External Whistleblowers’ Experiences of Workplace Bullying by Superiors and Colleagues

The purpose of this study was to investigate external whistleblowers’ experiences of workplace bullying by superiors and colleagues, and to analyze how the bullying was influenced by factors such as the support they received from government or NGOs, and whether colleagues understood the reasons for the whistleblower’s actions. For bullying by colleagues, we also examined to what extent this was influenced by superiors’ behavior towards the whistleblower. We reviewed the relevant literature on workplace bullying and whistleblowers’ experiences of negative or retaliatory actions and developed three hypotheses, which we tested using data gathered from Korean external whistleblowers. Results revealed that external whistleblowers experienced work-related bullying by superiors and social relation-related and person-related bullying by colleagues more frequently, and found it more distressing, than other types of workplace bullying. Superiors’ bullying was a dominant factor affecting bullying by colleagues. Colleagues’ understanding of the reason for the whistleblower’s actions was significant in reducing bullying frequency while support from government and NGOs was not significant in reducing it. Based on these findings, practical implications are discussed.


Introduction
Many studies have documented the extent to which whistleblowers suffer from hostile behavior (retaliation) in the workplace after disclosing wrongdoing (e.g., Peters et al. 2011;Jackson et al. 2010;Rothschild and Miethe 1999), with workplace bullying potentially a systematic and enduring form of retaliation. Workplace bullying has a detrimental impact on the health and well-being of victims as well as the work environment (Branch et al. 2013;Einarsen et al. 2009;Lewis 2006;Vartia 2001;Einarsen and Raknes 1997;Vega and Comer 2005), and  demonstrated that whistleblowers report more bullying than non-whistleblowers. Bullying behavior that whistleblowers suffer may be unique in being retaliation-related, that is at the opposite end of 'predatory bullying' directed at employees that did not do anything to deserve it 1 (Einarsen 1999). Perpetrators of bullying may be superiors (vertical or downwards workplace bullying, Vandekerckhove and Commers 2003) and/ or colleagues (lateral workplace bullying). There is strong evidence that external whistleblowers (i.e., those who blow the whistle to recipients outside of the organization) suffer the greatest level of retaliation (Dworkin and Baucus 1998), but the extent to which this retaliation takes the form of workplace bullying has not been widely studied. Most largescale studies of the experience of whistleblowers focus on employees, and external whistleblowers are less likely to be captured by such studies as they may be more likely to leave (or get fired) than internal whistleblowers. The present study is therefore unique in presenting data on workplace bullying from a substantial survey of external whistleblowers who remained with their employer. The findings contribute to our understanding of the damage suffered by whistleblowers and to the shaping of programs and policies for protecting whistleblowers, by exploring the bullying behavior experienced by external whistleblowers and the impact of the internal and external support for whistleblowers on bullying. Leymann (1996) defines workplace bullying as a systematic hostile and unethical form of communication performed by one or more individuals mainly directed at one person, who due to the bullying process is pushed to a helpless and defenseless position, often resulting in expulsion from the workplace. In his model of a typical bullying process, superiors may adopt the view of an employee which has previously been developed by employees (stage 3, personal management, Leymann 1996, p. 171); by contrast in a study of retaliation after whistleblowing, it was documented how employees may also adopt the image portrayed by management (Bjørkelo et al. 2008). Although many researchers have examined the extent to which whistleblowers are retaliated against at work (e.g., Jackson et al. 2010;Rothschild and Miethe 1999;Jos et al. 1989), few have gathered data on workplace bullying of external whistleblowers, who may be more likely no longer to be employed where they reported wrongdoing or perhaps even no longer able to work. According to a model of the predictors of retaliation (Miceli et al. 2008, p. 102), variables of influence are (a) the characteristics of the individual, (b) the situation (including the support of others), and (c) the group, organization, and society. The purpose of this study was to investigate (b) and (c) in relation to external whistleblowers' experiences of workplace bullying. Bjørkelo (2013) stated that although there is a relationship between whistleblowing and workplace bullying, whether this adheres to the nature of the reaction or the source of the bullying behaviors still remains unclear.
The following questions were addressed: (1) To what extent do external whistleblowers experience workplace bullying by superiors and colleagues in terms of frequency and distress? (2) How does bullying directed at external whistleblowers by superiors differ from that by colleagues? (3) How much support do external whistleblowers perceive they get, and from whom, inside and outside of the organization while they suffer bullying in the workplace? (4) What is the impact of bullying by superiors on bullying by colleagues? (5) How much does colleagues' understanding of the reasons for the whistleblower's actions contribute to reducing bullying by colleagues? (6) How much does the support of government and NGOs contribute to reducing bullying by colleagues?

