Scott Alexander Comes Out of the Closet on Race & IQ
Vibe Shift: Scott claims to have been "IQ-Pilled" since 2014, but still ...
From Scott Alexander’s prominent Substack Astral Codex Ten:
How To Stop Worrying And Learn To Love Lynn's National IQ Estimates
Jan 15, 2025
Richard Lynn was a scientist who infamously tried to estimate the average IQ of every country. Typical of his results is this paper, which ranged from 60 (Malawi) to 108 (Singapore).
People obviously objected to this, and Lynn spent his life embroiled in controversy, with activists constantly trying to get him canceled/fired and his papers retracted/condemned. His opponents pointed out both his personal racist opinions/activities and his somewhat opportunistic methodology. Nobody does high-quality IQ tests on the entire population of Malawi; to get his numbers, Lynn would often find some IQ-ish test given to some unrepresentative sample of some group related to Malawians and try his best to extrapolate from there. How well this worked remains hotly debated; the latest volley is Aporia’s
(they say no).
I’ve followed the technical/methodological debate for a while, but I think the strongest emotions here come from two deeper worries people have about the data:
First, isn’t it horribly racist to say that people in sub-Saharan African countries have IQs that would qualify as an intellectual disability anywhere else?
Second, isn’t it preposterous and against common sense to compare sub-Saharan Africans to the intellectually disabled? You can talk to a Malawian person, and talk to a person with Down’s Syndrome, and the former is obviously much brighter and more functional than the latter. Doesn’t that mean that the estimates have to be wrong?
But both of these have simple answers, which IMHO defuse the worrying nature of Lynn’s results. These answers aren’t original to me, but as far as I know, nobody has put them together in one place before.
I answered both questions in separate articles in 2002:
Putting the paywall in here with a couple of thousands of words below the lines:
Going over each in turn:
1: Isn't It Super-Racist To Say That People In Sub-Saharan African Countries Have IQs Equivalent To Intellectually Disabled People?
Here's what I wrote in my 2002 book review of Richard Lynn and Tatu Vanhanen's "IQ and the Wealth of Nations" on the nature vs. nurture question that gets everybody hot under the collar:
A clear example of how a bad environment can hurt IQ can be seen in the IQ scores for sub-Saharan African countries. They average only around 70. In contrast, African-Americans average about 85. It appears unlikely that African-Americans' white admixture can account for most of this 15-point gap because they are only around 17%-18% white on average, according to the latest genetic research. (Thus African-Americans white genes probably couldn't account for more than 3 points of the gap between African-Americans and African-Africans.) This suggests that the harshness of life in Africa might be cutting ten points or more off African IQ scores.
Similarly, West Africans are significantly shorter in height than their distant cousins in America, most likely due to malnutrition and infections.
Scott makes basically the same argument as I did 23 years ago.
Lynn and Vanhanen's book "IQ and the Wealth of Nations" came out in 2002. Since then, a number of other people have taken a crack at this question, both from IQ tests and from school achievement tests (e.g., the World Bank's Harmonized Learning Outcome database). Mostly Lynn’s findings have held up in the big picture.
However, regarding Africa, Wichert’s team of researchers came up with average scores much closer to 80 than to 70 (while others have vindicated Lynn's finding). I have not delved into the issue enough to have an opinion on who is right.
I would prefer the 70 figure to be true for humanitarian reasons: raising sub-Saharan IQ from 70 to 80 sounds a lot easier than raising it from 80 to 90 due to diminishing returns. South Koreans, for example, are now taller than Japanese on average, but it seems unlikely they'll eventually catch up with the people of the Dinaric Alps and become multiple NBA superstars (leaving aside genetic engineering for now).
Of course, Scott assumes that his explaining that the racial gap in IQ could well be due to nurture rather than nature will excuse him. On the other hand, I explained that in the early 2000s, but it's not like the Southern Poverty Law Center has thus announced, "Steve Sailer's moderate reasonableness should be a model for thinking about this topic."
