Introduction
Hierarchical bullying, where supervisors are the source of bullying, is a major concern within Australian public sector organizations (Azira-Montes, Leal-Rodriguez, & Leal-Millan, Reference Azira-Montes, Leal-Rodriguez and Leal-Millan2015; Crimp, Reference Crimp2017; Hurley, Hutchinson, Bradbury, & Browne, Reference Hurley, Hutchinson, Bradbury and Browne2016; Kleizen, Wynen, Boon, & De Roover, Reference Kleizen, Wynen, Boon and De Roover2021; Ng, Franken, Nguyen, & Teo, Reference Ng, Franken, Nguyen and Teo2022), with the incidence and impact being even more pronounced for those delivering public healthcare services (Einarsen, Hoel, & Zapf, Reference Einarsen, Hoel and Zapf2011; Hurley et al., Reference Hurley, Hutchinson, Bradbury and Browne2016; SafeWork, 2021). Healthcare workers make up one of the highest proportions (18%) of Australian serious workers' compensation claims for mental stress (SafeWork, 2021). Hierarchical bullying is a type of deviant and destructive behaviour that occurs when a supervisor targets an individual or a group of subordinates (Caillier, Reference Caillier2020; Demir & Rodwell, Reference Demir and Rodwell2012; Farr-Wharton et al., Reference Farr-Wharton, Brunetto, Hernandez-Grande and Brown2022; Norton et al., Reference Norton, Costa, Teixeira, Azevedo, Roma-Torres, Amaro and Cunha2017). Victims are often subjected to ridicule, social isolation, professional undermining, an excessive workload, malicious gossip and the assignment of menial tasks by perpetrators with greater perceived workplace power and authority (De Cieri, Sheehan, Donohue, Shea, & Cooper, Reference De Cieri, Sheehan, Donohue, Shea and Cooper2019; Omari & Paull, Reference Omari and Paull2016).
Unchecked, hierarchical bullying leads to detrimental outcomes for victims, witnesses and the teams and organizations where it occurs (De Cieri et al., Reference De Cieri, Sheehan, Donohue, Shea and Cooper2019; Kline & Lewis, Reference Kline and Lewis2019). Meta-analytic reviews link bullying with higher instances of depression, anxiety, suicidal ideation, insomnia, somatic stress and sick leave (Lever, Dyball, Greenberg, & Stevelink, Reference Lever, Dyball, Greenberg and Stevelink2019; Potter, Dollard, & Tuckey, Reference Potter, Dollard and Tuckey2016). For healthcare workers, recent research links hierarchical bullying with diminished employee wellbeing and increased intention to quit (Farr-Wharton et al., Reference Farr-Wharton, Brunetto, Hernandez-Grande and Brown2022); decreased employee engagement, satisfaction and organizational commitment and increased turnover (Demir & Rodwell, Reference Demir and Rodwell2012; Hurley et al., Reference Hurley, Hutchinson, Bradbury and Browne2016; Ng et al., Reference Ng, Franken, Nguyen and Teo2022; Rodwell, Brunetto, Demir, Shacklock, & Farr-Wharton, Reference Rodwell, Brunetto, Demir, Shacklock and Farr-Wharton2014).
The prominence of bullying in healthcare, despite it being at odds with the healthcare workforce's core function of advancing quality of life, is paradoxical (Adams & Maykut, Reference Adams and Maykut2015; Eley, Eley, & Rogers-Clark, Reference Eley, Eley and Rogers-Clark2010). The main antecedents of hierarchical bullying in healthcare are consistently identified as a poor social climate, lack of interpersonal trust and support, task conflict, status inconsistencies and power asymmetry (Rodwell et al., Reference Rodwell, Brunetto, Demir, Shacklock and Farr-Wharton2014).
Traditional strategies to address workplace bullying through training and harm reduction are negatively focussed, and have had limited effectiveness (Atkinson & Jones, Reference Atkinson and Jones2018; Hodgins, MacCurtain, & Mannix-McNamara, Reference Hodgins, MacCurtain and Mannix-McNamara2020). Hurley et al. (Reference Hurley, Hutchinson, Bradbury and Browne2016: 17) argue that the extant organizational bullying mitigation strategies:
…are in part creating the conditions for further victimization and exacerbation of the mental distress already generated by the bullying.
