1During the past few decades, far-reaching changes have taken place in the work environment in most Western countries. These changes are also reflected in the risks that employees are exposed to at work and in the conceptualizations of what is seen as a ‘good work environment’. More and more emphasis is put on the psychosocial work environment and on violence prevention. For example, the prevention of harassment and interpersonal hostility in the workplace is becoming increasingly important to both managers and organizational researchers (e.g., Einarsen, Hoel, Zapf & Cooper, 2003 ; Neuman & Baron, 1998).
2Interest in the prevention of harassment and interpersonal hostility in the workplace is largely motivated by the fact that a growing number of studies point to severe negative consequences associated with such phenomena. For example, interpersonal hostility may have serious effects on both the job satisfaction and well-being of victims (e.g., Ashforth, 1997 ; Einarsen & Raknes, 1997 ; Mikkelsen & Einarsen, 2001 ; Tepper, 2000). What is more, interpersonal hostility may also lead to high costs for organizations, in the form of increased absenteeism and higher turnover of personnel, decreased commitment and productivity, and negative publicity (e.g., Ashforth, 1997 ; Hoel, Einarsen & Cooper, 2003 ; Tepper, 2000). For society as a whole, this may lead to lower productivity, early retirements and increased health costs. As a result, several countries have adopted or are planning to adopt laws promoting dignity at work or banning different forms of work harassment.
3This paper addresses one aspect of harassment and interpersonal hostility in the workplace, i.e., workplace bullying. Workplace bullying can take many different forms and may be either work-related or person-related (e.g., Einarsen & Hoel, 2001). Work-related bullying may include, for example, unjustified criticism about a person’s work, unreasonable deadlines, somebody withholding information or somebody being excessively monitored. Person-related bullying may include, for example, insulting comments about a person’s private life or background, humiliating or intimidating behavior, rumors or false allegations, and exclusion or isolation. Most researchers emphasize that in order for negative acts to be classified as bullying there has to be an element of duration and/or repetitiveness of the behavior (cf. Einarsen & Skogstad, 1996 ; Hoel & Cooper, 2000 ; Vartia, 1996). What is more, researchers typically stress the centrality of power, or more precisely, power differences (cf. Einarsen & Skogstad, 1996 ; Hoel & Cooper, 2000 ; Vartia, 1996). In this paper, workplace bullying is therefore defined as
4repeated and persistent negative behavior, which involves a power imbalance and creates a hostile work environment (cf. Einarsen & Skogstad, 1996; Hoel & Cooper, 2000; Vartia, 1996).
‘However, this study also draws from research describing similar behaviors under different names, e.g., ‘mobbing’ (Leymann, 1996), ’workplace aggression’ (e.g., Neuman & Baron, 1998), ‘petty tyranny’ (e.g., Ashforth, 1997) and ’victimization’ (e.g., Aquino, Grover, Bradfield & Allen, 1999).’
5Since the early 1990s, a number of studies have been conducted to better understand the dynamics and causes of bullying (e.g., Aquino et al., 1999; Coyne, Seigne & Randall, 2000; Einarsen, Raknes & Matthiesen, 1994; Neuman & Baron, 1998; Vartia, 1996; Zapf, Knorz & Kulla, 1996) and to measure prevalence rates in different countries and professions (e.g., Einarsen & Skogstad, 1996; Hoel & Cooper, 2000; Leymann, 1992a; O’Moore, 2000; Soares, 2002). However, studies on bullying have typically been conducted in the public sector and among employees in non-managerial and often non-career-oriented jobs. For example, several studies have been undertaken fully or partially in the health care sector (e.g., Einarsen, Matthiesen & Skogstad, 1998 ; Kivimäki, Elovainio & Vahtera, 2000 ; Mikkelsen & Einarsen, 2001 ; Niedl, 1995 ; Ouden, Bos & Sandfort, 1999). Some other studies have been conducted among municipal or local council workers, mostly in administrative and fairly routine types of jobs (e.g., Ouden et al., 1999 ; Vartia, 1996), or among industrial workers (Einarsen & Raknes, 1997 ; Leymann & Tallgren, 1989). In contrast, highly educated employees in managerial positions in private companies have received little attention in bullying research.
