducted a Markov modeling to simulate the economic benefits of removing job strain-related depression. Estimated average benefit to eliminate depression attributed to job strain in Australia is $730 million in one year, with a confidence level ranging from $425 million to $882 million; and $11.8 billion over a lifetime, with a confidence level ranging from $6.9 billion to $14.3 billion.

Now we switch to a systematic review (Hassard et al., Reference Hassard, Teoh, Visockaite, Dewe and Cox2018a) that estimates annual cost of work stress, adjusted in 2014-valued US dollars, based on 15 cost-of-illness studies conducted in eight countries. Because different outcome criteria and approaches were used by each of 15 studies, annual estimated cost of occupational stress substantially varied within as well as across countries: Australia ($0.22 to $3.98 billion; cost per worker: $17.79 to $320.14), Canada ($2.59 to $9.59 billion; $131.31 and $486.33), Denmark ($0.379 to $2.27 billion; $130.07 to $777.26), France ($1.83 to $4.36 billion; $60.95 to $145.03), Sweden ($703.12 million; $136.71), Switzerland ($3.33 billion; $701.14), United Kingdom ($2.18 billion to 23.63 billion; $66.35 to $716.58), and the United States ($187 billion; $1211.84). Overall, the annual estimated cost of work-related stressful experience at a national level ranged from $221.13 million to $187 billion, of which productivity loss accounts for 70–90% and health care cost accounts for 10–30%.

There are a few comparative evaluation studies that focus on a specific type of occupational stressor. For example, Hassard et al. (Reference Hassard, Teoh, Visockaite, Dewe and Cox2018b) estimated annual costs attributed to psychosocial workplace aggression at a national level. Workplace aggression is defined as interpersonal and harmful behaviors such as bullying, mobbing, harassment, interpersonal conflict, discrimination, incivility, abusive supervision, or mistreatment. Based on 12 studies conducted in five nations – Australia, Italy, Spain, the United Kingdom (UK), and the United States (US) – the authors estimated the annual costs attributed to psychosocial workplace aggression ranged between $114.64 million and $35.9 billion across the five countries, adjusted to 2014 US dollar rates. Similarly, Hassard et al. (Reference Hassard, Teoh and Cox2019) estimated costs attributed to work-related violence (behaviors that are physical in nature such as homicide or assault) in the US (nine studies) and UK (one study). The annual estimated costs attributed to work violence, primarily in the US, varies from $2.36 million to $55.86 billion, adjusted to 2017 US dollar rates.

Impacts in Terms of Health Outcomes

In this section, we focus on epidemiological evidence in relation to health consequences of occupational stressors. Goh et al. (Reference Goh, Pfefer and Zenios2015) meta-analyzed 228 studies reporting relationships of occupational stressors (i.e., low job control, job insecurity, unemployment, job demand, organizational injustice, low social support at work, work–family conflict, long work hours, and shift work) with four health outcomes (morbidity or physician-diagnosed medical conditions, mortality, perceived poor physical health, and perceived poor mental health). They found that unemployment and low job control were significantly related to all of the health outcomes. In addition, all of the occupational stressors, except for work–family conflict, were significantly associated with morbidity.

To meaningfully interpret the above relationships, Goh et al. (Reference Goh, Pfefer and Zenios2015) used the odds ratio of secondhand smoking for the above four health outcomes as a frame of reference (i.e., 1.3 for morbidity, 1.15 for mortality, 1.47 for perceived poor physical health, 1.49 for perceived poor mental health, respectively). Tobacco is one of the health indicators in Health People 2020 (US Department of Health and Human Services, 2020). Regarding morbidity, it was found that organizational injustice, job demands, shift work, and unemployment exhibited stronger effects than secondhand smoking. Similarly, low job control, unemployment, long work hours, and work–family conflict exhibited stronger effects on mortality than secondhand smoking; work–family conflict, unemployment, and job insecurity showed stronger effects on poor physical health than secondhand smoking; and work–family conflict, unemployment, job demands, and organizational injustice showed stronger effects on poor mental health than secondhand smoking.

Goh et al. (Reference Goh, Pfeffer and Zenios2016) further estimated mortality and health care cost attributed to an absence of health insurance and the occupational stressors described in the previous section. Based on their conservative estimates, there are more than 120,000 deaths and more than $110 billion spent in health care per year in the United States. Lately, Goh et al. (Reference Goh, Pfeffer and Zenios2019) estimated 17,000 preventable deaths each year attributed to occupational stressors in the United States. They also estimated $44 billion in preventable health care costs each year attributed to occupational stressors in the United States. This cost is much higher than the annual direct/indirect health care cost, $10 billion, of secondhand smoke.

Suicide Attributed to Unemployment

Chen and Stallones (Reference Chen, Stallones, Kelloway and Cooper2011) have argued that self-inflicted fatality, or suicide, is a hidden occupational fatality for three reasons. According to the US Bureau of Labor Statistics, BLS (2019b), suicide is included in the BLS Census of Fatal Occupational Injuries if (1) the suicide occurred on the work premises, or (2) the suicide occurred off the work premises but can be definitively linked back to work. However, about 75% of suicides occur in a house or apartment, followed by natural areas, streets or highways, or in motor vehicles (Colorado Department of Public Health and Environment, 2007). Of 5,333 occupational fatalities in 2019, furthermore, there are only 12 fatalities coded as suicide. In addition, suicide has been a major public health concern and was ranked 2, 4, and 8 among the top 10 leading causes of death in 2018 for working age groups 25–34, 35–54, and 55–64, respectively, in the United States (US National Institute of Mental Health, 2020). Finally, suicide rates tend to be disproportionately higher in some industries than those in the general population (see Chen & Stallones, Reference Chen, Stallones, Kelloway and Cooper2011).

