Dealing with employees’ frustration in time saves your company from workplace bullying: The mediating roles of frustration and a hostile climate in the relationship between role stress and exposure to workplace bullying

Abstract

The development of workplace bullying, which involves negative behaviors occurring regularly and over a period of time, is explained by the work environment hypothesis, namely, that it is due to organizational factors, such as leadership practices and organizational climate. Although this has been the predominant theoretical framework for studying workplace bullying, the mechanism whereby particular organizational factors trigger exposure to bullying remains unclear. The present study aims to apply both the revised frustration—aggression theory and the social interactionist perspective of aggression to examine the mechanism responsible for the relationship between role stressors and exposure to bullying. In a two-wave longitudinal study, we collected data from 353 Polish employees. The double mediation analysis revealed the mediating role of both individual frustration, measured in wave 2, and perceived hostile work climate, measured in wave 2, in the relationship between the role stressors measured in wave 1 and exposure to workplace bullying measured in wave 2. The current study sheds light on the mechanism responsible for the relationship between organizational antecedents and exposure to workplace bullying, thus explaining the core assumptions of the work environment hypothesis. The findings suggest that to reduce the risk of workplace bullying development, it is crucial to identify and respond constructively to employees’ frustration and dissatisfaction by, for example, reorganizing work structures that may foster a hostile work climate and mistreatment.

Keywords:

• workplace bullying
• role stress
• hostile work climate
• work environment hypothesis
• frustration – aggression theory
• social interactionist perspective of aggression

1. Introduction

Workplace bullying is defined as emotional aggression in the form of insulting, intimidating, belittling, gossiping, or social excluding that is displayed regularly and over time by co-workers toward one or a few employees in situations where the target cannot defend themselves against this treatment (D’Cruz et al., 2021; Einarsen & Ågotnes, 2023; Einarsen et al., 2020). Recognized as a serious challenge for organizations all over the world (León-Pérez et al., 2021; Zapf et al., 2020), workplace bullying is a multifaceted construct that includes interpersonal, depersonalized, real, cyber, internal, and external bullying (D’Cruz & Noronha, 2014, 2021). For decades, a growing body of workplace bullying literature has provided evidence of its detrimental effects on both individuals (Conway et al., 2021; Hansen et al., 2021; Luo et al., 2023; Mikkelsen et al., 2020; Nielsen et al., 2020; Pauksztat et al., 2022) and organizations as a whole (Hoel et al., 2020; Høgh et al., 2021).

Two principal explanations of the antecedents of workplace bullying have been proposed: the vulnerability hypothesis (Bowling & Beehr, 2006; Bowling et al., 2010) and the work environment hypothesis (Hauge, 2010; Hauge et al., 2007; Leymann, 1996). The vulnerability hypothesis, which concentrates on the individual traits of workers exposed to bullying and assumes that these individual factors can explain the development of bullying at work, has gained some empirical support (e.g., Bowling et al., 2010; Gamian-Wilk et al., 2022; Reknes et al., 2021; see; Persson et al., 2021 for review). However, the findings are mixed and suggest that the shifts in traits in employees subjected to workplace bullying are a result rather than a cause of being bullied (e.g., Gamian-Wilk et al., 2022; Nielsen & Knardahl, 2015; Podsiadly & Gamian-Wilk, 2017). Conversely, the work environment hypothesis, which focuses on the impact of organizational causes and assumes that factors such as the type of leadership and social climate can explain the development of bullying at work, has been empirically evidenced (Balducci et al., 2021). A vast number of organizational factors have been identified as bullying antecedents (e.g., Hauge et al., 2011; see; Balducci et al., 2021, for a review); however, one of the most frequently examined risk factors is role stress in the form of role ambiguity and role conflict (Harlos & Holmvall, 2021; Van den Brande et al., 2016).

Although a great deal of evidence supports the relationship between role stress and workplace bullying, few studies have looked at the mechanism underlying this relationship (e.g., Hamre et al., 2023). In line with the frustration—aggression (FA) hypothesis (Berkowitz, 1969, 1981, 1989) and the social interactionist perspective (SIP) on aggression (Felson, 1992; Felson & Tedeschi, 1993; Neuman & Baron, 2011), Einarsen et al. (1994) argue that role stressors may generate employees’ stress and frustration, which, in turn, may worsen the social climate, leading to the development of bullying behaviors. To the best of our knowledge, no previous research has investigated these assumptions or tested the proposed causal chain of occurring factors. The aim of the present study is to examine the mechanism underlying the relationship between role stress and exposure to workplace bullying. In a longitudinal study, we examine the mediating roles of both individual employees’ frustration and the perceived hostile work climate in the relationship between role stressors and exposure to workplace bullying.

