Workplace incivility and bystanders’ helping intentions

Abstract

Purpose

Using observational and experimental designs, the purpose of this study was to explore if the power relation between the offender and the victim of incivility and the level of perceived severity of the incivility were associated with bystanders’ intentions to help when witnessing workplace incivility.

Design/methodology/approach

In Study 1, 160 participants completed a questionnaire where they described a recent uncivil incident they had witnessed, and completed measures of perceived severity and measures of their behavioural response as bystanders. In Study 2, 183 participants were randomised to read one of two vignettes (a manager being uncivil towards a subordinate or vice versa), and completed measures of perceived severity and of their motivation to intervene. The authors investigated whether the power relation between perpetrator and victim, and the perceived severity of the uncivil exchange, were associated with prosocial bystander behaviours in Study 1 and with motivation to defend the victim of incivility in Study 2.

Findings

Higher perpetrator power was significantly associated with the incident being perceived as more severe, and higher perpetrator power was directly related to greater tendency to confront, and lower tendency to avoid, the perpetrator. Perpetrator power was indirectly associated with social support according to the perceived severity. A supervisor acting in an uncivil manner was rated as more severe than a subordinate acting in such a way. Perceived severity mediated the relationship between perpetrator power and the witness’s introjected, identified and intrinsic motivation to intervene.

Originality/value

This study extends previous work by investigating how the perpetrator’s power influences both the bystander’s prosocial behaviour and their motivation to defend the victim. Furthermore, previous research has not considered how perceptions of severity might mediate the relationship between power, behaviour and motivation.

Citation

Jungert, T. and Holm, K. (2022), "Workplace incivility and bystanders’ helping intentions", International Journal of Conflict Management, Vol. 33 No. 2, pp. 273-290. https://doi.org/10.1108/IJCMA-08-2021-0131

Workplace incivility, which is defined as “low-intensity deviant behaviour with ambiguous intent to harm the target” (Andersson and Pearson, 1999, p. 457), has been described as a pervasive problem in organisational life (Cortina et al., 2017). In addition to the many negative outcomes for the victims of incivility (Schilpzand et al., 2016 for a review), several studies have demonstrated the adverse effects of witnessing incivility being exhibited towards others in the workplace. These negative effects for bystanders include negative emotions (Miner and Eischeid, 2012), lower job performance and reduced citizenship behaviours (Porath and Erez, 2009), higher levels of stress and lower levels of job satisfaction (Holm et al., 2019) and negative perceptions of organisational justice and well-being (Holm et al., 2021). In addition, workplace incivility has been shown to have detrimental behavioural outcomes for both victims and witnesses in the form of an increased tendency to act in an uncivil manner towards others (Meier and Gross, 2015; Rosen et al., 2016; Torkelson et al., 2016). This suggests that uncivil behaviours can spread in the workplace if they are not addressed (Foulk et al., 2016). Despite the detrimental consequences of incivility in the workplace, few studies have focused on exploring when and why bystanders intervene or exhibit prosocial behaviour to reduce workplace incivility (Hershcovis et al., 2017a, for an important exception), and more research has been requested on the effects of workplace incivility on bystanders (Holm et al., 2021; Schilpzand et al., 2016).

In the present study, we aim to address these issues by focusing on the bystanders’ helping intentions when witnessing incivility. Specifically, we investigate how situational factors such as the perpetrators’ position of power and the perceived severity of the incident relate to bystanders’ intentions to help. We explore helping intentions in the form of both bystanders’ behavioural response when witnessing incivility in the workplace and the bystanders’ motivation for intervening. By doing so, we are able to gain further information about predictors of prosocial behaviours that bystanders exhibit in response to witnessed incivility, as well as information about how motivational processes are influenced by situational factors such as the perpetrator’s power position and the perceived severity of the uncivil situation. Exploring both behavioural responses and motivational processes is an important step in understanding both the behavioural and cognitive dimensions of bystander intervention in response to witnessed workplace mistreatment.

Bystander behaviour and self-determined motivation to intervene

In the present work, we operationalise bystanders’ helping intentions in two ways. The first is through the behavioural roles that bystanders may or may not be prone to take when witnessing incivility in the workplace (Hershcovis et al., 2017a; Rosette et al., 2013). Second, we also consider their motivation to help, as described by the facets of motivation derived from self-determination theory (SDT; Ryan and Deci, 2017). Together, these two components of behaviour and motivation give a more complete picture of a bystander’s intention to help when witnessing incivility in the workplace.

In relation to bystander behaviour, there are several possible roles that a bystander may take when witnessing an uncivil interaction. In a recent study, Hershcovis et al. (2017a) explored how the observer’s position of power influences their intervention behaviour in terms of confronting the perpetrator, offering social support to the victim, or avoiding the perpetrator. They found that observers with high levels of power tended to confront the perpetrator more than observers with low levels of power, while those with low power avoided the perpetrator and supported the victim to a greater extent than those with high levels of power. However, their study only approached one aspect of the power dimension, which was the observer’s own power relative to that of the perpetrator. Hershcovis et al. (2017a) suggested that the perpetrator’s power position in relation to the victim might also be important for bystander behaviour, but the influence of perpetrator-victim power relations on prosocial bystander behaviour has not yet been explored.

