The relationship between work overload, narcissism, and work engagement in Korean firms: the moderated mediation of autonomy, grit, and stress

In dynamic contemporary workplaces, employees are expected to cope with persistently changing circumstances without suffering any decrease in effectiveness in their work. In these uncertain environments, work engagement has emerged as a key indicator of employee well-being and organizational success, and is widely recognized as a critical factor for positive organizational outcomes, such as innovation and employee effectiveness (Bakker et al., 2023; Keane et al., 2024). Whereas traditionally much focus has been placed on positive antecedents of work engagement, Job Demands-Resources (JD-R) theory explicitly acknowledges the role of negative antecedents in the form of hindrance demands (Bakker and Demerouti, 2017; Keane et al., 2024), and it is recognized that heightened demands profoundly impact employees' workplace behaviors and attitudes (Bakker and Demerouti, 2017). Despite this recognition, however, these demands are often viewed as being specifically job-related, while the role of personal demands remains relatively underexplored (Zeijen et al., 2021).

Narcissism represents one such negative antecedent of work engagement. Characterized by inflated self-importance and need for admiration, it is a normally-distributed trait within the population (Campbell et al., 2004, 2010). Within the JD-R framework, personal traits represent personal resources when they support goal attainment and well-being, thereby contributing to the motivational process (Bakker and Demerouti, 2017). In contrast with this, narcissism is associated with unrealistic goals, interpersonal conflict, and stress which influence the health-impairment process (Bakker and Demerouti, 2017; Campbell et al., 2004, 2010). As such, the current research conceptualizes trait narcissism as a personal demand which hinders individuals at work.

The mechanism through which narcissism, an internal employee trait, affects employee attitudes and behaviors has yet to be fully articulated (Andreassen et al., 2012; Biolik, 2025; Falco et al., 2020). Meanwhile, work overload, which is characterized by excessive job demands and unrealistic performance expectations, occurs externally to the individual and has been linked to adverse outcomes, including burnout and diminished job satisfaction (Brown and Benson, 2005). By examining narcissism and work overload together, we present a more holistic perspective of demand facets within JD-R theory than is typically seen.

To elaborate on the health-impairment process, the negative effects of narcissism and work overload contribute to job stress (Brown and Benson, 2005; Schwarzkopf et al., 2016). According to JD-R theory, job stress, defined as the harmful physical and emotional responses that occur when demands exceed workers’ resources, could negatively mediate the relationship between narcissism and work overload, and engagement (Bakker and Demerouti, 2017; Lazarus and Folkman, 1984). An understanding of how job stress channels the impact of work overload and narcissism on work engagement will help in developing comprehensive strategies to foster a healthier, more engaged workforce.

Beyond this, job autonomy and grit represent potentially moderating factors in our framework. Job autonomy represents the degree to which employees have control over their work and may ameliorate negative impacts of precursors, in this case narcissism and work overload, to job stress (Hackman and Oldham, 1975). Similarly, grit–the perseverance and passion for long-term goals–can enhance resilience towards stress triggers, thereby protecting engagement levels over sustained periods (Duckworth, 2016).

Research related to work engagement has identified relationships between variables based on JD-R theory, but several limitations and gaps exist. First, job demands and resources are well-documented and viewed as competing energies which impact the degree to which an employee can successfully fulfil their job roles while workers can also employ personal resources to exert control over their environment (Bakker et al., 2023). However, research examining personal demands tends to align with Barbier et al.'s (2013) initial narrow definition of the concept as “the requirements that individuals set for their own performance and behavior” (p.751). Narcissism, as a personality trait, cannot be defined as a self-imposed demand, but rather as a characteristic that is inherent to the individual (Miller et al., 2021). By examining narcissism as a personal demand, the current research expands the personal demands concept to become more encompassing.

