Subordinates in Western cultures generally prefer supervisors with a democratic rather than autocratic leadership style. It is unclear, however, whether more narcissistic subordinates share or challenge this prodemocratic default attitude. On the one hand, more narcissistic individuals strive for power and thus may favor a democratic supervisor, who grants them power through participation. On the other hand, similarity attracts and, thus, more narcissistic subordinates may favor an autocratic supervisor, who exhibits the same leadership style that they would adopt in a leadership position.

Method

Four studies (N_total = 1284) tested these competing hypotheses with two narcissism dimensions: admiration and rivalry. Participants indicated the leadership style they generally prefer in a supervisor (Study 1), rated their own supervisor's leadership style (Study 2a: individual ratings; Study 2b: team ratings), and evaluated profiles of democratic and autocratic supervisors (Study 3).

Results

We found a significantly weaker prodemocratic default attitude among more narcissistic subordinates: Subordinates' narcissism was negatively related to endorsement of democratic supervisors and positively related to endorsement of autocratic supervisors. Those relations were mostly driven by narcissistic rivalry rather than narcissistic admiration.

Conclusion

The results help clarify the narcissistic personality and, in particular, how more narcissistic subordinates prefer to be led.

1 INTRODUCTION

Narcissists tend to emerge as leaders (Brunell et al., 2008; Grijalva et al., 2015). An inherent feature of social hierarchies is, however, that most individuals find themselves in subordinate positions (Dufner et al., 2016; Schjelderup-Ebbe, 1922). Absolutely speaking, then, even narcissists can be found more frequently in subordinate than leadership positions. Yet, very little is known about the psychology of more (vs. less) narcissistic subordinates (Benson et al., 2016; Nevicka & Sedikides, 2021). We aim to help fill that gap in the literature by studying more narcissistic subordinates.

In particular, we address a fundamental question: How do more narcissistic subordinates prefer to be led? Subordinates in Western cultures generally prefer supervisors with a democratic rather than autocratic leadership style (Bass, 2008; Schoel et al., 2011). It is unclear, though, whether this prodemocratic default attitude (PDDA; Schoel et al., 2011) also holds true for more narcissistic subordinates. On the one hand, democratic supervisors share power with their subordinates, and narcissistic individuals strive for power and dominance (Krizan & Herlache, 2018; Sedikides, 2021). Thus, more narcissistic subordinates may show a particularly strong PDDA. On the other hand, autocratic supervisors exhibit a leadership style that more narcissistic individuals themselves would adopt in a leadership position (Matosic et al., 2017; Semenyna & Honey, 2015), and similarity attracts (Byrne, 1971; Montoya et al., 2008). Thus, more narcissistic subordinates may show a particularly weak PDDA. We test these competing hypotheses.

1.1 Narcissism

Subclinical narcissism (hereafter: narcissism)—a normally-distributed trait in the general population—is a multidimensional construct. Three narcissism dimensions that have consistently emerged are grandiosity/exhibitionism, entitlement/antagonism, and vulnerability/neuroticism (Krizan & Herlache, 2018; Miller et al., 2016). Given the dimensions' specific features, we consider the former two ones—also known as narcissistic admiration and narcissistic rivalry—particularly relevant for the present research. Narcissistic admiration is characterized by assertive self-promotion, manifesting in striving for uniqueness, grandiosity, and charming behavior. Narcissistic rivalry, by contrast, is characterized by antagonistic self-protection, manifesting in striving for supremacy, devaluation of others, and aggressive behavior (Back et al., 2013).

Narcissistic admiration and rivalry are linked differentially to outcomes such as interpersonal liking, leadership, and strategies to gain status. Specifically, narcissistic admiration fosters interpersonal liking in the early stages of acquaintance, but over time this liking weakens, whereas narcissistic rivalry fosters disliking (Leckelt et al., 2015). Moreover, narcissistic admiration, but not rivalry, predicts leadership emergence in social groups and being in a leadership position 2 years later (Härtel et al., 2023; Leckelt et al., 2019). Finally, narcissistic admiration and rivalry are positively associated with the status-seeking motive but differ in their unique links with strategies to gain status: Narcissistic admiration is primarily associated with willingness to rely on prestige-based strategies (e.g., displaying competence) and less so with willingness to rely on dominance-based strategies (e.g., fear and intimidation), whereas narcissistic rivalry has a positive association only with willingness to rely on dominance-based strategies (Zeigler-Hill et al., 2019).

1.2 Preferences for democratic versus autocratic leadership

Starting with Lewin et al. (1939), researchers have distinguished between two opposing clusters of leadership styles: democratic and autocratic (Bass, 2008).¹ Democratic leaders are concerned with subordinates' needs. They encourage subordinates to participate in decision-making processes as well as to initiate and select their own strategies for dealing with occupational tasks and interacting with each other. Autocratic leaders, by contrast, focus on getting the job done. They make most of the decisions alone, prescribe rules and goals, and control actions and interactions within the group. Democratic and autocratic leadership, then, are two poles of a continuum ranging from high to low levels of granted participation. Democratic leadership grants high levels of subordinate influence and autocratic leadership grants low such levels. Thus, democratic supervisors are more autonomy supportive, whereas autocratic supervisors are more dominant.

Subordinates in Western cultures evince a robust and prevalent preference for democratic over autocratic leadership—the PDDA (Schoel et al., 2011). Autocratic leadership is usually viewed with suspicion and aversion. It makes subordinates feel less valued by their organization (Chan et al., 2013), experience more negative emotions (Pellegrini & Scandura, 2008), and feel less satisfied with their job (Foels et al., 2000). Group members reject autocratic leaders when resolving public good conflicts (Van Vugt & De Cremer, 1999) and are more likely to exit the group under an autocratic than a democratic leader (Van Vugt et al., 2004). Even groups that start with an autocratic decision rule (e.g., the most senior member makes decisions alone) nearly always revert to a democratic rule (e.g., members make decisions together by choosing the majority preference), regardless of performance quality (Nielsen & Miller, 1997). Taken together, the less dominant and more autonomy supportive leadership style of democratic supervisors is more admired by subordinates than the more dominant and less autonomy supportive leadership style of autocratic supervisors (Slemp et al., 2018; Van den Broeck et al., 2016).

1.3 Leadership preferences of more narcissistic subordinates

The key difference between democratic and autocratic supervisors is that the former behave less dominantly than the latter (Bass, 2008; Lewin et al., 1939). Moreover, individuals higher in narcissism are characterized by strivings for dominance and power (Krizan & Herlache, 2018; Sedikides, 2021) and adopt a more autocratic leadership style (Matosic et al., 2017; Semenyna & Honey, 2015). We therefore considered the dominance dimension of interpersonal behavior most relevant for predicting how more narcissistic subordinates prefer to be led. Specifically, we derived two competing hypotheses regarding more narcissistic subordinates' endorsement of democratic versus autocratic supervisors.

The first hypothesis is based on the dominance-complementarity principle (Tiedens et al., 2007). Dominance complementarity has its roots in the interpersonal circumplex (Leary, 1957), represented by the two primary dimensions of interpersonal behavior: affiliation and dominance. Based on the interpersonal circumplex, interpersonal theory proposes more satisfaction with relationships characterized by “sameness on the affiliation dimension and oppositeness on the dominance dimension” (Sadler et al., 2011, p. 126). Regarding task-oriented relationships, such as supervisor–subordinate relationships, interactions characterized by dominance complementarity result in more positive relational and task outcomes (Estroff & Nowicki, 1992; Tiedens & Fragale, 2003). Particularly, dominance complementarity between a person's interpersonal goals (e.g., prefers to behave dominantly) and the partner's interpersonal behavior (e.g., behaves submissively) results in greater satisfaction (Dryer & Horowitz, 1997).² Given that democratic supervisors grant subordinates power through participation, their behavior is complementary to the interpersonal goal of more narcissistic individuals to dominate others. Thus, the first hypothesis states that subordinates' narcissism is positively related to endorsement of democratic supervisors and negatively related to endorsement of autocratic supervisors (stronger PDDA among more narcissistic subordinates). Indirect evidence for this hypothesis indicates that more narcissistic individuals disapprove of being in a subordinate position (Benson et al., 2016).

The second hypothesis is based on the similarity-attraction principle (Byrne, 1971). Individuals seek consensual validation of their characteristics (e.g., values, goals, and personality), and others with similar characteristics provide such validation (Festinger, 1954). The similarity-attraction principle therefore proposes that people evaluate others' characteristics more positively, the more similar those characteristics are to their own, resulting in increased attraction. There is meta-analytic evidence for the similarity-attraction association across different populations and contexts (Montoya et al., 2008), including supervisor-subordinate relationships (Kristof-Brown et al., 2005). Moreover, individuals favor for management positions fictitious job candidates whose leadership style (relationship-oriented vs. task-oriented) is similar to their own (Eagleson et al., 2000). The leadership style of autocratic supervisors is similar to the leadership style more narcissistic individuals would exhibit themselves if they were leaders. Thus, the second hypothesis states that subordinates' narcissism is negatively related to endorsement of democratic supervisors and positively related to endorsement of autocratic supervisors (weaker PDDA among more narcissistic subordinates). Indirect evidence for this hypothesis indicates that more narcissistic individuals support organizational hierarchies from a subordinate position when they expect to rise in rank (Zitek & Jordan, 2016).

According to existent theory, both hypotheses should apply more strongly to narcissistic rivalry than narcissistic admiration: Narcissistic admiration and rivalry are both related to strivings for dominance, but narcissistic admiration is also strongly related to strivings for admiration (Krizan & Herlache, 2018; Sedikides, 2021). Moreover, individuals higher in narcissistic admiration rely primarily on prestige-based strategies and less strongly on dominance-based strategies to gain status, whereas individuals higher in narcissistic rivalry rely only on dominance-based strategies to gain status (Zeigler-Hill et al., 2019). Democratic leadership is less dominant but typically more admired by subordinates than autocratic leadership (Van Vugt & De Cremer, 1999). It is therefore conceivable that for individuals higher in narcissistic admiration the interpersonal goal to dominate others (and accordingly the tendency to adopt an autocratic leadership style) competes against the interpersonal goal to be admired by others (and accordingly the tendency to adopt a democratic leadership style). Those competing goals might cancel each other out, resulting in no relation between narcissistic admiration and the PDDA. Alternatively, depending on which goal is stronger, results for narcissistic admiration might parallel results for narcissistic rivalry (if the goal to dominate others is stronger) or oppose them (if the goal to be admired by others is stronger).

1.4 Overview

We conducted four studies to test the two competing hypotheses that subordinates' narcissism—particularly their narcissistic rivalry—is (1) positively related to endorsement of democratic supervisors and negatively related to endorsement of autocratic supervisors (stronger PDDA) or (2) negatively related to endorsement of democratic supervisors and positively related to endorsement of autocratic supervisors (weaker PDDA). In all studies, participants reported their level of narcissistic admiration and rivalry. In Study 1, participants indicated the leadership style they generally prefer in a supervisor. In Studies 2a–b, we examined whether the link between those general leadership preferences and narcissism translates to subordinates' attraction to their own supervisor. These were field studies in which subordinates indicated their own supervisor's leadership style (individual ratings in Study 2a, team ratings in Study 2b) and evaluated their supervisor as a leader. In Study 3, we used tighter experimental control to test the critical role of supervisors' leadership style in how attracted more narcissistic subordinates are to supervisors. Participants evaluated a fictitious supervisor whose leadership style was either democratic or autocratic. In the latter three studies, participants also indicated the leadership style they would exhibit in a leadership position (i.e., own hypothetical leadership style). This practice allowed us to examine whether more narcissistic subordinates prefer supervisors whose leadership style is complementary or similar to their own hypothetical leadership style (and accordingly to their interpersonal goal to dominate others detached from their subordinate position). In other words, we tested the dominance-complementarity principle versus the similarity-attraction principle as a potential mechanism underlying the leadership preferences of more narcissistic subordinates.

