Do Billionaires Have Higher IQs? What the Swedish Data Shows

Do billionaires have higher IQs than everyone else? The best large-scale evidence available says something more specific, and more interesting, than a simple yes. Cognitive ability does rise with income for most of the income range, but the relationship flattens out well before anyone approaches billionaire-level wealth, and among the very highest earners, average scores tick down slightly rather than up. That is the finding of a 2023 study that matched Swedish wage records to cognitive-ability scores from mandatory military conscription testing, and it complicates the 'smart people get rich' story more than it confirms it.

What did the Swedish study actually measure?

Keuschnigg and colleagues, writing in the European Sociological Review in 2023 (doi 10.1093/esr/jcac076), used a data source most countries do not have: Sweden's population registers, linked to cognitive-ability tests that men were required to take at military conscription, typically around age 18 to 19, before Sweden ended mandatory service. That design records an ability score years before a career or income exists, then tracks the same people's earnings decades later, which is a stronger design than surveying adults and asking them to estimate their own intelligence. The tradeoff is coverage: conscription testing only captures men, and only in Sweden, so this is not a global or gender-balanced picture.

Where does the plateau kick in?

Average ability rises steadily with pay across most of the wage distribution. Above roughly 60,000 euros a year, the curve goes flat: further pay increases stop being matched by higher test scores.

• Below the plateau: cognitive ability tracks pay closely, step for step
• Around 60,000 euros a year and up: average ability plateaus at about plus 1 standard deviation, roughly 115 on the mean-100, SD-15 scale used by most modern IQ tests
• The top 1 percent of earners: score slightly lower than the bracket of earners immediately below them, despite often being paid about twice as much

That 'plus 1 SD' figure is doing real work. On a standard mean-100, SD-15 scale, one standard deviation above average is around 115, which is a solid but unremarkable score, at the 84th percentile rather than genius territory. Turning a raw test score into that kind of standard unit is exactly what our guide to IQ z-scores walks through.

Why would the very top earn more without scoring higher?

The study does not fully explain the reversal, but a few candidates fit the pattern. Extreme pay at the top increasingly comes from owning equity, running a business, or being paid for a role's scarcity and negotiating leverage, rather than from a further step up in measurable skill. A company founder's stake can be worth a great deal regardless of how that person scores on a conscription-era reasoning test. Risk tolerance, timing, inherited capital and network effects all shape who ends up in the top 1 percent of pay in ways a cognitive-ability test was never designed to capture.

Has this been checked outside Sweden?

A follow-up analysis extended the same approach to Finland and Norway, published in the same journal, and found a similarly steep ability-income relationship for most earners, with the gradient easing off at the very top. The newer paper frames the pattern as earnings climbing faster than ability rather than a flat reversal, but the direction, a weakening link at the extreme top, matches the original Swedish finding. Replication across three Nordic countries, each with its own mandatory testing data, is meaningfully stronger evidence than a single-country result.

Why does the 'genius billionaire' idea persist anyway?

Part of the answer is which proxies get repeated in profiles of famous entrepreneurs: an elite university degree, a chess or poker habit, a high SAT score. Those proxies are weaker evidence than they sound. SAT scores correlate with general cognitive ability at around r = 0.8 (Frey and Detterman, 2004, Psychological Science), which is a real but imperfect link, and getting into a selective university is shaped heavily by family wealth, through test preparation, private tutoring and legacy admissions, in a way a conscription test taken by nearly everyone is not. That is precisely the confound the Swedish design was built to avoid: nearly every young man took the same test regardless of his family's income, which is not true of the SAT or of elite college admission.

Does this mean smart people do not get rich?

No. The correlation between ability and pay is real and fairly strong across most of the range; this is specifically a story about what happens after a person is already earning well above average. The popular idea that the richest people must also be the most cognitively exceptional simply is not what the data shows once income gets extreme enough to be driven by capital and ownership rather than salary. For the more general, and much better supported, link between IQ and typical earnings, see our piece on does IQ predict income.

How does this fit alongside country-level IQ claims?

Claims about entire countries' average IQ run into a related trap: a national average says almost nothing about any specific rich or famous person from that country, and the methodology behind most cross-country IQ rankings is contested. Anyone comparing country-level figures should start with actual data rather than a viral list; see our average IQ by country page.

Any single test score, from any country's data, comes with a margin of error. Check yours on the percentile calculator after taking the IQ test.

Common questions