Knowledge Sharing,Abusive Supervision,and Support

Knowledge Sharing and Abusive Supervision

Given the discretionary nature of knowledge sharing, employees may decide whether to engage in knowledge sharing behaviors depending on how supervisors treat them. As supervisors control important decisions around promotions, incentives, and rewards, supervisors’ behaviors are likely to be influential in determining knowledge sharing behaviors (Wang & Noe, 2010). Previous studies have demonstrated that supervisor support is positively associated with the development of a knowledge sharing culture and a willingness by employees to share knowledge (Connelly & Kelloway, 2003; Lin, 2007), while abusive supervision brings negative attitudes and low OCB (Aryee et al., 2007; Aryee et al., 2008; Harris et al., 2007; Tepper, 2000, 2007; Tepper et al., 2006; Tepper et al., 2001; Zellars et al., 2002). These findings imply that abusive supervision has a negative effect on employee knowledge sharing behaviors, which are a type of discretionary behavior.

Social exchange theory can help explain the link between abusive supervision and knowledge sharing (Blau, 1964). According to social exchange theory and the norm of reciprocity (Blau, 1964; Cropanzano & Mitchell, 2005), when one person does another person a favor, there is an expectation of some future return for that favor, although the time and the form of the return may not be clear (Gouldner, 1960). According to social exchange theory (Blau, 1964; Cropanzano & Mitchell, 2005), individuals may change their attitudes or behaviors, depending on how they perceive they are being treated or on the need for reciprocity. Abusive supervisors are those who do not treat employees well, and instead publicly criticize them, shout at them, intimidate them, use aggressive eye contact, withhold needed information, ridicule their subordinates, and/or give them the silent treatment (Keashly, 1998; Tepper, 2000). Abused employees are likely to feel victimized, as compared with non-abused employees. These practices create difficulties in advancing employees’ discretionary behaviors such as knowledge sharing (Zellars et al., 2002). Given that knowledge is a unique resource held by each individual, employees may not share their knowledge when they cannot expect a fair evaluation of, or recognition for, their knowledge sharing behaviors. In this context, abused employees are likely to respond negatively, by decreasing their discretionary behaviors in accordance with social exchange theory (Gregory, Osmonbekov, Gregory, Albritton, & Carr, 2013; Zellars et al., 2002). In addition, reactance theory suggests that individuals who experience high stress levels tend to restore or increase their level of control by exercising discretion (Zellars et al., 2002). Similarly, abused employees may readily decide to decrease their knowledge sharing efforts, as a means of restoring the control they have lost through abusive supervision. As knowledge providers may lose their unique resources (A. Cabrera & Cabrera, 2002), abused employees may not wish to share these resources, when they are not certain they will receive fair or positive treatment from their supervisors. Therefore, we predict the following:

The Moderating Effects of Support

Although supervisors play a critical role, they are not the only exchange partners in the workplace (Chiaburu & Harrison, 2008; Chiaburu et al., 2013; Rhoades & Eisenberger, 2002). In fact, the organization and one’s coworkers are also important actors who can provide adequate resources for knowledge sharing, particularly when one’s supervisor is abusive (cf. Wu & Hu, 2009). A multifoci perspective of social exchange suggests that an employee’s positive perception of their organization or their coworker may help reduce the effects of a negative supervisor’s behavior, as individuals may form supportive relationships with multiple parties (Lavelle et al., 2007). Furthermore, the cross-domain buffering effect (e.g., Duffy, Ganster, & Pagon, 2002) suggests that support from one source (e.g., organizational or coworker support) may buffer the negative effects of another source (e.g., abusive supervision). Thus, we propose that individuals who receive support from their organization or their coworkers are less likely to reduce their knowledge sharing, even if they receive unfair or abusive treatment from their supervisors, as knowledge sharing might be beneficial for the organization or their coworkers.

Although individuals under abusive supervision do not receive positive treatment from their supervisors (Tepper, 2000), they are likely to feel an obligation to their organization when they perceive a high level of organizational support (Eisenberger, Huntington, Hutchison, & Sowa, 1986). Organizational support is defined as the “general belief that their work organization values their contributions and cares about their well-being” (Rhoades & Eisenberger, 2002, p. 698). Depending on how organizations demonstrate their concerns for employees’ contributions and well-being, the employees develop positive or negative general perceptions toward organizations (Eisenberger, Stinglhamber, Vandenberghe, Sucharski, & Rhoades, 2002). Employees who perceive high organizational support tend to believe that the organization values their contributions, and provides the necessary resources when the resources are requested (Rhoades & Eisenberger, 2002). Accordingly, employees with high organizational support are predicted to fulfill their sociological needs and have increased feelings of obligation that have been otherwise lowered by abusive supervision, which can attenuate the negative effects of abusive supervision on individuals’ knowledge sharing. On the contrary, a low level of organizational support can further decrease feelings of obligation that have already been reduced by abusive supervision, which can intensify the negative effects of abusive supervision on individuals’ knowledge sharing behaviors.

