The claim turns up every few years in the British press: tall people are cleverer, and tall people are paid more. Both statements are built on real research, and both are reported in a way the research does not support. The correlation between height and intelligence test scores, measured in Scotland on nearly 7,000 people, is 0.16. That is about two and a half per cent of the variation in scores. It is not nothing, and it is nowhere near a rule.

Britain happens to be the best place in the world to test what is left once you take the claim seriously, because three of the datasets that matter are British: two national birth cohorts followed since childhood, and UK Biobank, which holds genetic data on half a million people. Run the question through them and the height premium shrinks every time you add something you had previously failed to measure. This article walks through what survives.

The correlation is 0.16, and that number does the work

The study behind the headlines came from the University of Edinburgh. Riccardo Marioni and colleagues used Generation Scotland: the Scottish Family Health Study, a cohort with medical measurements and cognitive testing, and analysed 6,815 unrelated participants. Intelligence was assessed through four tasks: reaction time, verbal ability, processing speed and memory recall.

Their results, published in Behavior Genetics in 2014, were these. The observed correlation between height and general cognitive ability was 0.16, with a standard error of 0.01. The genetic correlation estimated from common variants was 0.28. And common additive genetic variants accounted for about 71% of the observed correlation between the two traits.

That last figure is the one that gets mangled. It is frequently reported as “70% of the differences in IQ and height are genetic”. It does not mean that. It means that of the small overlap between height and cognitive scores, most of the overlap appears to be genetic in origin. The overlap itself is small. Nothing in the paper claims that 70% of anyone’s intelligence comes from their genes, and the authors were careful to say their method cannot distinguish between shared genes affecting both traits and assortative mating, which is the tendency of taller people and more educated people to pair off, generation after generation, until the traits travel together in the population for reasons that have nothing to do with biology.

What a correlation of 0.16 looks like in a room

Take a hundred British adults at random and rank them by height, then by test score. A correlation of 0.16 means the two rankings agree slightly more often than chance and disagree constantly. The tallest person in the room is more likely to be in the bottom half on the test than at the top of it. If you picked the taller of any two people and bet that they would score higher, you would win roughly 55 times out of 100. You would be better off betting on who had stayed in education longer, whose parents had, or which part of the country they grew up in.

For comparison, the same literature that produces 0.16 for height and intelligence finds associations between greater height and lower risk of coronary heart disease, stroke, accidents and suicide (Batty and colleagues 2009, Lee and colleagues 2009, Paajanen and colleagues 2010, Whitley and colleagues 2010). Those are real, well-replicated and still modest. The pattern across all of them is the same: tall and short populations differ a little on average, and individuals differ enormously.

The British pay premium, measured and then dismantled

The pay question has a cleaner answer, because Britain has two national birth cohorts that measured children’s cognitive ability years before anyone knew what they would earn. Anne Case and Christina Paxson used both in Stature and Status: the 1958 National Child Development Study and the 1970 British Cohort Study.

Their first result is the one usually quoted. Among British men, earnings rose with height: about 2.4% per inch in the 1958 cohort and about 1.5% per inch in the 1970 cohort. Their second result is the one that matters, and it is almost never quoted.

1970 British Cohort Study, men Height coefficient on earnings Still significant?
No controls 0.010 Yes
Adding cognitive test scores at ages 5 and 10 0.004 No
Adding extended family controls as well -0.004 No

In the 1958 cohort the pattern repeats: a coefficient of 0.016 on height measured at 16 falls to 0.007 once test scores from ages 7 and 11 are included, a reduction of more than half. The authors’ own summary is that the height premium in adult earnings can be explained by childhood scores on cognitive tests.

Read that carefully, because it is not the same as saying tall people are cleverer. It is saying that the children who grew taller had, on average, already tested better at five and seven, which is before any employer, any interview panel and any salary negotiation could possibly have been involved. Whatever produced both was in place in early childhood. Height was a visible by-product of it, and employers in the 1980s and 1990s were paying for the thing underneath.

That reframing also kills the most popular mechanism. If the premium came from employers preferring tall candidates, controlling for a test taken at age five would not touch it. It does touch it. It removes it.

What UK Biobank adds: a genetic test of cause

Birth cohorts can still be confounded by something nobody measured. The next step is Mendelian randomisation, which uses genetic variants as a natural experiment: variants are fixed at conception, so they cannot be caused by a person’s later income or education.

Jessica Tyrrell and colleagues at the University of Exeter did exactly this with about 120,000 UK Biobank participants aged 40 to 70, using 400 genetic variants associated with height and 70 associated with body mass index. Their findings, published in the BMJ in 2016, were sex-specific and surprising.

  • For men, being genetically about 7.5 cm (3 in) shorter was associated with household income roughly £1,500 a year lower. The association ran through height, not weight.
  • For women, height did little. Being genetically about one stone (6.3 kg) heavier was associated with household income roughly £1,500 a year lower, at the same height.

UK Biobank recruited between 2006 and 2010 and collected income in broad bands, so £1,500 is a mid-2000s figure and an estimate rather than a payslip. For scale, ONS put median gross annual earnings for full-time employees at £39,039 in April 2025. A sum of that order is roughly a few per cent of a full-time salary, which is to say: detectable in a dataset of 120,000 people, invisible in any individual career.

The sex asymmetry is the part that should end the simple story. If the market rewarded size as a signal of competence, it would reward it in both sexes. What the British genetic evidence shows instead is a penalty attached to being short if you are a man and to being heavy if you are a woman, which looks far more like two different social prejudices than like one biological advantage. Taken together with Case and Paxson, the honest summary is that part of the pay gap by height is childhood circumstance showing through, and part of what remains looks like discrimination, with the discrimination pointed at different attributes depending on sex.