Second, existing studies often focus on measuring the positive predictors of work engagement while the negative variables that impact it have gained less attention or are limited to fragmentary causal relationships. For example, although authors have highlighted the significance of all types of demands and resources in explaining employee well-being and engagement outcomes, their findings were based on literature reviews, and measurement models were not presented (Bakker et al., 2023; Demerouti and Bakker, 2023; Keane et al., 2024). Elsewhere, Biolik (2025) examined how personality-based demands interact with structural job characteristics to shape stress and engagement and impact organizational functioning while, in the domain of personal resources, recent work established how grit, self-efficacy, and related constructs buffer against work demands in diverse contexts (Kim and Lee, 2022; Lytle and Shin, 2023).

JD-R theory, therefore, assumes that job and personal demands and resources interact to affect stress and work engagement (Bakker and Demerouti, 2017; Barbier et al., 2013). However, previous studies do not commonly include all four components in a single model, so there is a need to empirically verify the conceptual framework of JD-R theory, which states that job, and personal, demands and resources affect the stress and work engagement of employees. The present study builds upon recent literature by simultaneously incorporating all demand and resource types into a single model.

Thus, the purpose of this study is to empirically verify how work overload and narcissism affect work engagement through job stress, and whether job autonomy and grit moderate this relationship. Our analysis contributes to the broader work engagement discourse by utilizing JD-R theory to integrate these constructs into a single model and elucidate the pathways through which job and personal demands and resources influence it.

Work engagement, rooted in positive psychology, contrasts with earlier organizational theories focused on burnout and fatigue. Kahn (1990) introduced the need-satisfying approach, suggesting employees engage fully when their needs for meaningfulness, safety, and availability are met. Subsequently, various narratives emerged including Schaufeli et al.'s (2002) conceptualisation which defines work engagement as “a positive, fulfilling, work-related state of mind that is characterized by vigor, dedication, and absorption” (pp. 74). This definition is adopted in the current research as it clearly delineates work engagement from similar concepts and is specifically work-directed.

Narcissism is commonly characterized by exaggerated self-focus, perceived superiority, and a strong need for recognition (Ames et al., 2006; Campbell et al., 2010). Studies have found mixed results when examining the effects of narcissism on work engagement. Biolik (2025) found that narcissistic admiration may produce positive outcomes, such as harmonious passion and work focus, in the short-term whereas narcissistic rivalry may induce aggressive behaviors and a struggle for domination. Other research uncovered weak (Andreassen et al., 2012) and partial (Falco et al., 2020) support for a positive relationship between narcissism and work engagement while Burawat (2023) posited the strength of the relationship may be affected by age, or the type of narcissism measured (communal vs. agentic). Recognising these mixed results, and that overall narcissism is more commonly linked to negative attitudes and behaviors such as counterproductive work behaviors (CWB), job dissatisfaction, workplace incivility, and bullying (O’Boyle et al., 2012), the current research explores the possibility of narcissism having a negative impact on work engagement.

In general, narcissistic individuals possess high self-esteem and a desire to prove their superiority to others (Campbell et al., 2010). People with high narcissistic tendencies are susceptible to delusions of grandeur that may lead them to pursue tasks that are difficult for them to handle which, in turn, can prevent them from becoming fully immersed in their work (Campbell et al., 2004; O’Boyle et al., 2012). Ultimately, this could manifest in reduced work engagement as employees focus more of their energies on dealing with hindrance job demands rather than directing them towards their work. Thus, the following hypothesis was developed.

Work overload arises when employees face excessive demands, such as tight deadlines and unmanageable workloads, which hinder their ability to meet expectations (Brown and Benson, 2005). From a JD-R perspective, it reflects a mismatch between demands and an employee’s capacity to meet them, and leads to decreased engagement, burnout, and turnover intention (Bakker and Demerouti, 2017; Bakker and Sanz-Vergel, 2013). Thus, the following hypothesis was developed.