Unemployment has been shown to be associated with suicide. Loss of employment is a major stressful work and life event which has shown detrimental impacts on the physical and psychological well-being of workers and their families (McKee-Ryan et al., Reference McKee-Ryan, Song, Wanberg and Kinicki2005; Paul & Moser, Reference Paul and Moser2009). Nordt et al. (Reference Nordt, Warnke, Seifritz and Kawohl2015) tracked trends of suicide and unemployment between 2000 and 2011 in 63 countries across the Americas, Europe, Africa, and Asia and found an increase in suicide rate approximately six months after a rise in unemployment. They further estimated around 45,000 out of 233,000 suicides per year are likely related to loss of employment. Similarly, a study of suicide rate and business cycle from 1928 to 2007 in the US reveals that suicide rates generally rise during recessions (Luo et al., Reference Luo, Florence, Quispe-Agnoli, Ouyang and Crosby2011). Similar results were reported in Japan and European countries (Breuer, Reference Breuer2015).

Workers’ Compensation Legislation

Workers’ compensation legislation usually does not define occupational stress or specific occupational stressors because occupational stress is not a medical condition, even though it may be a cause or trigger of a medical condition. Occupational health and safety legislation in the United Kingdom, Canada, Denmark, and Australia places general duties upon all employers to ensure the health, safety, and welfare at work of all employees as long as it is reasonably practical. In Denmark, under the law on occupational safety and health (The Danish Working Environment Act), employers bear responsibility for workers not being exposed to any form of strain/harassment/discrimination at work. Typically, a workers’ compensation claim requires workers to establish that they have suffered from an “injury,” including illness, disease, and/or body injuries.

Claims in which a worker has suffered a recognized injury or disease caused by a work-related stressor are generally referred to by practicing lawyers as “mental stress” or “work-related stress” claims. For instance, in a decision of the High Court of Australia in Koehler v Cerebos (Australia) Ltd. (2005), the employee’s psychiatric injury was held not to have been foreseeable by the employer. Whilst Mrs. Koehler had frequently complained about her workload, it was found that her complaints were all “…directed to whether the work could be done; [and] none [of the complaints] suggested that the difficulties she was experiencing were affecting her health.” Given that the employee’s psychiatric injury was not foreseeable by the employer, there could be no duty of care (and, thus, no damages awarded). The High Court found that Mrs. Koehler had “voluntarily” accepted the position as a part-time merchandiser and by doing so had agreed to perform the duties that she later complained of as excessive. This decision therefore reflects a judicial view that employees and employers are free to contract as they agree about the work one party will do for the other party.

Age Discrimination Legislation

Over 50 countries have similar legislation for age discrimination (Lewis Silkin LLP, 2020). A sample of age discrimination legislation is described in Table 2.1. The US Age Discrimination in Employment Act 1967 (amended in 1974 and 1978) aims to promote employment decisions based on employees’ or job applicants’ qualifications rather than age, and age is specifically defined as 40 or older. Yet, the Act does not protect workers under the age of 40, although some states have laws that protect younger workers from age discrimination.

In contrast, other age discrimination legislation, such as the UK Equality Act 2010 or the Slovakia Labour Code & Antidiscrimination Act, emphasizes fair treatments regardless of age. In Hendrick v. Hercules, Inc. (1981), a reorganizational planning committee in Hercules determined that Hendrick did not possess the necessary qualifications for the new position, and he was terminated in April 1978 because his position as Plant Manager had been eliminated. Hedrick testified that Director of Operations in Hercules told him that “Hercules was going to get rid of the good ’ole Joes and get some younger folks in.” The court awarded Hedrick $52,887 for wage losses and $26,443.50 in liquidated damages.

A significant ruling highlights that EU member states (i.e., Germany in this case) and EU legislation were challenged when an organization failed to comply with the general principle of equal treatment. In Mangold v. Helm (2005), Mangold was a 56-year-old German man employed on a fixed term contract in a permanent full-time job. According to Sec. 14 para. 3 of the German Part-Time Work and Fixed-Term Employment Contracts Act, no objective reason is necessary to enter into an employment contract of limited duration if the employee has reached the age of 52. Mangold claimed that the lack of protection, over age 52, was age discrimination. The European Court of Justice sided with Mangold and held this provision as violating the Employment Equality Framework (Directive 2000/78/EC). Specifically, the European Court of Justice ruled that

Racial Discrimination Legislation

Over 169 countries have developed legislation for elimination of discrimination (racial/gender) to meet the standards set in Conventions of International Labour Organization (Fredman, Reference Fredman2012).

Title VII in the US Civil Right Act 1964 promotes equal employment opportunity and prohibits employment discrimination based on race, color, religion, sex, or national origin. In very narrowly defined situations, an employer is permitted to discriminate based on a reasonably needed trait that is a bona fide occupational qualification such certain knowledge, skills, and abilities needed to perform a job. Prior to 1972, Title VII lacked any strong enforcement provisions. After the US Congress passed the Equal Employment Opportunity Act 1972, the Act amended Title VII and gave Equal Employment Opportunity Commission (EEOC) authority to initiate its own enforcement litigation.

A landmark case, Griggs v. Duke Power Co (1971), argued before the US Supreme Court, has been considered the first employment discrimination case based on race. The Supreme Court ruled, under Title VII of the Civil Rights Act 1964, that Duke Power Co’s employment requirements (high school diploma, Bennett Mechanical Comprehension Test, and Wonderlic Cognitive Ability Test) not only showed disparate treatment of black employees but also that these did not pertain to applicants’ ability to perform the job. Thus, employment requirements unintentionally discriminate against black employees.