A systematic review of the literature about bullying in nursing by Hartina et al. (Reference Hartina, Tahir, Nurdin and Djafar2018) identified a significant gap as being the need for effective workplace practices to counter bullying. The emphasis has been on enhancing individual ‘resilience’ as a post-traumatic growth response (Heugten, Reference Heugten2012). Scholars argue, however, that this unfairly places responsibility on the victim, allowing organizations to neglect their obligation to protect employees from harm (Oliver, Reference Oliver2017).
More recently, researchers have postulated that an alternative way to alleviate the effects of workplace bullying is by activating social support systems within workplaces through specific programmes promoting kindness, empathy and compassion for colleagues (Ahmad, Islam, D'Cruz, & Noronha, Reference Ahmad, Islam, D'Cruz and Noronha2023; Simpson, Farr-Wharton, & Reddy, Reference Simpson, Farr-Wharton and Reddy2020). Organizational compassion is a positive interpersonal workplace process that involves colleagues (either individually, or as a group) noticing and feeling a co-worker's suffering, interpreting the situation and taking action to provide relief (Dutton, Workman, & Hardin, Reference Dutton, Workman and Hardin2014). As a result, the suffering co-worker perceives understanding and support from colleagues. It is important to note that organizational compassion is often expressed at a team-department level (Dutton, Worline, Frost, & Lilius, Reference Dutton, Worline, Frost and Lilius2006; Lilius, Worline, Dutton, Kanov, & Maitlis, Reference Lilius, Worline, Dutton, Kanov and Maitlis2011). It is distinct from related concepts – ‘self-compassion’, in which compassion is directed inward, and ‘compassion’, which involves feeling sympathy for another's suffering (Neff, Reference Neff2011), as these concepts are not necessarily contextualized within organizational settings.
The present study examines the effect of hierarchical bullying (by supervisors) on employee wellbeing, and the moderating effect of a team's compassionate response. A multilevel statistical model is employed to analyse data on employees' perceptions of hierarchical bullying, compassion and wellbeing, nested within their workgroups (in effect, the team of people they work with). The analysis uses the 1 X (1 → 1) design of Preacher, Zhang, and Zyphur (Reference Preacher, Zhang and Zyphur2016), and aggregates the nested data on compassion to the between-level of analysis, thereby moderating the interaction between hierarchical bullying and wellbeing at the within-level. This approach is consistent with the research of Wee and Fehr (Reference Wee and Fehr2021), who similarly modelled team compassion (at the between-level) as a moderator between individual perceptions of COVID-19 events and individual suffering (with the latter two variables analysed at the within-level). Accordingly, these phenomena are multilevel, involving both social interactions and individual emotional processing, nested within organizational contexts (Ashkanasy, Troth, Lawrence, & Jordan, Reference Ashkanasy, Troth, Lawrence and Jordan2017; Dunn, Masyn, Yudron, Jones, & Subramanian, Reference Dunn, Masyn, Yudron, Jones and Subramanian2014). The model is informed by social exchange theory (SET), which explains the interpersonal dynamics of hierarchical bullying, and compassion, and conservation of resource theory (CoRT), which explains the impact of hierarchical bullying and compassion on individuals' emotional processing and wellbeing. The study aims to answer the following research question:
Research question: To what extent does a team-level compassionate response moderate the relationship between hierarchical bullying and employee wellbeing?
This paper extends prior research about the bullying–wellbeing nexus (see, e.g., Demir & Rodwell, Reference Demir and Rodwell2012; Farr-Wharton et al., Reference Farr-Wharton, Brunetto, Hernandez-Grande and Brown2022; Ng et al., Reference Ng, Franken, Nguyen and Teo2022; Tummers, Brunetto, & Teo, Reference Tummers, Brunetto and Teo2016) by using multilevel structural equation modelling (MSEM) to explore the role of a team-level compassion response in shaping employee wellbeing outcomes at the individual level. The paper offers new insights into the behavioural and relational responses of employees exposed to bullying.
Background
Social exchange theory (SET)
The present study examines the impact of hierarchical bullying on employee wellbeing and the mitigating effect of team compassion, with SET as one of the informing theories. SET posits that positive interactions between organizational actors foster trustful relationships that engender positive social responses likely to promote mutual reciprocity over time (Brunetto et al., Reference Brunetto, Xerri, Trinchero, Beattie, Shacklock, Farr-Wharton and Borgonovi2018; Cropanzano & Mitchell, Reference Cropanzano and Mitchell2005). Conversely, negative acts, such as hierarchical bullying, likely promote negative reactions and have harmful effects on employees.
