The role of leadership practices in the relationship between role stressors and exposure to bullying behaviours – a longitudinal moderated mediation design

ABSTRACT Role conflicts and role ambiguity have been identified as important risk factors for exposure to workplace bullying, particularly when combined with inadequate leadership practices. Even though role ambiguity theoretically can be considered a causal precursor to role conflicts, previous research has mainly examined these role stressors as concurrent predictors of workplace bullying. The present study provides a more nuanced analysis by investigating role conflicts as a mediator in the relationship between role ambiguity and exposure to bullying behaviours. Adding to the understanding of the bullying process we also considered the possible moderating roles of laissez-faire and transformational leadership in the role stressor–bullying relationship. Employing a national probability sample of 1,164 Norwegian workers, with three measurements across a 12-month period, the results showed an indirect effect of employees’ role ambiguity on subsequent exposure to bullying behaviours through employees’ experience of role conflicts. Moreover, laissez-faire leadership exacerbated, while transformational leadership attenuated, the indirect relationship between role ambiguity and exposure to bullying behaviours through role conflicts. In summary, the present data shows that when the management of organisations neglects its inherent responsibility to adequately address employees’ experiences of role ambiguity and role conflicts, the risk of exposure to workplace bullying is likely to increase.

in the working environment, as well as employ those preventive measures that may stop, or at least halt, this detrimental situation.Stressful working environments, combined with inadequate leadership practices, are assumed to be the most prominent risk factors in this regard (Nielsen & Einarsen, 2018;Salin & Hoel, 2020).Specifically, exposure to bullying seems to be particularly prevalent in workplaces where employees experience organisational constraints and contradictory expectations and demands in the form of role ambiguity and role conflicts (Bowling & Beehr, 2006;Reknes et al., 2019;Skogstad et al., 2007;Van den Brande et al., 2016), as such reports possibly signify ambient stressors in the working environment that may afflict both targets and perpetrators alike (Hauge et al., 2011;Skogstad et al., 2011).Yet even more, being plagued with role stressors may make employees more vulnerable targets as they may fail to live up to expectations, breach norms of polite and respectful conduct, meet disapproval and negative reactions from a range of sources, while also being more vulnerable and less able to defend against the aggression and incivility of others.
Although most previous research has investigated role ambiguity and role conflicts as concurrent predictors of organisational outcomes (cf.Beehr, 1995), including exposure to workplace bullying, some theoretical notions and cross-sectional evidence suggest that the effect of role ambiguity on workplace bullying may be mediated through role conflicts (Hartenian et al., 1994;Notelaers et al., 2010).The present prospective study expands this research by testing the hypothesis that the experience of role conflicts is a prominent mechanism in explaining the relationship between role ambiguity and bullying.Building on the so-called "work environment hypothesis" of workplace bullying (Einarsen et al., 1994;Einarsen et al., 2020) in which the lack of leader involvement in stressful situations is argued to fuel and escalate the bullying process (Leymann, 1996) we also address whether laissez-faire and transformational leadership practices moderate the impact of the two role stressors on the risk being exposed to bullying behaviours.
The present study makes three important contributions to research on workplace bullying.First, by testing role conflicts as an intermediate variable in the relationship between role ambiguity and exposure to bullying behaviours, we examine a mechanism that can explain how role stressors relate to workplace bullying, in our case as perceived by targets.Such a mechanism has not previously been examined within a longitudinal research design and may thus have significant theoretical as well as practical implications.By knowing more about the specific processes via which stressors in the work environment are transformed into workplace bullying, organisations should be able to initiate the appropriate interventions needed to prevent workplace bullying from arising and developing.Second, by examining laissez-faire and transformational leadership as potential moderators of the association between role stressors and bullying behaviours we answer the call for more research on organisational conditions that may facilitate or prevent the development of workplace bullying (Rai & Agarwal, 2018).Finally, the use of prospective data from a 3wave national probability employee sample with 6 months between time-points allows us to substantiate a plausible causal direction between the study variables.

Theoretical background
Bullying in the workplace refers to the systematic and repeated exposure to negative behaviours at the workplace by other organisation members, taking place over a prolonged time period, in situations or instances in which the target(s) find it difficult to defend themselves, potentially leading to severe victimisation of the target over time (Einarsen et al., 2020).Hence, the concept of workplace bullying has a target perspective, where it is the total toll of the exposurepotentially from many sourcesthat is at the heart of the matter, more so than focusing on isolated perpetrators and their individual acts of incivility and aggression.Furthermore, exposure to bullying occurs on a dimension from occasional exposure to severe exposure (Reknes et al., 2021), supporting the notion that bullying exposure constitutes a gradually escalating process (Einarsen et al., 2020;Escartín et al., 2013).We may distinguish between "victimization" when looking at individuals who have been subjected to frequent and severe long term bullying, and "exposure to bullying behaviours' when including the whole range of such negative social experiencesfrom the occasional negative acts up to, and including, being a victim of severe bullying under the more strict formal definition (Nielsen et al., 2011).
Both work-and person-related factors may trigger processes that results in systematic exposure to workplace bullying (Nielsen & Einarsen, 2018).However, a growing body of evidence substantiates that role ambiguity and role conflicts are among the main risk factors in this regard (Hauge et al., 2007;Van den Brande et al., 2016).In the 1960s, role theory was developed to provide insight into the processes that affect the physical and emotional state of an individual in the workplace and, in turn, their workplace behaviour (Kahn et al., 1964).According to Katz and Kahn (1978), roles will contribute to the stability and consistency of organisations by giving direction to expected behaviour at work.However, when these roles are not well understood, or an individual must balance several different roles in an organisation which may be connected to conflicting sets of expectations, role stress in the form of role ambiguity and/or role conflicts may be experienced.Conceptually, role ambiguity denotes a lack of information or a lack of clarity in communication regarding what is expected of a given employee in a given position in the organisation (Kahn et al., 1964;King & King, 1990).More specifically, when individuals do not know what their responsibilities are, or do not have a clear understanding of what is expected of them, they may be experiencing role ambiguity.In contrast, role conflict refers to incompatible expectations and demands associated with a certain role, such that compliance with one set of expectations makes compliance with the other sets more difficult (Kahn et al., 1964;King & King, 1990).While many different forms of role conflicts have been identified, most are conceptualised as a subcategory of either inter-role conflicts (incompatible demands placed on an individual occupying multiple positions or several roles simultaneously) or intra-role conflicts (conflictual expectations associated with a single position or role) (Beehr, 1995;Kahn et al., 1964), where intra-role conflicts is the primary focus in the present study.An example of an intra-role conflict would be when a person is faced with incompatible requests from two co-workers at the same time, where one request cannot be carried out without it being at the expense of the other.According to Kahn and colleagues (1964), organisational members need sufficient information about what is expected of them to effectively carry out their work.Clearly defined roles help leaders to make followers responsible for their own task executions and achievements.Unclear descriptions of a given role may, in contrast, lead to experiences of stress, with associated negative consequences for both the individual, the workgroup, and the organisation (Katz & Kahn, 1978).In this, there is a risk of contradicting expectations, interpersonal tensions, and destructive social processes potentially leading to workplace bullying (Baillien & De Witte, 2009b;Hauge et al., 2007;Moreno-Jiménez et al., 2009).Supporting this view, Bowling and Beehr (2006), in a comprehensive meta-analysis including 90 independent samples, showed role conflicts and role ambiguity to be among the strongest predictors of bullying and harassment.Later studies employing prospective designs have further supported this notion (Reknes et al., 2014;Reknes et al., 2019).