Whistleblowers and Workplace Bullying
Whistleblowing research has documented a variety of negative and retaliatory acts that whistleblowers experience as a result of their disclosures, including negative performance appraisals, inappropriate work assignments or transfer, unnecessary refusals and delays of holiday leave, and threats of physical harm, harassment, humiliation, or isolation (Peters et al. 2011;Jackson et al. 2010;McDonald and Ahern 2002;Mesmer-Magnus and Viswesvaran 2005;Miceli and Near 1989, 1994, 2013Near and Jensen 1983;Miceli 1986, 2008;Parmerlee et al. 1982;Rehg 1998;Rehg et al. 2008;Rothschild and Miethe 1999;Soeken and Soeken 1987). Jackson et al. (2010) found whistleblowers suffered an exceptionally severe breakdown in working relationships, including hostile responses, marginalization, and exclusion in most interactions with other employees in the workplace. Superiors rejected requests for changes in work hours, assigned work outside targeted employees' competency, and treated them with suspicion and hostility. Further, colleagues would not speak to or work with them, and made sarcastic remarks.
Internal whistleblowing is usually applied to describe the act when an employee employed by an organization reports wrongdoing to a complaint recipient within the same organization (a superior or Ombudsman). External whistleblowing, on the other hand, implies reporting to "someone external to the organization" (Near and Miceli 1986, p. 327). Studies have repeatedly shown how reporting to internal complaint recipients most typically precedes external reporting . Dworkin and Baucus (1998) reported that although external whistleblowers "are more effective in eliciting change" they suffer more "extensive retaliation than internal whistleblower" (p. 1296). Rothschild and Miethe (1999, p. 120) reported external whistleblowers experienced "10-15% points" more retaliation than internal whistleblowers. Some scholars have defined repeated negative exposures to such acts as workplace bullying Bjørkelo et al. 2008). Bjørkelo and Matthiesen (2011, p. 135) documented that whistleblowing and bullying are related, and there is preliminary support for interpreting whistleblowing as a risk factor for later exposure to bullying (Bjørkelo et al. 2009(Bjørkelo et al. , 2015. Although there is no single agreed-upon definition of bullying at the workplace (Saunders et al. 2007;Lewis 2006), consistent elements have emerged over time. These include repeated negative and hostile acts in the workplace, persistence, an asymmetrical power struggle between perpetrators and victims, the inability of victims to defend themselves or control the hostile situation, and psychological or physical harm (Bjørkelo 2013;Branch et al. 2013;Ortega et al. 2009;Rayner and Keashly 2005). These behaviors are directed at the same employee at work and occurring on a regular basis for a lengthy period with the effect of "humiliating, intimidating, frightening or punishing the target" (Einarsen et al. 2009, p. 25;Einarsen 1999), and can come from colleagues, supervisors, or management towards a targeted employee (Vandekerckhove and Commers 2003). In this study, we define workplace bullying after external whistleblowing as consisting of recurrent negative acts in relation to work, personal identity, and social relations where the whistleblower feels unable to defend him or herself. Studies of workplace bullying have identified various forms of bullying, such as direct verbal abuse, physical intimidation, and death threats as well as indirect aggression such as slander, gossiping, or spreading unfounded rumors (McGlynn and Richardson 2014;Nolfe et al. 2010;Einarsen et al. 2009;Saunders et al. 2007;Niedhammer et al. 2006;Matthiesen and Einarsen 2004). The harmful effects may be deep-seated emotional and psychological distress, mental disorders, and symptoms of post-traumatic stress, as well as low productivity, job dissatisfaction, and thoughts of job exit.