Cropanzano, Anthony, Daniels, and Hall (Reference Cropanzano, Anthony, Daniels and Hall2017) identify a lack of differentiation between positive and negative impacts of an initial action in SET research and call for more in-depth analysis of how the target of an initial act (such as bullying by supervisors, or managers) informs the response of those affected (such as a victim, and their work team). Responding to this observation, our study assesses the extent to which team compassion can mitigate the harmful effects of hierarchical bullying on wellbeing. According to SET, a positive experience, such as that of receiving team compassion, has the potential to elicit a reciprocal positive emotional response in the form of wellbeing for the victim or victims. This positive exchange must be stronger than the negative exchange created by bullying, which previous research has shown decrease employee wellbeing (Demir & Rodwell, Reference Demir and Rodwell2012; Farr-Wharton et al., Reference Farr-Wharton, Brunetto, Hernandez-Grande and Brown2022; Ng et al., Reference Ng, Franken, Nguyen and Teo2022; Tummers, Brunetto, & Teo, Reference Tummers, Brunetto and Teo2016).
Cropanzano et al. (Reference Cropanzano, Anthony, Daniels and Hall2017: 488) note that bullying often ‘Start(s) at high levels of the organization and then trickles down’, leading to negative impacts on wellbeing. They argue that SET has limitations in capturing how an initial workplace interaction could trigger further behavioural and relational responses. For example, a victim of hierarchical bullying may withdraw from their work context (behavioural – and at the individual-level), inducing bystanders to step forward to offer support (relational – at the team or between-level). Our study seeks to address this gap in the literature and shed light on the ways in which team compassion can moderate the impact of hierarchical bullying on employee wellbeing.
Conservation of resources theory (CoRT)
CoRT explains behavioural motivation and is used in this paper to examine the emotional impact of hierarchical bullying on an individual employee and how compassionate support from co-workers may affect this process. CoRT posits that people seek to conserve resources, defined as anything an individual values (including physical assets, health, relationships and agency), and that perceived threats to resources, including through bullying, cause stress, which undermines wellbeing (Hobfoll, Reference Hobfoll1989, Reference Hobfoll2001; Hobfoll, Halbesleben, Neveu, & Westman, Reference Hobfoll, Halbesleben, Neveu and Westman2018). Relevant to this study, some scholars include supervisor support (Guan & Frenkel, Reference Guan and Frenkel2019; Wang, Li, Zhou, Maguire, Zong, & Hu, Reference Wang, Li, Zhou, Maguire, Zong and Hu2019) and co-worker support (Cordes & Dougherty, Reference Cordes and Dougherty1993) as valued resources.
The first principle of CoRT predicts that individuals react more strongly to a perceived loss of resources than to resource gains. A sustained loss of resources can lead to a resource ‘loss spiral’ and dysfunctional coping strategies (Buchwald & Shchwarzer, Reference Buchwald, Shchwarzer, Peterson, Baker and McGaw2010; Hobfoll, Reference Hobfoll2001). The argument of this paper is that in the case of bullying, when a subordinate is exposed to bullying from their supervisor, their emotional resources are depleted, and over time they adopt dysfunctional coping mechanisms such as lashing out and withdrawal, leading to diminished wellbeing (Bernstein & Trimm, Reference Bernstein and Trimm2016). Alternatively, if co-workers offer compassionate support, CoRT would predict that this could lead to resource conservation and replenishment, promoting a positive ‘resource gain spiral’ (Buchwald & Shchwarzer, Reference Buchwald, Shchwarzer, Peterson, Baker and McGaw2010; Simpson, Rego, Berti, Clegg, & Cunha, Reference Simpson, Rego, Berti, Clegg and Pina e Cunha2022). Compassionate actions by co-workers can enhance a sense of control, encourage functional coping strategies, and mitigate the impact of hierarchical bullying on personal wellbeing. However, if bystanders do not act, bullying can be legitimized (Hutchinson, Reference Hutchinson2013). Some co-workers may choose not to intervene directly due to the power imbalance (Berti & Simpson, Reference Berti and Simpson2021), but they can still offer interpersonal compassion to the victim, such as by consoling, comforting and generally helping them, which can aid in regulating the victim's emotions and provide a broader perspective on the systemic issues behind the bullying (Nel, Reference Nel2019).