Role stressors and workplace bullying
Following a social-interactionist perspective (Felson, 1992;Felson & Tedeschi, 1993), a stressful work environment is likely to create feelings of tension and frustration among employees, which may affect targets-to-be.That is, stressors in the workplace increase individuals' vulnerability by triggering emotional, cognitive, and behavioural changes, which may lead to aggressive outlets and breaches of established workplace norms for expected polite social interactionas well as less competent work performances (Baillien et al., 2011;Reknes et al., 2014).This, in turn, may cause others to react or even retaliate in such a way that the targeted employee reasonably perceives him or herself to be exposed to bullying behaviours (Neuman & Baron, 2011).These negative reactions may possibly arise from a range of colleagues and superiors.When accumulated, however, these more or less isolated negative responses may be perceived as a massive negative bullying situation by the target.Moreover, the experience of being in a role conflict denotes a situation where others have different and even conflicting expectations and therefore may pose sanctions on the focal person and target-to-be.Again, this may put a heavy burden on the target and thus explain the reported negative behaviour.
In line with this theoretical notion, role stressors are documented to be consistently related to self-reported strain, including anxiety, tension, anger, and depression (Jackson & Schuler, 1985;Örtqvist & Wincent, 2006;Spector & Goh, 2001).More specifically, role ambiguity has been hypothesised to increase feelings of stress because concerns and doubts about how to proceed with critical tasks may lead to frustration, resulting in increased tension (Schaubroeck et al., 1989).Role conflicts may have similar effects in the form of frustration and anxiety, because incompatible expectations from role senders diminish the focal person's self-perception of competence and effectiveness (Schaubroeck et al., 1993), and probably reduces one's experiences of decision latitude and self-control.Accordingly, one may both be unsure about how to execute one's job tasks as well as when to perform them.The uncertainty and ambiguity associated with such competing expectations inherent in role conflicts are expected not only to lead to work stress but also to interpersonal tensions and conflicts (Tidd & Friedman, 2002).
On top of this, targets' reports of role ambiguity and role conflicts may reflect ambient work stressors which may afflict most employees in the given environment.Hence, working in a stressful environment may, in general, be associated with tense as well as frustrated employees, who engage in more negative and aggressive behaviour towards others.This may particularly hold true when conflicting expectations and demands are present, thus also triggering escalating conflicts which ultimately may result in bullying (Balducci et al., 2012;Bowling & Beehr, 2006).This line of reasoning corresponds with the stressor-emotion model of counterproductive work behaviour (Spector & Fox, 2005), which states that stressful events will induce negative emotions in some or all affected individuals.Central to this model is the individual's appraisal of his or her abilities to cope with the given challenges (see also Lazarus, 1999;Perrewé & Zellars, 1999).Accordingly, this model posits that engaging in negative social behaviour towards others may in fact be an emotion-based response to stressful environmental conditions experienced by all employees in the collective work environment (Balducci et al., 2012).More specifically, a negative appraisal of a stressful situation triggers a negative emotion (i.e.anger or anxiety) which, in turn, is linked to a range of strain responsesincluding engaging in negative acts and bullying behaviours.Therefore, role stressors may result in more frequent negative behaviour in the working environment by triggering the effect of environmental factors on frustration and aggressive behaviour in general.
Taken together, employees experiencing high levels of role ambiguity and role conflicts probably also experience more frequent negative social acts from others who either react to them or who oppose their role enactment, as these stressors probably trigger (directly and/or indirectly) perceived negative behaviour from a range of sources that may mount a perceived bullying situation (Hauge et al., 2009).Accordingly, many of the stressors in the work environment that provoke aggression in perpetrators may be similar to those often reported by victims of bullying (Bowling & Beehr, 2006).Based on these theoretical notions and empirical findings, we state the following hypotheses: Hypothesis 1a: Reported role ambiguity is positively associated with subsequent reports of exposure to bullying behaviours at work.Hypothesis 1b: Reported role conflicts are positively associated with subsequent reports of exposure to bullying behaviours at work.

The relation between role ambiguity and role conflict
In line with Kahn and colleagues' (1964) early work on role ambiguity and role conflictsstating that "[…] conflict and ambiguity are independent sources of stress; either or both of them may be present in any given role" (p.89)the majority of research on role stress has to date treated role ambiguity and role conflicts as closely related yet separate and parallel constructs (Bowling et al., 2017;Fisher & Gitelson, 1983;Jackson & Schuler, 1985;King & King, 1990;Örtqvist & Wincent, 2006;Tubre & Collins, 2000).Accordingly, most studies on role stressors and workplace bullying have treated these two stressors as concurrent predictors of bullying (Bowling & Beehr, 2006).However, their concurrent and temporal interrelationships are still of relevance both from a theoretical and an applied perspective.In their original descriptions of the two role constructs, Kahn and colleagues (1964) were somewhat vague about how they might relate to each other (King & King, 1990).Nevertheless, the authors did argue that if a role is ambiguous for the focal person (i.e. the receiver of role expectations and demands), it is probably so for many of his/her role senders as well (Kahn et al., 1964).Furthermore, if a role is ambiguous, the focal person will have to fill in the missing information themselves.This may, in turn, be opposed and resisted by others, hence creating a state of intra-role conflict in the focal person.Similarly, in situations where role senders are vague or inconsistent in their expectations and demandsas captured in the concept of role ambiguitythe focal person may over time be increasingly more likely to receive conflicting pressures from colleagues and superiors.This is likely to result in experiences of intra-role conflicts for the focal person, which could, in the long run, be met with sanctions of varying intensity and legitimacy from frustrated role senders.Such reactions may even take the form of negative social acts escalating into bullying behaviours and as such spurring a bullying process.Thus, there is reason to believe that a working environment characterised by high levels of role ambiguity will create a fertile ground for intra-role conflicts to develop which again may increase the risk of exposure to bullying behaviours.
Despite this line of reasoning in early role theory (c.f.Kahn et al., 1964), studies examining causal relationships between role ambiguity and role conflicts and their outcomes, are scarce.While various studies show relatively moderate correlations between the two (see e.g.Skogstad et al., 2007, documenting a correlation of .35 in a representative sample of the Norwegian working population), findings from two cross-sectional studies suggest that the effect of role ambiguity on workplace bullying is indeed mediated by role conflicts (Hartenian et al., 1994;Notelaers et al., 2010).However, to our knowledge, no studies have examined their relationships longitudinally.Therefore, there is a need for studies investigating how these frequently studied role stressors relate to each other as antecedents in the developmental process of workplace bullying (Notelaers et al., 2010).In line with the theoretical reasoning found in early role theory, we propose that employees' experienced role ambiguity over time will lead to an increase in experienced intra-role conflicts.More specifically, we assume that employees who are uncertain about their role and what is expected of them at work are likely to be exposed to conflicting expectations and demands from other individuals in the work environment.Without access to the necessary role information, individuals are likely unable to adjust to these conflicting expectations, resulting in increased experiences of intra-role conflict.They may also develop their role in a more self-serving direction or develop goals and aims that turn out to be in opposition to those of important others in the organisation, laying the ground for more role-conflicts.This increase in role conflict, in turn, is likely to increase their feelings of anxiety and frustration, leading the individuals to change their behaviour in such a way that triggers a bullying process (i.e. the social interactionist theory; Felson & Tedeschi, 1993;Tedeschi & Felson, 1994).In support of such a notion, most studies on role stressors as antecedents of workplace bullying have shown role conflicts to be the stronger predictor over that of role ambiguity (Bowling & Beehr, 2006;Van den Brande et al., 2016), which may indicate it to be the most proximal antecedent of the two (Notelaers et al., 2010).Taken together, we propose that the relationship between role ambiguity and bullying behaviours is mediated by role conflicts.
Hypothesis 2: Reported role conflicts mediate the positive association between reported role ambiguity and subsequent reported exposure to bullying behaviours.

