Elsevier

Social Science & Medicine

Volume 67, Issue 6, September 2008, Pages 956-964
Social Science & Medicine

Going ill to work – What personal circumstances, attitudes and work-related factors are associated with sickness presenteeism?

https://doi.org/10.1016/j.socscimed.2008.05.022Get rights and content

Abstract

The aim of this study is to assess the impact of a broad range of possible factors relating to work, personal circumstances and attitudes towards sickness absence on a person's decision to go to work despite feeling ill, a phenomenon that has been termed sickness presence (SP), or ‘presenteeism’, in the literature. Using data from a random sample of 12,935 members from the core Danish work force the hypotheses were tested in a cross-sectional design utilising ordered logistic regression models. The results indicate that more than 70% of the core work force goes ill to work at least once during a 12-month period. This means that SP is just as prevalent a phenomenon as sickness absence. Many of the results from earlier studies of SP were replicated and new factors were discovered: for example time pressure (having a supervisory role and/or working more than 45 h per week) and relationship with colleagues (measured by working in a small company, having non-standard hours and degree of cooperation) both increase the likelihood of SP. However, personal circumstances and attitudes, e.g. treating work as home (cf. Hochschild's thesis) and being over-committed to work, were also found to lead to higher levels of SP. Finally, we found that those with a conservative attitude to absence were most likely to turn up ill at work. Overall, work-related factors seem to be slightly more important than personal circumstances or attitudes in determining people's ‘decision’ to go ill at work. However, the relatively low explanatory power of these combined factors suggests that there are still many unknowns in this field of research.

Introduction

In a recent article, Aronsson and Gustafsson (2005) reveal that research into the phenomenon of sickness presenteeism, in which an employee goes to work despite feeling ill, has become increasingly widespread over the last decade. However, a comparison of the amount of research devoted to sickness presenteeism (SP) as opposed to sickness absenteeism (SA) shows a staggering contrast: the number of articles dealing with SP is only 0.01% of the number dealing with SA (Dew, Keefe, & Small, 2005). While SP may be a relatively new research area, it is by no means a new phenomenon. In fact one might argue that SP both pre-dates and is more fundamental than SA. Before the advent of paid sick leave or publicly financed compensation for illness, it was probably more common to turn up ill at work than to stay at home if one was suffering from a non-fatal, non-incapacitating disease. In the early 20th century employees simply could not afford to stay home when ill.

One could argue that the lack of research into SP reflects its irrelevance. However, several factors count against this interpretation. First, the results from the Whitehall II study show that people with poor self-rated health who took no absence from work were twice as likely to experience serious coronary events when compared to those who took 1–7 sick days per year (Kivimäki et al., 2005). SP may thus cause serious health problems at a later stage. Second, illness at work may severely decrease productivity, making SP a costly affair for employers as well as society at large (Hemp, 2004). Finally, studying SP could enhance our understanding of SA, because both phenomena are ‘outcomes’ of the same ‘decision process’ (Kristensen, 1991).

An alternative explanation for the lack of SP research might lie in the difficulty of defining the phenomenon (Vingård, Alexanderson, & Norlund, 2004, chap. 10). If SA cannot be thought of merely as a function of a person's objective health status, because it involves her perception of her own health as well, this will also be true for SP. In addition, contextual factors play a major role, since the difficulty of working when already suffering the symptoms of illness will depend a good deal on the nature of one's work. This makes studies of SP quite different from those of SA, where more objective outcome measures, such as company registers, exist.

In this study, we define SP as the situation in which an employee goes to work despite perceiving herself to be sufficiently ill to have legitimately called in sick. Using legitimacy as the criterion for defining SP means that our focus is on the sociological aspects of the phenomenon (i.e. those relating to SP as a way of acting in a given context) rather than on the medical aspects (e.g. the long-term health consequences of SP).

A recent review of SA research devoted one small chapter to SP in which eight studies dealing exclusively with the causes of SP were described (Vingård et al., 2004, chap. 10). This confirms that current knowledge of the subject is very sparse. This is due partly to the small number of studies, and partly to the limited theoretical understanding of SP that these studies evince, most of them containing only tentative hypotheses put forward in various articles that do not deal with SP per se.

One exception is the study by Aronsson and Gustafsson (2005), which gives a brief summary of the main hypotheses relating to SP and proposes a theoretical model. The authors distinguish between ‘work-related demands for presence’ and ‘personally related demands for presence’, both external factors that may in some way influence the decision to go ill to work. While these two factors are certainly of importance, we believe it is necessary to draw further distinctions between different types of factors within these categories, as well as to incorporate the influence of attitudes, norms and values on the ‘decision’ to go ill to work. Norms cannot be attributed exclusively to a single individual but are instead internalised and appropriated by the individual in various ways that do not follow automatically (or deterministically) from the norms prevailing in society or the workplace. This is what Kristensen (1991) calls the voluntaristic aspect of SA theory. But voluntarism applies to SP as well, since SA and SP are part of the same ‘decision-making process’ in which the employee ‘decides’ whether to go ill to work or to call in sick. One way of describing this dialectic between external and internal forces can be found in the work of Pierre Bourdieu (1984) and the uses to which his theory has been put by Virtanen, Nakari, Ahonen, Vahtera, and Pentti (2000). From this vantage point SP would be seen as one element of an individual's SA practice rooted in habitus and in the interplay between, on one hand, the structure of social fields and the constraints they pose on action, and, on the other, the way these structures are appropriated and perceived by the individual.