In contrast to other countries’ legislation, described in Table 2.2, the Australian Parliament passed an Act that specifically addressed racial discrimination, the Racial Discrimination Act 1975. The Act prohibits treating someone less fairly than someone else in a similar situation because of his/her race, color, descent, or national or ethnic origin. The Act covers areas including employment, land, housing or accommodation, provision of goods and services, access to places and facilities for use by the public, advertising, or joining a trade union. In Bligh and Ors v. State of Queensland (1996), the Australian Human Rights Commission found past Aboriginal employees were underpaid by the Queensland Government, and A$7,000 was awarded to each plaintiff.

Although a definition of racial discrimination is absent from German General Equal Treatment Act, Berlin passed the first German state antidiscrimination law on June 4, 2020. This is the first antidiscrimination law to include discrimination factors of race, skin color, and German language skill.

Racial discrimination in employment is prohibited in Article 12 of the Constitution of the Republic of Singapore:

Similarly, racial discrimination is prohibited in Section 7 of the Constitution of South Africa.

Sex Discrimination Legislation

Title VII in the US Civil Rights Act promotes equal employment opportunity and prohibits employment discrimination based on race, color, religion, sex, or national origin. In United Automobile Workers v. Johnson Controls, Inc. (1991), the US Supreme Court ruled that Johnson Controls’ policy to exclude women with childbearing capacity from lead-exposed jobs explicitly discriminated against women based on their sex. In contrast to other countries’ legislation, described in Table 2.3, the Australian Parliament passed an Act that specifically addressed sex discrimination, the Sex Discrimination Act 1984.

In section 12, the Constitution of the Republic of Singapore defined that discrimination against women (only for discriminatory behaviors against women) shall

In France, any discrimination is considered as a criminal issue. Under the Criminal Code, where a legal entity has been found to have discriminated against an employee, it may be liable to a fine of €225,000 and possibly some additional sanctions such as the dissolution of the legal entity, the publication of the judgment, the ban of the exercise of a determined professional activity, or the closure of the company’s entities.

Sexual Harassment Legislation

In 2020, there are 140 countries with similar legislation for preventing and redressing sexual harassment (Arekapudi & Recavarren, Reference Arekapudi and Recavarren2020). According to Quick and McFadyen (Reference Quick and McFadyen2017), there is no universal legal definition of sexual harassment, although most definitions contain description of unwanted or unwelcome conduct that intimidates and humiliates victims, as shown in Table 2.4. US EEOC (2020b) defined sexual harassment as “unwelcome sexual advances, requests for sexual favors, and other verbal or physical harassment of a sexual nature.” Sexual harassment can include offensive remarks about a person’s sex; even harassment does not have to be a sexual nature. Furthermore, both victim and the harasser can be either a woman or a man, and the victim and harasser can also be the same sex. The US Supreme Court and federal courts support coverage of LGBT-related discrimination under Title VII of the Civil Rights Act (US EEOC, 2020c).

In Quick and McFadyen’s (Reference Quick and McFadyen2017) review, concerns about sexual harassment worldwide have also been addressed by the International Labor Organization, the International Confederation of Free Trade Unions, the European Union, and the United Nations Committee on the Elimination of Discrimination Against Women.

Similar to sexual discrimination legislation in France, sexual harassment is considered a criminal issue, and a legal entity may be liable to a fine of €225,000 and some additional sanctions, as described earlier. In contrast to other countries, the Australian court considers sexual harassment (covered under the Sex Discrimination Act 1984) in a much broader context, including social functions, work conferences, work-related trips, etc. In South Pacific Resort Hotels Pty Ltd v. Trainor (2005), the Full Court of the Federal Court of Brisbane confirmed that the conduct of a hotel worker constituted sexual harassment, despite the conduct occurring after hours. The case involved island hotel workers, residing in accommodation provided by the employer, at which no visitors were permitted to enter. A coworker entered a colleague’s bedroom uninvited on two occasions, while both employees were off duty. The Full Court held that the sexual harassment took place because of their common employment and subsequent accommodation and that it could not be said that the common employment was “unrelated or merely incidental to the sexual harassment of one by the other.” Ms. Trainor was awarded $5,000 in general damages and $7,500 in past and future economic loss.

Workplace Bullying Legislation

Finally, workplace bullying is a persistent pattern of mistreatment from others in the workplace that causes either physical or emotional harm. Currently, Cobb (Reference Cobb2017) has reviewed workplace bullying legislation from 50 countries across Europe, the Asia Pacific region, the Americas, and the Middle East and Africa. Comprehensive workplace bullying legislation in the United States has not been passed by the federal government, although some state legislatures have considered bills related to workplace bullying, such as Nevada, New York, Washington, Illinois, etc.

In the United Kingdom, although bullying is not specifically mentioned in workplace legislation, there are means such as the Protection From Harassment Act 1997 to obtain legal redress for bullying. Workplace bullying, or an employer tolerating bullying, is typically viewed as breaching contractual terms in the United Kingdom. Parties to the employment contract have a (legal) duty of trust and confidence to each other. Such a breach creates circumstances entitling an employee to terminate his or her contract of employment without notice, and an Employment Tribunal considers the termination as an unfair dismissal. An employee bullied in response to asserting a statutory right can be compensated for the damage under Part V of the Employment Rights Act 1996, and if dismissed, Part X of the same Act provides that the dismissal is automatically unfair. In Majrowski v. Guy’s and St Thomas’s NHS Trust (2006), William Majrowski (M) was formerly employed by St Thomas’s NHS Trust (D). M claimed that he had been bullied by his manager (L), which M claimed amounted to harassment under the Protection From Harassment Act 1997. It was held that D was vicariously liable for the offense committed by L. D appealed and the appeal was dismissed for the reason of breaching contract.

