Are Bullying Behaviors Tolerated in Some Cultures? Evidence for a Curvilinear Relationship Between Workplace Bullying and Job Satisfaction Among Italian Workers

Since the early 1990s, increasing attention has been paid to the impact of workplace bullying on employees’ well-being and job attitudes. However, the relationship between workplace bullying and job satisfaction remains unclear. This study aims to shed light on the nature of the bullying-job satisfaction relationship in the Italian context (n = 1,393 employees from different organizations). As expected, the results revealed a U-shape curvilinear relationship between workplace bullying and job satisfaction after controlling for demographic variables. In contrast to the curvilinear model, the results support a negative linear relationship between 1,*


Introduction
Scholars have paid increasing attention to workplace bullying and its severe negative consequences for employees' health and well-being since the introduction of the concept in the early 1990s. Although there is some discussion about possible overlaps between the concept of workplace bullying and other deviant behaviors displayed at work (e.g., incivility), researchers unanimously consider workplace bullying to be a harassment and social exclusion process wherein an individual is subjected to indirect and subtle forms of psychological violence-also referred as negative acts-in a systematic way and over a prolonged period of time (e.g., Einarsen et al. 2011 ). Examples of these acts include "being humiliated or ridiculed in connection with your work," "being ignored or facing a hostile reaction when you approach," and "being the subject of excessive teasing and sarcasm," among others (Einarsen et al. 2011 : p. 32). The more frequently individuals are exposed to these acts, the more they are considered as having been subjected to bullying.
However, one of the most significant current discussions in workplace bullying is the role that causal attributions play on the reactions of workers exposed to negative acts at work. In particular, considering the subtle and ambiguous nature of bullying behaviors, several authors have highlighted that workers may not perceive exposition to negative acts as a bullying situation (D'Cruz and Noronha 2010 ;Hoel et al. 2010 ;Parzefall and Salin 2010 ) or may even interpret such negative acts as part of their work environment and positive for their performance (Bulutlar and Ünler Öz 2009 ;Lee et al. 2013 ). Therefore, Samnani et al. ( 2013 ) questioned "whether bullying that is mistakenly perceived positively by the target can be labeled bullying." (p. 340). Furthermore, these authors have proposed "that bullying can paradoxically result in positive effects on target performance under certain conditions." (p. 337).
Thus, drawing on attributional models and considering contextual factors, this study aims to shed light on the relationship between workplace bullying, job satisfaction, and psychological well-being. In particular, our paper takes as a starting point the idea that victims of bullying usually try to determine the cause(s) and severity of the unwanted behavior received by considering this behavior in light of existing cultural norms and societal contexts. In turn, the perceptions generated within this sense-making process influence the magnitude and the direction of victims' reactions (Bowling and Beehr 2006 ;D'Cruz and Noronha 2010 ;Harvey et al. 2009 ;Hershcovis and Barling 2010 ;Parzefall and Salin 2010 ;Samnani et al. 2013 ).
We begin by analyzing the role of workplace bullying on job satisfaction. Based on the propositions made by Samnani et al. ( 2013 ) in their attributional model of bullying and the necessity of contextualizing studies in organizational behavior (Johns 2006 ), we argue that job satisfaction-as an attitudinal outcome-may reflect the way that victims of negative acts perceive the situation in which they are involved. Thus, in contrast to the mainstream findings from US and Northern European countries, we hypothesize that the bullying-job satisfaction relationship across Italian employees may be curvilinear. Under these assumptions, we then investigate the possible moderating role of job satisfaction on the association between bullying and psychological well-being. Last, we discuss how our results extend previous research by contextualizing workplace bullying, as well as we present some implications for implementing tailored interventions against workplace bullying.