Types of Workplace Bullying
Earlier studies (e.g., Waschgler et al. 2013;Einarsen 1999) have classified workplace bullying in various ways: vertical bullying by superiors versus lateral or horizontal bullying by colleagues; work-related bullying versus person-related bullying. The forms and patterns of bullying significantly vary depending on the bully and the bullied. Bullying by superiors (i.e., downwards bullying, Vandekerckhove and Commers 2003) is typically formal and work-related as superiors bully a victim by providing unreasonable deadlines and unmanageable workload, and assigning them work below their level of competence or more menial or unpleasant tasks. Conversely, bullying behavior by colleagues occurs in the form of more informal, social relation-related or personrelated behavior. Spreading negative gossip and unfounded rumors or inappropriate jokes and sarcasm are more common in bullying by colleagues. Upwards bullying describes situations where supervisors may be on the receiving end of repeated negative acts (Branch et al. 2005;Tepper et al. 2007). Preliminary qualitative results seem to suggest that in some cases supervisors who report wrongdoing at work may suffer even more than employees (O'Connor 2017). These differences suggest whistleblowers may experience bullying behaviors from superiors and colleagues that differ in type and intensity.

Workplace Bullying by Superiors and Colleagues
Many previous studies (Einarsen 1999;Einarsen and Raknes 1997;Hoel et al. 2001;Soeken and Soeken 1987) found employees are bullied more by superiors than by colleagues, although bullying studies in Scandinavian countries identified the number of perpetrators was 'approximately equal' between superiors and colleagues (Hoel et al. 2001, p. 445). Howard et al. (2016) found that employees' perception of how aggressive and threatening the bullying behavior is varies depending on the type of perpetrator. Employees perceived bullying by superiors as more aggressive than bullying by colleagues in equivalent or lower positions, perhaps because superiors' bullying behaviors violates the psychological contract in a way colleagues' bullying may not (Parzefall and Salin 2010).

The Relationship Between Bullying by Superiors and Colleagues
Bullying by colleagues can be influenced by bullying by other colleagues as well as bullying by superiors (Lutgen-Sandvik et al. 2007). In a whistleblowing situation, superiors' bullying may potentially initiate other people's bullying behaviors towards whistleblowers, as their role and asymmetrical position of power "allows" colleagues to engage in bullying, rather than discouraging them from attacking whistleblowers (Murray 2007). This can happen through role modeling (social learning theory, Miceli and Near 1992), through explicit or implicit encouragement from the superior for the whistleblower's colleagues to engage in bullying, or through signaling from the superior that s/he will not prevent acts of bullying by the whistleblower's colleagues. In this way bullying by superiors may fuel the spread of the bullying in the workplace (Lutgen-Sandvik et al. 2007) and influence the ethics and ethical infrastructure of an organization (Berry 2004;Einarsen et al. 2017;Senekal and Uys 2013). The damage to the victim worsens when colleagues engage in bullying with tacit acceptance or even approval from superiors. To test the relationship between bullying by superiors and colleagues, we proposed the following hypothesis.

H1
Bullying by superiors will significantly affect bullying by colleagues.