In Canada, under the Ontario Occupational Health and Safety Act 1979, all employers should take precaution to protect their workers in reasonable circumstances. In 2007, Ontario passed a new bill (Bill-29), which made an amendment to the Ontario Occupational Health and Safety Act. Bill-29 proposed that employees should be protected from harassment and violence in the workplace. In addition, protection for employees will include psychological abuse and bullying behaviors. Moreover, the Province of Quebec passed legislation addressing workplace bullying in 2004. Summaries of workplace bullying legislation are shown in Table 2.5.

Look Forward

We conclude our review by emphasizing preventive strategies applicable to workers, organizations, and nations. First, there have been more stress management interventions designed to reduce stress symptoms. These interventions, such as mindfulness training, cognitive-behavioral intervention, relaxation, personal resources training, or resilience-strengthening intervention, have generally shown expected outcomes (Bakker & van Wingerden, in press; Bartlett et al., Reference Bartlett, Martin, Neil, Memish, Otahal, Kilpatrick and Sanderson2019; Beehr, Reference Beehr2019; Falon et al., in press). In contrast, there are relatively fewer stress management interventions designed to modify occupational stressors (e.g., reducing abusive supervision; Gonzalez-Morales et al., Reference Gonzalez-Morales, Kernan, Becker and Eisenberger2018), increase support (e.g., family-supportive supervisor behaviors, Odle-Dusseau et al., Reference Odle-Dusseau, Hammer, Crain and Bodner2016; supervisor support behaviors, Brady et al., in press), or modify work contexts (e.g., increasing employee control through task restructuring, Bambra et al., Reference Bambra, Gibson, Sowden, Wright, Whitehead and Petticrew2009). Whatever the foci are, these evidence-based interventions tend to be highly intensive (i.e., time demands, high cost, required expertise). As a result, implementing a stress management intervention with high intensity, which is typically found in academic journals, is unlikely to align well with daily operations in organizations. It would also be a challenge for workers to sustain practicing individual-focused stress management interventions.

We argued that future intervention research shall not only ask “how does it work” and “does it work” but also answer “does it work with minimal intensity?” (Glasgow & Emmons, Reference Glasgow and Emmons2007) or “does or can it work under real-world conditions?” (Glasgow et al., Reference Glasgow, Lichtenstein and Marcus2003). Understandably, the latter questions raise the bar for future intervention research. Yet, these questions are indispensable to substantially advance occupational stress research and practices. Answers to these questions are also needed considering that most businesses establishments are small (defined as fewer than 500 employees). For instance, 99.9% of US businesses are small businesses with an average of ten employees (US Small Business Administration, 2019). It is well known that small establishments tend not to manage occupational safety and health (OSH) well, and often show low level of compliance in relation to OSH regulations (EU-OSHA, 2018). It should be emphasized that even large establishments such as National Health Service England (Health Education England, 2019) acknowledge the difficulty of implementing interventions because OSH interventions tend to be costly and time consuming and require considerable commitments from management. This sentiment was also echoed by progressive executives (Pfeffer et al., Reference Pfeffer, Vilendrer, Joseph, Kim and Singer2020), described below.

Research has consistently shown that management commitment plays a vital role to manage safety and health concerns (EU-OSHA, 2018; Schall & Chen, in press). Pfeffer et al. (Reference Pfeffer, Vilendrer, Joseph, Kim and Singer2020) interviewed executives from 20 progressive companies. Overall, these executives viewed themselves being responsible for well-being of employees and their families, and most of them viewed a healthy workplace as a means to gain profits. Similar to conventional view, they viewed occupational stressors as inevitable, a strong focus on productivity in business as imperative, and modifying sources of stress as unfeasible. These views highlight stiff challenges in future research and practice in managing or preventing stress. Kotter (Reference Kotter1995), in a seminal essay, pointed out that most organizational change initiatives fail because management fails to develop a sense of urgency, or to communicate the urgency broadly with key constituencies. Without the sense of urgency shared by the majority of managers/supervisors and employees pertaining to prevalence and severity of occupational stressors, management teams are less likely to commit and prioritize resources to minimize psychosocial risks, are less likely to institutionalize preventive strategies as daily, routine operations, and are less inclined to create shared values and norms with regard to healthy workplaces. To advance preventive stress management in future research and practice, it is essential to connect the urgency with impacts of occupational stressors. We acknowledge that it is a challenging goal in stress management. Next, we shall also consider some “low hanging fruit.”

Burke (Reference Burke, Burke and Richardsen2019) identified several organizational characteristics associated with unhealthy workplaces. These characteristics include toxic leadership, bullying, organizational politics that promote actors’ interest, uncivil behaviors at work, and discrimination. With relatively inexpensive strategies, organizations can build psychologically healthy workplace by creating climates for fairness and civility with follow-through, actionable polices guided by legislation, and utilizing evidence-based HR tools to select, monitor, appraise, and develop managers, supervisors, and employees. Scientific advancements over the past century in recruitment, selection, leadership, performance management and appraisal, or development (see Kozlowski et al., Reference Kozlowski, Chen and Salas2017) offer invaluable insights that occupational stress research and practices can tag along.

