This paper examines whether or not the psychological conditions of meaningfulness, safety and availability mediate the relationship between workplace bullying and intention to leave among South African employees. By doing so, the study seeks to contribute to the literature by clarifying the relational and meaning-based mechanisms through which bullying is associated with withdrawal cognitions. By analysing these psychological conditions, the study also aims to provide insight into how organisations can mitigate the harmful effects of workplace bullying and foster a work environment that enhances employee retention.

Workplace bullying refers to ‘harassing, offending, or socially excluding someone or negatively affecting someone’s work’ (Einarsen et al., 2011, p. 22). According to this view, the bullying label is applied to negative behaviour that has occurred frequently (such as weekly) and over a duration of time (such as over a 6-month period). Bullying behaviours can range from active actions, such as verbal threats and verbal aggression, to more passive forms like social isolation, gossiping or the spreading of rumours (Einarsen et al., 2020).

Workplace bullying is prevalent internationally. In a global workplace bullying prevalence rate meta-analysis, Nielsen et al. (2010) found that 15% of employees have experienced workplace bullying. Research indicates that workplace bullying is also a significant concern in the South African context. Cunniff and Mostert (2012), in a study including various sectors and regions, found that 31.1% of the sample had experienced bullying. Similarly, Kalamdien (2013) reported that between 30% and 50% of respondents from both public and private sectors in the Western Cape had been subjected to bullying. In the higher education sector, Conco et al. (2021) also found bullying to be highly prevalent, with 58% of participants reporting experiences of workplace bullying. More recently, Reddy and Naidu (2024) found that 38.7% of public-sector anaesthetists in eThekwini had experienced bullying in the workplace.

Bullying has been linked to several adverse consequences. Targets often experience issues such as depression (Ko et al., 2020) as well as anxiety and fatigue (Reknes et al., 2014). Research by Lahelma et al. (2012) indicates that exposure to bullying can lead to these mental health challenges even 5 years – 7 years after the experience. Emphasising its long-term impact, bullying has also been linked to increased levels of post-traumatic stress disorder (Laschinger & Nosko, 2015). Regarding job-related outcomes, workplace bullying has been identified as contributing to decreased job satisfaction and reduced organisational commitment (Nielsen & Einarsen, 2012). Research also suggests that bullying increases the risk of sickness absence (Lesuffleur et al., 2014), which is likely because of its physical and psychological effects. Victims of bullying may also feel insecure about their job stability and responsibilities (Glambek et al., 2014). This insecurity can result in career disruption and unemployment, as many employees may choose to leave their workplaces because of the unfavourable working conditions created by the bullying (Azizi-Qadikolaee et al., 2023).

Bullying is just as costly for the organisation as it is associated with decreased job satisfaction and organisational commitment, which point to generally low morale in the organisation (Nielsen & Einarsen, 2012; Paek et al., 2015). They also incur financial costs related to losses to productivity (Samnani & Singh, 2014), absenteeism, presenteeism (Naseem & Ahmed, 2020; Nielsen & Einarsen, 2012), additional recruitment and training processes (Hogh et al, 2011).

Intention to leave refers to employees’ thoughts of leaving his or her current organisation or employment willingly (Khan et al., 2013). Work-related factors that can lead to the intention to leave include the lack of job satisfaction (Gyensare et al., 2015), an unsupportive organisational culture, perceived lack of organisational justice (Halawi, 2014) and high levels of work-related stress (Salama et al., 2022). The relational dynamics at work, specifically, may also lead to increased intentions to leave. Some of these relational dynamics include factors such as the perception of high levels of organisational politics (Zhang & Lee, 2010), social exclusion by others in the workplace (Renn et al., 2013) and leadership behaviour leading to the belittling of employees and feelings of incompetence (Greenbaum et al., 2015).

Employees’ intention to leave the organisation has been associated with negative consequences, such as reduced engagement and productivity (Van Dyk, 2016). One of its most significant consequences is actual turnover, where employees ultimately leave the organisation (Sun & Wang, 2017). This is quite a notable consequence as it is associated with the costs related to the selection and training of new employees and the costs of improving or maintaining the morale and commitment of the remaining workforce (Khan et al., 2013).