According to the conservation of resources theory (COR; Hobfoll, 1989), employees are motivated to protect their current resources and acquire new resources (Hobfoll, 1989). Resources “either are centrally valued in their own right (e.g., self-esteem, close attachments, health, and inner peace), or act as a means to obtain centrally valued ends (e.g., money, social support, and credit)” (Hobfoll, 2002, p. 307). COR theory has two key tenets, which consist of resource conservation and resource acquisition (Ng & Feldman, 2012). The two tenets of COR theory are based on the idea that losses are more salient than gains, which suggests that individuals feel the experience of loss disproportionately more than they feel the experience of gain (Halbesleben, Neveu, Paustian-Underdahl, & Westman, 2014). This may imply that employees attach greater value to resource gains when they experience resource losses (Vinokur & Schul, 2002). Applying COR theory to our study, abused employees who experience reduced resources are more likely to perceive organizational support as a resource (Halbesleben et al., 2014). Although employees subject to abusive supervision do not receive the resources they need from their supervisor, they may receive the resources required for their work such as funding, technology, and physical assistance from the organization (Eisenberger et al., 1986). Furthermore, it is likely that abused employees may attach more value to the organizational support they receive in the context of the resources they have lost due to their abusive supervisors (cf. Vinokur & Schul, 2002). As a result, they may not reduce their knowledge sharing behaviors when there is a high level of organizational support, in accordance with COR theory (Hobfoll, 1989).

Similar logics can be applied to the moderating effects of coworker support on the relationship between abusive supervision and knowledge sharing relationships. Coworker support is defined as “the beliefs employees hold regarding the extent to which coworkers provide assistance” (Ng & Sorensen, 2008, p. 244). While organizational support is general in nature, coworker support may provide specific types of support, such as instrumental or emotional support (Chiaburu & Harrison, 2008; Kim & Yun, 2015). As one of the critical actors in the workplace, coworkers may provide instrumental or informational support by offering specific knowledge or information that are useful to perform tasks (Kim & Yun, 2015). Also, they may provide emotional support by listening to the employees’ problems and being friendly (Chiaburu & Harrison, 2008). This study focuses more on the emotional aspect of coworker support, as emotional support is likely to be helpful in reducing stress caused by abusive supervision.

Previous studies have noted that emotional parts of social support are generally beneficial to buffer the negative effects of stress on outcomes (Dormann & Zapf, 1999; Sloan, 2012). Consistent with the stress-buffering hypothesis, Sloan (2012) demonstrated that coworker support, specifically, emotional coworker support buffered the relationship between unfair treatment by a supervisor and job satisfaction, as well as the relationship between unfair treatment by a supervisor and psychological distress. Although individuals may perceive negative treatment from their abusive supervisors, they may feel a sense of obligation toward their coworkers when they have supportive coworkers who show concern and empathy for their difficulties. As knowledge sharing is beneficial not only for organizations but also for coworkers, individuals who receive high coworker support may limit their reduction of knowledge sharing as a form of positive reciprocity for their supportive coworkers, in line with social exchange theory.

Furthermore, these employees may view coworker support as a resource according to the COR theory (Hobfoll, 1989). Employees who have lost their resources due to an abusive supervisor may perceive that their resources could be replenished by coworker support. In particular, coworkers are likely to understand the difficulties facing abused employees and be able to provide psychological support (Wu & Hu, 2009). In this context, abused employees’ depleted resources could be recovered through psychological resources offered by coworkers that may mitigate the negative effects of abusive supervision on knowledge sharing. Overall, the negative effects of abusive supervision on knowledge sharing are likely to be weakened when the level of coworker support is high, based on social exchange and the conservation of resources perspectives. Thus, we predict the following:

Sample and Procedure

Data were collected using questionnaires distributed to several organizations located in South Korea. The industry sectors included in this sample were the manufacturing, finance, and service sectors. Most of participants in this study held positions in research and development, management, and the like, who need to share their knowledge and know-how to better perform their jobs. Furthermore, the companies we surveyed usually had Intranet-based knowledge sharing databases in common. Employees in these companies were encouraged to report their knowledge or information through the online knowledge sharing database. Employees had ready access to the database, and could update their information and acquire new knowledge from others at any time. For examples, employees may upload useful guidance or working tips that they have learned from their tasks or each project. They may share new ideas or solutions for performing certain tasks. Also, through this knowledge system, employees can suggest comments or acquire feedback for each other.