6A few studies have addressed the role of organizational status with regard to bullying. However, the results have been contradictory. In a Swedish nationwide study on bullying, senior managers were slightly over-represented among bullying victims, although the differences with regard to hierarchical position were not statistically significant (Leymann, 1992a). In addition, in a large Finnish survey, upper white-collar employees reported somewhat more bullying than lower white-collar employees or workers (Piirainen et al., 2000). However, contrasting findings have been presented by Hoel, Cooper and Faragher (2001) based on a large-scale study in the UK. Although they found few differences in the experience of self-reported bullying between employees in different hierarchical positions, they reported that workers and supervisors were more often exposed to predefined negative acts than managers. Given the scarcity of studies about bullying among managers and other highly educated business professionals and the inconclusive data concerning the effects of organizational status in bullying, further research on these groups of employees is needed. This article is thus one attempt to address this gap in existing research.
7What is more, so far the data on bullying and gender seem to be inconclusive, and contradictory findings have been reported. Most large-scale studies have reported fairly equal victimization rates for men and women (Einarsen & Skogstad, 1996 ; Hoel & Cooper, 2000 ; Leymann, 1992a ; Vartia, 1996). However, studies in specific occupational groups, for example among university employees (Björkqvist, Österman & Hjelt-Bäck, 1994), employees in child care (Lindroth & Leymann, 1993) and assistant nurses (Eriksen & Einarsen, 2004), have showed different patterns, typically that members of the underrepresented sex have reported significantly higher rates of bullying. Closer examination of the relationships between bullying and gender reveals even more complex patterns, for example in terms of the position and gender of the bullies. While men are typically bullied by male supervisors, women report being bullied by both superiors and colleagues and by both men and women in approximately equal proportions (Einarsen & Skogstad, 1996 ; Hoel & Cooper, 2000 ; Leymann, 1992b). What is more, Hoel et al. (2001) reported that while male workers and supervisors reported higher victimization rates than their female colleagues, the pattern was reversed for middle and senior management. The inconclusive data and the complex patterns between gender and bullying found in previous studies highlight the need for further studies on the significance of gender in bullying. This article contributes to this aim by studying in more detail the significance of gender in bullying in a sample of business professionals.
8Existing research has demonstrated strong links between the work environment and the prevalence of workplace bullying. For example, leadership style, job design, organizational norms and values, and communication climate have been found to have significant effects on the prevalence of bullying (Einarsen et al., 1994 ; Hoel & Cooper, 2000 ; Hoel & Salin, 2003 ; Vartia, 1996 ; Zapf et al., 1996). Still, bullying is typically seen as an ‘irrational’ behavior, due for example to unwanted personality traits or dissatisfaction in the workplace. Although some German researchers have described bullying as ‘foul game’ in organizations (Neuberger, 1999), as ‘personnel work with other means’ (Zapf & Warth, 1997) or as a ‘rent-seeking strategy’ (Kräkel, 1997), this perspective has gained little attention in the current bullying debate (e.g., Zapf & Einarsen, 2003). However, there are situations where it might be individually ‘rational’ or rewarding to bully somebody, for example if the potential victim is considered either as a burden for the department or as a personal rival (e.g., Kräkel, 1997). Furthermore, studies have shown a correlation between performance-based reward systems and bullying (Sutela & Lehto, 1998). It is thus possible to argue that bullying may be closely related to the phenomenon of organizational politics, i.e., the phenomenon when individuals or groups deliberately act in a way that will protect or enhance their own self-interests, when their actions may or may not be in the best interest of other individuals, groups or the organizations to which the actor belongs (Allen, Madison, Porter, Renwick & Mayes, 1979 ; Kacmar & Ferris, 1991). Thus, this study also aims to explore this neglected link between organizational politics and bullying.
9This article builds on the insights from the author’s doctoral thesis, which was entitled Workplace Bullying among Business Professionals : Prevalence, Organisational Antecedents and Gender Differences (Salin, 2003a). In the following section, the objectives of the thesis and this paper are described in more detail. Subsequently, the method employed in the study is described and the results are presented. Finally, the results and their implications are discussed.