Why the figures from abroad look so much larger

Studies from outside Britain report much bigger effects, and the reason is worth understanding before anyone quotes them.

The most-cited recent example is a Mendelian randomisation study of the height premium in China, by Jun Wang and colleagues, published in PLOS ONE in 2020 on 3,427 respondents recruited through a private genetic testing company. Its abstract states that an additional centimetre of height is associated with a 10 to 13% increase in annual earnings. Its own regression tables report 1.30% and 1.03%. And its genetic estimate, the one designed to test causation, is a coefficient of 0.0056, under 0.6% per centimetre, and not statistically significant at any conventional level.

So a paper routinely cited as proof that height causes higher pay contains an abstract that disagrees with its tables by a factor of ten, and a causal estimate that cannot be distinguished from zero. We are flagging it rather than repeating it. If a single centimetre really raised earnings 10%, a 10 cm advantage would double a salary, which nobody has ever observed in any labour market.

The Korean evidence is more solid and more interesting, because Korea is the country that grew fastest in the twentieth century. A study of the Korean Labour and Income Panel Study drew on annual surveys of 5,000 households between 1998 and 2012, giving 34,015 person-year observations for men and 20,233 for women. Average height in the sample was about 171 cm for men and 159 cm for women, with nearly 43% of men between 170 and 175 cm. The authors set the height premium in the context of rapid westernisation and a documented shift in Korean preferences towards taller stature, which is a social explanation, not a biological one.

The broadest European comparison comes from Felix Bittmann’s 2019 analysis of the European Social Survey across 19 countries, covering full-time employees aged 20 to 55. It found considerable differences between countries and a small effect on reaching a leadership position: about 0.15 percentage points per additional centimetre, for women. Country-by-country variation on that scale is itself evidence against a universal biological mechanism.

Two further studies are regularly cited in this area: Aller Zheng’s 2022 work on the Chinese General Social Survey, which links parental education to both children’s height and children’s education and then to pay, and Nargiza Ibragimova’s 2020 study of Russia in the Journal of Biosocial Science, which finds height a significant predictor of income there. The Zheng chain of reasoning is worth noting because it is explicit about the direction of travel: wealthier families produce taller and better-educated children, and better-educated children earn more. Height is a marker along the way.

The four stories, and which ones survive

The literature offers four explanations for a height premium. Setting them side by side makes clear how much of the field is still hypothesis.

Explanation What it claims How it holds up in British data
Health and development Height signals a well-nourished, healthy childhood, which also produced other capacities Strongly supported. This is what Case and Paxson find when test scores absorb the premium.
Ability Taller people have better interpersonal or cognitive skills Partly circular. The measured correlation with cognitive scores is 0.16, and the childhood scores precede the height advantage rather than following it.
Discrimination Employers and clients prefer taller people, particularly in sales and management Plausible but not isolated. The UK Biobank pattern, height for men and weight for women, fits prejudice better than productivity.
Self-perception Shorter people invest less in themselves because of stigma Weakest. Proposed since the 1960s, still supported mainly by small studies and inference.

One historical claim deserves retiring. The often-repeated finding that company directors are taller than average traces to Gowin in 1915 and was popularised again in the 2000s by a chapter of a bestselling book, not by a British dataset. It describes a cohort of American executives born in the nineteenth century, when the population-wide link between height and childhood advantage was far stronger than it is now. Quoting it as evidence about a modern British boardroom is a century out of date.

What this means if you are tall in Britain

Practically, very little, and that is the point worth taking away.

The documented effects are averages across tens of thousands of people and they are small compared with the things that actually determine British pay: occupation, sector, region, hours, and whether you went into higher education. Nobody is paid by the centimetre. If you are tall and doing well, your height is not the reason, and if you are short and doing badly, your height is not the reason either.

Where height does make a measurable difference in Britain is in selection for a handful of jobs with genuine physical requirements, and in sport, where a body’s proportions are part of the specification rather than a signal of something else. We have written about which sports suit tall bodies and about the tallest footballers on record, where height is paid for directly and nobody pretends otherwise.

The other honest caveat runs the other way. Studies of this kind measure pay and test scores, not cost. Being tall in Britain comes with expenses and daily friction that no earnings regression captures, from clothing and footwear to car and train seats, which is most of what this site exists to deal with. If you want the health side rather than the money side, our coaching notes on running when you are tall are more use than any of the papers above.

What the research does not say

  • It does not say tall people are cleverer. The measured correlation with cognitive test scores is 0.16, which leaves about 97% of the variation unexplained.
  • It does not say 70% of intelligence is genetic. That figure describes the share of the small height to intelligence overlap attributable to common genetic variants.
  • It does not show that being tall causes higher pay in general. In the British birth cohorts the premium disappears once childhood test scores are included.
  • It does not show a single mechanism applying to both sexes. In UK Biobank the causal signal runs through height for men and through weight for women.
  • It does not support the figure of 10 to 13% more pay per centimetre. That claim appears in one abstract and is contradicted by the same paper’s tables and its own causal estimate.

The accurate summary is duller and more useful. A childhood with enough food, little illness and enough attention produces adults who are both taller and better equipped. Britain delivered much more of that in 1990 than in 1930, which is why British height rose for a century and has now stopped. The pay gap by height is the trace of that history on a payroll, not a tax on being short. For the mechanics of how childhood conditions turn into adult height, see our piece on which country has the highest average height and what explains the differences.

Sources