Studies have consistently shown a positive relationship between workplace bullying and intentions to leave across different contexts and professions, where workplace bullying acts as a strong predictor of employees’ intention to leave. For example, Meriläinen and colleagues (2019) found bullying to predict turnover intention in the Estonian academic sector. Similarly, Coetzee and Van Dyk (2017) found that perceptions of bullying resulted in higher turnover intention in employees from various South African organisations. Interestingly, according to the study by Djurkovic et al. (2008), workplace bullying and intention to leave showed a positive relationship, even when the bullying employees experienced was considered less severe. This means that the experienced bullying behaviour did not necessarily involve overt behaviours such as physical abuse or threats of harm (Djurkovic et al., 2008). This noteworthy finding indicates the substantial impact of bullying on the intention to leave, even in situations of less severe types of bullying behaviour. Based on the above, the following hypothesis (H) was formulated for the study:

Previous studies have shown that different factors can mediate the relationship between workplace bullying and the intention to leave. For example, Ribeiro et al. (2023) found that workplace bullying increases turnover intentions among Portuguese employees, with burnout acting as a mediator. According to this study, targets of bullying experience high levels of burnout, which in turn leads to greater intentions to leave. Khan et al. (2021) demonstrated that workplace bullying positively influences employees’ intention to quit, with emotional exhaustion mediating this relationship. This indicates that employees who experience bullying often suffer from emotional exhaustion, which contributes to their decision to leave. Hsieh et al. (2019) identified self-efficacy as a partial mediator in the bullying intention to leave relationship among Taiwanese nurses. In this study, higher self-efficacy helped to mitigate the negative effects of bullying, reducing turnover intentions. Other identified mediators include work engagement (Coetzee & Van Dyk, 2017), psychosocial flourishing (Coetzee & Oosthuizen, 2017), job satisfaction (Mathisen et al., 2008) and affective commitment (McCormack et al., 2009).

Personal resources play a role in how individuals cope with difficult situations (Fredrickson, 2004). Personal resources may thus play a role in the degree to which employees express intention to leave as a result of experiencing bullying within the work context, which is a stressor. Kahn (1990) postulated that three psychological conditions represent resources that are influenced by both individual and organisational factors and consequently play a role in an individual’s decision to engage or disengage from work activities. This view was guided by the assumption that individuals continuously engage in behaviours of ‘bringing in and leaving out’ numerous parts of themselves during work to respond to the environment, express, and defend themselves (Kahn, 1990, p. 692). Kahn (1990) postulated that when people engage in any role behaviour, they ask themselves three questions: (1) How meaningful is this performance for me? (2) How safe is this performance? and (3) How available am I for this performance? These questions refer to the three psychological conditions of engagement and disengagement, namely: (1) psychological meaningfulness, (2) psychological safety and (3) psychological availability.

Psychological meaningfulness is defined as one’s subjective appraisal of his or her work as being valued, worthwhile, and instrumental to his or her personal growth, an appraisal that consequently drives an individual’s behaviour at work (Kahn, 1990). Experiences of psychological meaningfulness are also likely to be accompanied by favourable consequences such as increased satisfaction, motivation and commitment to the organisation (Wrzesniewski & Dutton, 2003). Employees may also report positive outcomes such as clarity of self-concept (Treadgold, 1999) and resilience under stressful conditions (Matuska & Christiansen, 2008).

Psychological safety, as the second psychological condition, refers to an individual’s state of ‘feeling able to show and employ oneself without fear of negative consequences to self-image, status, career’ (Kahn, 1990, p. 708). This safety, for example, refers to whether employees believe they will not be rejected for being themselves, expressing their thoughts or engaging in constructive disagreements with co-workers (Edmondson, 1999). They may also feel comfortable taking interpersonal risks and expressing opposing opinions (Ashauer & Macan, 2013). Psychological safety has also been found to be positively related to engagement (May et al., 2004).

Psychological availability, the third psychological condition, refers to an individual’s perception of having the physical, emotional and psychological resources necessary for engaging in work at a particular point in time (Kahn, 1990). It reflects a sense of being mentally focused, emotionally ready and physically energised to bring oneself into role performance (Jacobs, 2013; May et al., 2004; Rothbard & Patil, 2012). Prior research has shown that psychological availability is positively related to work engagement (Asiwe et al., 2017; May et al., 2004; Rothmann & Rothmann, 2010), suggesting that when employees feel depleted or distracted, they are less likely to invest themselves fully in their work roles.

