1. Introduction

More and more Generation Z employees dominate human resources in today's workplace [1]. As they grew up in an era of rapid information technology development and greater openness, the new employees are bolder, more passionate about personal values and have higher career expectations than their predecessors [1]. Thus, most Chinese business leaders differ from younger employees in mindset, behaviour and motivation. Leadership is an essential element that influences employees' sense of belonging, loyalty and embeddedness in the organization [2]. When leaders with dark personalities coordinate and plan organizational matters and employee work, such leaders could negatively impact employee attitudes and retention intentions [3]. The narcissistic personality is a trait more commonly found in top managers, which triggers low job satisfaction, reticence and turnover intentions in subordinates [4,5]. Narcissistic leaders are also more likely to be found in cultures with strong collectivist tendencies and high-power distance, such as China [6,7]. Therefore, this study will focus on the impact of narcissistic leadership on employee embeddedness in the Chinese context.

Over the years, job embeddedness has received a lot of attention from scholars to mitigate employee turnover. [8-10]. Previous studies have explored the impact of employment relationships, organizational commitment, and environmental factors in the talent market on employees' job embeddedness [11]. Less literature has focused on the influence of leadership on job embeddedness. Therefore, this study aims to enrich the existing literature by constructing a new theoretical model within the Chinese setting. Moreover, this study expands the research perspective by introducing employee stress and employee silence to analyse their mediating role. To achieve these objectives, this study uses a quantitative method of publishing an online questionnaire and analyzing the research data through SPSS.

2. Literature Review

2.1. New Generation Employee

New-generation employees are defined as human resources born after 1990, aged between 18 and 35, who have entered the workplace [12]. They are bold, open, and challenging authority [1]. They are highly educated, influenced by multiple values and can bring more information resources and innovative solutions to the organization [13,14]. However, they are keen on acquiring new knowledge and upgrading existing skills, have more career choices in the labour market, and have lower organizational loyalty, leading to floating job embeddedness and more likely to change jobs frequently [15,16]. Therefore, clarifying the factors that influence job embeddedness and reducing the turnover rate of new-generation employees become a topical issue for businesses to address.

2.2. Job Embeddedness

The concept of job embeddedness connects employees to their social networks and is a combined force that prevents employees from quitting and reduces turnover costs [17-19]. This interconnectedness can be explained by three dimensions: (a) Links, which is the informal or formal connection between an individual and their organization or individuals not belonging to the organization, such as co-workers, clients, family and friends; (b) Fit, which represents employees' feelings of comfort and compatibility with the organizational environment; (c) Sacrifice, which refers to the perceived cost of tangible and intangible benefits that employees lose when they leave the organization, including status, prestige, compensation, benefits and community clubs [8,20]. These three dimensions of job embeddedness cover both work and life aspects, that is, the attractiveness of keeping an employee in a position and the personal or community life aspects that enable the employee to remain geographically stable [21].

2.3. Narcissistic Leadership and Employee Job Embeddedness

The narcissistic trait is characterized by arrogance, hubris, aggression, lack of empathy, fragile self-esteem, and hostility towards others [22,23]. This personality usually focuses on success, competition, extreme pursuit of power and admiration [4,24]. Narcissistic leaders fit the profile of great leaders because they desire success and can provide a compelling vision for their followers [25-27]. However, narcissistic leaders' leadership beliefs and behaviors are often motivated by a desire for power rather than emotional caring for the organization and subordinates [5]. This means that narcissistic leaders are more inclined to make decisions in their self-interest rather than consider the needs and benefits of others [3,28]. Leaders with such personalities can exhibit excessively self-centred behaviors, manipulation, exploitation of others, and a dislike for accepting counter-arguments or reverse feedback from subordinates [22,29]. Thus, narcissistic leaders establish a competitive and unreliable organisational culture, ultimately hurting their followers’ career development and quality of work [30,31].

