Autonomy and job control are concepts with a long history in the study of work and health. Autonomy—the extent to which workers can exercise discretion in how they perform their work—is most closely associated with theories that are concerned with the challenge of designing work so that it is intrinsically motivating, satisfying and conducive to physical and mental well-being. In virtually all such theories, the concept of autonomy plays a central role. The term control (defined below) is generally understood to have a broader meaning than autonomy. In fact, one could consider autonomy to be a specialized form of the more general concept of control. Because control is the more inclusive term, it will be used throughout the remainder of this article.
Throughout the 1980s, the concept of control formed the core of perhaps the most influential theory of occupational stress (see, for example, the review of the work stress literature by Ganster and Schaubroeck 1991b). This theory, usually known as the Job Decision Latitude Model (Karasek 1979) stimulated many large-scale epidemiological studies that investigated the joint effects of control in conjunction with a variety of demanding work conditions on worker health. Though there has been some controversy regarding the exact way that control might help determine health outcomes, epidemiologists and organizational psychologists have come to regard control as a critical variable that should be given serious consideration in any investigation of psychosocial work stress conditions. Concern for the possible detrimental effects of low worker control was so high, for example, that in 1987 the National Institute for Occupational Safety and Health (NIOSH) of the United States organized a special workshop of authorities from epidemiology, psychophysiology, and industrial and organizational psychology to critically review the evidence concerning the impact of control on worker health and well-being. This workshop eventually culminated in the comprehensive volume Job Control and Worker Health (Sauter, Hurrell and Cooper 1989) that provides a discussion of the global research efforts on control. Such widespread acknowledgement of the role of control in worker well-being also had an impact on governmental policy, with the Swedish Work Environment Act (Ministry of Labour 1987) stating that “the aim must be for work to be arranged in such a way so that the employee himself can influence his work situation”. In the remainder of this article I summarize the research evidence on work control with the goal of providing the occupational health and safety specialist with the following:
- a discussion of aspects of worker control that might be important
- guidelines about how to assess job control in the worksite
- ideas on how to intervene so as to reduce the deleterious effects of low worker control.
First, what exactly is meant by the term control? In its broadest sense it refers to workers’ ability to actually influence what happens in their work environment. Moreover, this ability to influence the work setting should be considered in light of the worker’s goals. The term refers to the ability to influence matters that are relevant to one’s personal goals. This emphasis on being able to influence the work environment distinguishes control from the related concept of predictability. The latter refers to one’s being able to anticipate what demands will be made on oneself, for example, but does not imply any ability to alter those demands. Lack of predictability constitutes a source of stress in its own right, particularly when it produces a high level of ambiguity about what performance strategies one ought to adopt to perform effectively or if one even has a secure future with the employer. Another distinction that should be made is that between control and the more inclusive concept of job complexity. Early conceptualizations of control considered it together with such aspects of work as skill level and availability of social interaction. Our discussion here discriminates control from these other domains of job complexity.
One can consider mechanisms by which workers can exercise control and the domains over which that control can apply. One way that workers can exercise control is by making decisions as individuals. These decisions can be about what tasks to complete, the order of those tasks, and the standards and processes to follow in completing those tasks, to name but a few. The worker might also have some collective control either through representation or by social action with co-workers. In terms of domains, control might apply to such matters as the work pace, the amount and timing of interaction with others, the physical work environment (lighting, noise and privacy), scheduling of vacations or even matters of policy at the worksite. Finally, one can distinguish between objective and subjective control. One might, for example, have the ability to choose one’s work pace but not be aware of it. Similarly, one might believe that one can influence policies in the workplace even though this influence is essentially nil.
How can the occupational health and safety specialist assess the level of control in a work situation? As recorded in the literature, basically two approaches have been taken. One approach has been to make an occupational-level determination of control. In this case every worker in a given occupation would be considered to have the same level of control, as it is assumed to be determined by the nature of the occupation itself. The disadvantage to this approach, of course, is that one cannot obtain much insight as to how workers are faring in a particular worksite, where their control might have been determined as much by their employer’s policies and practices as by their occupational status. The more common approach is to survey workers about their subjective perceptions of control. A number of psychometrically sound measures have been developed for this purpose and are readily available. The NIOSH control scale (McLaney and Hurrell 1988), for example, consists of sixteen questions and provides assessments of control in the domains of task, decision, resources and physical environment. Such scales can easily be incorporated into an assessment of worker safety and health concerns.
Is control a significant determinant of worker safety and health? This question has driven many large-scale research efforts since at least 1985. Since most of these studies have consisted of non- experimental field surveys in which control was not purposely manipulated, the evidence can only show a systematic correlation between control and health and safety outcome variables. The lack of experimental evidence prevents us from making direct causal assertions, but the correlational evidence is quite consistent in showing that workers with lower levels of control suffer more from mental and physical health complaints. The evidence is strongly suggestive, then, that increasing worker control constitutes a viable strategy for improving the health and welfare of workers. A more controversial question is whether control interacts with other sources of psychosocial stress to determine health outcomes. That is, will high control levels counteract the deleterious effects of other job demands? This is an intriguing question, for, if true, it suggests that the ill effects of high workloads, for example, can be negated by increasing worker control with no corresponding need to lower workload demands. The evidence is clearly mixed on this question, however. About as many investigators have reported such interaction effects as have not. Thus, control should not be considered a panacea that will cure the problems brought on by other psychosocial stressors.
Work by organizational researchers suggests that increasing worker control can significantly improve health and well-being. Moreover, it is relatively easy to make a diagnosis of low worker control through the use of brief survey measures. How can the health and safety specialist intervene, then, to increase worker control levels? As there are many domains of control, there are many ways to increase workplace control. These range from providing opportunities for workers to participate in decisions that affect them to the fundamental redesign of jobs. What is clearly important is that control domains be targeted that are relevant to the primary goals of the workers and that fit the situational demands. These domains can probably best be determined by involving workers in joint diagnosis and problem-solving sessions. It should be noted, however, that the kinds of changes in the workplace that in many cases are necessary to achieve real gains in control involve fundamental changes in management systems and policies. Increasing control might be as simple as providing a switch that allows machine-paced workers to control their pace, but it is just as likely to involve important changes in the decision-making authority of workers. Thus, organizational decision makers must usually be full and active supporters of control enhancing interventions.