Below our my best estimates of the genetic IQs of many populations. Given the lack of precision, all estimates rounded to multiples of 5.
As one can see, there seems to be a pretty strong negative correlation correlation between IQ and skin colour. Those at the top of the totem pole are all light skinned. Those at the bottom are mostly dark skinned. This is likely because skin colour is a proxy for how far ones ancestors were able to migrate from the tropics and this likely would have selected for the cognitive ability to adapt to an environment unsuited to our African bodies.
It’s interesting that Arctic people have both lower IQs and darker skin than their ancestral climate predicts.
There also seems to be a negative correlation between antiquity and IQ. For example apes appear in the fossil record 25 million years ago and they average IQs of only 15. Ashkenazi Jews appeared only about 800 years ago and they average about 105.
By the end of the 20th century, Whites, Indians, Coloureds and Blacks in South Africa averaged IQs of 98, 92, 82, and 69 (UK norms). Although all four “races” were in school when tested and used to paper-pencil tests, and although the test used (Standard Progressive Matrices) was culture reduced, there were enormous difference in the quality of environment the four races were exposed to including the biological environment which affects brain growth. I tried to correct for this by comparing the different rates of stunting (low height), but as commenter “some guy” noted, this measure is too confounded with genetic height to be a good proxy for environment.
Perhaps a better proxy is Human Development Index (HDI) which combines income, education, and life span. While life span is confounded with genetics to some degree, being reared in a poor uneducated home is a clear environmental effect.
In 1991, South Africa’s Whites, Indians, Coloureds, and Blacks had HDIs of 0.901, 0.836, 0.663, and 0.50 respectively (see table III).
To put these numbers in perspective, I wanted to know the HDI of black Americans because black Americans have continued to score 15 points below white Americans for roughly a century, and it doesn’t seem to matter whether they’re reared by their biological parents or adopted into White professional homes. Thus black American HDI is perhaps a ceiling beyond which environment much affects IQ.
An article in The Atlantic (October 14, 2014) by Theodore R. Johnson reported their HDI but I don’t think a 2014 figure is comparable with the 1991 figure for South Africa’s races, both because living standards have changed over time and so has the method of calculating HDI. But assuming the relative ranking of black America has been similar over time (slightly above Saudi Arabia and slightly below Qatar) then in 1991 they had an HDI of about 0.705 (Saudi Arabia) to 0.745 (Qatar) or roughly 0.725.
Black Americans (1991 HDI 0.725) score 15 points lower than white Americans, but 16 points higher than South African blacks (1991 HDI 0.5). The former gap is arguably 100% genetic judging by the results of the Minnesota transracial adoption study suggesting HDI has no effect on IQ once you hit at least 0.725. The latter gap is probably 31% genetic, because Black Americans are only 75% Black. Adjusting for this reduces the latter gap to 11 points suggesting that for people with HDIs below 0.725, subtracting the HDI from 0.725 and then multiply by 48.9 gives an estimate of how much Raven IQ (taken by people in school) has been supressed by environment.
So the IQs of South African Whites and Indians (98 and 92 respectively) are probably not supressed because their 1991 HDIs were above 0.725 but Coloureds were 0.062 below this threshold, so multiplying by 48.9 suggests their IQs were supressed by 3 points. This would raise them from 82 to 85, the same as African Americans.
Native Americans
According to one study, “by 2001, the American Indian and Alaska Native population and the Canadian Indigenous population had…HDI scores comparable to South Korea or the Czech Republic and Belarus or Trinidad and Tobago, respectively.” In 1991 Czech Republic had an HDI score of 0.733 and Trinidad and Tobago, 0.67. Averaging just those two (couldn’t find 1991 data on the other countries) suggests American Indians and Arctic people had a 1991 HDI of 0.702. This suggests their Raven IQs are supressed by 1 point. This is consistent with a study that found that several years of foster care in white homes did not at all improve the IQ of these.
Stunting is defined as being at least 2 standard deviations (SD) below the average of the reference group for your height, sex and age. Wasting is defined as being at least 2 SD below average weight for your height and sex. The reference group is an international sample of socio-economically advantaged breast fed children.