It has been shown that leadership traits are significantly associated with subordinates' job embeddedness [32,33]. Based on the social exchange theory, narcissistic leadership styles are more likely to engage in negative managerial behaviors, assert their interests and authority through decision-making, break the positive reciprocal relationship between employers and employees, and make employees feel disrespected or exploited [34-36]. Besides, narcissistic leaders’ hypersensitivity leads to paranoia and lack of trust in subordinates, seriously weakening employees’ perceptions of organizational compatibility and leadership support [1]. Therefore, narcissistic leaders can lead to detrimental behavioral outcomes such as increased emotional exhaustion and decreased loyalty and trust in the organization, which reduces employees' job embeddedness. Accordingly, this study hypotheses that:

H1: Narcissistic leaders could negatively impact an employee's sense of job embeddedness.

2.4. The Mediating Effect of Job Stress

Job stress is influenced by factors such as the job itself, the individual's role in the organization, career progression, and the interrelationships between supervisors and subordinates [37,38]. Referring to the conservation of resources theory (COR), people have difficulty acquiring valuable resources and are sensitive to the possibility of losing such resources [39]. In organizations, leaders are usually in charge of some of the scarce resources [38]. However, selfish leaders do not empathize with subordinates who seek or protect resources [40]. Narcissistic leaders set goals based on their blindly arrogant performance assumptions instead of reality [38]. Since employees are performers, they are left to dedicate many resources to accomplish unrealistic goals, resulting in a significant drain on employee resources to cope with work assignments [41]. Besides, the abusive leadership style of narcissistic leaders can seriously undermine employees' self-resources, such as attention, willpower, and self-esteem, which ultimately leads to increased work stress [1]. Therefore, this study proposes a hypothesis as follows:

H2: Narcissistic leadership positively relates to employee job stress.

Traits such as the self-importance of narcissistic leaders attribute the resources and success of the organization to themselves, making it exceptionally challenging for employees to access resources and retain them without depletion [42,43]. Employees who experience significant work stress due to the exhaustion of resources are often tempted to protect those resources at all costs, including forgoing connections with their leaders [44,45]. Thus, employees ultimately choose to lower their job embeddedness or leave the organization to preserve and protect themselves from resource loss. Under the management of a narcissistic leader, work resources, as well as employees' self-resources, such as their emotional and respect needs, are significantly depleted, exacerbating job stress, which directly manifests itself in detachment from the organization and reduced intrinsic connectedness and integration with the organization [1,46]. Therefore, this study proposes the following hypothesis:

H3: Job stress as a mediator between narcissistic leadership and new-generation employees’ job embeddedness.

2.5. The Mediating Effect of Employee Silence

Employee silence is the reluctance of employees to express their opinions, consciously withholding vital information or avoiding providing advice [47,48]. Existing research reports that narcissistic leaders are negatively associated with employee vocal behaviour [42]. Narcissistic leaders' aggressive speech, abusive supervision, disrespect and other uncivilized work behaviors can largely deplete employees' emotional resources [49,50]. Based on COR theory, employees may choose avoidant coping strategies to preserve their emotional resources by distancing themselves from uncivil actors [51,52]. Moreover, employees' motivation to remain silent often relates to their belief that direct vocal behaviour will not change their work's content and status quo [48]. Therefore, subordinates under the leadership of narcissists may remain silent to prevent further incidents of abuse. With this in mind, this study proposes the following hypothesis:

H4: Narcissistic leadership is positively related to employee silence.

High silent behaviour triggered by narcissistic leaders reduces job embeddedness as employees become less connected to their leaders [51]. Employees also gradually lose interest in building social relationships at work, reducing social cohesion and adopting interaction avoidance [53]. Thus, employee silencing negatively influences the linkage dimension of job embeddedness. Besides, individuals who experience uncivil work behaviors from narcissistic leaders question whether they ‘fit in’ with the current organizational environment and weigh up whether the high sacrifices they make on the job are worth staying in the organization [51]. There is also a risk and concern that the identification of silent behaviour will lead to more serious abusive behaviour and repression [51]. Thus, employee silence exhibited under the management of a narcissistic leader reduces the employee's fitness for the organization and intensifies the employee's sense of sacrifice, which greatly reduces job embeddedness and ultimately has the potential to result in mass turnover. Conclusively, this study hypotheses that:

H5: Employee silence mediates the effect of narcissistic leadership on new-generation employees’ job embeddedness.

