"We already have abundant evidence that black Americans score about 85 on IQ tests, as compared to 100 for whites. If the average IQ in Sub-Saharan Africa is 80, this would mean the massive difference in environment between Sub-Saharan Africa and the US reduces IQ by only 5 points, yet the comparatively small difference in environment between black and white Americans somehow reduces it by 15 points."
However, the comparison of IQ between 'black Americans' and Sub-Saharan Africans only approaches validity if the admixture of black to white in the 'black American' group is known. Many 'black Americans' carry varying degrees of white genetics. This is more explanatory of IQ difference than the environment.
I thought it's well known that Black Americans average 15-20% white. Anyway, this estimate will be rough, as we don't know if there was selection in who was sold into slavery, and what selection you have had over two centuries. Sub-Saharan Africa is a big place. There's going to be a lot of variation between groups.
The strict environmentalist PoV is obviously wrong.
The PoV that national IQ numbers are perfect representations of genetic potential is obviously wrong.
Most things in life are caused by a combination of genetics and environment. That's a general rule. Of course some things are all genetics or all environment. But generally most things are a mix.
Most IQ gaps are much less well studied and much less well understood than the BW one in the US. That is the one we have studied a lot and where we have tried to correct the environment a lot. And understandably, because of all of the awful things that happened in the US like slavery, lynching, redlining, Jim Crow, and so on people are sensitive about it.
We know that one is mostly on g from adoption studies, studies on Spearman's hypothesis, and so on. This one has been studied to death. Of course this one is >50% genetic g differences. Whether it's 80% or 100% I don't know, you have the B are comparatively better at memory for instance so it will depend on your g-loading.
I don't think we're disagreeing to any great extent here. No-one, including Lynn, disputes that SSA's IQ is lowered by various adverse environmental factors (in his estimate, by about 13 IQ points.) I'm just a little skeptical that the environment in somewhere like Saudi Arabia is adverse enough to create a comparable gap.
Yeah I'm not so familiar with the Saudi IQ figures, but my guess is their genotypic IQ is about 90. Their PISA score is like 80-85, but they have inbreeding and a bad education system so I would guess the genetic IQ is about 90.
Too many in research today suffer from dead white male envy. Produce quality research if you want respect. Secondly, during research on mulattos in America, I discovered that the discrimination they encountered was understated and possibly there could be a link between skin colour and IQ. Lynn, did a study that was quite good, but I think we need more admixture studies. The MQ studies are okay, but I want bigger samples with people from elite mulatto families.
James Flynn once proposed to gather accurate data on Sub-Saharan IQ. Nobody wanted to be associated with such a study. Such a study is simply not possible today. It's just too inflammatory. I think tests like the PISA are the only measure that we will have for the foreseeable future.
The critiques of Lynn's methodology aren't entirely off-base, given he was essentially scraping data in a somewhat ad-hoc way from whatever studies were conducted in a given country, often for rather different purposes (one of my favourite examples is a study on the prevalance of Konzo where he took the unaffected control group as indicative of typical IQ in the region.) He was working off a shoestring budget, so I don't think he had much choice in most cases.
In any case, while you can poke holes in specific studies the overall trend in the numbers is just too overwhelming to ignore. (His critics, of course, have no interest in sending out an army of researchers armed with raven's matrices to go get higher-quality data on SSA's IQ. They know perfectly well what they would find.)
From what you say I can well imagine that scrupulous scientists in the herediarian camp and any scientist, scrupulous or not, from the environmentalist camp could pick holes in Lynn's data. But since I'm neither a scientist nor scrupulous I'm happy enough to give him a pass if things look more or less right. I suspect no amount of evidence would be good enough for the environmentalists.
Environmentalism is a straw man. But I find way too many of Lynn's numbers to be too low to be plausible. Iran 85 or so? Really? This does not make sense given the number of top scientists from Iran. There are twice as many Iranians as Black Americans, but many more than twice as many Iranian Fields Medalists, Nobel Prize winners, top scientists, and so on.