We report all manipulations, measures, and exclusions in these studies. If not otherwise indicated, we analyzed the data in R (version 4.2.1; R Core Team, 2022. All data, data-analysis scripts, and research materials are publicly available at https://madata.bib.uni-mannheim.de/427/. The studies were not preregistered.

2 STUDY 1

In Study 1, we examined the relation between narcissism and how subordinates generally prefer to be led. We asked participants to take a subordinate's perspective and indicate the leadership style they would like to see in their supervisor. We tested whether the relation of subordinates' narcissism—particularly their narcissistic rivalry—with endorsement of democratic versus autocratic supervisors is positive (stronger PDDA) or negative (weaker PDDA).

2.1 Method

2.1.1 Participants

We opted for 300 participants, because we sought to estimate stable relations with high precision (Schönbrodt & Perugini, 2013). In total, 322 MTurk workers completed the study (participation requirements: US resident; approval rating for past MTurk work >95%). We excluded six participants, because they stated that they did not take their participation seriously. The final sample comprised 316 participants (126 women and 190 men; age: 18–84 years, M = 35.37, SD = 10.88). The sample size provided 80% power to detect a unique relation of f² = 0.025 in a linear multiple regression (two to three predictors, α = 0.05, two-tailed; Faul et al., 2009). For reference, the median effect size of meta-analyses in social–personality psychology is f² = 0.033 (Richard et al., 2003).

2.1.2 Procedure and materials

Participants first completed questionnaires (in counterbalanced order) assessing narcissism and impression management (the latter served as a covariate to rule out that differences between more and less narcissistic individuals were simply due to social desirability concerns). Next, they imagined themselves being in a subordinate position and reflected on what it means and how it feels to be a subordinate. Then, they responded to an attention check asking them whether they had imagined themselves as leader, subordinate, or elephant. Those who failed this check were directed to an early-end page. The remaining participants indicated their leadership preferences and provided demographic information.

Narcissism

We administered the Narcissistic Admiration and Rivalry Questionnaire (NARQ; Back et al., 2013), comprising the 9-item Admiration subscale (α = 0.91; e.g., “I deserve to be seen as a great personality”) and the 9-item Rivalry subscale (α = 0.91; e.g., “I enjoy it when another person is inferior to me”; 1 = not agree at all, 6 = agree completely).³

Impression management

We administered the 8-item Impression Management subscale (α = 0.78; e.g., “I sometimes tell lies if I have to” [reverse-coded]; 1 = strongly disagree, 7 = strongly agree) of the Balanced Inventory of Desirable Responding Short Form (C. M. Hart et al., 2015).

Leadership preferences

We assessed leadership preferences with two complementary but highly correlated approaches, r(314) = 0.70, p < 0.001. First, participants read descriptions of democratic and autocratic leadership in counterbalanced order. We derived the descriptions from classic definitions of these leadership styles (Bass, 2008; Lewin et al., 1939). For example, the democratic leadership description stated that “the supervisor lets the subordinates initiate most actions and interactions within the group,” whereas the autocratic leadership description stated that “the supervisor directs most actions and interactions within the group” (for full descriptions, see the material file at https://madata.bib.uni-mannheim.de/427/). For each description, participants indicated how much they, as subordinates, would want their supervisor to show the described leadership style (1 = not at all, 6 = very much).⁴

Further, participants read 10 democratic leadership behaviors (e.g., “put suggestions made by the group into operation”) and 10 autocratic leadership behaviors (e.g., “decide what shall be done and how it shall be done”) in alternating order. We adopted these behaviors mainly from the Leader Behavior Description Questionnaire XII (The Ohio State Leadership Studies, 1962). For each behavior, participants indicated how frequently they, as subordinates, would want their supervisor to show it (1 = very seldom, never, 6 = very often, always).

Democratic and autocratic leadership can be construed as two poles of a continuum (Bass, 2008; Lewin et al., 1939). We therefore reverse-coded responses to the autocratic leadership description and behaviors before averaging all responses to a measure. The result was a single leadership descriptions measure and a single leadership behaviors measure. For both measures, high values indicated preference for clearly democratic leadership and low values indicated preference for clearly autocratic leadership, with values in between indicating preference for a mixture of democratic and autocratic leadership. We obtained highly similar results for the two leadership measures. Thus, for brevity, we report the results for the average of the two (z-standardized) measures (α = 0.82) here and the results of the single measures in Supplementary Material.

2.2 Results and discussion

Table 1 summarizes descriptive statistics and zero-order correlations. The sample means for narcissistic admiration and rivalry were comparable to past research (Back et al., 2013; Zeigler-Hill et al., 2019). The sample means of leadership descriptions and behaviors indicated that, on average, subordinates preferred democratic over autocratic supervisors, replicating the PDDA (Bass, 2008; Schoel et al., 2011). More important, narcissistic admiration and particularly narcissistic rivalry correlated negatively with leadership preferences, indicating a weaker PDDA among more narcissistic subordinates.

TABLE 1. Descriptive statistics and zero-order correlations in Study 1.

Variable	M	SD	1	2	3
ADM	3.35	1.20
RIV	2.66	1.20	0.54***
IM	4.34	1.17	−0.15**	−0.45***
Leadership preferences	0.00	0.92	−0.18**	−0.44***	0.10

Note: For leadership preferences, high values indicated preference for democratic leadership and low values indicated preference for autocratic leadership. M_{leadership preferences} = 0.00 equates to M_{leadership descriptions} = 4.35, SD = 1.01, and M_{leadership behaviors} = 4.12, SD = 0.66. The sample means of leadership descriptions and behaviors were significantly above the scale-midpoint of 3.5, ts(315) ≥ 15.00, ps < 0.001, ds ≥ 0.84.
Abbreviations: ADM, narcissistic admiration; IM, impression management; RIV, narcissistic rivalry.
** p < 0.01;
*** p < 0.001.

To test whether that relation was unique to each narcissism dimension, we conducted two regression models. In Model 1, we entered both narcissism dimensions as simultaneous predictors. In Model 2, we additionally entered impression management as third predictor. We report the results in Table 2. Narcissistic rivalry, but not narcissistic admiration, showed a unique negative relation to leadership preferences, f² = 0.20, ΔR² = 0.16, F(1, 313) = 62.59, p < 0.001; that is, only subordinates higher in narcissistic rivalry showed a weaker PDDA.⁵ This finding remained virtually unchanged when we controlled for impression management.⁶

TABLE 2. Unique relations of narcissistic admiration and rivalry to leadership preferences in Study 1.

Predictor	Model 1				Model 2
Predictor	β	95% CI	t	p	β	95% CI	t	p
ADM	0.07	[−0.04, 0.19]	1.23	0.221	0.09	[−0.03, 0.21]	1.51	0.131
RIV	−0.48	[−0.59, −0.36]	−7.91	<0.001	−0.55	[−0.68, −0.42]	−8.20	<0.001
IM^a					−0.14	[−0.25, −0.02]	−2.38	0.018
	R² = 0.19, F(2, 313) = 37.76, p < 0.001				R² = 0.21, F(3, 312) = 27.44, p < 0.001

Note: For leadership preferences, high values indicated preference for democratic leadership and low values indicated preference for autocratic leadership.
Abbreviations: ADM, narcissistic admiration; CI, confidence interval; IM, impression management; RIV, narcissistic rivalry.
^a The zero-order correlation between impression management and leadership preferences was not significant, r(314) = 0.10, p = 0.084 (Table 1). By contrast, the semi-partial correlation between impression management and leadership preferences in Model 2 was significant, sr(312) = −0.12, p = 0.034. That semi-partial correlation was also significant when narcissistic admiration was omitted from Model 2, sr(313) = −0.11, p = 0.048, but it was not significant when narcissistic rivalry was omitted from Model 2, sr(313) = 0.07, p = 0.213. These results suggest that narcissistic rivalry acts as a suppressor in the relation between impression management and leadership preferences (Paulhus et al., 2004).

In summary, Study 1 replicated the PDDA but showed that it is weaker among subordinates higher in narcissistic rivalry. Previous research suggests that narcissistic rivalry is associated with the inclination to adopt a more autocratic leadership style (Matosic et al., 2017). Thus, the weaker PDDA among subordinates higher in narcissistic rivalry is in line with a stronger preference for a supervisor who shows a leadership style that they would exhibit in a leadership position (similarity-attraction principle) rather than a supervisor who shows a leadership style that is complementary to their own (hypothetical) leadership style in terms of dominance (and accordingly to their interpersonal goal to dominate others; dominance-complementarity principle). Yet, a direct test of similarity-attraction as a process driving the relation between subordinates' narcissistic rivalry and their leadership preferences is needed. Moreover, it is unclear whether those leadership preferences translate to subordinates' attraction to their own supervisor. We addressed both issues in Studies 2a–2b.

3 STUDIES 2A–2B

Studies 2a–2b had two goals. The first goal was to test whether the relation of narcissistic admiration and rivalry with subordinates' endorsement of democratic versus autocratic supervisors in general translates to how attracted more narcissistic subordinates are to their own supervisor—that is, the real-life supervisor they actually have. Studies 2a–2b therefore examined the relation of narcissistic admiration and rivalry with subordinates' attraction to their own supervisor as a function of their own supervisor's leadership style. Participants occupying subordinate positions in organizational hierarchies indicated their own supervisor's leadership style and evaluated their supervisor as a leader. In Study 2a, assessment of supervisor's leadership style was based on individual ratings. In Study 2b, it was based on organizational team ratings to enhance reliability and objectivity of ratings. Following Study 1's results, we had a single hypothesis in the present study (recall that we had two competing hypotheses in Study 1). Our hypothesis was that subordinates' narcissistic rivalry is negatively related to endorsement of democratic supervisors and positively related to endorsement of autocratic supervisors (weaker PDDA).

The second study goal was to directly test similarity-attraction as a process driving the relation between subordinates' narcissistic rivalry and their leadership preferences. Therefore, in addition to the leadership style of participants' supervisor, we measured participants' hypothetical leadership style—that is, their leadership style if they were in a leadership position. Similarity attracts if the relation between a democratic own hypothetical leadership style and attraction to supervisor is positive for democratic supervisors and negative for autocratic supervisors. We hypothesized that a negative relation of subordinates' narcissistic rivalry with a democratic own hypothetical leadership style accounts for the weaker PDDA among subordinates higher in narcissistic rivalry.

3.1 Method of Study 2a

3.1.1 Participants

We gathered data using a snowballing procedure. A student assistant distributed the study invitation via organizational contacts and social networks (e.g., Facebook, Xing). We recruited as many participants as possible within 1 month. In total, 158 participants completed the German online survey (participation requirement: have been working with their supervisor for at least 1 year). We excluded two participants due to an error in data recording (inexplicably, their data included ratings of a female and a male supervisor). The final sample comprised 156 participants (103 women and 53 men; age: 19–62 years, M = 33.83, SD = 11.11). Participants had been working with their supervisor between 1 and 25 years (M = 3.95, SD = 4.01). Most participants (81%) reported seeing their supervisor at least once a week. The sample size provided 80% power to detect (1) a unique relation (main or interaction effect) of f² = 0.051 in a linear multiple regression (five predictors, α = 0.05, two-tailed; Faul et al., 2009) and (2) an indirect effect with each path's effect (predictor → mediator; mediator → outcome) being f² = 0.068 (bias-corrected percentile bootstrap method; Fritz & MacKinnon, 2007).