In conclusion, Figure 1 shows the research model constructed from the literature review for this study.

Figure 1: Research model.

3. Methodology

This cross-sectional study used a quantitative research methodology by publishing an online questionnaire to collect data from a random sample. A validated multi-item 5-point Likert scale was primarily used in this study to ensure consistency in question structure and scale validity. This study referenced the items developed by Hochwarter & Thompson to measure Narcissistic Leadership (NL) [54]. Job embeddedness (JE) was referenced in Crossley et al.'s [18] seven-item overall job embeddedness scale. Job Stress (JS) was measured through the stress scale of The American Institute of Stress [55]. Employee Silence (ES) was assessed using a five-item scale adapted from Tangirala & Ramanujam [56]. The questionnaire was distributed mainly through social media. Quantitative research follows pre-determined structured questions and, therefore, may not be flexible but contributes to the standardization and cost-effectiveness of the research process as it only takes 3-5 minutes to complete. The collected questionnaire data was first exported to SPSS software and decoded. Descriptive analyses of the sample characteristics were conducted in this study, and Cronbach's alpha and factor analysis were used to test the reliability and validity of the scales [57]. Besides, data analysis used correlation analysis and the Process model to examine influential relationships between variables [58]. The SPSS method of data analysis is a manually done operation and is relatively time-consuming, but it is relatively easy to master and produce visualized data.

4. Findings and Discussion

4.1. Sample Analysis

From 12th September to 15th September 2024, 201 valid questionnaires were received. Table 1 shows the profile of the sample. The sample was 50.2% males and 49.8% females, a relatively balanced gender distribution. The participants of this study were mainly aged 22-30 (56.2%) and 31-35 (31.3%). The education level of the sample was divided into college (40.3%) and bachelor’s degree (51.2%). Most participants had 3-5 years of working experience (56.2%). Overall, the characteristics of the sample collected were relatively consistent with the expectations of the target population of this study.

Table 1: The basic description of samples (n=201).

4.2. Reliability Analysis

This study checked the reliability of all the scales in the questionnaire and the degree of consistency of their dimensional measurements [59]. Cronbach's alpha above 0.7 means the scale is reliable [57]. Table 2 shows the results of the reliability analyses. The reliability value of the overall scale was 0.783, and the alpha of the scale items of the four dimensions was also around 0.9 and above. Therefore, the data collected based on the scales have better reliability.

Table 2: Reliability statistics.

4.3. Validity Analysis

Exploratory factor analysis in SPSS was adopted in this study to test the validity of the collected scale data and the reasonableness of the internal structural division of the questionnaire. Table 3 depicts the results. Firstly, the data of this study is suitable for exploratory factor analysis because the KMO is 0.924>0.7, and the significance of Bartlett's test of dispersion is 0.001<0.05 [59]. Besides, the results in Table 3 demonstrate that the cumulative variance explained by the factors extracted using principal component analysis is 67.875%, which exceeds the 60% criterion [60]. Therefore, the data collected in this study can effectively and completely restore the original data. Moreover, the rotated factor matrix derived from Varimax rotation by Kaiser's normalization method showed that a total of four factors were extracted, with each measure loading above 0.5, implying that the composition of the factors was consistent with the dimensionality of the scale and that no more than one variable shared a factor [61,62]. Therefore, the questionnaire data was valid.

Table 3: Exploratory factor analysis.

4.4. Correlation Analysis

Table 4 demonstrates the results of the correlation analysis. Narcissistic leadership (NL) was significantly negatively correlated with employee job embeddedness (JE) (r=-0.412, p<0.01). There was a significant positive correlation between narcissistic leadership and employee stress (JS) (r=0.404, p<0.01) as well as employee silence (ES) (r=0.337, p<0.01). There was a significant negative correlation between employee stress and employee embeddedness (r=-0.413, p<0.01) and between employee silence and job embeddedness (r=-0.403, p<0.01). Thus, the results of the correlation analyses provide an initial validation to test the research hypotheses.

