Through A Glass Ceiling Darkly: Racial IQ Disparities And The Wealth Of Nations

Now that I’m done with the Necessary Caveats, it’s time we had a look at why exactly HBD/IQ theories are both valid, and relevant to the real world. As I see it, their main import (as interpreted by me) can be distilled into a few logically consecutive, falsifiable statements:

  1. IQ tests are a valid, culturally fair measure of cognitive ability.
  2. It is hereditary.
  3. Race is real.
  4. There are racial/ethnic differences in average IQ that cannot be explained merely by reference to socio-economic or cultural factors.
  5. The US is an excellent “laboratory” to ascertain the average genetic IQ ceiling of different races and ethnicities.
  6. Average IQ influences prosperity, and general living standards.
  7. Consequently, knowing the racial constraints on average IQ’s – i.e., the IQ ceilings – we can estimate the relative development potential of different countries and regions.

All of them have have acquired a great deal of supporting evidence, even though they – or in particular, their linkage – remains taboo for the media and wider public discussion. By the numbers:

1. There is typically a large degree of correlation between various IQ tests, and academic achievement scores (1, 2). Nobody has yet discovered a test which has a negative correlation with a battery of other tests. This implies that there is a common “g factor” behind all types of cognitive ability.

Obviously this allows for very big variations within a single person. But within a group, someone who does well in one test will most likely also do well in another.

The argument that IQ tests are culturally biased is frequently made on the basis that they show differences in performance between racial/ethnic groups. This is a fallacy. In any case, there are IQ tests designed to be culturally fair insofar as they eschew words and test pattern recognition, such as Cattell Culture Fair III and Raven’s Progressive Matrices. These tests have a high correlation with the battery of other tests, i.e. they are valid reflections of g.

2. The scientific literature converges on a range from 40%-80% for IQ heritability. (1, 2) The correlation in scores between twins reared apart is around 75%-80%.

3. Is race a social construct? The commonsense answer is no. Consider that beyond the “my lying eyes” stuff (e.g. skin color/melanin content; epicanthic folds, etc.), there is also evidence of genetic differences in: Physical abilities (west African sprinters; Kenyan marathoners); conformity (Asians); lactose tolerance (whites); alcohol intolerance (Asians); even average penis size (large – Africa; Latin America; small – East Asia). The latter example isn’t nearly as… flippant as it seems, since testosterone levels have a large effect on behavior. There are multiple genetic disorders that only affect certain races or ethnicities, and race specific drugs are now coming online.

But ultimately, this is one of those cases where a picture is worth a thousand words.

This is a genetic map of Europe, superimposed on a geographic map (remarkable how it works out almost perfectly). Note that although there is some degree of overlap between European ethnicities, there are still clear clusters and centers of gravity corresponding to particular nationalities.

Now look at this genetic map (click to enlarge). See that dark blue oval to the center-left? That is Europe. Recall that even within that tiny space there are distinct clusters, with virtually zero overlap between, say, Greeks and Germans. Now note the vast distance that separates Europe from East Asia (center-right), and the three African clusters (bottom).

So it’s really just a minor matter of semantics. Some people shy away from using the word “race”, instead speaking of “genetic clusters”, “population groups”, “groups of common geographic ancestry”, etc. “Race” is short and convenient.PI

4. There is a vast body of global IQ scores (e.g. Lynn). They follow a consistent pattern: East Asian countries tend to score 105, European and Euro-settler countries 100, and sub-Saharan African countries 65-80.

The internationalized standardized PISA tests display the same pattern (as expected, since they load on the same g).

(Internationally, I think the culture argument makes some interesting points. E.g., the Protestant work ethic – reflected even now in the fact that the world’s richest and highest-IQ white countries tend to be Protestant. But there are two problems. First, possible narrative fallacy, e.g. Confucianism, with its connotations of traditionalism and conservatism, was once used to explain why Asian countries lagged behind Europe; but with their success in the past generation, the respect for learning, rule of law, etc. that it supposedly instilled is now sometimes used to explain their success! Can’t have it both ways. Second, what applies in one period can wane in another. Yes, Protestantism fostered human capital development by emphasizing independent Bible reading (i.e. more literacy!), which in turn helped early industrial growth. But today this effect seems to have receded into the past – see the success of south Germany, north Italy, France, Ireland.)

