Good soldiers or bad apples? Exploring the impact of employee narcissism on constructive and destructive voice

The study addresses how employee narcissism translates into different forms of challenge-oriented voice in organizations, specifically constructive versus destructive voice. Given rising levels of narcissism in the workforce and mixed evidence of its bright and dark sides, the paper asks whether narcissistic employees are more likely to express constructive suggestions or destructive criticism. Grounded in self-verification theory, the authors propose that narcissists’ inflated self-views lead to perceived overqualification, which in turn influences their voice behavior. The study further examines whether organizational identification, drawing on social identity theory, conditions these effects by steering overqualified employees toward constructive rather than destructive voice. The work formulates hypotheses that narcissism relates to both forms of voice, that perceived overqualification mediates these links, and that organizational identification moderates the relationships between perceived overqualification and both constructive and destructive voice, as well as the corresponding indirect effects.

The paper builds on research showing narcissism’s duality in organizations, linking it to both harmful outcomes (e.g., abusive leadership, counterproductive behavior) and potentially positive behaviors (e.g., innovation, prosocial acts, whistleblowing). Voice is conceptualized beyond its traditional positive framing to include constructive and destructive challenge voice (Maynes and Podsakoff, 2014). Self-verification theory suggests individuals enact behaviors that confirm their self-concept; for narcissists, grandiose self-views may foster perceived overqualification. Overqualification can prompt constructive voice through extra capacity and motivation, but also destructive voice via frustration and relative deprivation. Organizational identification is posited to channel overqualified employees’ expression toward organization-serving behaviors. Hypotheses: H1 narcissism → constructive voice (positive); H2 narcissism → destructive voice (positive); H3 narcissism → perceived overqualification (positive); H4 perceived overqualification → constructive voice (positive); H5 perceived overqualification → destructive voice (positive); H6/H7 mediation of narcissism → voice via perceived overqualification; H8/H9 moderation of perceived overqualification → destructive/constructive voice by organizational identification; H10/H11 moderated mediation of narcissism → voice via perceived overqualification by organizational identification.

Design and sample: A multiwave, three-wave longitudinal field study with leader–follower dyads was conducted in a high-tech company in Chengdu and Chongqing, China (March–July 2022). Time 1: employees reported demographics and narcissism (712 usable responses from 810 invited). Time 2 (3 weeks later): employees reported perceived overqualification and organizational identification (587 responses). Time 3 (3 weeks later): immediate leaders rated employees’ constructive and destructive voice over the prior month. The final matched dataset comprised 450 valid surveys. Participants received ~RMB 30 for participation. Ethical approval was obtained; informed consent and confidentiality were assured. Measures: All non-narcissism measures used 1–5 Likert scales (1 = strongly disagree, 5 = strongly agree). Translation/back-translation followed Brislin (1980).

• Employee narcissism: NPI-16 (Ames et al., 2006), sample item: “I truly like to be the centre of attention.” Cronbach’s alpha = 0.86.
• Perceived overqualification: 9-item scale (Maynard et al., 2006), sample item: “I have more abilities than I need in order to do my job.” Alpha = 0.85.
• Constructive voice: 5-item scale (Maynes & Podsakoff, 2014), sample item: “I frequently make suggestions about how to do things in new or more effective ways at work.” Alpha = 0.88.
• Destructive voice: 5-item scale (Maynes & Podsakoff, 2014), sample item: “I frequently make overly critical comments regarding how things are done in my organization.” Alpha = 0.84.
• Organizational identification: 5-item scale (Smidts et al., 2001), sample item: “I experience a strong sense of belonging to this organization.” Alpha = 0.87. Controls: Gender, age, education, and tenure. Analytic strategy: Confirmatory factor analysis (CFA) assessed measurement validity. Composite reliability (CR) and average variance extracted (AVE) examined convergent validity. Hypotheses were tested via regression analyses and PROCESS macro: mediation (Model 4), moderation (Model 7), and moderated mediation (Model 14) with 5000 bootstrap samples. Variables were mean-centered for interaction tests. Sample characteristics: Ages 20–55 (SD = 1.19 as reported); 51.3% women; 60.2% held a bachelor’s degree; 47.6% tenure >3 years.