Table 4: Correlation analysis.

Note: ** P<0.01, * P<0.05

4.5. Regression Analysis

This study first tested the regression model for multicollinearity through the Variance Inflation Factor (VIF) [62] before conducting the regression analysis. The results in Table 5 show that the VIF values of NL, JE and JS are all around 1.0-1.3, indicating no multicollinearity among these variables. Therefore, the correlation between variables is unlikely to affect the accuracy of the regression model.

Table 5: Multicollinearity test.

4.5.1. Regression Model Mediated by Job Stress

The study analyzed data from two mediation regression models using SPSS process model 4 [58]. The first model was based on narcissistic leadership as the independent variable, job stress as the mediating variable, and employee job embeddedness as the dependent variable. The results are shown in Table 6. The results show that narcissistic leadership had a significant positive effect on job stress (b=0.379, p=0.000), which verifies Hypothesis 2. Narcissistic leadership significantly negatively affects employees' job embeddedness (b=-0.386, p<0.001), supporting Hypothesis 1. Besides, Table 6 shows a significant negative effect of job stress on employees' job embeddedness (b=-0.315, p<0.001). In this case, the regression coefficient of narcissistic leadership on job embeddedness becomes b=-0.267 (p<0.001), influenced by the variable of job stress. Therefore, this study speculated that job stress has a mediating effect on the model.

Table 6: Mediation regression analysis.

Note: *p<0.05, **p<0.01, ***p<0.001

Table 7 shows the total, direct, and mediating effects from the coefficient product method to calculate the model and bootstrap sampling 5000 times for bias correction [58]. The total and direct effects of the process of narcissistic leaders' influencing job embeddedness were statistically significant, as none of the 95% confidence intervals were zero. The narcissistic leader's indirect effect on job embeddedness through job stress was -0.128, and the 95% confidence interval [-0.197 -0.067] also did not contain zero. Therefore, the impact of narcissistic leadership on job embeddedness was partly realized by exacerbating employees' job stress, with a mediating effect of 33.2%, validating Hypothesis 3.

Table 7: Mediation effects of job stress.

4.5.2. Regression Model Mediated by Employee Silence

The second regression mediation model was with employee silence as the mediating variable. The results in Table 8 show a significant positive regression effect of narcissistic leadership on employee silence (b=0.330, p<0.001), and Hypothesis 4 was established. Employee silence significantly negatively affected employee job embeddedness (b=-0.286, p<0.001). The regression coefficient of narcissistic leadership on job embeddedness was affected by employee silence and became b=-0.292 (p<0.001). Therefore, the research assumed that there was also a significant mediation effect in this model. Table 9 shows the results of the mediation effects analysis. The results show that the indirect impact of employee silence is significant because the 95% confidence intervals exclude zero. Thus, the mediating effect of employee silence was 24.4%, which tested Hypothesis 5.

Table 8: Mediation regression analysis.

Table 9: Mediation effects of employee silence.

5. Conclusion

Narcissistic leadership is leadership behaviour driven primarily by self-interested needs and arrogant beliefs. In this study, 201 new-generation Chinese employees were used as the research sample, and the model of narcissistic leadership's influence on employees' job embeddedness was constructed by combining factors such as job stress and employee silence. The research findings prove that narcissistic leadership negatively affects job embeddedness and that job stress and employee silence mediate the relationship. Although the study is cross-sectional and may not be able to track the trend of the variables of interest over time, the study has several contributions. In terms of theoretical significance, this study explores the impact of leaders’ narcissistic personalities on employees’ job embeddedness, enriching the research on leadership and employee behaviour in different contexts. Besides, this study references the conservation theory of resources and the social exchange theory to provide robust theoretical bases for the findings. This study also constructs a new research model that reveals the complex effects of job stress and employee silence on affective relationships. On practical implications, the findings of this study serve as a warning to organizations that they need to take timely steps to mitigate and constrain the dark influence of narcissistic leaders and motivate them to put the collective good above the individual good. Besides, HR practitioners need to communicate regularly with employees and help leaders coordinate their relationships to reduce employee stress and encourage vocal behaviour.