Our study contributes to the literature on the organizational antecedents of workplace bullying by using affective events theory (AET; Weiss & Cropanzano, 1996) as a theoretical background to explain how work environment generates affective events. Additionally, our study expands understanding of AET as a workplace bullying theoretical framework by implementing both the revised FA theory (Berkowitz, 1989) and the SIP on aggression (Felson & Tedeschi, 1993; Neuman & Baron, 2011) to indicate how role stressors trigger exposure to bullying via employees’ feelings of frustration and a perceived hostile work climate. These assumptions have previously been only hypothesized (Einarsen et al., 1994) or empirically examined in a fragmentary manner in either qualitative (Baillien et al., 2009) or cross-sectional (Van den Brande et al., 2017) research by taking either frustration (Baillien et al., 2009) or a hostile work climate (Zahlquist et al., 2023) as mediators of organizational factors and the development of workplace bullying. The present study is, to the best of our knowledge, the first attempt to verify the mechanism underlying both frustration and a hostile work climate as mediators of the relationship between role stress and exposure to workplace bullying in a longitudinal design. Moreover, as few studies have investigated possible organizational mediators in the relationship between the work environment and exposure to workplace bullying (Nielsen & Einarsen, 2018; Rai & Agarwal, 2018), the current research fills this gap by identifying two mediators, namely, individual frustration and a hostile work climate, that are responsible for bullying escalation. Therefore, the present study expands the literature on the work environment hypothesis (Einarsen et al., 2020, 1994; Hauge et al., 2011; Leymann, 1996) and on the relationship between role stressors and exposure to workplace bullying (Harlos & Holmvall, 2021).

1.1. Theoretical background

Although researchers have drawn on various prominent theories to explain the development of workplace bullying (e.g., social exchange theory (SET), the conservation of resources (COR) model, the cognitive activation theory of stress (CATS), the job demands—resources (JDR) model, social identity theory (SIT), and social rules theory (SRT)), Branch and colleagues’ (2021) theoretical review of the workplace bullying literature concludes that AET (Weiss & Cropanzano, 1996) is an especially relevant and comprehensive theory that describes the complexity of workplace bullying antecedents and outcomes. AET has been frequently used to investigate the workplace bullying phenomenon (e.g., Gamian-Wilk et al., 2022; Glasø et al., 2011; Jahanzeb et al., 2021). AET claims that individuals respond emotionally to affective events, such as workplace bullying. These work events, in turn, impact affective responses and work attitudes and behaviors. In light of AET, we assume that affective events may be directly or indirectly generated by a work environment. This approach resonates with the work environment hypothesis (Hauge et al., 2010; Leymann, 1996), which states that organizational features trigger affective events, such as workplace bullying exposure.

Additionally, based on both the revised FA theory (Berkowitz, 1969, 1981, 1989) and the SIP on aggression (Felson, 1992; Felson & Tedeschi, 1993; Neuman & Baron, 2011), we explain the mechanism underlying the relationship between work environment and exposure to bullying. According to the FA hypothesis, aggression is the result of hindering or frustrating an individual’s efforts to achieve their goals. As stated in the SIP on aggression, when individuals are confronted with injustice and conflicts, they respond by complaining and displaying aggression to exert social influence and enhance or maintain their desired social position in their groups. Thus, in line with the revised FA theory (Berkowitz, 1989), we assume that, as fulfilling job requirements is difficult or impossible because of role ambiguity and role conflict, employees feel frustrated and stressed. Then, as proposed by the SIP on aggression (Felson & Tedeschi, 1993), faced with role conflict and being stressed and frustrated, they tend to misbehave, violate social norms to influence their social surroundings, and thus provoke interpersonal conflicts that damage the overall social climate at the workplace. Moreover, according to the SIP of teasing and bullying (Kowalski, 2004; Neuman & Baron, 2011), by implementing negative behaviors bullies gain power and achieve personal goals. Bullying is therefore used for instrumental reasons to maintain position within a social hierarchy (Kowalski, 2004).

Situational factors, such as role stressors, are thus critical in activating aggression. Branch et al. (2021) argue that both the FA hypothesis (Berkowitz, 1989) and the SIP on aggression (Felson & Tedeschi, 1993; Neuman & Baron, 2011) are helpful in understanding the interactions within a system. Therefore, drawing on FA theory and the SIP to explain the work environment hypothesis mechanism, we assume that individual frustration and hostile work climate caused by role stressors predict exposure to workplace bullying.

1.2. Role stressors as workplace bullying predictors

The work environment hypothesis has undoubtedly been the predominant theoretical framework used to examine workplace bullying (Balducci et al., 2021). In line with this hypothesis (e.g., Einarsen et al., 2020), a poorly organized work environment, that is, lack of work design and proper leadership behaviors (Hauge et al., 2007, 2009, 2011), and permitting or even rewarding such negative practices (Einarsen et al., 2020), generates conditions fueling bullying activities.

With regard to work design, role stress has been one of the most reported and best documented causes of workplace bullying development (Bowling & Beehr, 2006; Harlos & Holmvall, 2021; Van den Brande et al., 2016). Generally, individuals respond with stress when they experience pressure from others’ expectations and work demands (Kahn et al., 1964), that is, role stressors in the forms of role ambiguity and role conflict (Beehr, 1995; Rizzo et al., 1970).

