Third party reactions to supervisor abuse : the roles of perceived responsibility and moral intensity

Abstract

Guided by attribution (Weiner, 1986, 1995) and moral intensity (Jones, 1991) theories, this dissertation develops and tests a moderated mediation model that accounts for third parties' emotional reactions (anger and sympathy) and dispositional factors (negative reciprocity and moral identity) in the relationship between cognitive evaluations (perceived responsibility, magnitude of consequence, and social consensus) and behavioural reactions (retaliation and emotional support) to supervisor abuse. It is proposed that upon witnessing supervisor abuse, third parties engage in cognitive evaluations in the form of assessing supervisor's responsibility, magnitude of consequence, and social consensus. These cognitive evaluations result in anger and retaliatory behaviours directed at the supervisor and sympathy and emotional support towards the target of abuse. Furthermore, these relationships are strengthened by negative reciprocity beliefs and moral identity. The proposed model was tested using scenario-based experiment (Study 1) and field survey studies (Studies 2 and 3). The primary objective of Study 1 was to test the relationship between perceived responsibility, magnitude of consequence, and social consensus (independent variables), and third parties' emotional response towards the abusive supervisor (anger) and abused subordinate (sympathy). Results of the study among 217 non-academic staff showed that perceptions of supervisor responsibility, magnitude of consequence, and social consensus had a significant main effect on anger towards the supervisor. Similarly, perceived responsibility and magnitude of consequence resulted in sympathy towards the target of abuse. Study 2 aimed to extend Study 1 by testing the full model of third parties' reactions to supervisor abuse and accounting for individual differences using 241 part-time graduate students. In Study 2, the conditional indirect effects of cognitive evaluations (both first and second stage moderation) in predicting retaliatory actions via anger were stronger for those with high as opposed to low moral identity (symbolization). Moreover, the conditional indirect effects of cognitive evaluations (first stage moderation) in predicting emotional support via sympathy were stronger for those with high as opposed to low moral identity (symbolization). Study 3 had two main objectives. First, it aimed to test the influence of both symbolization and internalization dimensions of moral identity on third party reactions to supervisor abuse. Results among 182 professional employees showed a differential moderating effects for the two sub-dimensions of moral identity within the cognitive evaluation-emotional response-behavioural reactions relationships. Second, examine the influence of negative reciprocity beliefs on the strength of the cognitive evaluations-anger-retaliation relationship. Results showed that the conditional indirect effects of cognitive evaluations in predicting retaliatory actions via anger were stronger for those with high as opposed to low negative reciprocity beliefs. This research program contributes to the abusive supervision literature in three important ways. First, it extends our understanding of supervisor abuse beyond the typical supervisor-subordinate dyads. Specifically, third party observers of abusive supervision can also be emotionally and behaviourally affected. Second, abusive supervision was examined using a moral perspective. Third, it considered how personality factors (moral identity and negative reciprocity beliefs) can predict third parties' reactions to supervisor abuse. Implications for research and practice are also discussed.