5. The US is an excellent “laboratory” to estimate the average genetic IQ ceiling of different races and ethnicities by virtue of its diversity; standardized education system that produces results that, when broken down by race, are superior to almost every other country in the world; decent equality of educational opportunity; no nutritional deficit among any population group; and post-Flynn effect status.

Within the US, all tests of cognitive ability – IQ, PISA (1, 2), SAT – replicate the global pattern. Though there is variance from test to test, but as a rule, the intelligence hierarchy is as follows: Asian-Americans; whites; Hispanics; blacks. The gaps between Asian-Americans and whites are narrowed than internationally (because Asian-Americans also include medium-IQ peoples like Filipinos and Vietnamese); and the gaps between whites and blacks are narrowed (because, unlike African or Haitian blacks, African-Americans enjoy better nutrition and education.

It must also be noted that whereas most of these tests indicate that Asians closed gaps (in reading, writing) and overtook whites (in math) over the years, there has been no sign of any significant convergence for blacks.

Is it because schools in poorer areas (inner city, where NAM’s cluster) are badly funded? No, per student funding tends to be broadly similar for both inner city and suburban schools. Besides, education funding doesn’t play a major role in results. In Italy, there is no correlation between school funding and performance by province. China gets PISA and IQ results higher than America’s despite spending a tiny fraction of the resources that the US lavishes on each of its pupils. Indeed, it probably doesn’t matter much, as IQ tends to be fixed by the age of the 5.

Is it because blacks come from poorer families on average? IQ is a far more persuasive explanation for why people are poor in the first place. Refer to The Bell Curve (Murray & Herrnstein).

US blacks (85-90) get far better IQ scores than Africans (65-80). These two facts are highly important because in Africa, the IQ’s of many populations are currently constrained by poor nutrition and (in some cases) the different psychologies of pre-industrial and illiterate peoples. In practice, and discounting variation (Africa is the world’s most genetically diverse continent, so it is not impossible that there will be some relatively high-IQ subgroups among them), it seems likely that the average IQ ceiling for African blacks is similar to the actual IQ’s of US blacks. I.e., maybe 85 (lower than 85-90, because US blacks have 20% admixture with whites).

US Hispanics score better than Mexicans or Central Americans, displaying IQ’s (or IQ equivalents) in the low 90’s (Mexico: 88-90; Panama: 84-80). Whites are at around 100. Asian-Americans tend to be in the low to mid 100’s, but there is huge variance (East Asians – higher; South-East Asians – lower).

7. Refer to Education as the Elixir of Growth III. There is a 0.43 correlation between the GDP (PPP) per capita of a country with its PISA/TIMMS scores, which rises to a stunning 0.84 once countries with a post-Communist legacy (low outliers), resource windfalls (high outliers), or offshore financial industries that constitute the bulk of their GDP (high outliers) are removed from the same. (Any correlation of >0.5 is an excellent one in social science). This implies a fairly rigid glass ceiling on GDP per capita for any one country in relation to its IQ. It’s largely invisible, as few people appreciate the importance of human capital to growth, but it’s most certainly there.

At the micro level, the g factor tends to be by far the best indicator of job performance (above grades, interviews, references, etc) – not only in “cognitive elite” jobs such as lawyers or physicians, but also to a significant extent even among menial workers. There is a correlation of 0.9-0.95 between employees in a certain profession and prestige ratings of those professions by the general public. The correlation between IQ and income is around 0.4-0.5.

The importance of average genetic IQ ceilings

We have a fairly strong and convincing (to me anyway) theory that average IQ ceilings depends on race, and that IQ (g, PISA scores, etc) are remarkable closely correlated with economic prosperity. Furthermore, it is almost certain that the causation is largely one way, at least once basic nutritional and literacy problems are solved; otherwise, the Chinese and Koreans would not be outperforming US Hispanics or African-Americans.

Following from the graph of Human Capital Index and income above, there seems to be a point past 450 – about 92.5, in IQ terms – at which (market-based) economies transition from middle-income status, to high-income. If Mexico could raise its human capital to about the levels of their compatriots in the US, this would (going by correlations) enable a massive expansion in its productivity.

Unfortunately, the average genetic IQ ceiling for African blacks is 85, maybe 90 at most. Nonetheless, if Africa could consistently raise it to even the former figure, it would then have a degree of human capital equivalent to today’s Brazil. Though true convergence with developed countries is precluded, reaching Brazil’s levels would be a gargantuan improvement for the living standards of the average African.