Leadership behaviour
Having proposed how role ambiguity may lead to bullying behaviours through role conflicts, we will now turn to leadership practices as a conditional factor that may moderate the effect of these role stressors on exposure to bullying.Bullying is unlikely to be explained by one factor alone and is probably a result of an interplay between various factors acting together in the work arena (Zapf, 1999).In line with the work environment hypothesis, bullying is caused by a combination of deficiencies in work design (also manifested in role stressors)causing stress and frustration among employeesand deficiencies in the leadership practices of immediate and more senior managers (see Einarsen et al., 2020).Accordingly, the combined impact of the two role stressors and the leadership style of one's immediate superior is likely to influence employees during their interactions (Cooper et al., 2001).
Leadership characteristics and styles have been shown to be linked to a wide variety of follower outcomes, both positive and negative (Harms et al., 2017;Kelloway & Barling, 2010), including exposure to workplace bullying (Hoel et al., 2010).The full range of leadership model (FLR; Bass & Avolio, 1994;Bass & Riggio, 2006), describes leadership behaviours and styles from the most passive and ineffective to the most active and effective styles of leadership.In line with this, we elaborate on the potential moderating effects of laissez-faire and transformational leadership, respectively, on the proposed relationships between the two role stressors and being exposed to bullying behaviours.These two leadership styles represent the extremes of the full range of leadership model, where transformational leadership is described as the most constructive and effective style of leadership.Laissez-faire leadership, described as the most passive and ineffective (Bass & Avolio, 1994), has even been argued and substantiated to be a destructive form of leadership (Craig & Kaiser, 2013;Fosse et al., 2019;Skogstad et al., 2017).
Based on Bass and Avolio's (1994) model and operationalisation of laissez-faire leadership (see also Bass, 1999;Hinkin & Schriesheim, 2008), Skogstad and colleagues (2014) have defined laissez-faire as "a follower-centred form of avoidance based leadership by focusing on subordinates' perceived situational need for leadership" (p.325).In line with this definition, laissez-faire leaders are absent, passive, and/or avoidant when followers are in need of leader help or assistance (Skogstad et al., 2014).In addition to being perceived as non-responsive, laissez-faire leadership behaviour may in such a context be conceived as a passive form of aggression (Buss, 1961;Parrott & Giancola, 2007).It may even be perceived as a type of social exclusion and ostracism (Fiset & Boies, 2018;Robinson & Schabram, 2017;Williams, 2007) often seen in cases of workplace bullying (Notelaers et al., 2019).Laissez-faire leadership is not restricted to only being ineffective for the organisation, but may in certain contexts also be experienced as a destructive type of leadership in and of itself, be it by subordinates' emotional and cognitive experiences or by their consequences (Skogstad et al., 2017).
One explanation for why laissez-faire leadership should influence the magnitude of the association between role stressors and exposure to workplace bullying is that laissez-faire leadership may serve to isolate and exclude individuals, create uncertainty and deny access to social support, thereby exacerbating the negative effects of other workplace stressors (Kelloway et al., 2005;Robinson et al., 2013).Such passive-avoidant leadership behaviours may increase subordinates' felt tension from existing stressorscausing employees to feel even more frustrated, anxious, and angry because of their superior's reluctance to provide direction and structure, failure to clarify expectations, and unwillingness to help solve problems (Donaldson-Feilder et al., 2013).Furthermore, passive and avoidant leaders do not actively manage task and interpersonal processes in such a way as to minimise experienced role conflicts among members of their organisation or team (Barling & Frone, 2017).In this, they neglect the legitimate responsibilities of their leader position and role to adequately address and amend stressful and conflicting working conditions, which is particularly troublesome when subordinates are in need of help and assistance (Skogstad et al., 2014).Thus, there is reason to believe that passive and avoidant laissez-faire leaders will increase the likelihood that followers' experiences of role conflicts go unmanaged, allowing them to escalate over time and, ultimately, even develop into workplace bullying.This line of reasoning is in accordance with the work environment hypothesis, where the lack of leader intervention in cases of unfavourable working conditions is hypothesised to create high-risk environments in which bullying may flourish (Einarsen et al., 1994;Einarsen et al., 2020;Leymann, 1996).To our knowledge, the moderating effect of laissez-faire on the relationship between role conflicts and bullying in the workplace has only been examined in one cross-sectional (Hauge et al., 2007) and one prospective (Rodriguez-Munoz et al., 2012) study.Both studies showed that the relationship between role conflicts and bullying is stronger for those respondents who reported high levels of laissez-faire leadership from their immediate supervisor.Thus, bullying seems to be more likely to occur when supervisors avoid or neglect intervening in stressful situations (Ågotnes et al., 2018), including working conditions characterised by high levels of role stressors.To our knowledge, no previous study has investigated the potential moderating effect of laissez-faire leadership on the relationship between role ambiguity and role conflicts.However, there are convincing arguments to be made for laissez-faire leadership behaviour to also influence this relationship.For example, laissez-faire leaders who do not intervene in ambiguous situations or do not provide their followers with the necessary information about their role when needed, are likely to leave these followers vulnerable to experiencing conflicting demands from different people in the organisation.Accordingly, there is no obvious theoretical reason to expect laissez-faire leadership to have a stronger moderating effect on either of the paths present in the relationship between role ambiguity and exposure to bullying behaviours, through role conflicts.Therefore, we propose that laissez-faire leadership will moderate this proposed indirect relationship, yet without specifying on which path we expect laissez-faire leadership to have the most impact.Accordingly, we propose the following explorative hypothesis: Hypothesis 3: The positive indirect effect of reported role ambiguity on subsequent reported exposure to bullying behaviours through role conflicts is stronger at high (vs.low) levels of reported laissez-faire leadership.
Contrasting laissez-faire leadership, transformational leadership is considered, and has been shown to be, an effective and particularly beneficial form of leadership (Arnold, 2017), even if also criticised and questioned by its measurement (e.g.Van Knippenberg & Sitkin, 2013).Theoretically, transformational leaders serve as role models for their employees, inspire and motivate their followers to do their best by providing meaningful and challenging work tasks, communicating a shared vision, and stimulating followers' innovation and creativity.Finally, and perhaps most important and relevant for the present study, transformational leaders show consideration for their followers' individual needs, concerns and well-being, through individualised consideration (Bass, 1985(Bass, , 1990b)).In this, transformational leadership behaviour includes the enactment of social support, whereby the leader provides both emotional (encouragement) and instrumental (advice) support (Kessler et al., 2013).In many instances, such supportive leadership makes the difference for employees when coping with work stress (Breevaart et al., 2014;Khalid et al., 2012).In strong contrast to laissez-faire leaders, transformational leaders pay individual attention to each of their followers' needs and should therefore be able to recognise and help followers who are faced with ambiguous or conflicting expectations and demands.Once a given leader is aware of such challenging working conditions, he or she is also likely to actively assist a given follower in managing the actual demands more effectively, including their experiences of role ambiguity and role conflicts (see also Bass & Riggio, 2006).In turn, followers will probably experience fewer negative emotionssuch as frustration and/or anxietyand their associated negative behavioural consequences (c.f. the social interactionist theory, Felson, 1992;Felson & Tedeschi, 1993), consequently buffering the risk of exposure to bullying in the workplace (Astrauskaite et al., 2015).Moreover, by showing effective conflict management skills (Baillien et al., 2011;Zapf & Gross, 2001), transformational leaders will probably deter the occurrence of follower negative behaviours and, ultimately, workplace bullying to take place (Baillien & De Witte, 2009a), especially in situations characterised by high levels of perceived role stress among subordinates.However, as with laissez-faire leadership, we do not have a clear expectation for where transformational leadership is likely to have the largest impact on the indirect effect of role ambiguity on exposure to bullying behaviours through role conflicts.It could be argued that transformational leaders, by communicating a clear and positive vision, as well as developing and supporting their staff, could attenuate the negative impact of role ambiguity by making it clear what is expected of them, thereby leading to a reduction in experienced role conflicts.However, transformational leaders also empower and support their followers and lead by example, meaning that these leaders may help their employers to cope with stressful situations without experiencing the negative emotions and subsequent behavioural consequences associated with increasing role conflicts.Additionally, transformational leadership has been associated with a problem-solving style of conflict management (Saeed et al., 2014), meaning that these leaders are likely to be more effective in conflict situations (including instances of role conflicts) because they explicitly pursue a satisfactory outcome that fully satisfies everyone's concerns.This style of conflict management has been associated with fewer instances of reported and enacted bullying in the workplace (Baillien et al., 2014;Baillien & De Witte, 2009a).As such, transformational leaders may also act as a buffer against exposure to bullying behaviours for those employees experiencing role conflicts.Accordingly, we propose the following explorative hypothesis: Hypothesis 4: The positive indirect effect of reported role ambiguity on subsequent reported exposure to bullying behaviours through role conflicts is weaker at high (vs.low) levels of reported transformational leadership.
In Figure 1, we summarise the proposed relationships and hypotheses to be investigated in the present study.