Occupational stress research has continued identifying timely occupational stressors such as telepressure (Barber & Santuzzi, Reference Barber and Santuzzi2015) and illegitimate tasks (Sonnentag & Lischetzke, Reference Sonnentag and Lischetzke2018). Of occupational stressors studied over the past century (Beehr & Newman, Reference Beehr and Newman1978; Bliese et al., Reference Bliese, Edwards and Sonnentag2017), some could be relatively easier to modify and use interventions such as those identified by the US NIOSH panel (Wiegand et al., Reference Wiegand, Chen, Hurrell, Jex, Nakata, Nigam, Robertson and Tetrick2012). Without utilizing many resources, organizations may gather surveillance data pertaining to these stressors. Accordingly, organizations can efficiently and effectively modify and intervene with these stressors.

Finally, governments should view occupational stress as a key public health concern. Goh et al. (Reference Goh, Pfefer and Zenios2015) has shown that negative impacts of occupational stressors are worse than those of tobacco uses, a health indicator in Health People 2020 (US Department of Health and Human Services, 2020). Health People 2020 identified 26 health indicators – injury and violence, access to health services (e.g., health insurances), clinical preventive services to treat cardiovascular diseases, mental health (e.g., suicide and depression), substance abuse (e.g., drinking) – which have been linked to occupational stressors. Yet, key occupational stressors, leading indicators that are responsible for mental health and physical health, are not included in Health People 2020, nor in US public health surveillance systems. Similarly, Leka et al. (Reference Leka, Jain, Iavicoli and Di Tecco2015) concluded that risk assessments were often not covered in most EU regulatory policies. Should key occupational stressors be monitored nationally and globally, surveillance data regarding these stressors will be extremely helpful to guide regulatory and organizational policies and raise awareness of urgency among the public. After all, it is inconceivable to expect nations, organizations, and individuals to diligently manage what has not been measured!

Explanatory Contribution

The first criterion concerns the model’s disciplinary background and its theoretical fundament. Does the concept define pathways linking the work environment with health and well-being, thus offering explanations of statistical associations? This also implies that features of the work environment are consistently measured within the model.

Assessment

These models are most often investigated using psychometrically validated scales of questionnaires or interviews. It is therefore important to include concepts based on scales that meet internationally established quality criteria.

Strength of Evidence

How extensive have these models been empirically tested? Were findings derived from appropriate study designs? How consistent is the body of evidence?

Generalization

This aspect concerns (a) the range of health and behavioral outcomes under study; (b) the model’s suitability for different types of occupations and employment sectors; and (c) its extension to the analysis of productive activities beyond paid work.

Policy Impact

As it is the overarching aim of related scientific knowledge to promote healthy work, it seems appropriate to explore the model’s policy impact.

Within a rich and extensive scientific literature dealing with psychosocial aspects at work and well-being, the following theoretical models made significant contributions to the advancement of knowledge and inspired new directions of analysis but were judged to provide insufficient evidence on one of the five criteria mentioned: ‘burnout’ (Maslach & Leiter, 2017), ‘sense of coherence’ (Antonovsky, Reference Antonovsky1987), ‘stressor-detachment’ (Sonnentag & Fritz, Reference Sonnentag and Fritz2015), ‘stress as offense of self’ (Semmer, Reference Semmer and Theorell2020), ‘person-environment-fit’ (Edwards et al., Reference Edwards, Caplan, Van Harrison and Cooper1998), ‘conservation of resources’ (Hobfoll, Reference Hobfoll1989), and ‘vitamin model’ (Warr, Reference Warr, Cooper and Quick2017). For instance, the strength of prospective evidence on adverse effects on health was considered to be restricted (‘stress as offense of self’; ‘stressor-detachment’; ‘person–environment fit’), or the elaboration of work-specific explanatory links was not obvious (‘burnout’; ‘vitamin model’; ‘sense of coherence’), or the operational measurement of the concept was not specific enough (for a critical appraisal of ‘conservation of resources’, see e.g. Halbersleben et al., Reference Halbesleben, Neveu, Paustian-Underdahl and Westman2014). As a result of this selection process, the following four theoretical models were judged to meet all five criteria to a sufficient extent: ‘demand-control’ (DC) or ‘job strain’ (Karasek, Reference Karasek1979; Karasek & Theorell, Reference Karasek and Theorell1990), ‘effort-reward imbalance’ (ERI) (Siegrist, Reference Siegrist1996; Siegrist & Wahrendorf, Reference Siegrist and Wahrendorf2016), ‘organizational justice’ (OJ) (Greenberg & Cohen, Reference Greenberg and Cohen1982; Elovainio et al., Reference Elovainio, Kivimäki and Vahtera2002), and ‘job demands-resources’ (JD-R) (Demerouti et al., Reference Demerouti, Bakker, Nachreiner and Schaufeli2001; Bakker & Demerouti, Reference Bakker and Demerouti2017). In the next section, these models are briefly described, followed by a comparative discussion of their similarities and differences.