Workplace Bullying and Job Satisfaction: Attributions and Contextual Factors resulting from the appraisal of one's job or job's experiences" (Locke 1976(Locke , p. 1304. Thus, the way that workers evaluate job satisfaction depends on their perceptions and judgments about their job's characteristics and their physical and interpersonal work environment. Accordingly, the prevailing assumption is that exposure to bullying behaviors will lead employees to perceive their work environment as hostile and negative and, therefore, bullying will be positively associated with job dissatisfaction (Bowling and Beehr 2006 ;Einarsen et al. 2011 ). However, research on the consequences of workplace bullying on employees' job satisfaction has neglected cultural and organizational contextualization explanations that are necessary to better understand such complex social phenomenon in organizations (Parzefall and Salin 2010 ). Indeed, Johns ( 2006 ) emphasized that context determines the meaning of organizational events. In that sense, our study takes an important step toward developing a better understanding of the bullying-job satisfaction relationship by contextualizing the way employees perceive their jobs and the meaning they imbue to workplace experiences (i.e., negative acts at work).
Regarding cultural factors, some authors have proposed that national culture may affect the relationship between workplace bullying and job satisfaction. For example, Loh et al. ( 2010 ) reported that workplace bullying was negatively related to job satisfaction in a sample of Australian and Singaporean workers, although this negative relationship was stronger among the Australians than Singaporeans due to Australians' lesser inclination for accepting power differences in their relationships. In similar vein, bullying behaviors seem to be more tolerated (or even accepted) in some cultures, especially in masculine cultures like Italy (e.g., Escartín et al. 2011a ;Power et al. 2013 ). Moreover, as reported by Javidan et al. ( 2006 ), the findings of the GLOBE (Global Leadership and Organizational Behavior Effectiveness) study reveal that Latin European countries (including Italy) rank among the lowest country clusters on "humane orientation," reflecting "the degree to which a collective encourages and rewards (and should encourage and reward) individuals for being fair, altruistic, generous, caring, and kind to others" (p. 69). This cultural practice corresponds with a view that such behavior does not greatly enhance leadership. Because employees in Latin European countries do not particularly expect compassion from others, workplace behavior that is not compassionate or kind (i.e., negative acts or bullying behaviors) may not be as disruptive to job satisfaction in Latin European contexts as they might be in countries scoring higher in humane orientation.
Similarly, according to the recent propositions made by Samnani et al. ( 2013 ) in their attributional model of bullying, employees will make contextbased attributions when there are high consensus (i.e., negative acts are common, which complicates identification of bullying), high consistency (i.e., negative acts are frequent over time, which lead to the normalization of bullying), and low distinctiveness (negative acts are directed to the whole group, which lead to perceive an in-group status). All these conditions seem to apply to the Italian organizations, leading employees to interpret bullying behaviors as culturally tolerated. Thus, in essence, we argue that Italian employees may be less likely to be dissatisfied with their jobs as a result of experiencing negative acts at work.
First, several studies have reported a high bullying prevalence in Italy, where more than 15 % of the employees have been exposed to at least two negative behaviors in a weekly or daily basis during the last 12 months (Giorgi 2009 ;Giorgi et al. 2011 ), which suggest that bullying is very common and frequent in Italian organizations (high consensus and consistency). Furthermore, when bullying is very widespread, some researchers have shown that bullying behaviors can be considered as part of the job and the work group culture or as a reasonable managerial practice in professions such as chefs (Bloisi and Hoel 2008 ) or police officers (Segurado et al. 2008 ). As Victor and Cullen ( 1988 ) pointed out, organizations shape the ethical or unethical behavior of their employees. The ethical climates "serve as a perceptual lens through which workers diagnose and assess situations'' (Cullen et al. 2003 : p. 129). Unfortunately, organizations often promote competition and reward behavior that is counter to what is generally accepted as ethical, supporting a competitive environment that rends all negative behaviors acceptable or even promotes people to exert negative acts to others (Bulutlar and Ünler Öz 2009 ). Thus, a higher tolerance of negative acts might rise as a consequence of perceiving that the organizational procedures and practices support deviant and unethical behaviors (i.e., bullying or unethical cultures).
