Overconfidence and the Pursuit of High-Status Positions: A Test of Two Behavioral Strategies

Abstract

Prior research demonstrates that overconfident people are more likely to attain high-status positions of leadership and influence. However, the underlying motivational and behavioral mechanisms driving this relationship remain largely unexplored. In the present research, we sought to fill this gap in the literature by proposing that overconfidence is associated with stronger status motives and the pursuit of high-status positions via dominance-based strategies. In Studies 1 and 2, we find overconfidence to be positively related to the pursuit of high-status positions of leadership. In Studies 3 and 4, we find overconfident individuals to lean towards dominance- over prestige-based status-seeking strategies. Finally, in Study 4, a field study among real-world supervisor-subordinate dyads, we find an indirect effect of overconfidence on expected social status advancement through dominance. Together, the current studies offer novel insight into the relationship between overconfidence and social status advancement by identifying previously unexplored explanatory mechanisms.

Overconfident people often find themselves appointed to high-status positions of leadership and influence (e.g., Anderson et al., 2012; Belmi et al., 2020; Banerjee et al., 2014; Goel & Thakor, 2008; Reuben et al., 2012; Ronay et al., 2019). Unfortunately, overconfidence and leadership can make for a hazardous combination, at times leading to “the most deleterious of outcomes” (Meikle et al., 2016, p. 129). For instance, overconfident leaders tend to overestimate returns to their investments (Malmendier & Tate, 2008), often fail to detect flaws and deficiencies in their plans (Shipman & Mumford, 2011), are more prone to underestimate the risks of projects (Gervais et al., 2011), more frequently persist in failing investments (Ronay et al., 2017), and are more likely to commit accounting fraud (Schrand & Zechman, 2012). Given these undesirable behaviors, it is important to understand the processes through which overconfident individuals rise to positions of leadership and influence.

Thus far, prior research on the relationship between overconfidence and the attainment of high-status positions has focused on the positive evaluative biases that overconfidence elicits. In particular, researchers have found that overconfident individuals tend to exhibit a number of behaviors (e.g., speaking more often and with a more factual vocal tone, establishing more direct eye contact, bearing a calmer and more relaxed demeanor, or making better use of gestures) that others read as signals of superior competence and leadership potential (Anderson et al., 2012; Belmi et al., 2020; Ronay et al., 2019), thereby favoring social status advancement. However, researchers have also observed that even when people gain clear objective information about overconfident individuals’ actual abilities and level of performance (so revealing them as less competent than previously perceived), the overconfident are still conferred higher status (Kennedy et al., 2013). These findings suggest that the relationship between overconfidence and the attainment of high-status positions is subject to additional explanatory mechanisms beyond exaggerated perceptions of superior competence and abilities. The nature of these mechanisms, nevertheless, has not yet been explored. To address this gap in the literature, we propose an alternative theoretical model, wherein we suggest that overconfidence is associated with stronger status motives and the active pursuit of high-status positions via dominance-based strategies. In order to test these ideas, we examined the relationship between overconfidence and the pursuit of high-status positions (Studies 1 and 2), assessed the relationships between overconfidence and two different status-seeking strategies – dominance and prestige (Studies 3 and 4), and conducted a field study investigating how dominance and prestige position the overconfident to climb their respective social ladders (Study 4).

The current research makes three main contributions to the existing literature. First, we extend current explanations for the relationship between overconfidence and the attainment of high-status positions. Specifically, drawing on self-verification theory (Swann, 1983, 1987, 1990), we propose and demonstrate that overconfident individuals are more prone to enter leadership selection competitions and self-select for leadership roles. Second, we provide the first investigation of the specific status-seeking strategies associated with overconfidence, with a focus on dominance and prestige. Here, by building on and connecting prior overconfidence research with dual-strategies theory (Cheng, 2020; Cheng et al., 2013; Henrich & Gil-White, 2001; Maner & Case, 2016; Van Vugt & Smith, 2019), we predict and find support for the idea that overconfident individuals tend to adopt dominance- over prestige-based status-seeking strategies. Third, by capturing supervisor- and subordinate-reports of dominance and prestige and mapping these onto supervisors’ expectations of subordinates’ social status advancement over time, we extend prior empirical research on dominance and prestige as strategies for navigating social hierarchies. To date, existing evidence in this regard is limited to lab studies involving short-term groups and hierarchies (Cheng et al., 2013; de Waal‐Andrews et al., 2014) and one longitudinal test among student groups (Redhead et al., 2019). The present research is thus the first to examine the influence of dominance and prestige within stable, longer-term hierarchies in actual organizational contexts.

Theoretical Framework

Overconfidence

In broad terms, overconfidence can be defined as an exaggerated perception of one’s competence or capabilities (see Moore & Healy, 2008 for a review). Importantly, overconfidence is not the same as self-confidence, which refers to an accurate, positive evaluation of one’s knowledge, skills, and abilities (Hollenbeck & Hall, 2004). Self-confidence is a strongly desired trait for leaders and people in general (Al-Hebaish, 2012; Bass, 1990; Magee & Frasier, 2014; Yukl, 2002). Overconfidence, on the other hand, occurs when individuals inaccurately evaluate themselves, believing and displaying a level of self-confidence that exceeds their real capabilities. Unlike well-calibrated self-confidence, overconfidence is generally regarded as an undesirable trait, especially among those in powerful positions (Meikle et al., 2016).

Overconfidence is also different from impression management or self-presentation, which involve the purposeful modification of one’s overt behavior with the intent of creating a desirable social image (Baumeister, 1982; Buehl et al., 2019; Goffman, 1959; Leary & Kowalski, 1990; Paulhus, 1984). Those who manage their impressions might or might not believe their own stories, and these stories might or might not reflect their true self (Bourdage et al., 2018), whereas overconfident individuals hold a genuine though overly positive image of their competence and abilities.

Overconfidence and the Pursuit of High-Status Positions

Self-verification theory (Swann, 1983, 1987, 1990) posits that people harbor a fundamental need to confirm and stabilize their self-views, as stable self-views provide individuals with a sense of psychological coherence that enables them to make sense of their worlds, predict future events, and guide their behavior (e.g., Cooley, 1902; Lecky, 1945; Mead, 1934). Indeed, numerous studies over the last three decades show that individuals tend to adopt behaviors that facilitate self-verification (e.g., Bosson & Swann, 1999; Cable & Kay, 2012; Kraus & Chen, 2014; Swann et al., 1994) and that their social functioning and psychological well-being flounder when self-views are compromised (e.g., Ayduk et al., 2008; Wood et al., 2005; see North & Swann, 2009 for a review).

Because self-views are shaped via social and interpersonal processes, self-verification theorists have stressed that individuals can confirm and stabilize their self-views only insofar as they receive self-verifying social responses from others and their social environments (Swann et al., 2002; Swann & Buhrmester, 2012; Swann & Read, 1981). For instance, someone who sees themselves as particularly competent might find it difficult to maintain their self-image if others do not seek or value their opinions, or if their professional achievements do not match self-perceived capabilities. Hence, according to self-verification theory (Swann, 1983, 1987, 1990), overconfident individuals should strive to engage in behaviors that help them confirm and stabilize their exaggerated self-views, and these efforts should be primarily expressed in the social domain.

Following these theoretical tenets, we point to professional contexts as one important social domain through which overconfident individuals might seek to validate their exaggerated self-views. In particular, we argue that the acquisition of greater power and social status in professional contexts represents a viable route for the verification of overconfident self-views, as both power and status have a strong influence on self- as well as social perceptions of competence and abilities. For example, Fast et al. (2012) provide experimental evidence that individuals’ confidence in their own competence increases after priming their subjective sense of power, and Belmi et al. (2020) found that higher social status, whether based on objective or subjective measures, was associated with overestimation of one’s actual competence, leading to the perception that one is better than others. Similar effects have been reported in the interpersonal sphere – Fiske et al. (2002) reported strong correlations of up to 0.88 between individuals’ social status and others’ perceptions of their competence, and Brambilla et al. (2010) found that prominent professional roles are associated with social perceptions of superior competence.

In sum, high-status positions tend to imbue people with an exaggerated perception of their own competence and abilities, and lead others to concur with such perception. This, we propose, might prompt overconfident individuals to the pursuit of socially prominent positions as a means to validate and reinforce their aggrandized self-image. Consistent with this proposition, self-verification theorists have argued that verification of one’s self-views can be accomplished through the position that individuals hold within social and organizational structures (e.g., Seyle & Swann, 2007).

