AbstractVoice endorsement—managers’ approval of employees’ voiced suggestions—has largely been regarded as positive, yet little research examines how managers’ emotional expressions during endorsement affect employees. Drawing on the Emotions as Social Information (EASI) model, we propose that not all instances of voice endorsements are necessarily positive events for employees. Specifically, we propose that managers’ anger while endorsing voice, negatively impact employees’ perceptions of psychological safety and their future voice behavior. Across two experimental studies, we find that participants who faced an angry manager endorsing their voice reported both lower psychological safety and future voice behavior than those whose voice was endorsed by a happy or neutral looking manager. Study 2 further found that employees experiencing voice endorsement with anger felt as psychologically unsafe as those whose voice was outright rejected, suggesting that anger-laden endorsements are just as detrimental as voice rejections. Our findings contribute to the literature by challenging the notion that all voice endorsements are inherently positive, emphasizing the role of managers’ emotions on managerial reactions to voice.
IntroductionVoice—a proactive work behavior that involves speaking up with suggestions for improvement1,2, can foster key outcomes, such as performance gains3,4,5 and innovation6,7,8. In order for voice to positively impact outcomes, someone with authority—often a manager—needs to endorse voice so that voice can be enacted and implemented9. While voice can positively impact work, managers sometimes need to turn down voiced ideas, which is not a pleasant event for employees10,11. In contrast, research shows that employees experience positive emotions when their voice is endorsed12,13. As such, researchers have regarded voice endorsement as an unequivocally positive event for employees10,11,14,15,16.However, the positive consequences implied in voice endorsement notwithstanding, managers may not always react unequivocally positively to voice as, for example, managers often display negative emotions when interacting with their employees17. Research shows converging evidence that managers can feel reluctance toward employee voice18,19,20. As managers sometimes may have limited mental resources to deal with voice15, depleted managers can react with negative emotions such as anger21, especially when confronted with challenging ideas22, or with ideas that uncover previous incompetence23. As managers’ emotional expressions, such as anger, can alter how employees interpret important work events24, the emotions managers display while endorsing voice may make more of a difference on employees’ attitudes and future voice expressions.Understanding how managers’ expression of anger (vis-à-vis positive emotions such as happiness, or neutral expressions) while endorsing voice affects employees is important for conceptual and practical reasons. Although voice endorsement has been generally regarded as a positive event for employees10,11,14,15,16, angry managers may harm employees24. In fact, employees do not appreciate managers when they display anger25 because employees perceive angry managers as though26, abusive27, and a threat to their sense of safety28. Ultimately, although the manager may be endorsing employees’ ideas, thus not rejecting their suggestion, employees may interpret their anger as a sign of a personal rejection29. Overall, this paradox in which employees may receive a positive signal when their voice is endorsed, but at the same time their angry manager may lead them to feel unsafe and unwilling to speak up again has not received enough scholarly attention. Therefore, it is important to understand how managers’ displays of anger affects employees’ voice attitudes and future voice expressions. From a practical perspective, if employees perceive an angry endorsement in a similar fashion to having their voice turned down, then these instances of endorsement may be as destructive as an instance of outright rejection.To understand how managers’ displays of anger while endorsing voice affects employees’ attitudes and future voice expressions, we draw from the emotions as social information (EASI) model30. The EASI model suggests that displayed emotions affect the way employees interpret social situations. As such, we propose and show that when angry managers endorse voice, employees experience lower levels of psychological safety compared to having their voice endorsed by a happy or neutral looking manager. In turn, this reduced sense of safety is associated with employees’ lower intentions to voice in the future.Our research contributes to the literature in important ways. First, our results challenge the prevailing notion that all instances of voice endorsement are unequivocally positive. Our results consistently reveal that endorsing voice with a happy or neutral expression produce more positive downstream consequences on employees compared to endorsing voice with an anger. Second, our research sheds light on understanding managerial reactions to specific voice instances. Whereas previous research has primarily explored managers’ reactions to voice behavior in general, we join recent scholarship that investigates managers’ reactions to specific voice instances10,11. Lastly, our research is among the first studies investigating managers’ emotional expressions when reacting to their employees’ voice. Whereas previous research has studied how managers’ and peers’ moods affect voice, we show that even managers’ subtle displays of anger can significantly affect employees after they voice.Theoretical backgroundThe EASI modelAs emotions convey critical information about others’ intentions, the emotions-as-social-information (EASI) model highlights the significant role that discrete emotions play in social situations30. According to the model, emotions act as relevant information cues that influence how individuals interpret and respond to their social environment allowing individuals to navigate complex social situations31. Thus, individuals interpret emotional signals and cognitively process them adjusting their reactions and behavior accordingly30. Research on the EASI model emphasizes the importance of discrete emotions, such as anger or happiness as each of them provides specific social signals that can lead to different responses. For instance, expressions of happiness can signal cooperation; whereas anger can signal toughness and treat32.Voice behavior and voice endorsementVoice is a proactive work behavior that focuses on employees’ verbal expression of constructive ideas, opinions, or suggestions aimed at improving the status quo33,34. Voice seeks to benefit the workplace by promoting constructive change35. However, as voice behavior is aimed at introducing changes into the organization, engaging in voice is risky because employees who speak up may suffer negative consequences for challenging the status quo1,36. Because voice is considered risky, the literature largely shows that employees assess their social environment when considering engaging in voice, and only decide to speak when they perceive they will not be harmed for challenging the status quo37. Thus, research shows that one of the main voice antecedents is psychological safety—employees’ perceptions that they will not be punished for speaking up38.From the managers’ perspective, studies show that managers often have negative reactions towards voice. For instance, as managers tend to focus more on short-term goals than on long term benefits, they dismiss voice as it can create disruptions that might only pay off in the long run18. Similarly, managers low on managerial self-efficacy react negatively to voice because voice threatens their ego19. Another study shows that managers are more likely to dismiss voiced ideas when employees speak up publicly (rather than privately) out of concern of the imagine they seek to portray as leaders20.Scholars have portrayed voice endorsement largely as an unequivocally positive event for employees. As such, when managers endorse employees’ voice, managers reinforce employees’ perceptions that speaking up is safe, that their ideas are not only welcome, but valuable and encouraged11. Conversely, when managers do not endorse voice, they signal that it is not safe to speak up and that their ideas are neither welcomed nor appreciated10,11. While the majority of research on voice endorsement has focused on its antecedents, one study that investigates its consequences shows that employees who have their voice endorsed experienced positive affective states as a result12.Hypothesis developmentEndorsing voice with angerExisting voice research largely conceptualizes voice endorsement as a generally positive experience that reinforces employees’ safety perceptions. We contend, however, that employees’ reactions to having their voice endorsed by a manager depend on the emotional expression that the manager displays at the moment of endorsing voice.From the employees’ perspective, the mere fact that a manager is endorsing their voiced ideas is in itself a critical event as most voiced ideas are ignored the first time they are communicated39. Given that endorsing voice is a critical event for employees, employees do pay close attention to the way their managers react to voice40. Research on voice rejection shows that the way managers reject voice