Internal and External Support for Whistleblowers
Whistleblowing has been classified as a prosocial behavior intended to increase public interest (Dozier and Miceli 1985). A bullied victim's antisocial or prosocial behavior is a determinant of colleagues' help-providing intention towards a victim, so the extent to which colleagues perceive the whistleblower's actions as prosocial depends on the extent to which they understand the reasons for those actions. The support of colleagues acts as a buffer against stress suffered by victims, whereas a lack of support may exacerbate the bullying situation (Desrumaux et al. 2016). Although whistleblowers at times receive support from their colleagues in 'private settings' (McGlynn and Richardson 2014, p. 213), most organizational members rarely support whistleblowers. Even in situations in which colleagues might wish to help they may fail to develop an intention to help in a bullying situation due to fear of retaliation (Báez-León et al. 2016) and potential stigma by association (Mulder et al. 2014). Support of organizational members for whistleblowing, however limited, may nevertheless contribute to reducing the frequency and level of distress associated with bullying by colleagues that whistleblowers experience. We therefore propose the following hypothesis: H2 Bullying from colleagues will be less frequent and distressing in situations where colleagues understand the reason(s) for the whistleblower's actions.
In many countries, laws have been enacted to protect whistleblowers against retaliation by organizations in situations where the whistleblowing is in the public interest (Fasterling and Lewis 2014). The State has therefore placed itself as a stakeholder in the whistleblowing process, and implicitly accepted an obligation to protect whistleblowers from retaliation, which would include bullying. NGOs (such as Government Accountability Project, Public Concern at Work, Blueprint) are also engaged in assisting whistleblowers through actions such as representing them in court, filing a petition instructing organizations to desist from continuing to engage in bullying, and appealing for public support. Although the involvement of government and NGOs might seem to be helpful for the whistleblower, there is limited evidence it would deter bullying. Indeed such support may actually prompt bullying, by reinforcing the tendency to view external whistleblowers as a threat to the organization, disloyal, separate from their colleagues, etc. 2 And since the actions of government or NGOs are likely to be focused on the employer, they are unlikely to affect the behavior of the whistleblower's colleagues. Thus, we hypothesized that: H3 External support for whistleblowers (from government and NGOs) while they suffer bullying in the workplace will not significantly affect bullying by colleagues.

Sample and Data Collection
To test the hypotheses above, we surveyed 72 external Korean whistleblowers, whose names were in the public domain through media reports. First, we compiled a list of whistleblowers that disclosed wrongdoing in the workplace to the media or authorities outside the organization using the databases of major daily newspapers covering the period 1992-2013. Next, we traced addresses or contact numbers through multiple sources: former colleagues and friends, civic groups advocating whistleblower protection, social networks on which whistleblowers share their experiences, and articles from newspapers and magazines that featured stories about the whistleblowers. This gave us a database of 143 external whistleblowers. Since we were interested in their experience of workplace bullying, it was important to survey only those who had remained with their employer after blowing the whistle. We screened out those who did meet this criterion through a simple initial question "Did you quit your job shortly before or after blowing the whistle?" This left us with 72 potential participants whom we invited to participate in this study via phone and email. Through considerable persistence over an extended period (December 2013-January 2017), including in some cases visiting the participants, we eventually managed to survey all 72 external whistleblowers.

