Algorithmic Decision-making, Employees’ Justice Perception and Deviant Behaviors: The Moderating Role of Decision Favorableness

Abstract

Algorithms have been increasingly applied in decision-making in organizations. It is of theoretical significance and managerial relevance to explore how algorithmic decision-making affects employees’ psychological states and behaviors. Integrating the literatures of algorithmic management and justice, we adopt the attribution perspective to investigate how the type of decision-maker (algorithm vs. leader) and decision favorableness interact to affect employees’ justice perception and their subsequent organizational and social deviant behaviors. We propose that decision favorableness moderates the effect of the type of decision-makers on employees’ overall justice perception and the indirect effect of the type of decision-makers on employees’ deviant behaviors. Specifically, when the decision is unfavorable for employees, employees will perceive higher levels of overall justice if the decision is made by algorithms rather than by human leaders. In contrast, when the decision is favorable, employees’ overall justice perception will not differ significantly between the algorithmic decision situation and the leader decision situation. In addition, we propose that employees’ overall justice perception is negatively related to their deviant behaviors in both organizational and social settings. Furthermore, decision favorableness moderates the indirect effects of the type of decision-makers on employees’ organizational and social deviant behaviors via their overall justice perception, such that these indirect effects are stronger when the decision is unfavorable rather than favorable for employees.We conduct a qualitative pilot study and an experimental study to test the theoretical model. The results provide substantial support for our hypotheses. This research contributes to the emerging literature on the relationship between algorithmic decision-making and employees’ justice perception by identifying the moderating role of decision favorableness. In a broader scale, it also enhances the understanding on how algorithms affect employees’ perception. Moreover, by revealing the indirect effect of the type of decision-maker on employees’ deviant behaviors via justice perception, this research extends the scope of literature on how employees respond to algorithmic management. The findings also provide practical implications for management decision-making in the algorithm era.