Sickness absence in female- and male-dominated occupations and workplaces

Abstract

Previous research suggests that both men's and women's level of sickness absence may be systematically related to the gender composition of their workplace as well as of their occupational category. The number of studies is, however, low and the composition of the occupational category has often been used as a proxy for the composition of the workplace. This paper employs a large data set broadly representative of the employed population of Norway. The data make it possible to take workplace and occupation simultaneously into account. Thus, the relationship between the gender composition of the workplace and sickness absence is estimated with detailed control for differences between occupational categories. Likewise, the importance of the gender composition of the occupation is assessed with control for between workplace variation. Men's sickness absence turns out to be largely unrelated to the gender composition of the workplace. For women the level of sickness absence tends to be higher in female-dominated workplaces, but the relationship is weak. These findings provide evidence against theories suggesting that the minority sex in the workplace faces special problems and is therefore more absent. They are to some extent consistent with the idea that female-dominated workplaces develop norms that are more tolerant towards sickness absence. The relationship of sickness absence to the gender composition of the occupational category is similar to the U-shaped pattern found in several previous studies (highest sickness absence both in strongly male-dominated and strongly female-dominated occupations), but again the relationship is weak.

Introduction

A large amount of research shows that women most often have higher rates of sickness absence than men (see, e.g., Barmby, Ercolani, & Treble, 2002; Mastekaasa & Olsen, 1998). In Norway, this gender difference has tended to increase over time, but in the last 10 years it has been quite stable with women 40–50% more absent than men.¹ More limited evidence suggests that there may also be systematic differences in sickness absence between female- and male-dominated occupations and/or workplaces. More specifically, several studies suggest that men as well as women tend to have higher absence rates in occupations or workplaces numerically dominated by the opposite sex (Alexanderson, Leijon, Åkerlind, Rydh, & Bjurulf, 1994; Evans & Steptoe, 2002; Hensing, Alexanderson, Åkerlind, & Bjurulf, 1995; Tsui, Egan, & O’Reilly, 1992).

Although the evidence on the relationship between gender segregation and sickness absence is quite limited, the more general issue of how gender segregation in the workplace affects men and women has received considerable attention in organizational research (for overviews, see, e.g., Williams & O’Reilly, 1998; Reskin, McBrier, & Kmec, 1999). Particularly influential has been Kanter's (1977) theory of “tokenism”. Kanter suggests that small minorities, like women in predominantly male workplaces, are faced with special problems. The basic issue is that members of small minorities are not perceived and treated as individuals but rather as representatives or “tokens” of their category. This may have a number of negative effects on social relationships in the workplace, and increase levels of stress (Hunt & Emslie, 1998). A related albeit different idea is that traditionally privileged majorities may feel that their advantaged position is threatened by the minority, and that the minority is therefore subject to various kinds of hostile behaviour (Blalock, 1967).

Much of the relevant empirical literature both within the field of sickness absence research and in the organisational literature suffers from a number of weaknesses. As far as the present author is aware, all previous studies measure the gender composition either within occupational categories or within workplaces. In studies measuring the gender composition of occupational categories, the effects of working in a male- or female-dominated environment will easily be confounded with broader differences in working conditions between traditionally male and female jobs (Glass, 1990). This problem will also be present in studies measuring workplace gender composition unless detailed control for occupation or working conditions is introduced. In addition, most studies are based on small and often unrepresentative samples (with some exceptions, e.g., Alexanderson et al., 1994; Hensing et al., 1995).

In this paper, I use a very large data set ($N=156,000$) that is broadly representative of the population of employees in Norway. The data makes it possible to take workplace and occupation simultaneously into account. Thus, the relationship between the gender composition of the workplace and sickness absence is estimated with detailed control for differences between occupational categories. Likewise, the importance of the gender composition of the occupation will be assessed with control for between workplace variation.