SDT (Deci and Ryan, 1985; Ryan and Deci, 2017) differentiates between various types of motivation and describes motivation as a continuum spanning from amotivation to intrinsic motivation (Deci and Ryan, 2000), and four motivational regulations can be found along this spectrum. These behaviours include external regulation, introjected regulation, identified regulation and integrated regulation, which progress from externally regulated behaviours to more autonomously regulated behaviours. External, or extrinsic, regulation is when a behaviour is motivated purely by external factors such as rewards or the avoidance of punishment. Introjected regulation is when behaviour is regulated primarily by maintenance of the ego or self-esteem rather than by internalised values associated with the behaviour. Identified regulation, which is more autonomous, refers to the regulation of behaviour because the action itself is valued by the person. In this case, the behaviour is seen as beneficial rather than just necessary to maintain one’s ego or self-esteem. Finally, integrated, or intrinsic, regulation involves processes of full assimilation with the self, such as striving to help others because the person perceives that helping is a part of their self. In previous studies on bystander behaviour in bullying situations among adolescents, more autonomous forms of regulation were found to be positively associated with defending the victim and negatively associated with passive bystander behaviour (Jungert et al., 2016, 2021). We therefore believe that these four types of regulation are also important facets to include in models of bystander incivility to understand not only if general motivation to intervene can be modified by situational characteristics but also in what way motivation is affected by such factors.

Power and severity

It has been suggested that high-power individuals have ample resources to act uncivilly without ramifications, whereas those with lower relative power in the organisation might engage in more covert types of incivility (Pearson and Porath, 2005). Previous research has found associations between the power and status of both the victim Porath and Pearson (2012) and the instigator Cortina and Magley (2009) and experiences of uncivil encounters in the workplace. For instance, incivility from a powerful perpetrator was appraised as more negative by targets (Cortina and Magley, 2009). When comparing incivility from supervisors to incivility from subordinates, it was found that supervisors engaged in such behaviours to a greater degree (Torkelson et al., 2016). Additionally, a previous study showed that supervisor incivility was related to negative outcomes via organisational factors such as increased demands and decreased control and support, whereas co-worker incivility had no association with negative outcomes such as detrimental effects on well-being, job satisfaction or turnover intentions (Holm et al., 2015). In models of witnessed co-worker and supervisor incivility, stronger associations with stress and lower job satisfaction were found when witnessing incivility from a supervisor rather than incivility from a co-worker (Holm et al., 2019). Taken together, this could indicate that incivility from a person of higher power could be perceived as more severe than incivility from someone with lower or equal power. In addition, Hershcovis et al. (2017b) demonstrated that perpetrator power moderated the relationship between incivility and embarrassment, and the relationship was stronger when the perpetrator was more powerful. This suggests that occurrences of incivility may be perceived as more severe if they are instigated by a powerful perpetrator. However, the relationship between the perpetrator’s power position and the perceived severity of the uncivil behaviour has not been explored more explicitly.

In situations of workplace bullying, perceived severity has been shown to be an important factor in determining helping behaviours, and situations perceived as more severe have been shown to be associated with greater attribution of responsibility to the perpetrator and increased helping intentions in bystanders (Desrumaux and De Chacus, 2007; Desrumaux et al., 2015). Similarly, Hellemans et al. (2017) found perceived severity to be a determinant of helping behaviour in situations of workplace bullying. In a study on bystander behaviour in response to incidences of incivility, it was found that the observer’s power position relative to the perpetrator was a significant determinant of whether they intervened, and more powerful bystanders were more likely to directly confront the perpetrator (Hershcovis et al., 2017a). However, how the power relationship between victim and perpetrator might influence helping intentions remains unexplored. Drawing from the literature on workplace bullying, it is possible that perceived severity is an important mechanism in why bystanders intervene after witnessing uncivil interactions at work, particularly when perpetrated by a more powerful individual.

Theoretical foundations and hypotheses development

Despite the theoretical assumptions described above, research on predictors of bystanders’ intentions to help when witnessing workplace incivility remains limited. Little is still known about how incivility impacts bystanders, and there has been a call for research models to be extended to include the bystander perspective to elucidate which impact incivility may have beyond those directly targeted (Holm, 2021; Holm et al., 2021). Although research recently has begun to explore attitudinal and behavioural outcomes of workplace incivility for witnesses, these studies have primarily focused on the negative impact incivility can have on bystanders, such as uncivil behaviour spreading to witnesses (Holm et al., 2019, 2021; Holm, 2021), or on more specific instances of mistreatment, such as how bystanders react against perceived gender discrimination (Sinclair, 2021). Holm (2021) recommended that future studies should explore if situational characteristics that are central to the construct of workplace incivility, such as perceived intensity of the incident, are associated with different appraisals and coping responses for either targets or witnesses. Consequently, more research is needed to explore whether situational factors are instrumental in influencing bystanders’ prosocial intentions in response to incivility in the workplace. To gain more understanding on bystanders’ intentions, we assert that new insights can be derived through both behavioural roles that bystanders may take Hershcovis et al. (2017a), Rosette et al. (2013) and their motivation to help (Ryan and Deci, 2017). We thereby complement the increasing body of research on bystanders’ intentions to help victims of workplace incivility, and address the question whether situational factors influence bystander reactions. By including both a field study and an experimental vignette study, we also answer the call for the use of multiple methods (Jensen and Raver, 2021; Sinclair, 2021). The present study takes advantage of the combined strengths of methods involving participant recall Mitchell et al. (2015) and experimental design Porath and Erez (2009) to enhance realism and assess actual behaviour.