In addition, bullying may be associated with job satisfaction, not only negatively but also positively because employees may perceive that receiving negative acts (e.g., personal jokes) suggests that she/he "fits in" with her/his work group or organizational culture (see Baillien et al. 2009 ), which increases "in-group" feelings that are incompatible to key bullying characteristics such as perceptions of isolation or out-group status (i.e., low distinctiveness). Furthermore, according to Samnani et al. ( 2013 ), the http://eproofing.springer.com/journals/printpage.php?token=fx-zw_rjwUruD5nZaXV_w8MjE39sPbQe3Ieznwa4PVMjIFF-3lpM6ju0Q_KDnV5K 6/26 combination of context-based attributions with the perception of positive work-driven intentions to a perpetrator of negative acts will lead to increase employees' performance since they perceive negative acts imbedded in a "challenging work culture" rather than interpreting them as bullying. In that sense, Yildiz et al. ( 2008 ) Vin a Turkish study among private sectors employees noted that victims perceived that there was no intention of harm behind bullying behaviors because these behaviors (e.g., "being exposed to workload", "being given tasks with unreasonable or impossible targets or deadlines", or "being ordered to do work below your level of competence") were interpreted somehow as a necessary sacrifice for their career development. Moreover, workers claimed that the experience of being subjected to negative actions on the job had strengthened their character.
Findings from a recent qualitative study by van Heugten ( 2013 ) seem to support this idea since 17 New Zealand social workers reported that they had developed greater resilience after being victims of bullying. Similar positive or neutral explanations for negative workplace acts may be particularly true in the current Italian's turbulent economic environment, in which diverse negative acts might be interpreted as sacrifices needed for job stability or career growth, and consequently viewed as satisfactory.
In sum, to the extent that negative acts are interpreted as occurring to all employees, such acts could be less damaging to job satisfaction than previously thought. In fact, employees' job satisfaction may be lowest when negative acts are inconsistently experienced. Because it is more challenging to construe such treatment as being generally applied to everyone or as having a positive undertone, employees may view these acts as indicating a misfit between themselves and their jobs. As a result, job satisfaction is likely to suffer. Taking together these arguments, we hypothesize a U-shape curvilinear relationship between workplace bullying and job satisfaction. That is, job satisfaction will drop when mid-levels of exposure to bullying behaviors are experienced and will remain high under both low and intense exposure to bullying behaviors (Hypothesis 1). effectively (e.g., Høgh and Dofradottir 2001 ;Zapf and Gross 2001 ). According to transactional stress theories (e.g., Lazarus and Folkman 1984 ), the inability to cope with this situation of being exposed to bullying behaviors at work leads to an excessive physiological activation during a prolonged period of time that triggers the negative effects of stress both at physical level (e.g., psychosomatic complaints or sleep disturbances) and at psychological level (e.g., reduced self-confidence, increased sense of vulnerability, and negative feelings like guilt or shame). Indeed, a recent meta-analysis has shown that employees exposed to bullying behaviors usually suffer health-related problems such as anxiety, depression, or posttraumatic stress symptoms . Longitudinal evidence also indicates victims of bullying have a higher risk of cardiovascular disease and depression compared to non-bullied employees (Brousse et al. 2008 ;Kivimäki et al. 2003 ). Consistent with these findings, we hypothesize that exposure to bullying behaviors is negatively related to employees' psychological well-being: higher exposition to bullying behaviors at work will be related to less psychological well-being (Hypothesis 2).
In addition, the negative consequences of workplace bullying on employees' attitudes and well-being seem to be moderated by employees' appraisals of their working environment. For example, Djurkovic et al. ( 2008 ), in a sample of 335 schoolteachers, found that workplace bullying was more strongly related to employees' intention to leave the organization among participants who perceived low organizational support than among participants who perceived high organizational support (in this condition the effects of bullying on intention to leave were non-significant indeed). In a similar vein, Cooper- Thomas et al. ( 2013 ) indicated that perceived organizational support buffered the relationship of bullying with self-rated job performance in a sample of 727 New Zealand employees from nine healthcare organizations.