Design and procedure
The sample was based on data from a three-wave representative survey of the Norwegian working force, with a six-month time lag between measurement points.A random sample of 5,000 employees was drawn from The Norwegian Central Employee Register by Statistics Norway, which is the official register of all Norwegian employees as reported by employers.The sampling criteria were adults between 18 and 60 years of age employed in a Norwegian enterprise of more than 5 employees.The first wave of data (T1) was collected during the spring of 2015.Questionnaires were distributed through the Norwegian Postal Service.Altogether 1,608 questionnaires were satisfactorily completed and included in this study, yielding a response rate of 32 percent.The mean age was 45.17 (SD = 10.02) years with a range from 21 to 61.The sample consisted of slightly more women (52%) than men (48%).Altogether 36% had a leadership position with personnel responsibilities, indicating an overrepresentation of leaders and managers in the sample.
The second wave of data (T2) was collected six months later following the same procedure as the first wave.There were no changes to the survey questionnaire.Only respondents who responded to the T1 survey were invited to participate at T2.Altogether 1,149 respondents participated in this follow-up survey (71.4%).The third wave of data (T3) was collected one year after T1 and six months after T2, following the same procedure as the previous assessments.All respondents who participated at T1 were invited to participate at T3, even if they had not participated at T2.Altogether 1,164 respondents participated in the third follow-up survey (72.4%).
The survey was approved by the Regional Committee for Medical Research Ethics for Eastern Norway.Responses were treated anonymously, and informed consent was given by the respondents.
Instruments.Exposure to bullying behaviours in the workplace was measured using the Norwegian version of the Short-Negative Acts Questionnaire (S-NAQ; Notelaers et al., 2019).Response categories ranged from 1-5 (never, now and then, monthly, weekly, and daily).Example items are "Being ignored or excluded", "Repeated reminders of your errors or mistakes," and "Someone withholding information which affects your performance."The internal consistency of the scale was good (α T1 = .86,α T3 = 87).The Skewness of exposure to bullying behaviours was found to be 4.097 at T1 and 4.019 at T3, indicating that the distribution in this sample is heavily skewed to the left.
Scales from the General Nordic Questionnaire for Psychological and Social Factors at Work (QPS Nordic ) (Dallner et al., 2000;Wännström et al., 2009) were used to measure role ambiguity (3 items, α Τ1 = .80)and role conflicts (3 items, α T1 = .63,α Τ2 = .57).Although the internal consistency for the role conflict scale was lower than the recommended value (α > .70),we regarded it as acceptable in the present study given that the scale only consisted of three items that showed moderate inter-correlations (.31, .31, and .34, respectively).The items measuring role ambiguity are "Have clear, planned goals and objectives been defined for you?", "Do you know what your responsibilities are?" and "Do you know exactly what is expected of you at work?"The items measuring role conflicts in the form of intra-role conflicts are "Do you have to do things that you feel should be done differently?","Are you given assignments without adequate resources to complete them" and "Do you receive incompatible requests from two or more people?"Respondents provided their responses on a four-point Likert scale ranging from 1-4 (never, sometimes, often, and always).Three items were phrased to reflect role clarity and reversed to represent role ambiguity in the present study.The Skewness of role ambiguity at T1 was found to be 0.826, indicating that the distribution in this sample was moderately skewed to the left.The skewness for role conflict was 0.432 at T1, and 0.309 at T2, indicating that the distribution for this variable is approximately symmetric.
Laissez-faire leadership was measured using four items from the Multifactor Leadership Questionnaire (Bass, 1990a;Bass & Avolio, 1990).In line with measurements of alternative forms of destructive leadership, such as abusive supervision (Tepper, 2000), the wording of each item was adjusted so as to emphasise the one-to-one relationship between the leader and the respondent (see e.g.Nielsen et al., 2019).Accordingly, the wording of the items are as follows: "My immediate supervisor … " " … avoids involving him/herself in tasks that are important for me and my work", " … is absent when I need him/her", " … avoids making decisions that are important for me and my work", " … delays responding to questions that I need urgent answers to".Response alternatives ranged from 1-5 (never, rarely, once in a while, quite often, and very often or always).The internal consistency of the scale was good (α Τ2 = 83).The skewness for laissezfaire leadership was 0.452 at T1, and 0.407 at T2, indicating that the distribution for this variable is approximately symmetric.
Transformational leadership was measured using the Global Transformational Leadership Scale (GTL; Carless et al., 2000).This seven-item short scale assesses transformational leadership as a unified construct and is designed to represent a global measure of perceived transformational leadership as portrayed by one's immediate leader/ manager (Carless et al., 2000).The items capture seven leadership behaviours: (i) communicates a clear and positive vision, (ii) develops staff, (iii) supports staff, (iv) empowers staff, (v) is innovative, (vi) leads by example, and (vii) is charismatic.Response alternatives ranged from 1-5 (never, rarely, once in a while, quite often, and very often or always).The GTL has good convergent validity with established lengthier scales such as the Multifactor Leadership Questionnaire (MLQ) and the Leadership Practices Inventory (LPI) (Carless et al., 2000).The internal consistency of the scale was very good (α Τ2 = .94).The skewness for transformational leadership was −0.493 at T1, and −0.653 at T2, indicating that the distribution in this sample is moderately skewed to the right.