Israel at 92? Cremieux had a post where he showed that the scores go up when you correct for psychometric bias and isolate the g-factor. It's more like 97 or 98, with 102-103 for Jews, majority of whom are non-Ashkenazi, whereas Lynn estimated 103 for Israeli Ashkenazi Jews based on some implausible hypothesis about selective migration.
Romania at 90 or so? Really low numbers for the Balkan countries? Albania and North Macedonia down into the 80s? These numbers are just not plausible. And it's noteworthy that post-Communist countries seem to have depressed IQs, you can look at West Germany versus East Germany.
The OECD has documented that people in some countries put in less effort on these tests and do better on the first half of the test whatever that is. I found this from the PISA website:
"In reading, on average across OECD countries, students who took the test in the second hour (in most cases, after completing an hour-long mathematics test) scored 14 points lower than students who took the test in the first hour – a large difference. Large performance declines during the test of between 20 and 30 score points were observed in Iceland, Israel, Latvia*, Albania, Qatar, Slovenia, Malta, Argentina and Norway (in descending order of the size of this difference) (Table I.A8.17)."
So I suspect that many of these national IQ numbers are also capturing something about trust rather than intelligence. In more than one Eastern European country and in Israel people will just try less hard on a test that counts for nothing.
Many of the numbers are undoubtedly somewhat inaccurate. However, this is true of all datasets. According to Gust and colleagues' dataset, Niger's student achievement is equivalent to an IQ of 56.
Mind you, but these IQ estimates by Lynn were never intended to reflect latent g. They don't assess for possible measurement bias using MGCFA and IRT or track low effort response or rapid guessing behavior using IRT. My point is that, if we follow your criticism, nearly all IQ papers should be deemed rubbish because they never tried to assess latent g, but instead, merely observed total IQ. Instead, a major concern should be about sample representativeness and test condition.
I agree and I certainly don't think that "all IQ papers should be deemed rubbish" or retracted. All I'm saying is their results should not always be taken at face value, and people should understand that Lynn national IQ is certainly an imperfect measure of national genetic g. Heck even Raven has a g-loading of like 0.5. Lynn agree with this, though he was wrong about some things. You agree with this. Aporia itself agrees with this and likes my comment. I'm not arguing with you. I'm arguing with Simon Laird.
Perhaps some context is in order. A few days ago I was arguing with Simon Laird on Substack that Iran is poor largely because a terrible government theocracy that is socialist and anti-American, and his response was that the reason that the only reason it's poor is that it has a low IQ and there are sanctions. And I said, you can't take the Lynn figure at face value. And I got a "the Lynn figure is just as likely to be an overestimate as an underestimate". I think he quoted a figure of about *80* to me. I can pull up the conversation. (edit: I found it, see here https://substack.com/@usuallywash/note/c-83086866?utm_source=activity_item ad here https://substack.com/profile/16148027-simon-laird/note/c-83086490 )
Simon Laird has actually written an article on Aporia https://www.aporiamagazine.com/p/africa-is-not-doomed-to-poverty about the quality of governance in developed countries and how it matters! In fact I agree with this article of his, and I think it's hilarious that he wrote this, in view of his comments about Iran that I linked to in the above paragraph.
Of course if woke people are trying to throw away all Lynn research because Lynn was imperfect and wrong about some things (whether because of sloppiness or because of some political motivation) then that is bad. He is definitely not the first person to be wrong and he did important work in the area! But all too often I do see people taking Lynn figures at face value in ridiculous ways, for instance using them to defend Iran's form of government and blaming the terrible state Iran is in on the supposedly 80-ish IQ of the Iranian people. (!).
You might as well defend communism by pointing out that East Germans did worse than West Germans on IQ tests and that therefore this must be genetic and must have nothing to do with communism.