• Hypothesis 1: Overconfidence is positively related to the pursuit of high-status positions.

Overconfidence, Dominance, and Prestige – Different Strategies Towards High-Status Positions

Dual-strategies theory (Cheng, 2020; Cheng et al., 2013; Henrich & Gil-White, 2001; Maner & Case, 2016; Van Vugt & Smith, 2019) is a theoretical framework that posits dominance and prestige as the two general and fundamental strategies people use to attain or maintain higher social rank. Dominance refers to the pursuit of high status through aggressive and forceful behaviors, such as intimidation, manipulation, or coerciveness. In the organizational context, dominance is exemplified via non-beneficial competition or self-serving use of common resources. Prestige, as opposed, refers to higher status “granted to individuals who are recognized and respected for their skills, success, or knowledge” (Cheng et al., 2010, p. 335). In organizational contexts, prestige is earned by successfully executing especially difficult tasks, or by assisting others and winning their respect and admiration. Despite representing qualitatively distinct strategies, both dominance and prestige have been shown to be effective for attaining higher social status within groups (e.g., Cheng et al., 2013). This creates an interesting juncture for the overconfident.

On the one hand, if overconfident individuals harbor a fulsome and unwavering belief in the superior nature of their competence and abilities, then we might expect the overconfident to follow a prestige-based route in their pursuit of higher social rank (i.e., presenting themselves as possessing the right knowledge and necessary abilities to advance collective goals). On the other hand, however, because overconfidence implies an exaggerated sense of competence and abilities, we reason that overconfident individuals might in practice fail to earn higher status via others’ respect and admiration, hence rendering prestige-based strategies towards higher social status ultimately ineffective. Indeed, in their review of the literature, Meikle et al. (2016) highlight how overconfident individuals can severely affect organizational functioning and performance through risky behavior and poor decision-making. Not only are these outcomes inconsistent with the image of superior competence that grants prestige-based status, they can also yield negative social reactions and thwart social advancement. For instance, Tenney et al. (2019) provide experimental evidence that when overconfident individuals’ self-confidence is revealed unwarranted, they are subsequently deemed less competent and less desirable collaborating partners than equally competent though humbler competitors.

Crucially, researchers have also found that overconfident individuals continue to attain higher-status positions despite such disappointing outcomes and negative impressions of competence (Kennedy et al., 2013), leading to question what mechanisms (other than misperceptions of superior competence) underlie the relationship between overconfidence and status attainment. One possible explanation is that overconfident individuals might strategically lean towards dominance- over prestige-based strategies in efforts of materializing their status aspirations – as the status-enhancing effects of dominance, unlike those of prestige, are not dependent on confirmed performance expectations (de Waal-Andrews et al., 2014) and actual contributions to the group (Cheng et al., 2013; Henrich & Gil-White, 2001; Maner & Case, 2016). Through dominance, individuals can successfully build routes towards higher-status positions via strategies that do not necessarily rely on others’ perceptions of superior competence and intrinsic value, such as the hiding and manipulation of relevant information, a self-serving use of valuable common resources, or by forming coalitions in order to control and coerce others (Maner, 2017). Therefore, we expect overconfident individuals to construe dominance-based strategies as better suited than prestige to serve their status aspirations.

Our theoretical argument that overconfidence is associated with dominance-based status-seeking strategies is consistent with numerous empirical findings on the attitudinal and behavioral profiles of the overconfident. For example, overconfident individuals tend to harbor individualistic worldviews in which one’s importance, interest, and well-being are emphasized over the group (e.g., Antonczyk & Salzmann, 2014), making them more prone to unethical behavior (Park & Chung, 2016). In the same line, prior studies among organizational managers have shown that overconfidence is related to executive hubris (Simon & Houghton, 2003), which often involves the imposition of one’s will through power and dominance (Delbecq, 2006). These attitudinal and behavioral tendencies represent fundamental features of dominance-oriented individuals, who similarly prioritize their self-interest over the public good, are arrogant and other-demeaning, and tend to resort to aggressive tactics in order to achieve their goals (e.g., Cheng et al., 2010; Maner & Mead, 2010; 2012). In contrast, prestige-oriented individuals tend to be humbler, self-deprecating, and exhibit prosocial traits such as altruism, agreeableness, helpfulness, strong moral convictions, and a concern for the public good (Cheng et al., 2010; Maner & Mead, 2010, 2012; Tracy & Robins, 2004; Weidman et al., 2018).

• Hypothesis 2: Overconfidence is positively related to dominance- but not prestige-based status-seeking strategies.

Researchers have consistently demonstrated the status-enhancing effects of dominance. For example, in a classic meta-analytic review of the literature, Lord et al. (1986) found the personality trait dominance, defined as the propensity towards forceful, assertive, and aggressive behaviors, to explain a substantial part of the variance in leadership perceptions. More recently, several studies have shown aggression, forcefulness, exploitation, and manipulation – behaviors typically associated with dominance (Cheng et al., 2013; Henrich & Gil-White, 2001; Maner & Case, 2016) – to represent effective tactics for improving one’s relative social status (e.g., Anderson & Kilduff, 2009; Cillessen & Mayeux, 2004; Grapsas et al., 2020; Kyl-Heku & Buss, 1996). Indeed, in a series of group- and dyadic-based studies, Cheng et al. (2013) observed that dominant individuals (based on self-reports, peer-reports, and outside observer-reports) tend to receive greater influence and social attention. Consequently, we expect dominance to feature as a mediating mechanism between overconfidence and social status advancement (Fig. 1).

• Hypothesis 3: Overconfidence has an indirect effect on social status advancement through dominance.

Fig. 1

Proposed Theoretical Model

Study 1

The goal of Study 1 was to provide an initial test of the relationship between overconfidence and the pursuit of high-status positions (Hypothesis 1). To do so, we captured individual differences in overconfidence and presented participants with a decision on whether or not to enter a selection procedure for a high-status leadership role within the context of an upcoming group task.

Method

Participants and Procedure

Based on an a priori power analysis for logistic regression analyses using G*Power software (Faul et al., 2007), we estimated a required sample size of at least N = 114 to achieve 80% power to detect a small effect (r = 0.18) with an alpha of 0.05. The initial sample consisted of 200 Amazon Mechanical Turk workers from the United States of America. Of these, we removed 45 cases due to incomplete responses, resulting in final sample size of 155 participants (56.8% male, M_age = 40.26, SD = 12.06). Most of them held at least a bachelor’s degree (69%), with the remaining having completed either high school education (29.7%) elementary education (0.6%), or no education (0.6%).

Participants first provided informed consent and completed a few demographic questions. To assess overconfidence, we then asked participants to complete the General Knowledge Questionnaire (GKQ; Michailova & Katter, 2014). Subsequently, we informed participants that the current study had finished and that we were interested in recruiting a limited number of participants for a series of upcoming studies. In these ostensible future studies, participants would have to act as leaders of a group of five, resembling a small organizational team. Their task would be to assess and offer a solution for an organizational challenge described to them, with the leaders having the final decision concerning their team’s submitted solution. Participants were also informed that leaders would receive twice the financial compensation for participating as leaders in such studies. However, in order to be considered for the role, candidates would need to complete a leadership selection procedure. Participants read the following fragment:

“If you want to be considered as a leadership candidate for the mentioned studies, we will need you to complete several selection tests. The selection procedure takes around 30 minutes. Based on the selection test scores, only 5% of the candidates will be selected. If you are not selected, you will not be paid for the time invested. If you are selected, however, your participation in the upcoming studies will be paid 2x the standard payment in Amazon Mechanical Turk.” If you decide to apply, please selected the option “Yes, I would like to apply for a leader participation in the upcoming studies” and in the following pages we will need you to go through a selection procedure in which you will:

• complete a leadership style questionnaire (~ 5 min)

• complete a previous work experience questionnaire (~ 5 min)

• explain in 3–4 sentences how you would act in response to several hypothetical situations that we will describe to you (~ 10 min)

• write a short text (between 100 and 200 words) in which you explain why you think you would be a good leader and why we should select you over other candidates (~ 10 min)

The description of the leadership role, the selection procedure, and the financial conditions were all intended to resemble those of real-world organizations, where high-status positions of leadership and influence come with higher salaries but are also more difficult to reach and involve greater responsibilities. The specific nature of the organizational-like challenge to solve was purposely kept vague in order to avoid differentially priming participants with different work-experience backgrounds. Participants indicated their decision on whether or not to enter the leadership selection procedure and were then redirected to a debriefing page, where we informed them of the research purpose and the ostensible nature of the selection procedure and future studies.

