Long-term sickness absence during pregnancy and the gender balance of workplaces

Aims: This study addresses how the gender composition of workplaces affects pregnant women's sickness absence. It also assesses whether an observed association may be explaine by differential selection to female- or male-dominated workplaces. Methods: The analyses are based on Norwegian registry data from 2003-2011. Using Poisson regressions with detailed control for occupational categories, I examine whether the number of absence days are associated with the proportion of females at the workplace. I address possible selection effects by Poisson regressions with fixed individual effects using only within-individual variation on women with two or more pregnancies during the time window. Results: The analyses indicate a positive and significant relationship between the female proportion in workplaces and sickness absence rates during pregnancy. Analyses limited to within-individual variation also show positive and significant effects of similar strength, indicating that the observed relationship is not due to differential selection of absence-prone pregnant workers to female-dominated workplaces. Conclusions: The proportion of female individuals at workplaces is positively associated with sickness absence rates during pregnancy. This association is not likely explained by occupational nor individual characteristics. The results are consistent with absence culture theory and more lenient norms concerning sickness absence during pregnancy at female-dominated workplaces.