can buffer the potential negative repercussions for employees for having their voice turned down. For instance, when managers turn down voice providing a sensible explanation11, or when managers use benevolent humor10, employees are more likely to voice again in the future.The way managers react to voice is an important signal for employees because it influences how employees will expect their managers to react to voice in the future. Aligned with the dominant views of the literature11,41, we expect that when managers endorse voice with a neutral expression, their employees will consider that it is safe to speak up again in the future because the act of endorsement is by itself a signal that the manager is open to voice. In this case, although the manager is not portraying any emotional cues, the act of endorsement will by itself provide a positive signal for employees. Similarly, when the manager endorses voice with happy expression, employees will receive additional positive signals from their manager. By endorsing voice with happiness, the manager signals that they are not only is open to voice by, but happy about it. Employees can perceive this cue as their managers way of showing that their voice is worthwhile. In both cases, we expect that employees will not think that future voice will bring them harm and thus they will perceive that it is safe to speak up again in the future.In contrast, when an angry manager endorses voice, their response to voice is not unequivocally positive. Although the manager is endorsing voice, which may be construed as a positive reaction, by displaying anger the manager’s emotions may be signaling that he or she does not fully and wholeheartedly support their employees’ voiced idea. There are several reasons why a angry manager may support voice. First, managers are more likely than other employees to display negative other-directed emotions, such as anger42. In fact, employees often tell stories about “angry bosses”43. Research on managerial anger shows that dealing with subordinates is one of the leading causes of their anger43. For instance, if managers perceive employees’ voice as disrespectful, managers are likely to react with anger44. Hence, while it may seem off-kilter to be angry when endorsing employee voice, managers may agree and endorse their employees’ suggestions while still experiencing anger for related or unrelated reasons. Indeed, voice is a constructive and deviant behavior45, implying that a manager’s response to voice is a conflict management behavior, and we know from research on negotiation and conflict management that anger is a common and effective negotiation tactic31,46.Voice behavior is largely driven by psychological safety and voice impact47. Yet, we expect that expressions of anger to be more closely related to safety than to impact (though we will statistically account for the role of voice impact in our studies). From the employees’ perspective, when they see their manager as angry, they will perceive their manager as conveying toughness26. Existing research also suggest that displays of anger will result in the manager appear as a threat to the employees28. Furthermore, when managers react with anger, their employees may perceive that even though their manager is endorsing their ideas, the manager is ultimately rejecting them29. We expect these reactions would make employees think about their social risks and their self-preservation, both of which are more related to their sense of safety rather than to their impression of whether voice would be acted upon. Not surprisingly, existing research shows that managers’ anger negatively affects employees’ sense of safety48.In sum, we expect that when angry managers endorse (relatively to happy or a neutral looking managers), employees will see their manager as tough and as a threat to them, which will impact how safe they will feel when considering speaking up again. Thus, we propose:
Hypothesis 1
Managers’ displays of anger (relative to happiness or a neutral expression) while endorsing voice are negatively related to employees’ perceptions of psychological safety.
Psychological safety refers to the employees’ perceptions that they will not be punished for voicing38. Previous research consistently shows that employees are more likely to speak up when they believe they will not face any harm for doing so37. Similarly, when employees do not feel safe, they tend to withhold voice49. Following this evidence, we expect that when employees experience high levels of psychological safety after having their voice endorsed, they will continue to voice in the future.Taken together, the previous argument in combination with Hypothesis 1 suggest an instance of mediation. Specifically, when angry managers endorse voice (relative to happy or neutral looking managers), employees will experience lower levels of psychological safety, which will negatively affect their future voice behavior. Our prediction is consistent with the EASI model insofar employees’ assessment of their managers’ emotional expressions not only affect employees’ attitudinal reactions, but also their future work behavior. Thus, we propose,
Hypothesis 2
Employees’ perceptions of psychological safety mediate the relationship between managers’ displays of anger while endorsing voice and employees’ future voice behavior.
MethodEthics statementEthical approval for this research was received from the Institutional Review Board of the research laboratory of the first author’s institution, the CNRS Research Center Lille Economie et Management, UMR 9221. As such, data-related procedures for the data presented in this article were conducted in accordance with the ethical standards of the European Union and with the 1964 Helsinki declaration and its later amendments or comparable ethical standards. Informed consent was obtained from all individuals participants included in this research.Transparency and opennessOur data was processed and analyzed utilizing Stata 17 MP. Our research design and hypotheses were not preregistered. Our data, syntax, and supplemental analyses are openly available as an Online Supplement in the Open Science Framework (OSF) at: https://osf.io/zrsuh/?view_only=dacdb9f95be24b80b54508dd36a4f180.Study 1We conducted Study 1 to provide initial evidence of the effect of managers’ displays of anger while endorsing voice on employees’ attitudes and voice behavior utilizing a between-subjects experimental design that conducted online on Qualtrics.com. Participants took the role of a coffee shop employee who voiced an idea during a meeting and got their idea endorsed by the manager. To ascertain whether the emotion the manager displayed while endorsing participants’ idea had our predicted effect, we randomly assigned participants into three conditions where we manipulated the managers’ emotional expression at the time of the endorsement by showing participants an image of their manager. Thus, participants saw a picture a man who either displayed an angry, happy, or neutral expression. Our study materials, including these images appear in the Online Supplement.Participants and designWe analyzed data from 175 USA-based individuals from Prolific.com. We used G*Power to conduct a power analysis. We followed the standard recommendations to detect a small effect size of 0.25, with a 0.05 alpha error probability and a power of 0.80, which results in a recommended sample size of 159 (i.e., 53 per cell). We thus collected a slightly larger sample size in case we needed to remove some participants. Therefore, although we initially recruited 180 participants, we removed 5 who failed at least one out of three attention checks. We paid all study participants 1.35 British pounds for their time. Participants self-reported their sex as female (49.1%), male (48.6%), or other (2.3%). Participants had a mean age of 39.43 years old (SD = 12.42). In terms of their employment, majority reported being employed (73.1%), looking for a job (7.4%), homemakers (7.4%), students (5.1%), or in another situation (6.9%).We divided participants into three experimental conditions: Facing an angry manager (N = 58), facing a happy manager (N = 59), and facing a manager with a neutral expression (N = 58). In all conditions, participants were asked to take the role of a coffee shop employee in a placed called Deluxe Beans. They read that Deluxe Beans is frequented by educated white-collar professionals who are willing to pay a higher prize for high-quality products and customer service. Participants further read that they have been working in there for the past two years, they know how to do their job, and that all employees (including them) report to Andrew, the general manager of the coffee shop.Next, we provided participants with some context about their job. Participants read that the general manager of the coffee shop holds a staff meeting at the beginning of every week to discuss the plan for the week, such as special products or new promotions, staff work schedules, and so on. Here, participants read that, “During the last staff meeting, Andrew talked about how baristas behind the counter are constantly bumping into one another, which slows down service during peak hours. To resolve this, your manager Andrew proposed a new ‘behind the counter plan’ that seeks to reduce the number of baristas employed during a shift, which in turn should reduce the number of baristas behind the counter and would allow