Measures

Overconfidence

To capture overconfidence, we used a previously validated version (Ronay et al., 2017) of the General Knowledge Questionnaire (GKQ; Michailova & Katter, 2014). The questionnaire consists of 24 general knowledge questions (e.g., “How many letters does the Russian alphabet consists of?”) with three given alternative answers each (e.g., “40 letters”, “33 letters”, or “26 letters”) from which participants must choose the correct one. For each question, participants were also asked to provide a number between 33% (“no certainty at all in the correctness of my answer, just guessing”) and 100% (“absolute certainty in the correctness of my answer”) indicating their level of confidence in their answer. Overconfidence was operationalized participants’ level of confidence while controlling for their actual performance.

Pursuit of High-Status Positions

To capture the pursuit of high-status positions, we presented participants with the option of entering a selection procedure for the leadership role described above. Participants were asked to select one of the two following possible responses: “Yes, I would like to apply for a leadership role in the upcoming studies” or “No, thanks, I do not wish to apply”. In total, 55 (35.5%) participants of the 155 that participated decided to go through the selection procedure and thus compete for one of the leadership roles.

Results

Table 1 provides means, standard deviations, and bivariate correlations among study variables. To assess the relationship between overconfidence and the pursuit of high-status positions, we ran a logistic regression analysis, introducing participants’ overconfidence as the predictor and their decisions on whether or not to enter the leadership selection procedure as the dependent variable. The results revealed a positive and significant relationship between overconfidence and the pursuit of high-status positions, β = 0.04, SE = 0.01, Nagelkerke R² = 0.10, Wald’s χ² (1) = 8.69, Exp(B) = 1.046, 95%CI[1.015,1.078], p < 0.01. In other words, participants were 3.4% more likely to enter the leadership selection procedure per each unit increase in their overconfidence score. Overall, those who decided to compete for the leadership position scored significantly higher on overconfidence (M = 13.12, SD = 20.69) relative to those who decided not to pursue the higher-status leadership position (M = 4.03, SD = 13.14), t(78.53), p < 0.01. To rule out the possibility that participants’ prior leadership experience had played a role in their decisions, we conducted again the same analyses while controlling for participants prior experience working as a leader (in years) and obtained similar results, β = 0.03, SE = 0.01, Nagelkerke R² = 0.11, Wald’s χ² (1) = 5.65, Exp(B) = 1.039, 95%CI[1.007,1.072], p = 0.01.

Discussion

Consistent with Hypothesis 1, we found participants’ overconfidence to be positively associated with the pursuit of high-status positions, operationalized as their decision to compete for a leadership role. This finding offers novel insight into the relationship between overconfidence and the attainment of high-status positions, which researchers have so far explained via the positive evaluative biases that overconfidence tends to elicit (i.e., misperceptions of higher competence) (Anderson et al., 2012; Belmi et al., 2020; Ronay et al., 2019). Specifically, our results point to an alternative (yet compatible) explanatory mechanism – i.e., the overconfident appear more inclined to actively pursue such prominent positions. From an organizational perspective, the present finding suggests that leadership selection panels might commonly encounter pools of overconfident leadership candidates, as it is the overconfident who are most likely to apply for these positions.

Despite our initial supporting evidence, it is important to note two possible alternative explanations for our results. First, because Amazon Mechanical Turk is a platform where people enroll for financial rewards, it is difficult to differentiate between participants’ motive to earn more (i.e., × 2 payment) from the motive to pursue high-status positions. Although we made this decision purposely to better resemble real-world scenarios (where high-status positions usually come with greater financial rewards), we acknowledge that a potential motivational confound limits the empirical support that Study 1 provides for our theoretical proposition – the overconfident pursue high-status positions not for financial gains (or at least not exclusively) but as an instrument to nourish their idealized self-image of superior competence. Therefore, overconfident individuals should pursue high-status positions irrespective of whether such positions come with associated financial gains. Second, because overconfidence leads to increased optimism and illusions of control (e.g., Cain et al., 2013; Camerer & Lovallo, 1999; Coelho, 2010; Durand, 2003; Moore & Cain, 2007), overconfident participants might have simply felt more encouraged to take their chances in a challenging leadership selection procedure from which only the top 5% applicants are selected. In Study 2, we thus sought to rule out these possible alternative explanations.

Study 2

The goal of Study 2 was to provide a second test of the prediction that overconfidence is associated with the pursuit of high-status positions (Hypothesis 1), this time adopting a methodological design that allowed us to rule out the aforementioned alternative explanations. To do so, we generally mirrored the design of Study 1 but implemented two modifications. First, we removed the financial bonus associated with attaining the leadership role. Second, we removed the selection procedure, asking instead participants to indicate the strength of their preference for the leader versus the team member in the subsequent group task. These changes intended to isolate the motive towards high-status positions from confounding financial motives, as well as to eliminate the potential effects of increased optimism and illusions of control associated with overconfidence.

In addition to these modifications, we also extended our methodological design in two manners. First, the proposition that attaining high-status positions can serve a self-verifying role for the overconfident implies that overconfident individuals perceive such prominent positions as reflecting their self-image. To test this idea, we included an additional measure to capture the extent to which overconfident people indeed perceive an overlap between their self-appraised capacities and those required to occupy high-status positions of power and influence. Second, we adapted the version of the General Knowledge Questionnaire (Michailova & Katter, 2014) that we used in Study 1 in a way that allowed us to capture and differentiate between two different manifestations of overconfidence – overestimation and overplacement (see Moore & Healy, 2008 for a review). Overestimation refers to an exaggerated belief in one’s abilities, performance, and chances of success, and overplacement refers to the exaggerated belief that one is better than others. In this sense, while Moore and Healy (2008) originally described overestimation and overplacement as representing theoretically and empirically different forms of overconfidence (as opposed to different manifestations of the same underlying construct), a more recent and comprehensive investigation involving longitudinal time-lagged data found no support for this idea, concluding instead that there is a dispositional component to overconfident beliefs and that the two forms are empirically related (Lawson et al., 2023). By capturing both overestimation and overplacement, our goal was to explore these contrasting pieces of evidence within the context of our own investigation.

Method

Participants and Procedure

Based on an a priori power analysis for logistic regression analyses using G*Power software (Faul et al., 2007), we again estimated a minimum sample size requirement of N = 114 to achieve 80% power to detect a small effect (r = 0.18) with an alpha of 0.05. The initial sample consisted of 188 Amazon Mechanical Turk workers from the United States of America. Of these, we removed 32 cases due to incomplete responses, resulting in a final sample size of 156 participants (57.7% male, M_age = 38.01, SD = 10.39). Most of them held at least a bachelor’s degree (87.0%), with the remaining having completed either high school education (11.5%) or elementary education (0.6%).

Participants first provided informed consent and completed a few demographic questions. To assess overconfidence, we asked participants to complete the General Knowledge Questionnaire (Michailova & Katter, 2014). Upon finalizing the questionnaire, participants were presented with the same ostensible task of assessing and offering possible solutions to an organizational-like challenge, again in groups of five participants with one of them acting as the leader. Once participants had read the task instructions, we asked them to indicate the strength of their preference for the leader role versus the team member role in the subsequent task. Finally, we also asked participants to indicate the extent to which they believed that the leader role reflected their level of competence and abilities. Participants provided their responses and were then redirected to a debriefing page, where we also informed them of the research purpose and the ostensible nature of the group task.

