Back in 2015 I wrote:
Many times on this blog I have claimed that the correlation between IQ and years of educations (indeed IQ and academic success in general) is 0.65. I have based many arguments on this figure which I had assumed was correct since it came from none other than the late great Arthur Jensen who cited it on page 279 of his 1998 book The g Factor. Well it turns out the figure is no longer true, and hasn’t been true for at least several decades. The standardization sample of adults (age 25+) on the WAIS-III revealed the correlation between full-scale IQ and years of education has sunk to 0.55. It seems back in the 1950s, when the first WAIS was standardized, the correlation between IQ and years of education was about 0.7, but by 1978, when the revised WAIS was normed, it had already sunk to the mid 0.5s.
It turns out this is only true when you look at all adults lumped together:
When you limit yourself to adults in specific age groups, the correlation remains remains around 0.7:
The massive correlation between IQ and years of education appears to be because high IQ people stay in school longer. It seems not at all because school raises IQ. This was proven by Dillon (1949):