The discipline of leadership is highly romanticized (Meindl, 1985). In particular, the popular press sensationalizes leaders by assigning them heroic qualities and crediting them with herculean feats of success. Common observation, however, suggests great people are almost always bad people (Acton, 1887) and that power is abused with surprising regularity (Kellerman, 2004). A relatively new wave of leadership research has exposed this phenomenon under a variety of banners, including petty tyranny, destructive leadership, and managerial derailment. Abusive supervision is one such area that focuses on the hostile actions perpetuated by a supervisor against their subordinates.
Although abusive supervision is a relatively low base-rate phenomenon (Tepper, Duffy, Henle, & Lambert, 2006), the annual damages it perpetuates in terms of health, productivity, retaliation, and employee withdrawal has been estimated to exceed $20 billion (Tepper, 2007). Clearly, the reduction of such behavior would greatly benefit followers and firms alike. While an impressive literature has been amassed on the consequences of abusive supervision (see Schyns & Schilling, 2013), relatively less empirical work has addressed why leaders – intentionally or otherwise – engage in subordinate mistreatment. At the time of his major review, Tepper (2007) noted only three studies on the antecedents of abuse, leading to calls for future research into its origins. Recognizing the likelihood abusive supervision is a multilevel and dynamic phenomenon, the goal of the current symposium is to address this question from a variety of vantage points, including leader characteristics, leader self-concepts, and environmental forces. Our ultimate aim is to help guide future research into this burgeoning arena.
The first study revisits the “great man,” or, in this case, “terrible man,” approach to leadership by pitting the normative side of personality (the Big Five) against its darker or maladaptive counterparts. Further, as a new trait model, Simonet, Bolen, and Nei argue derailing tendencies (as assessed by the Hogan Developmental Survey, HDS), owing to their interpersonally dysfunctional nature, should incrementally predict an abusive proxy above and beyond normative trait models. Using sequential logistic regressions and dominance analyses, they find multiple dysfunctional tendencies (e.g., excitable, cautious, leisurely, dutiful) increase the likelihood of classifying a supervisor as being too forceful and insensitive in their leadership style. Limitations are stressed and implications discussed.
Next, adopting a person-situation interactional approach, Schilling and Schyns provide a more likely portrayal of how supervisors’ dark side traits express themselves in harmful ways. Specifically, they found a main effect for Machiavellianism predicting abusive supervision. This finding was moderated by stress indicating Machiavellianism is less predictive of abusive supervision under high stress situations. They also found a main effect for narcissism predicting abusive supervision. This finding was moderated by procedural justice indicating that narcissism is more predictive of abusive supervision under low procedural justice situations. Collectively, findings suggest self-interested persons are more likely to mistreat subordinates, a tendency which is exacerbated by unfair procedures and potentially, at least for Machiavellian individuals, mitigated by stress.
The final panelist considers an array of macro-environmental factors which, to date, are woefully underrepresented in contemporary studies of abusive supervision. Using an ecological framework, Mulvey further develops the model of abusive supervision by considering the contextual factors of instability, perceived threat, cultural values, and an absence of checks and balances. Mulvey argues that this perspective allows for a richer and more useful set of research questions and conclusions. As such, this paper highlights the limitations of a purely behavioral perspective providing a contrast to the other papers.
This symposium will be held Thursday, May 15.
References available