Measures

Overconfidence

To capture overconfidence, we used the same previously validated version (Ronay et al., 2017) of the General Knowledge Questionnaire (GKQ; Michailova & Katter, 2014) that we had used in Study 1. However, to capture overestimation and overplacement as different operationalizations of overconfidence, this time we requested participants to indicate their confidence in two different ways. For overestimation, we asked them to indicate “how well they thought they had performed overall in the test”, on a scale ranging from 0 (“poorly, I think I answered all the questions incorrectly”) to 100 (“greatly, I think I answered all the questions correctly”). For overplacement, we asked them to indicate “how well they thought they had performed overall in the test relative to the other participants in the study”, on a scale ranging from 0 (“I am at the very bottom, I think that I got a worse final score than 99% of the other participants”) to 100 (“I am at the very top, I think that I got a better final score than 99% of the other participants”). Note that we moved from asking participants about their confidence at the item-level (Study 1) to the overall-level (Study 2), as measuring participants’ perceptions of performance relative to others (i.e., overplacement) at the item-level leaves little room for variance – on each single question, participants can only be either right or wrong. This approach is common in the existing literature (e.g., Anderson et al., 2012; Belmi et al., 2020; Logg et al., 2018), as it allows to better capture individual differences in overplacement. Overestimation was computed as participants’ level of confidence while controlling for their average accuracy, and overplacement was computed as participants’ self-perceived rank among all participants while controlling for their actual rank. To obtain participants’ actual rank, we transformed their scores into percentile rankings. Difference scores were also calculated in order to examine the correlation between overconfidence manifestations (i.e., overestimation and overplacement). Consistent with Lawson et al.’s (2023) findings, overestimation and overplacement were highly correlated ( r = 0.79, p < 0.001) and showed comparable relationships with all other variables, including our main dependent variable (i.e., pursuit of high-status positions). Similar correlations were observed in all subsequent studies (r’s ranging from 0.61 to 0.79). Therefore, we computed an overall overconfidence score by averaging participants’ overestimation and overplacement and used it as our predictor across studies and analyses. The results for overestimation and overplacement as independent predictors can be consulted in the supplementary materials.

Pursuit of High-Status Positions

To capture the pursuit of high-status positions, we relied on the same upcoming and ostensible group task described in Study 1, asking participants to indicate their preference for the leader versus team member role. To allow for a more precise measure of participants’ status-motives, this time the role preference was expressed via a slide bar ranging from 0 (“I absolutely prefer to participate as a team member”) to 100 (“I absolutely prefer to participate as a team leader”), hence moving from a dichotomous to a continuous dependent measure. Indeed, while two participants of Study 1 might have both opted to apply for the higher-status leadership role, they might have still differed to some extent in how much they desired such role. Transitioning to a continuous measurement of participants’ high-status motives allows us to better detect such potential differences (Cohen, 1983; DeCoster et al., 2009; MacCallum et al., 2002).

High-Status Positions as Self-Verifying

To examine whether the attainment of high-status positions can serve a self-verifying role for overconfident individuals, we asked participants to indicate the extent to which they believed that the leader role reflected their actual level of competence and abilities, on a scale ranging from 0 (“not at all”) to 100 (“very much so”).

Results

Table 1 provides means, standard deviations, and bivariate correlations among participants’ overconfidence, perceived match between their self-appraised competence and abilities and those required for the leadership role, strength of preference for the leadership versus the team member role, and perceived match between their self-appraised competence and abilities and those required for the leadership role.

To assess the relationship between overconfidence and the pursuit of high-status positions, we first regressed participants’ preference for the leader versus team member role onto their overconfidence scores. The results revealed a positive and significant relationship between overconfidence and the pursuit of high-status positions, β = 0.30, b = 0.31, SE = 0.08, 95%CI[0.15,0.47], t(154) = 3.91, p < 0.001.

In order to test the argument that high-status positions can serve a self-verification role for overconfident individuals, we then regressed participants’ perceived overlap between their competence and abilities and those required for the leadership role onto their overconfidence scores. The results revealed a positive and significant relationship between participants’ overconfidence and the belief that the leadership role reflected their actual level of competence and abilities, β = 0.36, b = 0.37, SE = 0.07, 95%CI[0.22,0.53], t(154) = 4.81, p < 0.001.

Finally, to further test our theorizing, we also explored a potential mediation pathway between overconfidence and the pursuit of high-status positions via participants’ perceived overlap between their self-appraised capacities and those required for the high-status leadership position. To do this, we used Model 4 of Process macro (Hayes, 2017), fitting participants’ overconfidence as the predictor, their pursuit of high-status positions as the dependent variables, and the perceived overlap in capacities as the mediator. This revealed a positive and significant indirect effect of overconfidence on the pursuit of high-status positions through participants’ perceived overlap between their self-appraised capacities and those required for the high-status leadership position, IE = 0.30, SE = 0.05, 95%CI[0.20,0.41]. To check the robustness of our theoretical logic, we also examined the reversed model – that is, introducing participants’ belief that the high-status leadership role reflected their self-perceived level of competence and abilities as the predictor and overconfidence as the mediator. The results revealed no significant indirect effect between participants’ belief that the high-status leadership role reflected their self-perceived level of competence and abilities and the pursuit of the high-status leadership position through overconfidence, IE = 0.001, SE = 0.014, 95%CI[-0.028,0.029]. While this does not completely mitigate the causality concerns of single-time point measures, it lends support for the proposed mediation model.

Discussion

Consistent with Hypothesis 1, we again found participants’ overconfidence to be positively associated with the pursuit of high-status positions, this time operationalized as the strength of their preference for a leader versus a team member role. The results of Study 2 thus replicate those of Study 1 whilst ruling out a potential confound between financial and purely status-seeking motives, as well as the possible alternative explanation that overconfident individuals are simply more optimistic about their chances of landing a prominent role in a heavily contested selection procedure. Indeed, within the Study 2 setting, taking on a leadership role was a matter of unrestricted choice.

Study 2 also offered support for our theoretical argument that overconfident individuals’ pursuit of high-status positions follows a self-verification logic. Specifically, we found that the overconfident perceive high-status leadership positions as reflective of their self-image in terms of competence and abilities. This finding represents an important addition to the results of Study 1, as self-verification theory (Swann, 1983, 1987, 1990) describes selective affiliation (i.e., the effort to interact with and be member of groups whose social image reflects one’s self-perceived image) as one means through which people seek to validate their self-views.

Studies 1 and 2, however, leave open the question of exactly what suit of strategies the overconfident tend to adopt in the pursuit of high-status positions. Indeed, while we observed overconfident individuals to be more likely to enter leadership selection competitions (Study 1) and self-select for leadership roles (Study 2), status-seeking strategies often precede and extend beyond these specific scenarios. For instance, in organizational contexts, employees who aspire to move up the ranks need to position themselves as potential leadership candidates (Schleu & Hüffmeier, 2021), and even well-established leaders might fall from grace (e.g., Bendersky & Shah, 2013). Understanding the specific strategies that overconfident individuals adopt would facilitate the possibilities to recognize and regulate the potential consequences of their actions (Heckhausen & Heckhausen, 2008). Therefore, in Study 3, we sought to investigate the specific strategies emerging from overconfident individuals’ status aspirations.

Study 3

The goal of Study 3 was to provide a first test of our prediction that overconfidence is positively associated with dominance-based status-seeking strategies (Hypothesis 2). To do so, we measured participants’ overconfidence and then offered them two consecutive behavioral choices designed to capture individual differences in dominance orientation.

First, we simultaneously presented participants with two organizational charts depicting either a horizontal and a vertical organizational structure, and asked them to indicate the extent to which they preferred to opt for a high-status leadership position in the horizontally- versus the vertically-structured organization. The logic underpinning this paradigm rests on prior theoretical (Ronay et al., 2020; van Vugt & Ronay, 2013; von Hippel et al., 2016) and empirical (Rus et al., 2012) studies suggesting that steep organizational structures (where power is usually concentrated in a small number of people at the hierarchical top) provide greater freedom and opportunities for individuals to engage in dominance-based status-seeking strategies (e.g., intimidation, forcefulness, or coercion), especially for those in powerful positions. Therefore, to the extent that overconfident individuals indeed harbor an inclination towards dominance, we should observe a preference among the overconfident for high-status positions in organizational contexts where such strategies can be better expressed (i.e., vertically-structured organizations).

Subsequently, we presented participants with a fictional leadership scenario built around the dictator game – a classic economic game (cf. Camerer, 2003) wherein participants are instructed to distribute an economic endowment between themselves (i.e., the allocator) and another individual (i.e., the recipient). In essence, the dictator game presents participants with the decision to act in a self-interested manner or to forgo self-interest in order to preserve the welfare of others (e.g., Forsynthe et al., 1994; Galinsky et al., 2015; Hilbig et al., 2015). Since dominance revolves around self-serving behaviors at the expense of others, the dictator game represents an ideal paradigm to capture one of the most fundamental aspects of dominance – i.e., self-serving manipulation of group resources (Cheng et al., 2013; Henrich & Gil-White, 2001; Maner & Case, 2016). Hence, to the extent that overconfident individuals are indeed inclined towards dominance, we should observe a self-interested allocation of the financial endowment.