Maybe you noticed but my comment was only directed at your second paragraph. It's the observation that the estimate can't be trusted because it doesn't account for measurement bias and latent g (which in this case means this criticism can be directed at nearly all IQ papers) that really throws me off.
In general I don't think I disagree with your other points. If you single out a few estimates, and this has been known for decades now, you might see some weird estimates. Some old estimates that you could tell were odd such as the Irish IQ at 87.
Ah OK, so when cremieux corrected Israeli IQ for "psychometric bias" (not even latent g, just bias internal to the test, whether caused by low effort, bad education, whatever) and it went up 5 points, you're saying we still report the lower number as the IQ, since that's what IQ means. Got it. IQ can be influenced by environment and various other things. Yeah, it's true and well-known that Lynn underestimated Irish IQ, probably because was a Northern Irish Protestant and had an axe to grind against the Irish.
I think that "can't be trusted" should be unpacked a little bit. Of course if IQ is defined by scores on various tests then IQ scores can be trusted by definition, if your goal is to accurately estimate the IQ. Insofar as IQ is predictive of various things though, I thought it's generally the g-factor, which is why Cremieux is always adjusting for psychometric bias. I'm sure you would know better than me. So of course you are right that these scores can be "trusted", but for most useful intents and purposes one does not want to take them at face value.
Ideally, all IQ reports should reflect latent g, if possible based on bifactor CFA model. And low-effort response, anxiety, rapid guessing behavior should be also accounted for as well. But obviously, it's easier said than done. It's so much more convenient to just collect FSIQ data, especially for meta-analysis. If you meta-analyze IQ estimates based on latent g, your data is reduced to maybe 1% at best. If you require studies that account for all other external factors, you get no data. This is why you might want to use national assessments as well, as a robustness check. The interesting outcome is that some of the estimates regarded as too low, such as African IQs, seem to match closely the estimates from national assessments. As opposed to IQ tests, there have been a few studies that have tested for possible national differences in rapid guessing behavior and other external factors using national assessments.
Do you know if there's a non-paywalled version of this article on East/West german IQ convergance? I don't want to dismiss any potential causal confound offhand.
I will say that while I'm a little puzzled by the enormous score disparity between the eurozone and much of the middle east, Saudi Arabia's national IQ does seem to be pretty consistent with their abysmal PISA scores. And on a per-capita-purchasing-parity basis they're spending roughly as much on education as Germany does.
Are the Saudis just lazy when it comes to taking tests, as you suggested was true for other countries? "Large performance declines during the test of between 20 and 30 score points" would only correspond to an IQ gap of 3-4 points or so. I've heard that a lot of Saudi citizens can fall back on government make-work positions or get economically-irrelevant degrees in theology, so maybe that has some effect on motivation, but is it really going to be ~20 IQ points worth of apathy?
Yes, I've considered that, but outside of a few isolated pockets of the muslim world with terrifying levels of consanguinuity it's not really common enough to account for more than a few IQ points.
Well Crem found a 5 point adjustment in Israeli IQ. If someone does 4 points worse on the second half than the first half, their genetic potential is probably even better than how they did on the first half, since they were somewhat lazy on the first half too. PISA also definitely doesn’t have a g loading of 1.0. Even Ravens matrices is 0.5 according to a paper Emil linked. Educational quality matters too.
The Saudi score on the PISA reading corresponds to an IQ somewhere in the 80-85 range. If you believe a few points are off due to low effort and/or bad education system and a few points are off due to inbreeding you get a genotypic IQ of 90 which makes sense.
I suspect PISA math is less g-loaded than reading. If you have a bad education system you won’t do well even if you have high g.
Inbreeding is a disadvantage for a population with genetic defects it needs to dilute, and it is an advantage for a population that needs to avoid importing genetic defects. Incest rules vary and are not based on genetic considerations. Outbreeding customs in settlements, clans, extended families, etc are based on the need for people collected in small groups to get along; and making insiders sexually off-limits depresses effects of jealousy, envy, and rejection resentment.