3.1.2 Procedure and materials

Participants first provided demographic information and completed the NARQ (Back et al., 2013; α_Admiration = 0.84; α_Rivalry = 0.80).⁷ Then, in counterbalanced order, they indicated (1) their supervisor's leadership style followed by an evaluation of their supervisor, and (2) their own hypothetical leadership style.

Supervisor's leadership style

Participants read the autocratic and democratic leadership descriptions from Study 1 (autocratic description first). For each description, they indicated the degree to which it corresponded with their supervisor's leadership style (1 = not at all, 7 = very much). Next, they read the 10 democratic and 10 autocratic leadership behaviors from Study 1 in alternating order. For each behavior, they reported how often their supervisor showed it (1 = never, 5 = almost always/regularly). As in Study 1, we first formed one leadership descriptions measure and one leadership behaviors measure and then averaged them (z-standardized) to obtain a single index (α = 0.90). We report results on this index here and results for the separate measures in Supplementary Material S2.

Subordinate's hypothetical leadership style

Participants imagined themselves being in their supervisor's position and read the same leadership descriptions and behaviors used to assess their supervisor's leadership style. For each description, they indicated the degree to which it corresponded with the leadership style they would show if they were in their supervisor's position (1 = not at all, 7 = very much). For each behavior, they reported how often they would show it if they were in their supervisor's position (1 = never, 5 = almost always/regularly). Parallel to supervisor's leadership style, we formed a single index of subordinate's hypothetical leadership style (α = 0.77).

Attraction to supervisor

Participants rated their supervisor on 23 items (e.g., “How much do you like your supervisor as a leader?”; 1 = not at all, 7 = very much) that covered eight aspects of attraction to supervisor, such as liking for their supervisor and their supervisor's competence (for all aspects and their items, see the material file at https://madata.bib.uni-mannheim.de/427/). Those aspects were so closely interrelated that only a single factor showed up in a principal components analysis. This factor explained 71.17% of the total variance, and all aspects manifested high loadings on it (all loadings ≥0.68). Hence, we z-standardized and averaged them to form the outcome: attraction to supervisor (α = 0.94).

3.2 Results of Study 2a

We organized this section in two parts. First, we conducted a moderation analysis to test whether the relation of narcissistic admiration and rivalry with subordinates' attraction to their own supervisor was moderated by their own supervisor's leadership style. In particular, we tested whether the relation of subordinates' narcissistic rivalry with attraction to their own supervisor was negative for democratic supervisors and positive for autocratic supervisors. Second, we conducted a mediation analysis with the R-package lavaan (v0.6-17; Rosseel, 2012) to test whether a negative relation of subordinates' narcissistic rivalry with a democratic own hypothetical leadership style accounts for those relations.

Table 3 summarizes descriptive statistics and zero-order correlations. The sample means for narcissistic admiration and rivalry were again comparable to past research (Back et al., 2013; Zeigler-Hill et al., 2019). The sample means of leadership descriptions and behaviors indicated that, on average, supervisors' leadership style was more democratic than autocratic, and subordinates would adopt a more democratic than autocratic leadership style if they were in their supervisor's position.

TABLE 3. Descriptive statistics and zero-order correlations in Studies 2a–2b.

Variable	Study 2a		Study 2b		1	2	3	4	5
Variable	M	SD	M	SD	1	2	3	4	5
ADM	3.19	0.81	3.06	0.80		0.47***	−0.26***	−0.20**	−0.09
RIV	2.25	0.76	2.04	0.75	0.31***		−0.48***	−0.30***	−0.18**
SHLS	0.00	0.90	0.00	0.89	−0.06	−0.45***		0.34***	0.25***
SLS	0.00	0.95	0.04	0.96	−0.10	−0.03	0.09		0.70***
Attraction	0.00	0.84	0.00	0.87	−0.08	0.03	0.07	0.67***

Note: Zero-order correlations of Study 2a (Study 2b) are reported below (above) the diagonal. For SHLS and SLS, high values indicated a democratic leadership style and low values indicated an autocratic leadership style. In Study 2a, M_SHLS = 0.00 equates to M_{leadership descriptions} = 5.44, SD = 1.33, and M_{leadership behaviors} = 3.51, SD = 0.47, and M_SLS = 0.00 equates to M_{leadership descriptions} = 4.38, SD = 1.82, and M_{leadership behaviors} = 3.23, SD = 0.75. In Study 2b, M_SHLS = 0.00 equates to M_{leadership descriptions} = 4.89, SD = 0.90, and M_{leadership behaviors} = 4.52, SD = 0.57, and M_SLS = 0.04 equates to M_{leadership descriptions} = 4.22, SD = 1.16, and M_{leadership behaviors} = 4.13, SD = 0.82. The sample means of all SHLS and SLS measures were significantly above the respective scale-midpoint, ts(155) ≥ 2.61, ps ≤0.010, ds ≥0.21, in Study 2a and ts(210) ≥ 9.05, ps <0.001, ds ≥0.62, in Study 2b.
Abbreviations: ADM, narcissistic admiration; Attraction, attraction to supervisor; RIV, narcissistic rivalry; SHLS, subordinate's hypothetical leadership style; SLS, supervisor's leadership style (individual ratings in Study 2a, team ratings in Study 2b).
** p < 0.01.
*** p < 0.001.

3.2.1 Moderation analysis

We simultaneously regressed attraction to supervisor on both narcissism dimensions, supervisor's leadership style, and their two-way interactions (all z-standardized). We report the results in Table 4. The effect of supervisor's leadership style was significant, indicating stronger attraction to supervisors with a more democratic leadership style (a replication of the PDDA). As hypothesized, narcissistic rivalry interacted with supervisor's leadership style, f² = 0.07, ΔR² = 0.03, F(5, 150) = 9.92, p = 0.002. The relation of subordinates' narcissistic rivalry with attraction to supervisor was negative for democratic supervisors, β = −0.32, 95% CI [−0.58, −0.05], t(150) = −2.35, p = 0.020, and positive for autocratic supervisors, β = 0.45, 95% CI [0.18, 0.73], t(150) = 3.28, p = 0.001 (weaker PDDA). We display simple slopes in Figure 1a.⁸ The main effects of both narcissism dimensions and the Admiration × Supervisor's Leadership Style interaction were not significant.⁹

TABLE 4. Unique effects of narcissistic admiration and rivalry on attraction to supervisor moderated by supervisor's leadership style in Studies 2a–2b.

Predictor	Study 2a				Study 2b
Predictor	β	95% CI	t	p	zPE	95% CI	z	p
(Intercept)					−0.06	[−0.18, 0.06]	−1.06	0.290
ADM	−0.01	[−0.13, 0.11]	−0.16	0.871	0.06	[−0.06, 0.17]	0.93	0.350
RIV	0.02	[−0.11, 0.14]	0.25	0.803	−0.03	[−0.22, 0.16]	−0.32	0.747
SLS	0.67	[0.55, 0.79]	11.02	<0.001	0.75	[0.67, 0.83]	18.07	<0.001
ADM × SLS	−0.03	[−0.16, 0.09]	−0.55	0.584	−0.01	[−0.13, 0.12]	−0.14	0.890
RIV × SLS	−0.20	[−0.32, −0.07]	−3.15	0.002	−0.11	[−0.22, 0.002]	−1.92	0.054
	R² = 0.50, F(5, 150) = 29.98, p < 0.001				$R_{total}^{2 (fvm)}$ = 0.53^a

Note: For SLS, high values indicated a democratic leadership style and low values indicated an autocratic leadership style. In Study 2b, the variance of the random intercept was 0.010, and the variance of the random slopes was less than 0.001 for narcissistic admiration and 0.007 for narcissistic rivalry.
Abbreviations: ADM, narcissistic admiration; CI, confidence interval; RIV, narcissistic rivalry; SLS, supervisor's leadership style (individual ratings in Study 2a, team ratings in Study 2b); zPE, standardized point estimate.
^a We estimated $R_{total}^{2 (fvm)}$ with the R-package r2mlm (v 0.3.7; Shaw et al., 2023). That package estimates $R_{total}^{2 (fvm)}$ based on model parameters computed in R. Those model parameters slightly differed from the reported model parameters computed in Mplus 8.5, but the differences were negligible.

Details are in the caption following the image — **FIGURE 1**
Open in figure viewer PowerPoint

Attraction to supervisor as a function of subordinate's narcissistic rivalry and supervisor's leadership style in Study 2a (a) and Study 2b (b). All variables were z-standardized. Supervisor's leadership style was based on individual ratings in Study 2a (a) and team ratings in Study 2b (b). Simple slopes were estimated for supervisors whose leadership style was clearly democratic or clearly autocratic—that is, for the leadership style values of the most democratic (Study 2a: standardized value = 1.83; Study 2b: standardized value = 1.54) and the most autocratic (Study 2a: standardized value = −2.40; Study 2b: standardized value = −2.24) supervisors in the studies (for corresponding values on the leadership descriptions and leadership behaviors measures, see Footnotes 8 and 12). Simple slopes were controlled for narcissistic admiration and the Admiration × Supervisor's Leadership Style interaction.

3.2.2 Mediation analysis

As depicted in Figure 2a, we tested a moderated mediation model. To this end, we extended the former regression model as follows: (1) We included subordinate's hypothetical leadership style as an additional predictor. (2) We specified the two-way interaction between subordinate's hypothetical leadership style and supervisor's leadership style. (3) We specified indirect effects of both narcissism dimensions on attraction to supervisor via subordinate's hypothetical leadership style (path a of indirect effect: narcissism dimension → subordinate's hypothetical leadership style; path b of indirect effect: Subordinate's Hypothetical Leadership Style × Supervisor's Leadership Style interaction → attraction to supervisor; model 15 in Hayes, 2018, p. 592). We report the results in Table 5.

TABLE 5. Results of the moderated mediation model in Studies 2a–2b.

Predictor	Study 2a				Study 2b
Predictor	β	95% CI	z	p	zPE	95% CI	z	p
Subordinate's hypothetical leadership style
ADM	0.09	[−0.06, 0.23]	1.22	0.221	−0.04	[−0.17, 0.09]	−0.60	0.550
RIV	−0.48	[−0.65, −0.31]	−5.46	<0.001	−0.46	[−0.62, −0.30]	−5.66	<0.001
	R² = 0.21, F(2, 153) = 20.67, p < 0.001				$R_{total}^{2 (fvm)}$ = 0.31^a
Attraction to supervisor
(Intercept)					−0.09	[−0.21, 0.03]	−1.44	0.149
ADM	−0.03	[−0.16, 0.09]	−0.49	0.624	0.02	[−0.11, 0.15]	0.32	0.747
RIV	0.06	[−0.06, 0.18]	0.90	0.368	−0.03	[−0.27, 0.22]	−0.22	0.829
SHLS	0.03	[−0.12, 0.18]	0.35	0.728	0.03	[−0.11, 0.18]	0.46	0.646
SLS	0.63	[0.50, 0.75]	10.22	<0.001	0.74	[0.66, 0.82]	18.19	<0.001
ADM × SLS	−0.05	[−0.19, 0.08]	−0.72	0.470	0.04	[−0.10, 0.18]	0.57	0.569
RIV × SLS	−0.04	[−0.17, 0.08]	−0.70	0.483	−0.03	[−0.15, 0.09]	−0.52	0.601
SHLS × SLS	0.26	[0.11, 0.41]	3.50	<0.001	0.17	[0.04, 0.30]	2.62	0.009
	R² = 0.55, F(7, 148) = 25.58, p < 0.001				$R_{total}^{2 (fvm)}$ = 0.59^a

Note: For SHLS and SLS, high values indicated a democratic leadership style and low values indicated an autocratic leadership style. In Study 2b, the variance of the random intercept was 0.010, and the variance of the random slopes was less than 0.001 for narcissistic admiration, 0.007 for narcissistic rivalry, and 0.055 for subordinate's hypothetical leadership style.
Abbreviations: ADM, narcissistic admiration; CI, confidence interval; RIV, narcissistic rivalry; SHLS, subordinate's hypothetical leadership style; SLS, supervisor's leadership style (individual ratings in Study 2a, team ratings in Study 2b); zPE, standardized point estimate.
^a We estimated $R_{total}^{2 (fvm)}$ with the R-package r2mlm (v 0.3.7; Shaw et al., 2023). That package estimates $R_{total}^{2 (fvm)}$ based on model parameters computed in R. Those model parameters slightly differed from the reported model parameters computed in Mplus 8.5, but the differences were negligible.

