A model of injustice, abusive supervision, and negative affect

Article history: Received 23 December 2011 Received in revised form 1 November 2012 Accepted 8 November 2012 Available online 6 December 2012 In this study we test a model of workplace interactional injustice, abusive supervision, and subordinate outcomes (work–family conflict and job performance) using affect to explain behavior. In a sample of 200 full-time workers from various industries, their supervisors, and workers' familymembers, for a total sample of 600 respondents, we position state negative affect as the explanatory mechanism for both how supervisors' perceptions of injustice are associated with subordinates' perceptions of abusive supervision, and also how abusive supervision, in turn, may be associated with subordinates' job performance and their family members' perceptions of work–family conflict. Organizational justice theory underpins our model. © 2012 Elsevier Inc. All rights reserved.

The idea of "bad bosses" has not only captured the attention of the popular media in recent years (Horrible Bosses; Online.wallstreetjournal.com, 2011) but has also garnered substantial research attention in the organizational behavior/ management literature. In the academic literature, many researchers have called attention to this social problem by testing the outcomes, and to a lesser extent the antecedents (Tepper, 2007), of what has been named abusive supervision, that is, subordinates' perceptions of the extent to which their supervisor engages in the display of hostile verbal and nonverbal behavior, excluding physical harm, toward them over time (Tepper, 2000).
Researchers have recently engaged in attempts to model what have been called "trickle-down" effects of abusive supervision. These models illustrate that factors specific to the organization (e.g., Mawritz, Mayer, Hoobler, Wayne, & Marinova, 2012;Restubog, Scott, & Zagenczyk, 2011) or perceptions and characteristics of the supervisor (e.g., Aryee, Chen, Sun, & Debrah, 2007;Hoobler & Brass, 2006;Tepper, Duffy, Henle, & Lambert, 2006) set in motion a chain of events that result in a supervisor's abusive behavior. Though these studies have largely been cross-sectional in nature, they suggest an active process whereby events "flow downhill" to and from supervisor abuse, resulting in negative outcomes for organizations, subordinates, and even subordinates' family members outside of the workplace (Aryee et al., 2007;Hoobler & Brass, 2006;Restubog et al., 2011). These models acknowledge that abusive supervision is not a simple act committed by isolated "bad apple" (Felps, Mitchell, & Byington, 2006) bosses in organizations, but rather a complex interplay between organizational, extraorganizational, and individual factors.
The purpose of this study is to test one such model of abusive supervision that addresses a common limitation of many models in organizational behavior research. We address calls from Weiss (2002) and Judge and Ilies (2004) for models that explain workers' behavior not just through cognitive means, e.g., through their appraisal of job factors, but rather through affective ones, that is, emotions and affective states. Especially in the context of abusive supervision, where people's actions may be considered hostile and damaging to others interpersonally, we feel that negative affect (a general dimension of negative mood) may be particularly important in determining when and in explaining how negative events are associated with work and nonwork outcomes for subordinates. Using a sample of 200 full-time workers from various industries, their supervisors, and workers' family members, for a total sample of 600 respondents, we test a model that positions supervisor negative affect as the explanatory mechanism for how supervisors' perceptions of interactional injustice are associated with subordinates' perceptions of abusive supervision, and subordinate negative affect as the explanatory mechanism for how abusive supervision is associated with subordinates' performance and their family members' perceptions of work-family conflict (the degree to which work spills over to negatively impact the home and family sphere). Please see Fig. 1.

Theoretical development and hypotheses
Prior trickle-down models of abusive supervision have done a good job of illustrating that abusive supervision does not occur in a vacuum. While we still know relatively little about why abusive supervision occurs (Tepper, 2007), what is reasonably clear is that supervisors who are perceived as abusive do not enact behaviors like telling subordinates their thoughts or feelings are stupid, lying to them, or not giving credit when it is due, as random events. More likely, in response to organizational events or norms, and in "kick the dog" (Hoobler & Brass, 2006;Restubog et al., 2011) or trickle-down fashion, supervisors take their aggression out on those whom they have power over in the organization-their subordinates. Aryee and colleagues (2007) found that when supervisors reported unfair treatment from others in the organization, their subordinates reported greater abusive supervision. Likewise, Hoobler and Brass (2006) found that when supervisors felt their employer had not lived up to what the supervisor felt he or she had been promised, that is, supervisors felt their psychological contracts had been violated, their subordinates judged them to be more abusive.
The reactivity inherent in these trickle-down models comes from the theory of organizational justice, from which Tepper's (2000) theory of abusive supervision was derived. Organizational justice comprises a large body of work that examines and predicts organizational factors and individual attitudes and behavior surrounding appraisals of unfairness and mistreatment in organizations (Greenberg, 1987). In his seminal work on abusive supervision, Tepper (2000) suggested that interactional justice, that is, the interpersonal dimension of organizational fairness, was particularly relevant to abusive supervision. Contextually rooted in early balance theories (e.g., Festinger, 1957;Heider, 1958), the idea is that fair relationships are those that are characterized by equitable inputs and outputs, such as a balanced ratio between a subordinate's work efforts and the reciprocal consideration and respect she receives from her supervisor. However, when a supervisor perceives injustice, that is unfair treatment in his or her work environment, this destroys the balance of the relationship. Disconcerting feelings of imbalance such as anxiety and other negative states may prompt resultant negative attitudes and behaviors (Greenberg, 1987) such as low job satisfaction and a desire to leave the organization (Tepper, 2000). While the organizational justice literature has done a good job of specifying the types of attitudes and behaviors that result from unfairness perceptions, less attention has been given to models such as ours that examine specific moods, affect, or emotions associated with interpersonal injustice (Weiss, Suckow, & Cropanzano, 1999).
As two exceptions, Tepper et al. (2006) and Restubog et al. (2011), together with their colleagues, supported models of abusive supervision that incorporated affective mechanisms to explain behavior. In Tepper and colleagues' (2006) study, supervisors' depression, an indicator of emotional distress, was the linking mechanism between their perceptions of injustice and subordinates' reports of abusive behavior. Moreover, this mediated relationship was stronger when subordinates were higher in trait negative affectivity. In essence, they explained that the depressive symptoms evoked by supervisors' feelings of injustice prompted the need to engage in the hostile behavior of abusive supervision, and that subordinates higher in trait NA seemed to be targeted as victims. In Restubog and colleagues' (2011) work, subordinates' psychological distress was the affective linkage between their perceptions of abusive supervision and their undermining of a spouse (criticizing, insulting them) at home. Our research integrates and extends these two studies in two ways. 1) We examine how shorter term, state negative affect on the part of both supervisors and subordinates, is associated with supervisor injustice perceptions and subordinate abusive supervision perceptions. In so doing, we add state negative affect to the short list of antecedents to abusive supervision just beginning to accumulate in the research (Tepper, 2007). Further, we offer reasons why employees (both subordinates and supervisors) experience agitation and distress, i.e., negative affect, at work-a question of merit to practitioners because subordinates high in negative affect tend to hold negative attitudes and engage in dysfunctional behavior (Brief & Weiss, 2002), and supervisors high  Fig. 1. Proposed model of injustice, abusive supervision, and negative affect. NA = negative affect. L indicates that the variable was reported by supervisors/ leaders; S indicates that the variable was reported by subordinates; F indicates that the variable was reported by subordinates' family members.