|
Article Excerpt
Do people experience greater psychological strain during workweeks that involve higher work demands and less opportunity to control those demands? According to Karasek's (1979) job strain model, psychological strain results from the 'joint effects of the demands of a work situation and the range of decision-making freedom (discretion) available to the worker facing those demands' (p. 287). Although the job strain model was formulated to explain why jobs with different characteristics have different psychological effects on well-being, it has typically been tested in relation to differences between individuals' perceptions of their job (Morrison, Payne, & Wall, 2003). The model has not, however, been tested as a potential explanation for ongoing changes in psychological strain within individual workers over time, even though it is clear that perceived demands and control can vary over time in many types of work. The present study therefore examined whether the job strain model applies to intra-individual changes in work characteristics and psychological strain. To test the model in this way, we studied a group of self-employed workers (without employees) who worked for multiple clients. This group--typically known as portfolio workers--was chosen because it seemed likely to experience week-to-week variations in demands and control. We also examined other important facets of the job strain model, including the influence of social support and individual differences (specifically, optimism and emotional stability). To introduce the study, we review the evidence for the job strain model, describe what is known about portfolio work and then summarize the aims and contribution of the study.
The job steam model
Over the last quarter century, Karasek's (1979) job strain model has become the predominant theory of work stress. The model identifies job characteristics as the principal source of distress in the workplace and specifically proposes that psychological strain is caused by the combination of high job demands and low job control (and for this reason it is also called the demands-control model). The underlying rationale of the model is that workers experience distress when this combination of circumstances exists because they are constrained from formulating effective responses to deal with the challenges of the job. Other researchers (Johnson & Hall, 1988; Payne, 1979) have extended the model to include lack of social support as a third potential source of job strain; this has become known as the demand-control-support model.
In practice, four versions of the job strain model have been tested because there is an additive and an interactive form for both the demand-control (DC) and demand-control-support (DCS) models. The additive form of the DCS model predicts that high demands and low control and low social support each cause psychological strain. The interactive form of the DCS model predicts that control and social support buffer the negative impact of high demands on well-being (i.e. they interact with demands to reduce its negative impact). Both forms predict that psychological strain will be greatest given the combination of high demands with low control and low social support (in the DCS model).
The evidence for the job strain model is mixed. For example, Van der Doef and Maes (1999) found that there was strongest support for the additive form of the DC model in their review of 63 empirical tests of the job strain model derived from 20 years of research. Support for the additive form of the DCS model and the interactive form of the DC model was only found in about half of studies, but there were an insufficient number of tests of the interactive form of the DCS model to reach any conclusion about it. The review also found that longitudinal studies were generally non-supportive of the job strain model. A more recent review by de Lange and colleagues (de Lange, Taris, Kompier, Houtman, & Bongers, 2003) selected 19 of the most methodologically rigorous from 45 longitudinal tests of the model. Eight of the 19 studies showed support for the DC or DCS models, usually in the form of additive effects. These high quality studies therefore provided only modest support for the job strain model.
Studies that have supported the model appear to differ from non-supportive studies in a number of ways. First, support for the model appears to partly depend on the way in which demands, control and support are conceptualized and measured. Van der Doef and Maes (1999), for example, reported that buffering effects were more likely when measures of control and support corresponded to the specific demands of the job. Job demands have often been measured in terms of worldoad or time pressure, but other types of demand, such as the amount or type of interaction with customers, may be more central in some jobs. Payne and Morrison (1999) also highlight the importance of knowing the affective meaning of job demands, because not all demands are stressful. Second, the choice of outcome measure may be important. For example, there has been greater support for the model when job-related well-being has been measured rather than more general measures of well-being (van der Doef & Maes, 1999). Third, certain subpopulations appear to be more susceptible to the effects of demands, control and social support. For example, the relationship between job characteristics and strain has been found to depend on personality characteristics and on the coping strategies that individuals typically use (Parkes, 1994).
To explain the influence of personality, it has been suggested that differences in the way people adapt to their environment can influence the work stress process (Parker & Sprigg, 1999). A vulnerability model of this process would suggest that only people with certain characteristics are able to take advantage of available job control to manage their job demands and thereby minimize their strain. However, such people will be unable to influence their job strain unless they are afforded control in their jobs. Alternatively, a person-environment-fit model would suggest that the relationship between demands and strain will depend on the fit between personal and work characteristics (see Parkes, 1994), such that some people may even benefit from low control. Both accounts predict that job strain depends on a three-way interaction between demands, control and personal characteristics. A number of personal characteristics have been found to exhibit a three-way interaction with demands and control, including proactive personality (Parker & Sprigg, 1999), self-efficacy (Schaubroeck, Lam, & Xie, 2000) and active coping (de Rijk, LeBlanc, Schaufeli, & de Jonge, 1998).
The influence of such personal characteristics suggests that employees who believe in their capability to achieve positive outcomes and act on that belief are most likely to benefit from job control in high demand situations. Optimism, which is a generalized expectancy of positive outcomes (or a tendency to look on the bright rather than the dark side of life), has also been shown to moderate the relationship between work stressors and mental and physical well-being (Makikangas & Kinnunen, 2003; Parkes, 1994; Riolli & Savicki, 2003). Usually, optimism has been found to buffer this relationship, but some studies have found it can exacerbate the negative impact of stressors (e.g. Chang & Sanna, 2003).
Optimism overlaps but is conceptually and empirically distinct from concepts based on belief in personal control over events such as self-mastery (Scheier, Carver, & Bridges, 1994). The beneficial health effects of optimism are thought to accrue from optimists' use of active coping strategies to focus on positive outcomes and overcome negative events (Geers, Handley, & McLarney, 2003)--although it has also been suggested that what counts is people's belief that they have the necessary coping skills rather than their actual use of coping skills (Ray, 2004). An alternative explanation is that the positive health effects are actually due to emotional stability rather than optimism, because there is a high negative correlation between optimism and neuroticism (see Parkes, 1994).
As well as a health advantage, optimism may also confer a performance-related advantage in jobs where success depends on external agents and rebuffs are common, such as in sales roles (Seligman & Schulman, 1986). This may also apply to the self-employed, who have been shown to have greater unrealistic optimism about financial outcomes than the employed (Arabsheibani, de Meza, Maloney, & Pearson, 2000).
We anticipate that optimism will have a moderating influence on the relationship between work stressors and job strain for portfolio workers because their work is a form of self-employment that depends on the vagaries of clients and therefore requires an optimist's ability to focus on positive outcomes and overcome negative events. In particular, we expect that optimism will reduce the negative impact of high demands and low control.
The research we have presented so far suggests that the success of the job strain model is affected by the measures chosen and by the inclusion of personal characteristics. Another consideration is the source of variance of the variables in the model. Specifically, multi-level tests of the job strain model have found that differences between jobs account for little of the variance in job strain, but the model is more successful when it is conceived in terms of individuals' perception of their job and their personal distress (Morrison et al., 2003).
Previous research on the job strain model has either made comparisons between the characteristics of different jobs or, more commonly, has compared individuals' perceptions of the characteristics of their job. These perceptions have been based on people's judgments of the general levels of demands, control and support in their job. Previous studies have therefore compared individuals with differing levels of DCS. Alliger and Williams (1993) observed that such research designs treat job-related...
|