Demand-Control

The demand-control (DC) model is one of the most widely applied and most well-known models of psychosocial stress at work. It was first proposed by American sociologist Robert Karasek (Reference Karasek1979), with elaborated links to psychobiological stress theory, as applied to cardiovascular disease (Karasek & Theorell, Reference Karasek and Theorell1990), and with an extension to include social support at work as a third dimension (Johnson & Hall, Reference Johnson and Hall1988). Its scientific roots are grounded in sociological alienation theory, where low control among workers was studied in relation to their well-being and personal development (Kornhauser, Reference Kornhauser1965), and in occupational health research on high workload (Frankenhaeuser & Gardell, Reference Frankenhaeuser and Gardell1976). Karasek’s original idea of combining the two dimensions of psychological demands and decision latitude (or control) to explain elevated health risks was supported by survey data. Soon, it became apparent that the control dimension was composed of two different aspects, the worker’s influence on performing the work (decision authority) and his or her ability to use own resources (skill level). Jobs defined by high psychological demands and low decision latitude or skill level were proposed to generate stress and to reduce health and well-being, whereas demanding jobs with a high level of control and autonomy were thought to buffer stress and to strengthen the experience of mastery and active learning (Karasek & Theorell, Reference Karasek and Theorell1990). A close link of these sociological components with psychobiological stress theory is considered a particular strength of this model. Specifically, conditions of high demand and low control were shown to simultaneously activate two stress axes within the organism, the sympatho-adrenal medullary and the hypothalamic-pituitary-adrenocortical axis, and this synergistic activation contributes to the pathology of stress-related disorders, such as ischemic heart disease (Henry & Stephens, Reference Henry and Stephens1977). Recently, Karasek has proposed a substantial multilevel extension of the original DC model, termed ‘associationalist demand-control theory’ (Karasek, Reference Karasek and Theorell2020). As empirical data on this new version are currently not available, it will not be included in this comparison.

The DC model is measured by different approaches, self-reported questionnaires, observational ratings, or features of administrative job classifications (‘job exposure matrices’). Yet, the job strain questionnaire (JCQ) has been established as the standard method. It is composed of five items measuring psychological demands, six items measuring decision authority, and four items measuring skill discretion. Items are assessed on a 5-point Likert scale according to experienced intensity (Karasek et al., Reference Karasek, Brisson, Kawakami, Houtman, Bongers and Amick1998). Psychometric properties of the scales and factorial validity of the model were tested in a number of methodological studies (e.g. Chungkham et al., Reference Chungkham, Ingre, Karasek, Westerlund and Theorell2013). An additional scale measuring social support at work was constructed in a similar way, but many epidemiological studies did not include this scale in their model tests. The main approach of analyzing the model is based on the ‘split half’ method, where scores of job demands above the median are scored ‘high’ and scores of job control below the median are scored ‘low’. However, additional analyses of the combined effect of the two dimensions were performed, applying additive or multiplicative interaction. It should also be mentioned that many investigations studied the effects of single scales rather than their combination. Overall, there is considerable variation in exploring the model’s propositions.

Strong evidence exists on the contributions of the DC model towards explaining elevated risks of stress-related disorders and ill health. This has been documented by several systematic reviews and meta-analyses of prospective cohort studies. They include the effects of job strain on ischemic heart disease (Theorell et al., Reference Theorell, Jood, Järvholm, Vingård, Perk, Östergren and Hall2016), high blood pressure (Gilbert-Ouimet et al., Reference Gilbert-Ouimet, Trudel, Brisson, Milot and Vézina2014), and depression (Madsen et al., Reference Madsen, Nyberg, Magnusson Hanson and Ferrie2017). A variety of additional reports demonstrate associations with metabolic disorders, musculoskeletal disorders, poor cognitive functioning, sickness absence, disability, and life years lost. It is beyond the scope of this chapter to give a representative account of this knowledge, but some recent reports from the original authors illustrate its significance (Theorell, Reference Theorell and Theorell2020; Karasek, Reference Karasek and Theorell2020). On balance, it seems that the single component ‘job control’ exerts relatively strongest effects on disease risks, and that the claim of a synergistic effect of high demand and low control is not consistently confirmed. On the other hand, some studies show that, according to the model, ‘active jobs’ defined by high demand and high control exert some beneficial health effect. Whereas abundant research on the DC model was conducted by epidemiologists, fewer findings are available from experimental and quasi-experimental studies, such as naturalistic studies of physiological activity in everyday working life. They provide some support with regard to increased blood pressure, reduced heart rate variability, elevated secretion of stress hormones, and impaired immune function (e.g. Steptoe & Kivimäki, Reference Steptoe and Kivimäki2012; Jarczok et al., Reference Jarczok, Jarczok, Mauss, Koenig, Li, Herr and Thayer2013; Theorell, Reference Theorell and Theorell2020).

In summary, the DC model has been investigated more thoroughly than any other similar concept, and its components, in particular job control, demonstrate robust effects on working people’s health and well-being. The model was applied to a wide range of health outcomes, including health-related behaviors (Kivimäki et al., Reference Kivimäki, Pentti, Ferrie, Batty, Nyberg, Jokela, Virtanen, Alfredsson, Dragano, Fransson, Goldberg, Knutsson, Koskenvuo, Koskinen, Kouvonen, Luukkonen, Oksanen, Rugulies and Deanfield2018), and the notion of low control was extended to explore health effects related to work at home (Chandola et al., Reference Chandola, Kuper, Singh-Manoux, Bartley and Marmot2004) and to school work (Gädin & Hammarström, Reference Gadin and Hammarström2000). To some extent, this latter information indicates that the model’s core notions can be generalized beyond the context of paid work. Finally, concerning policy impact, the DC model was investigated in a number of intervention studies, with a focus on job task redesign and other organizational improvements at the level of enterprises and businesses. Several such interventions resulted in positive effects on workers’ well-being (Brisson et al., Reference Brisson, Gilbert-Ouimet, Duchaine, Trudel, Vézina, Siegrist and Wahrendorf2016).

