Is There an IQ Sweet Spot for Leaders? What the 120 Study Found

Search “leader IQ” and one claim comes back again and again: the ideal IQ for a leader is about 120, and beyond that leaders get worse. The claim traces to a single study, and it is narrower than the headlines make it. It measured how colleagues rated a leader’s behavior, not how the leader’s team performed. This article goes through what the study did, how firm the number 120 is, and what larger datasets say about being “too smart to lead”.

Where does the “IQ 120 for leaders” claim come from?

It comes from a 2017 paper in the Journal of Applied Psychology by John Antonakis, Robert House and Dean Keith Simonton, titled “Can super smart leaders suffer from too much of a good thing?” The authors set out to test a prediction from Simonton’s 1985 models of intelligence and influence: a leader is seen as most effective when he or she sits about 1.2 standard deviations above the mean IQ of the group being led. A standard deviation is the usual unit of spread on IQ scales; on the common scale where it equals 15 points, 1.2 of them is 18 IQ points. So the idea is about a gap between leader and followers, not a fixed number. The proposed reason is communication: a leader far ahead of the group may be harder to follow.

What did the 2017 study actually measure?

The sample and scoring details below are taken from the authors’ accepted manuscript; the headline result is in the published abstract.

• Sample: 379 mid-level leaders from seven companies and two executive-education cohorts, in 30 countries. IQ was tested for only 171 of them, in four of the companies, and the main curve rests on those 171.
• Intelligence: the Wonderlic Personnel Test, a 12-minute test, converted to an IQ scale with the widely circulated rule of double the raw score plus 60. The leaders’ average of about 25 corresponds to an IQ near 111.
• Leadership: rated by others. About 2,900 raters in all, roughly 7.7 per leader, made up of subordinates, peers and the leader’s boss, filled in standard questionnaires on transformational leadership (including charisma), transactional leadership and related styles.
• Not measured: sales, profit, retention or team results, and the intelligence of the followers, so the “gap” itself could not be tested directly.

That IQ figure deserves a caution of its own. Our explainer on converting Wonderlic scores to IQ shows why double-plus-sixty is a rule of thumb rather than a published conversion, so an “IQ” derived from it is best read as an estimate.

How firm is the number 120?

The published abstract reports that ratings followed an inverted-U function of intelligence, peaking at an IQ of about 120. The figures underneath are softer than the headline:

• The inverted U appeared for 8 of the 10 active, constructive leadership measures: 7 at conventional significance and 1 at a weaker level. IQ had no effect on passive styles such as management-by-exception and laissez-faire.
• The peak fell between roughly IQ 117 and 122 depending on the measure, and the full-sample estimate was about 118.5. The theory’s own prediction was 124, and the tests could not rule it out, which at this sample size is a non-rejection rather than a confirmation.
• No confidence interval was reported for the peak. The downturn after it reached conventional significance only far along the curve, at a Wonderlic score of 35; at the predicted 124 it was negative but significant only at the looser 10% level.
• The curved term added about 3 percentage points of explained variance on average (R² = .034): a real but small effect.

Is 120 the ideal IQ for a leader?

Not as a universal figure. Because the theory is about a gap, the best IQ depends on who is being led. The study’s own prediction of 124 assumed followers averaging about 106 plus the 18-point gap. The same arithmetic gives about 118 for followers averaging 100, which is the figure the lead author used in later popular summaries, and about 128 for followers averaging 110. The 120 peak is also an estimate from one sample of mid-level managers whose own average was about 111, so it sits roughly three-quarters of a standard deviation above their mean, around the 78th percentile within that group.

Against the whole population, an IQ of 120 is 1.33 standard deviations above the mean: the 90.9th percentile, so about 1 person in 11 scores at or above it. Our percentile calculator converts any score, and what an IQ of 120 means covers the score in general.

Does real-world data show that very smart people make worse leaders?

The larger datasets on objective outcomes do not show a ceiling. A 2004 meta-analysis in the same journal, covering 151 independent samples, found an average correlation of .21 between intelligence and leadership, or .27 after correcting for range restriction, the distortion that comes from studying only people who have already been selected. That is a modest, positive relationship, lower than earlier reviews had suggested, and it was stronger when intelligence was rated by others than when it was tested.

A 2021 analysis in Perspectives on Psychological Science followed four cohorts and 48,558 people. Most effects of cognitive ability on life outcomes were linear or null, the authors reported no support for any downside to higher ability, and they found no sign of a threshold at IQ 100 or 120. Leadership there meant supervisor status and span of control, so it is not a replication of the 2017 perception study, and its data cannot test the gap theory. It simply does not find a ceiling in objective outcomes.

The most direct look at chief executives comes from Sweden. Adams, Keloharju and Knüpfer linked cognitive tests taken at the military draft at about age 18, for around 1.3 million men, to later careers, including about 26,000 who were CEOs between 2004 and 2010. The median chief executive of a large company (assets above about US$1.3 billion) had scored above roughly 83% of men his age, which on the mean-100, SD-15 scale is about IQ 114. A quarter of large-company CEOs scored below the 70th percentile and a quarter above the 92nd, about IQ 122. Fewer than one in five was in the top 5%, and CEO pay rose roughly in a straight line with cognitive ability, with no downturn at the top. The draft composite is not an IQ score, so treat the IQ-scale figures as an approximate translation.

What should you take from it?

For a hiring panel or a manager, a cut-off at 120 is not supported by any of this. What the 2017 study points to is a communication effect: colleagues rated leaders less favorably once leader IQ passed a certain point, which the authors attribute to a widening gap with followers. That explanation was proposed rather than tested, and follower IQ was never measured. For the straight-line evidence on job performance, see what cognitive ability tests predict at work. Because tests lose precision at the top of the scale, the IQ ceiling effect explains why very high scores are hard to compare, and presidents ranked by IQ shows how shaky leader IQ figures usually are.

So: is there an IQ sweet spot for leaders, about 120? In one study of mid-level managers, colleagues’ ratings of leadership peaked near 120 and then dipped, a small and imprecise effect that says more about how a large gap is perceived than about leadership results. Larger datasets on real outcomes find benefits that keep rising or level off, not a ceiling. To see where your own reasoning score sits, the full IQ test reports a percentile and a confidence range. More on work and scores: what quant trading firms look for and average IQ by occupation.

Common questions