Role ambiguity is characterized by uncertain job requirements and unclear information from role senders regarding what is expected of an employee and how tasks should be performed (Beehr, 1995; Rizzo et al., 1970). Role conflict concerns the existence of two or more inconsistent or contrary sets of expectations directed toward an employee such that performing one makes fulfilling the other difficult or impossible (Beehr, 1995; Rizzo et al., 1970). Several cross-sectional (e.g., López-Cabarcos et al., 2017; Notelaers et al., 2010) and prospective (e.g., Balducci et al., 2012; Hamre et al., 2023; Reknes et al., 2014; Salin & Hoel, 2020) studies as well as reviews (Harlos & Holmvall, 2021) and meta-analyses (Bowling & Beehr, 2006; Van den Brande et al., 2016) have consistently indicated the positive association between role stress and experiencing workplace bullying. Therefore, our first aim is to replicate previous patterns of results; thus, we anticipate that:

H1:

Role stress is positively related to exposure to workplace bullying.

1.3. Role stress, frustration at work, and a hostile work climate

As suggested in early works on the work environment hypothesis (Leymann, 1996), aversive organizational factors lead to physiological stress reactions and activate feelings of frustration. Therefore, role stressors, such as unclear requirements at work and incompatible or conflicting job demands, are viewed as organizational factors generating stress, frustration, and negative emotions within employees and thus facilitating the latter to vent or displace the accumulated pressure on co-workers (Einarsen et al., 2020; Leymann, 1996). This reasoning is consistent with the FA theory (Berkowitz, 1989), which has been used to explain how organizational antecedents lead to exposure to workplace bullying (Einarsen, 2000). In essence, being constantly confronted with role stressors leads to the inability to perform tasks, which is related to persistent exposure to stress and frustration. In line with stressor—strain theories, such stressors, if not adequately coped with, result in psychological (e.g., anxiety), physical (e.g., somatic symptoms), or behavioral (e.g., aggression) strain (e.g., Beehr, 1995; Jex & Beehr, 1991; Karasek & Theorell, 1990; Keashly et al., 1997). Indeed, role stressors are related to such outcomes as anxiety, depression, or job dissatisfaction (e.g., Örtqvist & Wincent, 2006; Spector, 1986; Spector & Jex, 1998). Frustration is therefore connected with negative emotions, particularly anxiety and tension caused by the inability to perform work duties. Thus, we anticipate that:

H2:

Role stress is positively related to individual frustration.

Furthermore, individual frustration caused by role stressors may create tensions within a workgroup as, in line with the SIP on aggression (Neuman & Baron, 2011), stressed employees tend to use inappropriate ways of coping (Reknes et al., 2014). Similarly, Baillien et al. (2009) highlight the role of frustration and the risk of irritation caused by work stressors in bullying development. Work environment features may form the basis for frustration and conflicts to which employees respond either actively or passively, but rather inefficiently, for example, by aggressively stopping discussions or avoiding tasks, thus violating norms, which, in turn, triggers further negative responses.

Therefore, generally, frustration and negative emotions may create the ground for a hostile work climate (Mawritz et al., 2012, 2014), which “refers to a social environment in the department characterized by escalated interpersonal conflicts and aggressive behaviour” (Zahlquist et al., 2023, p. 5). In workplaces with a hostile work climate, employees tend to be untrusting and aggressive (Mawritz et al., 2012). As a work climate is generally defined as a set of shared attitudes toward and perceptions of behaviors, values, and emotions in a certain working group (De Rivera, 1992), in workplaces characterized by a hostile work climate, destructive employee behaviors are perceived as a norm or “common practice” (Zahlquist et al., 2023, p. 4469). Employees who are stressed out, frustrated, insecure, and tense because of the dearth of clarity and conflicting demands and expectations tend to engage in interpersonal conflicts (De Rivera, 1992). Therefore, frustration resulting from role stressors may lead to a hostile work climate. Thus, we predict that:

H3:

Feelings of frustration are positively related to a hostile work climate.

Finally, a hostile work climate is related to interpersonal conflicts and aggressive behaviors. Therefore, employees working in a hostile work climate are more likely to be exposed to bullying (Zahlquist et al., 2023), and such a climate generally creates the grounds for deviant behaviors (Mawritz et al., 2012). We would like to replicate these results and therefore hypothesize that:

H4:

A perceived hostile work climate is positively related to exposure to workplace bullying.

1.4. Frustration and a hostile work climate in the relationship between role stress and workplace bullying

In early assumptions within the work environment hypothesis, Leymann (1996) and then Einarsen et al. (1994) suggested a causal process in which organizational features may trigger employees’ stress and frustration, which, in turn, may worsen the workplace climate, leading to bullying development. Therefore, the core of the work environment hypothesis is the assumption of a causal chain in which frustration and interpersonal tensions may play a mediating role in the pathway between organizational antecedents, such as role stressors, and exposure to bullying. The main aim of the present study is to test these assumptions.