Compassionate social exchange can activate emotional processing in the victim and help to prevent or reduce resource depletion. Through this process, the victim may compartmentalize the bullying as a symptom of broader systemic problems (particularly endemic to public sector, and especially healthcare sector contexts), instead of seeing it as a personal attack. In conclusion, the compassion support provided by co-workers has the potential to promote resource conservation and help to mitigate the impact of hierarchical bullying on an individual's wellbeing.
Independent variable: hierarchical bullying
The effects of workplace bullying on employee engagement, organizational commitment and turnover have been widely documented (Demir & Rodwell, Reference Demir and Rodwell2012; Rodwell et al., Reference Rodwell, Brunetto, Demir, Shacklock and Farr-Wharton2014). In severe cases, it can lead psychological distress and long-term health consequences such as depression and emotional exhaustion (Potter, Dollard, & Tuckey, Reference Potter, Dollard and Tuckey2016). In Australia, the impact of workplace bullying is seen in high sick leave rates and elevated stress-related workers compensation claims, particularly among Australian healthcare professionals (SafeWork, 2021). This results in low morale, productivity and wellbeing (Magee, Gordon, Caputi, Oades, Reis, & Robinson, Reference Magee, Gordon, Caputi, Oades, Reis and Robinson2014), which can negatively affect patient care and morbidity (Felblinger, Reference Felblinger2008; Johnson, Reference Johnson2009; Laschinger, Reference Laschinger2014).
A consistent theme in the delivery of health and social public services is the prevalence of hierarchical bullying, which begins with senior management under-resourcing departments in pursuit of efficiencies and permeates down the organizational hierarchy (Kerasidou, Reference Kerasidou2019; Tummers, Brunetto, & Teo, Reference Tummers, Brunetto and Teo2016). Under-resourcing creates an environment where line managers resort to bullying to coerce their subordinates to consistently deliver, despite insufficient resources. This dynamic, in turn, erodes employees' ability to give and receive honest feedback, hindering collaboration.
Austerity-led funding models exacerbate the strain on cohesion within healthcare teams, normalizing hierarchical bullying (Esteve, Schuster, Albareda, & Losada, Reference Esteve, Schuster, Albareda and Losada2017; Farr-Wharton et al., Reference Farr-Wharton, Brunetto, Hernandez-Grande and Brown2022; Pollitt & Bouckaert, Reference Pollitt and Bouckaert2017). Where they may be policies in place aimed at deterring and managing bullying, these policies are often not effectively implemented or have little impact due to the pervasive influence of austerity measures dominating workplace behaviours (Brunetto et al, Reference Brunetto, Xerri, Trinchero, Farr-Wharton, Shacklock and Borgonovi2015).
Dependent variable: employee wellbeing
Employee wellbeing is a complex concept that is defined and conceptualized differently across disciplines. In this paper, we define employee wellbeing as a combination of hedonic (mood) and eudemonic (consistency of work tasks with one's values) components (Brunetto, Teo, Shacklock, & Farr-Wharton, Reference Brunetto, Teo, Shacklock and Farr-Wharton2012; Forgeard, Jayawickreme, Kern, & Seligman, Reference Forgeard, Jayawickreme, Kern and Seligman2011). As noted above, viewed through the lens of CoRT, hierarchical bullying is seen to result in resource loss and leads to increased stress and decreased wellbeing (Bernstein & Trimm, Reference Bernstein and Trimm2016; Nel, Reference Nel2019). A body of literature provides consistent evidence of an association between workplace bullying and negative wellbeing outcomes in nursing and healthcare professionals globally. For example, Nelson et al. (Reference Nelson, Azevedo, Dias, Sousa, Carvalho, Silva and Rabelo2014) found a negative association between workplace bullying and wellbeing among Brazilian nurses, while Harb, Rayan, and Al. khashashneh (Reference Harb, Rayan and Al. khashashneh2021) and Sauer and McCoy (Reference Sauer and McCoy2017) reported workplace bullying exposure was related to lower mental health and wellbeing in Jordanian and US nurses, respectively. Burnout, a predictor of leave intentions and poor worker health, has also been positively linked to workplace bullying in various healthcare settings (e.g., Allen, Holland, & Reynolds, Reference Allen, Holland and Reynolds2015; Karatza, Zyga, Tziaferi, & Prezerakos, Reference Karatza, Zyga, Tziaferi and Prezerakos2016; Laschinger, Grau, Finegan, & Wilk, Reference Laschinger, Grau, Finegan and Wilk2010). Given these findings and, particularly the work of Farr-Wharton, Brunetto, Xerri, Shriberg, Newman, and Dienger (Reference Farr-Wharton, Brunetto, Xerri, Shriberg, Newman and Dienger2019), who found a strong negative correlation between harassment by supervisors and employee wellbeing in nursing contexts in the UK and USA (at the single/individual level), we propose the following hypothesis:
Hypothesis 1: Higher perceptions of hierarchical bullying are associated with lower levels of employee wellbeing.
