Engaged managers are not workaholics: Evidence from a longitudinal personcentered analysis

The aims of this two-year follow-up study among Finnish managers (n = 463) were twofold: first, to investigate the relation between work engagement and workaholism by utilizing both variable- and person-centered approaches and second, to explore whether and how experiences of work engagement and workaholism relate to job change during the study period. The variable-centered analysis based on Structural Equation Modelling revealed that the latent factors of work engagement and workaholism did not correlate with each other, thereby suggesting that they are independent constructs. The person centered inspection with Growth Mixture Modelling indicated four work engagement-workaholism classes: 1) 'high decreasing WE - low stable WH' (18%), 2) 'low increasing WE - average decreasing WH' (7%), 3) 'low decreasing WE - low stable WH' (6%), and 4) 'high stable WE - average stable WH' (68%). Overall, these results suggest first that also at the intra-individual level work engagement and workaholism were largely independent psychological states (changes in work engagement and workaholism were related only in the class 'low increasing WE - average decreasing WH', 7%); second, job conditions had an impact on the levels of both work engagement and workaholism as, typically, the participants in the class 'low increasing WE - average decreasing WH' had typically changed their job during the study period. The fact that work engagement and workaholism are sensitive to job changes suggests that both psychological conditions depend - at least partly - on the individual's work situation.