Isn’t it the other way around? Latent g has been shown many times to correlate fairly strongly with IQ. Because g is hard to measure, they instead tend to measure IQ, assuming that with the correlation already proven, they are also in a sense measuring g, directionally if not numerically.
All IQ tests measure g, but some measure g better than others. If nuisance factors such as lack of motivation, unclear test conditions, difficulty to holding a pen, stressed by the instructor or instructions (such as stereotype effects), etc, these factors would account for a non-trivial portion of the IQ variance/gap. This would act to diminish the variance that would be otherwise attributed to g.
"If the average IQ in Sub-Saharan Africa is 80, this would mean the massive difference in environment between Sub-Saharan Africa and the US reduces IQ by only 5 points, yet the comparatively small difference in environment between black and white Americans somehow reduces it by 15 points. "
The first and second statements of that sentence are not inconsistent with each other. One can believe that the 15 point IQ difference between black and white Americans is genetic in origin, while believing that the environment has caused only a 5 point IQ difference between black Americans and Sub-Saharan Africans. An inconsistency exists only if one believes that all population differences are environmental in origin. Where is the inconsistency for those of us who think differently?
I would also question that a "massive" difference in environment exists between black Americans and Sub-Saharan Africans. What are you talking about here? Nutrition? Some African countries suffer from malnutrition, but others don't. It's also debatable whether malnutrition has a substantial effect on IQ. The 1944-45 famine in Holland had no measurable effect on the IQ of Dutch people born during that time.
In addition, mean IQ probably differs among Sub-Saharan populations, just as it does among European populations. Cognitive ability is a product of selection by the cultural environment. Some African environments have imposed higher cognitive demands than others, so mean IQ should vary accordingly.
The argument of this post is summed up in the title "Are Richard Lynn's national IQ estimates flawed?" One can believe that Lynn's estimates are flawed, especially those for Sub-Saharan Africa, without being a "pure environmentalist."
Do you consider Heiner Rindermann to be a pure environmentalist? In a review article, he argued against the accuracy of Lynn's African data, saying that "the best guess for an African average is IQ 75."
His conclusion is worth quoting:
"The described IQ means vary between IQ 68 and 78. Averaging the given means for 2010 results in an estimated IQ of around 75 for African majority countries. This result is based on many assumptions and corrections, such as properly given tests and understood test instructions, sample representativity and school enrollment rates and their corrections, mean IQ growth per year at school and per age year, and a higher African FLynn-effect. Researchers diverge in selection and correction criteria. Given the quality of the data, it is not possible to come to a really precise result."
Rindermann, H. (2013). African cognitive ability: Research, results, divergences and recommendations. Personality and Individual Differences, 55(3), 229-233. https://doi.org/10.1016/j.paid.2012.06.022
Rindermann's criticisms were not directed at a specific estimate. In his 2013 article, he cited a range of possible IQ estimates for Sub-Saharan Africa, varying from 68 to 78. There is no reason to accept an estimate of 70 as more credible than one of 75.
Could you please provide a reference to this 2018 study? (I don't see any likely ones for that year in Google Scholar). I doubt very much that he changed his opinion on this subject between 2013 and 2018.
That's a secondary reference. If we go to the primary reference, we read the same caveats that Heiner Rindermann made in his 2013 article:
"The African data base is quite unsatisfactory: Samples are rather small and
frequently not representative, there are only very few internationally comparable SAS. Therefore data for African countries were added from three studies ... The average of the three African studies SACMEQ, MLA and PASEC is given for 29 countries. These are all student assessment studies, but the usually information given by SAS publications, especially on age, participation and school attendancy rates, is missing. The representativity of data is unclear; an international norm was not given. Therefore, we did not include them in our international SAS measure. However, conventional SAS are not given for the majority of African countries and psychometric intelligence test samples are not always convincing. Thus we used this data set for all African countries and added them to our cognitive ability grand mean"
Nothing in this document indicates that he changed his mind after writing his 2013 article.
