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

Narcissism, characterized by grandiosity, self-love, and dominance, has become increasingly prevalent in modern workplaces. While often associated with negative consequences like abusive leadership and counterproductive behavior, recent research suggests narcissism may also have positive aspects, such as innovation and whistle-blowing. However, the role of narcissism in employee voice—the expression of ideas and opinions to improve the organization—remains under-explored. This study addresses this gap by investigating the impact of narcissism on both constructive (positive, helpful suggestions) and destructive (critical, harmful comments) voice. The study draws on self-verification theory, suggesting that narcissists engage in behaviors confirming their self-concept, and social identity theory, which posits that organizational identification influences behavior. The study hypothesizes that employee narcissism indirectly influences both constructive and destructive voice through perceived overqualification (the feeling of having more skills than the job requires), with organizational identification moderating this relationship.

Existing literature on narcissism highlights its dual nature, encompassing both positive and negative characteristics. Research on employee voice has primarily focused on positive forms, neglecting the potentially destructive aspects. This study bridges this gap by exploring the impact of narcissism on both constructive and destructive voice, drawing on the NARC model (Back et al., 2013) which distinguishes between narcissistic admiration and rivalry. Self-verification theory provides a framework for understanding how narcissists seek to confirm their inflated self-views through their behavior, while social identity theory explains how organizational identification moderates the expression of these behaviors. Studies showing the link between narcissism and negative workplace behaviors, and the relatively unexplored area of narcissism and voice behaviors, are reviewed to establish the theoretical foundation for the study.

A three-wave longitudinal study was conducted with 450 employees from a high-tech company in China. Data were collected over three weeks using questionnaires. Employee narcissism was measured using a 16-item scale (Ames et al., 2006), perceived overqualification using a 9-item scale (Maynard et al., 2006), constructive and destructive voice using 5-item scales (Maynes and Podsakoff, 2014), and organizational identification using a 5-item scale (Smidts et al., 2001). Control variables included gender, age, education, and tenure. Confirmatory factor analysis (CFA) was used to assess the measurement model's fit, and convergent and discriminant validity were examined. Moderated mediation analysis (PROCESS, Model 4 and 7; Hayes, 2013) was employed to test the hypotheses, examining the indirect effects of narcissism on voice through perceived overqualification, with organizational identification as a moderator.

CFA results supported the five-factor model, demonstrating good fit and adequate convergent and discriminant validity. A significant positive correlation was found between employee narcissism and perceived overqualification (r = 0.47, p < 0.01), and between narcissism and destructive voice (r = 0.45, p < 0.01). Perceived overqualification also positively correlated with destructive voice (r = 0.50, p < 0.01). Hypothesis testing revealed a significant indirect effect of narcissism on destructive voice through perceived overqualification (indirect effect = 0.24, SE = 0.05, 95% CI [0.13, 0.36]). Organizational identification significantly moderated the relationship between perceived overqualification and both constructive and destructive voice. At high levels of organizational identification, the relationship between perceived overqualification and constructive voice was positive (b = 0.24, p < 0.05), while the relationship with destructive voice was weaker compared to low organizational identification. Moderated mediation analysis indicated a stronger indirect relationship between narcissism and destructive voice at high levels of organizational identification (Estimate = 0.67, 95% CI [0.522, 0.831]) than at low levels (Estimate = 0.39, 95% CI [0.268, 0.524]). The hypothesis regarding the relationship between narcissism and constructive voice was not supported.

The findings support the hypothesized mediating role of perceived overqualification in the relationship between narcissism and destructive voice, consistent with self-verification theory. Narcissists, with their inflated self-views, may perceive themselves as overqualified, leading them to express negative views to validate their self-concept. The moderating role of organizational identification suggests that a strong sense of belonging can mitigate the negative effects of narcissism by channeling overqualified employees' energy into constructive rather than destructive behaviors. The lack of support for the relationship between narcissism and constructive voice may be due to narcissists' tendency to express their self-importance more readily through criticism than constructive suggestions. These findings provide a nuanced perspective on narcissism, suggesting it is not simply a 'bad apple' but a complex trait whose impact is shaped by both individual perceptions and organizational context.

This study contributes to the literature by examining the dual impact of narcissism on employee voice, highlighting the mediating role of perceived overqualification and the moderating effect of organizational identification. The findings have practical implications for managing narcissistic employees, suggesting the importance of careful recruitment, addressing perceived overqualification, and fostering strong organizational identification. Future research could explore the causal relationships using experimental designs, investigate the role of social exchange, and examine cultural variations in these relationships.

The study's cross-sectional design limits causal inferences. The sample was drawn from a single company in China, potentially limiting generalizability to other contexts. Future studies could address these limitations by employing longitudinal designs, examining various organizational cultures, and utilizing diverse methodologies such as experiments and qualitative methods.