things to run more smoothly.” Participants then read that after their manager had finished explaining their plan, they raised their hand with a suggestion because they were not sure their manager’s new procedure would work because there would not be enough people to prepare drinks when demand was high. This is where participants were made aware that they voiced an idea to their manager. Thus, we told participants that they spoke up to their manager saying “[I think we need to] keep the same number of employees but redesign the workspace layout to create clear, designated zones for different tasks (e.g., espresso machine area, milk steaming area, syrup station, etc.). Then, each barista will only handle one section of the process at a given time, which would make their work smoother and allow them to work as a team and keep up with demand during peak hours.”After participants finish reading about the idea they proposed to their manager, we asked them to advance to the next screen to see an image of how their manager reacted to their suggestion, as well as how he responded. In the next screen, all participants read that their manager had endorsed their idea: “Andrew: Thanks for bringing this up. I see the benefits that your idea could have for us and our customers. Let’s try your way and see how it goes.”At this point we separated participants into the three experimental conditions (i.e., facing an angry manager vs. facing a happy manager vs facing a manager with a neutral expression). In each condition participants saw a picture of an individual displaying the emotion corresponding to their condition. These images were developed and validated previously50, and have been used to investigate interpersonal effects of emotions51. Afterwards, participants responded to the rest of the survey, and received payment.MeasuresUnless otherwise indicated, participants responded utilizing a 5-point Likert type scale (1 = Strongly disagree, 5 = Strongly agree). The full item list appears in the Online Supplement.To assess psychological safety, we utilized the 4-item scale by Hamstra and colleagues10 that was based on the measure by Wei and colleagues52. A sample reversed-coded item is, “if I were to speak up to my supervisor, he would evaluate my performance negatively.” (α = 0.95).We measured voice in line with other experimental voice studies53,54. Study participants were presented with another situation in which they had the chance to speak up to their manager, and they reported their willingness to do so. Specifically, participants read that two weeks had passed since the last time they spoke up to their manager at the Deluxe Beans coffee shop. We then asked them to “imagine that you are participating again in your scheduled regular staff meeting. This time, your manager suggested that in order to increase sales the coffee shop will start advertising to college students so that they can come to the coffee shop when they are not in class. Clearly, you can see that your manager has not thought through this new plan as college students are not the target audience of your coffee shop (i.e., you typically serve young professionals).” We then asked participants if they were willing to express their concerns and/or make suggestions, (“yes” = 1, “no” = 0). This operationalization of voice follows a decision-making perspective and represents a behavioral measure of voice54. In line with previous research53, we believe this measure is appropriate in our study because it is focused on whether participants are willing to speak up to their manager on a specific situation (i.e., shortly after their manager had endorsed their voice displaying a specific emotion).To ascertain whether participants perceived the appropriate emotion from the facial images we displayed to them, we asked participants to indicate the “extent to which you think your manager at Deluxe Beans felt this way when reacting to your input.” We then presented participants with three items intended to assess their managers’ perceived anger (angry, irritated, and furious, α = 0.90) and happiness (happy, enthusiastic, cheerful, α = 0.89).Results Study 1Table 1 depicts Study 1 descriptive statistics, reliabilities, and correlations.Table 1 Descriptive statistics, reliabilities, and correlations (Study 1).Full size tableA one-way analysis of variance (ANOVA) revealed that participants perceived their manager as angrier (M = 2.58, SD = 1.13) in the facing an angry manager condition compared to the facing a happy manager condition (M = 1.80, SD = 0.72, F(1, 115) = 20.07, p < 0.001) and the facing a manager with a neutral expression condition (M = 2.10, SD = 0.75, F(1, 114) = 7.40, p = 0.008). Similarly, participants perceived their manager as significantly happier (M = 3.61, SD = 0.80) in the facing an happy manager condition compared to the facing a angry manager condition (M = 3.01, SD = 1.07, F(1, 115) = 11.84, p < 0.001) and marginally happier than the facing a manager with a neutral expression condition (M = 3.34, SD = 0.84, F(1, 115) = 3.05, p = 0.083). Overall, these results confirm the effectiveness of our visual stimuli and its intended manipulation effect.We tested Hypothesis 1 utilizing ANOVA with psychological safety as the dependent variable and the experimental conditions (2 = facing an angry manager, 1 = facing a happy manager, and 0 = facing a manager with a neutral expression) as the independent variable. We plotted these results in Fig. 1. In support of Hypothesis 1, there was a significant effect of anger, F(2, 172) = 18.74, p < 0.001, partial eta-squared = 0.18. Tukey post-hoc contrasts indicated that participants in the facing an angry manager condition reported lower psychological safety (M = 3.31, SD = 1.10), than those who faced a happy manager (M = 4.35, SD = 0.74, t = 6.06, p < 0.001) and a manager with a neutral expression (M = 3.97, SD = 0.91, t = 3.79, p = 0.001), but there were no significant differences between those in the facing a happy or neutral manager (p = 0.065).Fig. 1Means of Psychological safety by condition (Study 1).Full size imageRegarding mediation, we followed the best practice recommendations by Hayes and Preacher55 when testing mediation with multicategorical independent variables. We began by dummy coding all experimental groups, using the voice endorsement with a neutral expression group as the reference group. As a result, we created two dummy variables to represent the two remaining conditions: endorsing voice with happiness and endorsing voice with anger. We then conducted OLS regression with voice behavior as the dependent variable, and the independent variables included psychological safety and the two dummy variables. Additionally, the dummy variables predicted psychological safety. To test for mediation, we computed indirect effects by calculating bias-corrected confidence intervals (CI) utilizing 5,000 bootstraps replications. Mediation is present when zero is absent from the confidence interval56.The results of our mediation analysis show that psychological safety positively mediates the relationship between the happiness dummy on voice (indirect effect = 0.07, bias corrected CI = (0.02, 0.15)). The total effect for both the anger dummy (coefficient = -0.03, bias corrected CI = (-0.21, 0.15)) and the happiness one (coefficient = 0.11, bias corrected CI = (-0.07, 0.28)) included zero in their respective confidence intervals. Similarly, the direct effect on voice behavior of both the anger dummy (coefficient = 0.09, bias corrected CI = (-0.07, 0.25)) and the happiness dummy (coefficient = 0.04, bias corrected CI = (-0.15, 0.21)) also included zero in their respective confidence intervals. These results are consistent with the bivariate correlations between voice behavior and the happy dummy (r = 0.12, p = 0.107) and the anger dummy (r = -0.09, p = 0.257) that are reported on Table 1. Moreover, and consistent with Hypothesis 2, psychological safety negatively mediates the indirect effect of the anger dummy on voice behavior (indirect effect = -0.12, bias corrected CI = (-0.22, -0.05)), further reinforcing psychological safety’s key role in shaping voice.Robustness checksWe conducted additional analyses to investigate the robustness of our findings. First, to rule out the possibility that facing an angry manager would affect voice for reasons unrelated to psychological safety, we explored whether other potential mediators—voice impact and positive affect—could explain our findings. Voice impact has to do with employees’ perceptions that their voice can significantly influence their manager57. We assessed voice impact because employees who have their voice endorsed may believe they can further impact their supervisor with their future voice behavior. We assessed voice impact with the 3-item measure by Tangirala and Ramanujam57. A sample item is, “my ideas, opinions, and suggestions have significant influence on my supervisor” (α = 0.85). Moreover, we assessed high-activated positive affect because having one’s ideas endorsed is a positive emotional event, and thus it is possible that participants engage in future voice behavior because they feel a surge of positive affective energy58. We assessed high-activated positive affect with four items developed by Warr and colleagues59(happy, enthusiastic, inspired, and cheerful, α = 0.93).We repeated our mediation analyses predicted by Hypothesis 2 while including voice impact and positive affect as alternative