Data analysis
We conducted a correlation and a regression analysis SPSS (version 25) to test the proposed relationships between role ambiguity and role conflict, respectively, and subsequent exposure to bullying behaviours (hypotheses 1a and 1b).In order to test the proposed mediation effect of role conflicts in hypothesis 2, we estimated a simple mediation model (model 4), using the SPSS PROCESS macro version 3.4 provided by Hayes (2013).This macro facilitates estimation of the indirect effect (ab) with a bootstrap approach to obtain confidence intervals (CIs).The application of bootstrapped CIs is preferred over the Sobel tests because the bootstrapping approach does not make any assumptions regarding the sampling distributions of the indirect effects, and also reduces the likelihood of Type 1 errors (MacKinnon et al., 2004;Preacher & Hayes, 2008).Bootstrapping was set to 5,000 subsamples.
Subsequently, we predicted that laissez-faire and transformational leadership, respectively, would moderate the indirect path between role ambiguity and bullying, through role conflict.We started by carrying out an explorative analysis (model 45), in which the two leadership styles were included as potential moderators on both the path between role ambiguity (T1) and role conflicts (T2) and on the path between role conflicts (T2) and bullying behaviours (T3).Subsequently, the result from this first analysis was used to further inform our analysis strategy.These moderated-mediation models were tested using the above-mentioned SPSS macro.This SPSS macro facilitates the implementation of the recommended bootstrapping methods and permits the probing of the significance of conditional indirect effects at different values of the moderator variable.