Method

Participants and Procedure

Based on an a priori power analysis for linear regression analyses using G*Power software (Faul et al., 2007), we again estimated a required sample size of at least N = 150 to achieve 80% power to detect a small effect (r = 0.20) with an alpha of 0.05. Participants were recruited through an open survey link that was advertised on different social media sites (i.e., Facebook, Instagram, and WhatsApp) for a period of two weeks. These various sites were selected in order to increase demographical variance in our sample, as prior research indicates differences in their user base (e.g., in age, gender, and educational level; Gambo & Özad, 2020). To further increase demographical heterogeneity, the advert also encouraged people to distribute the study link within their own personal networks (i.e., snowball sampling technique). In total, 243 individuals accessed the link and started participation. Of these, we removed 78 cases due to incomplete responses, resulting in final sample size of 165 individuals (64.1% female, M_age = 32.01, SD = 12.86) from the general population. Most participants had at least a bachelor’s degree (61.1%), with the remaining having completed either high school education (36.5%), elementary education (7.1%), or no education at all (0.6%).

Participants first provided informed consent and completed a few demographic questions. In order to assess overconfidence, we asked participants to complete the General Knowledge Questionnaire (Michailova & Katter, 2014). Once participants had completed the questionnaire, we presented them with the two scenarios described above, subsequently and respectively – i.e., asking participants to express the strength of their preference for the horizontal versus the vertical organizational structure in the context of seeking a leadership position; and, once effectively in a leadership role, to allocate a financial endowment between themselves (as leaders) and others (i.e., their subordinates) (see below for more details) – with the former question carrying into the latter. Participants provided their responses and were then redirected to a debriefing page.

Measures

Overconfidence

To capture overconfidence, we used the same previously validated (Ronay et al., 2017) and modified version of the General Knowledge Questionnaire (GKQ; Michailova & Katter, 2014) that we used in Study 2.

Dominance Orientation Measure 1

Our first measure consisted of observing whether overconfidence predicts a preference for “dominance-favoring” organizational contexts. More specifically, we instructed participants to imagine that they were currently seeking a leadership opportunity within the context of their professional career, and showed them the organizational charts of two ostensible organizations with vacant leadership positions – APZ Singular and Blue Skylark. Participants were told that these organizations were real and that the images of the organizational charts that we showed them were screenshots directly taken from the organizations’ actual webpages. However, the organizational charts were manipulated to reflect either a horizontally- or a vertically-structured organization (see Fig. 2 for the images of the organizational charts that participants saw). These organizational charts have been previously validated (Zitek & Tiedens, 2012) and adapted (Fath & Kay, 2018) as experimental manipulations of equal versus unequal distributions of power within organizations. In order to ensure that participants unequivocally understood what these structures implied in terms of intraorganizational dynamics, we accompanied the images of the organizational charts with the following fragments:

Fig. 2

Hierarchical Structures (Vertical vs. Horizontal) Shown to Participants in Study 3

For the horizontal organization: “APZ-Singular has a horizontal and decentralized organizational structure. This means that in this organization power and decision-making processes are relatively equally distributed among departments and organizational members. Therefore, leaders in this organization have less unilateral power over organizational dynamics relative to those who occupy lower-ranking positions.

For the vertical organization: “Blue Skylark has a vertical and centralized organizational structure. This means that in this organization power and decision-making processes are concentrated in a relatively small number of departments and organizational members at the top of the organizational hierarchy. Therefore, leaders in this organization have more unilateral power over organizational dynamics relative to those who occupy lower-ranking positions.

Participants were asked to indicate the strength of their preference for these organizations via a slide bar ranging from 0 (“I would absolutely prefer to opt for a leadership position at APZ-Singular – horizontal organization”) to 100 (“I would absolutely prefer to opt for a leadership position at Blue Skylark – vertical organization”). Because the organizations differed in their names as well as their brand colors (see Fig. 2), we crossed the two organizations with the two hierarchical structures and randomly presented to participants one of the possible resulting versions. In total, 83 participants saw APZ-Singular as the horizontal organization and Blue Skylark as the vertical organization, while the remaining 73 participants saw the reverse combination.

Dominance Orientation Measure 2

Our second measure consisted of observing whether participants’ overconfidence predicts authoritarian, self-serving behavior when presented with the power and opportunity to do so. To do this, we instructed participants to think of themselves as leaders of an organizational team consisting of four people that had just earned a performance-based financial bonus, and to distribute this bonus between themselves, as leaders, on one side, and their subordinates, on the other side, as team members. The financial bonus was therefore divided in two parts, with participants indicating the amount of the bonus that they wanted to reserve for themselves. More precisely, participants read the following text:

“Imagine that you are working at an organization where you are the leader of a team of four people (yourself and three subordinates). Your team just successfully completed a project and earned a $4000 preestablished financial bonus for the good performance. As the team leader, you directed the team and were in charge of the final decisions, but everyone strived in their respective tasks and did a good job. Since you are the leader of the team, the organization leaves it up to you to decide how to distribute the $4000 bonus between yourself and the rest of the team.”

The purpose of this scenario was to harness the logic of the dictator game while framing it within an organizational context. Participants were asked to indicate “how much of the $4000 bonus you would keep for yourself in such scenario” and indicated their responses via a slide bar that could be set at any value between $0 and $4000.

Results

Table 2 provides means, standard deviations, and bivariate correlations among participants’ overconfidence, preference for the horizontal versus the vertical organizational structure, and the amount of the teams’ financial bonus that they decided to keep for themselves.

First, we examined the relationship between overconfidence and the preference for horizontal versus vertical organizational structures. To do so, we regressed participants’ preference for these two organizational types onto their overconfidence scores. The results revealed a positive and significant relationship between overconfidence and the preference for vertical organizational structures, β = 0.17, b = 0.27, SE = 0.12, 95%CI[0.02,0.52], t(154) = 2.17, p = 0.03.

Second, we examined the relationship between overconfidence and self-serving behavior as expressed via the dictator game. To do so, we regressed the amount of the financial bonus that participants decided to keep for themselves onto their overconfidence scores. The results revealed a positive and significant relationship between overconfidence and self-serving allocation of the financial bonus, β = 0.32, b = 10.22, SE = 2.42, 95%CI[5.45,15.05], t(154) = 4.22, p < 0.001.

Discussion

Consistent with Hypothesis 2, we found overconfidence to be positively related to dominance-based status-seeking strategies, alternatively operationalized as participants’ (1) preference for vertical over horizontal organizational structures as well as (2) self-serving allocation of group resources. In other words, we observed a preference among the overconfident for organizational contexts favoring dominance through unilateral power, as well as self-serving use of such power, both of which have been amply documented as characteristic of dominance-based processes (e.g., Cheng et al., 2013; Henrich & Gil-White, 2001; Maner & Case, 2016; Ronay et al., 2020, Rus et al., 2012: van Vugt & Ronay, 2013; von Hippel et al., 2016). Therefore, the results of Study 3 provide initial evidence for the theorized relationship between overconfidence and dominance. To finalize the present investigation, in Study 4 we sought to replicate our findings among a sample of working professionals (i.e., supervisor-subordinate dyads) where we could (1) more directly compare dominance versus prestige strategies, (2) in a more ecologically valid context, (3) and examine the proposed indirect effect of overconfidence on social status advancement through dominance (Hypothesis 3).

Study 4

The goal of Study 4 was threefold. First, we wanted to replicate our findings in Study 3 by providing a second test of the proposed relationship between overconfidence and dominance – with our prediction again being that overconfidence is associated with dominance but not prestige (Hypothesis 2). Second, we wanted to examine the influence of dominance and prestige in terms of status-attainment within real-world organizational contexts. This is an important and necessary complement to the results of Study 3, as researchers have stressed how organizational research, especially on power and status, has typically over-relied on fictional, context-poor settings that fail to reflect the complexities of real-world organizational processes (Schaerer et al., 2018), leading to “little or no ecological validity” (Sturm & Antonakis, 2015, p. 150). Based on prior reports (Brand & Mesoudi, 2019; Cheng et al., 2013; de Waal‐Andrews et al., 2014), we expected both dominance and prestige to be positively related to status attainment. However, based on the results of Study 3 and as per Hypothesis 2, we expected dominance to represent the behavioral tactic preferred by the overconfident. Lastly, we wanted to test our prediction that overconfidence has an indirect effect on social advancement through dominance (Hypothesis 3). To these ends, we recruited a field sample of direct supervisor-subordinate dyads and measured (1) subordinates’ overconfidence and self-reported dominance and prestige, as well as (2) supervisors’ perceptions of their subordinates’ dominance and prestige and expected changes in social status within the next five to ten years.

