Two conflicting hypotheses exist on how work values change across time – Are they modelled in contrast to a perception of lack or surplus (contrast hypothesis) or reinforced by respective work rewards (reinforcement hypothesis)? We suggest that the results may depend on the level of analysis that is used to explore the data (sample-average, between-person, within-person) and investigated this question performing a bivariate latent growth curve model with intrinsic work values and rewards data from 999 individuals followed across 16 years from the age of 21/22 to 37/38 extended by an autoregressive latent trajectory model structured residuals. We further relate the levels and trajectories in work values and rewards to work centrality and job satisfaction. We find support for a divergent development of work values (contrast) on the sample-average level, a conjoint development (reinforcement) on the between-sample level, and no clear relations on the within-person level of analysis. Our findings help bring clarity to the inconsistent empirical evidence on how individuals’ work values change over time. Moreover, they can inform employers regarding the typical development of work values in their workforce and how work rewards may be best designed to maintain high intrinsic work values and satisfaction.