As far as health and well-being are concerned, Hoel et al. ( 2004 ) indicated that bullying tends to manifest in mental and physical health problems when a victim attributes intention to harm in the bullies behaviors. Indeed,  concluded, after conducting a longitudinal study in a representative cohort sample of 1,775 Norwegian employees, that "the effect of workplace bullying on subsequent distress is mainly explained by the subjective feeling of being victimized by the bullying, and not by mere exposure to bullying behaviors." (p. 42). Additionally, Faragher et al. ( 2005 ), after conducting a systematic review and meta-analysis of 485 studies, concluded that job satisfaction is a crucial factor influencing employees' psychological well-being. Thus, considering that Italian workers may tolerate negative acts to a certain extent, together with the fact that individuals may not perceive themselves as specific targets of bullying (since bullying behaviors are more widespread in these contexts), it seems reasonable to assume that job satisfaction may buffer the negative impact of bullying behaviors on employees' psychological well-being. In other words, we hypothesize that the relationship between workplace bullying and psychological well-being is moderated by job satisfaction (Hypothesis 3).
In summary, this study addresses the relationship between workplace bullying, job satisfaction, and psychological well-being. In particular, we propose a curvilinear relationship between bullying and job satisfaction, which, in turn, will moderate the linear relationship between bullying and employees' psychological well-being in a sample of Italian workers.

Method Participants
In this study, 1,393 employees participated from 10 medium sized Italian organizations that were spread throughout Italy, representing different organizational sectors (e.g., luxury, sales, and manufacturing). Participants completed a questionnaire that was administered both for research purposes and to fulfill work-related stress obligations imposed by Italian regulations on occupational health and safety (Legislative Decree no. 81/2008 and subsequent amendments). In return, each organization received a report to be included in their "Risk Assessment Document" (Documento di Valutazione dei Rischi).
Most of the participants (44 % female, 56 % male) were working in private sector companies (1,241 vs. 152 in a public organization) and reported job tenure higher than 7 years (59 vs. 41 % with less than 7 years in their actual job). In addition, blue-collar employees represented 65.4 % of our participants, followed by white-collar employees at 27 %, and managers at 7.6 %.

Procedure
Around 50 Italian companies, which are established throughout Italy, were http://eproofing.springer.com/journals/printpage.php?token=fx-zw_rjwUruD5nZaXV_w8MjE39sPbQe3Ieznwa4PVMjIFF-3lpM6ju0Q_KDnV5K 9/26 contacted and selected out of convenience. Ten companies agreed to survey their employees. Thus, after obtaining the permission of the CEOs, employees were contacted and requested to voluntarily participate in the present study. Depending on the size of the organization, the whole organizational population (in organizations with less than 100 employees) or a stratified sample of employees that was representative from the different organizational departments and job positions within the organization (in organizations with more than 100 employees) were involved in the study. Data were collected through paper-and-pencil questionnaires during working hours in rooms provided by the organizations. A research assistant was present during the process to answer any doubts of the participants.
Participants were informed that their responses were anonymous and that completing the questionnaires would take about 30 min, although no time limit was imposed. The mean response rate was 78 %, ranging from 65 to 94 %.

Measures
Exposure to Workplace Bullying The reduced Italian version of the Negative Acts Questionnaire-Revised (NAQ-R: Einarsen et al. 2009 ) validated by Giorgi et al. ( 2011 ) was used to measure the frequency of exposition to 17 specific negative acts (bullying behaviours) at work (response categories were 1: Never, 2: Now and then, 3: Monthly, 4: Weekly, and 5: Daily) within the last 6 months (e.g., "being withheld information which affects your performance").