Surveys were prepared and distributed to all 250 supervisor–subordinate dyads, in sealed survey packets. Every unique supervisor–subordinate dyad received a survey packet, which included supervisor and subordinate surveys. As supervisors in each company often have more than one subordinate, we had to make selection criteria to maintain a subordinate–supervisor ratio of 1:1. When making a list of supervisor and subordinate pairs, our contacts were informed that supervisor–subordinate dyads had to meet the following criteria. First, supervisors had to be in positions where they could closely observe and accurately evaluate their subordinates’ behaviors. That is, the distance between supervisors and subordinates should not be large in order for them to be paired with each other. Second, supervisors and subordinates had to work together regularly, which meant that they should have regular interactions. If there was a case that more than one subordinate meet these criteria, our contacts selected one subordinate randomly. Thus, we predicted that supervisors and subordinates included in this research were in a good position to evaluate subordinates’ knowledge sharing and observe supervisors’ abusive behaviors.

Participants were asked to enclose their completed surveys in the envelopes provided, and to seal them, to maintain the confidentiality of their responses. All the supervisors and subordinates in this study participated voluntarily and were assured of anonymity. As a single supervisor completed a questionnaire for only one subordinate, this study had no nesting issues. From 250 dyadic samples, 178 supervisor–subordinate dyads were returned, giving a response rate of 71.2%. However, some questionnaires could not be used for the analyses, either because only one partner participated or because the answers were not completed properly. As a result, 128 supervisor–subordinate dyads were used for the final analysis. In terms of demographics, of the subordinates, 67.4% were male, and their average age was 32.23 years (SD = 5.51). Most participants were full-time employees (93%). Of the supervisors, 87.8% were male, and the average age was 43.33 years (SD = 5.63). The most frequently mentioned level of education for both subordinates and supervisors was university (78% for subordinates and 71.2% for supervisors). Subordinates had worked with their supervisors on average for 2.54 years (SD = 2.81).

Measures

The English language questionnaires were translated into Korean, using the conventional method of back translation (Brislin, 1980). The translation and back translation of the questionnaires were completed independently by two Korean bilingual academics. All the variables were measured, using established measures. The employees were asked to provide information with regard to their perceptions of abusive supervision, organizational support, and coworker support. To reduce concerns about common method bias, their immediate supervisors were asked to provide their evaluations of the subordinates’ levels of knowledge sharing. All the items were measured on a seven-point Likert-type scale (ranging from 1 = strongly disagree, to 7 = strongly agree).

Analytical Procedures

We conducted hierarchical regression analyses to test our hypotheses. First of all, to test the hypotheses of the main and moderating effects between abusive supervision and knowledge sharing, we included control variables in Step 1. Next, in Step 2, we included abusive supervision, followed by organizational support and coworker support in Step 3. Finally, in Step 4, the two interaction terms (Abusive supervision × Organizational support, Abusive supervision × Coworker support) were entered. We conducted a test for multicollinearity and reviewed the Variance Inflation Factor (VIF) levels. All VIF values were far less than 10, which is the upper-limit cutoff for VIF scores. These results suggest that multicollinearity is not a serious concern in the present study (Chatterjee & Price, 1991).

Theoretical Implications

This study offers four major theoretical implications. First, our study contributes to the knowledge sharing literature, demonstrating the important influence of supervisors on individuals’ knowledge sharing behaviors. Despite the critical role of supervisors, there is a paucity of research examining how a supervisor’s behavior may relate to knowledge sharing by individuals (Srivastava et al., 2006). Applying social exchange theory (Blau, 1964), this study shows the negative effects of abusive supervision on knowledge sharing. As our results demonstrate, how supervisors treat their subordinates is critical to promoting or inhibiting individuals’ knowledge sharing. As Srivastava et al. (2006) found, empowering leadership positively relates to knowledge sharing at the team level. It might be worthwhile to examine how supervisors’ other behaviors relate to knowledge sharing in future research. For example, a supervisor’s cooperative or ethical behavior might facilitate knowledge sharing by individuals by reducing the risks of knowledge sharing.