Measurement model: Five-factor CFA (narcissism, perceived overqualification, organizational identification, constructive voice, destructive voice) fit well and better than alternatives: χ² = 1492.81, df = 568, CFI = 0.93, TLI = 0.92, IFI = 0.93, NFI = 0.91, RMSEA = 0.059. Convergent/discriminant validity was acceptable (e.g., CR: 0.86–0.93; AVE: 0.52–0.74). Descriptive correlations: Narcissism correlated positively with perceived overqualification (r = 0.47, p < 0.01) and destructive voice (r = 0.45, p < 0.01). Perceived overqualification correlated positively with destructive voice (r = 0.50, p < 0.01). Direct and mediated effects: Narcissism predicted perceived overqualification (b = 0.47, p < 0.001) and destructive voice (b = 0.44, p < 0.001). Perceived overqualification predicted destructive voice (b = 0.36, p < 0.001). The indirect effect of narcissism on destructive voice via perceived overqualification was significant (indirect effect = 0.24, SE = 0.05, 95% CI [0.13, 0.36]; mediation effect size ≈ 0.22). The hypothesized positive effect of narcissism on constructive voice (H1) was not supported. Moderation by organizational identification: The interaction between perceived overqualification and organizational identification was significant for constructive voice (b = 0.60, p < 0.05) and for destructive voice (b = −0.84, p < 0.001). Simple slopes showed that perceived overqualification related positively to constructive voice at high organizational identification (b = 0.24, SE = 0.10, p < 0.05) but was not significant at low identification. For destructive voice, perceived overqualification related positively at both levels but more strongly at low identification (b = 0.85, SE = 0.10, p < 0.001) than at high identification (b = 0.53, SE = 0.06, p < 0.001), indicating organizational identification weakens the POQ–destructive voice link and strengthens the POQ–constructive voice link. Moderated mediation: Conditional indirect effects of narcissism on destructive voice via perceived overqualification were stronger at low organizational identification (Effect = 0.677, Boot SE = 0.078, 95% CI [0.522, 0.831]) than at high identification (Effect = 0.396, Boot SE = 0.065, 95% CI [0.268, 0.524]). Conditional indirect effects on constructive voice were not significant (Low ID 95% CI [−0.516, 0.035]; High ID 95% CI [−0.052, 0.404]). Overall, H2, H3, H5, and H7 were supported; moderation patterns were consistent with H8 and H9; H11 was not supported.

The findings indicate that employee narcissism functions as a driver of destructive voice in organizations, and this effect operates through perceived overqualification. This supports a self-verification perspective in which narcissistic employees, motivated to affirm grandiose self-views, perceive themselves as overqualified and use voice to validate those views, often in destructive ways. Organizational identification acts as a contextual boundary condition, channeling overqualified employees toward more constructive forms of voice and dampening destructive expressions. The absence of support for a positive direct link from narcissism to constructive voice suggests that narcissists’ entitlement and grandiosity may be more readily expressed through critical or demeaning commentary rather than constructive suggestions. The results enrich dark personality and voice literatures by clarifying when and how narcissism translates into different voice forms and by identifying perceived overqualification as a key psychological mechanism moderated by organizational identification.

This study advances understanding of narcissism’s double-edged implications for employee voice. It demonstrates that narcissism promotes destructive voice via perceived overqualification and that organizational identification can attenuate the pathway to destructive voice while facilitating constructive voice from overqualified employees. The work offers a holistic view of employee narcissism by examining both constructive and destructive challenge voice, introduces perceived overqualification as a self-verification-based mediator linking narcissism to voice, and identifies organizational identification as a situational moderator. Future research should employ experimental, longitudinal panel, and experience sampling designs to establish causality, explore additional mechanisms (e.g., social exchange processes), and test generalizability across cultures and organizational contexts.

The three-wave design cannot conclusively establish causal direction. The study was conducted in a single Chinese company, limiting generalizability, particularly given cultural features (collectivism, high power distance). Additional theoretical lenses (e.g., social exchange) and diverse samples and methods (qualitative, experiments, ESM) are recommended. Some reported statistics and table narratives show inconsistencies, and the study relies on self-reports for some constructs and leader ratings for outcomes, which may introduce measurement limitations.