Specifically, we performed two studies examining the impact of situational factors on a bystander’s intentions to help when witnessing incivility in the workplace. The purpose was to investigate the impact of the power relation between the offender and the victim of incivility as well as the level of perceived severity of the incivility to determine if these are associated with the bystanders’ intentions to help. Study 1 looked at prosocial or passive roles when facing uncivil interactions, while Study 2 looked at their motivation to intervene to defend the victim. Specifically, the aim of the two studies was to explore whether perpetrator power was positively associated with higher levels of perceived severity and if perceived severity in turn mediated the relationship between perpetrator power and the outcomes. In Study 1, we gathered participants’ reports of their actual behaviours in response to witnessing an uncivil situation by asking them to describe a critical incident of incivility that they had witnessed in their workplace. In Study 2, we presented participants with a detailed, context-laden vignette of an uncivil situation capable of eliciting emotional responses and appraisals similar to those that would be experienced in real-life bystander situations. Based on the background provided above, we hypothesise that perpetrator power and perceived severity is related to bystanders’ helping intentions. In both studies, we hypothesised that higher perpetrator power was associated with incivility being perceived as more severe (H1a and H2a). In addition, we hypothesised that perpetrator power was indirectly related to helping intentions via perceived severity across both studies (HH1 b and 2 b). See section “Study 1” and “Study 2” for detailed descriptions of each respective study hypotheses.

Study 1

This study was designed to examine how the relative power of a perpetrator of workplace incivility is related to the observer’s perception of the severity of the situation and to bystander behaviour in terms of confronting the perpetrator, avoiding the situation, or socially supporting the victim. The study was designed as a retrospective survey study and was built on a critical incident technique similar to that used in the study by Hershcovis et al. (2017a). The rationale behind conducting a field study to gather retrospective reports of bystander behaviour was that such a study would provide information about actual behaviours exhibited in response to witnessing incivility in the workplace, rather than intended behaviours, to strengthen the ecological validity of the conclusions. We hypothesised that higher perpetrator power is directly related to higher levels of the perceived severity of the situation (H1a). Second, we hypothesised that perpetrator power is indirectly related to increased prosocial behaviour, such as higher levels of confronting the perpetrator and offering social support to the victim, and lower levels of avoidance, based on the perceived severity of the situation (H1 b).

Method

Participants and procedure

This field study consisted of a sample of 160 employees working in 13 different scopes of practice in Northern Europe, including construction, education, administration, transport, retail/catering, real estate, banking/finance, health care and media. The majority of the participants were employed in retail and catering (25%), education (18%), health care (16%) and construction (8%). The average age of the participants was 33.21 (SD = 12.83) years, the participants had been working in their current field for an average of 5.66 years (SD = 6.26) years, and 48% were female. The data were collected through convenience sampling from several organisations as part of the course work in an undergraduate psychology course. We used a critical incident technique in which we first defined the term “workplace incivility” and gave a couple of examples of incivility. Thereafter, we asked participants to recall an incident of witnessed incivility between two employees (either a manager behaving in an uncivil manner towards a subordinate in their organisation or a subordinate behaving in an uncivil manner towards a manager or a co-worker). Participants were then requested to briefly describe the incident and to answer questions about the power of the perpetrator, their perception of the severity of the incident and the extent to which they confronted the perpetrator, supported the victim or avoided the situation. Participants were made aware that their participation was voluntary and that all responses would be anonymous, and they all provided informed consent prior to participating.

Measures

Participants completed three measures in Swedish.

Perpetrator power.

We asked about the perpetrator’s power using three items adapted from Hershcovis et al. (2017a), which were back-translated to Swedish. Participants were asked to indicate how much power the perpetrator had in comparison to the victim in the organisation on a scale from 1 (none) to 5 (a great deal). An example item is “How much influence did the perpetrator have over the victim?” Cronbach’s alpha was 0.75.

Perception of the severity of the situation.

We asked a single question about the severity of the situation:

Q1.

In your opinion, how severe was the situation?

Participants were asked to indicate their perception of the severity on a scale from 1 (not severe at all) to 5 (highly severe).

Prosocial behaviours.

This is a 12-item scale measuring the participant’s prosocial behaviour when they witnessed the incident of incivility in their workplace. The scale was a development by Hershcovis et al. (2017a) of three subscales constructed by Fitzgerald (1990) to measure how much observers of incivility in the workplace engage themselves in confronting the perpetrator, avoiding the situation, or socially supporting the victim, which were back-translated to Swedish. With regard to confrontation, participants were asked to report to what extent they confronted the perpetrator using four items. An example item is “I confronted the perpetrator”. Cronbach’s alpha = 0.93. With regards to avoidance, the participants were asked to report the extent to which they avoided the perpetrator. An example item is “I just ignored it”. Cronbach’s alpha = 0.73. Finally, participants were asked to report how much social support they gave the victim. An example item is “I showed my support for the victim”. Cronbach’s alpha = 0.93. For each statement, participants responded along a five-point scale of agreement (1 = completely disagree, 5 = completely agree).