mediators. In this case, similar to our main analyses the total effect for both the anger dummy (coefficient = -0.03, bias corrected CI = (-0.21, 0.16)) and the happiness one (coefficient = 0.11, bias corrected CI = (-0.06, 0.28)) as well as the direct effect on voice behavior of both the anger dummy (coefficient = 0.12, bias corrected CI = (-0.03, 0.28)) and the happiness dummy (coefficient = 0.03, bias corrected CI = (-0.12, 0.20)) included zero in their respective confidence intervals.In terms of our proposed alternative mediators, voice impact does not mediate the indirect effect on voice from anger (indirect effect = -0.04, bias corrected CI = (-0.13, 0.02)) or happiness (indirect effect = 0.02, bias corrected CI = (-0.01, 0.08)) when the other two mediators are involved. High-activated positive affect does mediate the indirect effect on voice from both anger (indirect effect = -0.12, bias corrected CI = (-0.24, -0.03)) and happiness (indirect effect = 0.07, bias corrected CI = (0.01, 0.17)) when the other two mediators are involved. Lastly, our results for psychological safety remain consistent with our main analyses when the two alternative mediators are included since it mediates the indirect effect from both anger (indirect effect = -0.06, bias corrected CI = (-0.16, -0.01)) and happiness (indirect effect = 0.03, bias corrected CI = (0.00, 0.11)).We also examined whether participants’ gender had any effect on our findings. We repeated our main analyses including gender as a covariate, and our results and conclusions did not change with its inclusion. Furthermore, we employed the Wald test for joint significant to investigate whether gender moderated the effect of our experimental condition on psychological safety (F(2, 165) = 1.02, p = 0.365) and voice behavior F(2, 165) = 0.76, p = 0.468), and found that it does not significantly affect these outcomes. In sum, we conclude that participants’ gender did not meaningfully affect our study findings. Overall, these additional analyses bolster the confidence in our study findings.Discussion (Study 1)Study 1 provides support for our main predictions. Study participants experienced lower levels of psychological safety when an angry manager endorsed their voice, relative to a happy or neutral looking manager. Furthermore, facing an angry manager negatively affected participants future voice behavior due to participants’ lower levels of psychological safety.Previous research on angry managers have largely shown the detrimental effects on subordinates resulting from managers’ public displays of anger43. Study 1 results uniquely contribute to existing research because they show that the negative effects of anger are context dependent. That is, unlike situations where an angry manager expresses anger because they are dealing with an incompetent or immoral employee, where such anger could be expected and even justified, voice endorsement is typically considered a positive work event. This creates a unique situational incongruence where managers are simultaneously signaling approval and expressing anger, which challenges employees’ interpretations and expectations as voice endorsement should foster psychological safety.Although Study 1 provided evidence of our predicted effects, a key question remains unanswered: how detrimental is for employees to have their voice endorsed voice by an angry manager? As Study 1 only allowed us to compare different emotional expressions during a relatively positive situation for employees (i.e., having their voice endorsed), it remains unclear whether the consequences of having their voice endorsed with anger is comparable to a relatively negative situation for employees (i.e., having their voice rejected). We expect that participants who have their voice endorsed with anger will experience similarly negative consequences on their psychological safety and future voice behavior relative to participants who have their voice rejected. Our prediction is consistent with the EASI model as managers’ displays of anger have been shown to produce stronger negative reactions in subordinates compared to receiving negative feedback without expressing anger60.Study 2The objective of Study 2 is twofold. First, we conducted Study 2 to replicate our results from Study 1. Additionally, we wanted to investigate how detrimental is for employees to have their voice endorsed voice by an angry by expanding our model to include two managerial reactions to voice—a positive reaction, voice endorsement, and a negative reaction, voice rejection. We thus conducted a 3 (manager emotional display: anger vs happy vs neutral) × 2 (manager reaction: voice endorsement vs voice rejection) between-subjects experimental design.Participants and designWe analyzed data from 218 USA-based individuals from Prolific.com. We used G*Power to conduct a power analysis. Aligned with Study 1, we sought to detect a small effect size of 0.25, with a 0.05 alpha error probability and a power of 0.80, which results in a recommended sample size of 216 (i.e., 36 per cell). Therefore, we initially recruited 240 participants and removed 22 who failed at least one out of four attention checks. We did not allow Study 1 participants to participate in Study 2. All participants received 1.35 British pounds for participating. Most participants self-reported their sex as female (51.8%), and the rest as male (48.2%). Participants had a mean age of 38.32 years old (SD = 12.23). In terms of their employment, majority reported being employed (78.9%), looking for a job (7.3%), homemakers (3.7%), students (4.6%), or in another situation (5.5%).We divided participants into six experimental conditions: Facing an angry manager who endorsed their voice (N = 37), facing a happy manager who endorsed their voice (N = 32), facing a manager with a neutral expression who endorsed their voice (N = 38), facing an angry manager who rejected their voice (N = 40), facing a happy manager who rejected their voice (N = 36), and facing a manager with a neutral expression who rejected their voice (N = 35).We adapted the scenario and cover story from Study 1 with one main difference. Right after participants finish voicing their plan aimed at resolving the issues baristas faced behind the counter, they were also asked to advance to the next screen to see an image of how their manager reacted to their suggestion. Here is where we introduced Study 2 manipulations. Specifically, participants in the voice endorsement condition read the same reaction from Study 1: “Andrew: Thanks for bringing this up. I see the benefits that your idea could have for us and our customers. Let’s try your way and see how it goes.” However, participants in the voice rejection condition read, “Andrew: Thanks for bringing this up. I don’t see the benefits that your idea could have for us and our customers. Let’s try my way and see how it goes.”Participants in both the voice endorsement and the voice rejection conditions saw one of three images (i.e., facing an angry manager vs. facing a happy manager vs facing a manager with a neutral expression). We utilized the same images from Study 1. Afterwards, participants in all conditions answered the rest of the survey which included the same questions from Study 1.MeasuresWe utilized the same measures from Study 1 to assess the Study 2 variables.Results Study 2Table 2 depicts Study 2 descriptive statistics, reliabilities, and correlations.Table 2 Descriptive statistics, reliabilities, and correlations (Study 2).Full size tableAn ANOVA test showed that participants perceived their manager as significantly angrier (M = 2.94, SD = 1.04) in the facing an angry manager condition compared to the facing a happy manager conditions (M = 2.24, SD = 1.09, F(1, 143) = 15.87, p < 0.001) and the facing a manager with a neutral expression conditions (M = 2.52, SD = 0.99, F(1, 148) = 6.24, p = 0.014). Participants also perceived their manager as significantly happier (M = 3.04, SD = 0.98) in the facing an happy manager condition compared to the facing an angry manager conditions (M = 2.56, SD = 1.05, F(1, 143) = 8.26, p = 0.005) but not significantly happier when the facing a manager with a neutral expression conditions (M = 2.78, SD = 1.08, F(1, 148) = 1.57, p = 0.213). Although we did not find significant differences between the happy and neutral condition for the happiness manipulation, we still decided to continue with our analyses as our key predictions are about differences relative to the angry condition.The results of a 3 × 2 ANOVA test shows a significant main effect of the managerial reaction condition (F(1, 212) = 94.50, p < 0.001, partial eta-squared = 0.31), the emotional display condition (F(2, 212) = 21.36, p < 0.001, partial eta-squared = 0.17), and a significant interaction between the two conditions (F(2, 212) = 6.49, p < 0.001, partial eta-squared = 0.07). We plotted this interaction in Fig. 2. Tukey post-hoc contrasts show that among participants in the voice endorsement condition, those who faced an angry manager reported lower psychological safety (M = 3.06, SD = 1.20) than those who faced a happy manager (M = 4.58, SD = 0.47, t = 7.06, p < 0.001) and a manager with a neutral expression (M = 4.06, SD = 0.95, t = 4.85, p < 0.001), but there were no significant differences between those facing a happy or neutral manager (p = 0.147). Thus, Hypothesis 1 is supported.Fig. 2Means of Psychological safety by condition (Study 2).Full size