To date, few studies have concentrated on mediation models to understand how workplace bullying is related to other variables (Nielsen & Einarsen, 2018; Rai & Agarwal, 2018). Emotions and psychological stress have usually been regarded as outcomes of being bullied (Bowling & Beehr, 2006) and have hardly ever been studied as mediators in the workplace bullying process. When emotions have been studied in this way, they have been perceived to have a mediating role in the relationship between exposure to bullying and health problems (e.g., Casimir et al., 2012; Mikkelsen & Einarsen, 2002; Rodríguez-Muñoz et al., 2011; Vie et al., 2012) or between exposure to bullying and job satisfaction (Glasø et al., 2011).

However, strain and frustration may also be considered mediators in the relationship between organizational antecedents, such as role stressors, and exposure to bullying. Working under frustration, stress, and negative emotions may constitute a kind of vulnerability; that is, strained employees may become victims for aggressive behaviors (Baillien et al., 2009). Applying the SIP on aggression (e.g., Neuman & Baron, 2011), Baillien and colleagues (Baillien et al., 2009), in a qualitative study, indicated that workplace bullying may be triggered by inefficient coping with frustration. Frustrated employees tend to verbalize and ventilate emotions to their co-workers, slowing down task performance and thus violating workplace norms. In reply, co-workers may display aggression towards frustrated employees. Indeed, in a cross-sectional study, Van den Brande et al. (2017, p. 4) revealed that emotion-focused coping (“focus on and venting of emotions”) increased the relationship between role conflict and exposure to workplace bullying. Therefore, we argue that:

H5:

Feelings of frustration mediate the relationship between role stress and exposure to workplace bullying.

In previous research, organizational climate, including a hostile work climate, has been studied as a moderator between organizational antecedents and workplace bullying (Hamre et al., 2023; Zahlquist et al., 2019, 2023). Thus, work climate has been perceived as a relatively stable contextual factor. However, a hostile work climate may also be fueled by an uncertain and insecure work environment, such as a lack of clarity on how to behave and how to perform duties, which triggers interpersonal conflicts (De Rivera, 1992). Thus, role stressors may serve as antecedents of a hostile work climate perceived as a state at a certain moment. We may therefore assume a dynamic process in which certain role stressors generate frustration, tensions, and perceived shifts in a hostile work climate (De Rivera, 1992; Mawritz et al., 2012, 2014), which, in turn, trigger exposure to workplace bullying (Zahlquist et al., 2023). Thus, we hypothesize that:

H6:

A hostile work climate mediates the relationship between role stress and exposure to workplace bullying.

Figure 1 presents the conceptual model and summarizes the hypotheses.

Figure 1. Conceptual model of the present study.

2. Method

2.1. Procedure and participants

To test our hypotheses, we collected data across two waves. In the first wave, data were collected among 716 employees (aged 18–63, M_age = 29.34, SD_age = 10.79, 71.8% women) from several private and public corporations, including working students. We used convenience sampling: participants were approached by the first author in their workplaces (e.g., state forest workers, firefighters, and bank accountants) in five Polish cities and asked to complete surveys. University students with working experience were recruited via SONA systems (Participant Pool Management for Universities; a platform allowing researchers at universities to build their own participant pools, run lab and online studies, and grant credit or cash to participants). The selection criteria required participants to be employed in full-time positions and to be in contact with their superiors or subordinates at least three times per week (e.g., Glasø & Einarsen, 2008). All respondents were invited to participate in the second wave of the study.

In total, 353 employees took part in both waves (aged 19–63, M_age = 30.02, SD_age = 11.01, 63.5% women). The participants’ tenure varied from 3 to 412 months (M_tenure = 34.55, SD_tenure = 56.12). The majority of participants (N = 289) worked as subordinates. Only those participants whose working situation did not change from wave 1 to wave 2 (i.e., they did not change job) took part in the second wave (the results of three participants were excluded on this basis). We analyzed the drop-out group—second wave differences; that is, we compared between participants in the second wave and those who dropped out after the first wave. The differences were insignificant in t-test comparisons (the drop-out group: for role stress, t = 1.28, p = .31; for exposure to bullying, t = .87, p = .56; for individual frustration, t = .78, p = .49; for hostile work climate: t = .29, p = .82).

Prior to data collection, all participants signed an informed consent form in which they agreed to take part in the study. The study was approved by the Ethical Committee at the SWPS University (decision No. 07/P/12/2021).

2.2. Instruments

2.2.1. Bullying

Workplace bullying was measured by a Polish version of the Negative Acts Questionnaire—Revised (NAQ-R) (Warszewska-Makuch, 2007), developed by Einarsen and colleagues (Einarsen et al., 2009). The NAQ-R consists of 22 items and describes different behaviors that may be perceived as bullying if they occur regularly. All items were formulated in behavioral terms, with no reference to the phrase “bullying and harassment.” The NAQ-R contains items referring to both direct (e.g., open attack) and indirect (e.g., social isolation, slander) behaviors. It also contains items referring to personal as well as work-related forms of bullying. For each item, the participants were asked how often they had been exposed to the behavior at their present workplace during the last six months on a 5-point scale (1– “never,” 2–“now and then,” 3–“monthly,” 4–“weekly,” and 5– “daily”). The 22 NAQ items were summarized (in line with, e.g., Nielsen et al., 2011).