Moderator: team-level compassion
Engaging in compassion relations has been widely recognized as an important interpersonal competency for healthcare professionals, contributing to higher hospital ratings and patient satisfaction (McClelland & Vogus, Reference McClelland and Vogus2014). Research conducted among nurses in China found that receiving compassion, regardless of its source (colleagues, superiors or patients), leads to improved subjective health and reduces work-related stress (Zhang et al., Reference Zhang, Liu, Wang, Shi, Xie, Cang and Fan2018). Further, research on the compassion capabilities of healthcare workers ‘that enable employees of a collective unit to notice, feel and respond to members' suffering’ has been linked to ‘high quality connections’ (Lilius et al., Reference Lilius, Worline, Dutton, Kanov and Maitlis2011: 873) and higher levels of collective commitment and positive emotion (Lilius, Worline, Maitlis, Kanov, Dutton, & Frost, Reference Lilius, Worline, Maitlis, Kanov, Dutton and Frost2008).
Recently, research has narrowed its focus more explicitly to the concept of ‘team compassion’, which refers to the ‘extent to which team members as a whole engage in empathetic reactions to members’ suffering’ (Wee & Fehr, Reference Wee and Fehr2021: 1085). This idea has been explored further by Vanstone et al. (Reference Vanstone, Sadik, Smith, Neville, LeBlanc, Boyle and Toledo2020) who focus on team dynamics that give rise to collective compassionate acts within palliative care workgroups.
Viewed through an SET lens, team compassion as a process of noticing, empathizing, making sense of and responding to a co-worker's suffering to alleviate their distress (Dutton, Workman, & Hardin, Reference Dutton, Workman and Hardin2014; Simpson, Farr-Wharton, & Reddy, Reference Simpson, Farr-Wharton and Reddy2020) will likely induce a positive response of consoling the team member. Receiving compassion from workgroup members can assist a victim of hierarchical bullying in emotionally processing the event, potentially de-legitimizing the bully's actions and mitigating resource loss. Thus, we hypothesize that team compassion moderates the relationship between hierarchical bullying and employee wellbeing, buffering or reducing the negative impact of bullying on wellbeing as a result of the compassion-induced social–individual emotional processing.
Hypothesis 2: Team compassion moderates the relationship between hierarchical bullying and employee wellbeing.
The hypothetical model is displayed in Figure 1.
Figure 1. Hypothetical model.
Methods
Population sample
The research was conducted within a large public health organization, in Sydney, Australia, consisting of six hospitals and 14 community healthcare centres, all in proximity and with a central administering body. The organization was designed to service a local population of approximately 650,000 people, with a total of 11,000 employees. All staff were invited to participate in a survey through several means, including emails, an internal newsletter, QR codes displayed on posters at work sites and electronic tablets with research assistants present to assist. In total, 632 valid responses were received, representing a response rate of 6% across the organization. The respondents included 211 nurses, 57 physicians/surgeons, 130 allied health staff, 58 staff from corporate and management services, 114 from administrative services, one oral hygienist, seven hospital volunteers and 50 employees classed as ‘other’ (mainly support roles). The gender distribution was 508 identified as female, 103 male and 21 nonbinary or transgender. The age distribution was 122 participants aged 29 and under, 253 aged 30–49 and 256 aged 50 and over.
The respondents were nested within 48 separate workgroups, with an average cluster size of 13.17 individuals per group. The workgroups represented functional teams (i.e., employees who worked together in the same unit), and included teams of staff involved in medical, dental, nursing, pathology and diagnosis, community health and aged care functions. As part of the survey, respondents were asked to identify their workgroup from a drop-down list.
Measures, multilevel reliability and validity
Previously validated survey instruments were used to measure the constructs of hierarchical bullying, organizational compassion and employee wellbeing. The original wording of the measures was maintained in the survey, with the data and analysis nested within the workgroup/team setting in accordance with multilevel analysis. The survey items retained in the analysis are provided in Appendix 1.
