| There have been many attempts to define destructive leadership (e.g., Craig & Kaiser, 2012; Einarsen et al., 2007; Krasikova, Green, & LeBreton, 2013). An important question is whether the label of destructive leadership should be reserved for actively counterproductive leader behaviors, or whether it should also include the absence of desirable leader behaviors (i.e., general leader ineffectiveness). Researchers have not yet reached a consensus as to whether destructive leadership and leader (in)effectiveness are distinct from one another. The main goal of this study was to determine whether or not destructive leadership and leader ineffectiveness are independent from one another. To conduct this investigation, destructive leadership and related constructs were separated into values-centric and performance-centric constructs. Values-centric constructs (e.g., destructive leadership, abusive supervision, ethical leadership) refer to the types of values underlying a leader's goals and the behavior that he or she demonstrates in the process of achieving those goals. Performance-centric constructs (e.g., managerial derailment, laissez-faire leadership) refer to how effective a leader is at leading the group to goals, irrespective of any values-based judgments about those goals. This study examined whether values-centric constructs were distinct from performance-centric constructs in terms of conceptual definition, factor structure, and external correlates, and how many unique dimensions underlay the values-centric constructs. Data were collected from 17 subject matter experts (SMEs) and 1,648 full-time employees. SMEs agreed that items belonged to either values-centric or performance-centric constructs, not both. Employee data revealed that supervisors did not simultaneously exhibit highly destructive behaviors and high performance with meaningful frequency, nor score low in both categories. Exploratory factor analyses revealed a four-factor solution among the constructs: destructiveness, derailment, prosocial behavior, and task orientation. Follow-up confirmatory factor analyses confirmed an acceptable fitting four-factor model indicating that these factors are distinct, but related. Correlations indicated that the relationships between destructive leadership and several outcome variables were significantly different from derailment, prosocial behavior, and task orientation's relationships with those outcome variables. However, several performance-centric relationships with outcomes were not significantly different from destructive leadership's correlations. Hierarchical multiple regression analyses revealed that destructiveness predicted significant additional variance above and beyond performance-centric factors for all outcome variables except for follower well-being. Overall, these findings suggest that values-centric and performance-centric constructs are distinct in terms of conceptual definition and factor structure, but show similar patterns of relations to external correlates. Factor analyses revealed that two dimensions underlie the set of values-centric constructs: active destructive behavior and active prosocial behavior. Implications, limitations, and future directions for research are discussed. |