From a COR perspective, psychological availability can be viewed as a personal energy resource that is vulnerable to depletion when employees face chronic interpersonal stressors (Hobfoll, 1989). Workplace bullying often involves ongoing exposure to hostility, humiliation and social exclusion (Einarsen et al., 2011, 2020), which can consume substantial emotional and cognitive resources as targets ruminate about incidents, anticipate future confrontations and engage in self-protection. Over time, such vigilance and emotional strain may erode employees’ perceived capacity and energy to engage in their work, thereby diminishing psychological availability. Within the Transactional Model of Stress and Coping (Lazarus & Folkman, 1984), repeated exposure to bullying is likely to be appraised as a threat to one’s self-worth and interpersonal security; these appraisals demand continued coping efforts that draw on physical, emotional and cognitive resources, leaving fewer resources available for constructive role engagement (Dewe, 1997). In line with this reasoning, targets of bullying may experience not only reduced relational safety and meaning, but also a felt sense of being ‘too drained’ to invest themselves fully in their work, even when they remain in the organisation.

On this basis, it is expected that workplace bullying will be negatively associated with psychological availability, as the ongoing demands of coping with hostile interpersonal encounters deplete employees’ perceived physical, emotional and cognitive capacity to engage in their roles.

Research evidence has suggested that workplace bullying may result in the depletion of the psychological conditions of meaningfulness, safety and availability (Fountain, 2016). Thus, the following hypothesis is proposed:

The psychological resources of meaningfulness, safety and availability enable an individual to engage or disengage in certain activities at work (Kahn, 1990). Employees’ perceived access to mental, emotional and physical resources to engage at work may influence their ability to regulate emotions and engage in active problem-solving during stressful situations (Bondarchuk et al., 2024). A psychologically safe environment enables employees to draw on social support at work to buffer against the harmful effects of challenging experiences such as bullying (Farley et al., 2023; Nielsen et al., 2020). Similarly, when individuals perceive their work as meaningful and aligned with personal values and goals, they are more likely to employ effective coping strategies in handling stressors (Hager, 2018). It is thus worthwhile to explore the relationship between these psychological conditions and their potential mediation effect in the bullying intention to leave link.

Following from the above, the following hypotheses were formulated:

The Transactional Model of Stress and Coping (Lazarus & Folkman, 1984) holds that stress is relational or dynamic in nature, arising from the interplay between an individual and the individual’s environment. Stress arises from the judgement that environmental demands have the likelihood of exceeding an individual’s physical or psychological resources and therefore threaten the individual’s well-being (Dewe, 1997). According to this theory, the degree to which a situation is perceived as stressful and the resultant action taken, is a function of the presence of resources that individuals use to cope and respond to situations (Lazarus & Folkman, 1984). In the context of this study, the theory suggests that the extent to which employees experience bullying – and their subsequent responses, such as expressing intentions to leave – is shaped by the availability of personal resources (psychological conditions).

The COR (Hobfoll, 1989) also provides a theoretical basis for the consideration of mediating factors in the association between bullying and the intention to leave. The COR theory suggests that people have a desire to ‘retain, protect, and build resources’ (Hobfoll, 1989, p. 516). Having surpluses of these resources is likely to lead to positive well-being. The potential or actual loss of these resources, on the other hand, is viewed as a threat and leads to behaviour aimed at minimising the net loss of resources (Hobfoll, 1989). The theory also suggests that there are stressful circumstances in the environment that threaten or deplete people’s resources. When individuals face these stressful circumstances, their behaviour is aimed at minimising the net loss of resources (Hobfoll, 1989). Applied to this study, the experience of bullying represents a stressful situation which could potentially threaten individuals’ resources. This could potentially lead to negative outcomes, such as the expression of intentions to leave, in a bid to minimise resource loss.

If evidence suggests that psychological conditions of engagement and disengagement weaken the impact of workplace bullying on the intention to leave, it becomes clear that organisations can intervene in this relationship by cultivating these psychological conditions. This can be achieved by creating and sustaining a work environment that supports the development and preservation of these resources. These measures will ultimately help organisations retain more employees, as proactive efforts to address intention to leave have been suggested to encourage employees to stay (Dysvik & Kuvaas, 2009). However, this does not imply that workplace bullying should be overlooked; its harmful effects also require the implementation of preventive and supportive structures within organisations. Interventions to improve psychological conditions will provide employees who experience bullying with additional resources to manage such experiences.

Figure 1 graphically represents the hypothesised model of the study based on the available literature.

This study used a quantitative, cross-sectional research approach with a survey research design. This approach allows for the exploration of patterns of association between two or more variables through the analysis of numerical data (Bryman, 2012).

