Background: Sickness presenteeism, or presenteeism, refers to working while ill, instead of being on sick leave. Presenteeism is associated with adverse health including poor self-rated health, depression, and future health adversity. Moreover, factors of the psychosocial work environment, including demands, have be related to presenteeism. Work in higher education is increasingly characterized by high demands such as job insecurity. This may lead faculty to work longer and harder, in efforts to keep the job. This may, in turn, be associated with greater presenteeism. Conversely, resources, including autonomy are related to employee health. Work in higher education is often characterized by high autonomy. This may, in turn, be associated with lower presenteeism. Job insecurity and autonomy may have direct relationships to presenteeism. However, the relationships between demands and resources and health behaviours such as presenteeism may be mediated through a process of health impairment, namely burnout. This study aimed to investigate if and how job insecurity (a demand) and autonomy (a resource) are associated with sickness presenteeism, and whether burnout mediated the relationship, among faculty in Swedish academia. Specifically, four hypotheses were tested: 1) job insecurity is positively associated with presenteeism, 2) autonomy is negatively associated with presenteeism, 3) the relationship between job insecurity and presenteeism is mediated by burnout: job insecurity has a positive relationship with burnout, and burnout has a positive relationship with presenteeism, and 4) the relationship between autonomy and presenteeism is mediated by burnout: autonomy has a negative relationship with burnout, and burnout has a positive association with presenteeism.

Method: Self-reports of quantitative job insecurity, autonomy, burnout, and presenteeism were collected via online questionnaires. The analytic sample included 1899 individuals (55% women; mean age: 48 years), with a doctoral degree, working in Swedish higher education institutions. Structural equation modelling was performed to test the hypotheses. Direct and indirect effects were tested separately in two models.

Results: The first direct effects model showed that increasing job insecurity was associated with increasing presenteeism, thus confirming hypothesis 1. Moreover, increasing autonomy was associated with decreasing presenteeism which confirmed hypothesis 2. The second model, testing burnout as a mediator, provided support for hypothesis 3, showing that job insecurity had a positive, indirect effect on presenteeism via burnout. Finally, burnout was found to mediate the relationship between autonomy and presenteeism so that autonomy had a negative effect on presenteeism via burnout, which confirmed hypothesis 4. Overall, the indirect model fit the data better and explained 24 percent of the variance in burnout and 6 percent of the variance in presenteeism.

Conclusion: This cross-sectional study suggests that faculty in Swedish academia facing job insecurity may experience increasing burnout and increasing presenteeism. Importantly, however, autonomy seems a beneficial resource. This is important given any future implications for the work ability of faculty who work when ill.