Job Satisfaction
As Souza-Poza and Souza-Poza ( 2000 ) noted, there are some determinants of job satisfaction that apply to all countries (such as having an interesting job and good relations with management) and others that are country specific (such as pay and job security). Thus, to allow further cross-cultural research, job satisfaction was assessed by using five items from Hartline and Ferrel ( 1996 ) that analyses the satisfaction with different dimensions of work (salary/wage, job security, social support, supervision, and global satisfaction) on a scale from 1 ("very dissatisfied") to 5 ("very satisfied"). Furthermore, this scale was chosen because seems particularly parsimonious and representative of both extrinsic and intrinsic aspects of the job that might be particularly relevant in the actual social and economical Italian context.

Psychological well-being
The Italian version of the 12-items Goldberg's General Health Questionnaire (GHQ-12) was used to assess the perceptions of the employees regarding their general health and psychological well-being (Fraccaroli et al. 1998 ). Each item is rated on a four-point scale (less than usual, no more than usual, rather more than usual, or much more than usual). This questionnaire gives a total score ranging from 0 to 36 when a Likert scoring method is used (0-1-2-3), in which a higher score indicates a greater degree of psychological distress (less psychological well-being or mental health).

Results
Descriptive statistics are given in Table 1 , which shows scale reliabilities according to the Cronbach's alpha on the diagonal. As it can be seen, all measures had a satisfactory internal consistency. To test the relationship between workplace bullying and employees' job satisfaction and psychological well-being, two different hierarchical regression analyses were conducted, in each one the squared term for workplace bullying (negative acts) was computed to test for possible curvilinear effects (for similar procedures, see De Dreu 2006 ; Janssen 2001 ). Job satisfaction and psychological well-being (measured with the GHQ) were, respectively, introduced as the dependent variable in each regression. Then, demographical variables, dummy codified following the suggestions provided by Aiken and West ( 1991 ), were entered as control variables in the first step (Model 1). Workplace bullying was entered in Step 2 (Model 2), and the squared term for workplace bullying (negative acts) was entered in the third step (Model 3).
In the case of job satisfaction, results are summarized in Table 2 . As it can be seen, the control variables explained a significant proportion of variance in job satisfaction scores [R = .02, F(4,1059) = 6.39, p < .001]. In that sense, blue-collars reported less job satisfaction (β = −.14, t = −4.28, p < .001) than white-collar and managers. Adding the linear term for workplace bullying in Model 2 produced a significant effect that explained a 26 % increase of variance in job satisfaction [∆R = .26, F(1,1058) = 376.64,p < .001]. Thus, the linear term for workplace bullying was negatively and significantly related to job satisfaction (β = −.51, t = −19.41, p < .001). Finally, the quadratic term for workplace bullying added in the third step produced a significant increase in explained variance [∆R = .04, F(1,1057) = 62.72, p < .001]. Contrary to the previous linear model, the squared term for workplace bullying was positively and significantly related to job satisfaction (β = .97, t = 7.92, p < .001). In addition, the relationship between being a blue-collar worker and job satisfaction remained significant and negative across all models, suggesting a main effect of job position on job satisfaction. Step 1 (control variables)  Step 2 (linear effect) Negative acts --−0.80* 0.04 −2.31* 0.19 Step 3  Unstandardized coefficients, the standardized beta coefficients that were significant are reported in the text; *p < .001 Overall, these results supported Hypothesis 1, suggesting that, whereas an increase from low to moderate levels of negative acts at work (workplace bullying) is associated with a decrease in job satisfaction, job satisfaction increases and is not negatively affected at relatively high levels of workplace bullying. Indeed, the quadratic model is summarized in Fig. 