imageTo test for mediation, we followed mediation procedures from Study 155. Specifically, we first created dummy variables to code all six experimental groups. We then conducted OLS regression with voice behavior as the dependent variable, and the independent variables included psychological safety and five of the six dummy variables. Additionally, the selected five dummy variables predicted psychological safety. We conducted these analyses six times, such that we were allowed to compare our experimental condition against all other experimental conditions. We achieved this by selecting five dummy variables at a time and leaving out the dummy variable representing the condition we used as the baseline for the five selected dummy variables.Table 3 summarizes the results of our indirect effects. According to Table 3, the dummy corresponding to endorsing voice with anger negatively affects the indirect relationship with voice behavior via psychological safety when the reference group was endorsing voice with a neutral expression (Table 3, M1: indirect effect = -0.17, bias corrected CI = (-0.31, -0.08)) and when the reference group was endorsing voice with a happy expression (Table 3, M2: indirect effect = -0.26, bias corrected CI = (-0.42, -0.15)). These results support Hypothesis 2.Table 3 Indirect effects results of the mediation analysis of psychological safety predicting voice behavior with multicategorical independent variables (Study 2).Full size tableTo provide additional information regarding our mediation analyses, Table 4 depicts the total effects and Table 5 the direct effects of our mediation with multicategorical independent variables. While the direct and total effects of our dummy variables generally do not have a significant impact on voice, the indirect effect through psychological safety is significant, reinforcing the idea that psychological safety is a key mechanism explaining the relationship. Notably, according to Table 1 we found a significant negative correlation between voice and the anger dummy (r = -0.20, p = 0.003), but not with the happy dummy (r = 0.08, p = 0.229), suggesting that although anger has a bivariate association with voice, their relationship is better explained by their indirect relationship via psychological safety, which aligns with our core prediction.Table 4 Total effects results predicting voice behavior with multicategorical independent variables (Study 2).Full size tableTable 5 Direct effects results of the mediation analysis of psychological safety predicting voice behavior with multicategorical independent variables (Study 2).Full size tableRegarding our last prediction about the consequences of endorsing voice with anger vis-à-vis rejecting voice, Fig. 2 suggests that participants who got their voice endorsed by an angry manager experienced similar levels of psychological safety than participants who got their voice rejected. Tukey post-hoc contrasts show that participants who got their voice endorsed by an angry manager experienced similar levels of psychological safety (M = 3.06, SD = 1.20) than those who got their voice rejected by an angry (M = 2.58, SD = 0.67, t = 2.44, p = 0.148), happy (M = 2.99, SD = 1.01, t = 0.39, p = 0.999), or neutral looking manager (M = 2.64, SD = 0.80, t = 2.07, p = 0.308). It is important to note that the mean values for angry endorsement and happy rejection are virtually identical, clearly showing that endorsement is not universally better than rejection. Furthermore, according to Table 3, the voice endorsement with anger dummy did not indirectly predict voice via safety when the reference group was rejecting voice with a neutral expression (Model M3: indirect effect = 0.07, bias corrected CI = (-0.00, 0.17)) or rejecting voice with a happy expression (Model M4: indirect effect = 0.01, bias corrected CI = (-0.08, 0.10)). However, the voice endorsement with anger dummy did affect voice via safety when the reference group was rejecting voice with an angry expression (Model M5: indirect effect = 0.08, bias corrected CI = (0.02, 0.18)). In sum, we found no significant differences predicting psychological safety between the dummy representing endorsing voice with anger and three dummy variables representing voice rejection. In the case of the indirect effect on voice via psychological safety, although we did not find differences between the endorsing voice with anger dummy and the two of the endorsement conditions (i.e., neutral and happy expression), we find that endorsing voice with anger with a positive effect on voice behavior relative to rejecting voice with anger. As such, we conclude that endorsing voice with anger has similar consequences than rejecting voice with an angry or neutral expression but is less negative than rejecting voice with anger.Robustness checksAs a robustness check of our main prediction that voice endorsement with anger can be just as devastating as rejecting voice, we compared participants’ levels of psychological safety between the voice endorsement with anger condition to the combination of the three voice rejection conditions. Providing further support for our key prediction, an ANOVA test shows that employees whose voice was endorsed with anger experienced similar levels of psychological safety (M = 3.07, SD = 1.20) than those whose voice was outright rejected (M = 2.72, SD = 0.84, F(1,146) = 3.61, p = 0.060).Following Study 1, we repeated our mediation analyses including voice impact and positive affect as alternative mediators of our indirect effects predicted by Hypothesis 2. A summary of these analyses appears on Table 6. In terms of our proposed alternative mediators, Table 6 shows that voice impact does not mediate the indirect effect of the endorsing voice with anger condition on voice when the baseline is endorsing voice with a neutral expression (Table 6, M1: indirect effect = -0.02, bias corrected CI = (-0.09, 0.01)) and endorsing voice with happiness (Table 6, M2: indirect effect = -0.02, bias corrected CI = (-0.09, 0.01)) when the other two mediators are involved. Moreover, high-activated positive affect mediates the indirect effect of the endorsing voice with anger condition on voice when the baseline is endorsing voice with a neutral expression (Table 6, M1: indirect effect = -0.09, bias corrected CI = (-0.20, -0.03)) and endorsing voice with happiness (Table 6, M2: indirect effect = -0.13, bias corrected CI = (-0.26, -0.04)) when the other two mediators are involved. Furthermore, the dummy corresponding to endorsing voice with anger negatively affects the indirect relationship with voice behavior via psychological safety when the reference group was endorsing voice with a neutral expression (Table 6, M1: indirect effect = -0.10, bias corrected CI = (-0.22, -0.03)) and when the reference group was endorsing voice with a happy expression (Table 6, M2: indirect effect = -0.15, bias corrected CI = (-0.30, -0.04)) even when the other two variables were included as alternative mediators in the model. In sum, the results of this robustness check boosters the confidence in our findings by showing that psychological safety mediates the relationship between our experimental conditions and voice behavior even when other alternative mediators are included.Table 6 Indirect effects results of the mediation analysis of psychological safety, voice impact, and high-activated positive affect predicting voice behavior with multicategorical independent variables (Study 2).Full size tableIn line with Study 1, we also examined whether participants’ gender had any effect on our findings. We repeated our main analyses including gender as a covariate, and our results and conclusions did not change with its inclusion. Aligned with Study 1, we utilized the Wald test for joint significant to investigate whether gender moderated the effect of the managerial reaction condition on psychological safety (F(1, 214) = 0.35, p = 0.554) and voice (F(1, 214) = 2.42, p = 0.122), and of the emotion conditions on psychological safety (F(2, 212) = 0.48, p = 0.622) and voice (F(2, 212) = 1.02, p = 0.363). We also tested the possibility of a three-way interaction including participants’ gender and both experimental conditions affecting psychological safety (F(2, 206) = 0.38, p = 0.688) and voice behavior (F(2, 206) = 0.26, p = 0.771). We find that gender does not significantly moderate any of these effects. In sum, we conclude that participants’ gender did not meaningfully affect our study findings.Discussion (Study 2)Study 2 provides support for our predictions. In line with Study 1, Study 2 participants experienced lower levels of psychological safety when an angry manager endorsed their voice, and this in turn affected participants’ future voice behavior. Furthermore, results from contrasting the different experimental groups and our mediation analyses largely shows that endorsing voice with anger produces similar consequences than rejecting voice.General discussionOur research provides important insights into the role of emotions in the effect of managerial reactions to voice. In two experimental studies, we found that while voice endorsement is generally considered a positive event, managers emotional expression displayed while endorsing voice significantly shapes employees’ psychological safety and their future willingness to speak up. Specifically, we demonstrate that when angry managers endorse voice, employees report lower levels of psychological