Effort-Reward Imbalance

The effort-reward imbalance (ERI) model represents a complementary approach towards analyzing a health-adverse psychosocial environment. It focuses on the employment contract as the core element of employment relations, emphasizing the role of social reciprocity in costly transactions for health and well-being (Siegrist, Reference Siegrist1996). Rooted in sociological and social psychological theories of exchange and equity (Gouldner, Reference Gouldner1960; Adams, Reference Adams and Leonard1965), and referring to sociological role theory (Merton, Reference Merton1968), the model assumes that the principle of equivalent reward in return for effort spent at work is often violated, such that workers are exposed to ‘high cost–low gain’ conditions (imbalance) in a core social role. Importantly, three basic types of reward are transmitted, salary or wage (financial reward), career promotion or job security (status-related reward), and esteem or recognition (socioemotional reward). Imbalance is likely to occur if workers have no alternative choice in the labor market and if they are exposed to heavy competition. In addition to these extrinsic factors, an intrinsic motivational factor, the working person’s inability to withdraw from work obligations and excessive striving (‘overcommitment’), contributes to this imbalance. This risky pattern of coping may be due to an underlying need for approval and esteem at work, but it can also be reinforced by group pressure at work. The model’s central proposition states that each component, high effort, low reward, and high overcommitment, is associated with an elevated health risk, and that the combined effect of high effort and low reward contributes to adverse health in addition to the main effects. The ERI model has a strong transdisciplinary link to neuroscience-based stress theory as sustained experience of reward deficiency at work activates distinct brain reward circuits, including nucleus accumbens, anterior cingulate cortex, and insula (Schultz, Reference Schultz2006). This threat or loss of significant social reward is associated with sustained activation of the stress axes mentioned above, and it may additionally stimulate addictive behavior. In the long run, allostatic load is triggered within several regulatory systems of the body, thus precipitating the development of stress-related disorders (McEwen, Reference McEwen1998).

The ERI model is measured by the three scales ‘effort’, ‘reward’, and ‘overcommitment’, containing 4-point Likert-scaled items. The original version has been supplemented by a short version with a total of 16 items. Extensive psychometric tests of both versions were performed in several languages. The factorial validity was tested by confirmatory factor analysis, and the model fit was recurrently improved when a second order model representing the theoretical structure underlying the scales was analyzed (e.g. Leineweber et al., Reference Leineweber, Wege, Westerlund, Theorell, Wahrendorf and Siegrist2010). Additionally, longitudinal measurement invariance of the model’s scales was demonstrated (Montano et al., Reference Montano, Li, Siegrist, Siegrist and Wahrendorf2016). Hypotheses were tested using regression analyses with single scales, and additionally with their combination. To this end, a logarithmic ratio of the scales ‘effort’ and ‘reward’ was constructed that quantifies the imbalance at the individual level. Alternative approaches include additive and multiplicative interaction analysis of scales.

The evidence of prospective associations of ERI with elevated risks of stress-related disorders concurs with the one demonstrated for the DC model. Although based on a lower number of studies, the effect sizes of relative risks or odds ratios are comparable in the case of ischemic heart disease (Siegrist & Li, Reference Siegrist, Li and Theorell2020), depression (Rugulies et al., Reference Rugulies, Aust and Madsen2017), and high blood pressure (Gilbert-Ouimet et al., Reference Gilbert-Ouimet, Trudel, Brisson, Milot and Vézina2014). Again, a broad spectrum of additional indicators of health and well-being were examined with regard to effort and reward at work (Siegrist & Wahrendorf, Reference Siegrist and Wahrendorf2016). Further support of the model’s hypotheses comes from a series of experimental and quasi-experimental studies using cardiovascular and autonomic nervous system markers, measures of stress hormone release, and measures of inflammatory and immune function (Eddy et al., Reference Eddy, Wertheim, Hale and Wright2018; Siegrist & Li, Reference Siegrist, Li and Theorell2020). When summarizing the results, the model’s propositions are mainly supported, yet to a different extent. Concerning the effects of single scales, ‘low reward’ exerts relatively strongest effects on health, followed by ‘effort’ and ‘overcommitment’. The summary measure combing the two extrinsic components is associated with health in a majority of studies, but adjusting for the effects of single scales has often not been performed or reported in publications, thus leaving room for further analyses. Few studies so far have tested the moderating role of ‘overcommitment’ in this latter association with health.

Given a fundamental role of the norm of reciprocity in social exchange, the ERI model has been applied to several types of costly social transactions other than paid work. Associations with well-being were observed in household and family work, voluntary work, caregiving, and informal help, and among students in the context of educational work (Siegrist & Wahrendorf, Reference Siegrist and Wahrendorf2016). It is therefore assumed that the model’s hypotheses are valid for a larger range of productive social activities. Concerning policy impact, the ERI model was investigated in several intervention studies, in combination either with or without the DC model, with effects similar to those reported (Brisson et al., Reference Brisson, Gilbert-Ouimet, Duchaine, Trudel, Vézina, Siegrist and Wahrendorf2016). Yet, the model’s significance for policy is not restricted to the organizational level but extends to the level of national labor and social policies. Cross-country investigations demonstrated that lower levels of effort-reward imbalance at work were documented in countries with high investments in active labor market policies, thus pointing to their potential mitigating effect (Lunau et al., Reference Lunau, Siegrist, Dragano and Wahrendorf2015). In conclusion, research based on the ERI model has extended the evidence of health-adverse psychosocial work environments above and beyond the demand-control approach.