Of course, Scott tends to bore the chuds before getting to the red meat of his argument, while I try to reduce my contention of the fewest possible words.
Then:
2: Can’t You Talk To A Malawian And An Intellectually Disabled Person And Notice That The Former Is Obviously More Functional Than The Latter?
In 2002, I also wrote about this question that got Arthur Jensen interested in IQ and race in the 1960s.
Analysis: IQ defenders feel vindicated
By STEVE SAILER, UPI National Correspondent
LOS ANGELES, June 24 (UPI) -- Several IQ researchers, accustomed to having their field of expertise ignored or denounced as racist and fraudulent, were bemused by Thursday's vote by six Supreme Court Justices to ban the execution of murderers, in effect, who score poorly on IQ tests.
As staunch defenders of the much-maligned concept of the intelligence quotient, these scientists found vindication in the Supreme Court's embrace of intelligence testing, though they cautioned that the Justices' understanding of the complex subject was simplistic.
The IQ experts were particularly amused that newspapers that routinely condemn IQ tests as biased and meaningless were quick to endorse intelligence exams in this case. The New York Times, for example, editorialized, "[I]nflicting the death penalty on individuals with I.Q. scores of less than 70 who have little understanding of their moral culpability violates civilized standards of justice."
Linda S. Gottfredson, co-director of the University of Delaware-Johns Hopkins Project for the Study of Intelligence and Society, said, "Just about the only time I see journalists and liberals take IQ seriously is when it meets their ideological predilections. For example, they treat IQ as real when anyone claims that early intervention raises it, but not when evidence goes the other way. And so it is with crime. We are told we must not link IQ with crime, unless low IQ can be used to roll back the death penalty." …
This is an issue of surprisingly broad social importance. The intelligence researchers noted that while IQs below 70 or 75 are extremely rare in the kind of circles that modern Supreme Court justices travel in, they are much more common in other social settings.
The researchers said that the majority of low IQ individuals do not suffer from medical problems such as Down's Syndrome. Gottfredson noted, "About 75 percent-80 percent of mental retardation is called 'familial,' because it mostly just represents the unlucky combinations of genes that are passed in the normal manner from parents to children. Only a small proportion of mental retardation is due to organic problems, such as chromosomal abnormalities or brain damage. This is just like height. Most very short people are perfectly normal."
The stereotype that most low IQ children are what obstetricians often callously refer to in their notes as FLKs - "Funny Looking Kids" is not true. Elite members of American society tend not to realize this because when an extremely high-IQ person, such as a Supreme Court justice, has a retarded child, it's generally due to organic causes.
As children, these "familial" low-IQ individuals fit in well on the playground, where they may be indistinguishable from their higher-IQ friends. They are normal, except that they run into problems when they need to do the higher-order, abstract thinking that a modern society rewards.
Familial low IQ has been quite common in the past, in other countries today, and in segments of American society.
Interestingly, raw IQ scores have been rising around the world for as far back as testing has been performed on a national basis. To deal with this, test developers periodically raise the number of correct questions a test taker must answer correctly to achieve an average score of 100. This means a vast number of Americans in the past would have scored under 70 on today's tests. Lynn conservatively estimated, "In terms of today's IQs, about 16 percent of whites and about half of blacks" would have scored below 70 in the 1920s. Looking back to the early 20th Century, Gottfredson mused, "It is far-fetched to say that half the population then did not know that murder was wrong."
The causes of this rise in unadjusted IQ scores remain in dispute. Both Gottfredson and Brand doubt whether it reflects a real increase in intelligence. Brand has argued that contemporary people are simply better test takers. In contrast, Lynn has published studies showing that improved nutrition has probably caused some of the rise in raw IQ scores.
Other theories trying to account for the improvement, sometimes called the Lynn-Flynn Effect, include better health, more schooling, and a generally more stimulating mental environment due to more intensive urbanization, pervasive media, abundant shopping choices, and the like.