As expected, subordinates' narcissistic rivalry (but not admiration) was negatively related to a democratic own hypothetical leadership style, f² = 0.27, ΔR² = 0.21, F(1, 153) = 40.63, p < 0.001.The Subordinate's Hypothetical Leadership Style × Supervisor's Leadership Style interaction on attraction to supervisor was also significant, f² = 0.10, ΔR² = 0.05, F(1, 148) = 15.41, p < 0.001. Consistent with the similarity-attraction principle, the relation between a democratic own hypothetical leadership style and attraction to supervisor was positive for democratic supervisors and negative for autocratic supervisors (Figure 2a).

Most important, the index of moderated mediation (Hayes, 2018) for narcissistic rivalry was significant, estimate = −0.13, 95% CI [−0.24, −0.04], z = −2.58, p = 0.010. The indirect effect of narcissistic rivalry on attraction to supervisor via subordinate's hypothetical leadership style was negative for democratic supervisors, estimate = −0.24, 95% CI [−0.48, −0.07], z = −2.40, p = 0.016, and positive for autocratic supervisors, estimate = 0.29, 95% CI [0.10, 0.56], z = 2.47, p = 0.014. By contrast, the index of moderated mediation for narcissistic admiration was not significant, estimate = 0.02, 95% CI [−0.01, 0.07], z = 1.15, p = 0.249.

3.3 Method of Study 2b

3.3.1 Participants

We gathered data through a snowballing procedure. A student assistant distributed the study invitation via organizational contacts and social networks (e.g., Facebook, Xing). To obtain more reliable and objective ratings of the leadership style of supervisors, we sought to recruit members of organizational teams. Specifically, we asked participants to invite other members of their organizational team who had the same supervisor to take part in the study. We recruited as many participants as possible within 1 month. In total, 226 participants completed the German online survey (participation requirements: have been working with a supervisor, but to maximize sample size we did not specify the minimum duration; have at least two colleagues who had been working with the same supervisor and were willing to take part in the study). Of these, 212 could be matched with at least one other team member. In one team, however, one participant indicated a different gender of the supervisor and was therefore excluded from analyses. The final sample comprised 211 participants (123 women, 86 men, and 2 undisclosed; age: 18–63 years, M = 33.00, SD = 11.26) across 70 teams with two to six team members (M = 3.01, SD = 0.84). Except for 12 participants (6%), they had been working with their supervisor for at least 1 year (1–35 years, M = 4.33, SD = 4.65). Most participants (73%) reported seeing their supervisor at least once a week.

It is possible to estimate power for cross-level interactions with Monte Carlo simulations (Green & MacLeod, 2016). However, that method requires a number of parameters that we were unable to specify for our hypothesized cross-level interactions without prior data. Instead, we report observed power and evaluate evidence from this study within the broader evaluation of our research in General Discussion (cf. Wegener et al., 2022).

3.3.2 Procedure and materials

Participants first rated their supervisor's leadership style followed by an evaluation of their supervisor. Next, they completed the NARQ (Back et al., 2013; α_Admiration = 0.83; α_Rivalry = 0.82) and indicated their own hypothetical leadership style. Finally, they provided demographic information.

Supervisor's leadership style

We used the leadership behaviors and descriptions from Studies 1–2a with minor changes. To increase participation rate, we reduced the study duration by using only half of the leadership behaviors and shorter leadership descriptions. Moreover, we introduced 6-point rating scales for both measures. We aggregated each measure's individual values on team level (reliability of team-level means: ICC[2]_behaviors = 0.87; ICC[2]_descriptions = 0.83).¹⁰ Then, congruent with Study 2a, we averaged team-level means of both measures (z-standardized) to obtain a single index for the leadership style of each team's supervisor (α = 0.95). As in Studies 1-2a, we focus on this index here and report results for the separate measures in Supplementary Material S2.

Subordinate's hypothetical leadership style

Participants imagined themselves being in their supervisor's position and read the leadership behaviors and descriptions from Studies 1-2a but—as for supervisor's leadership style—with fewer leadership behaviors, shorter leadership descriptions, and 6-point rating scales. As in Study 2a, we formed a single index of subordinate's hypothetical leadership style (α = 0.75).

Attraction to supervisor

Participants rated their supervisor on 12 items (e.g., “How much do you like your supervisor as a leader?”; 1 = not at all, 7 = very much) that covered six aspects of attraction to supervisor, such as liking for their supervisor and their supervisor's competence (for all aspects and their items, see the material file at https://madata.bib.uni-mannheim.de/427/). As in Study 2a, those aspects were so closely interrelated that only a single factor showed up in a principal components analysis. This factor explained 75.64% of the total variance, and all aspects manifested high loadings on it (all loadings ≥0.80). Hence, we z-standardized and averaged them to form the outcome: attraction to supervisor (α = 0.93).

3.4 Results of Study 2b

As in Study 2a, we first conducted a moderation analysis to test whether the relation of subordinates' narcissistic rivalry with attraction to their own supervisor was negative for democratic supervisors and positive for autocratic supervisors. Next, we conducted a mediation analysis to test whether those relations were attributable to a negative relation of subordinates' narcissistic rivalry with a democratic own hypothetical leadership style. To account for the nested data structure (participants nested in teams), we conducted random-intercept random-slope models (Barr et al., 2013). We computed the models with Mplus 8.5 (Muthén & Muthén, 1998–2017), because the currently available R-packages do not allow testing indirect effects of multilevel moderated mediation models that include random slopes. To achieve model convergence, we set the correlations between the random intercept and the random slopes to zero (Bates et al., 2015). We z-standardized all variables to obtain standardized coefficients.¹¹

3.4.1 Moderation analysis

We simultaneously regressed attraction to supervisor (Level 1) on both narcissism dimensions (Level 1), supervisor's leadership style (Level 2), and their cross-level interactions. We report the results in Table 4. The effect of supervisor's leadership style was significant, indicating stronger attraction to supervisors with a more democratic leadership style (a replication of the PDDA). The Rivalry × Supervisor's Leadership Style interaction was not significant, but the pattern of results paralleled that of Study 2a: Descriptively, the relation of subordinates' narcissistic rivalry with attraction to supervisor was negative for democratic supervisors, standardized point estimate (zPE) = −0.20, 95% CI [−0.44, 0.04], z = −1.62, p = 0.104, and positive for autocratic supervisors zPE = 0.21, 95% CI [−0.12, 0.54], z = 1.23, p = 0.220 (trend for a weaker PDDA). We display simple slopes in Figure 1b.¹² The main effects of both narcissism dimensions and the Admiration × Supervisor's Leadership Style interaction were not significant.

3.4.2 Mediation analysis

The total effect of the Rivalry × Supervisor's Leadership Style interaction was not significant. Yet, our predicted indirect effect via subordinate's hypothetical leadership style might be present because the total effect is the sum of the direct and the indirect effect. If the direct and the indirect effect differ in their direction, the total effect might be null, whereas the indirect effect might be significant (Hayes, 2018).

As in Study 2a, we tested a moderated mediation model (Figure 2b). To this end, we extended the former multilevel model as follows: (1) We included subordinate's hypothetical leadership style as an additional (Level 1) predictor. (2) We specified the cross-level interaction between subordinate's hypothetical leadership style and supervisor's leadership style. (3) We specified indirect effects of both narcissism dimensions on attraction to supervisor via subordinate's hypothetical leadership style. We report the results in Table 5.

Again, subordinates' narcissistic rivalry (but not admiration) was negatively related to a democratic own hypothetical leadership style. The Subordinate's Hypothetical Leadership Style × Supervisor's Leadership Style interaction on attraction to supervisor was also significant. Consistent with the similarity-attraction principle, the relation between a democratic own hypothetical leadership style and attraction to supervisor was positive for democratic supervisors and negative for autocratic supervisors (Figure 2b).

Most important, the index of moderated mediation for narcissistic rivalry was significant, estimate = −0.08, 95% CI [−0.14, −0.02], z = −2.55, p = 0.011. The indirect effect of narcissistic rivalry on attraction to supervisor via subordinate's hypothetical leadership style was negative for democratic supervisors, estimate = −0.14, 95% CI [−0.24, −0.03], z = −2.55, p = 0.011, and positive for autocratic supervisors, estimate = 0.16, 95% CI [0.00, 0.32], z = 1.96, p = 0.050. By contrast, the index of moderated mediation for narcissistic admiration was not significant, estimate = −0.01, 95% CI [−0.03, 0.02], z = −0.56, p = 0.572.

3.5 Discussion of Studies 2a–2b

Studies 2a–2b provided evidence that the general leadership preferences of more narcissistic subordinates found in Study 1 translate to how attracted more narcissistic subordinates are to their own supervisor. Specifically, subordinates' narcissistic rivalry was negatively related to endorsement of democratic supervisors and positively related to endorsement of autocratic supervisors (albeit not significant in Study 2b). Moreover, consistent with the similarity-attraction principle, a negative relation of subordinates' narcissistic rivalry with a democratic own hypothetical leadership style accounted for those relations. These findings held for assessments of supervisor's leadership style at the individual level (Study 2a) and at the team level (Study 2b).

A strength of Studies 2a–2b is their high ecological validity, as participants occupied subordinate positions in real organizational hierarchies. Yet, all data were cross-sectional. It can therefore not be ruled out that subordinates had adopted their supervisors' leadership style over time when they were attracted to their supervisor, instead of being more attracted to their supervisor when their own hypothetical leadership style was similar to their supervisor's leadership style. In an effort to provide firmer evidence for the critical role of supervisors' leadership style in how attracted more narcissistic subordinates are to supervisors, we used tighter experimental control in Study 3.

4 STUDY 3

In Study 3, we tested whether a low democratic and high autocratic leadership style elicits stronger attraction to a supervisor for subordinates higher in narcissistic rivalry. We presented participants with a fictitious supervisor whose leadership style was either clearly democratic or clearly autocratic. Following our findings from Studies 1–2, we hypothesized that subordinates' narcissistic rivalry is negatively related to endorsement of the democratic leader profile and positively related to endorsement of the autocratic leader profile. We again expected the similarity between subordinate's hypothetical leadership style and the fictitious supervisor's leadership style to mediate those relations.

4.1 Method

4.1.1 Participants

We opted for 300 participants per condition (600 in all), because we sought to (1) estimate stable relations with high precision within each condition (Schönbrodt & Perugini, 2013), and (2) ensure high reliability of indirect effects (Fritz & MacKinnon, 2007). In total, 601 MTurk workers (318 women, 282 men, and 1 undisclosed; age: 18–75 years, M = 36.85, SD = 11.20) completed the study (participation requirements: US resident; approval rating for past MTurk work >95%).¹³ The sample size provided 80% power to detect (1) a unique relation (main or interaction effect) of f² = 0.013 in a linear multiple regression (five predictors, α = 0.05, two-tailed; Faul et al., 2009) and (2) an indirect effect with each path's effect being f² = 0.02 (bias-corrected percentile bootstrap method; Fritz & MacKinnon, 2007).

