The mental health of workers: the role of job quality and physical activity

Bracci, Anna and Soldini, Emiliano and Lisi, Angela and Mangili, Francesca and Bonesana, Claudio and Kahlmeier, Sonja (2021) The mental health of workers: the role of job quality and physical activity. Project Report (Unpublished)

Full text not available from this repository.


This research aims at investigating the link between mental health, job quality and healthy behaviours among Swiss workers. Despite the relevance of considering the joint effect of work- and non-work activities as health factors, these relationships have been widely neglected in the literature. The study focuses on the relation between job quality and mental health in the Swiss context, where the number of new claimants for disability benefits with a mental disorder continues to increase and spending on sickness and disability benefits remains high. Following a multidimensional definition of job quality, we identify a plurality of both extrinsic and intrinsic conditions at work. Thus, we estimate a Bayesian network to explore the relationship structure of the variables considered using the Swiss Household Panel (SHP) data from the years 2006 to 2017. Exploiting the panel nature of the SHP, we build a dynamic Bayesian network with lagged values to explore the longitudinal relationship structure of data (useful to account for path dependence) and fixed-effects linear models with two mental health outcomes (Mental Health Index and Vitality Index) in order to control for bias from unobservable omitted variables bias. The contribution to the literature of this study is related to two elements: on the one hand, the comparison of the results obtained using two different statistical methods (Bayesian networks and fixed-effects linear models), on the other hand, the consideration of work- and non-work conditions simultaneously. On the basis of the results, we finally develop a pilot support tool, by building a hybrid dynamic Bayesian network model, with the aim of helping employers and policymakers identifying psychosocial risk factors of mental problems. The empirical evidence seems to corroborate the positive relations between specific characteristics of job quality and self-perceived mental health, whereas non-work activity seems to play a minor role in mental health. And, concerning job quality, a range of working conditions seem to be more correlated to workers' wellbeing compared to the traditional employment conditions (pay and contractual status). We identify, indeed, many working conditions connected to the security, ergonomic and working time quality dimensions as potential drivers for the Mental Health Index, while security, skill and ergonomic dimensions are relevant in the Vitality Index. Similar results are obtained through the Bayesian networks model, whereby variables significantly related to the outcomes are distributed in all job quality dimensions, although the specific variables are slightly different when compared to the econometric results. Additionally, our results seem to support the hypothesis that occupational physical activity – captured by specific ergonomic risks - is detrimental to mental health. The findings are relevant for policy purposes as the structural change in the labour market has transformed the social patterns and imply new health and safety risks for workers. Methodologically, this study highlights how unobservables deserve attention in both the psychological and epidemiological literature, and also confirms an autoregressive process in mental health that should be considered in literature in order to avoid biased estimates. We conclude by designing potential research strategies at the Department level, where we are regularly interested in investigating simultaneous interactions between economic, social and health phenomena.

Actions (login required)

View Item View Item