Antecedents and outcomes of abusive supervision: Multi-level perspectives

Previous research has focused on the consequences of abusive supervision and the effects of moderators on the relationship between it and employee outcomes. Only a few studies have looked at the antecedents of this behavior and these have found that supervisors’ perceptions of fairness may lead to abusive supervision. Extending this research, this study examined antecedents of abusive supervision such as supervisors’ personal characteristics (e.g., narcissism), and supervisors’ level of stressors (e.g., job demands, work-family conflict). Also, I examined supervisors’ experience of abusive supervision as an antecedent to test a trickle-down model of this negative behavior. In addition, this study investigated subordinates’ dispositional characteristics such as neuroticism and implicit person theories as antecedents of abusive supervision. Using data from both supervisors and subordinates, the current study conducted multi-level analyses and tested the effect of abusive supervision on employee unit-level organizational citizenship behaviors and performance. Further, this study examined the moderating effects of coping styles on the relationship between abusive supervision and employee psychological distress. Ninety three supervisors and 402 subordinates were surveyed in order to examine the proposed relationships among constructs. Results showed that supervisors who have experienced abusive supervision from their current bosses may treat their subordinates with same negative behaviors. Also, I found that subordinates high in neuroticism perceive more abusive supervision than those low in neuroticism. Results indicated that abusive supervision is positively related to subordinates’ psychological distress. Findings also showed that employees who have experienced abusiveness by their supervisor tried various coping strategies. Unfortunately, individual coping strategies did not buffer the impact of abusive supervision. Implications for practice and limitations of the study are discussed.