Further, there are quite a few impoverished countries right now with inadequate nutrition, health, and education, where the national average IQ is in the 70s or even the 60s. In Lynn's new book "IQ & the Wealth of Nations" (co-authored with Tatu Vanhanen of the University of Helsinki), he calculated average IQ scores for 81 countries based on 184 studies published in scientific journals. Lynn found that in 15 of the 81 nations, the mean IQ was below 75. (At the other end of the spectrum, Hong Kong came in highest, averaging nine IQ points more than the United States).
Roughly half of the populations of these low-IQ countries would be considered mildly retarded by the U.S. Supreme Court and thus of diminished moral culpability. Yet, while these destitute societies have undeniable trouble competing in global technology markets that require advanced skills, they would certainly reject the Supreme Court's implication that their average adult is of such limited moral capacity that if he murders, he should be treated like a child.
Below are comments I made in Scott’s comment section in fairly random order:
I knew the late Richard Lynn.
He was a frustrating fellow because, at least in his old age, he was somewhat sloppy, but his mistakes tended to be random enough that his basic conclusions were highly accurate.
For example, when reviewing his 2011 book on the high achievements of Jews, "The Chosen People: A Study of Jewish Intelligence and Achievement," I discovered so many mistakes that I redid his entire analysis of Jewish Nobel Prize winners in physics, chemistry, and medicine-physiology. I discovered numerous mistakes, but also that his conclusion that Jews win an incredibly disproportionate share of hard science Nobels, was absolutely on the money
Yeah, I'd guess that Lynn's Protestant prejudice against Irish Catholics played a role.
In general, Lynn's mistakes tended to be random and thus his overall findings have held up well. But his aversion toward Irish Catholics probably played a role in underestimating Ireland’s IQ, which now does quite well on the PISA.
Russell Warne looked into old data and concluded that Ireland was always pretty close to England in IQ.
But, I dunno ... I could also imagine that the agriculture-oriented culture promoted by the Irish Free State tended to hold down IQ scores a little in the 20th Century.
Perhaps it's kind of like how Israel doesn't have quite the test scores you'd expect. But that's partly by design. The Zionist founders wanted to socially construct a "normal" country for Jews where they would be farmers and soldiers rather than financiers and intellectuals. My impression is that they pretty much got what they wanted, which could be why Israeli Ashkenazis seem less intellectually-inclined than American Ashkenazis.
The notion that IQ test scores represent genetic potential while school achievement test scores represent nurture strikes me as dubious.
That's a common assumption among people who participate in acrimonious debates over the average test scores data from around the world. But as I've been arguing since 2002, the much lower scores seen in sub-Saharan Africa than among African-Americans suggests that impoverished nurture plays a role in holding down African scores.
The high correlation globally between IQ test scores and school achievement test scores suggests to me that both nature and nurture play a role in both kinds of scores.
On the other hand, IQ tests weren't designed to predict genetic potential, but instead individual potential. An adolescent in Malawi who scores a 70 on an IQ test is likely to grow up to be about as economically productive as a Westerner who scores a 70.
There is an interesting study of a district in Kenya where life improved notably on most measures between 1984 and 1998 (e.g., more electricity, more schooling, better nutrition, better healthcare, and other good things), and the children's IQ scores rose by an average of 11 points in just 14 years:
IQ tests evolved to measure potential for performing in the modern world.
The first American IQ test, the Stanford-Binet, was introduced in 1916 by Stanford professor Lewis Terman in what became the heart of Silicon Valley. Lewis's son Fred Terman, Dean of Engineering at Stanford, is probably the best candidate for Father of Silicon Valley, although others point to Fred's friend William Shockley.
By the way, all these people are considered deplorable eugenicists today: e.g., Lewis & Fred Terman Middle School in Palo Alto had its name changed to that of a city councilwoman who introduced bike lanes to Palo Alto, rather than a compromise proposal to drop outspoken eugenicist Lewis's name from the school and just name it after Fred Terman, because the hereditary taint of eugenics is disqualifying unto the seventh generation or something.
