Testing a causal model of job insecurity and job satisfaction : do dispositions matter?

Abstract

In the context of widespread downsizing and restructuring, job insecurity has emerged as a prolific stressor in modern working life. While previous research has linked job insecurity to diminished job satisfaction, scholars rarely examine the role of personality dispositions and have yet to establish the strength and direction of causation. To address these gaps, the present research develops and tests a causal model describing the influence of dispositional positive and negative affect (PA and NA) on the relationship between job insecurity and job satisfaction. To test this model, the author analyzes a three-wave longitudinal panel data set (n = 258) collected at one-year intervals using structural equations modelling (SEM). Prior to this investigation, two ancillary studies established a robust measurement model. The first study developed and validated a new job insecurity measure, randomly splitting the Time 1 data set (n = 1185) into two equal sub-samples, with one used for exploratory factor analysis (n = 502) and the other split into two subsamples of n = 251 each for confirmatory factor analysis with tests of measurement invariance. Results strongly supported a measurement model consisting of four correlated but distinct job insecurity dimensions: job loss, job changes, marginalization, and organizational survival.Further analyses of longitudinal invariance supported the structural invariance and stability of three of the four job insecurity scales and the intrinsic and extrinsic job satisfaction scales. The organizational survival construct was found to be unstable and was, therefore, excluded from further consideration. The second study, a psychometric meta-analysis of the job insecurity - job satisfaction relationship (K = 101; N = 70,957), showed stronger correlations between job insecurity and intrinsic (ρ = -.48) and extrinsic (ρ = -.44) job satisfaction relative to global measures (ρ = -.40). These results provided empirical justification for measuring job satisfaction at the dimension-specific level. Finally, SEM supported a parsimonious structural model, where dispositional PA at Time 1 increased satisfaction with pay and promotions at Time 2, which led to a reduction in marginalization insecurity and enhanced intrinsic job satisfaction at Time 3. The model also included structural paths from job changes insecurity at Time 1 to not only job changes insecurity but also to job loss and marginalization insecurity at Time 2. However, this influence diminished by Time 3. All direct and indirect autoregressive paths were statistically significant over time. Partial support was found for a less robust model with autoregressive paths removed and an additional path from extrinsic job satisfaction at Time 2 to job loss insecurity at Time 3. Although previous dispositional research has primarily focused on NA, this disposition did not impact job insecurity or job satisfaction longitudinally. The author concludes by discussing theoretical and practical implications of the model as well as directions for future research.