Thus at this stage of research it is not possible to embark from one comprehensive theory of SP from which one could derive a set of hypotheses to be tested. Instead we can examine the validity of the various scattered remarks dealing directly and indirectly with SP, and on this basis suggest what needs to be examined more closely.

In this article, we propose to subdivide the hypotheses tested and the factors influencing the decision to turn up ill at work into three main areas: (1) work-related factors, (2) personal circumstances and (3) attitudes.

  • (1)

    There are at least four different types of work-related factors that have been hypothesised to influence the decision to turn up ill at work. (a) Time pressure: One set of factors is related to the results of downsizing/rationalisation and includes ‘lack of resources’/‘time pressure’. This may be linked directly to SP because employees who lack resources will be reluctant to take sick leave on the grounds that they will face a backlog of tasks when they return to work (Aronsson & Gustafsson, 2005). In addition, those who work in some supervisory capacity and/or routinely work a higher-than-average number of hours per week are likely to feel under greater time pressure if they take time off for illness. (b) Control over work tasks: Johansson and Lundberg (2004) have proposed a model of ‘illness flexibility’ where they hypothesise that employees with a high degree of control over their work tasks (‘adjustment latitude’ in their terminology) are more likely to go ill to work because they can modify their work tasks in such a way as to be able to carry on despite ill health. (c) Relationship with colleagues: Grinyer and Singleton (2000) point to the influence of team-work and pressure from colleagues on the decision to turn up ill at work. Higher levels of cooperation in performing work tasks are therefore hypothesised to be associated with higher levels of SP. The size of the firm/institution may also be a factor in this regard. In small companies, employees will be more dependent on each other because the tasks of a person on sick leave will have to be shared among fewer colleagues. One would thus expect that the smaller the size of the firm, the higher the level of SP. Moreover, employees who enjoy a high degree of social support from their colleagues would be more motivated by these bonds to show up ill at work than those who lack such support. Finally, those who work non-standard hours (e.g. doing shift work) may be more likely to come ill to work because they do not want to oblige their colleagues to work at odd hours of the day. (d) Employment conditions: In a few studies, job insecurity has been shown to be the most plausible explanation for sudden drops in SA rates during periods of lay-offs. In two Finnish studies, an ‘epidemic of good health’ broke out after a round of lay-offs had been announced (Virtanen, 1994). Job insecurity is thus hypothesised to influence the decision to go ill to work in situations where people fear losing their jobs. Employees with contingent employment (e.g. fixed-term appointments) also face this type of fear because their job situation is chronically insecure; we would thus expect higher levels of SP amongst this group as well (Virtanen, Kivimaki, Elovainio, Vahtera, & Cooper, 2002).

  • (2)

    Factors outside work may also influence the ‘decision’ to go ill to work. As was the case with the work-related factors we should distinguish between several types of personal circumstances. (a) Financial situation: As noted above, we can assume that before the advent of paid sick leave, it was more common to turn up ill at work than it is today, when the majority of working people have fewer financial incentives for SP. In Denmark, all employees have the right to sick pay from their first day of illness. White collar workers are entitled to their full salary during SA. However, many blue collar and unskilled workers, and employees whose wage structure is heavily dependent on pay supplements due to e.g. working at odd hours, are entitled only to sick pay equalling 90% of the minimum wage in their trade. For this group there are financial incentives for going ill to work. (b) Family life: In order to find out whether family life influences the decision to go ill to work we need to examine the situations in which it can be a ‘negative presence factor’ (Kristensen, 1991), i.e. a factor that encourages people to turn up at work despite ill health instead of staying at home. Hochschild (1997) has put forward the thesis that for some people ‘work becomes home and home becomes work’: in other words, they find it more taxing to stay at home than to go to a stimulating and interesting workplace. Such people, and those who are generally dissatisfied with family life, would be more likely to go to work when ill. A large number of children or a sick spouse could also be seen as factors that make home life more exhausting and that would be likely to prompt SP. (c) Psychological factors: Finally, psychological characteristics are among the personal factors that may influence SP. Aronsson and Gustafsson (2005) note the influence of what they call ‘individual boundarylessness’, a personality characteristic that makes it difficult for people to say no to other people's wishes and demands. Along similar lines, Siegrist (1996) has described what he calls ‘over-commitment’. Behind both theories is the idea that a strong commitment to work will increase the likelihood of SP.