Why don't you simply say that mean SSA IQ falls somewhere between 68 and 78? Why do you want to create an impression of certainty that is not justified by the data?
Did you actually read the article, or my own reply for that matter?
The excerpt you quoted is immediately surrounded by two other sentences:
"Indeed, it really doesn’t make sense for [environmentalists] to argue that the average IQ in Sub-Saharan Africa is as high as 80... ...This point has been made a number of times, including by Lynn himself, but environmentalists still don’t seem to get it."
Noah is going out of his way to point out how the pure-environmentalist perspective is incoherent on it's face. Why are you sperging out over an article that agrees with your own position?
I did read the article, and I'm not "sperging." It's important to point out that disagreement with Lynn's estimates is not confined to pure environmentalists. I gave the example of Heiner Rindermann, and I could give others (including myself). That isn't a trivial point.
If there is one thing that draws curious onlookers to "our side", it's that we act like academics who are willing to debate the facts without getting angry, without trying to impose a uniform position, and without trying to throw the other side in jail.
> It's important to point out that disagreement with Lynn's estimates is not confined to pure environmentalists.
The post never said otherwise. That part of the post was specifically addressing a problem for the critics who happen to be environmentalists.
> If there is one thing that draws curious onlookers to "our side", it's that we act like academics who are willing to debate the facts without getting angry, without trying to impose a uniform position, and without trying to throw the other side in jail.
The only impositions I see are in your comments. The article never says that every critic of Lynn's figures are environmentalists.
The article implies that Jelte Wicherts attributes the black-white IQ difference to environmental causes. Admittedly, the article doesn’t actually say that, but it wouldn’t make much sense otherwise.
I’m familiar with Wicherts’ work. He is interested in genetic influences on intelligence and has argued against the environmentalist position, notably the theory of “stereotype threat.” In one paper, he concluded that the literature on stereotype threat might be seriously distorted by publication bias. (https://doi.org/10.1016/j.jsp.2014.10.002).
In another paper, he expressed skepticism about the Flynn effect:
“The Flynn effect raises many questions: How can IQ be substantially heritable, yet show such strong gains that appear to be due to environmental factors? Were Dutch males in 1982 so much smarter than Dutch males in 1952?”
Yes, he has also criticized the hereditarian position. It doesn’t follow, however, that he rejects the possibility of group differences in intelligence:
“This does not mean, however, that the group differences were not genetic in origin. Rather, it means that Rushton et al.'s analyses shed no light on the question of the origins of the group differences.
… at present, the methods available to address the question of the extent of involvement of genetic determinism in group differences in intelligence are not sufficient to resolve it. This is because we do not have sufficient understanding of how genes are involved in intelligence to interpret the heritability statistics we obtain, regardless of the groups within which we obtain them. https://doi.org/10.1098/rspb.2009.0238
This is not the position of an activist or an ideologue. It's the position of an honest academic. You characterize him as a “pure environmentalist” when he is nothing of the sort.
I followed the link; I found the content on that page troubling.
But woke indoctrination of that sort is not necessarily confined to universities nor even community colleges. My first psychology course was in high school.
Frankly the most interesting Black national IQ data wouldn't be from SSA, but from the West Indies (where data quality would likely be much higher). Has anyone remarked on these?
My question is always to be the same, how can individuals coming from a population with an IQ below 80 manage to live in Western society? And their offspring to partake in the western educational system? Aren’t individuals with a IQ below 80 not considered “legally” retarded?