Not surprisingly, the former Terman Middle School had the highest test scores of any public middle school in the state of California. It's almost as if real estate values in Palo Alto are so stratospheric because Stanford and its offshoots like HP and Google have successfully carried out the Termans' strategy of finding and recruiting smart people.
My point is that early IQ test designers correctly guessed in which direction life was moving, which helps explain the Flynn Effect: the world has gotten more like what some Stanford professors hypothesized it would a century ago, so subsequent generations have more practice dealing with machine logic and the like in daily life, so they score higher on IQ tests than their forefathers.
But the implication of this is that IQ test designers in the 1905-1940 did a good job of anticipating in what direction daily life would evolve. The Ravens test, for example, looked super alien in the 1930s but by 50 years later, it seemed less bizarre, so the Flynn Effect on it was huge.
Similarly, on the Wechsler, the subtests with the smallest Flynn Effects tended to be in the most culturally loaded ones, such as vocabulary and general knowledge and were bigger on the weirder-looking more culturally-fair subtests. If the good friends Charles Dickens and Charles Babbage came back to life in 2000, Dickens would probably find that people with literary interests didn't seem notably more clever than in 1850. But I suspect that Babbage would quickly find that a much larger fraction of the population could grasp what he was interested in than in 1850.
Are Africans incredibly diverse?
It’s commonly argued that Africans are the most diverse people and therefore, for poorly explained reasons, Richard Lynn was WRONG.
People like to say this, but the truth is that Descendants of American Slaves (e.g., African-Americans whose ancestors arrived via the Atlantic slave trade) are pretty homogenous in ancestry (other than their share of white admixture). Most are descended from Iron Age farmers who expanded over much of Africa, roughly the western half of the continent, from a spot near the Cameroon-Nigeria border, mostly displacing the hunter-gatherers who had lived there previously.
Right. The Bantu, along with their West African relatives up into, say, Senegal, are rather genetically homogeneous. For example, Lagos, Nigeria and Lusaka, Zambia are about 3,500 miles of bad road apart. Yet, as Harvard geneticist David Reich notes:
"…the frequencies of mutations in groups in Nigeria and in Zambia are more similar than the frequencies of mutations in Germany and Italy despite the former two countries being separated by a far greater geographic difference.
Nigeria has a national test for college admissions. The highest scoring Nigerian state is one that's all Igbo. The lowest scoring states are in the Muslim north. Nigeria has big regional affirmative action preferences for northerners to keep southerners from getting all the spots in public universities.
The Igbo appear to be culturally a little like the Parsis in Bombay: a minority group who welcomed the British Empire and molded their culture to emphasize European learning. In contrast, the northern Muslims did not look avidly on the new culture introduced by the British.
Whether the Igbo were seen as smarter than other Nigerians before the British Empire introduced the chance of European-style schooling seems to be up in the air. Historian David Hackett Fischer's recent book African Founders on the African origins of American blacks notes that American slaveowners had differing opinions of the Igbo's economic potential. Slaveowners around the Chesapeake Bay sought out Igbo slaves, while slaveowners further north and south didn't like them. So, I dunno...
Leaving aside more recent arrivals from Africa like the Obama line, African-Americans appear to be a pretty representative sample of sub-Saharans from, roughly, the western half of America. I recently reviewed David Hackett Fischer's "African Founders" about where African-Americans came from in Africa, hoping to find "Albion's Seed" like regional patterns. But, instead, African Americans tend to be fairly mixed from most parts of western Africa.
E.g., the Gullah of the Sea Islands in South Carolina, whose dialect is the most African of various African American dialects and who appear to have mostly lived there at least since the trans-Atlantic slave trade was banned in the early 1800s, are a mixture of people from Sierra Leone and from Congo and Angola, which are a long way apart.
I've been looking for years for evidence that African-Americans include descendants of the more exotic sub-Saharans such as Pygmies and Khoi-San, but haven't found much.
Harvard geneticist David Reich points out that the popular talking point that Africans are the genetically diverse people doesn't really imply what people think it implies.
What about differences in white admixture percentage in African Americans?