Data analysis

Data were analysed using IBM SPSS v24 and R v3.4.3. Because previous studies have shown prosocial and passive behaviours to be correlated (Hershcovis et al., 2017a), we sought to model the dependent variables simultaneously in order to achieve more unbiased estimates of the relationships. We therefore estimated a path model using the lavaan package in R (Rosseel, 2012), to test both direct and indirect effects. In the model, perpetrator power was the predictor of all constructs, i.e. the perceived severity of the uncivil situation and the reports of confronting the perpetrator, avoiding the perpetrator or offering social support to the victim of incivility. Paths from severity to confrontation, avoidance and social support were also added to the model to test the indirect effects of perpetrator power on the three behaviours according to the perceived severity. The Maximum Likelihood estimator was used in the model estimation because the factors did not demonstrate any large violations from a normal distribution. When testing the hypothesis concerning mediation, bootstrapped (5,000 draws) standard errors were generated in lavaan to create bias-corrected confidence intervals for the indirect effects in the model (Preacher and Hayes, 2008).

Results

Table 1 shows the means, standard deviations and Pearson’s correlations of the observed study variables.

To test H1a and 1b – that higher perpetrator power is positively associated with perceived severity and that severity mediates the positive relationship between perpetrator power and prosocial behaviours such as confrontation and offering social support, as well as the negative relationship with passive behaviours such as avoidance – a path model was estimated. The full model is shown in Figure 1. The model revealed significant paths from perpetrator power to severity (β = 0.48, p < 0.001), confrontation (β = 0.56, p < 0.001) and avoidance (β = −0.28, p = 0.001). H1a was therefore supported. The direct path from power to social support was, however, not significant (β = 13, p = 0.087). Although the path from severity to social support was significant (β = 0.46, p < 0.001), perceived severity did not significantly predict confrontation or avoidance. The model explained 23% of the variance in severity, 34% of the variance in confrontation, 7% of the variance in avoidance and 28% of the variance in offering social support.

Next, to test the mediation hypothesis, the indirect effects from perpetrator power to the prosocial or avoidant roles, via perceived severity, were calculated. The effect from power to social support via severity was the only significant indirect effect in the model (β = 0.22, p < 0.001). Bias-corrected confidence intervals (based on 5,000 bootstrap draws) were calculated to test the robustness of the indirect effect, and this yielded confidence intervals of [−0.04, 0.09] for confrontation, [−0.07, 0.08] for avoidance, and [0.12, 0.31] for social support. Because the only confidence interval not ranging over zero was that for social support, this suggests that perpetrator power was only indirectly linked to social support via severity, whereas there was a direct effect between perpetrator power and the two other dependent variables of confrontation and avoidance. H1 b was therefore only partly supported. Overall, the results showed that higher perpetrator power was related to higher levels of confrontation, lower levels of avoidance and higher levels of perceived severity. Furthermore, incidents where the perpetrator had higher power were perceived as more severe, which in turn was associated with an increased tendency to offer social support to the victim.

Brief discussion

Study 1 showed that higher perpetrator power was directly and positively associated with confronting the perpetrator and negatively associated with avoidance. These effects did not appear to be transmitted by perceived severity, which is in contrast to findings on bystander intervention in cases of workplace bullying (Desrumaux et al., 2015; Hellemans et al., 2017). Conversely, there was no main effect from perpetrator power to offering social support, but this effect was significantly mediated via perceived severity. However, the helping behaviours studied by Hellemans et al. (2017) primarily consisted of offering emotional or public support to the victim, and the likelihood of engaging in such behaviours was associated with the perceived severity of the encounter. Interestingly, perceived severity only appeared to be a mechanism in the relationship between perpetrator power and social support, and it did not explain why higher perpetrator power is associated with other prosocial behaviours such as confronting the perpetrator or exhibiting less avoidance. These findings would appear to be consistent with Hellemans et al. (2017) by suggesting that perceived severity is an important mechanism behind why bystanders offer social support to the victim. It is possible that a higher level of perceived severity makes the suffering of the victim more salient and elicits a stronger focus on the victim who is being exposed to incivility. This might prompt the bystander to engage in victim-focused prosocial behaviours, such as offering social support, rather than confronting the perpetrator. Although this study provides novel information about actual bystander behaviour in response to witnessed incivility, the motivational mechanisms involved in helping, which explain why the bystander becomes motivated to intervene, remain to be explored.