Apologies for referencing something practical , but this Lynn nuff nuff wouldn’t last 5 minutes in sub saharan africa as his “IQ” and utter uselessness would mean he would die in a very short time. I suppose he could eat his thesis, not much nutrition in that though (physical or mental)
It is ludicrous to believe that the Japanese have an average IQ of 105, most of the population is very old and primitive, we are talking feudal people who are very short due to deficits in nutrition growing up, imagine some Babushka from 300 years ago in a village of Russia, that's the level.
I came to Japan in part thinking they were gonna be intelligent, and after 2 years here and having learned the language and met hundreds of them I can tell you they are very unintelligent, they are extremely superficial, gossipy and never ever do they speak about anything intellectual, they have no interest in sciences or engineering or religion, they just play their social role in the feudal primitive system that they have, then use the money for vices. They can't speak languages, they have never visited foreign countries, this people are medieval.
Lynn's estimates must be very unscientific, my guess is he assumed sophistication because they are obedient cowardly slaves as opposed to problematic warring Africans.
There is also the possibility of foul play by the Asian groups performing the testing, in the cases where there is any testing at all. Their cultures value saving face and pretending to be better than they actually are above all else, plus their societies and languages are some of the most impenetrable in the world, and they lack freedom, so having foreigners do the testing seems unlikely.
Sure , no fake it till you make it types faking their IQ in advanced western countries they are all to busy trying to figure out how many sexes there are. For each year of exposure to woke, their IQ drops by 5 points according to a recent study by the University of the Inconsequential’s Department of Rat Studies based in Brussels
Noah Carl and his like accept IQ estimates derived from academic achievement when they confirm the racial IQ gap, but not when they disconfirm it. Eric Turkheimer collects a lot of academic achievement evidence that the racial IQ gap is diminishing, but no “hereditarian” that I’ve come across has ever been willing to accept Turkheimer’s argument. https://www.vox.com/the-big-idea/2017/6/15/15797120/race-black-white-iq-response-critics
It's hard to believe that rural Chinese are all that bright . . . they don't act intelligently. They are being dragged out of the Middle Ages by government programs and formal education. I also don't think people living four hours off the grid are being assessed on IQ scores. If the results indicate an IQ of 105 as the national average, it's probably safe to knock 5 or 10 points off it.
This article has an excellent point: that you can take the results for American ethnic groups and shave some points off for very poor countries of origin (in terms of ancestry). A huge portion of Chinese live in isolated rural villages with bad nutrition (esp. for anyone over 25), terrible educational opportunities and very little intellectual life. That ABC's (American-Born Chinese) would have IQ's of 105, on average, and China itself would be just as high, seems highly implausible.
Noah, why was Irish IQ once thought to be 87, and now it is effectively the same as the UK? See Keith Woods substack on this. Could purely environmental factors or unfamiliarity with testing have led to such a low result in earlier tests?
Great article, Noah. Thanks for the analysis.
"We already have abundant evidence that black Americans score about 85 on IQ tests, as compared to 100 for whites. If the average IQ in Sub-Saharan Africa is 80, this would mean the massive difference in environment between Sub-Saharan Africa and the US reduces IQ by only 5 points, yet the comparatively small difference in environment between black and white Americans somehow reduces it by 15 points."
However, the comparison of IQ between 'black Americans' and Sub-Saharan Africans only approaches validity if the admixture of black to white in the 'black American' group is known. Many 'black Americans' carry varying degrees of white genetics. This is more explanatory of IQ difference than the environment.
I thought it's well known that Black Americans average 15-20% white. Anyway, this estimate will be rough, as we don't know if there was selection in who was sold into slavery, and what selection you have had over two centuries. Sub-Saharan Africa is a big place. There's going to be a lot of variation between groups.
> Many 'black Americans' carry varying degrees of white genetics. This is more explanatory of IQ difference than the environment
Possibly, but from the strict-environmentalist PoV this shouldn't matter either way, so the critique still stands.
The strict environmentalist PoV is obviously wrong.
The PoV that national IQ numbers are perfect representations of genetic potential is obviously wrong.