1 , which gives the predicted outcomes on the basis of the regression equation: y(x) = ax + bx +c; where a = .4; b = −2.3; and c = 6 (see Table 2 ). Regarding employees' psychological well-being, as can be seen on Table 3 , the control variables did not contribute to explain variance in employees' psychological well-being. Adding the linear term for workplace bullying (β = .53, t = 21.46, p < .001) explained a significant proportion of variance in psychological well-being scores, R = .29, F(5,1165) = 94.50, p < .001. In addition, gender (β = −.05, t = −2.07, p < .05) and job position "b" (β = −.06, t = −2.38, p < .05) significantly predicted psychological well-being scores, suggesting that men reported a better mental health than women (since higher scores in the GHQ means lower psychological well-being and vice versa) as well as managers reported a better mental health than bluecollar and white-collar employees. On the other hand, the quadratic term for workplace bullying introduced in Model 3 did not further explain the variance in psychological well-being scores. Therefore, according to Hypothesis 2, results supported a linear relationship between workplace bullying and psychological well-being, in which higher exposition to negative acts at work is related to a decrease in psychological well-being. Step 2 (linear effect) Negative acts --6.45** 0.30 8.62** 1.44 Step 3  Control variables were dummy categorized: gender (1 = male; 0 = female); job tenure (1 = less than 7 years; 0 = more than 7 years); job position (1 = manager; 0 = white-collar; 0 = blue-collar); job position (0 = manager; 0 = white-collar; 1 = blue-collar) Unstandardized coefficients, the standardized beta coefficients that were significant are reported in the text; *p < .05; **p < .001 Finally, a hierarchical regression analysis was performed to assess whether or not job satisfaction moderates the relationship between exposure to bullying behaviors and psychological well-being. Psychological well-being was introduced as a criterion variable. Then, after centering the independent variables, job satisfaction was introduced in the first step, exposure to negative acts in the second step, and the interaction effect between job satisfaction and negative acts in the third step. Our results suggested that both job satisfaction (β = −.20, t = −7.39, p < .001) and exposure to negative acts (β = .41, t = 15.03, p < .001) have a main effect on psychological well-being. Considering the nature of mental health scores, a higher job satisfaction is related to higher levels of psychological well-being; whereas higher exposure to bullying behaviors is related to lower levels of psychological well-being. In contrast, the interaction effect was not significant, which did not support Hypothesis 3.

Discussion
Previous theorizing has suggested that the attributions of workers exposed to bullying behaviors determine their reactions to such mistreatment at work (e.g., Bowling  considering a contextual perspective, our results confirm the expected curvilinear relationship between workplace bullying and job satisfaction. That is, job satisfaction decreases under higher exposure to bullying behaviors, but only up to a certain point, in which intense exposure to bullying behaviors is related to high levels of job satisfaction. Although this result seems to contradict previous studies in domain of bullying (e.g., Bowling and Beehr 2006 ;Loh et al. 2010 ), cultural factors together with coping and motivation literatures can help to explain this apparently counterintuitive finding. Moreover, Johns ( 2006 ) emphasized the necessity of contextualizing studies in organizational behavior because the context plays a key role in explaining study-to-study variation. For example, one of Johns' basic assumptions is that context can prompt curvilinear effects because the organizational context provides "situational opportunities and constraints that affect the occurrence and meaning of organizational behavior as well as functional relationships between variables" (p. 386). Indeed, several studies that have compared non-linear and linear models have noticed that non-linear models usually have a better data fit than linear models in organizational contexts (e.g., Borg et al. 2000 ;Karanika-Murray et al. 2009 ).