safety, which negatively influences future voice behavior. Our findings challenge the dominant view that all instances of voice endorsement are inherently beneficial and offer a nuanced understanding of how managers’ emotional expressions (even subtle ones) can shape employee voice, providing meaningful implications to the literature.Theoretical implicationsOur manuscript challenges the widely held assumption that all instances of voice endorsement are unequivocally positive for employees. While the prevailing narrative in the literature portrays voice endorsement as a constructive event that fosters psychological safety and encourages future voice expressions10,11,14,15,16, our findings reveal that managers’ emotional display accompanying voice endorsement can significantly alters its impact. Specifically, we demonstrate that when managers endorse voice with anger, the positive effects typically associated with endorsement are mitigated. This evidence suggests that voice endorsement is not a uniformly positive experience, but one that depends critically on the emotional expressions managers display when it occurs. Thus, we extend existing voice research by highlighting the complex interplay between endorsement and managerial emotions, offering a more updated view of how employees interpret and react to their managers’ reactions to voice.Second, our study also contributes to the emerging research that explores managerial reactions to specific voice instances. Whereas prior research often examines how managers react to employee voice behavior in general1,33, our study aligns with emerging scholarship that focuses on investigating voice over time by studying discrete voice episodes10,11,13. By examining how managers react to a specific voiced idea with different emotional expressions, we contribute to our understanding of voice as a more dynamic and iterative process. This shift from studying average managerial responses to investigating particular instances of voice provides a deeper insight into the interpersonal factors that influence employees’ decision to speak up. In doing so, our research enhances voice knowledge by emphasizing the importance of managerial emotions in the context of how managers react to specific voiced ideas.Lastly, our study is among the first to explore how managers’ emotional expressions when reacting to employee voice affect employees’ attitudes and future voice behavior. While prior research has predominantly focused on how managers’61 or peers’62 emotional states influence employee voice, our findings shift the focus to investigating the impact of managers’ emotional displays after employees have spoken up. Importantly, although our manipulation checks for anger yielded a moderate intensity (M = 2.58 for Study 1 and M = 2.94 for Study 2 on a 5-point scale), our results indicate that even these subtle displays of anger are sufficient to disrupt the positive effects typically associated with voice endorsement. In other words, while previous studies have highlighted the detrimental effects of high levels of managerial anger43, our findings suggest that even subtle angry expressions can significantly undermine employees’ psychological safety and prevent future instances of voice behavior. This underscores the sensitivity of employees to incongruences between verbal messages (i.e., endorsement) and nonverbal emotional cues. Thus, in line with the Emotions as Social Information (EASI) model, our study expands current understanding by demonstrating that it is not only the content of managers’ responses to voice that matters, but also the emotional manner in which these responses are conveyed.Limitations and future research directionsDespite the valuable insights provided by this research, several limitations should be acknowledged, some of which may spark future research. First, because our study is composed of two experiments, which provide high internal validity and allow to draw causal inferences, they are limited in external validity. Therefore, future studies should replicate our findings in organizational settings to increase the generalizability of our results. Similarly, because of the experimental nature of our studies, we were unable to investigate alternative and potentially key contextual factors, such as the organizational culture or important characteristics of the manager-employee relationship, such as LMX. These contextual variables could make more of a difference in the relationship between managerial emotional expressions and employee outcomes. Second, in line with the previous limitation, our study focused on short-term reactions to managers’ subtle emotional displays, leaving the long-term effects of repeated exposure to such displays unexplored. As such, future studies utilizing diary methods or other types of experience sampling methodologies could investigate long term-effects of angry managers on voice. Third, the voice process in supervisor-subordinate dyads may unfold qualitatively different than in leaderless teams62,63. As such, future research may seek to investigate whether our findings can be replicated in contexts where endorsement and rejection is carried by peers with similar amounts of power, such as leaderless teams. Fourth, another limitation worth considering is the relatively small sample size of our experimental conditions in Study 2, which may have led to low statistical power, increasing the likelihood of a Type II error. Lastly, our study focused on a limited range of emotional expressions—anger and happiness—excluding other potentially relevant emotions, such as pride or disappointment, that could also influence employees’ perceptions and behavior. Therefore, future studies could expand our findings by investigating other emotional expressions that managers could display while endorsing or rejecting voice.Participants read the manager from our experimental materials said, “thanks for bringing this up,” before endorsing or rejecting their voice, which might have manipulated gratitude. Because we did not collect data on perceived gratitude, we cannot rule out this possibility, as such, this is a potential limitation of our design. That being said, we think that the manager saying thanks does not necessarily threaten our findings. Gratitude expressions are effective when recipients perceive them as genuine, and not as conforming to social norms64. We would argue that a supervisor who acknowledges the reception of an employee’s voiced idea by thanking them is more likely to be construct as a social norm, rather than a genuine expression of gratitude.Practical implications and conclusionOur research also offers recommendations for managers. Managers should be mindful of their emotional expressions (even subtle ones) while endorsing voice to avoid undermining employees’ voice attitudes and future voice behavior. To do so, organizations could provide managers with training programs to help managers develop their emotional intelligence and emotion regulation to ensure constructive emotional displays during key employee interactions. Thus, organizations could develop guidelines on how to respond to employee suggestions with either a positive or neutral emotional expression. Managers should also be aware on the potentially destructive nature of their anger. Hence, organizations can provide managers with anger management techniques or mindfulness training to help them recognize and regulate their negative emotional displays towards employees. By the same token, receiving timely feedback from employees can also help managers to recognize the impact of their emotional expression.In conclusion, our research investigated how managers’ emotional expressions during voice endorsement affects employees’ downstream attitudinal and behavioral consequences. Results from two experimental studies show that when angry managers endorse voice (relative to happy or neutral looking managers), their subtle display of anger negatively impacts employees’ psychological safety and future voice behavior, and these negative consequences are almost as negative as to those experienced by employees who got their voice rejected. We hope that our research encourages future studies exploring how managers’ emotional expressions can further affect employee voice.
Data availability
Study materials, as well as the data and syntax to replicate all of our findings are openly available on the Open Science Framework repository. Link appears in the manuscript.
ReferencesDetert, J. R. & Burris, E. R. Leadership Behavior and Employee Voice: Is the Door Really Open?. Acad. Manag. J. 50, 869–884 (2007).MATH
Google Scholar
Van Dyne, L. & LePine, J. A. Helping and voice Extra-Role behaviors: evidence of construct and predictive validity. Acad. Manag. J. 41, 108–119 (1998).MATH
Google Scholar
Frazier, M. L. & Bowler, W. M. Voice climate, supervisor undermining, and work outcomes: A Group-Level examination. J. Manage 41, 841–863 (2015).MATH
Google Scholar
Lam, C. F. & Mayer, D. M. When do employees speak up for their customers? A model of voice in a customer service context. Pers. Psychol. 67, 637–666 (2014).MATH
Google Scholar
Li, A. N. & Tangirala, S. How employees’ voice helps teams remain resilient in the face of exogenous change. J. Appl. Psychol. 107, 668–692 (2022).PubMed
MATH
Google Scholar
Guzman, F. A. & Espejo, A. Introducing changes at work: How voice behavior relates to management innovation. J. Organ. Behav. 40, 73–90 (2019).