Results
The scales' means, standard deviations, and correlations are reported in Table 1.
Positive correlations (see Table 1) were found for role ambiguity T1 and role conflicts T1 & T2 with exposure to bullying behaviours T3 .These results are in support of hypotheses 1a and 1b.Additionally, laissez-faire leadership T2 was positively correlated with role ambiguity T1 , role conflicts T1 & T2 and exposure to bullying behaviours T3 , as expected.Finally, negative correlations were found for transformational leadership T2 with all other study variables.
We also estimated a linear regression model to analyse the main effects of role ambiguity and role conflicts at T1 on subsequent exposure to bullying behaviours at T3, when also controlling for the stability of bullying exposure at T1.These results showed that there was not a significant main effect of role ambiguity T1 on exposure to bullying behaviours T3 (β = .002,n.s.).Taken together with the results of the correlation analysis, our findings indicate that while there is an association between role ambiguity and subsequent exposure to bullying behaviours, this association is only significant if we do not control for the stability of bullying exposure.There was, however, a significant main effect of role conflicts T1 on subsequent exposure to bullying behaviours T3 (β = .106,p < .001),providing further support for hypothesis 1b.Finally, as expected, the stability in exposure to bullying behaviours from T1 to T3 was high (β = .593,p < .001).
In order to analyse the mediating effect of role conflicts in the relationship between role ambiguity and exposure to bullying behaviours, we applied the PROCESS macro by Hayes (2013).
Table 2 shows the results of the mediation analysis (PROCESS model 4).Here, there were significant direct associations between role ambiguity T1 and role conflicts T2 (b = 061, SE = .022,p < .01),as well as between role conflicts T2 and exposure to bullying behaviours T3 (b = .119,SE = .021,p < .001).The direct effect of role ambiguity T1 on exposure to bullying behaviours T3 was not significant (b = -.003,SE = .004,n.s.)There was, however, a significant indirect effect of role ambiguity T1 on exposure to bullying behaviours T3 , through role conflicts T2 (b = .007,SE = .003,95% CI [.002, .014]),supporting hypothesis 2. In our analyses, we controlled for the stability of role conflicts T1 , which significantly predicted subsequent role conflicts six months later (b = .46,SE = .03,p < .001).We also controlled for the stability of exposure to bullying behaviours T1 , which significantly predicted both subsequent role conflicts T2 (b = .16,SE = .04,p < 01), and later exposure to bullying behaviours T3 (b = .52,SE = 03, p < .001).The explained variance of the total effects model was 40%.We also tested for the possibility for reversed causation between the two role stressors by estimating a model in which role ambiguity T2 is included as a mediating mechanism between role conflicts T1 and bullying behaviours T3 .However, we did not find support for this model. 1This further supports our theoretical assumption that the level of experienced role conflicts is in fact the mediating mechanism at play.In order to test the moderating effects of laissez-faire and transformational leadership as proposed in hypotheses 3 and 4, we conducted several moderated mediation analyses using PROCESS for SPSS (Hayes, 2013).First, we conducted an analysis that included both laissez-faire leadership T1 & T2 and transformational leadership T1 & T2 as concurrent moderators in the same moderated mediation model.As our hypotheses did not specify on which path leadership is expected to act as a moderator, we first conducted an analysis where the two leadership styles were included as moderators both in the path between role ambiguity and role conflicts (path a), and the path between role conflicts and exposure to bullying behaviours (path b) (see PROCESS model 45).The result of this analysis revealed interaction effects of both leadership styles, but only on the path between role conflicts and exposure to bullying behaviours.We therefore decided to only report the most parsimonious analysis, where laissez-faire and transformational leadership, respectively, are only included as moderators in path b 1i (PROCESS model 14, see Tables 3 and 4), in line with the suggestions of Cohen and colleagues (2003).
Table 3 shows the conditional indirect effect (through role conflict) between role ambiguity and exposure to bullying behaviours (S-NAQ) at different levels of laissezfaire leadership.
First, the analysis revealed significant direct associations between role ambiguity T1 and role conflicts T2 (b = .06,SE = .02,p < .05),as well as between role conflicts T2 and exposure to bullying behaviours T3 (b = .09,SE = .02,p < .001).The direct effect of role ambiguity T1 on exposure to bullying behaviours T3 was not significant (b = -.007,SE = .01,n.s.).This provides additional support for hypothesis 2. Furthermore, the effect of role conflicts T2 on exposure to bullying behaviours T3 was stronger at high compared to low levels of laissez-faire leadership T2 (b = 08, SE = 02, p < .001).Additionally, the results showed a significant index of moderated mediation (b = .004,SE = .003,95% CI [.0002, .0101]),indicating that the indirect effect of role ambiguity T1 on exposure to bullying behaviours T3 , through role conflicts T2 , is contingent on the level of laissez-faire leadership T2 .These results support hypothesis 3. Controlling for the stability of experienced role conflicts T1 and exposure to bullying behaviours T1 (see Table 3), the full model explained 43.9% of the variance in exposure to bullying behaviours T3 .The interaction term alone explained 1.2%, and is illustrated in Figure 2, showing the effect at different levels of laissez-faire leadership (+/-1 SD).
In line with hypothesis 3, Figure 2 indicates a positive relationship between role conflicts T2 and exposure to bullying behaviours T3 for employees reporting high levels of laissez-faire leadership T2 from their immediate supervisor.For employees reporting low levels of laissez-faire leadership T2 , the figure indicates no increase in exposure to bullying behaviours T3 at higher levels of role conflicts T2 .This is also supported by simple slope tests, where the positive slope for high levels of laissez-faire leadership was significant (Slope = 0.151, t = 5.816, p < .001.), whereas the slope for low levels of laissez-faire leadership was not (Slope = 0.023, t = 0.850, n.s.).
Table 4 shows the conditional indirect effect (through role conflicts) of role ambiguity on exposure to bullying behaviours at different levels of transformational leadership.The  association between role conflicts T2 and exposure to bullying behaviours T3 was weaker at (vs. low) levels of transformational leadership T2 (b = -.09,SE = .02,p < .001).Additionally, the results showed a significant index of moderated mediation (b = -.005,SE = .003,95% CI [-.01, -.001]), indicating that the indirect effect of role ambiguity on exposure to bullying behaviours, through role conflicts, is contingent on the level of transformational leadership, supporting hypothesis 4. Controlling for the stability of experienced role conflicts T1 and exposure to bullying behaviours T1 (see Table 4), the full model explained 44.7% of the variance in exposure to bullying behaviours T3 .The interaction term alone explained 1.4%, and is illustrated in Figure 3, showing the effect at different levels of transformational leadership (+/-1 SD).
Figure 3 indicates no increase in exposure to bullying behaviours T3 at higher levels of role conflicts T2 for employees who report high levels of transformational leadership T2 from their immediate supervisor.For employees reporting low levels of transformational leadership T2 , however, there is a positive relationship between role conflicts T2 and exposure to bullying behaviours T3 .Hypothesis 4 is further supported by the results of  the simple slopes test, showing that the positive slope for low levels of transformational leadership was significant (Slope = 0.172, = 6.572, p < .001.), whereas the slope for high levels of transformational leadership was not (Slope = 0.017, t = 0.610, n.s.).

