EMPLOYEE CONTROL IN THE ERA OF FLEXIBILITY: A STRESS BUFFER OR A STRESS AMPLIFIER?

ABSTRACT

Traditionally, employee control has been considered a buffer against stress, but debate about the transformation of work raises new questions. Can individual control really counteract the high demands in downsized organisations? Or does it make flexible work even more difficult to define and delimit, physically and emotionally? This article, based on survey data from 800 Swedish employees, studies the effect of job control on work hours, work-family conflict and psychological wellbeing. The results show that high job demands are associated with longer work hours, more work-family conflict and lower wellbeing, while control has positive effects, even when demands are high.

Keywords:

• flexibility
• job control
• job demands
• psychological wellbeing
• work/family conflict
• work hours

Notes

¹To my knowledge, only Voydanoff (1998) has tested the demands/control model on work/family conflict. However, with data from 1977, the study does not capture the characteristics of modern working life. Also, Voydanoff does not discuss the main effects of demand and control (only interactions).

²The share of housework is an index of two questions asking how large a share of the housework the respondent′s carries out on weekdays and weekends, respectively. Swedish men have a mean value of 0.38 on a this 0-1 scale, which is significantly higher than in other European countries.

³Both regarding the service sector and white-collar employment, Belgium, the UK and the Netherlands rank about as high as Sweden. This is true also for Luxembourg and France regarding service-sector employment and for Switzerland regarding white-collar employment (OECD 2000).

⁴Interaction variables confirm that gender differences are significant at the two middle levels of control.

⁵In all groups, employees with very high control have higher psychological wellbeing than those with very low control, and a variable measuring curvilinearity is not significant. The idea that control has a curvilinear effect has previously been rejected in studies of depression and anxiety (Warr 1990; Fletcher and Jones 1993).

⁶This gender difference is significant or almost significant (0.06) at the two middle levels of control.

⁷The conclusion holds even if we focus on employees with very high demands (the 4^th quartile). Even in this group, job control (as a continous variable) significantly reduces work/family conflict and increases wellbeing.