Psychology says the trait we most admire as "confidence" in leaders overlaps disturbingly often with narcissism, which helps explain why the wrong people so often rise to the top.

Watch how people are chosen for leadership positions and a pattern shows up. The person who speaks first in the meeting. The person who acts like they know the answer whether they do or not. The person who takes up space and assumes their opinion matters. That collection of traits reads, to almost everyone in the room, as confidence. And confidence, in the shorthand most organisations use, reads as leadership material.

The problem is that the same collection of traits also reads, to psychologists who have spent thirty years studying it, as something else. It reads as narcissism.

Not the clinical version. Not narcissistic personality disorder, which is a diagnosable condition affecting a small fraction of the population. What the research has been documenting is the everyday sub-clinical trait, the ordinary personality profile that scores high on the standard measurement instruments. And the overlap between that profile and the profile of people who successfully rise to leadership positions is, on the accumulated research record, considerably tighter than most organisations would like to admit.

How narcissists win the room

According to a 2008 paper by Amy B. Brunell at Ohio State University with William Gentry, Keith Campbell, Brian Hoffman, Karl Kuhnert and Kenneth DeMarree, published in Personality and Social Psychology Bulletin under the title “Leader Emergence: The Case of the Narcissistic Leader”, the researchers ran three separate studies asking a simple question. If you put four strangers in a room with no assigned hierarchy and give them a task to complete, who ends up in charge?

The answer, across all three studies, was the narcissist. Participants who scored higher on the Narcissistic Personality Inventory before entering the group were consistently more likely to be identified as the group’s leader afterwards. Both by themselves and by the other members of the group. The effect held even after controlling for sex, self-esteem, and the Big Five personality traits. The specific dimension of narcissism doing most of the work was the power dimension, meaning the drive to be in charge and to have others defer to you.

Watch here – the real difference between self-esteem and ego:

Study three tested the same pattern in a sample of practising executives evaluated by expert raters. It held there too. Narcissism predicted who would be seen as a leader in a room full of actual working leaders.

What the Brunell team argued from those results is that narcissists are unusually good at the specific behavioural signature that groups read as leadership. They speak first. They project certainty. They frame their preferences as decisions. They seem, in the crucial early moments of any group interaction, to already know what should happen next. Other people, faced with someone who seems to know, defer. The narcissist ends up in charge not because anyone consciously decided they were the best person for the role, but because they took the role and nobody else contested it.

The finding has since been replicated many times. According to a 2015 meta-analysis by Emily Grijalva, Peter Harms, Daniel Newman, Blaine Gaddis and R. Chris Fraley, published in Personnel Psychology under the title “Narcissism and Leadership: A Meta-Analytic Review of Linear and Nonlinear Relationships”, the emergence pattern is one of the more robust findings in the whole organisational-psychology literature. The meta-analysis pooled results across dozens of previous studies and found a consistent positive relationship between narcissism scores and the likelihood of being identified as a leader by others. The mechanism, on Grijalva and colleagues’ analysis, runs primarily through extraversion. Narcissists are outgoing, socially dominant, and comfortable with attention, and those specific traits are exactly what most groups mistake for the underlying capacity to lead.

What happens after they get the job

The problem the same body of research documents is that the traits that get narcissists into leadership positions are not the traits that make someone good at the job.

The Grijalva meta-analysis also examined leadership effectiveness. Meaning, once the person is in the role, are they any good at it? The linear relationship between narcissism and effectiveness, on the pooled data, was essentially zero. Narcissists were rated by their subordinates, peers, and supervisors as no better and no worse than non-narcissists at actually leading. There was one exception. Narcissists rated themselves as considerably more effective than other people rated them. Which is exactly the self-enhancement pattern the underlying research has documented at length. Narcissists believe they are excellent leaders. The people they lead, on the observer-rated evidence, do not agree.

The effectiveness picture is slightly more interesting than a flat zero suggests. On further analysis of the meta-analytic data, the relationship between narcissism and effectiveness followed an inverted-U shape. Moderate narcissism appeared to be optimal. Very low narcissism produced ineffective leaders. Very high narcissism produced ineffective leaders. The best performers sat in the middle. Which is a nuance that gets lost in most popular coverage of the topic, but which reflects the messier reality that some narcissistic traits, in moderation, actually help.

The core problem the whole body of research points at is not that narcissists are always bad leaders. It is that selection processes reward narcissism at a much higher rate than performance justifies. According to a 2011 study by Barbora Nevicka, Annebel De Hoogh, Annelies Van Vianen and colleagues at the University of Amsterdam, published in the Leadership Quarterly under the title “All I Need Is a Stage to Shine: Narcissists’ Leader Emergence and Performance”, narcissists in laboratory work teams reliably emerged as leaders. Their teams, on the objective performance measures, did not do any better than teams led by less narcissistic members. The narcissists thought they had done brilliantly. The teams around them, quietly, had not.

The gap between how narcissists are perceived on the way in and how they actually perform once installed is the specific mechanism by which the wrong people so often rise. The trait that gets you the promotion is not the trait that makes you good at the job. Both facts are well established. The organisations doing the promoting have, for the most part, not yet caught up.

What all of this suggests, on the combined evidence of the past two decades, is not that confidence is bad or that confident people should be viewed with suspicion. Genuine self-assurance, grounded in actual competence, is one of the more valuable qualities a leader can have. The problem is that the shorthand the ordinary human eye uses to identify confidence in the first place is unreliable. It picks up the visible surface behaviour and cannot easily distinguish between the person whose confidence reflects genuine capacity and the person whose confidence reflects a self-assessment that is, on any external measure, not particularly accurate. The second person, on the accumulated research, is disproportionately likely to be a narcissist. And the second person, on the same research, is disproportionately likely to end up in charge.

Which is not, on the current best reading, an easy problem to fix. The behavioural signature is genuinely persuasive. The people around it are not being lazy or stupid when they respond to it. The signature is doing what signatures evolved to do, which is convey something quickly about the person carrying it. The information it conveys just turns out to be, in this particular case, misleading often enough that it should probably be trusted less than most workplaces currently trust it.

Kiran Athar is a writer, not an organisational psychologist. This piece draws on peer-reviewed research in Personality and Social Psychology Bulletin, Personnel Psychology, and the Leadership Quarterly.