Commenter “Someguy” pointed out that using stunting rates to measure how malnourished a certain ethnic group was is faulty because some groups might be genetically shorter than others. I agree but was unsure if this genetic difference would show up in young children. After all, the World Health Organization (WHO) claims children of all ethnic groups grow similarly when breast fed and born and raised healthy, at least up to age five, and so a single reference group can be applied internationally.
Further, Arthur Jensen stated that (in his population) height has a heritability of 0.95 in early adulthood, but only 0.30 in infancy. With genes having only moderate effects in the first years of life, this made it seem quite plausible that all races (with the exception of pygmies) could use the same growth chart to measure nutritional status.
However even comparing different races in First World countries show large differences in early childhood height. For example averaging across ages zero, 0.5, 1.5, 2.5, 3.5 and 4.5 years, Dutch boys (see table 1) are 1.02 SD taller than Japanese boys (Japanese SD used since the Dutch one not provided).
To be sure this is much smaller than the 1.93 SD height gap that shows up at age 17.5, consistent with the square root of height heritability being almost twice as high in early adulthood compared to infancy.
It is sometimes said that in most ethnic groups, well nourished kids are within half an SD of the WHO growth chart so close enough, but if one population is half an SD above and another half an SD below, that’s a 1 SD difference!
And because small differences in the mean create huge differences at the extremes, a 1 SD gap means the shorter group will show about 7% of their population stunting while the taller group will show only 0.6% despite the fact that both groups have equal nutrition! That’s a little too much error for comfort.
If the World Health Organization is serious about measuring malnutrition, they should invest in getting us better polygenic scores for height. Only then could they say with accuracy that a given population is below their genetic height potential, and thus malnourished.
By the end of the 20th century, Whites, Indians, Coloureds and Blacks in South Africa averaged IQs of 98, 92, 82, and 69 (UK norms) and malnutrition rates of 5.7, 10.9*, 18, and 32 percent respectively (see table 2.13 below):
From The Global Bell Curve (2008) by Richard Lynn
Malnutrition is defined here as the percentage of the population that is stunted. Stunted is defined as two or more standard deviations (SDs) shorter than comparable healthy populations. In theory only 2% of the healthy reference population should be stunted.
But nutrition is not a discrete variable. In theory there’s a perfect continuum between optimum and suboptimum processing of nutrients and each population has their own bell curve.
So if 32% of South African blacks have sub-optimum nutrition compared to only 2% of well nourished populations, then that tells us that the 32nd percentile (-0.47 SD) on the former’s bell curve equates to the 2nd percentile (- 2 SD) on the latter’s. Assuming roughly equal standard deviations, it suggests South Africa’s black bell curve is shifted 1.53 SD to the left of what was considered optimum at the time.
So the stunted children are just the tip of the iceberg. The average black child in South Africa should be 1.53 SD below his genetic potential in physical growth. What about brain size? Also 1.53 SD below genetic potential? Given the 0.4 correlation between IQ and brain size, this would predict IQ would be 0.4 (1.53 SD) = 0.61 SD below genetic potential. Multiplying by the IQ standard deviation of 15 points, this gives 9 points of impairment caused by malnutrition. Adding this to the IQ 69 of black South Africans gives IQ 78.
I applied the same calculations to all the major groups of South Africa:
It is interesting to compare the estimated black genetic IQ of 78 to the average IQ of 85 for African Americans, a presumably well nourished population judging from the fact that they are virtually indistinguishable from white Americans in height. But genetically African Americans are only 75% sub-Saharan, so adjusting for this, they go from 15 points below IQ 100, to 15/0.75 = 20 points below 100 = IQ 80. In other words, virtually all of the difference between unmixed American and South African blacks vanishes when we adjust for presumably stunted brain size.
Similarly, the IQs of Indians and whites rise to their corresponding levels when reared in the UK (a well nourished country).
*The malnutrition of Indians were estimated from the line of best fit predicting poverty rates in table 2.13 to malnutrition rates in table to 2.13: Malnutrition = 0.65(poverty) – 2.79.