Study 2

This study was designed to examine the impact of the power relation between the offender and the victim of incivility, the level of perceived severity of the incivility by the bystander, and if the perceived severity is associated with motivation to intervene to defend the victim of incivility. In contrast to Study 1, this study was designed as an experimental study, and participants were randomised to read a vignette of a situation where either a manager or a subordinate behaved uncivilly towards a co-worker. Because the dependent variables of interest in this study were motivation factors, we believed that retrospective recall of previously witnessed incidents might be too influenced by recall bias. Recalling which behaviour was exhibited likely requires less effort than attempts to recall which particular kind of motivation that the individual was feeling at that time, which might be quite difficult to remember accurately. Thus, we asked participants about their motivation to intervene in response to the hypothetical situation described in a vignette. We hypothesised that higher perpetrator power is directly related to the situation being perceived as more severe (H2a). Second, we hypothesised that higher perpetrator power is indirectly associated with increased motivation to defend the victim via the perceived severity of the situation (H2 b).

Method

Participants

Participants were 183 Swedish workers with a mean age of 32.24 (SD = 12.80) years, and 62% were females. Their average tenure was 6.30 years (SD = 5.60). The data were collected through convenience sampling via Facebook, and participants were randomly assigned to the experimental conditions as part of the course work in an undergraduate psychology course. Participants were made aware that their participation was voluntary and that their responses would be anonymous. All participants provided informed consent prior to participation.

Stimuli

The two vignettes had identical descriptions of a workplace incivility situation except for who was the perpetrator and who was the victim. The participants were asked to imagine that they were in the staff room of their organisation having a coffee break and had witnessed everything that happened in the vignette. In the first condition, the manager behaved uncivilly towards a subordinate co-worker, and in the second condition, the subordinate behaved uncivilly towards the manager. The vignettes were about 250 words long (Appendix).

Measures

Participants completed four measures and a comprehension check question in Swedish.

The motivation to defend scale (MDS).

This is a 15-item scale measuring motivation to defend a victim of incivility. The scale is a Swedish development of the MDS and is designed to measure motivation to intervene in a school bullying situation (Jungert et al., 2016). Participants were asked to report “why they would engage in helping the victim of incivility” and to choose from responses on the MDS representing four types of motivation: “Because I like to help other people” (intrinsic), “Because I think it is important to help people who are treated badly” (identified), “Because I would feel like a bad person if I did not try to help” (introjected), and “To be praised by my manager or co-workers” (extrinsic). For each statement, participants responded along a five-point scale of agreement (1 = completely disagree, 5 = completely agree). Cronbach’s alpha was 0.67 for intrinsic motivation, 0.72 for identified motivation, 0.61 for introjected motivation and 0.61 for extrinsic motivation.

Perception of the severity of the situation.

We asked a single question about the severity of the situation in the vignette:

Q2.

In your opinion, how severe is the situation between person A and the manager in this story?

Participants were asked to indicate their perception of the severity on a scale from 1 (not severe at all) to 4 (highly severe).

Comprehension check.

We asked a single comprehension question:

Q3.

Was the manager/person A subjected to disrespectful behaviour by the co-worker/manager?

with the answer options “yes” and “no”.

Procedure

Data were collected with an electronic survey, and participants were randomised to either condition 1 or condition 2. Two participants failed to answer the comprehension check question correctly and were removed from further analyses.

Data analysis

To test the study hypotheses, a path model was estimated in the same way as in Study 1. The chief advantage of employing path analysis rather than other multivariate techniques such as MANOVA for exploring group differences for several dependent variables is that path analysis is more flexible and allows for the possibility to test the indirect effects from categorical and ordinal predictors on multiple correlated dependent variables. Additionally, structural equation modelling-techniques, such as path analysis, have been shown to be superior to standard regression procedures in testing mediation models (Iacobucci, 2008). Because the predictors were ordinal and categorical, weighted least squares with means and variance-adjusted χ²-values were used for model estimation. This is the preferred choice when analysing data organised into ordered categories (Schumacker and Lomax, 2010). The power vignette was entered as a dummy-coded manifest predictor variable in the model (co-worker uncivil to supervisor coded as 0, supervisor uncivil to co-worker coded as 1) for predicting all other variables in the model. The severity item was also entered as a single-item manifest variable in the model that was predicted by the dummy-variable and also predicting the four dependent variables of extrinsic motivation, introjected motivation, identified motivation and intrinsic motivation. When testing the mediational hypothesis, standard errors were bootstrapped (5,000 draws) to generate bias-corrected confidence intervals of the indirect effects.

Results

Table 2 shows the means, standard deviations and Pearson’s correlations of the observed study variables.

To test H2a and H2b – that perpetrator power is related to higher levels of perceived severity and that severity mediates the relationship between perpetrator power and motivation to defend – a path model was estimated. The model revealed four significant paths, including the path from perpetrator power to severity (β = 0.39, p < 0.001) and the paths from severity to intrinsic (β = 0.21, p = 0.012), identified (β = 0.33, p < 0.001) and introjected motivation (β = 0.29, p < 0.001). The strongly significant path from power to severity demonstrated that the condition where a co-worker was targeted by a supervisor was rated as more severe than the other condition, in support of H2a. The model, which is depicted in Figure 2, explained 16% of the variance in severity, 6% of the variance in intrinsic motivation, 12% of the variance in identified motivation, 11% of the variance in introjected motivation and 2% of the variance in extrinsic motivation.