2.2.2. Role stress

We used two subscales used by Hauge et al. (2011) back-translated into Polish. Hauge based his subscales on the General Nordic Questionnaire (QPSNordic) for Psychological and Social Factors at Work (Dallner et al., 2000). In line with this, role ambiguity was measured by three statements referring to the clarity of behavioral requirements at work (“Clear, planned goals and objectives have been defined for your job”; “You know what your responsibilities are”; “You know exactly what is expected of you at work”). Role conflict consisted of three statements referring to inconsistencies in the requirements of participants’ work roles (i.e., “You have to do things that you feel should be done differently”; “You are given assignments without adequate resources to complete them”; “You receive incompatible requests from two or more people”). Responses were measured by seven categories of responses ranging from “very seldom or never” to “very often or always” for both scales. Role ambiguity items were reverse scored, thus, the sum of both scales, the Role Stress scale indicate high level of role stress (role ambiguity and role conflict).

2.2.3. Frustration

To measure individual feelings of frustration, three items focusing on emotion-focused coping (“focus on and venting of emotions”) based on the brief-COPE scale were used (Carver et al., 1989; Polish adaptation by Juczyński & Ogińska-Bulik, 2009), following Van den Brande et al. (2017). Respondents were asked to consider three statements “I express my negative feelings at workplace,” “I express my frustration and irritation at the workplace,” and “I talk with my colleagues about difficulties at my work.” Participants could respond to the statements using seven categories of responses ranging from “very seldom or never” to “very often or always.” High scores on the Frustration scale indicate a high level of individual frustration at work.

2.2.4. Hostile work climate

We used eight items back-translated into Polish based on the Short Inventory to Monitor Psychosocial Hazards (Notelaers et al., 2007; Zahlquist et al., 2023) and Hostile Work Climate taken from the Psychosocial Work Environmental Questionnaire (PSYWEQ, Blomberg & Rosander, 2020; Rosander & Blomberg, 2018) to measure hostile climate in the workplace. The items are as follows: “My workplace is characterized by aggressiveness from colleagues,” “My workplace is characterized by aggressiveness from my boss,” “My workplace is characterized by conflicts with my colleagues,” “My workplace is characterized by conflicts with your my,” “My workplace is characterized by suspicion and rudeness,” “My workplace is characterized by conflicts and misunderstandings,” “At our workplace, the atmosphere is bad,” and “I feel unsafe at my workplace.” Responses were measured by seven categories of responses ranging from “very seldom or never” to “very often or always.” High scores on the Hostile Work Climate scale indicate high level of perceived hostile work climate.

3. Results

First, the means and standard deviations were calculated, and correlation analyses (Table 1) were conducted.

Table 1. Correlations and descriptive statistics between particular variable, reliabilities in parentheses (N = 353)

		1	2	3	4	5
1	Exposure to bullying T1	(.93)
2	Role stress T1	.33*	(.72)
3	Frustration T2	.34*	.29*	(.82)
4	Hostile climate T2	.37*	.61*	.59*	(.94)
5	Exposure to bullying T2	.59*	.22*	.39*	.40*	(.94)
	Min	22.00	6.00	3.00	8.00	22.00
	Max	100.00	42.00	21.00	56.00	101.00
	M	35.45	23.58	9.70	27.14	35.72
	SD	12.89	5.32	3.95	10.73	13.83

*p < .01

The correlation analysis revealed that the higher the stress role measured in wave 1, the higher the exposure to workplace bullying (in both waves 1 and 2), which supports H1. In line with H2, the higher the role stress measured in wave 1, the higher the individual frustration measured in wave 2. As predicted in H3, the higher the individual frustration measured in wave 2, the higher the perceived hostile work climate measured in wave 2. Hostile work climate was also related to exposure to bullying, as suggested by H4. Moreover, the higher exposure to bullying (in both waves 1 and 2), the higher the level of individual frustration and hostile work climate measured in wave 2.

To test H5 and H6, we ran a double mediation analysis (model No. 6) with role stress measured in wave 1 as an independent variable, frustration at work measured in wave 2 and hostile climate at work measured in wave 2 as mediators, and workplace bullying exposure measured in wave 2 as a dependent variable. We used sampling with replacement and with a bias-corrected bootstrapping procedure (5,000 samples) using the PROCESS macro (Hayes, 2013). The results indicate that both frustration at work and a hostile climate play the mediating role in the relationship between role stress and exposure to workplace bullying. We controlled for age, gender, and exposure to bullying measured in wave 1 (Figure 2). The overall (total effect) model explained R² = 36.1% of the variance in the dependent variable; F_{(4, 348)} = 49.17, p < .001.

Figure 2. Frustration and hostile work climate as a mediators of the role stress—exposure to bullying relationship (standardized coefficients).