Job stress occurs when individuals perceive a mismatch between their abilities and external demands. It incorporates work-related anxiety, tension, and exhaustion resulting from threats at work which negatively affect work atmosphere and relationships (Hunter and Thatcher, 2007; Lambert and Paoline, 2008). Both environmental factors (job demands) and individual characteristics (personal demands) have an impact on job stress. The job demand of work overload leads to stress and increases negative health-related outcomes, affective and behavioral responses to one’s work, and performance-related metrics among employees (Okuhara et al., 2021), while the personal demand of narcissism can lead to job stress due to a disconnect between an employee’s unrealistic expectations and inflated levels of self-esteem when compared with the actuality of their work situations (Schwarzkopf et al., 2016). Furthermore, job stress is commonly associated with reduced work engagement (Bakker et al., 2023; Bakker and Demerouti, 2017). Thus, the following hypotheses were developed.

JD-R theory proposes a relationship in which job demands, such as work overload, are associated with psychological and physical costs (Bakker et al., 2023; Bakker and Demerouti, 2017). When individuals cannot deal with these demands, they are viewed as hindrances and manifest in stress and reduced motivation (Bakker and Demerouti, 2017). This stress can be mitigated through the application of job resources which refer to physical, psychological, social, or organizational aspects which facilitate the attainment of work goals, lead to personal or professional growth, or reduce job demands and their negative costs (Bakker and Demerouti, 2017). As such, the negative impact of the relationships between narcissism and work overload, stress, and work engagement may be alleviated through the acquisition and utilization of job and personal resources.

One such job resource is job autonomy which refers to the freedom, discretion, and independence that allows members to make their own choices when deciding on work methods and tasks (Clausen et al., 2022) and contributes to employees’ optimism, hope, and self-efficacy. High levels of autonomy may buffer the negative effects of stress, promoting greater engagement even in the face of demanding job conditions as employees with greater autonomy are necessarily freer to explore stress-relieving methods in their pursuit of organizational goals (De Spiegelaere et al., 2014; Hämmig and Vetsch, 2021). From a JD-R perspective, an absence of rigid parameters or constraints allows employees to identify routes towards organizational goals which circumvent stress-inducing hindrance demands.

In this way, job autonomy can have a positive effect on the psychological and behavioral responses of an employee towards their work (Gözükara and Çolakoğlu, 2016) as employees may compensate for excessive barriers or demands by autonomously developing alternative pathways to fulfil their work obligations (Gordon et al., 2015; Scharp et al., 2022). Therefore, it can be predicted that job autonomy will moderate the detrimental impact of work overload and narcissism on job stress and, consequently, reduce the level of decline in work engagement.

Grit comprises persistence and passion toward long-term goals, enabling individuals to overcome setbacks and maintain effort (Duckworth, 2016). As such, it could be argued that grit can act as a compensatory mechanism whereby gritty employees push through difficult situations at work which may otherwise be stressful or debilitating (Kim and Lee, 2022; Schwepker and Good, 2002).

In certain contexts, such as the COVID-19 pandemic, members of SMEs experience more severe negative impacts when compared to members of large corporations (Kalemli-Ozcan et al., 2020). These negative impacts may be mitigated in gritty individuals, though, as they endeavour to control their emotions and make sustained efforts towards their goals (Lytle and Shin, 2023). Furthermore, gritty employees are likely to control their level of personal stress in order to achieve the goals they have set, thereby minimising the negative impact on work engagement (Suzuki et al., 2015).

This study investigates the effects of work overload and narcissism on job stress and work engagement, and examines whether job autonomy and grit moderate these effects. The conceptual model is illustrated in Figure 1.