To test H2 b concerning mediation, the indirect effects from power via severity to the motivation variables were calculated. The model revealed significant indirect effects of power on intrinsic (β = 0.08, p = 0.026), identified (β = 0.13, p < 0.001) and introjected motivation (β = 0.12, p = 0.001) via perceived severity. Bias-corrected confidence intervals (based on 5,000 bootstrap draws) for the indirect effects from power via severity showed confidence intervals of [0.02, 0.29] for intrinsic, [0.07, 0.26] for identified, [0.07, 0.28] for introjected and [−0.14, 0.03] for extrinsic motivation. The confidence intervals for the indirect effects on intrinsic motivation, identified motivation and introjected motivation did not contain zero, indicating that the relationships between power and three of the motivation variables were mediated by perceived severity. There was no significant indirect effect of power on extrinsic motivation via severity (β = −0.05, p = 0.179). H2 b was therefore only partly supported. The strongest indirect relationships were from power to introjected and identified motivation via the perceived severity. There were no significant main effects from the vignette to the motivation variables. Overall, the results showed that an uncivil interaction with a perpetrator in a relatively higher power position was perceived as more severe, and greater severity was related to higher levels of intrinsic, identified and introjected motivation to intervene.

Brief discussion

In Study 2, H2a, that incivility from a more powerful perpetrator is associated with higher ratings of perceived severity of the uncivil situation, was again supported, in this case in a complementary experimental design. Additionally, H2 b was partly supported because the positive relationship between perpetrator power and intrinsic, identified and introjected motivation was mediated by the perceived severity of incivility in the workplace. As a result, when the perpetrator has high power, the transgression is seen as more severe, and the bystander is more motivated to defend the victim. These results provide support for previous findings that situations perceived as severe are associated with increased helping intentions in bystanders (Desrumaux and De Chacus, 2007; Fischer et al., 2011). However, Study 2 extends those previous findings by demonstrating that bystanders’ intrinsic, identified and introjected motivation to defend victims of incivility depends on the perceived severity, whereas bystanders’ extrinsic motivation to defend is not associated with power or perceived severity. Bystanders seem to be motivated to defend a victim either because it is in line with their values or because they desire to maintain self-esteem and avoid self-sanctions (Weinstein and Ryan, 2010).

General discussion

We investigated how situational factors, such as the perpetrator’s power position and how severe the incivility is perceived, are associated with bystanders’ intentions to help in the form of participants’ behaviour and motivation to intervene when observing incivility in the workplace. Specifically, we aimed to explore whether higher perpetrator power was associated with incivility being perceived as more severe and whether perceived severity mediated the relationship between perpetrator power and the bystanders’ helping intentions. Higher perpetrator power was consistently shown to be associated with higher levels of perceived severity in both studies. Perceived severity was also found to be an important mechanism for transmitting the effect of high perpetrator power to prosocial behaviour such as offering social support and to introjected, identified and intrinsic motivation to intervene.

Our findings have important theoretical and practical implications. From a theoretical perspective, previous research on bystanders has almost solely investigated the consequences of observing incivility towards others in work contexts Holm et al. (2015, 2019; 2021), Miner and Cortina (2016), Reich and Hershcovis (2015) and found that observations of workplace incivility showed declines in psychological well-being and job satisfaction, heightened levels of stress and negative affect and lowered perceived justice in the organisation. Furthermore, bystanders’ prosocial behaviour has only previously been investigated in a study by Hershcovis et al. (2017a), which focused on bystanders’ reactions to incivility and the bystanders’ power position. This study extends previous work by investigating how the perpetrator’s power influences both the bystander’s prosocial behaviour and their motivation to defend the victim. Furthermore, previous research has not considered how perceptions of severity might mediate the relationship between power, behaviour and motivation.

Power relationships play an essential role in organisations (Aquino and Bradfield, 2000; Pitesa and Thau, 2013), and an offender’s power might influence how severe the bystander perceives acts of incivility. Understanding these power dynamics is critical to understanding bystanders’ behaviours and motivations in response to acts of incivility in organisations. Therefore, our studies provide valuable contributions by contextualising theory and research on bystanders’ reported reactions in the dynamic relationships in which they occur.

Moreover, this study also extends previous work on incivility as we consider how power and perceived severity are associated with different types of motivation to intervene and different types of behaviour by a bystander observing incivility. Previous research has found that bullying situations perceived as severe are associated with increased helping intentions in bystanders (Desrumaux and De Chacus, 2007; Desrumaux et al., 2015), while Hershcovis et al. (2017a) found that bystanders who are more powerful are more likely to directly confront the perpetrator. Our results expand these findings by showing that there is a relationship between perpetrator power and motivation to defend the victim if incivility is perceived as severe. In other words, when the perpetrator has high power, incivility is perceived as more severe, and the bystander is more motivated to defend the victim. Similarly, we found that when the perpetrator has high power, the bystander is more prone to consider the incident to be more severe and is more prone to confront the perpetrator and less likely to avoid intervening. Additionally, when the act is perceived as more severe, the bystander is more likely to offer the victim social support. This corroborates and extends the findings of Hellemans et al. (2017) demonstrating that bystanders might engage in different forms of prosocial behaviours depending on the context of the situation. Specifically, if the incivility is perceived as severe, the bystander might engage in more victim-oriented prosocial behaviour such as offering social support to the victim. However, the perpetrator’s power relative to the victim, rather than the severity of the incivility, better explained prosocial behaviours such as confrontation and less explained avoidance behaviours. This suggests that other mechanisms than perceived severity might be relevant in explaining the tendency to confront a powerful perpetrator. One possibility is that bystanders are more prone to confront the perpetrator due to other reasons, such as perceiving the interaction to be unjust Porath et al. (2011) or because of negative emotional reactions towards the perpetrator (Reich and Hershcovis, 2015).

