«UNDERSTANDING EMPLOYEES’ BEHAVIORAL REACTIONS TO AGGRESSION IN ORGANIZATIONS by MARIE S. MITCHELL B.S., George Mason University, 1993 M.A.H.R., ...»
Coworker aggressive modeling did not significantly influence reactions of organization or customer displaced aggression to perceived supervisor aggression. At first, it may seem that these results reflect individuals’ reactions to negative social exchange. Because social exchange theory suggests negative reciprocity norms promote destructive behavior (i.e., the negative norm of reciprocity, Gouldner, 1960), it may appear individuals victimized by an aggressive supervisor, who are also exposed to aggressive coworkers, may have reacted against coworkers as a form of retaliation. However, in the analysis, I control for coworker aggression; therefore, the influence on coworker displaced aggression appears to be something more than reciprocity.
Alternatively, according to social psychology research, they may have been displacing aggression against targets that appeared similar to the source of the aggression (i.e., the aggressive supervisor). A recent meta-analysis in social psychology demonstrated that individuals were more likely to displace aggression against targets that were more similar to the source of the harm (see Marcus-Newhall et al., 2000). Therefore, individuals may have been more likely to displaced aggression against coworkers when the coworkers were aggressive because they perceived the coworkers as similar to the aggressive supervisor.
Further, coworker aggressive modeling may also be viewed as an individual perception of the organization’s work climate. Organizational climate involves shared perceptions of formal and informal work norms, policies and practices (Reichers & Schneider, 1990). Thus, it is possible that the combination of an aggressive supervisor with aggressive coworkers suggests the climate of the organization is aggressive (i.e., an aggressive climate). Nevertheless, had aggressive modeling represented an aggressive climate of the organization, one would expect it to have influenced all reactions to perceived supervisor aggression. This was not the case.
Further, in general, climate perceptions are “shared” among organizational members, meaning the perceptions are collectively understood by the majority as accepted norms and practices.
Given the sample and measures used in this study, it was not possible to aggregate responses to assess “shared perceptions” of aggression. Nevertheless, understanding the influence of climate on reactions to perceived aggression should be considered in future research.
Research by Aquino and colleagues (Aquino & Douglas, 2003; Aquino et al., 2001; in press) suggests individuals’ absolute hierarchical status influences the types of reactions individuals engage in as a consequence of perceived aggression. Aquino and colleagues’ research suggests that individuals in low-status positions are more likely to respond to aggression with aggression and that those in high-status positions do not engage in behaviors that can be
viewed by others as unprofessional. The results of this study, however, suggest the opposite:
individuals in high-status positions were more likely to respond to supervisor aggression with aggression than individuals in low-status positions. Specifically, when absolute hierarchical status was high, the relationships between perceived supervisor aggression and organization and customer displaced aggression were positive. In contrast, when absolute hierarchical status was low, the relationship with organization displaced aggression was negative and the relationship with customer displaced aggression was not significant. Overall, the results suggest that highstatus employees were more likely to displace aggression toward the organization and customers than low-status employees, and low-status employees were far less likely to displace aggression (particularly against the organization).
Because high-status individuals were more likely to react with aggression against the organization and customers, the findings of this study suggest that perhaps high-status employees were more likely to displace aggression on targets when it can go unnoticed. In particular, when individuals are in high-status positions, aggression against the organization or customers may not be reported or checked, unlike acts of aggression against organizational members (e.g., aggressing against one’s supervisor or coworkers). In contrast, aggressing against the organization or customers might be particularly difficult for lower-status employees who are more closely supervised. Essentially then, higher-status employees may aggress against the organization and customers “because they can,” unlike lower-status employees. In short, displacing aggression toward the organization and customers may allow high-status victims to vent hostilities about an aggressive supervisor without negatively impugning their status.
These contradictory findings may be explained by comparing the studies by Aquino and colleagues (Aquino & Douglas, 2003; Aquino et al., 2001) to this study. First, the samples used in the Aquino studies differ from the sample used in this study. In Aquino and colleagues’ studies, they used more bureaucratic and government-oriented samples. For example, Aquino and Douglas (2003) used three organizations: a transportation company, a school system, and a municipality. In Aquino et al. (2001), they used government employees. Due to the bureaucratic nature of these samples, perhaps it is easier for lower-status employees to engage in aggressive behaviors than employees in higher-status positions, who are more closely watched by the general public. As a consequence, higher-status employees might be particularly careful not to engage in aggressive behaviors (particularly against the organization or customers). In contrast, this study used a random sample of individuals called for jury duty, who represented a widerange of positions and job types. Thus, the broader range of organizations and job suggests individuals in higher-status positions might be less constrained and more willing to engage in aggression toward the organization and customers.
A last difference in the Aquino studies from this study involves the overall design of the studies themselves. Aquino et al. (2001) and Aquino et al. (in press), used a critical incident technique. Specifically, they asked respondents to reflect back on a time when someone offended them in their current organization, and then asked them about revenge and reconciliation reactions to that particular offense. Asking respondents to recall past events (and reactions to those events) may promote demand characteristics and self-enhancement bias (Aronson, Ellsworth, Carlsmith, & Gonzales, 1990). In this study, respondents were not asked about a specific aggressive event or specific reactions to that aggressive event. Rather, they were asked about the aggressive treatment by others (from different sources), as well as various behaviors that they may have engaged in throughout the course of a year. None of the questions specifically asked about revenge or retaliation, unlike the Aquino et al. studies, making demand characteristics and self-enhancement bias less likely. Therefore, the different findings from the Aquino studies and this study may also be due to asking respondents about a specific event rather than behaviors experienced and engaged in over the course of a year.