When compared with employees in large organizations, SME workers are more affected by negative workplace outcomes in difficult contexts (Kalemli-Ozcan et al., 2020), and SMEs may not have the resources to support their workforce. This is particularly relevant in economies such as Korea where SMEs make up 99% of the number of enterprises and account for 83% of total employment (Ministry of SMEs and Startups, 2023). To facilitate focused measurement of this subpopulation, snowball sampling of SME workers in a range of industries throughout Korea was employed. Informed consent was obtained and data anonymized per Sehan University ethics policy. Based on Bentler and Chou (1987) and Kline (2023), a sample size of 390–780 was appropriate for the 78 free parameters in the model. Data were collected via Google Forms over 8 days (October 18–26, 2022), yielding 547 responses. After excluding 19 for insincerity, 528 valid cases remained (validity rate: 96.5%). 59.8% were male, and the most common age group was in their 40s (31.8%). Most had a college degree (67.6%) and were full-time employees (87.5%). Entry-level workers were the largest group (29.2%), with strategy/planning the most common job type (26.5%). 27.3% had over 21 years’ experience in total, while 46.8% had 1–5 years in their current role. Participants were deliberately drawn from a diverse range of industries, including manufacturing (21.4%), strategy and planning (26.5%), sales and marketing (19.1%), education and training (13.3%), and other industries (19.7%).

All measurement instruments utilized 5-point Likert scales ranging from 1 - “not at all”, to 5 - “very much”. Common method bias (CMB) was mitigated against through procedural remedies including the provision of clear instructions to participants. In addition, results of Harman’s Single-Factor Test showed poor model fit (RMSEA 0.218; SRMR 0.168; TLI 0.497; CFI 0.447) which suggests an absence of CMB.

Work engagement was measured using the Utrecht Work Engagement Scale (UWES-9) originally developed by Schaufeli et al. (2006) adapted for Korean by Baik (2016). The 9 items represent 3 sub-factors of work engagement: vigor (3 items), dedication (3 items), and absorption (3 items). Sample item: “At work, I feel bursting with energy”.

Narcissism was measured using the 16-item Narcissistic Personality Inventory (NPI) developed by Ames et al. (2006), adapted for Korean by Ha (2019). Sample item: “I am apt to show off if I get a chance”.

Work overload was measured using a 6-item tool utilized by O’Connor et al. (2018) and adapted into Korean by Kim (2020). Sample item: “I have unachievable deadlines”.

Job stress was measured using a 6-item instrument based partly on the research of Lambert and Paoline (2008) and adapted for Korean by Ko (2022). Sample item: “When I’m at work, I often feel tense or uptight”.

Job autonomy was measured using the 9-item, 3-factor Work Design Questionnaire (WDQ) (Morgeson and Humphrey, 2006), adapted for Korea by Im (2019). Sub-factors: work plan autonomy (3 items), decision-making autonomy (3 items), and work method autonomy (3 items). Sample item: “The job allows me to decide on the order in which things are done”.

Grit was measured using the 8-item, 2-factor Short Grit Scale (Grit-S) developed by Duckworth and Quinn (2009) translated to Korean by Kim and Hwang (2015). Sub-factors: consistency of interest (4 items), and perseverance of effort (4 items). Sample item: “I finish whatever I begin”.

Data was analyzed using SPSS 27, AMOS 22, and PROCESS macro 4.2. First, model fit indices (RMSEA, SRMR, CFI, TLI) were used to assess construct validity, and Cronbach’s α was used to assess reliability. Second, descriptive statistics tested skewness, kurtosis, and normality. Third, correlations were analyzed to assess relationships and multicollinearity. Fourth, Hayes’ PROCESS macro examined the moderated mediation effects of job autonomy, grit, and stress on links between narcissism, work overload, and work engagement.

First, validity of the research model was verified as all minimum acceptable values (RMSEA ≤0.08; SRMR ≤0.10; CFI ≥0.90; TLI ≥0.90) were exceeded. The obtained absolute fit indices (RMSEA = 0.058; SRMR = 0.044), and incremental fit indices (TLI = 0.961; CFI = 0.967), were all found to meet the criteria for excellence (Hu and Bentler, 1999).