Most things in life are caused by a combination of genetics and environment. That's a general rule. Of course some things are all genetics or all environment. But generally most things are a mix.
Most IQ gaps are much less well studied and much less well understood than the BW one in the US. That is the one we have studied a lot and where we have tried to correct the environment a lot. And understandably, because of all of the awful things that happened in the US like slavery, lynching, redlining, Jim Crow, and so on people are sensitive about it.
We know that one is mostly on g from adoption studies, studies on Spearman's hypothesis, and so on. This one has been studied to death. Of course this one is >50% genetic g differences. Whether it's 80% or 100% I don't know, you have the B are comparatively better at memory for instance so it will depend on your g-loading.
Probably preaching to the choir here.
I don't think we're disagreeing to any great extent here. No-one, including Lynn, disputes that SSA's IQ is lowered by various adverse environmental factors (in his estimate, by about 13 IQ points.) I'm just a little skeptical that the environment in somewhere like Saudi Arabia is adverse enough to create a comparable gap.
Yeah I'm not so familiar with the Saudi IQ figures, but my guess is their genotypic IQ is about 90. Their PISA score is like 80-85, but they have inbreeding and a bad education system so I would guess the genetic IQ is about 90.
Too many in research today suffer from dead white male envy. Produce quality research if you want respect. Secondly, during research on mulattos in America, I discovered that the discrimination they encountered was understated and possibly there could be a link between skin colour and IQ. Lynn, did a study that was quite good, but I think we need more admixture studies. The MQ studies are okay, but I want bigger samples with people from elite mulatto families.
James Flynn once proposed to gather accurate data on Sub-Saharan IQ. Nobody wanted to be associated with such a study. Such a study is simply not possible today. It's just too inflammatory. I think tests like the PISA are the only measure that we will have for the foreseeable future.
Nice that someone bothered to come to Richard Lynn's defense now that he's unable to defend himself.
A couple of years ago I read his memoir/autobiography. I thought he seemed a nice, intelligent, courageous, down-to-earth bloke.
The critiques of Lynn's methodology aren't entirely off-base, given he was essentially scraping data in a somewhat ad-hoc way from whatever studies were conducted in a given country, often for rather different purposes (one of my favourite examples is a study on the prevalance of Konzo where he took the unaffected control group as indicative of typical IQ in the region.) He was working off a shoestring budget, so I don't think he had much choice in most cases.
In any case, while you can poke holes in specific studies the overall trend in the numbers is just too overwhelming to ignore. (His critics, of course, have no interest in sending out an army of researchers armed with raven's matrices to go get higher-quality data on SSA's IQ. They know perfectly well what they would find.)
From what you say I can well imagine that scrupulous scientists in the herediarian camp and any scientist, scrupulous or not, from the environmentalist camp could pick holes in Lynn's data. But since I'm neither a scientist nor scrupulous I'm happy enough to give him a pass if things look more or less right. I suspect no amount of evidence would be good enough for the environmentalists.
Environmentalism is a straw man. But I find way too many of Lynn's numbers to be too low to be plausible. Iran 85 or so? Really? This does not make sense given the number of top scientists from Iran. There are twice as many Iranians as Black Americans, but many more than twice as many Iranian Fields Medalists, Nobel Prize winners, top scientists, and so on.
Israel at 92? Cremieux had a post where he showed that the scores go up when you correct for psychometric bias and isolate the g-factor. It's more like 97 or 98, with 102-103 for Jews, majority of whom are non-Ashkenazi, whereas Lynn estimated 103 for Israeli Ashkenazi Jews based on some implausible hypothesis about selective migration.
Romania at 90 or so? Really low numbers for the Balkan countries? Albania and North Macedonia down into the 80s? These numbers are just not plausible. And it's noteworthy that post-Communist countries seem to have depressed IQs, you can look at West Germany versus East Germany.
