In a new paper at the (conveniently open) journal The Winnower (h/t @whyvert), building on his earlier work, geneticist Davide Piffer has tried to calculate the genotypic IQs of various world populations, and how they compare to measured phenotypic IQ:
Here is the abstract:
Factor analysis of allele frequencies was used to identify signals of polygenic selection on human intelligence. Four SNPs which reached genome-wide significance in previous meta-analyses were used. Allele frequencies for 26 population were obtained from 1000 Genomes. The resulting factor scores were highly correlated to average national IQ (r=0.92). A regression of IQs on genetic factor scores of developed countries was used to estimate the predicted genotypic IQs of developing countries. The residuals (difference between predicted and actual scores) were negatively correlated to per capita GDP and Human Development Index, implying that countries with low socioeconomic conditions have not yet reached their full intellectual potential.
As far I can see, the methodology is sound (perhaps apart from a few quibbles over phenotypic IQ sources). But this is exceedingly minor, and doesn’t really change anything in a material way. So I will focus here mostly on the real world impacts these findings would imply.
As one might expect, there is a gap – usually a very significant one – between calculated genotypic and measured phenotypic IQ in developing countries. This is only logical, since developing countries frequently suffer from a variety of maladies, such as malnutrition and parasitic disease load, that are almost entirely absent in the First World. These maladies have a negative impact on IQ. (To a very large extent this also explains the Flynn Effect of secular rises in IQ in the developed world. Effectively, developing nations may be considered as living in the the First World’s past).
Below is a table showing measured IQ in developed countries and predicted IQ from the paper.
|IQ developed countries||Predicted (G.wich) IQ|
|Gujarati Ind. Tx||97.1|
|Indian Telegu UK||95|
|Mende Sierra Leo||83.7|
And here is another table, displaying, for peoples in developing nations, predicted IQ (relative to the standard “Greenwich mean” of 100 for the UK); 100 in the UK); the difference between the predicted and the measured IQ; and GDP per capita in purchasing power terms. They are arranged in order of the size of the phenotypic/genotypic difference.
|Predicted (G.wich) IQ||“Pseudoresiduals” (Predicted minus measured IQ)||GDP per capita PPP (2010-2013)||HDI (2012)|
|Mende Sierra Leo||83.7||19.7||1432||0.368|
|Gujarati Ind. Tx||97.1|
|Indian Telegu UK||95|
Some observations we can consequently make:
Africa: The biggest gaps are all in West Africa. Not only is the region grindingly poor, but it also has perhaps the world’s most acute parasitic disease load, thanks to the hot, humid equatorial climate and low-lying, swampy geography (which the region’s disorganized and resource-pool governments are unable to mitigate) . The gap is lower in Kenya, which as a hilly country can be expected to have a lower parasitic disease load, and non-existent amongst Afro-Caribbean Barbadians, who live in a relatively prosperous country (likely in large part thanks to its “smart fraction”) with one of the most salubrious climates on the planet. On average, it appears that their phenotypic IQ is ~high 60s and their genotypic IQ is ~low 80s. US Black IQ is given as 85, but bear in mind that they have 20% admixture with Caucasoids. (Though on the other hand, US Blacks do slightly better according to PISA, at ~88. If this figure is substituted for in the calculations, then the genotypic estimate for Africans would also rise, though not by very much). Either way, there is thus very substantial room for improvement, but even were that to happen, the overall outlook for self-sustained African convergence to developed world living standards would remain grim.
Latin America: Has a phenotypic IQ of ~mid 80s and genotypic IQ of ~low 90s. As expected, the gap is smaller than in Africa or India (Latin American countries are after all far more socially developed than in West Africa or India, albeit one should should treat straight GDP per capita figures with caution due to the massive levels of inequality). In the developed US, it is basically non-existent, what with Latinos scoring ~low 90s in the PISA tests. The big gap seen in Puerto Rico is intriguing, considering that its close economic ties with the US has allowed it to have a very high GDP per capita relative to its IQ, so lack of money can’t be a limiting factor. But in general, Latin America is already pretty much “where it should be” in terms of prosperity as implied by its level of human capital.
South Asia: Has a phenotypic IQ of ~low 80s and genotypic IQ of ~low to mid 90s. The gap is much bigger than for Latin America, – indeed, comparable to West Africa’s – which is perhaps explainable by dint of India’s greater parasitic disease load, high rates of malnutrition (which is perhaps even higher than in Sub-Saharan Africa), and, in the case of the Punjabis and Bengalis, a strong tradition of FBD marriage, which has very strong negative effects on IQ [AK edit: See also Razib’s comment]. But on the whole, this is positive news. Countries with an average IQ of ~95 include Romania, Greece, Turkey, and Israel (!). If the South Asian continent could successfully resolve its malnutrition, parasitic disease load, and inbreeding issues – admittedly, no small challenge – then it could well expect to eventually rise close to southern European living standards.
Vietnam: Phenotypic IQ of 99, versus a genotypic IQ of 106. Certainly a major surprise, considering it is even higher than China. The gap is substantial, but smaller than in India or Africa. This is not surprising, since although Vietnam has the GDP per capita (PPP) of India, it is led by conscientious Communists and is much better off in terms of social development and nutrition (e.g. meat consumption per person is equivalent to that of neighboring, much richer countries). This makes its excellent performance in PISA 2012, which I wrote about in my introductory post on this site, much easier to explain. Consequently, it would also be a strike against Ron Unz’s theory of the East Asian Exception (i.e. that East Asian IQs are very resilient to negative socio-economic and environmental factors). There would still be a substantial gap between Vietnamese genotypic and phenotypic IQ; it’s just that the former are so phenomenally high that the latter can’t help but be very high as well, since Vietnam is at least in terms of social provision no longer a truly Third World country.
China: No gap. Phenotypic IQ (~105) actually higher than genotypic (~104), which is very unusual for a developing country. Here, however, I must stress two things. First of all, with a GDP per capita (PPP) of $12,000, China has already substantially passed the point at which wealth or the absence of it is a significant limiting factor to realizing genotypic IQ potential. Consult this post where I go into this in greater detail in my debate with Ron Unz. Second, I believe that 105 is, at least today, a substantial overstatement of Chinese IQ. My own estimate based on declassified PISA data is 102.5. So that’s already a gap, even if a very small one. But note also that Asian-Americans scored ~107 in PISA 2009, and Asian-Americans in the US include relatively lower IQ Thais, Filipinos, etc. If we set that as the genotypic IQ of the Han people, then there is still very substantial room for further improvement (with the consequence that the Flynn Effect really does apply very much to East Asians too).
Regardless, short of them embarking on some new Maoist adventure, or getting flooded off the world map by runaway global warming, or getting nuked, or some other similarly apocalyptic scenario, China’s and Vietnam’s convergence to at least Japan’s level is all but certain in the long run.