Google Scholar
Li, A. N., Liao, H., Tangirala, S. & Firth, B. M. The content of the message matters: The differential effects of promotive and prohibitive team voice on team productivity and safety performance gains. J. Appl. Psychol. 102, 1259–1270 (2017).PubMed
Google Scholar
Liang, J., Shu, R. & Farh, C. I. C. Differential implications of team member promotive and prohibitive voice on innovation performance in research and development project teams: A dialectic perspective. J. Organ. Behav. 40, 91–104 (2019).
Google Scholar
Brykman, K. M. & Raver, J. L. Persuading managers to enact ideas in organizations: The role of voice message quality, peer endorsement, and peer opposition. J. Organiz. Behav. https://doi.org/10.1002/job.2703 (2023).Article
MATH
Google Scholar
Hamstra, M. R. W., Guzman, F. A., Qian, S., Schreurs, B. & Jawahar, I. M. Turning down employee voice with humour: A mixed blessing for employee voice resilience?. J. Occup. Organ. Psychol. https://doi.org/10.1111/joop.12530 (2024).Article
Google Scholar
King, D. D., Ryan, A. M. & Dyne, L. V. Voice resilience: Fostering future voice after non-endorsement of suggestions. J. Occup. Organ. Psych. 92, 535–565 (2019).
Google Scholar
Wu, S., Kee, D. M. H., Li, D. & Ni, D. Thanks for your recognition, boss! a study of how and when voice endorsement promotes job performance and voice. Front. Psychol. 12, 706501 (2021).PubMed
PubMed Central
Google Scholar
Huai, M. et al. Does voice endorsement by supervisors enhance or constrain voicer’s personal initiative? countervailing effects via feeling pride and feeling envied. J. Appl. Psychol. https://doi.org/10.1037/apl0001191 (2024).Article
PubMed
Google Scholar
Burris, E. R. The risks and rewards of speaking up: Managerial responses to employee voice. Acad. Manage. J. 55, 851–875 (2012).MATH
Google Scholar
Li, J., Barnes, C. M., Yam, K. C., Guarana, C. L. & Wang, L. Do not like it when you need it the most: Examining the effect of manager ego depletion on managerial voice endorsement. J. Organ. Behav. 40, 869–882 (2019).
Google Scholar
Lam, C. F., Lee, C. & Sui, Y. Say it as it is: Consequences of voice directness, voice politeness, and voicer credibility on voice endorsement. J. Appl. Psychol. 104, 642–658 (2019).ADS
PubMed
Google Scholar
Lewis, K. M. When leaders display emotion: how followers respond to negative emotional expression of male and female leaders. J. Organ. Behav. 21, 221–234 (2000).MATH
Google Scholar
Sherf, E. N., Tangirala, S. & Venkataramani, V. Why managers do not seek voice from employees: The importance of managers’ personal control and long-term orientation. Organ. Sci. 30, 447–466 (2019).
Google Scholar
Fast, N. J., Burris, E. R. & Bartel, C. A. Managing to stay in the dark: managerial self-efficacy, ego defensiveness, and the aversion to employee voice. Acad. Manage. J. 57, 1013–1034 (2014).
Google Scholar
Isaakyan, S., Sherf, E. N., Tangirala, S. & Guenter, H. Keeping it between Us: Managerial endorsement of public versus private voice. J. Appl. Psychol. 106, 1049–1066 (2021).PubMed
Google Scholar
Tsai, M. & Li, N. P. Depletion manipulations decrease openness to dissent via increased anger. Br. J. Psychol. 111, 246–274 (2020).PubMed
MATH
Google Scholar
Nugier, A., Niedenthal, P. M., Brauer, M. & Chekroun, P. Moral and angry emotions provoked by informal social control. Cogn. Emot. 21, 1699–1720 (2007).
Google Scholar
Lee, R., van der Ellemers, N. & Scheepers, D. Mastering moral misery: Emotional and coping responses to intragroup morality (vs. competence) evaluations. Cogn. Emot. 30, 51–65 (2016).PubMed
Google Scholar
Cropanzano, R., Dasborough, M. T. & Weiss, H. M. Affective events and the development of leader-member exchange. Acad. Manag. Rev. 42, 233–258 (2017).MATH
Google Scholar
Silard, A. & Dasborough, M. T. Beyond emotion valence and arousal: A new focus on the target of leader emotion expression within leader–member dyads. J. Organ. Behav. 42, 1186–1201 (2021).
Google Scholar
Sinaceur, M. & Tiedens, L. Z. Get mad and get more than even: When and why anger expression is effective in negotiations. J. Exp. Soc. Psychol. 42, 314–322 (2006).MATH
Google Scholar
Tepper, B. J. Consequences of abusive supervision. Acad. Manag. J. 43, 178–190 (2000).MATH
Google Scholar
Sinaceur, M., Van Kleef, G. A., Neale, M. A., Adam, H. & Haag, C. Hot or Cold: Is communicating anger or threats more effective in negotiation?. J. Appl. Psychol. 96, 1018–1032 (2011).PubMed
Google Scholar
Heerdink, M. W., van Kleef, G. A., Homan, A. C. & Fischer, A. H. On the social influence of emotions in groups: Interpersonal effects of anger and happiness on conformity versus deviance. J. Pers. Soc. Psychol. 105, 262–284 (2013).PubMed
Google Scholar
Van Kleef, G. A. How Emotions regulate social life: The emotions as social information (EASI) model. Curr. Dir. Psychol. Sci. 18, 184–188 (2009).
Google Scholar
Van Kleef, G. A., De Dreu, C. K. W. & Manstead, A. S. R. The interpersonal effects of anger and happiness in negotiations. J. Pers. Soc. Psychol. 86, 57–76 (2004).PubMed
Google Scholar
Van Kleef, G. A. & Côté, S. Emotional dynamics in conflict and negotiation: Individual, dyadic, and group processes. Annu. Rev. Organ. Psychol. Organ. Behav. 5, 437–464 (2018).MATH
Google Scholar
Liang, J., Farh, C. & Farh, J. L. Psychological antecedents of promotive and prohibitive voice: A two-wave examination. Acad. Manag. J. 55, 71–92 (2012).
Google Scholar
Burris, E. R., Detert, J. R. & Chiaburu, D. S. Quitting before leaving: the mediating effects of psychological attachment and detachment on voice. J. Appl. Psychol. 93, 912–922 (2008).PubMed
Google Scholar
Klaas, B. S., Olson-Buchanan, J. B. & Ward, A. K. The determinants of alternative forms of workplace voice an integrative perspective. J. Manage 38, 314–345 (2012).MATH
Google Scholar
Grant, A. M. Rocking the boat but keeping it steady: The role of emotion regulation in employee voice. Acad. Manag. J. 56, 1703–1723 (2014).MATH
Google Scholar
Morrison, E. W. Employee voice behavior: Integration and directions for future research. Acad. Manag. Ann. 5, 373–412 (2011).MATH
Google Scholar
Edmondson, A. C. & Bransby, D. P. Psychological safety comes of age: Observed themes in an established literature. Annu. Rev. Organ. Psychol. Organ. Behav. 10, 55–78 (2023).MATH
Google Scholar
Satterstrom, P., Kerrissey, M. & DiBenigno, J. The voice cultivation process: How team members can help upward voice live on to implementation. Adm. Sci. Q. 66, 380–425 (2021).