4.1.2 Procedure and materials

Participants first completed the NARQ (Back et al., 2013; α_Admiration = 0.88; α_Rivalry = 0.88) and indicated their own hypothetical leadership style. Next, they were randomly assigned to the democratic or autocratic leader profile of a fictitious supervisor and evaluated this person. Finally, they stated how they perceived the supervisor's leadership style and provided demographic information.

Supervisor's leadership style

Participants read the ostensible answers of a male leader to the democratic and autocratic leadership descriptions and behaviors from Studies 1–2a and imagined that they were given by a leader who was their supervisor (but not the real-life supervisor they actually have).¹⁴ The ratings of the fictitious supervisor were either clearly democratic (“6” for the democratic description and “2” for the autocratic description on 7-point scales; M = 4.50 for democratic behaviors and M = 1.50 for autocratic behaviors on 5-point scales) or clearly autocratic (“6” for the autocratic description and “2” for the democratic description; M = 4.50 for autocratic behaviors and M = 1.50 for democratic behaviors). We presented the democratic description together with the democratic behaviors and the autocratic description together with the autocratic behaviors to help participants notice that the supervisor's leadership style was clearly democratic or autocratic. Participants viewed the democratic items and the autocratic items in counterbalanced order.

Subordinate's hypothetical leadership style

Participants imagined themselves being in a leadership position and read the same items that we used for the leader profile of the fictitious supervisor (in the same order as in the supervisor's leader profile). Participants indicated correspondence with the two leadership descriptions (1 = not at all, 7 = very much) and frequency of each leadership behavior (1 = never, 5 = always). As in Studies 2a–2b, we formed a single index of subordinate's hypothetical leadership style (α = 0.80).

Attraction to supervisor

Participants rated the fictitious supervisor on 21 items (e.g., “How much would you like this supervisor as a leader?”; 1 = not at all, 7 = very much) that covered seven aspects of attraction to supervisor, such as liking for the supervisor and the supervisor's competence (for all aspects and their items, see the material file at https://madata.bib.uni-mannheim.de/427/). As in Studies 2a–2b, those aspects were so closely interrelated that only a single factor showed up in a principal components analysis. This factor explained 83.38% of the total variance, and all measures manifested high loadings on it (all loadings ≥0.85). Hence, we z-standardized and averaged them to form the outcome: attraction to supervisor (α = 0.97).

Perceived leadership style

Participants stated how they perceived the leadership style of the fictitious supervisor (1 = democratic, 6 = autocratic).

4.2 Results and discussion

First, we tested whether participants perceived the supervisor's leadership style as intended. Second, as in Studies 2a–2b, we conducted a moderation analysis to test whether subordinates' narcissistic rivalry is negatively related to endorsement of the democratic leader profile and positively related to endorsement of the autocratic leader profile. Finally, we conducted a mediation analysis to test whether a negative relation of subordinates' narcissistic rivalry with a democratic own hypothetical leadership style accounted for those relations.

Table 6 summarizes descriptive statistics and zero-order correlations. The sample means for narcissistic admiration and rivalry were again comparable to past research (e.g., Back et al., 2013; Zeigler-Hill et al., 2019). The sample means of leadership descriptions and behaviors indicated that, on average, subordinates would adopt a more democratic than autocratic leadership style if they were in a leadership position.

TABLE 6. Descriptive statistics and zero-order correlations in Study 3.

Variable	Democratic condition (n = 299)		Autocratic condition (n = 302)		1	2	3	4
Variable	M	SD	M	SD	1	2	3	4
ADM	3.31	1.00	3.18	1.03
RIV	2.35	1.08	2.27	0.95	0.42***
SHLS	0.00	0.94	0.00	0.89	−0.19***	−0.38***
SLS	0.50	0.00	−0.50	0.00	0.07	0.04	0.004
Attraction	0.66	0.58	−0.65	0.69	0.10*	0.07	−0.03	0.72***

Note: For SHLS, high values indicated a democratic leadership style and low values indicated an autocratic leadership style. M_SHLS = 0.00 equates to M_{leadership descriptions} = 5.18, SD = 1.28, and M_{leadership behaviors} = 3.57, SD = 0.53. The sample means of both SHLS measures were significantly above the respective scale-midpoint, ts(600) ≥ 22.60, ps < 0.001, ds ≥ 0.92.
Abbreviations: ADM, narcissistic admiration; Attraction, attraction to supervisor; Autocratic condition, participants viewed the autocratic leader profile of the fictitious supervisor; Democratic condition, participants viewed the democratic leader profile of the fictitious supervisor; RIV, narcissistic rivalry; SHLS, subordinate's hypothetical leadership style; SLS, supervisor's leadership style (predetermined; −0.5 = autocratic, 0.5 = democratic).
* p < 0.05.
*** p < 0.001.

4.2.1 Perceived leadership style

As intended, participants perceived the supervisor's leadership style as democratic in the democratic condition (M = 1.94, SD = 1.38), t(298) = −19.50, p < 0.001, d = −1.13, 95% CI [−1.27, −0.98], and as autocratic in the autocratic condition (M = 5.18, SD = 1.47), t(301) = 19.79, p < 0.001, d = 1.14, 95% CI [0.99, 1.28], resulting in a large difference between the two leader profiles, t(599) = 27.77, p < 0.001, d = 2.27, 95% CI [2.06, 2.47].¹⁵

4.2.2 Moderation analysis

We simultaneously regressed attraction to supervisor on both narcissism dimensions, supervisor's leadership style (−0.5 = autocratic, 0.5 = democratic), and their two-way interactions (all z-standardized). We report the results in Table 7. The effect of supervisor's leadership style was significant, indicating stronger attraction to the democratic than the autocratic supervisor (a replication of the PDDA). As hypothesized, the Rivalry × Supervisor's Leadership Style interaction was significant, f² = 0.02, ΔR² = 0.01, F(1, 595) = 11.35, p < 0.001. The relation of subordinates' narcissistic rivalry with attraction to supervisor descriptively tended to be negative for the democratic supervisor, β = −0.07, 95% CI [−0.16, 0.01], t(595) = −1.71, p = 0.089, and was positive for the autocratic supervisor, β = 0.14, 95% CI [0.05, 0.23], t(595) = 3.02, p = 0.003 (weaker PDDA). We display simple slopes in Figure 3. The main effects of both narcissism dimensions and the Admiration × Supervisor's Leadership Style interaction were not significant.¹⁶

TABLE 7. Results of the moderation model and the moderated mediation model in Study 3.

Predictor	Moderation model				Moderated mediation model
Predictor	β	95% CI	t	p	β	95% CI	z	p
Subordinate's hypothetical leadership style
ADM					−0.03	[−0.11, 0.05]	−0.78	0.433
RIV					−0.36	[−0.44, −0.28]	−9.13	<0.001
					R² = 0.14, F(2, 598) = 49.31, p < 0.001
Attraction to supervisor
ADM	0.05	[−0.01, 0.11]	1.61	0.109	0.04	[−0.03, 0.11]	1.11	0.267
RIV	0.03	[−0.03, 0.09]	1.06	0.289	0.02	[−0.04, 0.09]	0.68	0.495
SHLS					−0.03	[−0.08, 0.03]	−0.97	0.331
SLS	0.71	[0.66, 0.77]	25.26	<0.001	0.71	[0.66, 0.77]	27.16	<0.001
ADM × SLS	−0.04	[−0.10, 0.03]	−1.16	0.248	−0.03	[−0.10, 0.04]	−0.78	0.434
RIV × SLS	−0.11	[−0.17, −0.04]	−3.37	<0.001	−0.01	[−0.08, 0.06]	−0.30	0.767
SHLS × SLS					0.26	[0.20, 0.31]	9.27	<0.001
	R² = 0.53, F(5, 595) = 134.61, p < 0.001				R² = 0.59, F(7, 593) = 121.20, p < 0.001

Note: For SHLS, high values indicated a democratic leadership style and low values indicated an autocratic leadership style.
Abbreviations: ADM, narcissistic admiration; CI, confidence interval; RIV, narcissistic rivalry; SHLS, subordinate's hypothetical leadership style; SLS, supervisor's leadership style (predetermined; −0.5 = autocratic, 0.5 = democratic).

4.2.3 Mediation analysis

As depicted in Figure 4, we tested the same moderated mediation model as in Study 2a. We report the results in Table 7. Again, subordinates' narcissistic rivalry (but not admiration) was negatively related to a democratic own hypothetical leadership style, f² = 0.13, ΔR² = 0.11, F(1, 598) = 74.76, p < 0.001. The Subordinate's Hypothetical Leadership Style × Supervisor's Leadership Style interaction on attraction to supervisor was also significant, f² = 0.14, ΔR² = 0.06, F(1, 593) = 83.19, p < 0.001. Consistent with the similarity-attraction principle, the relation between a democratic own hypothetical leadership style and attraction to supervisor was positive for the democratic supervisor and negative for the autocratic supervisor (Figure 4).

Most important, the index of moderated mediation for narcissistic rivalry was significant, estimate = −0.09, 95% CI [−0.13, −0.07], z = −6.41, p < 0.001. The indirect effect of narcissistic rivalry on attraction to supervisor via subordinate's hypothetical leadership style was negative for the democratic supervisor, estimate = −0.08, 95% CI [−0.12, −0.06], z = −5.07, p < 0.001, and positive for the autocratic supervisor, estimate = 0.10, 95% CI [0.07, 0.14], z = 5.42, p < 0.001. By contrast, the index of moderated mediation for narcissistic admiration was not significant, estimate = −0.01, 95% CI [−0.03, 0.01], z = −0.77, p = 0.439.

In summary, Study 3 provided firmer evidence that a low democratic and high autocratic leadership style elicits stronger attraction to a supervisor for subordinates higher in narcissistic rivalry. Moreover, consistent with the similarity-attraction principle, a negative relation of subordinates' narcissistic rivalry with a democratic own hypothetical leadership style accounted for this stronger attraction.

5 GENERAL DISCUSSION

We replicated previous findings showing that subordinates generally prefer democratic over autocratic supervisors (Bass, 2008; Schoel et al., 2011). However, we also found that this PDDA is significantly weaker among more narcissistic subordinates—especially those higher in narcissistic rivalry. Our findings were highly consistent across four studies using different methods (hypothetical scenarios and evaluations of own supervisors in organizations).

5.1 Implications

Our findings attest to the utility of investigating specific narcissism dimensions. In particular, the findings indicate that the weaker PDDA among more narcissistic subordinates is mostly driven by the entitlement/antagonism dimension (NARQ-Rivalry) and less so by the grandiosity/exhibitionism dimension (NARQ-Admiration). When narcissistic rivalry was controlled, narcissistic admiration did not predict the proclivity to adopt a more autocratic leadership style, nor did it predict the proclivity to adopt a more democratic leadership style. Hence, similarity-attraction did not result in a weaker or stronger PDDA for subordinates higher in narcissistic admiration.¹⁷ By contrast, narcissistic rivalry predicted the proclivity to adopt a more autocratic leadership style. Moreover, in line with the similarity-attraction principle, the size of the PDDA was reduced by about one half (Study 3) up to three quarters (Study 2a) for subordinates higher (M +1 SD) versus lower (M −1 SD) in narcissistic rivalry. Yet, the PDDA did not reverse. One reason may be that democratic principles constitute such important social values in Western cultures that even most individuals higher in narcissism adhere to them.