Organizational Justice

The third concept selected for this contribution, organizational justice (OJ), has a long tradition in research on social and organizational psychology, where three basic notions of justice are distinguished: distributive, procedural, and relational justice (Greenberg & Cohen, Reference Greenberg and Cohen1982). The first notion points to the fairness of distribution of resources and rewards between members of a group, whereas the second notion stresses fair decision-making procedures and equal application of rules in organizational behavior. The third notion concerns respectful and trusting behavior between colleagues, and specifically between leaders or managers and employees within the organization, thus avoiding prejudice, bullying, and discrimination. Through the decades, research on organizational justice was devoted to the study of its effects on working people’s behavior within their companies and businesses. Analyses revealed that perceived injustice within organizations increased the likelihood of dissatisfaction, absenteeism, turnover intention, and attitudes of disengagement or even revenge (Greenberg & Colquitt, Reference Greenberg and Colquitt2005). Therefore, these findings were of direct practical interest for improving leadership behavior, management styles, teamwork collaboration, and work motivation. In 2002, Finnish psychologist Marko Elovainio and his team applied this model to the explanation of stress-related ill health and disease, documenting associations with mental health and sickness absence (Elovainio et al., Reference Elovainio, Kivimäki and Vahtera2002). The hypothesis states that all three notions of perceived injustice in an organization weaken employees’ sense of security, trust, and control, even in the absence of personally experienced offense, thus eliciting feelings of anxiety, tension, and resulting physiological stress reactions.

The OJ model is commonly measured by a set of standardized questionnaire items developed in 1991 (Moorman, Reference Moorman1991), but additional operationalizations have been used. Compared with the measurement of the DC and the ERI models, fewer reports on the scales’ psychometric properties have been published so far. The empirical evidence of associations of organizational justice with health and well-being has been summarized in two reviews, one focusing on mental health (Ndjaboué et al., Reference Ndjaboué, Brisson and Vézina2012), and one providing a general overview (Elovainio & Virtanen, Reference Elovainio, Virtanen and Theorell2020). Solid prospective evidence was reported for effects of relational and procedural injustice on reduced mental health, where effect sizes were comparable to those reported by the other two models. In addition, several investigations document an association with cardiovascular disease risk (Theorell et al., Reference Theorell, Jood, Järvholm, Vingård, Perk, Östergren and Hall2016). As preliminary evidence points to links with behavioral risk factors and with inflammatory markers (blood-based indicators of disease), there is reason to assume that perceived injustice affects the cardiovascular and other bodily systems via enhanced stress reactions and negative emotions threatening people’s sense of integrity and self (Elovainio & Virtanen, Reference Elovainio, Virtanen and Theorell2020).

Given the general significance of perceptions of organizational injustice, one can assume that health effects are observed in organizations other than companies and businesses. In fact, at least for one different type of organizations, schools, this assumption has been successfully tested (Elovainio & Virtanen, Reference Elovainio, Virtanen and Theorell2020). Overall, relational and procedural injustice are considered independent predictors of reduced mental health and increased cardiovascular risk, while the contribution of distributive justice deserves further clarification. Additional prospective investigations using a standardized assessment and including interventions are desirable to increase the impact of this promising conceptual approach.

Job Demands-Resources

As a shared characteristic, the models discussed so far define particular organizational features that affect human health and well-being. With their hypotheses, they offer explanations that can be empirically supported or falsified. The job demands-resources (JD-R) model provides a more general conceptual framework that includes a wider set of job demands and job resources and that addresses negative (impaired health) as well as positive motivational outcomes (performance, personal growth). This model was developed in several steps, starting as a concept that aimed at explaining burnout (Demerouti et al., Reference Demerouti, Bakker, Nachreiner and Schaufeli2001), being extended to include work engagement as a motivational process stimulated by job resources, and finally integrating personal resources as mediators and moderators of relations between job characteristics and well-being (Schaufeli & Taris, Reference Schaufeli, Taris, Bauer and Hämmig2014). As a general framework, the JD-R model calls on distinct psychological theories of coping with demands and of strengthening resilience and recovery in order to explain the observed associations. Such approaches include conservation of resources, social cognitive, and self-determination theories (Schaufeli & Taris, Reference Schaufeli, Taris, Bauer and Hämmig2014). The model has also inspired more specific applications, in particular the concept of psychosocial safety climate, a concept maintaining that managerial investments into psychosocial safety are instrumental in reducing job demands, increasing job resources, and promoting workers’ well-being (Dollard et al., Reference Dollard, Dormann and Idris2019). With its emphasis on direct and indirect effects of job resources on personal and organizational outcomes, the JD-R model adds an important dimension to the field of occupational health research, and it links this field with research on job design and human resource management (Bakker & Demerouti, Reference Bakker and Demerouti2017).

Abundant empirical support was documented for the model’s central hypotheses of direct positive effects of job resources on work engagement and of job demands on reduced well-being, while job resources also protect against reduced well-being. Fewer studies tested moderation and mediation effects, including personal resources. A recent meta-analytic review of 74 longitudinal studies found substantial support for the central hypotheses (Lesener et al., Reference Lesener, Gusy and Wolter2019). Similar to previously discussed concepts, demands and resources were also analyzed in contexts other than paid work, in particular with respect to home and family work (Bakker et al., Reference Bakker, Ten Brummelhuis, Prins and Van Der Heijden2011). Despite its strengths, a general limitation of the model concerns the use of self-report measures in assessing work aspects and well-being, the latter being measured in terms of burnout in most studies. Therefore, the findings are subject to the bias of common method variance, and independent prospective epidemiologic cohort studies with objectively assessed health outcomes are still needed to overcome this limitation. In summary, with its broad application, the JD-R model has produced considerable practical impact, for instance as a monitoring tool for evaluating organizational performance, and as a frame of reference for developing job redesign or job crafting interventions. With the differentiation between organizational and personal resources, and with the dual concept of positive and negative outcomes, this model has substantially enriched research in this field. A controversial issue concerns its epistemological status, i.e. the question whether it serves as an open, heuristic framework that includes any type of demands and resources, or whether it aims at reducing the potential complexity to more specific analytic dimensions, with an inherent benefit of increasing the model’s predictive power (Schaufeli & Taris Reference Schaufeli, Taris, Bauer and Hämmig2014).