This design also allowed us to offer the first examination with regards to the correlations between people’s self-reported dominance and prestige (measured through subordinates’ self-reports) and externally observed expressions of these strategies (measured through supervisors’ other-reports) in organizational contexts. To our knowledge, Cheng et al. (2010) provided thus far the only test of convergence between self- and other-ratings of dominance and prestige within student sports teams, reporting correlations of 0.33 for dominance and 0.40 for prestige. However, it is unclear whether these results generalize to the organizational context, as dynamics within sports teams are inherently different to those taking place between subordinates and their supervisors. For instance, in sports teams, tasks and expected behaviors are extremely well defined, and goals are clear and collective (Devine, 2002). In contrast, organizational employees typically have considerably more room to craft their own tasks and to display various strategic behaviors and impression management tactics to influence their supervisors’ perceptions (Tims et al., 2014; Wayne & Liden, 1995). Therefore, there is a clear need to gain a deeper understanding of dominance and prestige as strategies for social advancement within organizations.

Method

Participants and Procedure

Based on an a priori power analysis for mediation effects using MedPower (Kenny, 2017), we estimated a minimum sample size requirement of at least N = 113 to achieve 80% power to detect a medium-sized effect (an ab path of 0.08) with an alpha of 0.05. Participants were recruited via the social and professional networks of numerous bachelor students at one Spanish and two Dutch and universities. Students were asked, in the context of a course assignment, to contact supervisor-subordinate dyads from a variety of real-world organizations. In total, 271 supervisors and 288 subordinates accessed the surveys. However, 149 individuals failed to complete the process. Excluding these participants and their dyadic partners from the data resulted in a final sample of 205 complete supervisor-subordinate dyads (N = 410, 53.1% male, M_age = 38.11, SD = 12.85). Supervisors (59.2% male) were on average 42.14 years old (SD = 12.24), with most of them having completed a bachelor’s degree or higher (77.1%) and the remaining having completed either intermediate vocational education (10.7%), high-school education (7.8%), elementary education (1.0%), or other (3.4%). Subordinates (43.7% male) were on average 35.48 years old (SD = 12.87), with most of them having completed a bachelor’s degree or higher (69.0%) and the remaining having completed either intermediate vocational education (19.4%), high-school education (8.7%), elementary education (0.5%), or other (2.4%).

Both supervisors and subordinates first provided informed consent and completed a few demographic questions. Subordinates then completed the General Knowledge Questionnaire (Michailova & Katter, 2014) as well as self-reports questionnaires of dominance and prestige (Cheng et al., 2010). Supervisors also reported on subordinates’ dominance and prestige, and then indicated their (1) perceptions of subordinates’ current socioeconomic status and their (2) expectations of subordinates’ future socioeconomic status (i.e., in 5 to 10 years from now). These supervisors’ perceptions of their subordinates’ social status at two different points in time acted as our dependent variable in assessing the influence of dominance and prestige strategies in terms of status attainment.

Measures

Overconfidence

To limit the burden for the subordinates, we measured overconfidence with a subset of 12 items from the same General Knowledge Questionnaire (Michailova & Katter, 2014) that we used in all prior studies. This 12-item subset was taken from prior research on overconfidence (Ronay et al., 2017) that required the division of the full 24-item questionnaire into two blocks. The authors reported a high and significant correlation between the two blocks, r = 0.67, p < 0.01.

Self- and Other-Reports of Dominance and Prestige

To capture participants’ dominance and prestige, we employed the dominance (8 items) and prestige (9 items) scales developed by Cheng et al. (2010). Subordinates completed a self-report version and supervisors completed a peer-report version with regards to their subordinates. Sample items for the dominance self- and peer-report versions are “I am willing to use aggressive tactics to get my way” and “He/she is willing to use aggressive tactics to get his/her way”. Sample items for the prestige self- and peer-report versions are “My unique talents and abilities are recognized by others” and “His/her unique talents and abilities are recognized by others”. Responses were given in a 7-point Likert scale anchored by 1 (“strongly disagree”) and 7 (“strongly agree”). Cronbach’s alpha reliabilities from the dominance and prestige self-report (α = 0.71 and α = 0.71, respectively) and supervisor-report (α = 0.79 and α = 0.77, respectively) versions were acceptable.

Social Status Advancement

To measure of the influence of dominance and prestige in terms of social status advancement, we relied to supervisors’ expectations of their subordinates’ social status advancement over time, as expressed via the McArthur scale of Subjective Socioeconomic Status (SSS; Adler et al., 2000). The scale consists of a pictorial representation of a "social ladder" with ten steps accompanied by the following text: "Think of this ladder as representing where people stand in your country. At the top of the ladder are the people who are the best off – those who have the most money, most education, and the best jobs. At the bottom are the people who are the worst off – those who have the least money, least education, and the worst jobs or no job". In this sense, while some studies have modified the SSS scale by asking participants to rank themselves within a profession or community (e.g., Giatti et al., 2012), contextualizing oneself within local environments reflects idiosyncrasies associated with those local environments, as opposed to social status as it is commonly understood. These idiosyncratic effects tend to result in lower test–retest reliability at the local community level (Giatti et al., 2012). Because our supervisor-subordinate dyads came from a number of different countries and organizations, we expected there to be substantial idiosyncratic variation across contexts. For this reason, we wanted to limit the influence of local effects and instead captured supervisors’ perceptions of subordinates’ potential for socio-economic advancement at the broader societal level, where the ceiling for career advancement is less bounded. Supervisors were instructed to indicate the rung on which they thought their subordinates (1) “stand at this point in their life” and (2) “are likely to stand in five to ten years from now”. Our dependent variable (i.e., social status advancement) was thus operationalized as subordinates’ future socioeconomic status controlled for subordinates’ current socioeconomic status.

Results

Table 3 provides the bivariate correlations among subordinates’ age, gender, educational level, overconfidence, and dominance and prestige (i.e., both self- and supervisor-reported), as well as supervisors’ expectations of their subordinates’ current and future social status.

Overconfidence and Dominance and Prestige

Self-Reported Dominance and Prestige

To examine the relationships between overconfidence and self-reported dominance and prestige, we regressed subordinates’ self-reported dominance and prestige scores onto their overconfidence. The results revealed a positive and significant relationship between overconfidence and dominance, β = 0.13, b = 0.006, SE = 0.003, 95%CI[0.000,0.011], t(203) = 1.99, p = 0.04, and no significant relationship between overconfidence and prestige, β = -0.06, b = -0.002, SE = 0.002, 95%CI[-0.006,0.002], t(203) = -0.85, p = 0.39.

Supervisor-Reported Dominance and Prestige

To examine the relationships between overconfidence and supervisor-reported dominance and prestige, we regressed subordinates’ dominance and prestige onto their overconfidence. The results revealed no significant relationships between overconfidence and supervisors’ reports of their subordinates’ dominance and prestige (p’s > 0.31). However, further explorative analyses revealed a significant interaction effect, p = 0.01, such that overconfidence was positively and significantly related to supervisor-reported dominance at low levels (16th percentile) of supervisor-reported prestige, β = 0.20, b = 0.009, SE = 0.004, 95%CI[0.00,0.01], t(201) = 2.19, p = 0.02, but not at medium (50th percentile), β = 0.01, b = 0.000, SE = 0.003, 95%CI[-0.00,0.00], t(201) = 0.24, p = 0.80, nor high (84th percentile) levels of supervisor-reported prestige, β = -0.14, b = -0.006, SE = 0.00, 95%CI[-0.01,0.00], t(201) = -1.36, p = 0.17. See Fig. 3 for a graphical representation of the simple slopes analyses.

Fig. 3

Simple Slopes Analysis for Interaction between Overconfidence and Prestige in Predicting Dominance (Study 4)

Dominance and Prestige and Expected Social Status Advancement

Self-reported Dominance and Prestige

To examine the relationships between self-reported dominance and prestige and status advancement, we regressed supervisors’ expectations of subordinates’ social status in the future onto subordinates’ self-reported dominance and prestige, while controlling for supervisors’ perceptions of their subordinates’ current social status. The results revealed positive and significant relationships between both dominance, β = 0.14, b = 0.21, SE = 0.08, 95%CI[0.05,0.37], t(202) = 2.61, p = 0.01, and prestige, β = 0.11, b = 0.23, SE = 0.11, 95%CI[0.001,0.464], t(202) = 1.99, p = 0.04, and supervisors’ expectations of subordinates’ social status advancement. Examining a potential interaction between dominance and prestige revealed no significant effect, p’s > 0.76.