* p < .01Standardised indirect effects: Total: = .13, se = .04, [.06, .21] role stress—frustration – exposure to bullying: = .03, se = .02, [.004, .07] role stress—hostile climate—exposure to bullying: = .08, se = .03, [.02, .15] role stress—frustration – hostile climate—exposure to bullying: = .02, se = .008, [.003, .04]

Role stress in wave 1 predicted frustration in wave W2 (coeff = .20, se = .04, t = 3.87, p < .001, CI [.07, .22]), controlling age (which predicted frustration in wave 2, coeff = .12, se = .02, t = 2.29, p = .02, CI [.006, .08]), gender (which did not predict frustration in wave 2, coeff = .01, se = .42, t = .26, p = .80, CI [−.72, .93]), and exposure to bullying in wave 1 (which predicted frustration in wave 2, coeff = .28, se = .02, t = 5.38, p < .001, CI [.05, .12]). These results provide additional support for H2.

Both role stress in wave 1 (coeff = .45, se = .08, t = 11.84, p < .001, CI [.76, 1.06]) and frustration in wave 2 (coeff = .44, se = .10, t = 11.58, p < .001, CI [1.002, 1.41]) predicted a hostile climate in wave 2, controlling age (which predicted a hostile climate in wave 2, coeff = -.09, se = .04, t = -.2.64, p = .009, CI [−.16, −.02]), gender (which did not predict a hostile climate in wave 2, coeff = -.04, se = .81, t = -.1.11, p = .27, CI [−2.50, .69]), and exposure to bullying in wave 1 (which marginally predicted a hostile climate in wave 2, coeff = .07, se = .03, t = 1.75, p = .08, CI [−.007, .12]). These results provide additional support for H3.

Finally, as predicted in H1, role stress in wave 1 marginally predicted exposure to bullying in wave 2 (coeff = -.10, se = .14, t = -.1.92, p = .056, CI [−.53, .007]), while both frustration in wave 2 (coeff = .15, se = .18, t = 2.88, p = .004, CI [.17, .90]) and a hostile climate in wave 2 (coeff = .18, se = .08, t = 2.88, p = .004, CI [.07, .39]) predicted exposure to bullying in wave 2 (in line with H4), controlling age (which predicted exposure to bullying in wave 2, coeff = -.12, se = .05, t = -.2.67, p = .008, CI [−.25, −.04]), gender (which did not predict exposure to bullying in wave 22, coeff = .009, se = 1.23, t = .21, p = .83, CI [−2.16, 2.66]), and exposure to bullying in wave 1 (which predicted exposure to bullying in wave 2, coeff = .50, se = .05, t = 10.93, p < .001, CI [.44, .63]).

Notably, while the direct effect of role stress in wave 1 on exposure to bullying in wave 2 was marginally significant (effect = -.26, se = .14, t = -.1.92, p = .056, CI [−.53, .007]), the total effect of role stress in wave 1 on exposure to bullying in wave 2 is no longer significant (effect = .07, se = .12, t = .60, p = .55, CI [−.16, .30]), controlling age (which predicted exposure to bullying in wave 2, coeff = -.11, se = .06, t = -.2.38, p = .02, CI [−.24, −.02]), gender (which did not predict exposure to bullying in wave 2, coeff = .005, se = 1.28, t = .11, p = .92, CI [−2.39, 2.65]), and exposure to bullying in wave 1 (which predicted exposure to bullying in wave 2, coeff = .58, se = .05, t = 12.62, p < .001, CI [.52, .71]).

Most importantly, the double mediation analysis revealed a significant indirect effect (full mediation) of role stress in wave 1 on exposure to bullying in wave 2 via frustration and hostile climate (total standardized indirect effect = .13, se = .04, [.06, .21]). All of the indirect effects that occurred were significant, which supports both H5 and H6 (Figure 2).

4. Discussion

Workplace bullying, with its detrimental negative effects (e.g., Conway et al., 2021; Hansen et al., 2021; Luo et al., 2023; Mikkelsen et al., 2020; Nielsen et al., 2020; Pauksztat et al., 2022), involves a gradually escalating process caused primarily by the immediate working environment and the design and management of work (Einarsen & Ågotnes, 2023). The present longitudinal study focused on identifying the mechanism underlying the well-documented relation between role stressors and exposure to workplace bullying. We examined and found support for six hypotheses. Firstly, applying AET (Weiss & Cropanzano, 1996) and based on the work environment hypothesis (Leymann, 1996), we replicated the previously obtained pattern of results (Harlos & Holmvall, 2021; Van den Brande et al., 2016) indicating that working under role stress is related to exposure to bullying six months later (H1). In line with H2, working under role stress is linked with experiencing frustration. As predicted in H3, individual frustration was related to a perceived hostile work climate. Finally, a hostile work climate was associated with exposure to bullying (H4). Most importantly, basing on both the revised FA theory (Berkowitz, 1989) and the SIP on aggression (Felson & Tedeschi, 1993; Neuman & Baron, 2011), we anticipated and revealed the double mediation effect of both individual frustration and a perceived hostile work climate in the relationship between role stressors and exposure to workplace bullying (H5 and H6). The study provides original findings with important theoretical and practical implications.