Returning to the role of the context in low human orientation countries where there are a very high prevalence of bullying exposure such as Italy (around 16 % according to Giorgi 2009 ;Giorgi et al. 2011 ) or Turkey (more than 40 % according to Bilgel et al. 2006 ;Yildiz et al. 2008 ), negative acts are usually very widespread and are accepted rather than not tolerated and discouraged. Thus, after a certain point of experiencing negative acts in an organizational culture that tolerates bullying, individuals might interpret bullying as less dissatisfying that thought before. In that sense, taking into consideration macro-level variables, such as the current negative socioeconomic situation and the high unemployment rates in some countries, negative behaviors at work may be more accepted. Under difficult economic situations, negative acts might become spread and more subtle and ambiguous and, therefore, is more difficult to clearly perceive the intention to harm behind these bullying behaviors (Hoel and Beale 2006 ). Furthermore, employees might accept bullying if it is functional to their career development or job stability, particularly under financial crisis circumstances in which extrinsic motivation seems to be more important than intrinsic motivation at work (Yildiz et al. 2008  Our findings are also in line with the premises of few empirical studies that suggest potential positive effects of workplace bullying on employee performance under specific situations and contexts. For example, a recent Korean study (Lee et al. 2013 ) demonstrated a curvilinear (inverted Ushaped) relationship between supervisors abusive behaviors and employees creative performance. Particularly, employees obtained higher creativity scores when abusive supervision was at a moderate level rather than at very low or very high levels. Accordingly, Ferris et al. ( 2007 ) found a positive side of bullying suggesting that "the leader often needs to engage in coercive power" with employees who are "unable and unwilling to take responsibility" (p. 201). Correspondently, Ma et al. ( 2004 ) highlighted that leader's sarcastic remarks or tyrannical leadership might be associated positively with job performance and productivity.
On the other hand, our results confirm that higher exposure to bullying behaviors is related to lower levels of mental health (measured as employees' perceptions of general psychological well-being), which is in line with past research that have highlighted the severe negative consequences of workplace bullying on both employees' physical and psychological health (e.g., Høgh et al. 2012 ;. It is interesting to notice that in our study men and managers reported better mental health that women and employees in lower job positions. This finding is congruent with previous literature that has shown that: (a) women reported more negative health effects, such as post-traumatic symptoms, as a result of exposure to bullying behaviors (e.g., Rodríguez-Muñoz et al. 2010 ); and (b) perpetrators of bullying are likely to be men and employees in managerial positions than women and employees in other job positions (e.g., Hauge et al. 2009 ;Lee and Brotheridge 2011 ). Thus, interventions to counteract workplace bullying should take into account the job position and gender of perpetrators and victims since these variables seem to play a pivotal role on how workplace bullying is perceived (e.g., Escartín et al. 2011b, Hauge et al. 2009Lee and Brotheridge 2011 ).
Finally, our results revealed that perceptions of job satisfaction do not moderate the bullying-mental health relationship, suggesting that exposure to bullying behaviors has a detrimental effect on health regardless of the employees' job satisfaction. As Vie et al. ( 2011 ) reported, perceiving oneself as a victim of bullying moderated the relationship between exposure to bullying behaviors and health outcomes only in cases of low exposure; whereas high exposure to bullying behaviors was related to higher levels of health complaints that were independent of reporting of being a victim of bullying or not.

Limitations and further research
Although our findings may promisingly contribute to explain bullying reactions, we should note some limitations of our study that further research needs to overcome by: (a) using different methods to triangulate the data since our study relied exclusively on self-report measures (e.g., social network and sociometric analyses to better capture target attributions: see Coyne et al. 2004 ; or including self-evaluations to determine who is victim of bullying: see Leon-Perez et al. 2013 ); and (b) gathering longitudinal data in order to infer causality since it is also possible that job satisfaction played a potential suppression role in our study, thus those employees that indicated a high job satisfaction may not perceive negative acts at work because they can cope with them or because they are particularly satisfied with their job. Indeed, time can play an important role in the bullying-satisfaction relationship. For example, Rodríguez-Muñoz et al. ( 2009 ) conducted two studies among Belgian workers using two-wave panel designs with different lengths of time (2 years vs. 6 months). When the time lag between measures was 2 years, workplace bullying (T1) was negatively related to subsequent job satisfaction (T2); however, when the time lag between measures was 6 months there was no significant effect of workplace bullying (T1) on job satisfaction (T2). In a similar vein, spillover hypotheses between bullying and health should be longitudinally assessed since different studies have shown that harassment processes and health can mutually influence each other (Høgh et al. 2012 ; ).