This can be overstated as well. At least until recently, most people who self-identified in the US as black tended to have some but not all that much white ancestry, the great majority being, say, 3% to 35% white. There weren't all that many Americans outside of Louisiana (and to a lesser extent South Carolina) who were more than 50% African but less than say 90+% white.
This is due to the workings of the One Drop Rule in America, which strongly encouraged marriages to not cross the color line. In Brazil, in contrast, there are plenty of people who are, say, 75% white and 25% black.
Recent studies of racial admixture have found a modest IQ gradient depending upon percentage of white admixture among kids identified by their parents as having two African-American parents. But because most historic African Americans are pretty similar in terms of their white admixture percentage, it's not that easy to see in daily life. My impression is that people sincerely disagree on this question of whether you can notice a Light and Bright pattern in daily life.
Of course, the admixture studies, while they could have (but didn't) falsify the hereditarian hypothesis, don't prove it either because perhaps the world in some fashion favors fairer-looking African Americans.
My impression is that whites in America don't pay much attention to subtle differences in apparent racial ancestry between blacks: e.g., I've seen lots of movies starring Denzel Washington, Samuel L. Jackson, or Morgan Freeman, but I couldn't tell you who is fairest and who is darkest. I just have them all mentally categorized as African-American.
I'd imagine blacks, on the other hand, pay more attention.
Why do people get so angry over this data?
A huge fraction of well-educated people are simply ignorant of the social science data on racial differences in test scores and crime rates. For example, Bowdoin professor Tyler Austin Harper is a bright and fair-minded guy, but in his review in The Atlantic of Richard Hanania's book on affirmative action, most of his ire was concentrated on a few of Hanania's lines that presume that racial differences in average test scores exist. Harper simply didn't know about the social science findings, nor, it appears, did his editors at The Atlantic:
https://www.theatlantic.com/ideas/archive/2023/09/richard-hanania-origins-of-woke-book/675348/
What about lead poisoning reducing IQ?
If lead poisoning from living near freeways was so deleterious, then rural Southern blacks should have higher IQs than urban Northern blacks. That does not appear to be true.
Similarly, I went to school from 1964-1976 in Sherman Oaks, CA, home to the 101-405 freeway interchange, at that time the busiest in the world. I have looked for evidence that kids who grew up in Sherman Oaks during leaded gasoline are notably stupid, but so far have no findings to report.
A couple of generations ago, the Chicago Tribune editorialized in favor of building the Cabrini-Green public housing project on the grounds that it would stop poor kids from eating lead paint flakes.
It was built and quickly became notorious and now is demolishd.
"an extremely high-IQ person, such as a Supreme Court justice,"
That didn't age well.
It's sort of hard to miss that broadly speaking, higher performing groups in the US tend to have higher performing homelands and the inverse is also true. Obviously in the case of India and some other places we tend to skim off the cream and get unusually intelligent immigrants compared to their national averages but overall it's a good rule of thumb. A lot of right-thinking white Americans like to point to massive outlier success stories of non-whites as what should be possible on a broad scale, a sort of thinking that they would never apply to their own race where most are quite comfortable with the idea that there are a lot of unexceptional or even downright dumb whites out there and they themselves or their kids are in fact quite special.
However, just anecdotally it does seem to me that the reality of IQ and the related performance metrics have a significant genetic component seems to be seeping into places it used to be walled off from. I was recently socializing with some white progressive friends from NYC and completely unbidden during my conversation with the husband he brought up race and IQ and the implications for social policy (we weren't even talking about domestic politics at the time). Obviously knowing my personal politics I suppose he thought he could vocalize a few thoughts/ideas that are probably totally verboten in his circle back home. Now, he did cling to the idea that perhaps there are some social interventions like early childhood education that could meaningfully move the needle on outcomes, but he did accept that they would not eliminate inherent cognitive disparities and that from a moral perspective the left needs to drop its obsession with equity because it's an impossible goal.
This doesn't mean the progressive left will drop its love of DEI, but it does probably mean the overall attachment and willingness to fight for it is weaker.