In general, it should be possible to construct a “potential IQ” (and corresponding potential GDP per capita level) for each country. They would look something like this (PISA/IQ format):

  • US (500/100) – All racial groups are already performing very close to their genetic potential (Asian-Americans (NOT East Asians) – low 100’s, whites – 100, Hispanics – low 90’s, blacks – high 80’s). The challenge will be in maintaining it (due to Hispanic immigration).
  • China (550/108) – It currently has 103(PISA converted into IQ)-105(IQ), but may still eke out a few more points by totally eliminating malnutrition. Should have no problem in becoming as rich as Korea or Japan with one more generation.
  • India (450/92.5) – Very low (and puzzling) 75(PISA)-82. But also malnutrition is still extremely high, as are endemic diseases; the vegetarianism of a significant portion of the population may also have a negative effect (protein aids brain development). My estimate is that average Indian genetic IQ ceiling is similar to Hispanics, but with huge variance due to caste/ethnic diversity.
  • Russia (500/100) – Moderate score at 95(PISA)-97(IQ). Smaller than ethnically similar Poland (99/100), due to 2 possible factors: (1) Academic focus to exclusion of more general g (Russia does much better on TIMMS, PIRLS); (2) Possible effects of highly prevalent alcoholism on the current cohort being tested. Poland’s score may represent a more accurate “Slavic ceiling.” Natural GDP per capita level would seem to be in between the Germanic countries and the Med, but resource windfall would nudge it closer to the former.
  • Brazil (460/94) – Today at 87(PISA)-85(IQ). Capping off the BRIC’s, the potential is derived by taking the estimated genetic ceilings of US blacks (with adjustment for Brazilian blacks having more admixture) with Med Europe ceiling, and weighing by population ratios.
  • Germanic Europe (520/103) – based on current scores, but will face challenge in maintaining it (due to immigration policies).
  • Med Europe (490/99) – based on Spain, Portugal, Italy results. In a way, one can see the Euro crisis – which is mostly affecting the PIGS – as the invisible hand’s way of bumping down the Med countries to a level more in line with their lower human capital (relative to the Germanic countries).
  • Turkey (470/96) – now at 91(PISA)/93(IQ), but potential based on Greece’s current scores; Turks and Greeks genetically similar.
  • Japan (530/105) – as now, no significant change as immigration is low. But economic difficulties due to debt, negative population growth, ballooning elderly dependency ratios.
  • Israel (?) – is globally irrelevant but a really fascinating case study, complicated by the discrepancy between Ashkenazi Euro-Jewish IQ scores (c.115) and Israel’s mediocre performance of 95(IQ)/94(PISA) which is hard to explain as Ashkenazi Jews still make up more than half of Israel’s population. Also lots of demography has to be taken into account. Deserves a separate post.
  • Australia, Canada (520/103) – as now, but will NOT face difficulties maintaining them because of their Cognitive Elitist immigration policies; as they also enjoy resource windfalls, they will probably do economically better than the rest of the developed world.

Based on the figures above, we can expect that the BRIC’s nations should in principle be able to converge to developed country levels. However, there are two major groups within the BRIC’s. China and Russia are already at a human capital level that enables convergence (China’s is significantly higher, but Russia is already richer, and also has the added bonus of a resource windfall). In contrast, Brazil and India are not currently at human capital levels that enable convergence; however, they do both have the potential to raise them to just the levels needed to break out of the “middle-income” trap and converge. But for that they have to develop their human capital to its full genetic potential. The quantity of the needed change is very significant and, even under rosy assumptions (e.g. +3 IQ points a decade until equalization with genetic potential), the process will take decades. Until they at least the low 90’s, they will remain stuck – especially Brazil, because it is already very rich for its human capital level – with fairly low long-term growth rates.

Whatever their average genetic IQ ceilings, it is highly advisable for all the poorer and low-IQ nations to: (1) improve childhood nutrition (e.g. free vitamin pills at schools seems especially low-cost/high-impact); (2) anti-vegetarian propaganda, where such applies; (3) study the experience of foreign countries, esp. the US which has had a lot of success with maximizing NAM scores, and Finland which (at least on PISA) manages to maximize white scores; (4) explore pharmaceutical and technological means of bridging their IQ gaps with the developed world (e.g. nootropics).

If you like the words I write, and want me to write more of them, consider donating or supporting me on Patreon.