Second, our study contributes to the abusive supervision literature by demonstrating the negative effects of abusive supervision on discretionary behaviors, such as knowledge sharing. Most studies in the abusive supervision literature have investigated the relationship between abusive supervision, employees’ job attitudes, and stress-related variables (Harris et al., 2007). This study contributes to that literature by focusing on an element of discretionary behaviors—knowledge sharing behavior—that is critical for organizational effectiveness. Extending from this work, future research should investigate the mechanisms between abusive supervision and knowledge sharing. In addition, the findings of this study suggest that abusive supervision may contaminate organizational climates (Tepper, Duffy, Hoobler, & Ensley, 2004). Although supportive coworkers are present in a workplace, abused employees may not fully enjoy the benefits of their coworkers’ additional resources in the presence of abusive supervision. Instead, they may misinterpret their coworkers’ intentions or believe coworker support is an instrumental tactic used to enhance their self-image. Future research should examine whether and how abused employees attribute the support of coworkers as being self-serving rather than as being well-intentioned.

Third, our study gives insights to the support literature by examining the effects of receiving support from different sources in a workplace. Our study found that only organizational support, not coworker support, has a moderating effect on the relationship between abusive supervision and knowledge sharing. Previous studies have examined the interaction effects of the different types of support on outcomes (Duffy et al., 2002). For example, the negative relationship between organizational support and turnover was strengthened in the context of low supervisor support, which may imply that the role of organizational support becomes more critical when there is lack of supervisor support (Maertz, Griffeth, Campbell, & Allen, 2007). Duffy and colleagues (2002) investigated the interaction effects of social undermining and social support in the workplace. Their results showed that coworker support failed to buffer the relationship between supervisor undermining and counterproductive behaviors, both active and passive (Duffy et al., 2002). Our results extend this line of research by examining the moderating effects of organizational support and coworker support in the relationship between abusive supervision and knowledge sharing. As our study shows, not all types of support are necessarily beneficial, but organizational support is an effective way of alleviating the negative impacts of abusive supervision on knowledge sharing.

Our study does not support a moderating effect of coworker support. There are several plausible reasons for this. First, the effect of coworker support might be weaker than that of supervisor support because of power or value differences between coworkers and supervisors (Chiaburu et al., 2013; Ng & Sorensen, 2008). Given the critical role of the supervisor, coworker support might not be sufficiently influential to decrease the negative effects of abusive supervision (Kim & Yun, 2015; Ng & Sorensen, 2008). The literature on coworker exchange relationships also indicates the possible influence of supervisor–subordinate relations on exchange quality between coworkers (Sherony & Green, 2002). In this respect, it is possible that abused employees will experience tension or conflict with other non-abused coworkers, which may prevent them from getting the full benefits of coworker support. In addition, our finding might be interpreted in a cultural context. Unlike Western cultures, where power distance is low, South Korea has a high power distance culture (Hofstede, 1980). In this context, the ability of coworker support to buffer the effects of supervisors’ abusive behaviors may be much weaker.

Furthermore, the impact of coworker influence on knowledge sharing behaviors could be stronger for instrumental coworker support rather than for emotional coworker support. Separately from organizational support, we measured the emotional aspect of coworker support, such as listening behaviors and showing concerns to employees, rather than directly helping to solve problems (Wu & Hu, 2009). In other words, coworker instrumental support could be a very powerful tool, which should be investigated in future research. In addition, it may be possible that abused employees who interact with their coworkers and discuss their stressful issues may feel even more stress; these interactions may unintentionally lead employees to concentrate on their problems and lower their self-image (Beehr, Bowling, & Bennett, 2010; Wu & Hu, 2009). Recent research found a “reverse buffering effect” from coworker support such that the negative relationship between abusive supervision and emotional exhaustion was strengthened with a higher level of coworker support (Wu & Hu, 2009). Thus, there is a possibility that the potential negative and positive effects of coworker support may offset each other, leading to our non-significant results here.