Measures
This study used the Negative Acts Questionnaire-Revised (NAQ-R), with slight modifications, to measure frequency and distress of bullying suffered by external whistleblowers in the workplace. The NAQ-R is an improvement of the Negative Acts Questionnaire developed by Einarsen and Raknes (1997) and is one of the most widely used instruments to measure workplace bullying ). The NAQ-R does not differentiate between bullying by superiors and colleagues and so, drawing on the work of Waschgler et al. (2013), we divided the 22 NAQ-R items to enable an exploration of the frequency and distress of bullying by the two groups. For example, "Some coworkers criticize the way I work" was classified into collegial workplace bullying behavior. We then consulted with five whistleblowers who had experienced bullying in the workplace and two members of staff from whistleblower protection NGOs that had worked with whistleblowers for over a decade to get their views on the items. This process resulted in two subscales: a group of 13 bullying items by superiors and a group of 11 items by colleagues. Based on recommendations from the consultation groups, two NAQ-R items-"withholding information that affects your performance" and "threats of violence or physical abuse or actual abuse"-were included in both subscales. In total the questionnaire consisted of 24 items, all drawn from the NAQ-R, but with two items used in both scales. Finally, the items were translated from English into Korean, with back translation then used to check the faithfulness of the translation (cf. Schaffer and Riordan 2003).
Participants were presented with the 24 items and asked to indicate: "Over the last 6 months since you were identified as a person who disclosed wrongdoing within the organization, how often did each event happen to you?" as well as to "How distressed did you feel as a result of the event at your workplace?" Participants were asked to rate items for bullying frequency on a 5-point scale: 5 = almost every day; 4 = one to three times per week; 3 = one to three times per month; 2 = one to three times for 6 months; 1 = never, and for distress from bullying on a 5-point scale (4 = extremely distressed to 0 = not at all distressed).
To assess perceived support, participants were asked to indicate what type of backing they had experienced in the situation from the organization or societal actors. Example: "How would you rate each of the following social actors' support or help after exposure to negative consequences after reporting wrongdoing at work?" Participants were then asked to rate the perceived extent of support from each of these actors (e.g., government and NGOs) using a five-point Likert scale response format, ranging from 1 = not at all to 5 = very much. This section of the survey also included an item asking participants whether their colleagues understood their reason(s) for blowing the whistle externally.

Demographic Characteristics of the Sample
In the full sample (N = 72), 65 (90.3%) participants were male and 7 (9.7%) were female. (It is unclear why such a high proportion of the whistleblowers were male). Age was measured in five categories: 1 = less than 30, 2 = 30-39, 3 = 40-49, 4 = 50-59, 5 = more than 59. The largest age group included 38 (52.8%) that were 40-49, followed by 23 (31.9%) 50-59. Education level was categorized into three groups: 1 = less than a high school degree or equivalent, 2 = junior college/four-year university degree, and 3 = postgraduate. Forty-one (56.9%) of respondents had junior college/four-year university degrees, compared to just three (4.2%) of the total respondents with a high school education or less. Table 1 presents data on the extent to which external whistleblowers experience workplace bullying from superiors and colleagues in terms of frequency and distress.

Perceived Frequency and Distress of Workplace Bullying
The frequency of bullying by superiors was the highest for the action "excessive monitoring of your work" (m = 3.90. cf. 4 = one to three times per week). The lowest was the action "being exposed to an unmanageable workload" (m = 2.19. cf. 2 = one to three times for 6 months). The action that triggered the highest level of bullying distress was "excessive monitoring of your work" (m = 3.29. cf. 3 = very distressed), while that of the lowest bullying distress was the action "being exposed to an unmanageable workload" (m = 2.39. cf. 2 = moderately distressed). The overall results revealed that respondents were more exposed to and distressed by superiors' work-related bullying (e.g., "excessive monitoring of your work," "withholding information Table 1 Frequency and distress of bullying by superiors (N = 72) Percent affected is the percentage of the respondents who experienced workplace bullying at least "one to three times for 6 months" For distress, the respondents who answered they were distressed from workplace bullying although they never experienced it were omitted from its calculation that affects your performance") than person-related bullying (e.g., being shouted at or being the victim of spontaneous anger/rage) or physical bullying (e.g., threats of violence or physical abuse or actual abuse). Prior to an exploratory factor analysis to identify latent dimensions of bullying actions by superiors, we conducted KMO and Bartlett's Test for the 13 bullying by superiors items to examine if the sample is appropriate for factor analysis. Analysis revealed that KMO statistics for the bullying actions in terms of frequency and distress were 0.884 (approx. χ 2 = 784.540, df = 78, sig. = 0.000) and 0.889 (approx. Chi square = 780.928, df = 78, sig. = 0.000), respectively, revealing that the sample was adequate for factor analysis. Applying the Kaiser rule of principal component analysis (PCA) (eigenvalues > 1) and the rule of parallel analysis (PA) (PCA eigenvalues > PA eigenvalues from corresponding random data), we conducted a factor analysis on actions of superior bullying. Factor analysis extracted one factor for perceived frequency of and distress from bullying by superiors, indicating no significantly different forms or patterns of bullying were identified, suggesting that participants experienced bullying as a single phenomenon, despite the diverse ways in which bullying may be performed.
For bullying by colleagues, we analyzed the items with the same procedures described above. Table 2 shows the results.
Frequency and distress of bullying by colleagues was the highest for the action "being ignored or excluded" (m = 4.19. cf. 4 = one to three times per week) and "having insulting or offensive remarks made about your person, your attitude or your private life" (m = 3.21. cf. 3 = very distressed), respectively, but the lowest on the action of "threats of violence or physical abuse or actual abuse" (m = 2.14, m = 2.38, respectively). Overall results revealed that respondents perceived social relation-related or person-related bullying (e.g., "being ignored or excluded," "spreading of gossip and rumors about you") as more frequent and distressful than physical bullying (e.g., "threats of violence or physical abuse or actual abuse," "intimidating behavior such as finger-pointing, invasion of personal space, shoving, blocking your way").