Participants were recruited from different South African organisations through direct contact (via in-person visits) and through placing advertisements on social media platforms. The invitation included details about the study and what participation entailed. Participants could complete the survey either on paper or online via a link shared through email and social media. Participants were treated with respect and care, and measures were taken to ensure no harm occurred. This included requesting that all research participants have free access to EAPs within their organisations. As some of the items of the measurement instruments of the study could possibly elicit some discomfort or emotional feelings, having access to an EAP would thus enable participants to receive assistance following their participation in the study, if needed. The ethical principles of written informed consent, voluntary participation, the right to withdraw and confidentiality were also strictly followed (Mtshali, 2018). The researcher also ensured that participant data were accurately represented.

Data analysis was conducted on the SPSS version 30 programme (IBM, 2024). A descriptive statistics analysis was conducted to summarise the study variables and to assess their central tendency, variability and distributional properties. The reliability of the scales used in the present study was analysed by the examination of Cronbach’s alpha coefficients. Pearson’s correlation analysis was utilised to determine the existence, strength and direction of the relationships between the constructs of the study (Mtshali, 2018), in relation to H1, H2, H3, H4, H5, H6 and H7. Mediation analysis was conducted using Model 4 of the PROCESS Macro for SPSS (Hayes, 2018), for the purpose of examining the mediating effect of the psychological conditions on the relationship between workplace bullying (as the independent variable) and intention to leave (as the dependent variable), as per H8. Mediation analysis followed the four-step regression process by Baron and Kenny (1986) which entailed looking at the regressions of: (1) workplace bullying on intention to leave, (2) workplace bullying on each of the psychological conditions, (3) each of the psychological conditions on intention to leave and (4) workplace bullying and the psychological conditions on intention to leave. For all of the above, the level of statistical significance was set at p ≤ 0.05. A 95% percentile bootstrap confidence interval (CI) was further utilised to draw inferences about the indirect effect of workplace bullying on intention to leave (Hayes, 2018).

An application for full ethical approval was made to the College of Business Economics Research Ethics Committee at the University of Johannesburg, and ethics consent was received on 25 June 2018. All procedures performed in studies involving human participants were in accordance with the ethical standards of the institutional research committee and with the 1964 Helsinki Declaration and its later amendments or comparable ethical standards.

Descriptive statistics were computed to describe the central tendency, variability and distributional properties of the study’s variables. These results are presented in Table 2. As indicated, respondents reported relatively low mean levels of workplace bullying and moderate to high levels of psychological meaningfulness, psychological safety and psychological availability. Intention to leave was reported at a moderate level. Skewness and kurtosis values fell within acceptable ranges.

Internal consistency reliability was assessed using Cronbach’s alpha (see Table 2). Of all the scales utilised in the study, the psychological safety subscale had a Cronbach’s alpha below the recommended 0.70 (Nunnally, 1978). As this scale comprises only three items, the mean inter-item correlation coefficient was also investigated, as recommended by Pallant (2011). The mean inter-item correlation for psychological safety was 0.31, which falls within the recommended range of 0.20 to 0.40 (Briggs & Cheek, 1986), supporting the adequacy of the scale’s internal consistency.

Pearson’s product-moment correlation coefficients are presented in Table 3. The results indicate a statistically significant positive correlation between workplace bullying and intention to leave (r = 0.55, p < 0.001), supporting H1. Workplace bullying was also significantly and negatively correlated with psychological meaningfulness (r = −0.32, p < 0.001) and psychological safety (r = −0.46, p < 0.001), supporting H2 and H3, respectively. In contrast, workplace bullying was not significantly correlated with psychological availability (r = −0.12, p = 0.082); therefore, H4 was not supported.

In terms of the associations between each of the psychological conditions and intention to leave, psychological meaningfulness (r = −0.41, p < 0.001) and psychological safety (r = −0.52, p < 0.001) were both significantly and negatively related to intention to leave, supporting H5 and H6. Psychological availability was also significantly, though weakly, negatively correlated with intention to leave (r = −0.16, p < 0.05), supporting H7.

A parallel mediation analysis was conducted using Model 4 of the PROCESS macro for SPSS to examine whether psychological meaningfulness, psychological safety and psychological availability mediated the relationship between workplace bullying and intention to leave. The results of the mediation analysis are provided in Table 4 and Table 5.

