How do we know genes gain ground with age? #
In 2010, a team led by Claire Haworth (King's College London) pooled raw general-intelligence data from eleven thousand twin pairs across six independent studies — three from the United States, one each from the United Kingdom, Australia, and the Netherlands — and split them into three age groups, with mean ages of nine, twelve, and seventeen. The result was published in Molecular Psychiatry.
The twin method separates three sources of the intelligence differences between people. Identical (monozygotic) twins share all of their genes; fraternal (dizygotic) twins share, on average, half of the genes that vary between people, the same as ordinary siblings. By comparing how similar twins of each kind are to one another, at each age, researchers calculate three shares that add up to one hundred percent: how much of the difference between children traces to genes, how much to shared family environment — everything siblings from the same house have in common — and how much to experiences including measurement error, which siblings don't share.
By age nine, genes already explain 41% of intelligence differences, and family only 33%.
By age twelve, the split is 55% genes to 18% family. By seventeen, 66% to 16%. Thomas Bouchard named this pattern the "Wilson Effect" after Ronald Wilson, who first documented it. Bouchard's (2013) synthesis, combining other twin studies and studies of adopted siblings raised together, describes the same rise in the genetic share with age, without establishing an exact crossing date valid for every child. In the Haworth data above, genes are already above family at the youngest reported age, nine.
How much do genes explain at your age? #
Pick an age between 9 and 24 — perhaps your own. Values are interpolated between the cited estimates and describe differences between people of the same age, not one person’s intelligence.
At age 9:
| Age | Genes | Family | Own experience |
|---|---|---|---|
| 9 | 41% | 33% | 26% |
| 12 | 55% | 18% | 27% |
| 17 | 66% | 16% | 19% |
| 20+ | 80% | 10% | 10% |
Separate studies of Dutch twins (Boomsma et al., 2002) put shared family environment at 55% around age five — more than one and a half times the age-nine figure above. The two are different cohorts and methods, so they cannot be joined to calculate an exact crossing point. From each source on its own, only this can be said: in the Dutch study, at five, the shared environment was 55%; in the Haworth data, at nine, genes were already larger, 41% against 33%. The chart starts at nine because that's the youngest age at which Haworth et al. reported all three shares from the same method, on the same sample.
Why does the genetic share rise with age? #
One well-supported explanation is that as children grow, they increasingly choose what they read, who they befriend, what activities they seek out and which they avoid. Psychologists call this mechanism gene-environment correlation: a child with a genetic pull toward verbal curiosity will seek out books and conversations that exercise exactly that pull, regardless of how many or how few books sit on the shelf at home. Every year, the child builds more of their own environment instead of simply inheriting one from their parents — and that chosen environment, in turn, reflects the child's genes. Statistically, the effect shows up as a rising genetic share, though the real cause is an interaction, not a competition.
Shared family environment doesn't become unimportant to the child — it stays practically significant, especially while the child is young and depends on the decisions adults make. What the data show is something narrower: that the difference between families matters less and less for the differences between children, because siblings who grow up in different households end up, as adults, resembling each other almost as much as siblings raised under the same roof. Ronald Wilson himself, who first described this pattern in the 1970s, insisted that the finding doesn't reduce parents to irrelevance — it relocates their role, from determining a child's trajectory to supporting the potential already there.
Method note
The genes line and the family line in the instrument above connect, with straight segments, the four points cited in the text — at nine, twelve, and seventeen, from Haworth et al. (2010), and at twenty and beyond, from Bouchard's (2013) synthesis, which combines studies of identical twins raised apart with Dutch data (Boomsma et al., 2002) to estimate a genetic share of roughly 80% and a shared-environment share of roughly 10%, stable from around age twenty well into adulthood. The displayed value is a linear interpolation between these points, not additional raw data. The “own experience” share in the instrument is calculated as the remainder to 100% and includes measurement error. At seventeen, that calculated remainder is 18%; the paper reports 19%, and its rounded estimates sum to 101%.
Haworth, C. M. A., Wright, M. J., Luciano, M. et al. (2010). The heritability of general cognitive ability increases linearly from childhood to young adulthood. Molecular Psychiatry, 15(11), 1112–1120. DOI: 10.1038/mp.2009.55.
Bouchard, T. J., Jr. (2013). The Wilson Effect: The increase in heritability of IQ with age. Twin Research and Human Genetics, 16(5), 923–930. DOI: 10.1017/thg.2013.54.
Plomin, R., & Deary, I. J. (2015). Genetics and intelligence differences: Five special findings. Molecular Psychiatry, 20(1), 98–108. DOI: 10.1038/mp.2014.105.