  • (3)

    Finally, attention should also be given to how these structures relating to work and family are appropriated and interpreted by the individual in defining and legitimating their ‘choice’ between SA and SP. McKevitt et al. asked a group of medical doctors and ‘company fee earners’ to indicate what their reasons had been for going ill to work. In addition to arguing that it would be ‘unfair to their colleagues’ to stay home and that ‘patients were already booked’ (i.e. reasons relating to the work-related factors discussed above) many mentioned ‘work ethics’ (e.g. ‘commitment to work’ and ‘taking sick leave is disapproved of’) as important reasons (McKevitt, Morgan, Dundas, & Holland, 1997). In a Finnish study substantial differences were found in the perception of SA across three different communities. For those who identified strongly with the working class, SA was seen as a (legal) right obtained through years of struggle with employers. Work commitment was stronger among those of middle-class origin, and entailed a sense of duty towards the workplace that might prompt people to turn up ill at work instead of taking SA (Virtanen et al., 2000). It is thus hypothesised that ‘work ethics’ and differences in attitude towards SA in different situations are related to SP.

An important feature of some of the above-mentioned factors is that they are related to SP by more than one causal mechanism, a phenomenon that has been dubbed ‘double risk factors’ by Aronsson and Gustafsson (2005). Some of these factors have already been identified as ‘stressors’ that are detrimental to health. This is the case for example with time pressure, the degree/absence of control over one's own work tasks and having a sick spouse. This means that in principle any factor found to be a stressor will be related to SP, albeit indirectly, because we can assume that worsening health status leads to more situations in which the employee has to ‘decide’ between taking SA or going ill to work, thereby increasing the probability of the latter. However, some of these factors are also associated directly with SP, i.e. they are associated with the ‘decision’ to turn up ill at work (e.g. being a supervisor or working closely together with colleagues). In line with our sociological perspective, we are interested here mainly in the direct effect of a given factor on SP. However, we cannot ignore the total effect of each of these factors on SP, which is arrived at by combining the direct and indirect effects of each factor. Unless we take this total effect into account, we are likely to underestimate the importance of ‘double risk factors’. But it is also interesting to examine those factors that have a greater direct than total effect because they highlight situations in which circumstances at work or at home encourage the ill person to ‘choose’ presence over absence, leading to levels of SA lower than their health status would lead one to expect.

To sum up, this study aims to replicate existing results relating to SP as well as to test new hypotheses that have only been hinted at in the literature so far. The broad focus of the study, moreover, enables us to examine the relative influence of the three different types of factor. In an area where we still lack basic knowledge, it is particularly interesting to see whether work-related factors, personal circumstances or attitudes play the greatest role. Finally, by distinguishing between the direct and the total effect of a factor on SP we are able to identify ‘double risk factors’ and those factors that conceal a group's ‘true’ levels of SA.

Section snippets

Methods

The data were taken from a postal questionnaire sent to a random sample of the Danish core work force between the ages of 19 and 64 years. Only employees who had actually been in employment for at least 80% of the time during the previous year (i.e. core members of the work force) were included in the analysis (18,902 respondents). Employees who had taken SA for more than 10 weeks in the 12 months preceding the baseline questionnaire (sent out in September 2004) were thus excluded from the

Statistical analysis

Ordered logistic regression models were used to examine the associations between the IVs and SP. Four models were tested: we calculated the relationships between each of the IVs and SP, adjusted for socio-demographic variables only (Model 0), then a model with all factors relating to work, personal circumstances and attitudes adjusted for socio-demographic variables (Model 1), one model with all the health measures adjusted for socio-demographic variables (Model 2) and finally one model that

The prevalence of SP

Table 1 contains descriptive statistics of all variables. Only 27% of working Danes have no episodes of SP over a one-year period. Of those respondents who had episodes of SP, 38% had gone ill to work 2–3 times while approximately 8% had done so more than five times. These figures were somewhat lower for employees who reported no serious health problems (defined as those indicating no disease, no significant musculoskeletal pain and high mental health): of this group, 44% had no SP episodes and

The prevalence and measurement of SP

The results presented above show that SP is just as prevalent in the working population as SA. This confirms Dew et al.'s (2005) claim that SP is under-prioritised as a research subject. If we compare our results to those from the Swedish studies by Aronsson we find that they are quite similar (Aronsson and Gustafsson, 2005, Aronsson et al., 2000). The strong association between SA and SP indicates that the two phenomena are indeed outcomes of the same ‘decision process’ and that taking sick

References (28)

  • G. Esping-Andersen

    The three worlds of welfare capitalism

    (1990)
  • H. Ganzeboom et al.

    Three internationally standardised measures for comparative research on occupational status

  • A. Grinyer et al.

    Sickness absence as risk-taking behaviour: a study of organisational and cultural factors in the public sector

    Health, Risk & Society

    (2000)
  • Hansen, C., Andersen, J., & Mikkelsen, S. Attitudes towards sickness absence – what effect do they have on future...
  • Cited by (359)

    • Health uninsurance premium and mortgage interest rates

      2023, International Review of Financial Analysis
    • Auditor health and audit outcomes before COVID-19

      2023, Journal of Accounting and Public Policy
    View all citing articles on Scopus

    The authors would like to thank The Danish Health and Safety Research Fund for financing the study. The authors also wish to thank the anonymous reviewers for their many helpful suggestions.

    View full text