Our findings also advance knowledge on traits that attenuate the PDDA and on the underlying processes. Previous research has shown that subordinates with a lower, unstable self-esteem have a weaker PDDA under conditions of self-uncertainty (Schoel et al., 2011, 2015). Presumably, they are more likely than subordinates with a higher, stable self-esteem to doubt their own ability to handle an uncertain situation well and, thus, prefer to leave decisions solely to a leader (dominance-complementarity principle). We found a weaker PDDA for subordinates higher in narcissistic rivalry when self-uncertainty was probably neither high nor low on average, and this effect was driven by similarity-attraction. Given that narcissistic rivalry is associated with a rather low, unstable self-esteem (Geukes et al., 2017), however, subordinates higher in narcissistic rivalry may show a weaker PDDA under conditions of self-uncertainty because of strivings for dominance-complementarity rather than similarity-attraction. Consequently, traits that attenuate the PDDA may do so through distinct processes depending on situational variables.

Our findings also have implications for the type of supervisor more narcissistic subordinates support. In two field studies (Studies 2a–b), subordinates' narcissistic rivalry was negatively related to a democratic own hypothetical leadership style, and consistent with the similarity-attraction principle, this resulted in a relation of subordinates' narcissistic rivalry with attraction to their supervisor that was negative for democratic and positive for autocratic supervisors. Subordinates who are more attracted to their supervisor should also be more likely to support their supervisor. Therefore, more narcissistic subordinates may pose a threat to the support for democratic supervisors and contribute to support for autocratic supervisors.

The positive relation of subordinates' narcissism with attraction to autocratic supervisors also has implications regarding support for narcissistic leaders. Given that narcissistic leaders are inclined to adopt a more autocratic leadership style (Matosic et al., 2017), they should receive broad support particularly from their more narcissistic subordinates. Moreover, more narcissistic individuals are more accepting of other's narcissistic characteristics (W. Hart & Adams, 2014) and are more often friends with other more narcissistic individuals (Maaß et al., 2016), rendering it likely that narcissistic leaders prefer to hire more narcissistic subordinates. The combination of narcissistic leaders and more narcissistic subordinates, however, may be detrimental to an organization. Sooner or later, narcissistic leaders will clash with their organization because of their abrasiveness and contentious financial or ethical decisions (Sedikides & Campbell, 2017). Support from more narcissistic subordinates may obscure that conflict in the short run.

5.2 Limitations and future research

We investigated two narcissism dimensions: grandiosity/exhibitionism in the agentic domain (NARQ-Admiration) and entitlement/antagonism (NARQ-Rivalry). Future research may examine the grandiosity/exhibitionism dimension in the communal domain (i.e., communal narcissism) or the vulnerability/neuroticism dimension (i.e., vulnerable narcissism). Communal narcissists perceive themselves as both dominant and highly communal (Gebauer et al., 2012). Accordingly, their interpersonal goal to dominate others is likely at odds with their interpersonal goal to care for the needs of others. How that conflict affects the leadership preferences of communal narcissists awaits empirical testing. Vulnerable narcissism is unrelated to dominance (Edershile & Wright, 2021) but negatively related to self-esteem and stable self-views (Krizan & Herlache, 2018). Building on previous research on subordinates with a lower, unstable self-esteem (Schoel et al., 2011, 2015), vulnerable narcissists might show a weaker PDDA particularly under conditions of self-uncertainty. This hypothesis also awaits empirical testing.

We focused on the dominance dimension of interpersonal behavior to predict how more narcissistic subordinates prefer to be led. The dominance dimension is related to narcissism as well as democratic and autocratic leadership. It is also related to status hierarchies, rendering it particularly relevant in task-oriented relationships such as supervisor-subordinate relationships (Tiedens et al., 2007). The second primary dimension of interpersonal behavior—affiliation—plays a tangential role in such relationships and is not included in the definition of democratic and autocratic leadership. Moreover, warmth is not only a key feature of the affiliation dimension but also an ingredient of attraction. Thus, it is unclear when higher warmth ratings result in higher attraction and not vice versa. Beyond that, there are certainly processes unrelated to the two primary dimensions of interpersonal behavior that additionally contribute to the attraction of more narcissistic subordinates to supervisors. For example, the weaker PDDA among subordinates higher in narcissistic rivalry may additionally be driven by perceiving the typically admired democratic supervisors as a threat to the inflated ego. Future research may test such additional processes.

Except for supervisor's leadership style in Study 3 (which we predetermined), each participant provided data for all variables at one time point (albeit we aggregated data on supervisor's leadership style in Study 2b across team members to yield higher objectivity). This might have produced common method variance, particularly because self-reports of more narcissistic individuals may be biased to maintain their overly positive self-view. For example, communal narcissists see themselves as highly prosocial, but they do not behave more prosocially than their nonnarcissistic counterparts in the eyes of their peers (Nehrlich et al., 2019). Importantly, controlling for impression management in Study 1 did not change the results. Moreover, in Studies 2a–2b and 3, we tested interaction effects, and common method variance cannot account for interaction effects, although it can deflate them (Siemsen et al., 2010). Nonetheless, future research may examine whether leadership preferences translate to observable behavior that naturally occurs in group or organizational settings.

We provided converging evidence for a weaker PDDA among more narcissistic subordinates across two vignette-type studies and two field studies. Results showed that the vignette-type studies (Studies 1 and 3) and the first field study (Study 2a) were adequately powered, whereas the second field study (Study 2b) was partly underpowered (observed power: 41% for the Rivalry × Supervisor's Leadership Style interaction, 100% for path a and 70% for path b of the indirect effect). Yet, given that Study 2b conceptually replicated the results of the other studies, we consider the results all in all reliable. Regardless, as larger datasets increase statistical power, we hope that further data will be added to our publicly available dataset. Moreover, new data may test situational moderators. For example, more narcissistic subordinates support organizational hierarchies only if they expect to rise in rank (Zitek & Jordan, 2016). Maybe more narcissistic subordinates are more attracted to autocratic supervisors especially in light of promotion prospects.

6 CONCLUDING REMARKS

The PDDA has long been considered a robust, prevalent preference in organizational settings (and not only there). Our findings indicate that the PDDA is weaker for more narcissistic subordinates—an effect predominantly driven by narcissistic rivalry and accounted for by the similarity-attraction principle. Given narcissistic leaders' inclination to more autocratic leadership, they are likely to receive stronger support from more narcissistic subordinates. Together with narcissistic leaders' inclination to hire more narcissistic subordinates, this practice may be harmful to an organization, because stronger support from subordinates may shroud narcissistic leaders' controversial behavior and decisions.

AUTHOR CONTRIBUTIONS

Jennifer Eck: Conceptualization, methodology, formal analysis, visualization, validation, data curation, writing—original draft preparation, writing—review and editing. Christiane Schoel: Conceptualization, methodology, project administration, investigation, formal analysis, visualization, writing—original draft preparation, writing—review and editing. Constantine Sedikides, Jochen E. Gebauer, and Dagmar Stahlberg: Conceptualization, methodology, writing—review and editing.

ACKNOWLEDGMENTS

We thank Maren Oberländer and Jula Grünewald for their help in collecting the data of Study 2a and Study 2b, respectively. Open Access funding enabled and organized by Projekt DEAL.

FUNDING INFORMATION

The authors received no financial support for the research, authorship, and/or publication of this article.

CONFLICT OF INTEREST STATEMENT

The authors declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

ETHICS STATEMENT

We conducted all studies in accordance with the Ethical Standards of the American Psychological Association. As per local ethics committee regulations, none of the studies required ethical review.

Endnotes

¹ The meaning of “democratic” and “autocratic” in leadership research (referring to the leadership style of supervisors in groups and organizations) differs from the meaning of those terms in political science (referring to patterns of government). Thus, findings on democratic and autocratic leadership are inapplicable to democratic and autocratic governments, and vice versa.

² Interpersonal complementarity can occur during a particular interaction or across a relationship. We focus on the supervisor–subordinate relationship, because interpersonal goals (also known as trait interpersonal styles; Sadler et al., 2011) are relatively stable and do not need to be reflected in observable behaviors during every interaction (Dryer & Horowitz, 1997).

³ We report the results on the NARQ total score in Supplementary Material. In all studies, we also administered the most widely used measure of narcissism—the Narcissistic Personality Inventory (NPI; Raskin & Terry, 1988)—for supplementary analyses reported in Supplementary Material. Results on the NARQ total score and NPI conceptually replicated most results on narcissistic rivalry. NPI results were mostly driven by the entitlement/exploitativeness facet of the scale.

⁴ Additionally, we asked participants whether they wanted their supervisor to show a more autocratic or more democratic leadership style. Given that we asked this question only in Studies 1 and 2b, we report the results in Supplementary Material S1 and S2. This measure yielded results similar to the reported ones.

⁵ The PDDA was reversed (i.e., preference for autocratic over democratic leadership) for narcissistic rivalry scores ≥ M + 2.07 SD (leadership descriptions measure) and M + 2.06 SD (leadership behaviors measure).

⁶ Presentation order of the democratic and autocratic leadership descriptions was not a significant moderator (Supplementary Material S1).

⁷ In Studies 2a–2b, we measured the Big Five and self-esteem as covariates. Including the Big Five and self-esteem into analyses did not change the reported results (Supplementary Material S2).

⁸ To derive simple slopes for supervisors whose leadership style was clearly democratic or clearly autocratic, we had to rely on the leadership style values of the most democratic (standardized value = 1.83, composed of leadership descriptions = 7.00, and leadership behaviors = 4.75) and the most autocratic (standardized value = −2.40, composed of leadership descriptions = 1.00, and leadership behaviors = 1.20) supervisors in the study. No leadership style values were outliers.

⁹ Completion order of supervisor's leadership style and subordinate's hypothetical leadership style was not a significant moderator (Supplementary Material S2).

¹⁰ The variance in individual values that could be explained by team membership (i.e., the ratio of between-team variance to total variance) was ICC[1]_behaviors = 0.69 and ICC[1]_descriptions = 0.62. The average within-team agreement was r_{WG(behaviors)} = 0.87 and r_{WG(descriptions)} = 0.72.

¹¹ Level-1 predictors can be grand-mean or group-mean centered: “The choice between the two options for centering can only be made on a theoretical basis” (Kreft et al., 1995, p. 1). We decided to z-standardize level-1 predictors (i.e., all grand means = 0, all SDs = 1), because our hypotheses concerned participants' level of narcissism relative to the entire sample rather than relative to the few other participants of their team.

¹² To derive simple slopes for supervisors whose leadership style was clearly democratic or clearly autocratic, we had to rely on the leadership style values of the most democratic (standardized value = 1.54, composed of leadership descriptions = 5.75, and leadership behaviors = 5.50) and the most autocratic (standardized value = −2.24, composed of leadership descriptions = 2.50, and leadership behaviors = 1.63) supervisors in the study. No leadership style values were extreme outliers.

¹³ Six participants completed the study twice (original N = 607). We entered only their first round of data into analyses.

¹⁴ The fictitious supervisor was male for all participants. People might perceive successful leaders as possessing characteristics typically ascribed to men rather than women (think-manager-think-male phenomenon; Schein, 1973). Moreover, female supervisors might be evaluated less favorably than male supervisors when perceived as dominant (Ma et al., 2022).

¹⁵ An additional regression analysis revealed that subordinates higher in narcissistic admiration perceived less pronounced differences between the two leader profiles. Yet, the perceived differences were substantial and highly significant (Supplementary Material S3).

¹⁶ Presentation order of democratic and autocratic leadership items was not a significant moderator (Supplementary Material S3).