The aim of the present study was to expand our understanding of work-related features responsible for exposure to workplace bullying. Although the work environment hypothesis is well-documented (Balducci et al., 2021), the mechanism underlying the relationship between role stress and exposure to bullying remained empirically unsupported. Einarsen et al. (1994) assumed that role stress may cause frustration and stress, generating tensions and aggressive behaviors. An increased level of strain and interpersonal conflicts may, in turn, provide a reason for perpetrators to counteract when faced with frustrated co-workers’ norms violation (Bowling & Beehr, 2006). Applying the revised FA theory (Berkowitz, 1989), we indicated the mediating role of individual frustration in the relationship between working under contradictory and unclear demands and experiencing bullying.

Next, drawing on the SIP on aggression (Felson & Tedeschi, 1993; Neuman & Baron, 2011), we showed the mediating role of a hostile work climate between role stressors and exposure to bullying. The double mediation effect suggests a chain process in which role stressors trigger exposure to workplace bullying because of both individual frustration and a perceived hostile work climate. Thus, when information and requirements regarding work-related roles in the organization are unclear, and employees do not know what is expected from them or feel that they cannot perform their obligations because information is contradictory or absent, employees tend to express frustration, irritation, and disappointment, in line with FA theory (Berkowitz, 1989). Employees who vent their frustration may provoke tensions and interpersonal conflicts or be perceived as “notorious complainers” by co-workers (Baillien et al., 2009, p. 2), which is consistent with the SIP on aggression (e.g., Neuman & Baron, 2011). Such social circumstances may be perceived as a hostile work climate characterized by bitter, unfriendly, suspicious communication and deviant, aggressive behaviors within the work group (Mawritz et al., 2012, 2014). Finally, a hostile work climate, as found in previous studies (Zahlquist et al., 2023), is related to workplace bullying.

Accordingly, the findings not only support the work environment hypothesis (Einarsen et al., 2020; Hauge et al., 2010; Leymann, 1996), but provide an explanation of the mechanism underlying the association between work-related stressors, such as role ambiguity and role conflict, and experiencing bullying. Therefore, our results indicate that bullying may develop because of individual frustration and a perceived hostile work climate generated by organizational deficiencies.

To the best of our knowledge, this is the first study to show the mediating effects in the relationship between work environment and exposure to bullying (Nielsen & Einarsen, 2018; Rai & Agarwal, 2018). Previous research focused on the link between exposure to bullying and its consequences, for example, health problems (e.g., Casimir et al., 2012; Mikkelsen & Einarsen, 2002; Rodríguez-Muñoz et al., 2011; Vie et al., 2012) or job satisfaction (Glasø et al., 2011). Therefore, our findings are novel in identifying mediators between organizational antecedents and experiencing bullying. It is crucial to understand why certain organizational features trigger workplace bullying. The current study provides an explanation of how role ambiguity and role conflict generate exposure to bullying. The present study therefore contributes to the literature on workplace bullying antecedents by identifying new mediators in the process of bullying development, thus meeting the expectations of research on mediation and moderation effects (Nielsen & Einarsen, 2018).

4.1. Limitations and future research directions

Investigating mediators explaining the relationships between organizational antecedents and exposure to workplace bullying within a longitudinal design with a six-month time lag is certainly a strength of the present study from both a theoretical and a methodological perspective. A relatively large and heterogeneous sample including employees from various occupations is also a strength. However, the present study also has some limitations that must be addressed. Firstly, the collected responses in the study come from employees working in Poland. There may be some national cultural issues in the organizational context, especially in relation to workplace climate. Thus, future studies should either consider international samples or replicate the present results in different countries. Secondly, our data come from self-reported measures. Common method bias (Podsakoff et al., 2003) is a risk when examining self-report questionnaire data. However, a longitudinal two-wave design may decrease the risk of common method variance (Podsakoff et al., 2003). Moreover, in the present study, participants filled out questionnaires outside from the workplace context; thus, we may reason that socially desirable answers were low.

Next, in our study, frustration was measured by a scale focusing on emotion-focused coping (“focus on and venting of emotions”). We were inspired by Van den Brande et al. (2017) use of this scale to show that frustration measured in this way amplified the relationship between role conflict and exposure to workplace bullying. This method of measuring frustration is close to the theoretical assumptions of SIP on aggression (e.g., Neuman & Baron, 2011) where frustration is related to verbalizing and ventilating emotions to co-workers, which are met with aggression from other employees. Frustration can, however, be conceptualized and measured differently. Future research might concentrate on other conceptualizations of frustration and include the expression of various stress-related negative emotions, such as anxiety, irritation, disappointment, tension, depression, or job dissatisfaction, which have been shown to be outcomes of role stress (e.g., Örtqvist & Wincent, 2006; Spector, 1986; Spector & Jex, 1998). Based on stress—strain research, psychological (e.g., anxiety), physical (e.g., somatic symptoms), or behavioral (e.g., aggression) strain (e.g., Beehr, 1995; Jex & Beehr, 1991; Karasek & Theorell, 1990; Keashly et al., 1997) might also be included to measure frustration as a result of inability to perform work duties. Including such strains would be closer to the FA theoretical conceptualization of frustration (Berkowitz, 1989). Therefore, further conceptual replications of the current study are needed.

