I have previously complained that some scholars overestimate the average IQ of black Africa by using samples that have access to a lot more schooling than many people in the region. A commenter suggested I do my own estimate of the average IQ of sub-Saharan Africa by correcting for this problem.

One of the best studies of black African IQ was done by the late J.P. Rushton who decided to test university students under the best testing conditions imaginable. Why university students? Because if he could show that even the best and brightest black Africans, tested under superb testing conditions, still scored very low, then he could be certain the average IQ in Africa was every bit as low as scholar Richard Lynn had claimed. Rushton describes the study here:

Rushton found that on the Raven IQ test, the average black university student in South Africa has an IQ of 84 (U.S. norms), though this might be reduced to 83 since the test norms may have been inflated by the Flynn effect (the students were tested about seven years after the Raven’s 1993 norming).

An IQ of 83 on U.S. norms is equivalent to about 79 on U.S. white norms. The reason the IQ is so much lower using white norms is that the white American mean is not only higher the American mean, but the white American SD is narrower than the American SD (see here for the math and rationale).

Estimating the average IQ score of sub-Saharans

Now if the average South African black University student has an IQ of 79, what does that tell us about the average IQ of black South Africans as a whole? Assuming the average first year university student has about 13 years of schooling (and given the high dropout rate in South Africa, the average first year university student never gets much beyond the first year) and the average South African had about 8.2 years of schooling at the time of this study, then the students Rushton tested were 4.8 years more educated that the mean of their country.

Now in the U.S., for whatever reason, educated people score higher on IQ tests, by about 3.54 points (3.66 points using white norms) per each year of schooling. Assuming the same pattern applies in South Africa, black South Africans who are 4.8 years more schooled than the average South African, should have IQs that are 4.8(3.66) = 18 points higher than the average South African. So if the black South African first year university students average IQ 79, the average black South African should average IQ 79 – 18 = 61.

Estimating the real intelligence of sub-Saharans

Now, just because I estimate the average IQ of black Africa to be 61 does not mean I believe this is an accurate reflection of their intelligence. The average South African adult (circa 2000) hardly attended high school let alone uinversity (largely for structural reasons, independent of his ability) and it is known that each year of missed schooling shaves 1.8 points off one’s IQ score (though probably not one’s real intelligence), so if South Africans stayed in school as long as Westerners, they would probably average 70 instead of 61.

Some readers might be confused as to why I’m arguing each year of missed schooling shaves 1.8 points off IQ, when above I said people with more schooling average 3.66 IQ points higher for each year more schooled. The 1.8 figure is the independent effect of more schooling on IQ, while the 3.66 figure is just the difference in IQ between people with different schooling (including both the effect of schooling on IQ and the effect of IQ on schooling)

Estimating the genetic IQ of sub-Saharans

So IQ 70 probably reflects the actual level of real intelligence of black South Africans. Of course their real intelligence could probably be higher if they had received First World nutrition, particularly during the prenatal stage when the brain is most growing. I’ve previously documented that black Africans are probably about 0.8 SD below their genetic potential for height. Assuming they are also 0.8 SD below their genetic potential for intelligence would raise their IQs by an addition 12 points, bringing them to about 82.

Comparison with white Victorians

It is interesting to compare the adult Raven IQ of 61 for black Africans with the adult Raven IQ of 70 for white Victorians. Like black Africans, lack of schooling made the Raven unfair to Victorians and correcting for the schooling effect, raises their IQs to about 78, which might be a good estimate of their real intelligence.

Also like modern Africans, even real intelligence was stunted. Victorians appear to be even much more malnourished than modern Africans, being about 1.5 SD shorter than whites today. If we assume malnutrition also stunted their real intelligence by 1.5 SD (23 IQ points), that fully explains why their real intelligence was nearly 23 points lower than whites today who score 100.

Of course if the dysgenics theory is true, it would imply that Victorians should be genetically much smarter than modern whites and thus be much higher than modern whites after correcting for nutrition. Of course if dysgenics is true, then people have probably also become genetically shorter despite becoming phenotypically taller, which means I under-corrected IQ for nutrition and doing the full correction would make white Victorians much smarter than modern Whites.

While I’m open-minded to the dysgenics theory, I think that dysgenics should have caused a major decline in a test called Digit Span given the insensitivity of this test to both nutrition and schooling, and that apparently hasn’t happened.