¹⁷ Separate analyses for each narcissism dimension conceptually replicated the reported results with one exception: In Studies 2b and 3, subordinates' narcissistic admiration had a significant negative relation with a democratic own hypothetical leadership style, resulting in a significant negative index of moderated mediation (Supplementary Material S2 and S3). In other words, subordinates higher in narcissistic admiration reported a more autocratic own hypothetical leadership style than subordinates lower in narcissistic admiration, and this culminated in a weaker PDDA among subordinates higher in narcissistic admiration. However, these results were likely driven by the relation between narcissistic admiration and narcissistic rivalry.

Open Research

DATA AVAILABILITY STATEMENT

Deidentified data, data-analysis scripts, and research materials for all studies are publicly available on the University of Mannheim's Research Data Repository (https://madata.bib.uni-mannheim.de/427/).

Supporting Information

REFERENCES

Back, M. D., Küfner, A. C. P., Dufner, M., Gerlach, T. M., Rauthmann, J. F., & Denissen, J. J. A. (2013). Narcissistic admiration and rivalry: Disentangling the bright and dark sides of narcissism. Journal of Personality and Social Psychology, 105(6), 1013–1037. https://doi.org/10.1037/a0034431
10.1037/a0034431
PubMed Web of Science® Google Scholar
Barr, D. J., Levy, R., Scheepers, C., & Tily, H. J. (2013). Random effects structure for confirmatory hypothesis testing: Keep it maximal. Journal of Memory and Language, 68(3), 255–278. https://doi.org/10.1016/j.jml.2012.11.001
10.1016/j.jml.2012.11.001
Web of Science® Google Scholar
Bass, B. M. (2008). The Bass handbook of leadership: Theory, research, and managerial applications ( 4th ed.). Free Press.
Google Scholar
Bates, D., Kliegl, R., Vasishth, S., & Baayen, R. H. (2015). Parsimonious mixed models. https://arxiv.org/pdf/1506.04967.pdf
Google Scholar
Benson, A. J., Jordan, C. H., & Christie, A. M. (2016). Narcissistic reactions to subordinate role assignment: The case of the narcissistic follower. Personality and Social Psychology Bulletin, 42(7), 985–999. https://doi.org/10.1177/0146167216649608
10.1177/0146167216649608
PubMed Web of Science® Google Scholar
Brunell, A. B., Gentry, W. A., Campbell, W. K., Hoffman, B. J., Kuhnert, K. W., & DeMarree, K. G. (2008). Leader emergence: The case of the narcissistic leader. Personality and Social Psychology Bulletin, 34(12), 1663–1676. doi:10.1177/0146167208324101
10.1177/0146167208324101
PubMed Web of Science® Google Scholar
Byrne, D. E. (1971). The attraction paradigm. Academic Press.
Google Scholar
Chan, S. C. H., Huang, X., Snape, E., & Lam, C. (2013). The Janus face of paternalistic leaders: Authoritarianism, benevolence, subordinates' organization-based self-esteem, and performance. Journal of Organizational Behavior, 34(1), 108–128. https://doi.org/10.1002/job.1797
10.1002/job.1797
Web of Science® Google Scholar
Dryer, D. C., & Horowitz, L. M. (1997). When do opposites attract? Interpersonal complementarity versus similarity. Journal of Personality and Social Psychology, 72(3), 592–603. https://doi.org/10.1037/0022-3514.72.3.592
10.1037/0022-3514.72.3.592
Web of Science® Google Scholar
Dufner, M., Leising, D., & Gebauer, J. E. (2016). Which basic rules underlie social judgments? Agency follows a zero-sum principle and communion follows a non-zero-sum principle. Personality and Social Psychology Bulletin, 42(5), 677–687. https://doi.org/10.1177/0146167216640902
10.1177/0146167216640902
PubMed Web of Science® Google Scholar
Eagleson, G., Waldersee, R., & Simmons, R. (2000). Leadership behaviour similarity as a basis of selection into a management team. British Journal of Social Psychology, 39(2), 301–308. https://doi.org/10.1348/014466600164480
10.1348/014466600164480
PubMed Web of Science® Google Scholar
Edershile, E. A., & Wright, A. G. C. (2021). Grandiose and vulnerable narcissistic states in interpersonal situations. Self and Identity, 20(2), 165–181. https://doi.org/10.1080/15298868.2019.1627241
10.1080/15298868.2019.1627241
PubMed Web of Science® Google Scholar
Estroff, S. D., & Nowicki, S., Jr. (1992). Interpersonal complementarity, gender of interactants, and performance on puzzle and word tasks. Personality and Social Psychology Bulletin, 18(3), 351–356. https://doi.org/10.1177/0146167292183012
10.1177/0146167292183012
Web of Science® Google Scholar
Faul, F., Erdfelder, E., Buchner, A., & Lang, A.-G. (2009). Statistical power analyses using G*Power 3.1: Tests for correlation and regression analyses. Behavior Research Methods, 41(4), 1149–1160. https://doi.org/10.3758/BRM.41.4.1149
10.3758/BRM.41.4.1149
PubMed Web of Science® Google Scholar
Festinger, L. (1954). A theory of social comparison processes. Human Relations, 7(2), 117–140. https://doi.org/10.1177/001872675400700202
10.1177/001872675400700202
Web of Science® Google Scholar
Foels, R., Driskell, J. E., Mullen, B., & Salas, E. (2000). The effects of democratic leadership on group member satisfaction: An integration. Small Group Research, 31(6), 676–701. https://doi.org/10.1177/104649640003100603
10.1177/104649640003100603
Web of Science® Google Scholar
Fritz, M. S., & MacKinnon, D. P. (2007). Required sample size to detect the mediated effect. Psychological Science, 18(3), 233–239. https://doi.org/10.1111/j.1467-9280.2007.01882.x
10.1111/j.1467-9280.2007.01882.x
PubMed Web of Science® Google Scholar
Gebauer, J. E., Sedikides, C., Verplanken, B., & Maio, G. R. (2012). Communal narcissism. Journal of Personality and Social Psychology, 103(5), 854–878. https://doi.org/10.1037/a0029629
10.1037/a0029629
PubMed Web of Science® Google Scholar
Geukes, K., Nestler, S., Hutteman, R., Dufner, M., Küfner, A. C. P., Egloff, B., Denissen, J. J. A., & Back, M. D. (2017). Puffed-up but shaky selves: State self-esteem level and variability in narcissists. Journal of Personality and Social Psychology, 112(5), 769–786. https://doi.org/10.1037/pspp0000093
10.1037/pspp0000093
PubMed Web of Science® Google Scholar
Green, P., & MacLeod, C. J. (2016). SIMR: An R package for power analysis of generalized linear mixed models by simulation. Methods in Ecology and Evolution, 7(4), 493–498. https://doi.org/10.1111/2041-210X.12504
10.1111/2041-210X.12504
Web of Science® Google Scholar
Grijalva, E., Harms, P. D., Newman, D. A., Gaddis, B. H., & Fraley, R. C. (2015). Narcissism and leadership: A meta-analytic review of linear and nonlinear relationships. Personnel Psychology, 68(1), 1–47. https://doi.org/10.1111/peps.12072
10.1111/peps.12072
Web of Science® Google Scholar
Hart, C. M., Ritchie, T. D., Hepper, E. G., & Gebauer, J. E. (2015). The Balanced Inventory of Desirable Responding Short Form (BIDR-16). SAGE Open, 5(4), 1–9. https://doi.org/10.1177/2158244015621113
10.1177/2158244015621113
Web of Science® Google Scholar
Hart, W., & Adams, J. M. (2014). Are narcissists more accepting of others' narcissistic traits? Personality and Individual Differences, 64, 163–167. https://doi.org/10.1016/j.paid.2014.02.038
10.1016/j.paid.2014.02.038
Web of Science® Google Scholar
Härtel, T. M., Leckelt, M., Grosz, M. P., Küfner, A. C. P., Geukes, K., & Back, M. D. (2023). Pathways from narcissism to leadership emergence in social groups. European Journal of Personality, 37(1), 72–94. https://doi.org/10.1177/08902070211046266
10.1177/08902070211046266
Web of Science® Google Scholar
Hayes, A. F. (2018). Introduction to mediation, moderation, and conditional process analysis: A regression-based approach ( 2nd ed.). Guilford Press.
Google Scholar
Kreft, I. G., de Leeuw, J., & Aiken, L. S. (1995). The effect of different forms of centering in hierarchical linear models. Multivariate Behavioral Research, 30(1), 1–21. https://doi.org/10.1207/s15327906mbr3001_1
10.1207/s15327906mbr3001_1
CAS PubMed Web of Science® Google Scholar
Kristof-Brown, A. L., Zimmerman, R. D., & Johnson, E. C. (2005). Consequences of individuals' fit at work: A meta-analysis of person-job, person-organization, person-group, and person-supervisor fit. Personnel Psychology, 58(2), 281–342. https://doi.org/10.1111/j.1744-6570.2005.00672.x
10.1111/j.1744-6570.2005.00672.x
Web of Science® Google Scholar
Krizan, Z., & Herlache, A. (2018). The narcissism spectrum model: A synthetic view of narcissistic personality. Personality and Social Psychology Review, 22(1), 3–31. https://doi.org/10.1177/1088868316685018
10.1177/1088868316685018
PubMed Web of Science® Google Scholar
Leary, T. (1957). Interpersonal diagnosis of personality. Ronald Press.
Google Scholar
Leckelt, M., Küfner, A. C. P., Nestler, S., & Back, M. D. (2015). Behavioral processes underlying the decline of narcissists' popularity over time. Journal of Personality and Social Psychology, 109(5), 856–871. https://doi.org/10.1037/pspp0000057
10.1037/pspp0000057
PubMed Web of Science® Google Scholar
Leckelt, M., Richter, D., Wetzel, E., & Back, M. D. (2019). Longitudinal associations of narcissism with interpersonal, intrapersonal, and institutional outcomes: An investigation using a representative sample of the German population. Collabra Psychology, 5(1), Article 26. https://doi.org/10.1525/collabra.248
10.1525/collabra.248
Web of Science® Google Scholar
Lewin, K., Lippitt, R., & White, R. K. (1939). Patterns of aggressive behavior in experimentally created “social climates”. Journal of Social Psychology, 10(2), 271–299. https://doi.org/10.1080/00224545.1939.9713366
10.1080/00224545.1939.9713366
Google Scholar
Ma, A., Rosette, A. S., & Koval, C. Z. (2022). Reconciling female agentic advantage and disadvantage with the CADDIS measure of agency. Journal of Applied Psychology, 107(12), 2115–2148. https://doi.org/10.1037/apl0000550
10.1037/apl0000550
PubMed Web of Science® Google Scholar
Maaß, U., Lämmle, L., Bensch, D., & Ziegler, M. (2016). Narcissists of a feather flock together: Narcissism and the similarity of friends. Personality and Social Psychology Bulletin, 42(3), 366–384. https://doi.org/10.1177/0146167216629114
10.1177/0146167216629114
PubMed Web of Science® Google Scholar
Matosic, D., Ntoumanis, N., Boardley, I. D., Sedikides, C., Stewart, B. D., & Chatzisarantis, N. (2017). Narcissism and coach interpersonal style: A self-determination theory perspective. Scandinavian Journal of Medicine and Science in Sports, 27(2), 254–261. https://doi.org/10.1111/sms.12635
10.1111/sms.12635
CAS PubMed Web of Science® Google Scholar
Miller, J. D., Lynam, D. R., McCain, J. L., Few, L. R., Crego, C., Widiger, T. A., & Campbell, W. K. (2016). Thinking structurally about narcissism: An examination of the five-factor narcissism inventory and its components. Journal of Personality Disorders, 30(1), 1–18. https://doi.org/10.1521/pedi_2015_29_177
10.1521/pedi_2015_29_177
PubMed Web of Science® Google Scholar
Montoya, R. M., Horton, R. S., & Kirchner, J. (2008). Is actual similarity necessary for attraction? A meta-analysis of actual and perceived similarity. Journal of Social and Personal Relationships, 25(6), 889–922. https://doi.org/10.1177/0265407508096700
10.1177/0265407508096700
Web of Science® Google Scholar
Muthén, L. K., & Muthén, B. O. (1998–2017). Mplus user's guide ( 8th ed.). Muthén & Muthén.
10.1146/annurev.ecolsys.28.1.341
Web of Science® Google Scholar
Nehrlich, A. D., Gebauer, J. E., Sedikides, C., & Schoel, C. (2019). Agentic narcissism, communal narcissism, and prosociality. Journal of Personality and Social Psychology, 117(1), 142–165. https://doi.org/10.1037/pspp0000190
10.1037/pspp0000190
PubMed Web of Science® Google Scholar
Nevicka, B., & Sedikides, C. (2021). Employee narcissism and promotability prospects. Journal of Personality, 89(4), 847–862. https://doi.org/10.1111/jopy.12619
10.1111/jopy.12619
PubMed Web of Science® Google Scholar
Nielsen, M. E., & Miller, C. E. (1997). The transmission of norms regarding group decision rules. Personality and Social Psychology Bulletin, 23(5), 516–525. https://doi.org/10.1177/0146167297235007
10.1177/0146167297235007
Web of Science® Google Scholar
Paulhus, D. L., Robins, R. W., Trzesniewski, K. H., & Tracy, J. L. (2004). Two replicable suppressor situations in personality research. Multivariate Behavioral Research, 39(2), 303–328. https://doi.org/10.1207/s15327906mbr3902_7
10.1207/s15327906mbr3902_7
PubMed Web of Science® Google Scholar
Pellegrini, E. K., & Scandura, T. A. (2008). Paternalistic leadership: A review and agenda for future research. Journal of Management, 34(3), 566–593. https://doi.org/10.1177/0149206308316063
10.1177/0149206308316063
Web of Science® Google Scholar
R Core Team. (2022). R: A language and environment for statistical computing. R Foundation for Statistical Computing. https://www.R-project.org/
Google Scholar
Raskin, R., & Terry, H. (1988). A principal-components analysis of the Narcissistic Personality Inventory and further evidence of its construct validity. Journal of Personality and Social Psychology, 54(5), 890–902. https://doi.org/10.1037/0022-3514.54.5.890
10.1037/0022-3514.54.5.890
CAS PubMed Web of Science® Google Scholar
Richard, F. D., Bond, C. F., & Stokes-Zoota, J. J. (2003). One hundred years of social psychology quantitatively described. Review of General Psychology, 7(4), 331–363. https://doi.org/10.1037/1089-2680.7.4.331
10.1037/1089-2680.7.4.331
Web of Science® Google Scholar
Rosseel, Y. (2012). lavaan: An R package for structural equation modeling. Journal of Statistical Software, 48(2), 1–36. http://www.jstatsoft.org/v48/i02/
10.18637/jss.v048.i02
Web of Science® Google Scholar
Sadler, P., Ethier, N., & Woody, E. (2011). Interpersonal complementarity. In L. M. Horowitz & S. N. Strack (Eds.), Handbook of interpersonal psychology: Theory, research, assessment, and therapeutic interventions (pp. 123–142). Wiley.
Google Scholar
Schein, V. E. (1973). The relationship between sex role stereotypes and requisite management characteristics. Journal of Applied Psychology, 57(2), 95–100. https://doi.org/10.1037/h0037128
10.1037/h0037128
CAS PubMed Web of Science® Google Scholar
Schjelderup-Ebbe, T. (1922). Beiträge zur Sozialpsychologie des Haushuhns [Observation on the social psychology of domestic fowls]. Zeitschrift für Psychologie und Physiologie der Sinnesorgane. Abt. 1. Zeitschrift für Psychologie, 88, 225–252.
Google Scholar
Schoel, C., Bluemke, M., Mueller, P., & Stahlberg, D. (2011). When autocratic leaders become an option—Uncertainty and self-esteem predict implicit leadership preferences. Journal of Personality and Social Psychology, 101(3), 521–540. https://doi.org/10.1037/a0023393
10.1037/a0023393
PubMed Web of Science® Google Scholar
Schoel, C., Stahlberg, D., & Sedikides, C. (2015). Psychological insecurity and leadership styles. In P. J. Carroll, R. M. Arkin, & A. L. Wichman (Eds.), Handbook of personal security (pp. 55–73). Psychology Press.
Google Scholar
Schönbrodt, F. D., & Perugini, M. (2013). At what sample size do correlations stabilize? Journal of Research in Personality, 47(5), 609–612. https://doi.org/10.1016/j.jrp.2013.05.009
10.1016/j.jrp.2013.05.009
Web of Science® Google Scholar
Sedikides, C. (2021). In search of narcissus. Trends in Cognitive Sciences, 25(1), 67–80. https://doi.org/10.1016/j.tics.2020.10.010
10.1016/j.tics.2020.10.010
PubMed Web of Science® Google Scholar
Sedikides, C., & Campbell, W. K. (2017). Narcissistic force meets systemic resistance: The Energy Clash Model. Perspectives on Psychological Science, 12(3), 400–421. https://doi.org/10.1177/1745691617692105
10.1177/1745691617692105
PubMed Web of Science® Google Scholar
Semenyna, S. W., & Honey, P. L. (2015). Dominance styles mediate sex differences in Dark Triad traits. Personality and Individual Differences, 83, 37–43. https://doi.org/10.1016/j.paid.2015.03.046
10.1016/j.paid.2015.03.046
Web of Science® Google Scholar
Shaw, M., Rights, J. D., Sterba, S. S., & Flake, J. K. (2023). r2mlm: An R package calculating R-squared measures for multilevel models. Behavior Research Methods, 55(4), 1942–1964. https://doi.org/10.3758/s13428-022-01841-4
10.3758/s13428-022-01841-4
PubMed Web of Science® Google Scholar
Siemsen, E., Roth, A., & Oliveira, P. (2010). Common method bias in regression models with linear, quadratic, and interaction effects. Organizational Research Methods, 13(3), 456–476. https://doi.org/10.1177/1094428109351241
10.1177/1094428109351241
Web of Science® Google Scholar
Slemp, G. R., Kern, M. L., Patrick, K. J., & Ryan, R. M. (2018). Leader autonomy support in the workplace: A meta-analytic review. Motivation and Emotion, 42(5), 706–724. https://doi.org/10.1007/s11031-018-9698-y
10.1007/s11031-018-9698-y
PubMed Web of Science® Google Scholar
The Ohio State Leadership Studies. (1962). Leader Behavior Description Questionnaire–Form XII Self. Fisher College of Business, The Ohio State University.
Google Scholar
Tiedens, L. Z., & Fragale, A. R. (2003). Power moves: Complementarity in dominant and submissive nonverbal behavior. Journal of Personality and Social Psychology, 84(3), 558–568. https://doi.org/10.1037/0022-3514.84.3.558
10.1037/0022-3514.84.3.558
PubMed Web of Science® Google Scholar
Tiedens, L. Z., Unzueta, M. M., & Young, M. J. (2007). An unconscious desire for hierarchy? The motivated perception of dominance complementarity in task partners. Journal of Personality and Social Psychology, 93(3), 402–414. https://doi.org/10.1037/0022-3514.93.3.402
10.1037/0022-3514.93.3.402
PubMed Web of Science® Google Scholar
Van den Broeck, A., Ferris, D. L., Chang, C.-H., & Rosen, C. C. (2016). A review of self-determination theory's basic psychological needs at work. Journal of Management, 42(5), 1195–1229. https://doi.org/10.1177/0149206316632058
10.1177/0149206316632058
Web of Science® Google Scholar
Van Vugt, M., & De Cremer, D. (1999). Leadership in social dilemmas: The effects of group identification on collective actions to provide public goods. Journal of Personality and Social Psychology, 76(4), 587–599. https://doi.org/10.1037/0022-3514.76.4.587
10.1037/0022-3514.76.4.587
Web of Science® Google Scholar
Van Vugt, M., Jepson, S. F., Hart, C. M., & De Cremer, D. (2004). Autocratic leadership in social dilemmas: A threat to group stability. Journal of Experimental Social Psychology, 40(1), 1–13. https://doi.org/10.1016/S0022-1031(03)00061-1
10.1016/S0022-1031(03)00061-1
Web of Science® Google Scholar
Wegener, D. T., Fabrigar, L. R., Pek, J., & Hoisington-Shaw, K. (2022). Evaluating research in personality and social psychology: Considerations of statistical power and concerns about false findings. Personality and Social Psychology Bulletin, 48(7), 1105–1117. https://doi.org/10.1177/01461672211030811
10.1177/01461672211030811
PubMed Web of Science® Google Scholar
Zeigler-Hill, V., Vrabel, J. K., McCabe, G. A., Cosby, C. A., Traeder, C. K., Hobbs, K. A., & Southard, A. C. (2019). Narcissism and the pursuit of status. Journal of Personality, 87(2), 310–327. https://doi.org/10.1111/jopy.12392
10.1111/jopy.12392
PubMed Web of Science® Google Scholar
Zitek, E. M., & Jordan, A. H. (2016). Narcissism predicts support for hierarchy (at least when narcissists think they can rise to the top). Social Psychological and Personality Science, 7(7), 707–716. https://doi.org/10.1177/1948550616649241
10.1177/1948550616649241
Web of Science® Google Scholar