10Based on the gaps and inconsistencies in previous research on bullying, the objectives of this article have been defined. The main objectives are to describe the prevalence of bullying experienced by business professionals and to further the understanding of bullying by analyzing to what extent gender aspects and organizational politics may contribute to bullying in knowledge-intensive career-oriented jobs. More precisely the aims are :
to measure the prevalence of bullying among business professionals. This study thus aims to describe the frequency of bullying among business professionals and how it is expressed among professionals, i.e., what kinds of negative acts professionals report.
to explore the significance of gender in bullying. The study thus aims to analyze whether there are gender differences in the prevalence and perceptions of bullying, and if so, how these differences can be explained.
to analyze the role of the work environment in bullying among business professionals. In particular, this study aims to analyze how the fact that the respondents work in career-oriented jobs (which are often characterized by a highly politicized climate) may affect the presence of bullying.
11The research questions were studied using a multi-method approach, combining both quantitative and qualitative data. Quantitative data were used in order to study the prevalence and forms of bullying and the relationship between bullying and organizational politics, while qualitative data were used in order to explore the subjective experiences of bullying, in particular gender differences in perceptions and interpretations of bullying episodes.
12A questionnaire was mailed to 1000 randomly drawn members of SEFE - The Finnish Association of Graduates in Economics and Business Administration, a central Finnish nation-wide professional organization for employees holding a university degree in business studies (please see www.sefe.fi for more information). The organization plays an active role in national social and income policies and participates in bargaining and contract negotiations concerning professional and managerial staff. In addition, the organization aims to support the lifelong learning and career planning of members, by providing them with professional information and arranging seminars, by conducting regular salary surveys, and by developing recruitment services. Various leisure activities are arranged on a regional level. Altogether the organization has approximately 28,000 ordinary members, divided into 26 different regional associations, and 12,000 associated student members. The questionnaire was sent to ordinary members from all the regional sub-associations.
13A total of 385 questionnaires were returned, giving a response rate of 38.5 %. Owing to extensive missing data, 8 individuals were excluded from all analyses. Of the respondents, the majority (82 %) worked within the private sector, and most of them identified their organizational status as either ‘managerial’ or ‘expert’. More precisely, 13.6 % classified themselves as managers, 30.7 % as middle managers, 37.1 % as experts, 16.8 % as officials/clerks and 1.8 % as ‘other’ (e.g., teachers). They represented a variety of sectors, with the highest numbers of respondents representing the manufacturing (23.4 %), wholesale and retail trade (16.2 %), and business services sectors (15.4 %). Mean job tenure was 6.9 years and almost all (97.6 %) were employed full-time. Of the respondents, 57.3 % were females and 42.7 % males, with women somewhat underrepresented among managers and overrepresented among clerks and officials. Compared to the overall 50-50 gender ratio of the surveyed professional organization, women were slightly overrepresented among those responding, indicating a higher interest in the subject among women.
14To obtain a more comprehensive picture of bullying, two complementary methods were used to measure bullying (cf. Hoel et al., 2001). First, the respondents were asked to indicate on a scale from 1 (never) to 5 (daily) how often they had experienced 32 listed negative and potentially harassing acts over the past 12 months. The 32 potentially harassing acts were primarily taken from the Negative Acts Questionnaire (NAQ) (Einarsen & Raknes, 1997 ; Einarsen & Hoel, 2001). Second, the respondents were introduced to a short definition of bullying and asked whether they had been subjected to such behavior within the past 12 months, and if so, how often. In the questionnaire the respondents were given the following definition of bullying :
‘Bullying’ is used here to denote repeated and persistent negative acts that are directed towards one or several individuals, and that create a hostile work environment. When subjected to bullying, the targeted person has difficulties defending him or herself ; it is therefore not a conflict between parties of equal strength.
15Perceived organizational politics was measured with the Perceptions of Politics Scale (POPS) developed by Kacmar and Ferris (1991). However, some minor changes in wording suggested by Nye and Witt (1993) were adopted. The instrument consisted of 12 statements concerning the perceived degree of organizational politics, and included, for example, statements such as ‘Some people build up themselves by tearing others down’ and ‘Favoritism, not merit, gets people ahead’. Respondents were asked to judge all the statements on a Likert-type scale from 1 (strongly disagree) to 5 (strongly agree). The internal consistency of the scale was high : Cronbach’s alpha was 0.88.
