Multilevel Job Demand-Control-Support: A between- and within-person examination of work characteristics and employee well-being

Research goals and why the work was worth doing: Research on the characteristics of the work environment has mainly focused on between-persons differences, investigating whether employees dealing with high job demands and low job resources generally experience impaired well-being (Morrison et al., 2003). A major shortcoming is that this approach assumes that variables are completely stable over time. Conversely, recent studies highlighted that the perceptions of work characteristics and well-being are likely to vary over the course of employees’ work experiences (e.g., Totterdell et al., 2006). Moreover, recent evidence showed that within-person fluctuations in work characteristics’ levels could have contingent effects on well-being (e.g., Pindek et al., 2019). Thus, the present study is aimed at: (a) investigating the degree of stability and fluctuations of the three main components of the Demand-Control (-Support) model (DCS; Karasek & Theorell, 1990) and well-being across a three-months period; and (b) examining the association among work characteristics and well-being at the between-persons and the within-person levels. In doing so, we provide an updated representation of the DCS by testing whether the model is also suitable for understanding how well-being may result from interactions between employee and more situational, dynamic characteristics. Moreover, we integrate the between-persons and the within-person perspectives to investigate whether and how work characteristics are associated with well-being at both levels of the employee functioning. Theoretical background: We rely on the DCS as our theoretical framework. The underlying rationale of the DCS is that impaired well-being occurs when specific combinations of workload, control and support make it difficult for employees to provide effective strategies to deal with the challenges of their job (Karasek & Theorell, 1990). This assumption has been tested at between-persons (or trait-like) level and, albeit only rarely, at within-person (or state-like) level. While the between-persons level reflects inter-individual variability and capture differences between employees, the within-person level reflects intra-individual variability and capture processes that happen within the individuals. So far, little is known on to what extent work characteristics and well-being (and their associations) differ with respect to their trait-like and state-like features, although most personality and organizational researchers contend that these variables are determined by both stable and dynamic factors. Design: We recruited a sample of Italian employees to set up a shortitudinal research design (Dormann & Griffin, 2015). At time 1 (T1) we had a total of 1308 participants. One month later (T2) we had 638 participants. After one month, the final sample at T3 included 317 workers. Workload, control and support were investigated through the Management Standards Indicator Tool (Rondinone et al., 2012), while well-being was assessed with the General Health Questionnaire (Goldberg & Williams, 1988). We used a Doubly Latent Multilevel Structural Equation Modelling approach (Marsh et al., 2009) to investigate our research questions. Results About two-thirds of the true variance in workload, control, support, and well-being was accounted for by the trait-like component, revealing that, although relatively stable, considerable within-person variations occur in the four variables. At between-persons level, trait-like workload, control and support were all significant predictors of trait-like well-being, explaining a total variance of 26.7%. Moreover, at within-person level, a non-negligible proportion of variance in well-being (6.4%) was accounted for by work characteristics, with higher-than-usual workload and higher-than-usual support significantly associated with state-like well-being. Limitations: The constructs were measured with self-report scales. Moreover, we limited our investigation to the three main components of the DCS and we did not include individual characteristics in our investigation. Conclusions and relevance to the congress theme: Our study demonstrates that the perception of work characteristics can fluctuate over time, and that such fluctuations can trigger certain levels of well-being on a more contingent basis. As the costs of poor mental health in the workforce are significant, proper occupational health interventions are needed to promote and protect employee well-being. These should enable employees to properly perform their tasks on a day-to-day basis, and to protect them against particularly intense periods of work which may lead to well-being issues.