Support for Whistleblowers
We next turned to examine how much support the whistleblowers received inside and outside an organization when they suffered bullying in the workplace. We asked respondents how much they received assistance from each of four different sources (see Table 3).
Respondents reported the most support from family members (m = 4.14), followed by colleagues' understanding of the whistleblower's reasons for acting, support from NGOs, and finally support from government. The result of t-test revealed the mean of family members' support was significantly greater than that colleagues' understanding of the whistleblower's reasons for acting (mean Table 2 Frequency and distress of bullying by colleagues (N = 72) Percent affected is the percentage of the respondents who experienced workplace bullying at least "one to three times for 6 months" For distress, the respondents who answered they were distressed from workplace bullying although they never experienced it were omitted from its calculation difference = 0.833, t = 5.186, p < 0.001), which in turn was greater than that of NGOs' support (mean difference = 0.611, t = 3.144, p < 0.001), which in turns was greater than that of government support (mean differences = 0.778, t = 4.446, p < 0.001). The four different types/sources of support are thus separated from the next highest/lowest form of support by a statistically significant amount.

Impact of Internal and External Support
We conducted multiple regression analyses to examine the impact of superiors' bullying, internal support (colleagues' understanding of the whistleblower's reasons for acting), and external support (from government and NGOs) on bullying by colleagues. Concerned that the sample of 72 external whistleblowers may be insufficient for multiple regression analysis, we examined if the sample size fits for regression analysis using a sample size calculator (see http://www. danie lsope r.com/statc alc/calcu lator .aspx?id=1). Analysis revealed that a minimum sample size of 71 was required for the regression model with four predictors, given the probability level of 0.05, the anticipated effect size of 0.18, and the desired statistical power level of 0.8. Based on this result, we considered the sample adequate for regression analysis. Table 4 details the results of the analysis.
Regression models of frequency and distress of colleagues' bullying were significant for the data (F = 56.432, p < 0.001; F = 94.935, p < 0.001, respectively). Frequency and distress of superiors' bullying accounted for most of the variance of frequency and distress of bullying by colleagues (b = 0.862, p < 0.001; b = 0.938, p < 0.001), revealing that superiors' bullying is a dominant factor that affects frequency and distress of colleagues' bullying. Colleagues' understanding of the whistleblower's reasons for acting is a significant factor in reducing bullying frequency (but not distress) by colleagues (b = − 0.118, p < 0.05), while the impact of support from government or NGOs was insignificant for both frequency and distress of bullying by colleagues. Based on these results, hypotheses H1 and H3 were supported, and H3 was partially supported.

