From a restless pillow to a ruffled mind: testing a moderated mediation model of off-the-job antecedents of abusive supervision

Purpose Although work-related antecedents of abusive supervision are well-known, knowledge on the cross-domain antecedents of this destructive leadership behavior is scarce. Accordingly, this study aims to investigate off-the-job supervisors’ after-work experiences that may influence their work behavior. The authors explore how and when a supervisor’s poor recovery experiences lead to abusive supervisory behaviors through a negative start-of-workday mood for high vs low supervisor sleep quality. Design/methodology/approach The authors conducted a single-source, three-phase field study (N = 422) to test the proposed moderated mediation model for participants from a large telecommunications company located in Anhui province, People’s Republic of China. Findings Poor recovery experiences in the supervisor’s personal life can spill over to their work domain and provoke abusive supervisory behavior through the mediating effect of a negative start-of-workday mood. Moreover, a supervisor’s good night’s sleep (i.e. first-stage moderator) serves as a key mitigating factor to diminish the negative start-of-workday mood resulting from a lack of relaxation, mastery experiences and control experiences (except for the lack of psychological detachment from work) and lessens abusive supervision. Originality/value This study contributes to the literature examining off-the-job events to understand the antecedents of abusive supervision that are beyond organizations’ control but significantly influence workplace behaviors, showing that not all antecedents of abusive supervision exist in the workplace; some are transferred from the home domain through nonvisible moods. Finally, the inclusion of sleep quality as a first-stage moderator provides insights on preventing abusive supervision caused by nonwork-related events. This adds a unique dimension to the abusive supervision literature by highlighting factors in the home domain that can prevent negative spillovers to the work domain. The authors conclude with some theoretical and practical implications for researchers and practitioners.