Interestingly, neither the perpetrator’s power position nor the perceived severity appeared to influence the bystanders’ extrinsic motivation. Rather, they enhanced the more autonomous forms of introjected, identified and intrinsic motivation. This might be positive from the perspective of bystander intervention because previous research among adolescents has shown that autonomous motivation is related to more active prosocial behaviours and is negatively associated with passive bystander behaviours (Jungert et al., 2016). Consequently, bystanders might be more autonomously motivated to intervene when witnessing a powerful perpetrator engage in incivility instead of relying on external incentives to act. This is partly consistent with the findings from Study 1, where more active and less passive bystander behaviours were predicted by perpetrator power and perceived severity.

Practical implications

From a practical perspective, our findings that perceived severity might be an instrumental factor in why bystanders offer social support to the victim, and might be a factor for increased motivation to intervene, can be used to increase prosocial bystander behaviours. For example, how victims and others in an organisation talk about and define incivility might increase motivation to defend victims and might encourage them to take a more supporting role, for instance, managers could be stressing the severity of incivility in the workplace and emphasising the moral responsibility to intervene. In particular, it could be effectual for managers to highlight the negative consequences of workplace incivility that have been established in the literature, to demonstrate its potential severity (Holm, 2021; Schilpzand et al., 2016). This may otherwise not be evident due to the low-intensity nature of the behaviour. Support could also be offered to observers of incivility to help them define the severity of the situation, and this might foster identified motivation in particular because identified motivation has strong advantages over introjected motivation in terms of its stability, persistence and affective embellishments, while introjected motivation can both diminish intrinsic motivation and create distress (Leary, 2004; Ryan and Deci, 2017). Overall, our findings show a novel way of applying SDT concepts and further confirm the theory’s relevance in a work context that has not been investigated previously.

Moreover, our findings show that lower perpetrator power is associated with a smaller tendency to confront the perpetrator and a higher tendency to avoid the perpetrator. However, being a victim of incivility is related to many negative outcomes even if the perpetrator has low power (Schilpzand et al., 2016). If there is less chance of bystander intervention when the perpetrator has low power, it may be even more important for managers to take action in cases where incivility is being exhibited by a low power perpetrator, as it may otherwise remain unaddressed by bystanders. This finding highlights the importance for organisations and managers to have a strategy on how to perceive and intervene in incidents of incivility even when there is low perpetrator power involved. By identifying factors that are relevant to bystanders’ helping intentions, the present study contributes with new important insights into how incivility can be addressed practically. The knowledge derived from the study can also be integrated into larger intervention programs that are currently being developed to enhance bystander intervention initiatives in response to workplace bullying (Einarsen et al., 2020).

Limitations and future directions

Some considerations in this study must be noted. Social desirability is a likely factor that can affect how people choose to act in a particular situation, even if people do not consciously see such desirability as their motivation for their actions. We tried to eliminate social desirability through the anonymous surveys, but we cannot completely rule out that it might have influenced the results. Furthermore, in Study 2, we used a small variety of fictional vignette with different kinds of incivility, and it is possible that the participants might have had difficulty putting themselves entirely in the shoes of a bystander via the vignette. On the other hand previous studies on prosocial interventions have shown that intentions powerfully indicate real behaviour (Smith and McSweeney, 2007), and the present study used mixed methodologies by combining an observational study with an experimental study to assess intentions to help. Nevertheless, on the basis of these results we cannot conclude how the bystanders’ motivation corresponds to actual intervention behaviour. This would be an interesting area to explore in future studies.

Another limitation is that Chronbach’s alpha showed low reliability for some of the motivation variables, as well as the avoidance measure. The relatively low reliability could to some degree limit the validity of the findings. Furthermore, severity was measured with a one-item question assessing how severe the participants considered the described behaviour. Because single-item measures are sensitive to measurement error, this could potentially have distorted the study’s results. Nevertheless, single-item measures may be acceptable if the item represents a homogenous and unidimensional construct (Wanous et al., 1997). In addition, it is important to consider that statistical power may have been limited by the sample size in the present study. Larger samples would be warranted to reduce the risk of type I and type II errors. Future studies should attempt designs that ameliorate these shortcomings in order to further validate the findings of the present study.

A further limitation is that there was no neutral condition in Study 2 where both parties had similar power. Future studies should include a neutral condition in addition to the high and low power conditions.

