Why Two Strangers Raised in the Same House End Up Nothing Alike in Intelligence

Two identical twins, separated at birth and raised by families that never met, end up more alike in IQ than two ordinary siblings who grew up under the same roof. And two children with no genetic relationship, adopted as babies into the same family, reach adulthood barely alike in IQ at all.

It sounds like the order got flipped, but it didn’t. That counterintuitive crossing of data is what has made twin and adoption studies one of the most cited tools for understanding where differences in intelligence come from.

The research design: comparing what’s shared and what isn’t

In The g Factor (1998), psychologist Arthur Jensen explains why certain pairs of people are especially valuable to science:

• Monozygotic (identical) twins share 100% of their genes.
• Dizygotic (fraternal) twins share, on average, 50%, like any pair of siblings.
• Adopted people, raised alongside genetically unrelated siblings, share the family environment but not the genes.

By crossing these groups (raised together, raised apart, genetically related or not), it’s possible to estimate, in a given population, what proportion of the differences in IQ between people is associated with genetics and what proportion with the environment shared within the family.

The best-known study of this kind is the Minnesota Study of Twins Reared Apart, led by Thomas Bouchard from 1979. In its 1990 report in Science, the team concluded that around 70% of the variation in IQ in their sample was associated with genetic differences, and that raising twins together or apart barely changed how alike they were on many psychological traits.

The numbers that don’t fit intuition

Jensen gathers the average correlations, weighted by the number of pairs studied across dozens of investigations. These are the most revealing:

• Identical twins raised together: correlation of +0.86.
• Identical twins raised apart: correlation of +0.75.
• Fraternal twins raised together: correlation of +0.60.
• Ordinary biological siblings raised together: correlation of +0.49.
• Genetically unrelated people raised together from infancy: correlation of +0.25 in childhood, and -0.01 (practically zero) in adulthood.

Two identical twins who never shared a home are more alike in IQ (+0.75) than two biological siblings who did grow up together (+0.49). And two people with no genetic kinship, who shared a house, parents and routine from infancy, end up as adults with no appreciable correlation between them.

The Wilson effect: heredity weighs more with age

The pattern has a time dimension that often goes unnoticed. Psychologist Ronald Wilson was the first to show clearly, in longitudinal twin studies in the 1970s and 1980s, that the heritability of IQ increases from childhood to adulthood. Thomas Bouchard named the phenomenon the Wilson effect in a 2013 review:

• The heritability of IQ rises with age until it levels off at around 0.80 by age 18 to 20, and stays at that level in adulthood.
• The weight of the shared environment (what siblings have in common from growing up in the same house) falls in parallel, settling at around 0.10 at those same ages.

Bouchard specifies that these conclusions apply to Western industrial democracies, where almost all the studies have been done.

One way to understand it: in childhood, the shared environment (the same house, the same routine, often the same school) leaves a mark on how cognitively alike siblings are. Over time, each person chooses their own environments, hobbies and challenges to a greater extent, and their level aligns more with their own inclinations than with the house they grew up in.

What the home does change: the level, not the resemblance

The fact that the resemblance between adoptive siblings fades doesn’t mean the family has no influence on intelligence. Correlation and average level are different things.

The study that shows this best was published in 2015 in PNAS. Kenneth Kendler, Eric Turkheimer and colleagues used Swedish registries to find 436 pairs of male siblings in which one had been raised by his biological parents and the other adopted by another family. When they compared their scores on the military conscription aptitude test, at ages 18 to 20, the adopted brothers scored on average 4.4 IQ points higher than their non-adopted brothers. The higher the educational level of the adoptive parents, the larger the difference.

In other words, the home can shift the average level even if it doesn’t make two children raised together end up alike.

Heritability depends on the environment

There’s a second important nuance. Heritability isn’t a fixed property of IQ, but of a population under specific conditions.

In 2003, Eric Turkheimer and his team analyzed 7-year-old American children from families with different income levels. In better-off families, IQ differences were associated mainly with genetics; in the poorest families, the weight of genetics was much smaller and that of the environment much larger.

A later meta-analysis by Elliot Tucker-Drob and Timothy Bates (2016), covering almost 25,000 pairs of twins and siblings from six countries, confirmed that pattern in the United States, but didn’t find it in Western Europe or Australia. One possible explanation is that where healthcare and education are more accessible, poverty limits the development of each child’s abilities less.

What you can’t conclude from this, however tempting

This kind of data lends itself to more categorical headlines than it deserves, so the limits are worth pinning down:

• These are population averages, calculated over thousands of pairs. They can’t predict anything about a specific family or person.
• Jensen himself notes a caveat: twins raised “apart” often share the prenatal environment and the first weeks or months of life, which may slightly inflate the +0.75 correlation.
• Adoptive families tend to have a more uniform level of resources than the general population, which can make the weight of the shared environment seem smaller than it would be in a more diverse sample.
• None of this says anything about personality, values, emotional wellbeing or the quality of the bond with the family, dimensions these IQ-focused studies don’t measure.

What this crossing of data does leave is an idea worth thinking over slowly: having grown up under the same roof says less than it seems about how cognitively alike two people will be in the future, even if that roof may well have influenced the level each of them reaches.

Another equally counterintuitive finding about the g factor is in the test that predicted road deaths better than medical history, a large-scale study on accident rates that reached an equally surprising conclusion.

Sources

• 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.
• Bouchard, T. J., Jr., Lykken, D. T., McGue, M., Segal, N. L., & Tellegen, A. (1990). Sources of human psychological differences: The Minnesota study of twins reared apart. Science, 250, 223-228.
• Bouchard, T. J., Jr., & McGue, M. (1981). Familial studies of intelligence: A review. Science, 212, 1055-1059.
• Jensen, A. R. (1998). The g Factor: The Science of Mental Ability. Praeger.
• Kendler, K. S., Turkheimer, E., Ohlsson, H., Sundquist, J., & Sundquist, K. (2015). Family environment and the malleability of cognitive ability: A Swedish national home-reared and adopted-away cosibling control study. Proceedings of the National Academy of Sciences, 112(15), 4612-4617. doi:10.1073/pnas.1417106112
• Tucker-Drob, E. M., & Bates, T. C. (2016). Large cross-national differences in gene × socioeconomic status interaction on intelligence. Psychological Science, 27(2), 138-149.
• Turkheimer, E., Haley, A., Waldron, M., D’Onofrio, B., & Gottesman, I. I. (2003). Socioeconomic status modifies heritability of IQ in young children. Psychological Science, 14(6), 623-628.