Leadership traits

Almost everything written about leadership traits answers the wrong question, and does it so smoothly that the substitution is hard to notice. The question people think they are asking is what makes someone good at leading. The question the trait research can mostly answer is who ends up leading.

Those are not the same question, they do not have the same answer, and the distance between them is the most useful fact in this whole literature. It explains the confident mediocrity three levels above you. It explains the person on your team everyone would follow who has never been offered the job. And it explains why a leadership questionnaire can be genuinely informative while telling you nothing at all about whether you would be any good.

Traits predict emergence better than they predict effectiveness. Hold onto that sentence and the rest of the field becomes readable.

For the first half of the twentieth century, leadership research was a hunt for the qualities of great men. Ralph Stogdill ended it in 1948 with a review of more than a hundred studies, and his conclusion is still quoted because it was correct: a person does not become a leader by virtue of possessing some combination of traits, because the traits that mattered kept changing with the situation. Leadership research spent the next thirty years on behaviours and contingencies instead, and the trait question was something you did not ask in print.

It came back in the 1990s for an unglamorous reason. The Big Five gave the field a common vocabulary, and meta-analysis gave it a way to pool decades of small, noisy studies that had each measured something slightly different. Judge, Bono, Ilies and Gerhardt did exactly that in 2002 and found that the five factors together correlated with leadership at around .48 — a substantial multiple relationship, and the rehabilitation of the trait idea in a much more modest form than the one Stogdill buried.

Extraversion was the most consistent single predictor. Conscientiousness and openness followed, emotional stability contributed, and agreeableness — the trait most people name first when asked what a good boss is like — was essentially unrelated. That last result has held up and is worth sitting with.

The same 2002 review did the thing the popular summaries drop. It separated studies of who emerged as a leader from studies of how well leaders performed, and the numbers came apart: the trait relationship was clearly stronger for emergence than for effectiveness.

Look at how emergence is measured and the reason is not mysterious. A typical emergence study puts strangers in a leaderless group discussion and asks afterwards who led. Ensari and colleagues pooled forty-five such studies in 2011 and found extraversion the standout predictor, alongside the more uncomfortable finding that a dominant, authority-oriented style also predicts being picked. What is being measured is a judgement made by observers, usually within minutes, on the basis of how someone talks.

Two well-replicated facts follow. Sheer talking time predicts being seen as the leader, largely independent of the quality of what is said — which is why extraversion does so well in these designs, since extraverts talk. And Judge, Colbert and Ilies found in 2004 that measured intelligence relates to leadership at about .27, weaker than a century of assumption, while perceived intelligence relates to it considerably more strongly. Looking smart beats being smart, in the part of the process that decides who gets the role.

This is not cynicism. It is a measurement result, and it has a practical edge: the selection system that promotes people is running on the same cues these studies isolate. Emergence is a real gate even when it is a bad gate.

The moment you ask whether a leader worked, the trait signal weakens and the context takes over.

The cleanest demonstration is Grant, Gino and Hofmann's 2011 study, which found the extraversion advantage reverses. Extraverted leaders produced better results with passive teams who needed direction, and worse results with proactive teams, because an extraverted leader responds to subordinate initiative as competition for the floor rather than as input. The same trait, opposite sign, depending entirely on who is in the room. No trait score can carry that, which is precisely Stogdill's point arriving seventy years later with better statistics.

Judge, Piccolo and Kosalka made the second point in 2009: the traits that get people the job include several that damage their tenure. Narcissism is the standard example — it predicts emerging as a leader well and predicts leader effectiveness poorly, and the trait-level story is a curvilinear one where a little confidence helps and a lot destroys judgement. Dominance, risk appetite and charisma behave similarly. The bright side and the dark side of a leader trait are frequently the same trait.

One more finding that is usually reported badly. Arvey and colleagues estimated in 2006, using twins, that around a third of the variance in whether someone occupies a leadership role is heritable. That is not a gene for leadership. It is largely the heritability of the traits that shape how people are perceived, flowing into a selection process — which means it is a statement about the system doing the selecting as much as about the people it selects.

Read this way, a leadership questionnaire stops overclaiming and becomes useful for something specific.

It is a reading of how you tend to operate and how that lands, which is the emergence side. That is legitimate information, because the emergence side is what promotion runs on and because most people are wrong about their own. The Leadership Potential test scores four dimensions and deliberately reports the failure mode of your strongest one, which is the only responsible way to present a trait that is a liability at the high end as often as it is an asset.

It is not a prediction that you will be good at the job. Nothing self-report can do that, and the honest reason is in the paragraphs above: effectiveness lives in the match between a person and a particular team at a particular moment, and a questionnaire only has one of those three.

Two directions worth taking from a result. If your profile is low on the dimensions that drive emergence and you want the role anyway, the lever is not personality change — it is the narrow set of behaviours the emergence research keeps isolating, mostly speaking early and being visible on things that matter, which is learnable and does not require becoming an extravert. And if you emerge easily, the useful reading of your result is the shadow line rather than the strengths, since the failure modes of confident leaders are invisible from the inside by construction.

If the gap between the two outcomes is landing personally — if you can see that you would be effective and are not being picked — that is a common and well-documented position rather than a defect, and impostor phenomenon is the entry for what it does to people over time.