Next, convergent validity and reliability of the variables were confirmed as all standardized factor loadings and average extracted variances (AVE) were greater than 0.5, and all construct reliability (CR) and Cronbach’s α values were found to exceed 0.7 (lowest obtained values–CR: Narcissism (0.821); Cronbach’s α: Work engagement (0.911)).

Finally, the largest value among the squared correlations was 0.279 (work overload and narcissism), while the smallest value among AVEs was 0.606 (narcissism). As no squared correlation coefficient exceeded the relevant AVE, discriminant validity between constructs was confirmed.

The mean value of work overload was 3.064 (SD = 0.802), job autonomy was 3.513 (SD = 0.703), narcissism was 2.230 (SD = 0.811), grit was 3.324 (SD = 0.825), job stress was 3.326 (SD = 0.994), and work engagement was 3.296 (SD = 0.719). All values of skewness were found to be between −0.724 and 0.701, while the values of kurtosis were between −1.123 and 1.008, satisfying the assumption of normality.

Multicollinearity may be present if the bivariate correlation coefficient between variables is above 0.9 (Kline, 2023). The bivariate correlation coefficients in this study ranged from a low of −0.506 (work engagement and narcissism) to 0.528 (narcissism and work overload) and confirm the absence of multicollinearity.

Hypotheses were presented in the literature based on the order in which they logically emerged from the theoretical framework. However, due to the relationships between variables specified in the hypotheses, it proved necessary to test them in a different order to which they appeared in the literature review. The outcome of each hypothesis is included in Table 1.

To verify the mediating effect of job stress and the moderated mediating effect of job autonomy and grit in the relationship between work overload and work engagement, Hayes’ (2022) PROCESS macro version 9 (bootstrapping 5,000 times and 95% confidence interval) was used as nonparametric bootstrapping is considered an appropriate method to obtain confidence intervals for indirect effects (Williams and Mackinnon, 2008). The seed number was fixed to 781127 to calculate the same bootstrap confidence interval for the indirect effect. Results are shown in Table 2.

In the mediating variable model, the independent variable of work overload (b = 0.512, p < 0.001) had a significant direct positive effect on job stress confirming H3, while the control variables of job autonomy (b = −0.229, p < 0.001) and grit (b = −0.151, p < 0.01) had significant effects on job stress. The interaction term between work overload and job autonomy (b = 0.212, p < 0.01) was significant, which indicates a moderating effect of job autonomy and confirms H9, while the interaction term for work overload and grit was non-significant which means H11 was rejected. In the dependent variable model, both work overload (b = −0.222, p < 0.001) and job stress (b = −0.279, p < 0.001) had significant direct negative effects on work engagement confirming H2, and H5.

The change in R² due to the addition of the interaction term between work overload and job autonomy was significant, again confirming the moderating effect while the conditional effect of work overload according to job autonomy and grit was confirmed. Results are shown in Table 3.

Table 3 shows the moderated effect of work overload on job stress (M±SD, M) was significant in all conditions (low, medium, high) of job autonomy and grit. This indicates that the relationship between work overload and job stress varies depending on the conditions of job autonomy and grit. Figure 2 which shows that as work overload increases, job stress also increases in the three conditions of grit. This tendency appears stronger when job autonomy is low rather than high.

Mediation indices were calculated to test the statistical significance of the moderated mediating effect of job autonomy and grit on the relationship between work overload and work engagement through job stress. Results are shown in Table 4.

The conditional indirect effect of work overload on work engagement was significant in all conditions of job autonomy and grit supporting H7. The adjusted mediation index of job autonomy is −0.059, and the lower limit (−0.091) and upper limit (−0.025) in the 95% confidence interval do not include 0, confirming the moderated mediation effect of job autonomy, supporting H9. For grit, the 95% confidence interval included 0, indicating that there was no moderated mediation effect meaning H11 was rejected.

To empirically verify the mediating effect of job stress in the relationship between narcissism and work engagement, and the moderated mediating effect of job autonomy and grit, Hayes’ (2022) PROCESS macro version 9 (bootstrapping 5,000 times and 95% confidence interval) was used. The seed number was fixed to 781127 to calculate the same bootstrap confidence interval for the indirect effect. Results are shown in Table 5.