Discussion
This study investigated external whistleblowers' experience of workplace bullying following their disclosure, in terms of frequency and distress, and the impact of bullying from superiors on bullying by colleagues, colleagues' understanding of the whistleblower's reasons for acting, and support from government and NGOs. Our results are consistent with findings of studies on non-whistleblowers (e.g., Waschgler et al. 2013;Einarsen 1999) that bullying by superiors is work-related, while that by colleagues is social relationrelated and person-related. Victims experience less frequent physical bullying compared to psychological and social bullying-"physical abuse or threats of physical abuse" was one of the least frequently experienced form of workplace bullying (cf. Einarsen and Raknes 1997). Differences between bullying by superiors and colleagues may be attributed to job roles and behavioral patterns in the workplace. Previous studies of workplace bullying using the NAQ or NAQ-R survey (e.g., Hoel et al. 2001;Lutgen-Sandvik et al. 2007 and;Tsuno et al. 2010) reported that approximately 10-25% of employees were exposed to workplace bullying involving at least one or more of the 22-29 negative acts over the last 6 months. Compared to these findings, the external whistleblowers in our study reported much higher frequencies-the vast majority of our participants had experienced at least one form of workplace bullying in the aftermath of blowing the whistle. This clearly had an impact, as although 34 (47.2%) of the 72 respondents were still with the same employer, 38 (52.8%) reported they had eventually left due to the aftermath of their disclosure of wrongdoing.
Our study supports previous findings regarding the different levels of retaliation of internal versus external whistleblowers (see, e.g., Dworkin and Baucus 1998) as well as preliminary results from longitudinal studies of whistleblowing and workplace bullying (Bjørkelo et al. 2009(Bjørkelo et al. , 2015. The study reveals that both superiors and colleagues may perform workplace bullying perceived as targeted towards external whistleblowers. Bullying by superiors had a profound impact on bullying by colleagues in terms of frequency and distress. In line with Leymann's (1996) model and the work of others regarding the potential impact of retaliation and bullying on work environment and organizations (Berry 2004;D'Cruz and Bjørkelo 2016;Einarsen et al. 2017;Senekal and Uys 2013), bullying by superiors may lead to bullying by colleagues through provoking or at least allowing subordinates to retaliate against whistleblowers. Colleagues' understanding of the whistleblower's reasons for acting had a significant effect on reducing bullying frequency by colleagues, but support from government and NGOs did not. The result may be interpreted as an indication that the government and NGOs fail to protect whistleblowers from being bullied by superiors and colleagues. Although employment protection for whistleblowers is in place in many countries to shield them from retaliation by employers, this may not always be sufficient to protect whistleblowers from bullying, which can be carried in subtle ways that nevertheless inflict serious social or psychological damage to whistleblowers.
Our results suggest governments need to develop strategies to protect whistleblowers from being bullied in the workplace to reflect the differences between bullying by superiors and colleagues. NGOs continue to lobby governments to introduce or strengthen whistleblowing protection laws, and a legal requirement making an employer more liable for bullying by superiors and colleagues (already in place in some countries) could be effective in protecting whistleblowers from bullying at work. Norway, described by Skivenes and Trygstad (2010) as an example of where whistleblowing works, provides some useful insights. Setting aside cultural and structural elements (e.g., low power distance, high union density) which are not readily replicated, Norway's legal approach offers a potential way forward. There are two separate legal provisions which apply, one protecting whistleblowers against retaliation (Lewis and Trygstad 2009), and one which protects all employees against harassment (the 2012 Work Environment Act). Whistleblowers experiencing bullying are thus in theory doubly protected, which encourages employers to take particular care to ensure these individuals do not experience a hostile work environment, although sadly this is not always the case (Bjørkelo 2017). Further research is required to examine the relationship between government and NGO efforts to protect whistleblowers and the bullying they receive from superiors and colleagues-comparative studies looking at different national contexts and legal systems would be invaluable.