Moreover, as a hostile work climate is defined in terms of a group phenomenon, it should be measured on the group level, as Zahlquist and colleagues (Zahlquist et al., 2023) did. As we did not have aggregated data that would enable a multilevel approach, we measured a hostile work climate on the individual level by implementing a perceived hostile work climate scale, following, inter alia, Blomberg and Rosander (2020). However, to gain more insight into the process of bullying development, future studies should focus on including group-level measurements of hostile work climate. Moreover, when examining the relationship between a hostile climate and bullying exposure, it might be beneficial to include individual perpetrators’ traits, such as Machiavellianism (Turska & Pilch, 2016), and leadership practices (e.g., Stapinski et al., 2023) as moderators. As a positive correlation between an employee’s Machiavellianism and the displaying of bullying activities was found only among employees working in negative culture organizations (Turska & Pilch, 2016), it seems that a hostile work climate might provide ground for individuals with specific tendencies.

Last but not least, although our findings shed some light on the mechanism underlying the relationship between role stress and exposure to workplace bullying, the question of a causal relationship between particular variables needs further investigation. Future studies might include four instead of two waves to measure the double mediating effect of both experienced frustration and a hostile work climate in the relationship between role stress and exposure to workplace bullying with each variable measured in separate waves. To measure the workplace bullying process and the dynamics between working under role stressors, expressing frustration, a hostile work climate, and exposure to bullying, we might think of a study that uses a multi-wave design with shorter time lags (Nielsen & Einarsen, 2018). Such procedures might include more crucial mediators or moderators and would provide insight into bullying dynamics. Moreover, increasing personal growth by having ethical leaders in teams could help to reduce the effects that bullying may have on implicated individuals (Ruiz-Palomino et al., 2021).

4.2. Practical implications

Generally, our findings indicate that both individual frustration and a perceived hostile work climate caused by role stress are responsible for exposure to workplace bullying. These findings have several practical implications for managers and HR personnel. Significantly, our results suggest a hazardous chain of negative outcomes, here, employees’ frustration and a hostile work climate, of a dearth of well-structured working conditions. This result implies that bullying interventions should be planned at various stages of the bullying development process. Most importantly, interventions should concentrate on creating clear requirements and demands by reducing uncertainty. By examining organizational risk factors, that is, identifying frustrating insecurities, managers may not only reduce grievances but have more social control over task performance. Furthermore, our results point to the fact that managers should pay attention to identifying the sources of frustration, thus examining which employee goals or values are blocked. To decrease bullying development, signals of dissatisfaction, irritation, and frustration should be detected and solved. Thus, managers should let employees express complaints instead of suppress frustration and try to discuss alternatives to resolve the problem (Glasø et al., 2011; Mikkelsen & Einarsen, 2002).

Moreover, managers should focus on the organizational climate. The findings of the present study indicate that a hostile work climate is extremely harmful because it increases the chances of workplace bullying development. Shaping the workplace climate, as a set of shared attitudes, policies, practices, and values in a certain working group (De Rivera, 1992), might constitute a challenge. Thus, workplace bullying interventions should include enhancing the positive aspects of workplace climate, namely, by changing certain attitudes, introducing policies and procedures encouraging to use good practices, and influencing commonly shared values. This suggestion resonates with the notion that ethical infrastructure crucial in workplace bullying interventions (e.g., Einarsen et al., 2017), based on both formal (e.g., standardized procedures, such as codes of conduct) and informal (e.g., informal signals of organizational values, beliefs, traditions) characteristics. Other research focusing on workplace climate also indicates that, for example, a psychosocial safety climate (Zahlquist et al., 2019) or conflict management climate (Hamre et al., 2023) may hinder bullying progress and promote ethical behavior (Ruiz-Palomino et al., 2013). Moreover, a climate of responsibility weakens the relationship between unethical leadership and subordinates’ personal growth satisfaction (Ruiz-Palomino et al., 2021). Therefore, it seems crucial to investigate the interplay between particular organizational workplace bullying antecedents, such as role stressors, individual emotions, and frustration, as well as the workplace climate to create successful interventions.

5. Conclusions

In the present two-wave study, we documented the mechanism underlying the relationship between role stressors and exposure to workplace bullying. In line with AET, role stressors, that is, role ambiguity and role conflict, acted as indirect risk factors of exposure to bullying. Drawing on both the revised FA theory (Berkowitz, 1989) and the SIP of aggression (Felson, 1992; Felson & Tedeschi, 1993; Neuman & Baron, 2011), our findings indicated that individual feelings of frustration at work as well as a perceived hostile work climate mediate the relationship between role stressors, that is, role ambiguity and role conflict, and exposure to workplace bullying. We believe that our findings will inspire further studies on the interplay between particular workplace bullying organizational risk factors and shed more light on the process of responsible workplace bullying development. We also hope that our research findings will serve managers as guidelines to protect their employees from exposure to workplace bullying.

Disclosure statement

No potential conflict of interest was reported by the author(s).