Instruments used
Hierarchical bullying was measured using the items from the ‘Work-Related Bullying’ (WRB) subscale of the ‘Negative Acts Questionnaire Revised’ (NAQ-R) developed by Einarsen, Hoel, and Notelaers (Reference Einarsen, Hoel and Notelaers2009). The items within this subscale encompass a variety of hierarchical bullying behaviours, such as excessive monitoring, unreasonable deadlines, ignored opinions and pressure not to claim entitled benefits. Respondents rated their experiences with behaviours over the past 6 months using a 5-point Likert scale (1 = never, 2 = now and then, 3 = monthly, 4 = weekly, 5 = daily). The scale is presented in Appendix 1.
Organizational compassion was measured using the NEAR Organizational Compassion Scale (Simpson & Farr-Wharton, Reference Simpson and Farr-Wharton2017), which captures respondents' perceptions of four sub-processes: noticing a co-workers' suffering at work, empathizing, assessing and responding to the suffering. A higher order organizational compassion variable was created by loading the individual items onto it, as is recommended by Simpson and Farr-Wharton (Reference Simpson and Farr-Wharton2017). To improve the overall strength of the model and increase parsimony, several items from the original 22-item NEAR Organizational Compassion Scale with low factor loads in the between-level of analysis were removed. The remaining scale consisted of four items for noticing, three items for empathizing, three items for assessing and five items for responding. A 6-point Likert scale was used (1 = strongly disagree, 2 = disagree, 3 = slightly disagree, 4 = slightly agree, 5 = agree and 6 = strongly agree), in line with the original publication of the scale.
Employee wellbeing was measured using a four-item scale developed by Brunetto et al. (Reference Brunetto, Farr-Wharton and Shacklock2011), which captures both the hedonic and eudemonic elements (mood and values alignment). A sample item is ‘Overall, I am reasonably happy with my work life’. Consistent with the original publication, a 6-point Likert scale was used (1 = strongly disagree, 2 = disagree, 3 = slightly disagree, 4 = slightly agree, 5 = agree and 6 = strongly agree). Recent research has concluded that no improvement in psychometric precision is identified past six response options, and thus this even-spreaded Likert is gaining increasing acceptance (Simms, Zelazny, Williams, & Bernstein, Reference Simms, Zelazny, Williams and Bernstein2019).
Geldhof, Preacher, and Zyphur (Reference Geldhof, Preacher and Zyphur2014) argue that the assessment of instrument reliability in multilevel statistical models cannot be performed using traditional, single-level approaches or aggregated mean factor loads, as they do not account for factor load variation at both the within and between levels of analysis. They propose a multilevel confirmatory factor analysis (MCFA) process to test instrument reliability at both levels of analysis. Using this framework, the reliability alpha for both the within- and between-level constructs was calculated, and the results are presented in Table 1. All items possessed appropriate reliability with alpha scores above ‘.7’, as indicated by Geldhof, Preacher, and Zyphur (Reference Geldhof, Preacher and Zyphur2014: 72), who state that the two-level reliability alpha score ‘performs relatively well in most settings’.
Table 1. Within- and between-level instrument reliability
To assess the discriminant validity of variables in the multilevel model (MLM), we compared two models: a freely estimated MCFA and a model where all correlations are constrained to 1. The results showed that the fit of the freely estimated model was superior, with a χ2 over degrees of freedom (CMIN/df) of .00 (p value = 1.00), a corrected fit index (CFI) of 1.00, a Tucker–Lewis index (TLI) of 1.00, a standardized root mean square residual (SRMR) within of .00 and SRMR between of .00. The constrained model had a significantly worse fit, with a CMIN/df of 6.54 (p value = .00), a CFI of .86, a TLI of .91, SRMR within of .06 and an SRMR between of .55. These results suggest that the discriminant validity of the variables in the MLM was supported as the constrained model was statistically inferior.
To reduce the likelihood of method bias, we took several precautions. The survey was brief, the collection method was convenient (either online or face-to-face) and organizational leadership endorsed participation. Furthermore, we conducted a posthoc Harmon's single-factor analysis, which explained 40.49% of the variance, indicating a low chance of method bias.