Supervisor-reported Dominance and Prestige

To examine the relationships between supervisor-reported dominance and prestige and status advancement, we regressed supervisors’ expectations of their subordinates’ social status in the future onto subordinates’ self-reported dominance and prestige, while controlling for supervisors’ perceptions of subordinates’ current social status. The results revealed no significant relationships between dominance, β = 0.02, b = 0.03, SE = 0.07, 95%CI[-0.10,0.18], t(202) = 0.53, p = 0.59, nor prestige, β = 0.04, b = 0.07, SE = 0.09, 95%CI[-0.11,0.26], t(202) = 0.73, p = 0.46, and supervisors’ expectations of subordinates’ social status advancement. However, further explorative analyses again revealed a significant interaction effect, p = 0.04, such that supervisor-reported dominance was positively and significantly associated with their expectations of their subordinates’ social status advancement at high levels (84th percentile) of supervisor-reported prestige, β = 0.20, b = 0.26, SE = 0.12, 95%CI[0.01,0.51], t(201) = 2.08, p = 0.03, but not at medium (50th percentile), β = 0.08, b = 0.11, SE = 0.07, 95%CI[-0.04,0.26], t(201) = 1.38, p = 0.16, nor low (16^thpercentile), β = -0.05, b = -0.06, SE = 0.09, 95%CI[-0.41,0.08], t(201) = -0.69, p = 0.14, levels of supervisor-reported prestige. See Fig. 4 for a graphical representation of the simple slopes analyses.

Fig. 4

Simple Slopes Analysis for Interaction between Dominance and Prestige in Predicting Social Status Advancement (Study 4)

Self-Reported by Supervisor-Reported

To examine the importance of self- and supervisor-reports agreement in terms of dominance and prestige, we also examined potential interaction effects between self- and supervisor-reports of dominance and prestige in predicting supervisors’ expectations of social status advancement. The results revealed no significant interaction for dominance nor prestige, all p’s > 0.32.

Indirect Effects Analyses

Finally, in order to test the prediction that dominance operates as a mediator between overconfidence and social status advancement, we used Process (Hayes, 2017) Model 4, first fitting subordinates’ overconfidence as the independent variable, supervisors’ expectations of their subordinates’ social status in the future as the dependent variable, and subordinates’ self-reported dominance as the mediator, while controlling for supervisors’ perceptions of their subordinates’ current social status. The results revealed a positive and significant indirect effect of subordinates’ overconfidence on expected social status advancement through dominance, IE = 0.001, SE = 0.000, 95%CI[0.000,0.002]. No indirect effects were found for self-reported prestige nor for supervisor-reported dominance and prestige. To check the robustness of the proposed theoretical model, we also examined the reversed model – that is, introducing dominance as the predictor and overconfidence as the mediator. The results revealed no significant indirect effect between dominance and expectations of social status advancement through overconfidence, IE = -0.001, SE = 0.012, 95%CI[-0.029,0.021]. While this does not completely mitigate the causality concerns of single-time point measures, it lends support for the proposed mediation model.

Discussion

Consistent with Hypothesis 2 and the results of Study 3, we again found overconfidence to be positively related to dominance, as self-reported by subordinates as well as their direct supervisors. Notably, supervisors rated their overconfident subordinates as dominant only insofar as they were perceived low in prestige. In terms of social status advancement, although we did not observe a direct relationship with overconfidence, we found support for the proposed indirect effect (Hypothesis 3), by which dominance (though not prestige) facilitated the expected social status advancement of overconfident subordinates.

In addition, the dyadic design of Study 4 allowed us to examine the relationships between self- and other-reports of dominance and prestige and social status advancement. In this regard, and consistent with prior reports (Cheng et al., 2013; de Waal‐Andrews et al., 2014; Redhead et al., 2019), we found both self-reported dominance and prestige to be positively related to supervisors’ expectations of subordinates’ social status advancement over time. Interestingly, however, supervisor-reported dominance and prestige significantly interacted – only when supervisors perceived their subordinates as prestigious (i.e., knowledgeable, competent, and successfully contributing to collective goals) did subordinates’ dominance positively influence their prospects of social status advancement.

General Discussion

Prior research demonstrates that overconfident people are more likely to attain high-status positions of leadership and influence. However, the underlying motivational and behavioral mechanisms driving this relationship remain largely unexplored. In the present research, we sought to fill this gap in the literature by proposing that overconfidence is associated with stronger status motives and the pursuit of high-status positions via dominance-based strategies. In Studies 1 and 2, we found overconfidence to be positively related to the pursuit of high-status positions. In Studies 3 and 4, we found overconfident individuals to lean towards dominance- over prestige-based status-seeking strategies. Lastly, in Study 4, we found an indirect effect of overconfidence on expected social status advancement through dominance. Together, these findings generate a number of theoretical and practical contributions.

Theoretical and Practical Contributions

First, our findings provide novel evidence that overconfidence is associated with the active pursuit of high-status positions. Specifically, building on self-verification theory (Swann, 1983, 1987, 1990), we predicted and demonstrated that overconfident individuals are more inclined to enter leadership selection competitions as well as self-select for leadership roles. From a theoretical standpoint, these findings contribute to a better understanding of the explanatory mechanisms underlying the relationship between overconfidence and social status advancement, which thus far limited to the positive evaluative biases that overconfidence elicits (Anderson et al., 2012; Belmi et al., 2020; Ronay et al., 2019). In particular, our research suggests that overconfident people are more prone to the pursuit of high-status positions – an effect that we observed irrespective of whether or not such prominent positions were accompanied by financial benefits. This is consistent with the theoretical argument that the relationship between overconfidence and the pursuit of high-status positions follows from a fundamental human need to confirm and validate one’s (exaggerated) self-views, and so emerges independently of the typically associated financial privileges.

From a practical perspective, the finding that overconfidence is associated with the pursuit of high-status positions bears at least two important implications for organizations and selection professionals. First, this self-selection bias might skew leadership candidate pools towards greater levels of average overconfidence, hence making overconfidence (and the associated misperceptions – e.g., Anderson et al., 2012; Belmi et al., 2020; Ronay et al., 2019) more normative in leadership selection contexts. Second, equally or even more competent candidates whose self-perceptions are humbler might feel disinclined to compete for high-status positions of leadership, or find themselves at a disadvantage if they do compete. Indeed, because confidence holds strong currency in the context of leadership selection (e.g., Hogan et al., 1994; Kirkpatrick & Locke, 1991), a higher proportion of overconfident candidates runs the risk of obscuring the entry and selection of more competent candidates whose self-perceptions are better calibrated. In order to minimize possibilities of suboptimal selection outcomes, organizations and selection professionals should thus be mindful of this self-selection effect. Relatedly, prior theoretical and empirical studies indicate that organizational environments characterized by higher degrees of inequality and competition tend to exacerbate the expression and behavioral consequences of overconfidence (Johnson & Fowler, 2011; Ronay et al., 2022; Szamado, 2003). This suggests that organizations could also seek to counteract the self-selection effects of overconfidence by developing organizational cultures that promote a more equitable distribution of resources and maintain competition levels in check.

The present research also examined the behavioral strategies that overconfident individuals tend to adopt in service of their status aspirations. Here, drawing on prior overconfidence research in organizational contexts and dual-strategies theory (Cheng, 2020; Cheng et al., 2013; Henrich & Gil-White, 2001; Maner & Case, 2016; Van Vugt & Smith, 2019), we predicted and found support for the notion that overconfident individuals tend to engage in dominance- over prestige-based status-seeking strategies. This finding helps explain prior reports that the overconfident attain high-status positions even when objectively revealed as less competent that initially perceived (Kennedy et al., 2013), and so offers new insight into the explanatory mechanisms underlying the relationship between overconfidence and status-attainment. Indeed, while prestige-based status is granted based on rational evidence of superior competence, dominance-based status is gained by those, competent or not, who successfully adopt dominant and self-serving behaviors (e.g., Cheng et al., 2013; Henrich & Gil-White, 2001; Maner & Case, 2016), such as those that we here found associated with overconfidence. In this sense, and consistent with our theorizing, our results suggest that overconfident individuals may at some level be aware that their exaggerated self-image is insufficient to garner higher social status through prestige, thus leading them to construe dominance as the most viable means of navigating their respective social hierarchies.