The Influence of Individual Factors on Reactions to Perceived Supervisor Aggression The findings for trait anger suggest trait-angry individuals were more likely to respond to supervisor aggression will retaliatory aggression. As predicted, the positive relationship between perceived supervisor aggression and retaliatory aggression was stronger when trait anger was high than low. The results are consistent with the argument that because high trait-angry individuals have the tendency to perceive a variety of situations as anger-provoking and also react intensely to provocation, they are more likely to react aggressively. However, trait anger did not moderate the relationship of perceived supervisor aggression and displaced aggression.
Previous research suggests that trait anger enhances cathartic expressions of anger (see Perrowé & Spector, 2002, for a review). Yet, the results of this study suggest that trait anger heightened targeted reactions against the provoking source of aggression rather than aggressive reactions in general. Previous research has provided results consistent with these findings. For example, Martinko and Douglas (2001) found that when respondents believed the harmdoer purposefully and wrongfully offended them, high trait anger strengthened revenge. Further, in his review of trait anger, Deffenbacher (1992) found that individuals high in trait anger experienced more intense negative reactions to personal attacks (i.e., verbal threats, intimidation). Consequently, Deffenbacher Oetting et al. (1997) investigated these findings further and found that, although trait anger increases the intensity of anger over various situations, high trait-angry individuals became particularly antagonistic (both verbally and physically) against the provoker of personal attacks. Therefore, perhaps high trait anger also heightens motivations to get even with the transgressor.
The results in the study did not provide support for moderating effects of need for social approval on the relationship between perceived supervisor aggression and non-aggressive reactions (i.e., constructive problem-solving or withdrawal). However, some support was found for its influence on the relationship between perceived supervisor aggression and aggressive reactions. Specifically, the relationship of perceived supervisor aggression and organization displaced aggression was negative when need for social approval was high. The relationship was not significant when need for social approval was low. Therefore, individuals with a high need for social approval were less likely to displace aggression on the organization when dealing with an aggressive supervisor compared to individuals with a low need for social approval. Further, although the relationship between perceived supervisor aggression and customer displaced aggression was not significant when need for social approval was high, the relationship was positive when need for social approval was low. For individuals low in need for social approval, as perceived supervisor aggression increased, customer displaced aggression increased. Overall, the pattern of results supports the notion that individuals who have a high need for social approval avoid threatening and conflictive situations (like aggressing against others). Thus, high need for social approval individuals may be less likely to displace aggression (particularly against the organization) in response to an aggressive supervisor, whereas low need for social approval individuals may be more likely to displace aggression (particularly against customers) in response to an aggressive supervisor.
Because research has shown that high need for social approval-individuals are more likely to engage in help-seeking and approval-seeking activities, it is surprising that need for social approval did not moderate the relationship between perceived supervisor aggression and constructive problem-solving. It could be that constructive problem-solving behaviors might instigate discussions about a situation. Research on need for social approval has found that those high in this need retreat from situations or interactions that imply negative affiliation (e.g., Exline, 1963; Mehrabian & Ksionzky, 1974; Terhune, 1968). Therefore, individuals high in need for social approval may feel that discussing the problem might reflect negatively on them, particularly for discussions directly with the aggressive supervisor, and therefore they may not engage in constructive problem-solving.
It is important to highlight that none of the variables moderated the relationship between perceived supervisor aggression and withdrawal. Rather, it seemed as if variables that contribute to more destructive work environments directly influenced withdrawal behaviors. In particular, perceived coworker aggression and coworker aggressive modeling were significantly and directly related to withdrawal. These results may be explained through principles of reactance and learned helplessness (deCharms, 1968; Seligman, 1975; Wortman & Brehm, 1975). When individuals are dealing with threatening situations (like perceived supervisor aggression), psychological reactance occurs (Brehm, 1966). Generally speaking, reactance involves the victim trying to change the situation effectively; however, if change is not possible, individuals then feel helpless (Wortman & Brehm, 1975). As a consequence, Rothbaum, Weisz, and Synder (1982) argue that helplessness promotes individuals to modify their own behavior in order to regain a sense of personal control (e.g., withdrawal). Perhaps victims of supervisor aggression, who are also exposed to coworker aggression—meaning they are victim to coworker aggression and coworkers actively act out aggressively at work (modeling)—feel helpless, and the only thing they can do to cope with their circumstances is to withdraw.
Limitations As with all studies, this research has some limitations. First, because cross-sectional data was used, inferences of causality cannot be made. Although it is argued that supervisor aggression elicits different types of reactions from employees, employees’ reactions (particularly aggression and withdrawal) may elicit what is perceived by the employee as aggressive behavior by the supervisor. Therefore, future research should examine the causal dynamics of reactions to aggression within an experimental setting or through a longitudinal design.
Second, the use of single-source data raises concerns about common method variance.