Moreover, the effects of power may have been influenced by culture. Given that Sweden is a country with a comparatively low power distance, which means that less powerful members of organisations expect power to be distributed equally, the results may have been affected by its cultural context. The cultural setting has been found to affect individuals’ acceptability of incivility at work (Moon and Sánchez‐Rodríguez, 2021), such that high power distance orientation individuals were less likely to perceive their supervisors as behaving unfairly Lian et al. (2012) and accepting subtle forms of mistreatment (Moon et al., 2021). In other words, the results of the current study may be limited to cultures low in power distance. A limitation in Study 2 was that the imaginary scenario was based on the role of occupational position (manger and subordinate coworker), not power. Although a higher position normally is associated with higher power in a work context, we cannot know whether all participants interpreted the scenario in that way. Future research should therefore include a manipulation check of power to determine to what extent results are due to the power difference. Finally, both Study 1 and Study 2 focused on only one incident of incivility in the workplace. Future research should be undertaken to investigate whether a longer time span (e.g. incivility over several days or weeks) will result in similar associations between perpetrator power, severity, bystander behaviour and motivation.

Conclusion

Taken together, the results of this study showed that higher perpetrator power was related to uncivil situations being perceived as more severe. Higher perpetrator power was also directly associated with a greater tendency to confront the perpetrator and a lower tendency to avoid the perpetrator. Perceived severity was also shown to be an important mechanism in the relationship between the perpetrator’s power position and the bystanders’ tendency to offer social support to the victim, as well as their motivation to intervene. In particular, introjected, identified and intrinsic forms of motivation to intervene were influenced by perpetrator power via higher levels of perceived severity.

Figures

Figure 1.

Path model of the relationships between perpetrator power, perceived severity and confrontation, avoidance and offering social support

Figure 2.

Path model of the relationships between perpetrator power, perceived severity and extrinsic, identified, introjected and intrinsic motivation

Table 1.

Pearson’s correlations, means and standard deviations of observed study variables

Variable	M (SD)	1	2	3	4	5
1. Power	2.78 (1.23)	–
2. Severity	3.56 (1.23)	0.48***
3. Confront	2.21 (1.26)	0.58***	0.31***
4. Avoid	2.30 (1.06)	−0.27***	−0.12	−0.49***
5. Social support	3.83 (1.20)	0.35***	0.52***	0.25**	−0.12	–

Notes:

*p < 0.05, **p < 0.01, ***p < 0.001

Table 2.

Pearson’s correlations, means and standard deviations of observed study variables

Variable	M (SD)	1	2	3	4	5	6
1. Power	–	–
2. Severity	3.22 (0.73)	0.39***
3. External regulation	1.84 (0.62)	−0.06	−0.12
4. Introjected regulation	3.25 (0.75)	0.20**	0.33***	0.35***
5. Identified regulation	4.37 (0.66)	0.18*	0.34***	−0.11	0.35***
6. Integrated regulation	3.38 (0.95)	0.14	0.23**	0.11	0.33***	0.51***	–

Notes:

*p < 0.05, **p < 0.01, ***p < 0.001

Appendix. Study 2

Manager acting in an uncivil way

Imagine that you are at your current workplace. At your workplace, there is a staff room where the staff can go to have water or coffee during their breaks. Your workplace also has a boss. One day when you are having coffee in the staff room, you are involved in the following situation. You and three other workers are sitting at the coffee table. Two of them, your boss and Person B, are sitting close to you, while the third person, Person A, is sitting a bit further away, but can hear what is being said. Your boss, Person B, and you are talking to each other in a quite friendly way. Your boss says, “What a great time we had yesterday afternoon when we had a beer together at the bar!” Person B answers, “Yes, it was great that only our really fun co-workers were there”. Both of them glance at Person A, who looks down at the floor and looks uncomfortable. A few minutes later, you and your boss are standing by the coffee machine drinking coffee. Person A walks up to the two of you, turns to your boss, smiles hesitantly and says “Do you guys know if anything fun is going on with everyone else after work soon?” Your boss answers, “Yeah, we’re going to arrange something, but do you really have time for those kinds of things, being recently divorced from your wife and all? Or what did you call it, you just took a break from each other?” Person A looks hurt and ill at ease.

Subordinate acting in an uncivil way

Imagine that you are at your current workplace. At your workplace, there is a staff room where the staff can go to have water or coffee during their breaks. Your workplace also has a boss. One day when you are having coffee in the staff room, you are involved in the following situation. You and three more workers are sitting at the coffee table. Two of them, Person A and Person B, are sitting close to you, while the third person, your boss, is sitting a bit further away, but can hear what is being said. Person A, Person B and you are talking to each other in a quite friendly way. Person A says, “What a great time we had yesterday afternoon when we had a beer together at the bar!” Person B answers, “Yes, it was great that only our really fun co-workers were there”. Both of them glance at your boss, who looks down at the floor and looks uncomfortable. A few minutes later, you and Person A are standing by the coffee machine drinking coffee. Your boss walks up to the two of you, turns to Person A, smiles hesitantly and says, “Do you guys know if anything fun is going on with everyone else after work soon?” Person A answers, “Yeah, we’re going to arrange something, but do you really have time for those kinds of things, being recently divorced from your wife and all? Or what did you call it, you just took a break from each other?” Your boss looks hurt and ill at ease.