In the mediating variable model, the independent variable narcissism had a significant direct positive effect on job stress confirming H4 (b = 0.538, p < 0.001) while the control variables job autonomy (b = −0.225, p < 0.001) and grit (b = −0.154, p) < 0.01) had significant direct negative effects on job stress. In the dependent variable model, both narcissism (b = −0.299, p < 0.001) and job stress (b = −0.240, p < 0.001) had significant direct negative effects on work engagement confirming H1 and H5. The interaction term between narcissism and job autonomy (b = 0.104, p > 0.05) and the interaction term between narcissism and grit (b = 0.087, p > 0.05) were not significant, indicating that there was no moderating effect. Moderated mediation indices were analysed to test the statistical significance of the effect of job autonomy and grit on the effect of narcissism on work engagement through job stress. Results are shown in Table 6.

The conditional indirect effect of narcissism on work engagement was significant for all conditions of job autonomy and grit supporting H6. However, the 95% confidence interval of the adjusted mediation effect index of job autonomy and grit contained 0 in both the lower and upper limits, indicating that there was no moderated mediation effect and H8 and H10 were rejected.

Based on JD-R theory, this study explored how narcissism and work overload affect work engagement via job stress and the moderating effects of job autonomy and grit on the relationships.

First, the results confirmed that narcissism negatively and directly affects work engagement (H1), consistent with our view that narcissism functions as a personal demand within the JD-R framework. Narcissistic employees’ tendency to pursue unrealistic goals and engage in counterproductive interpersonal behaviors appears to drain their cognitive and emotional resources, ultimately reducing their work engagement. Similarly, work overload directly and negatively predicted work engagement (H2), affirming that excessive job demands deplete employees’ resources. Both narcissism and work overload were confirmed to positively predict job stress (H3, H4), and job stress negatively predicted work engagement (H5), thereby establishing mediating pathways. Furthermore, narcissism and work overload exerted indirect negative effects on work engagement via job stress (H6, H7), underscoring the role of the health-impairment process in linking both personal and job demands to reduced engagement.

Second, job autonomy did not moderate the indirect effect of narcissism on work engagement via job stress (H8). These findings contradict reports that people with high narcissism perceive a high level of job autonomy at work, and that autonomy may act as a moderator and contribute to decreased job stress and increased engagement (Clausen et al., 2022; De Spiegelaere et al., 2014). From a theoretical standpoint, this non-significant finding may also point to the nature of narcissism as a personal demand that operates through internal psychological processes rather than situational work conditions. Unlike work overload, which is externally imposed and may be alleviated by structural job resources such as autonomy, narcissism generates stress through internally driven mechanisms, such as a distorted self-perception or inflated entitlement. Job autonomy, as a job resource, may be insufficient to moderate these internal stress processes, suggesting that interventions targeting narcissism-driven stress may need to focus on psychological support or individual attributes rather than job design.

Third, job autonomy was found to moderate the mediating effect of stress in the relationship between work overload and engagement (H9). Specifically, when autonomy is low, the level of stress due to work overload increases further. This supports the findings from previous research that suggested that job autonomy may moderate the relationship between work overload and stress, buffer against the health-impairment process, and ultimately result in desirable outcomes such as reduced burnout, turnover intentions, and workplace bullying (Chia, 2024). Additionally, it was confirmed that in all conditions of grit, stress increases as work overload increases. As such, if greater autonomy is granted at work, the resulting level of job stress will be reduced even if the amount of work is excessive. As a result, the negative impact of job stress on work engagement caused by work overload will be alleviated.