16Finally, all respondents who had either experienced or observed bullying were encouraged to write down their stories in their own words. Altogether 79 respondents did this, 60 of them women and 19 men. The qualitative data were analyzed using N’Vivo software for qualitative data. The process started by entering all the written stories as documents into N’Vivo and by defining suitable attributes for the documents (for example, the respondent’s gender and position) and by coding the text sequences (for example, in terms of different forms of bullying and different explanations for bullying). The analysis then continued by identifying relationships between the different categories.
17In the analysis of these stories or narratives, attention was paid to the complex relationship between facts and stories. Gabriel (2000 : 5) argues that “stories interpret events, infusing them with meaning through distortions, omissions, embellishments and other devices, without, however, obliterating the facts.” In other words, stories can be seen as part of a sense-making process, indicating that stories are influenced both by facts and the meanings and interpretations that the teller gives them. Instead of them being seen as mere factual descriptions dealing with ‘facts-as-information’, they represent ‘facts-as-experience’ (cf. Gabriel, 2000). For example, although stories may convey some insights into actual gender differences in bullying, they also convey information about gender differences in perceptions of bullying and the meanings attributed.
18The first aim of the study was to investigate the prevalence and forms of bullying. As for the reported prevalence of bullying, it can be noted that when provided with a definition of bullying, 8.8 % of the respondents reported that they had at least occasionally been bullied in the past 12 months. 1.6 % reported that they had been bullied at least weekly. A considerably higher number had been affected by bullying at least indirectly. Of all the respondents, 30.4 % reported that they had witnessed bullying in their present workplace in the past 12 months.
19However, in this study, self-classifications of being bullied or not were also compared with another strategy for measuring bullying. When using a slightly modified version of the Negative Acts Questionnaire (Einarsen & Hoel, 2001) containing a list of 32 predefined negative and potentially harassing acts, as many as 24.1 % of the respondents reported that they had been subjected to at least one of the negative acts on a weekly basis. Thus, the difference in prevalence level using the two different methods for measuring bullying (8.8 % vs. 24.1 %) was considerable. However, there was a degree of consistency between the two methods in the sense that the majority of those classifying themselves as bullied were also identified when using the Negative Acts Questionnaire.
20As for the forms of bullying, the respondents had experienced predominantly work-related negative acts. The negative acts that the highest number of respondents had been subjected to at least weekly were 1) being ordered to do work clearly below their level of competence (13.7 %), 2) someone withholding information (7.4 %), 3) being given tasks with unreasonable or impossible targets or deadlines (5.3 %), and 4) having one’s own opinions and views ignored (5.1 %). Of the non-work related items, being ignored or excluded was reported by the highest number of respondents (2.1 %) (See Salin, 2001 for more results.).
21The second aim of the thesis was to explore the significance of gender in bullying. In this sample, 11.6 % of the women as compared to 5 % of the men classified themselves as bullied. This difference was statistically significant (p<0.05). However, when measuring weekly exposure to the predefined negative acts, the gender difference was notably lower, or 26.4 % versus 21.1 %, and no longer statistically significant. As for the perpetrators of bullying, the results showed that men were typically bullied by their superiors, whereas women were bullied by superiors and colleagues in approximately equal proportions. What is more, only female respondents reported being bullied by subordinates. (See Salin, 2003b for more results).
22When discussing and trying to explain bullying, the written stories indicated that men were more likely than women to emphasize the victim’s role and responsibility. Men typically saw victims as ‘weak’ and ‘incompetent’. In contrast, women seemed to emphasize the personality of the perpetrator and group dynamics. Women tended to see tyrannical aggressors or to explain bullying in terms of a scapegoat phenomenon. What is more, women were more likely than men to see bullying as intentional and as related to internal competition, whereas men tended to see it as an unintended consequence of stress.