Citing Literature

Volume93, Issue2

April 2025

Pages 503-523

Which leadership style do more narcissistic subordinates prefer in supervisors?

Abstract

Background and Objective

Method

Results

Conclusion

1 INTRODUCTION

1.1 Narcissism

1.2 Preferences for democratic versus autocratic leadership

1.3 Leadership preferences of more narcissistic subordinates

1.4 Overview

2 STUDY 1

2.1 Method

2.1.1 Participants

2.1.2 Procedure and materials

Narcissism

Impression management

Leadership preferences

2.2 Results and discussion

3 STUDIES 2A–2B

3.1 Method of Study 2a

3.1.1 Participants

3.1.2 Procedure and materials

Supervisor's leadership style

Subordinate's hypothetical leadership style

Attraction to supervisor

3.2 Results of Study 2a

3.2.1 Moderation analysis

3.2.2 Mediation analysis

3.3 Method of Study 2b

3.3.1 Participants

3.3.2 Procedure and materials

Supervisor's leadership style

Subordinate's hypothetical leadership style

Attraction to supervisor

3.4 Results of Study 2b

3.4.1 Moderation analysis

3.4.2 Mediation analysis

3.5 Discussion of Studies 2a–2b

4 STUDY 3

4.1 Method

4.1.1 Participants

4.1.2 Procedure and materials

Supervisor's leadership style

Subordinate's hypothetical leadership style

Attraction to supervisor

Perceived leadership style

4.2 Results and discussion

4.2.1 Perceived leadership style

4.2.2 Moderation analysis

4.2.3 Mediation analysis

5 GENERAL DISCUSSION

5.1 Implications

5.2 Limitations and future research

6 CONCLUDING REMARKS

AUTHOR CONTRIBUTIONS

ACKNOWLEDGMENTS

FUNDING INFORMATION

CONFLICT OF INTEREST STATEMENT

ETHICS STATEMENT

Endnotes

Open Research

DATA AVAILABILITY STATEMENT

Supporting Information

REFERENCES

Citing Literature

Figures

References

Related

Information