Fourth, grit failed to moderate the narcissism–engagement relationship (H10). This is inconsistent with JD-R-based research showing that grit buffers stress effects (Kim and Lee, 2022). The predominance of millennials in the sample may partially explain this, as this cohort tends to prioritize self-expression and may be less reactive to external judgment or failure (Myers and Sadaghiani, 2010). More fundamentally, however, the non-significant moderating effect of grit on this pathway may reflect a conceptual incompatibility between grit and narcissism as personal characteristics in certain circumstances. Grit involves sustained effort and passion towards long-term goals, yet narcissistic individuals’ motivation is often externally oriented by a need for admiration and validation (Campbell et al., 2004). As a result, the perseverance associated with grit may be insufficient to counteract the relational friction and cognitive distortions inherent to narcissism. This suggests that, in the context of personal demands, the moderating effect of personal resources may be limited when the demand itself conflicts with the psychological mechanisms through which the resource operates.

Finally, grit was not found to moderate the mediating effect of job stress in the relationship between work overload and work engagement (H11). This contradicts previous research which suggested that grit lowered job stress and that job stress had a significant effect on work engagement (Schwepker and Good, 2022; Suzuki et al., 2015). Within JD-R theory, job autonomy is viewed as a job resource while grit is a personal resource which buffer the impact of demands on stress by supporting employees’ capacity to meet those demands and stay motivated (Bakker and Demerouti, 2017). This may be more effective when the demand is motivational in nature, though. As work overload can be seen as a misalignment between task volume and available time, capacity, and resources it is quantitative in nature and represents a structural, rather than motivational, barrier. Grit, as a motivational personal resource, may lack the functional properties necessary to offset this type of demand. A gritty employee facing work overload does not simply need to try harder or persist, they need more time, more support, or fewer tasks, none of which grit can provide.

In addition, Korea’s cultural context, as well as changes in the work environment and situational constraints, may offer an explanation. An increase in the number of retirees due to COVID-19 and Korea’s aging demography, an increase in the number of tasks that must be handled, or a change from working in the workplace to working from home, could increase the work overload of organizational members, contributing to job stress and lowering work engagement.

Based on the above findings, we present several theoretical and practical implications. First, this study advances JD-R theory by incorporating all four components–job demands, job resources, personal demands, and personal resources–within one model. However, only the interaction between job demands and job resources was statistically supported. This suggests that in certain contexts, personal resources such as grit may have limited moderating influence compared to structural job features.

Second, from a practitioner’s perspective, enhancing job autonomy remains a key strategy for improving engagement, and subsequent organizational performance. This may be especially relevant in organizations with high power distance cultures, and rigid vertical hierarchies, where excessive control over subordinates is exerted. In order to increase work engagement, practitioners may consider moving away from rigid vertical hierarchies, affording members increased autonomy in decision-making, and establishing a systematic division of work. Accordingly, it may be beneficial to provide training and support to strengthen members' autonomy so they can better work towards organizational goals.

Lastly, stress generally occurs when a person is assigned a task that is beyond their scope to perform. HR practitioners may choose to assign roles that match the work abilities of employees to minimize negative effects. This could be reinforced by utilizing systems that identify weaknesses and support employees through competency development projects. Finally, institutions should also prepare for potential disruptions, such as those that occurred due to COVID-19, by establishing flexible work policies and contingency frameworks.

Limitations of this study include the fact that narcissism was used as a single example of personal demands, limiting generalizability. Future research should explore more personal traits that may hinder performance. Furthermore, the focus on SMEs in Korea restricts applicability across different organizational sizes. Broader sampling in varied contexts would strengthen the external validity of the findings. Finally, the cross-sectional design limits our ability to definitively rule out CMB, and to make causal inferences. However, the model presented was based on strong theoretical foundations and was supported by prior literature. Finally, this study measured narcissism as an overall construct. As recent research has shown, grandiose and vulnerable narcissism may have distinct and even opposing effects on work-related outcomes (Biolik, 2025). Future research may differentiate between narcissism types to more precisely identify the conditions under which narcissism functions as a personal demand and when it may act as a resource.