Discussion
Earlier studies have consistently shown positive associations between role ambiguity and role conflicts, respectively, and exposure to workplace bullying (see Van den Brande et al., 2016, for an overview).The present study extends this knowledge by demonstrating a mediating effect of role conflicts in the relationship between role ambiguity and subsequent exposure to bullying behaviours.Moreover, our data showed that laissez-faire leadership exacerbated, while transformational leadership attenuated, the indirect relationship between role ambiguity and subsequent exposure to bullying behaviours through role conflicts.
Hence, in support of hypotheses 1a and 1b, role ambiguity and role conflicts were positively associated with subsequent exposure to acts of workplace bullying (see Table 1).However, when controlling for baseline exposure to bullying behaviours at T1, only role conflicts, but not role ambiguity, remained a significant predictor for subsequent exposure to bullying behaviours at T3.Thus, although we find support for hypothesis 1a, it should be noted that only role conflict was associated with an actual change in levels of bullying over time.This finding is in line with the theoretical notion that experiencing role conflicts denotes a situation where one may face sanctions and negative responses from others in the working environment, which of course may be experienced as bullying behaviours.Moreover, the findings from the correlation analysis are in line with the theoretical underpinnings of the social interactionist theory (Felson, 1992;Felson & Tedeschi, 1993), where stressors in the work environment are thought to trigger emotional and behavioural changes in employees, leading them to violate normbased expected work behaviours, indirectly unleashing negative workplace behaviour from others.Note, however, that this explanation includes unobserved mediators (i.e.employees' emotional response to stressors and subsequent norm violation) that have not been explicitly tested in the present study.
Alternatively, in line with the stressor-emotion model of counterproductive work behaviour (Spector & Fox, 2005), the presence of ambient stressors (i.e.role stressors) in the work environment is likely to elicit a negative emotional response in exposed individuals.This emotional response may, in turn, motivate and energise subsequent behavioural, physical, or psychological changes in the affected individuals (Spector, 1998;Spector & Goh, 2001).More specifically, individuals who experience high levels of role ambiguity and role conflict, and who perceived these situational constraints as threats to their individual well-being, are likely to experience negative emotional reactions such as anger and anxiety (Spector, 1998;Spector & Goh, 2001).Engaging in negative social behaviour towards other organisational members may be one outcome of this stress process (Spector & Fox, 2005).Accordingly, the stressor-emotion model posits that work stressors (including role ambiguity and role conflict) may lead to bullying by increasing the overall levels of employees' vulnerability and aggressiveness.Again, however, this explanation remains theoretical in nature, as we have not measured the potential perpetrators' response to ambient stressors in the work environment.
Supporting hypothesis 2, and in line with early role theory (Kahn et al., 1964), our findings indicated an indirect effect of role ambiguity exposure to bullying a year later through an increase in employees' perceptions of role conflicts after six months (see Tables 2-4).Employees who do not have a clear picture of what is expected of them are likely to experience increased levels of role conflictsthat is increasingly more conflicting expectancies from different role-sendersprobably because they have not received the necessary information and resources to be able to adjust to the requirements of other roles (Kahn et al., 1964).In line with the social interactionist theory (Felson & Tedeschi, 1993;Tedeschi & Felson, 1994), this increase in role conflicts, should, in turn, lead to increased levels of frustration and negative affect in the individual which may act to trigger the bullying process, as well as trigger tension and conflict with and among others in the working environment.To the best of the authors' knowledge, this is the first study to empirically demonstrate the mediating mechanism of role conflicts in the role ambiguity-negative acts relationship by employing longitudinal data.Hence, employees' experienced ambiguity constitutes a risk factor for bullying by creating fertile soil for intra-role conflicts to develop.
In support of hypothesis 3, laissez-faire leadership exacerbated the indirect relationship between role ambiguity and exposure to bullying through role conflicts (see Table 3).This findingalong with studies showing the exacerbating effect of laissez-faire leadership on the negative relationship between other work stressors and bullying exposure (Ågotnes et al., 2018; Ågotnes et al., 2021) support the notion that the combined effect of workplace stressors may in fact create an elevated risk of bullying over and above their additive effects (Cooper et al., 2001).This adds further support to the work environment hypothesis by showing that the combination of poorly organised work conditions (i.e.role stressors) and inadequate leadership seems to create a fertile ground for the development of bullying at work (Einarsen et al., 1994;Einarsen et al., 2020;Leymann, 1996).Specifically, the lack of active and constructive interventions on the part of a superior in conflicting and demanding working situations where subordinates need leader support and intervention is likely to increase subordinates' feelings of frustration and stress.This allows interpersonal tensions and conflicts to escalate, leaving the focal employee (s) at an increased risk of exposure to bullying (c.f. the social-interactionist theory).
Likewise, hypothesis 4 was supported, as transformational leadership attenuated the indirect relationship between role ambiguity and exposure to bullying behaviours one year later, through increased role conflicts after six months (see Table 4).That is, employees who reported higher levels of superiors' transformational leadership did not report this increased exposure to bullying behaviours even when the level of role conflicts was high.The individual consideration and supportive behaviour enacted by transformational leaders may constitute a decisive mechanism in the role stressors-bullying behaviour relationship.Superiors who are attentive to their followers facing ambiguous and/or conflicting expectations and demands, and who actively support and assist their followers in coping with these demands, seem to be effective in averting subordinate negative social interactions from occurring and developing into escalated conflicts and bullying.Similarly, transformational leaders who show effective management skills (Baillien et al., 2011;Zapf & Gross, 2001) thereby making it clear that negative social behaviour is not tolerated in the organisation (Baillien & De Witte, 2009a, seem to be effective in preventing instances of negative social acts from developing in the workplace as a consequence of experienced role conflicts.Furthermore, this result may indicate that high levels of transformational leadership reflect the presence of organisation-level variables such as a strong for conflict management (Rivlin, 2001) or a strong psychological safety climate (Dollard & Bakker, 2010).To this point, there is some evidence that a strong conflict management climate is not only related to lower reports of bullying behaviours in general but also that this organisation-level variable acts as a buffer on the negative consequences of work stressors on perceived bullying (Zahlquist et al., 2019).In such a climate, organisations have implemented proper policies for managing instances of bullying and negative behaviour among subordinates.As such, it is possible that these organisation-level policies and practices may in fact be the moderating factors in play, manifesting in the behaviour of the individual leader.This possibility should be further investigated in future studies.
An important question not answered in this study is whether this finding only applies to transformational leadership, or if it is also present for other forms of constructive leadership practices, such as transactional forms of leadership.Likewise, in addition to testing the detrimental effects of passive-avoidant forms, such as laissez-faire leadership, active-confronting forms, such as abusive supervision, should be tested.Future studies should therefore explore the potential moderating effects of alternative constructive and well as destructive forms of leadership in the role stressor-bullying behaviour relationship.
It is worth noting that while our analyses included leadership behaviour as a moderator on both the path between role ambiguity and role conflicts and between role conflicts and exposure to bullying behaviours, our results showed that laissez-faire behaviour exacerbated, while transformational leadership attenuated, the association between role conflicts and subsequent exposure to bullying behaviours only.One explanation for this finding may be that proper conflict management behaviour by the leader is especially important when there are signs of a conflict situation arising and/or some level of exposure to bullying behaviours is present.In his model of conflict escalation, Glasl (1982) argued that when conflicts are in the first main phase of escalation, the parties involved still consider the task and content related aspect of their conflict to be central, and they are normally motivated to co-operate to resolve the conflict.This is likely also the case in situations with increasing levels of role conflicts.The longer the conflict persists and escalates, however, the associated tension and frustration is likely to become high, making it increasingly difficult to achieve a resolution of the conflicting situation.Thus, when a conflict situation has reached a higher level of escalation, it is arguably less likely that the subordinate(s) in question will be able to resolve the situation themselves, making it paramount for the leader to step in and manage the conflict.In contrast, it could be argued that leadership is likely to have a stronger impact on the presence of role ambiguity in general (Skogstad et al., 2014), as the experience of role ambiguity is theorised to arise from the expectations and subsequent communications emanating from a role sender, typically the leader (Kahn et al., 1964).However, it seems that leadership is not the critical conditional factor when it comes to the development of increased role conflicts from role ambiguity, at least not in the context of explaining exposure to bullying in the workplace.In this, leadership behaviour seems to be more important when it comes to exacerbating or attenuating the further development of role conflicts into subsequent perceived exposure to bullying behaviour.As such, it is reasonable to expect active involvement and conflict management strategies on the part of the leader to be particularly important in preventing instances of role conflict from tiggering a subsequent process.