Google Scholar
Poulton, E. C. et al. My manager endorsed my coworkers’ voice: understanding observers’ positive and negative reactions to managerial endorsement of coworker voice. J. Appl. Psychol. 109, 1250–1270 (2024).PubMed
MATH
Google Scholar
Hamstra, M. R. W., Schreurs, B., Jawahar, I. M. J., Laurijssen, L. M. & Hünermund, P. Manager narcissism and employee silence: A socio-analytic theory perspective. J. Occupat. Organ. Psychol. 94, 29–54 (2021).
Google Scholar
Anderson, C. & Brown, C. E. The functions and dysfunctions of hierarchy. Res. Organ. Behav. 30, 55–89 (2010).MATH
Google Scholar
Shao, B. & Guo, Y. More than just an angry face: A critical review and theoretical expansion of research on leader anger expression. Hum. Relat. 74, 1661–1687 (2021).MATH
Google Scholar
Fitness, J. Anger in the workplace: an emotion script approach to anger episodes between workers and their superiors, co-workers and subordinates. J. Organ. Behav. 21, 147–162 (2000).
Google Scholar
Vadera, A. K., Pratt, M. G. & Mishra, P. Constructive deviance in organizations: Integrating and moving forward. J. Manage 39, 1221–1276 (2013).MATH
Google Scholar
When Communicating Anger Pays and When It Backfires. Dijk, E. van, Kleef, G. A. van, Steinel, W. & Beest, I. van. A social functional approach to emotions in bargaining. J. Pers. Soc. Psychol. 94, 600–614 (2008).
Google Scholar
Morrison, E. W. Employee voice and silence: Taking stock a decade later. Annu. Rev. Organ. Psych. 10, 1–29 (2023).MATH
Google Scholar
Ocampo, A. C. G. et al. A multimethod investigation of the interpersonal effects of leader perfectionism and anger expression on employee psychological safety and creativity. J. Organ. Behav. https://doi.org/10.1002/job.2822 (2024).Article
MATH
Google Scholar
Van Dyne, L., Ang, S. & Botero, I. C. Conceptualizing employee silence and employee voice as multidimensional constructs*. J. Manag. Stud. 40, 1359–1392 (2003).MATH
Google Scholar
Tracy, J. L., Robins, R. W. & Schriber, R. A. Development of a FACS-verified set of basic and self-conscious emotion expressions. Emotion 9, 554–559 (2009).PubMed
MATH
Google Scholar
van Kleef, G. A. & Côté, S. The social effects of emotions. Annu. Rev. Psychol. 73, 1–30 (2021).MATH
Google Scholar
Wei, X., Zhang, Z.-X. & Chen, X.-P. I will speak up if my voice is socially desirable: A moderated mediating process of promotive versus prohibitive voice. J. Appl. Psychol. 100, 1641–1652 (2015).PubMed
MATH
Google Scholar
Shepherd, D. A., Patzelt, H. & Berry, C. M. Why didn’t you tell me? Voicing concerns over objective information about a project’s flaws. J. Manage 45, 1087–1113 (2019).
Google Scholar
Hussain, I., Shu, R., Tangirala, S. & Ekkirala, S. The voice bystander effect: How information redundancy inhibits employee voice. Acad. Manag. J. 62, 828–849 (2019).
Google Scholar
Hayes, A. F. & Preacher, K. J. Statistical mediation analysis with a multicategorical independent variable. Br. J. Math. Stat. Psychol. 67, 451–470 (2014).MathSciNet
PubMed
MATH
Google Scholar
Shrout, P. E. & Bolger, N. Mediation in experimental and nonexperimental studies: New procedures and recommendations. Psychol. Methods 7, 422–445 (2002).PubMed
MATH
Google Scholar
Tangirala, S. & Ramanujam, R. Ask and you shall hear (but not always): Examining the relationship between manager consultation and employee voice. Pers. Psychol. 65, 251–282 (2012).
Google Scholar
Parker, S. K., Bindl, U. K. & Strauss, K. Making things happen: A model of proactive motivation. J. Manage 36, 827–856 (2010).
Google Scholar
Warr, P. A conceptual framework for the study of work and mental health. Work Stress 8, 84–97 (1994).MATH
Google Scholar
Johnson, G. & Connelly, S. Negative emotions in informal feedback: The benefits of disappointment and drawbacks of anger. Hum. Relat. 67, 1265–1290 (2014).MATH
Google Scholar
Liu, W., Song, Z., Li, X. & Liao, Z. Why and When Leaders’ Affective States Influence Employee Upward Voice. Acad. Manag. J. 60, 238–263 (2017).MATH
Google Scholar
Liu, W. et al. How and when peers’ positive mood influences employees’ voice. J. Appl. Psychol. 100, 976–989 (2015).PubMed
MATH
Google Scholar
Guzman, F. A. & Reiche, B. S. A chorus of different tongues: Official corporate language fluency and informal influence in multinational teams. Organ. Behav. Hum. Decis. Process. 182, 104334 (2024).
Google Scholar
Silver, I., Kelly, B. A. & Small, D. A. Selfless first movers and self-interested followers: Order of entry signals purity of motive in pursuit of the greater good. J. Consum. Psychol. 31, 501–517 (2021).
Google Scholar
Download referencesAuthor informationAuthors and AffiliationsIESEG School of Management, UMR 9221 - LEM - Lille Économie Management, 3 Rue de La Digue, 59000, Lille, FranceFelipe A. Guzman & Melvyn R. W. HamstraUniv. Lille, UMR 9221 - LEM - Lille Économie Management, 59000, Lille, FranceFelipe A. Guzman & Melvyn R. W. HamstraCNRS, UMR 9221 - LEM - Lille Économie Management, 59000, Lille, FranceFelipe A. Guzman & Melvyn R. W. HamstraDepartment of Management and Humanities, EDHEC Business School, 59057, Roubaix, FranceAntonella Tempesta FernándezAuthorsFelipe A. GuzmanView author publicationsYou can also search for this author in
PubMed Google ScholarAntonella Tempesta FernándezView author publicationsYou can also search for this author in
PubMed Google ScholarMelvyn R. W. HamstraView author publicationsYou can also search for this author in
PubMed Google ScholarContributionsFG: Writing—Original draft, writing—review and editing, software, conceptualization, methodology, formal analysis, visualization, data curation; AT: writing—review and editing, conceptualization, methodology; MH: writing—review and editing.Corresponding authorCorrespondence to
Felipe A. Guzman.Ethics declarations
Competing interests
The authors declare no competing interests.
Additional informationPublisher’s noteSpringer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.Rights and permissions
Open Access This article is licensed under a Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International License, which permits any non-commercial use, sharing, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if you modified the licensed material. You do not have permission under this licence to share adapted material derived from this article or parts of it. The images or other third party material in this article are included in the article’s Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by-nc-nd/4.0/.
Reprints and permissionsAbout this articleCite this articleGuzman, F.A., Tempesta Fernández, A. & Hamstra, M.R.W. Endorsing voice behavior can be as devastating as rejecting it when the endorsement is laced with a hint of anger.
Sci Rep 15, 8454 (2025). https://doi.org/10.1038/s41598-025-93636-8Download citationReceived: 18 September 2024Accepted: 07 March 2025Published: 12 March 2025DOI: https://doi.org/10.1038/s41598-025-93636-8Share this articleAnyone you share the following link with will be able to read this content:Get shareable linkSorry, a shareable link is not currently available for this article.Copy to clipboard
Provided by the Springer Nature SharedIt content-sharing initiative
KeywordsVoice behaviorVoice endorsementVoice rejectionAngerEASI