23Finally, the stories also revealed differences in handling and responding to bullying. Women often chose to contact their supervisors or health care professionals, whereas men more often actively confronted their aggressors.
24The third aim of the study was to investigate the relationship between bullying and the degree of perceived organizational politics. In the analysis, bullying was measured as the respondent’s own perception of how often he or she had been bullied, from 1 (never) to 5 (daily), and organizational politics were measured with the Perceptions of Organizational Politics Scale (Kacmar & Ferris, 1991). The correlation between bullying and organizational politics was 0.30, which turned out to be significant at the 0.01 level.
25Additionally, in order to verify the victims’ reports, observers’ reports were analyzed to see if their reports supported the assumed relationship between bullying and organizational politics. A t-test was performed to compare the means of perceived degree of organizational politics between the group consisting of those who had observed bullying in their workplace and the group consisting of those who had not observed bullying in their workplace in the past 12 months. The reported mean of organizational politics for those who had observed bullying was 3.07 (s.d. = 0.71), as compared to 2.64 (s.d. = 0.72) for those who had not observed bullying. This difference was also statistically significant (t =-5.221, df =350, p<0.00), which provides further support for a relationship between a high degree of organizational politics and bullying. In line with this, the stories written by observers of bullying also showed that bullying was sometimes seen as a deliberate attempt to get rid of unwanted people, for example employees who were considered as threats or a burden. (See Salin, 2003c for more results.)
26The aim of this study was to describe the prevalence of bullying experienced by business professionals and to further the understanding of bullying by analyzing to what extent gender aspects and organizational politics may contribute to bullying in knowledge-intensive career-oriented jobs. The aims were addressed by conducting a survey study among members of a professional organization for business professionals and by further encouraging the respondents to provide written stories about their experiences of bullying.
27This study clearly showed that bullying is also a widespread and visible problem among highly educated professionals. 8.8 % of the respondents classified themselves as bullied and the respondents reported high levels of work-related negative acts. In addition, there were significant differences in self-labeling between men and women, and men and women also provided different explanations for bullying, with men emphasizing the role of the victim and women emphasizing the role of the perpetrator or group processes. In addition, the study indicated a positive relationship between organizational politics and bullying.
28When compared to previous studies on bullying, this study yielded two rather surprising findings. First of all, in contrast to most other bullying studies in Scandinavia (e.g., Einarsen & Skogstad, 1996 ; Leymann, 1992b ; Vartia & Hyyti, 2002), women were significantly overrepresented among victims in this group. Second, compared to other studies (e.g., Einarsen & Raknes, 1997 ; Hoel & Cooper, 2000), the respondents in this study reported very high levels of exposure to work-related negative acts. Since some of these work-related acts of bullying (as for example, withholding information and ignoring views and opinions (cf. Allen et al., 1979)) are also described as tactics of micro-politics, this further indicated that bullying might sometimes be linked to organizational politics in career-oriented jobs.
29Several possible explanations can be proposed for the higher bullying rate reported by women. First of all, the fact that women are still underrepresented in managerial positions in Finnish business life (e.g., Hearn, Kovalainen & Tallberg, 2002), and are therfore more visible and exposed, might make them more vulnerable to negative acts. Second, the stories written by women also indicated that they felt more defenseless when confronted with hostility and less able to ‘pay back’. What is more, women seemed more willing than men to use the label ‘bullying’ to describe abusive behavior. This might be partly explained by the fact that women tended to see hostile perpetrators as the main reason for bullying, whereas men explained it in terms of weak victims. Men’s propensity to emphasize the victim’s role and responsibility would make them more likely to consider bullying a ‘personal failure’ for themselves. It can also be argued that all of these explanations are mediated by gendered perceptions of power.
30The fact that bullying among business professionals often manifested itself in the form of work-related acts and the fact that some of these acts (as for example, withholding information and ignoring views and opinions) are also described as tactics of micro-politics, indicated that bullying may sometimes be linked to organizational politics in career-oriented jobs. In addition, the results showed a positive, albeit not very strong, relationship between bullying and a generally politicized climate. The work environment may thus enable bullying, as for example, though weak leadership or indirect acceptance of bullying behavior. However, certain structures and processes may in fact encourage and motivate bullying (see Salin, 2003d). Certain reward systems and high internal competition may actually possibly increase the value of trying to eliminate colleagues or subordinates who are considered as burdens or rivals, and may thus lead to more negative forms of micro-political behaviors and bullying (cf. Kräkel, 1997).