What might organizations themselves do to signal their support for whistleblowers and protect them from bullying by superiors and colleagues? Bullying of whistleblowers does not take place in isolation, and is logically more likely to occur in organizations where bullying and harassment is already common. If an organization wishes to ensure whistleblowers are not bullied, then they will first need to ensure bullying in general is not tolerated. Organizations can seek to create a culture that supports whistleblowers through a work training sensitivity program that raises employees' awareness about bullying and empowers them to be better prepared to interact with whistleblowers (see, e.g., Berry 2004;Senekal and Uys 2013). Previous studies on bullying in general, for instance, suggest focusing on ethical understanding (Lavan and Martin 2008), climate (Bulutlar and Öz 2009), and leadership (Stouten et al. 2010).
One of the challenges for organizations seeking to assure potential whistleblowers that they will be supported is that effective management responses to whistleblowing may involve HR processes that are confidential. To illustrate, a member of staff may be given a written warning about their conduct because of information brought to management attention by a whistleblower, but neither the whistleblower nor anyone else in the organization will be aware of this (Blenkinsopp and Edwards 2008;Vandekerckhove et al. 2016). Senior managers in their study were very aware of the problem, and keen to make the workforce aware of action taken in response to whistleblowing. This is easiest to do when the wrongdoing results in dismissal, as the outcome is very visible. One senior manager commented that when an executive is "thrown off the fourth floor" (the C-suite) as a result of wrongdoing the firm would metaphorically "leave the body there for a while" to ensure everyone gets the message that management will act when wrongdoing is brought to their attention.
There are some potential limitations to the study which should be acknowledged. First, the modified version of NAQ-R used in our study was developed to measure bullying in Anglo-American cultures ), and so there may be issues with using it for a Korean sample, though we note that the measure is one of the most widely used globally (e.g., Power et al. 2013). Second, internal and external support for whistleblowers was measured by a single item. Third, the NAQ-revised way of measuring bullying from leaders may have underestimated the frequency. The scale for frequency of bullying runs from 5 (almost every day) to 1 (never); however, Boddy et al. (2015) found some forms of bullying can occur more than once a day, leading Boddy and Taplin (2017) to recommend that frequency of bullying might best be measured by actual numerical frequency. Future research on bullying of whistleblowers could usefully adopt this approach. Finally, the fact this was an entirely Korean sample may limit our ability to generalize findings to other cultural settings. On this last point, the study by Park et al. (2008) provides us with insights to Korean attitudes to different forms of whistleblowing. Comparing Korea to the UK and Turkey, they found statistically significant differences between the countries, but also a degree of consistency-in all three countries participants preferred internal over external whistleblowing, anonymous over identified whistleblowing, and formal over informal whistleblowing, and in all cases external whistleblowing was the least preferred route. Given also previous evidence that external whistleblowers face greater retaliation, it seems reasonable to suggest our findings from the Korean sample are in the direction which might be expected in other countries.

Conclusion
External whistleblowers are exposed to higher levels of retaliation than internal whistleblower (see, e.g., Dworkin and Baucus 1998), and our data suggest they may also be exposed to higher levels of workplace bullying (see also Bjørkelo 2013). However, less has been known about the nature of workplace bullying, as perceived by external whistleblowers. Despite the limitations acknowledged above, this study adds practical information to the literature about damages that whistleblowers suffer after exposing wrongdoing by exploring bullying behavior by superiors and colleagues and the severity perceived by whistleblowers and the impact on bullying by colleagues of superiors' bullying and support for whistleblowers from inside and outside organizations. Our study provides insights into the workplace bullying that external whistleblowers experienced, which is informative in capturing the intensity of bullying behavior. We found that bullying by superiors had a close link to bullying by colleagues. Colleagues' understanding of the whistleblower's reasons for acting had a significant effect on lowering frequency of bullying by colleagues while government and NGO support were insignificant.

Compliance with Ethical Standards
Conflict of interest All authors declare that they have no conflicts of interest.
Open Access This article is distributed under the terms of the Creative Commons Attribution 4.0 International License (http://creat iveco mmons .org/licen ses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made.