The finding that overconfident individuals tend to enlist dominance-based strategies also has practical implications for organizations. Specifically, by exhibiting these morally questionable behaviors (e.g., manipulation, coercion, or self-serving allocation of collective resources), overconfident leaders might over time contribute to the emergence of toxic organizational cultures in which abuse of power, exploitation, and personal agendas become the norm (e.g., Farrell & Dane, 2020; Gilbert et al., 2012; Hodson et al., 2006). Indeed, researchers have consistently demonstrated how those in leadership roles play a pivotal role in shaping organizational culture and dynamics (e.g., Bass, 1990; Bass & Avolio, 1994; Brykman & Maerz, 2023; Gilbert et al., 2012; Ofori, 2009; Ogbonna & Harris, 2000). Hence, our research suggests that the perils of selecting overconfident leaders might extend beyond the well-known negative consequences of overconfident decision-making (see Meikle et al., 2016 for a review), contributing as well to the development of undesirable organizational cultures that ultimately hamper optimal organizational-level performance. This possibility, nevertheless, should be directly tested in future studies.

Third, the current research examined the roles of dominance and prestige in predicting expectations of social status advancement. Prior studies have reported positive associations between both dominance and prestige and the attainment of higher social rank (Cheng et al., 2013; de Waal‐Andrews et al., 2014; Redhead et al., 2019). These studies, however, were conducted in lab settings involving short-term groups and hierarchies or among student groups, and so researchers have called for generalization to field settings, such as organizations (Cheng et al., 2013). Our data from real-world organizational contexts shed new light in this regard, since the status-enhancing effects of dominance and prestige that we observed differed depending on whether dominance and prestige were self- or supervisor-reported. In particular, while we found both self-reported dominance and prestige to be positively related to expectations of social status advancement, supervisor-reported dominance and prestige did not predict expectations of social advancement. Crucially, however, we found supervisors’ reports of their subordinates’ dominance and prestige to interact, such that dominance was positively related to expectations of social status advancement only insofar as subordinates were also rated as high in prestige.

This is a novel and relevant finding as it suggests that the dynamics of status allocation in real-world organizational contexts differ from those previously observed in lab studies or among student samples – wherein dominance and prestige have been reported sufficient to predict the attainment of higher social status independently (Cheng et al., 2013; de Waal‐Andrews et al., 2014; Redhead et al., 2019). This may be due to several reasons. For instance, it is possible that professional contexts impose greater demands for status attainment than groups of students do in the context of course assignments (Redhead et al., 2019), requiring superior competence (i.e., prestige) but also the desire to command others or the willingness to enlist aggressive tactics (i.e., dominance). It is also possible that the longer-term nature of work relationships, as well as the higher stakes involved, influence people’s reactions to dominance and prestige relative to lab contexts (Cheng et al., 2013), as participants are typically strange to each other and decisions bear no real or significant consequence. It is important to note, however, that the observed interaction between dominance and prestige emerged from our exploratory analyses, and so replication by future studies is needed.

This leads to our fourth contribution, which concerns the extent to which self- and other-reports of dominance and prestige converge. In this regard, while we observed positive associations between subordinates’ self-reports of dominance and prestige and their supervisors’ perceptions of such constructs, the observed correlations were nonetheless modest, r_dominance = 0.26, p < 0.01 and r_prestige = 0.24, p < 0.01. This result indicates that there is some noise between intention and expression, or between the expression and interpretation of dominance and prestige strategies. This highlights the importance of carefully considering which dominance and prestige measurement method (i.e., self-reports, other-reports, or both) best fit researchers’ specific focus and purposes.

Limitations and Future Research Directions

There are also limitations to the current studies that can serve as avenues for future research. First, our reliance on a correlation approach prevents us from nailing down questions of causality – that is, whether overconfidence is a causal antecedent of the pursuit of higher social rank. This problem is intrinsic to all overconfidence research (e.g., Anderson et al., 2012, Belmi et al., 2020; Betzer et al., 2022; Kennedy et al., 2013; Moore & Healy, 2008; Reuben et al., 2012; Ronay et al., 2017; Ronay et al., 2019; Shipman & Mumford, 2011), as overconfidence represents a stable individual difference (similar to personality dimensions) (e.g., Johnson & Fowler, 2011) and is therefore difficult to manipulate experimentally. Even if some methods could be devised, such as providing participants with false feedback on competence in order to produce changes in self-perceptions, it is improbable that such variations would reflect a deep genuine change in self-views. Indeed, prior studies indicates that single-point interventions on self-views (e.g., to improve confidence and self-esteem) often result in “fanciful and ephemeral” changes (e.g., Crocker & Park, 2004; Swann et al., 1994, 2007, p. 90) unlikely to produce meaningful effects. These inherent difficulties to assessing the causal effects of overconfidence should urge researchers to explore novel alternative methods to capture individual differences in overconfident self-views, as this would be highly valuable for future research endeavors. In the meantime, scholars should interpret empirical reports with the necessary level of caution, including those that we present here.

Relatedly, it is important to note that our test of mediation in Study 4 (i.e., overconfidence on expected social status advancement through dominance; Hypothesis 3) relies on single-time point measures, limiting the strength of evidence. In order to address this issue within the limitations of the current methodology, we examined the possibility of reversed causality – that is, introducing dominance as the predictor and overconfidence as the mediator. The results showed no significant effects. While this lends additional support for the proposed mediation model, the cross-sectional nature of our data remains a limitation of the present research. Given the difficulties of experimentally manipulating overconfidence (as we discussed above), future studies could thus seek to replicate the current findings via longitudinal or time series methodologies. For instance, researchers could capture individuals’ overconfidence and dominance at multiple time points, as this would allow to examine whether fluctuations in overconfidence correspond with concurrent and future changes in individuals’ tendency toward dominance.

Third, while overconfidence and narcissism are recognized to represent distinct theoretical constructs predicting different sets of outcomes (see Brunzel, 2021 for a review), prior studies indicate that overconfidence and narcissism are often, though not always (e.g., Shipman & Mumford, 2011), positively associated (e.g., Macenczak et al., 2016). This implies the possibility that some of the effects of overconfidence reported here might be partially explained, or exacerbated, by narcissistic attributes (e.g., entitlement or a sense of superiority). Unfortunately, we did not measure narcissism in the current research, precluding us from partitioning out individual differences in narcissism. Therefore, future studies might consider the potential role that narcissistic attributes may play in the relationship between overconfidence and the pursuit of high-status positions. This would also help gain a more fine-grained understanding of the interaction between overconfidence and narcissism, contributing to “knowledge accumulation” in the field (Brunzel, 2021, p. 585).

Finally, although our samples comprised participants and organizations from three different countries (i.e., US, Spain, and The Netherlands), it is important to test our hypotheses in other cultural contexts where the effects of overconfidence, dominance, and prestige might differ. For instance, self-serving exploitative behaviors characteristic of dominance often meet more severe negative social reactions in collectivistic cultures, such as in China and Japan (e.g., Adair & Semnani-Azad, 2011; Kowner & Wiseman, 2003; Ohbuchi et al., 1999). Cultural differences might therefore attenuate or even reverse the status-enhancing effects of dominance that we observed. Unfortunately, existing evidence for the relationship between dominance and social status advancement, including the present studies, is thus far limited to individualistic cultures (e.g., Cheng et al., 2013; de Waal‐Andrews et al., 2014; Redhead et al., 2019). Given the remarkable rate at which organizations and the labor market in general are becoming multicultural (Derous, 2017; Kim et al., 2022; Manroop et al., 2013), we believe this to be an important avenue for future research.

Conclusion

Thus far, the ascension of overconfident individuals to high-status positions of leadership and influence has been attributed to the mistaken perceptions of superior competence and abilities that overconfidence tends to elicit. The studies presented here suggest an alternative (though compatible) explanatory mechanism – overconfident individuals are more prone to actively pursue such prominent positions, doing so via self-serving, often morally questionable dominance-based strategies.

Appendix 1

Tables 1, 2 and 3

Table 1 Means, Standard Deviations, and Correlations between Studies 1 and 2 Variables

Table 2 Means, Standard Deviations, and Correlations between Study 3 Variables

Table 3 Means, Standard Deviations, and Correlations between Study 4 Variables