Methodological strengths and limitations
The current study has several strengths, as especially indicated by its three-wave prospective design.Studies utilising three or more waves of data are rare (Ployhart & Vandenberg, 2010), and there has long been a call for research employing longitudinal data when studying the antecedents and mechanisms in relation to workplace bullying (Nielsen & Einarsen, 2018) as well as to detrimental and destructive forms of leadership in general (Martinko et al., 2013;Tepper et al., 2017).In this regard, the present study represents a unique contribution to the fieldin parallel to the increasing number of dairy studies which also advance the designs used in research on workplace bullying (see e.g.Ågotnes et al., 2021;Rodríguez-Muñoz et al., 2022).By employing prospective data from a national probability employee sample with three measurement points over a total of 12 months, we were able to test the proposed indirect effect of role conflicts in the role ambiguity-bullying relationship with a proper longitudinal designcontrolling for the T1 values of the dependent and mediating variables as well as testing for any reversed relationships between role conflicts and role ambiguity.
Furthermore, all study variables were measured using psychometrically sound instruments.To fit the individual-level hypothesis, the indicator of laissez-faire leadership was revised to emphasise a one-to-one relationship between the leader and the subordinate, rather than being a global measure of laissez-faire leadership behaviours across subordinates and situations (see Nielsen et al., 2019).
However, some limitations are also worth considering.First, the response rate at timepoint 1 was only 32%.This is lower than the average rate earlier documented for survey studies (Baruch & Holtom, 2008), and could limit the external validity of the findings.However, it should also be noted that there is a declining trend in response rate in survey research, and our response rate is in line with other survey studies from the last few years (Stedman et al., 2019).Furthermore, it has recently been shown that response rates in surveys have no noticeable effect on the magnitude of correlations and therefore also on theory development and testing (Beehr et al., 2022).Finally, our sample closely resembles the working population in Norway in terms of age and gender (National Institute of Occupational Health, 2021), also mitigating the generalizability threat.
Second, the problem of common method variance (CMV) due to self-reports always exists in such survey data (Podsakoff et al., 2003;Spector et al., 2017).However, the risk of common method variance is probably reduced by the use of a longitudinal research design with multiple measurement points (Podsakoff et al., 2003).Finally, the use of the PROCESS macro in SPSS to analyse our data, rather than estimating a structural equation model with latent variables, could be considered a limitation.There are limitations associated with standard regression analyses using PROCESS, including the fact that we do not have the opportunity to control for potential measurement error in our data.However, simulation studies indicate that latent variable interactions cannot be reliably estimated with skewed data using latent moderated structural equations (LMS) methods (Cham et al., 2012;Maslowsky et al., 2015).Thus, given that our dependent variable is highly skewed and that the utility of the LMS procedure, as a result, is limited, we have deemed PROCESS as an appropriate of analysis in the present study.

Implications for theory and practice
By examining role conflict as a mediator in the role ambiguity-workplace bullying relationship employing longitudinal data, our findings contribute important and nuanced knowledge regarding the inter-relationship between role ambiguity and role conflicts, at least with regards to exposure to bullying behaviours, which so far has been lacking in the field.Furthermore, the fact that our analyses failed to support a reverse causal relationshipi.e. that role ambiguity did not function as a mediator in the relationship between role conflict and workplace bullyingstrengthens our hypothesis that the presence of role conflicts in the work environment is a crucial and proximal risk factor for workplace bullying to occur and develop.It should be noted, however, that bullying over time may have long term effects such as an increase in subsequent role stress, or that this relationship may in fact be cyclical in nature (Balducci et al., 2012;Hauge et al., 2011).Yet, we argue that the presence of poor and stressful working conditions (i.e.role ambiguity and role conflict) are prerequisites for bullying to exist and develop in the first place.
From a practical standpoint, the implementation of measures reducing role stressors, starting with experienced role ambiguity, is likely a decisive measure to prevent negative acts and bullying from developing.Yet, as some presence of role ambiguity and role conflicts in organisations is probably inevitable, the organisational goal should probably not be to eliminate these role stressors completely, but rather to keep them at a tolerable level (Kahn et al., 1964).In line with this, we have to keep in mind that role expectations have a very bright side in the form of role privileges and gratification (see e.g.Sieber, 1974).Important interventions as regards the dark sides of role expectations include the implementation of work designs that ensures clarity about work roles and responsibilities, research-based strategies for the selection and development of considerate and responsive leaders, a culture that rewards considerate and helpful behaviours, and the development of a strong climate for conflict management (Zahlquist et al., 2019).If these interventions are successful, organisations will probably reduce the risk of negative social interactions as well as prevent workplace bullying from arising and developing.
Furthermoreas the present study underscoresescalating negative acts and bullying will only occur within organisations that condone and/or accept such behaviours to take place (Brodsky, 1976).More specifically, our findings show that it is only under high levels of laissez-faire leadershipas well as under low levels of transformational leadershipthat role stressors pose a risk for workplace bullying to develop.Consequently, it is critical to improve the competencies and sharpen the focus of supervisors, so that they are better equipped and tuned in to intervene and manage the stressful situations facing their subordinates.Therefore, organisations must intervene, not only in the organisation of work tasks and the clarification of roles but also inform and train leaders about the type of leadership behaviours that are expected and reward those behaviours accordingly.In this regard, leadership development programmes should include the development of competencies in identifying and dealing with role ambiguities and role conflicts and early signs of conflict escalation and bullying (Salin, 2006).approach may be an even more effective intervention than merely focusing on reducing role stressors in general.A noteworthy issue and finding in this respect is that transformational and laissez-faire leadership may work in tandem, which indicates that one should not only train leaders to be more transformational but also focus on reducing their laissezfaire leadership behaviours, that is leaders' disregard, inattention, and avoidance in concrete situations where subordinates are in need of leadership (Skogstad et al., 2014).

Conclusion
The present study extends previous research on role stressors and workplace bullying by showing the indirect effect of role ambiguity on subsequent exposure to bullying behaviours, through increased levels of reported role conflicts.In this, we add knowledge both to role theory in general and more specifically to the processes and conditions via which workplace stressors are transformed into bullying scenarios.Our study also shows that this indirect effect is moderated by two distinct leadership practices, namely laissez-faire and transformational leadership.As such, managers' leadership behaviours will eventually determine whether ambiguous and conflicting working conditions give rise to a working environment in which negative social interactions such as bullying behaviours are prevalent.Earlier data show that if the management avoids or neglects its inherent responsibility to adequately address stressful work conditions, as is the case with laissez-faire leadership, this may constitute a particularly high-risk situation in terms of the development of bullying at work (Leymann, 1996).Our results support these findings, but also show a buffering effect of transformational leadership on the relationship between role conflicts and exposure to bullying behaviours.Therefore, one may argue that it is important not only to reduce the occurrence of laissez-faire leadership behaviour in order to prevent employees from being exposed to bullying behaviours, but that the presence of transformational leadership behaviours also is important in this regard as a preventive measure.

Figure 1 .
Figure 1.Overview of the proposed relationships to be investigated in the present study.

Table 1 .
Descriptive statistics and correlations among the study variables.Cronbach's alpha is reported in bold (N = 1,103-1,594).

Table 2 .
Analyses for the conditional indirect effect (through role conflicts) between role ambiguity and exposure to bullying behaviours (S-NAQ) (PROCESS model 4).

Table 3 .
Analyses for the conditional indirect effect of role ambiguity (through role conflicts) on exposure to bullying behaviours (S-NAQ) at different levels of laissez-faire leadership (PROCESS model 14).

Table 4 .
Analyses for the conditional indirect effect of role ambiguity (through role conflicts) on exposure to bullying behaviours (S-NAQ) at different levels of transformational leadership (PROCESS model 14).