Analysis
The data were analysed using the MSEM framework, as proposed by Preacher, Zhang, and Zyphur (Reference Preacher, Zhang and Zyphur2016), within the Mplus software. This framework, specifically the 1 × (1→1) design, was used to test multilevel moderation. In this design, all constructs were collected at the within-level (employee level), while the moderator, which represented the team/between-level effect when aggregated, was drawn from the between-level. This type of multilevel moderation model has been recently applied in other research exploring team-compassion (see Wee & Fehr, Reference Wee and Fehr2021), where team-compassion was modelled as a moderator between COVD-19 events and individual suffering.
In the path model, the role of each respondent served as a control variable at the within-level, while the workgroup's location was used as a control variable at the workgroup level. However, neither variable was significant. The final path model employed two-level random intercepts and utilized the recommended Bayes estimator (Preacher, Zhang, & Zyphur, Reference Preacher, Zhang and Zyphur2016). The interclass correlation for employee wellbeing was 22.6%, indicating a significant multilevel effect. Table 2 provides the mean, correlation and variance statistics for the within- and between-level variables. To maintain model parsimony, a linear composite was created for each construct.
Table 2. Descriptive statistics and correlations
*p > .05, **p > .01, ***p > .001. N.S., not significant, within n = 632, between n = 48.
Results
The means, standard deviation (SD), and the within- and between-level correlations of the constructs is displayed in Table 2.
The model fit indices for the final path model, which utilized linear composites, were robust, with a CMIN/df score of .00 (p value = 1.00), a CFI of 1.00, a TLI of 1.00, an SRMR of .00 within and between. The moderation effects, as well as the results for the final path model, are depicted in Figures 2 and 3.
Figure 2. Results from multilevel moderations 1 × (1−>1) model. *p > .5, **p > .01, ***p > .000. Within n = 632, between n = 48. Estimator = Bayes (one tailed), integration = 4.
Figure 3. Interaction testing – hierarchical bullying on wellbeing moderated by team-level compassion
A significant negative relationship was observed between hierarchical bullying and employee wellbeing, supporting the first hypothesis. The results also indicated a significant moderating effect of team compassion on relationship between hierarchical bullying and employee wellbeing, thus providing evidence of support for the second (moderation) hypothesis. In sum, when the level of team compassion was high, the detrimental impact of hierarchical bullying on employee wellbeing was mitigated.
Discussion
This study aimed to examine the moderating role of team compassion on the impact of hierarchical bullying on the wellbeing of healthcare workers using SET and CoRT as theoretical frameworks to develop hypotheses. The research builds on work of previous scholars (Atkinson & Jones, Reference Atkinson and Jones2018; Hodgins, MacCurtain, & Mannix-McNamara, Reference Hodgins, MacCurtain and Mannix-McNamara2020; Hurley et al., Reference Hurley, Hutchinson, Bradbury and Browne2016) who have highlighted the importance of finding successful organizational strategies to address hierarhical bullying in the workplace as a matter of urgency. In line with Simpson, Farr-Wharton, and Reddy (Reference Simpson, Farr-Wharton and Reddy2020), the study aimed to examine the effectiveness of promoting team compassion to buffer against the negative effects of hierarchical bullying.
The results showed a negative association between hierarchical bullying and wellbeing and significant moderation effect from team compassion on the path linking hierarchical bullying and employee wellbeing (Figures 2 and 3). The findings suggest that higher levels of team compassion provide greater protection for individual employees against the negative effects of hierarchical bullying. Indeed, for teams with high levels of compassion and high levels of hierarchical bullying, wellbeing was higher compared to teams with low hierarchical bullying. This finding highlights the important exchange between team compassion and individual wellbeing and prompts the need for further research to validate these findings.
The findings of this study have important implications for evidence-based best practice in organizations. Investing in cultivating compassion within work teams can act as a buffer against some of the harmful effects of hierarchical bullying. Team members showing compassion towards a bullied colleague constitutes a pathway for offering support without eliciting unwanted organizational/managerial responses (Simpson et al, 2017). SET would also suggest that the act of supporting a colleague may also positively enhance the giver's wellbeing, an idea that appears to have found support in clinical trials (Cosley, McCoy, Saslow, & Epel, Reference Cosley, McCoy, Saslow and Epel2010), although it requires further investigation in an applied setting. Additionally, further research should examine the antecedents of team compassion, including the potential impact of role/profession homogeneity in enhancing social identity and strengthening team compassion.