31In addition to contributing to our understanding of gender differences in bullying and the role of organizational politics, the study also contributed to bullying research in methodological terms. First of all, it should be noted that the study made use of both quantitative and qualitative data, whereas the vast majority of previous studies relied on quantitative data only. What is more, in the quantitative part of the survey, two different methods were used for measuring bullying, i.e., both exposure to predefined negative acts and the respondent’s own classification of being bullied or not. This allowed for a comparison of the two widely used techniques and showed that although some consistency was found, using lists with predefined acts seems to overestimate the prevalence of bullying. An important implication of this is that great care needs to be taken when comparing studies based on different ways of measuring bullying. Another implication may be that not all of the negative acts included in the questionnaire are necessarily perceived as bullying by the respondents. It is thus possible that some of the acts, although perceived as negative and undesirable, may still be seen by many respondents as ‘natural’ aspects of working life that ‘simply have to be accepted’ rather than as ‘actual bullying’.
32As for the implications for managers, the results first of all showed that managers need to be aware that bullying can also occur at higher levels in organizations. In particular, managers need to be aware of the risks associated with a political climate, where everybody pursues his own interests, even at the expense of others. When encouraging competition in order to increase efficiency and reduce costs, managers need to be aware of the possible increase in micro-political behavior and possible detrimental effects on co-worker relations, which may even result in bullying. As for enabling/disabling structures in the work environment, managers need to ensure that employees perceive the costs associated with bullying someone as being high. Thus, management needs to have a clear policy on bullying and to show that perpetrators are both identified and punished. What is more, when trying to take appropriate prevention measures, management needs to ensure that there are no motivating factors in the reward system that may directly or indirectly encourage bullying. In other words, bullying and sabotaging the work performance of rivals must not be accepted and certainly not rewarded. For example, care should therefore be taken not to promote employees who have succeeded by using bullying tactics on others.
33Although this thesis has shed new light on bullying in a group that has been neglected in previous research, there are several limitations that must be kept in mind. First of all, the study was of fairly small scale and the response rate was only 38.5 %. It is therefore likely that the actual prevalence of bullying is slightly lower, since it may be assumed that respondents with experiences of bullying are slightly more willing to take part in this kind of study than those not concerned by the phenomenon. On the other hand, the researcher noted that bullying victims were overrepresented among the late respondents, indicating that responding to this kind of survey may be quite difficult for victims. Furthermore, when interpreting the results, it should also be remembered that the response rate was slightly higher among women. The higher response rate might partly be explained by the fact that a higher number of women felt affected by this phenomenon.
34Second, like most other studies on bullying, this study has mainly relied on a victim perspective and the victim’s subjective perceptions of being bullied or not. This study has not offered the possibility of comparing victim and observer accounts for the same bullying situations. What is more, information about bullying and organizational factors, such as the degree of organizational politics, was collected from the same respondents. There is thus the risk of a common method problem, with employees being subjected to or observing bullying also having a negative attitude towards the organization in general.
35Third, when conducting studies on bullying, it is important to remember the highly sensitive nature of this kind of investigation. The highly sensitive nature of bullying makes it difficult to obtain access to ‘objective’ data on bullying, since the victims’ stories are often colored by the fact that the victims are very strongly emotionally engaged in and affected by the bullying.
36Future studies should thus aim to incorporate victim, bystander and perpetrator accounts of the same bullying episodes. In addition, as pointed out earlier, further studies are needed on the labeling process and on the perceptions of which acts are actually construed as bullying in different national and occupational contexts. What is more, up to now, much research has focused on personal experiences of bullying, while the organizational perspective has received only limited attention. For example, little is known about the perceptions of bullying among HR professionals and their strategies to deal with the problem. Given the serious negative consequences associated with bullying, an important avenue for further research would be to analyze the prevention and intervention strategies chosen by companies and the strategies’ success rates.