Another possible explanation is that coworker support may be seen as political, rather than altruistic (Ng & Sorensen, 2008). As individuals who provide support to others may be viewed as more competent and valuable, coworkers could use support as a tactic for increasing their image or status in the workplace (Tepper et al., 2004). Furthermore, given a similar hierarchical position, employees may feel low self-esteem or that they are not able to utilize coworker support (Tepper et al., 2004). Due to these possibilities, the effects of coworker support might not be related to favorable outcomes (Ng & Sorensen, 2008). Different from organizational support, employees may interpret coworker support as a political behavior or impression management (Ng & Sorensen, 2008). In particular, in the context of abusive supervision, climates are likely to be competitive and hostile (Tepper et al., 2004; Tepper, Moss, & Duffy, 2011). In this context, there is a greater chance that employees may not regard coworker support as a valuable resource but instead view it as instrumental or political, which may reduce the buffering effect of abusive supervision on knowledge sharing. In this situation, abused employees might face the so-called “receiver’s dilemma,” where it is not easy to form accurate impressions of the coworker or interpret the true intentions behind their support (Gurevitch, 1985). Therefore, unlike organizational support, abused employees might be suspicious of coworkers’ intentions for offering support, which may decrease the role of coworker support in the relationship between abusive supervision and knowledge sharing. Future research may need to examine this issue further in different research settings.

Finally, our findings fall in line with social exchange theory. According to social exchange theory, individuals may limit their efforts at improving discretionary behaviors, depending on their perceived benefits, or on their treatment by their exchange partners (Cropanzano & Mitchell, 2005). Our study is meaningful in that it includes all major exchange partners in a workplace—organizations, supervisors, and coworkers—in one framework, investigates how negative treatment from one exchanging partner may affect individuals’ knowledge sharing, and investigates how favorable treatment from another exchanging partner may buffer this negative effect. As our results indicate, individuals decrease their level of knowledge sharing when they are subject to abusive treatments from their supervisors, but the negative effect of a supervisor’s detrimental behavior could be mitigated by support from the organization. Extending from our research, it will be worthwhile to use the social exchange perspective to examine other discretionary behaviors in future research.

Limitations and Future Research

This study has limitations. First, our cross-sectional design cannot infer causality; reverse causality might exist. For example, employees may receive abusive treatments owing to their low participation in knowledge sharing. In addition, longitudinal research design should examine how individuals may change their perceptions of organizational or coworker support after they experience abusive supervision, as there is a chance that abusive supervision might influence followers’ perceptions of such support. Second, common method bias might be a concern. We have tried to reduce this potential problem by collecting data from two different sources—the employees and their supervisors—and used the supervisors’ data on knowledge sharing to increase the objectivity of data. Furthermore, common method bias is less likely to explain our significant interaction effects (Evans, 1985). Therefore, common method bias may not affect our major findings. However, it would be valuable to measure knowledge sharing from diverse sources—including supervisor, self, and a coworker—and compare how these data sources may influence results differently in future research. Third, the sample in our study was male dominated and highly educated. We have tried to minimize this issue by controlling for gender and education in our model. Nevertheless, it might be beneficial to use a more balanced sample in future research. Finally, our study included only a limited number of variables, as we intended to conduct a focused rather than a comprehensive study. Future research may need to include more variables to deepen the understanding of knowledge sharing and abusive supervision. For example, justice, trust, and psychological contract violation might be investigated as mechanisms pertinent to abusive supervision and knowledge sharing in future research.

Managerial Implications

Our study also offers practical implications. First, recognizing the negative consequences of abusive supervision, organizations should make efforts to prevent abusive supervision in workplaces. For example, organizations should arrange coaching or constructive leadership training that could be helpful for preventing this dark side of a supervisor’s behavior from manifesting. Second, organizations should pay more attention to finding ways to decrease the negative consequences of abusive supervision on job-related performance. As it is often not easy or realistic to remove all abusive supervisors in organizations, providing social contexts that can decrease the negative effects of abusive supervision could be more practical remedies. As our study suggests, organizational support could be one of the effective solutions for reducing the harmful effects of abusive supervision. Scholars suggest that favorable work experiences and job conditions—including training, job security, reward expectations, and high autonomy—are effective ways for cultivating employees’ high organizational support (Eisenberger & Stinglhamber, 2011; Shanock & Eisenberger, 2006). Previous studies have noted that organizational justice and supportive human resource policies that provide feelings of inclusion and recognition from upper management are critical antecedents of organizational support (Wayne, Shore, Bommer, & Tetrick, 2002). Thus, companies can design human resource practices and evaluation systems that increase perceptions of fairness, recognition, or inclusion. For example, a 360° feedback system or mentoring programs might increase perceptions of organizational support, by enhancing fairness and personal growth.