As shown in Table 4 and Table 5, workplace bullying was a significant and negative predictor of psychological meaningfulness (a₁-path; b = −0.12, t = −4.40, p < 0.001) and psychological safety (a₂-path; b = −0.09, t = −7.24, p < 0.001). In contrast, workplace bullying did not significantly predict psychological availability (a₃-path; b = −0.02, t = −1.22, p = 0.223). Workplace bullying was a significant predictor of intention to leave (c’-path; b = 0.11, t = 5.14, p < 0.001).

The consideration of the effect of each of the psychological conditions on intention to leave indicated that psychological meaningfulness (b₁-path; b = −0.23, t = –4.31, p < 0.001) and psychological safety (b₂-path; b = −0.58, t = −4.72, p < 0.001) were both statistically significant predictors of intention. Psychological availability did not significantly predict intention to leave (b₃-path; b = 0.05, t = 0.60, p = 0.549).

After accounting for the mediators, workplace bullying remained a significant predictor of intention to leave, indicating partial mediation. The overall model explained 45% of the variance in intention to leave, R² = 0.45, F(4,178) = 37.05, p < 0.001.

The indirect effect was tested using a bootstrap estimation approach with 1000 samples. These results are provided in Table 6. The total indirect effect of workplace bullying on intention to leave via psychological meaningfulness and psychological safety was statistically significant (b = 0.08, standard error [s.e.] = 0.01) with a 95% bias-corrected CI that did not include zero (0.05, 0.11). Psychological availability did not contribute significantly to the indirect effect.

All proposed hypotheses were supported, with the exception of H4. With regard to H8, psychological availability did not emerge as a significant mediator of the relationship between workplace bullying and intention to leave, nor did it significantly predict intention to leave when controlling for the other psychological conditions.

Firstly, the strong direct effect of workplace bullying on intention to leave underscores the need for organisations to prioritise the prevention and active management of bullying behaviours. Leadership accountability is critical, as unaddressed bullying signals organisational tolerance and accelerates withdrawal intentions.

Secondly, the partial mediation effects highlight the importance of restoring psychological meaningfulness and psychological safety. Interventions should therefore focus on cultivating a work environment that will enhance employees’ sense that their work remains valuable and that interpersonal risk-taking is safe. This may include supervisor training in respectful leadership, inclusive communication practices and consistent feedback, as well as team-level interventions that clarify role purpose and reinforce shared values. When psychological meaningfulness and safety are cultivated and protected within the organisation, employees facing workplace bullying may be more inclined to seek constructive coping strategies rather than opting to leave the organisation.

These findings suggest that effective retention strategies in bullying-affected contexts require a dual focus: eliminating bullying at its source while simultaneously restoring key psychological conditions that support sustained employee investment.

Several limitations of this study should be acknowledged. Firstly, the use of a cross-sectional research design precludes causal inferences regarding the relationships between workplace bullying, psychological conditions and intention to leave. Future research should therefore employ longitudinal designs to examine how experiences of workplace bullying influence psychological conditions and withdrawal cognitions over time.

Secondly, the reliance on self-report measures introduces the potential for socially desirable responding. The use of cross-sectional, self-report data also raise the possibility of common method variance, which may have inflated the strength of the observed associations. Participants may have thus underreported or misrepresented their workplace experiences. Future studies could mitigate these limitations by incorporating multi-source data, such as supervisor reports, peer ratings or organisational records, where appropriate.

Thirdly, although established instruments were used, only the anchor points of the 5-point Likert-type scales were explicitly defined. This may have resulted in varied interpretations of the intermediate response options. Future research is encouraged to provide fully labelled scale points to enhance measurement clarity and response consistency.

Fourthly, the study focused on overall exposure to workplace bullying, rather than examining specific bullying behaviours or sub-dimensions of the NAQ-R. This approach limits insight into whether particular forms of bullying differentially affect psychological conditions and intention to leave. Future studies should therefore examine the distinct effects of bullying subtypes on psychological meaningfulness, psychological safety and psychological availability.

Fifthly, the study was context-specific, as it was conducted among South African employees with access to an EAP. This sampling criterion may have excluded employees from smaller or less resourced organisations, thereby limiting the generalisability of the findings. Future research should adopt broader sampling strategies that include employees from organisations without formal support structures to capture a more diverse range of workplace experiences.

Finally, future research could extend the present model by examining the mediating or moderating roles of additional personal, relational and organisational resources, such as perceived organisational support, resilience or leadership behaviours.