In addition, future research should also replicate our findings in other cultural contexts, considering for instance lower power distance countries. Power distance refers to the degree "to which a society accepts the fact that power in institutions and organizations is distributed unequally" (Hofstede 1980 : p. 45), which might have a strong implication in the bullying-job satisfaction relationship as well as the acceptability of bullying behaviors (see Power et al. 2013 ).

Theoretical and practical implications
This study shows that a curvilinear model outperformed linear models to explain the relationship between workplace bullying and job satisfaction. Although this study is not exempt from limitations, the results highlight the importance of contextualization on explaining workplace bullying, which may have important implications for theory and practice. First, our study revealed the existence of a group of employees who are simultaneously bullied and satisfied, suggesting that workplace bullying can be embedded in the organizational culture and is tolerated in some organizations and, in turn, workplace bullying does not negatively affect employees' job satisfaction. These findings have important implications for research on target reactions to bullying, which is still in its early stages (e.g., Leck andGalperin 2006 , Samnani 2013 ). In this regard, future research should compare the effects of bullying among organizational cultures after understanding for the frequency of negative acts within the organizational culture (Power et al. 2013 ;Samnani 2013 ). For example, victims of bullying can perceive negative acts as a systemic organizational issue when bullying is widely spread and, therefore, they may reconstruct and reinterpret bullying behaviors as necessary or even satisfying, ascribing, for instance, alternative motives to the perpetrator (Anand et al. 2004 ;Samnani 2013 ).
Second, derived from these results, we can argue that it is important to identify employees who are satisfied or unsatisfied at work for accurately recognizing the bullying risks and formulate appropriate solutions. Indeed, job satisfaction may relate to whether employees that are bullied seek assistance or react passively. For example, satisfied workers that received negative acts can be considered as "easy" targets because of the lack of potential repercussions. Moreover, tolerating bullying behaviors may be a risk to employees' health and may facilitate perceiving other types of harassment and violence at work as acceptable as well, promoting the development of spiraling negative effects of bullying behaviors in organizations (e.g., Giorgi 2012 ;. However, future studies should test whether victims' perceptions of a high prevalence and reoccurrence of negative acts is threatening or at a certain point does it become a resilient factor (van Heugten 2013 ).
Finally, from a managerial point of view, organizations aiming to reduce bullying should consider more in depth the cultural and the organizational context that drive these behaviors, and, in turn, fostering more positive ethical climates, so that when their employees are confronted with the bullying experience, they know how to deal with it without considering it satisfying. Particularly, organizations should develop codes of acceptable behaviors and anti-bullying policies, inform about the risk of tolerating bullying and highlight the importance of adhering to formal codes of conduct (i.e., establishing a code of ethics). Furthermore, in such organizations where employees are satisfied even when they are exposed to frequent negative acts, training to raise awareness about the harmful consequences of bullying and their unethical implications for dignity need to be introduced in order to change employees' attitudes and, in turn, reduce the prevalence and the acceptability of bullying.

Conclusion
The exploration of a nonlinear relationship of between bullying withand job satisfaction seems interesting in ordercan help to explain how negative acts are still tolerated in spite of health consequences. In that sense, our findings have important implications for the emerging research issue of bullying acceptability (i.e., understanding why people withstand or tolerate this kind of negative and unethical climate from a contextual perspective is a compelling endeavor: see also, Escartín et al. 2011a ;Power et al. 2013 ).
Our results show that whether bullying might be to a certain extent satisfying, it is not healthy since the relationship between bullying and health is linear. Employees, who tolerate too much bullying, even if they do not appraise negative acts so threatening to be unsatisfied, can impair their psychological well-being. Consequently, the full understanding of the acceptability of bullying behaviors in the workplace and its diverse implications can help organizations to counteract bullying in a more effective way. AQ3