The findings contribute to SET theorizing by highlighting a multi-dimensional employee response to bullying. Cropanzano et al. (Reference Cropanzano, Anthony, Daniels and Hall2017) previously pointed out that SET scholars have failed to differentiate between employees' initial reaction to bullying and their subsequent behavioural and relational responses, assuming instead a linear response process. Prior research assumed that an initial act of hierarchical bullying would result in a negative behavioural response by employees. However, the current findings provide greater insight into how such a response can be moderated under certain conditions. Specially, the results indicate that, under certain circumstances, an initial hierarchical bullying interaction does not necessarily result in a negative impact on wellbeing when considering a team-level dynamic of compassion.
The current study advances our understanding of employee responses to hierarchical bullying by highlighting the moderating role of group compassion. Previous research assumed a linear response process, where hierarchical bullying would result in a negative response from employees. However, the findings of this study indicate that in work groups with high levels of compassion, other employees may provide compassion to a victim, which somewhat reduces the initial negative impact of bullying. The use of multi-level moderation analysis enabled the differentiation of employee responses to hierarchical bullying in work groups with high levels of compassion. These insights offer new perspectives on how organizations can leverage employees’ interpersonal resources as a complementary approach to managing the effects of hierarchical bullying. The findings call for further research into effective workplace interventions for fostering team-level compassion in organizations.
The research findings also contribute to CoRT (Hobfoll, Reference Hobfoll and Folkman2011) by explaining that employees who experience hierarchical bullying within highly compassionate teams are likely to have multiple responses. As expected, employees do probably respond initially to hierarchical bullying according to the first principle of CoRT which is that they likely perceive a loss of resources (Hobfoll, Reference Hobfoll2001). For those employees that work in groups that demonstrate low compassion, it is likely that over time, continual bullying will invoke the second corollary that employees will perceive a spiral loss of resources (likely evident in continual decreases in their wellbeing). In the worst case, the fourth corollary may come into play as employees adopt dysfunctional coping strategies (lashing out, withdrawal and avoidance, leading to turnover, depression and ill-health) as a response to a continual negative loss spiral (Buchwald & Shchwarzer, Reference Buchwald, Shchwarzer, Peterson, Baker and McGaw2010). The underlying mechanism being eroded is employee wellbeing, which may lead to further negative outcomes such as reduced workdays, income, social activities and relations, etc. (Bernstein & Trimm, Reference Bernstein and Trimm2016). Such outcomes appear increasingly plausible, particularly because healthcare workers make up one of the highest proportion (18%) of serious workers' compensation claims for mental stress in Australia (SafeWork, 2021). Although other variables such as bullying by patients or patient's families are also likely a factor (Spector et al., 2014), in this study we specifically measured the relationship between workplace compassion, wellbeing and hierarchical bullying.
In contrast, those victims who belong to teams that demonstrate high compassion are likely to have multiple responses available. The impact of social resources in the form of compassion from colleagues can positively enhance an employees' individual cognition and counter the harmful impact which could have initiated a downwards spiral from the emotional processing of being exposed to hierarchical bullying. The new knowledge is that team-compassion is a social resource that can limit a victim's emotional resources from becoming depleted. Alternatively, should work colleagues undertake compassionate actions towards a colleague who is a victim of bullying, according to CoRT, there is potential for resource maintenance, and perhaps even resource gains (Buchwald & Shchwarzer, Reference Buchwald, Shchwarzer, Peterson, Baker and McGaw2010).
Limitations
The limitations of study include reliance on cross-sectional data collected at a single time point through surveys, which may introduce the risk of common method bias. However, the results of Harmon's single factor test suggest a low probability of method bias affecting the results. Notwithstanding these limitations, replication of the findings is needed, both within and outside of healthcare contexts. Until then, the findings should be considered exploratory in nature.
Conclusions
The prevalence of negative work acts, particularly hierarchical bullying, has been a persistent issue in healthcare workplaces, and efforts to curb this behaviour have had limited success. The current study explored the role of team compassion as a moderator of the relationship between hierarchical bullying and employee wellbeing in a healthcare setting. Using multilevel statistical analysis, the study found a strong moderating effect of team compassion, suggesting that teams with high levels of compassion can buffer against the negative effects of hierarchical bullying on individual wellbeing. Despite the encouraging results of this exploratory study, further research is required to confirm these findings and to investigate the antecedent factors that contribute to the development of team compassion.
Appendix 1
Questionnaire items
Open access
