Thread by @KirkegaardEmil

Education attainment is often given as the best example of heritability not being noticeably stronger than shared environment. But this conclusion is somewhat incorrect because of the assortative mating bias. In this study of Finnish and Dutch families, heritabilities were estimated at 55% and 66%, compared with shared environment of 16% and 13%.

frontiersin.org/articles/10.33…

In this Norwegian family study, heritability for education was estimated at 55%, close the above. Shared environment was 27%.

nature.com/articles/s4146…

As usual, beware that variances are deceptive. If heritability is 60% and shared environment is 15%, this doesn't mean that genetics is 4x as important. You need to take the square root to get the path coefficients. These are 0.77 and 0.39. Thus, genetics is about 2x as powerful.

To fully set it out in numbers. If you increase the genetic potential for a person by 1 standard deviation, there education goes up by 0.77. If you somehow increase the relevant shared environment factors, education goes up by 0.39 standard deviation.