The OECD has documented that people in some countries put in less effort on these tests and do better on the first half of the test whatever that is. I found this from the PISA website:
"In reading, on average across OECD countries, students who took the test in the second hour (in most cases, after completing an hour-long mathematics test) scored 14 points lower than students who took the test in the first hour – a large difference. Large performance declines during the test of between 20 and 30 score points were observed in Iceland, Israel, Latvia*, Albania, Qatar, Slovenia, Malta, Argentina and Norway (in descending order of the size of this difference) (Table I.A8.17)."
So I suspect that many of these national IQ numbers are also capturing something about trust rather than intelligence. In more than one Eastern European country and in Israel people will just try less hard on a test that counts for nothing.
Many of the numbers are undoubtedly somewhat inaccurate. However, this is true of all datasets. According to Gust and colleagues' dataset, Niger's student achievement is equivalent to an IQ of 56.
–NC
Mind you, but these IQ estimates by Lynn were never intended to reflect latent g. They don't assess for possible measurement bias using MGCFA and IRT or track low effort response or rapid guessing behavior using IRT. My point is that, if we follow your criticism, nearly all IQ papers should be deemed rubbish because they never tried to assess latent g, but instead, merely observed total IQ. Instead, a major concern should be about sample representativeness and test condition.
I agree and I certainly don't think that "all IQ papers should be deemed rubbish" or retracted. All I'm saying is their results should not always be taken at face value, and people should understand that Lynn national IQ is certainly an imperfect measure of national genetic g. Heck even Raven has a g-loading of like 0.5. Lynn agree with this, though he was wrong about some things. You agree with this. Aporia itself agrees with this and likes my comment. I'm not arguing with you. I'm arguing with Simon Laird.
Perhaps some context is in order. A few days ago I was arguing with Simon Laird on Substack that Iran is poor largely because a terrible government theocracy that is socialist and anti-American, and his response was that the reason that the only reason it's poor is that it has a low IQ and there are sanctions. And I said, you can't take the Lynn figure at face value. And I got a "the Lynn figure is just as likely to be an overestimate as an underestimate". I think he quoted a figure of about *80* to me. I can pull up the conversation. (edit: I found it, see here https://substack.com/@usuallywash/note/c-83086866?utm_source=activity_item ad here https://substack.com/profile/16148027-simon-laird/note/c-83086490 )
Simon Laird has actually written an article on Aporia https://www.aporiamagazine.com/p/africa-is-not-doomed-to-poverty about the quality of governance in developed countries and how it matters! In fact I agree with this article of his, and I think it's hilarious that he wrote this, in view of his comments about Iran that I linked to in the above paragraph.
Of course if woke people are trying to throw away all Lynn research because Lynn was imperfect and wrong about some things (whether because of sloppiness or because of some political motivation) then that is bad. He is definitely not the first person to be wrong and he did important work in the area! But all too often I do see people taking Lynn figures at face value in ridiculous ways, for instance using them to defend Iran's form of government and blaming the terrible state Iran is in on the supposedly 80-ish IQ of the Iranian people. (!).
You might as well defend communism by pointing out that East Germans did worse than West Germans on IQ tests and that therefore this must be genetic and must have nothing to do with communism.
Maybe you noticed but my comment was only directed at your second paragraph. It's the observation that the estimate can't be trusted because it doesn't account for measurement bias and latent g (which in this case means this criticism can be directed at nearly all IQ papers) that really throws me off.
In general I don't think I disagree with your other points. If you single out a few estimates, and this has been known for decades now, you might see some weird estimates. Some old estimates that you could tell were odd such as the Irish IQ at 87.
Ah OK, so when cremieux corrected Israeli IQ for "psychometric bias" (not even latent g, just bias internal to the test, whether caused by low effort, bad education, whatever) and it went up 5 points, you're saying we still report the lower number as the IQ, since that's what IQ means. Got it. IQ can be influenced by environment and various other things. Yeah, it's true and well-known that Lynn underestimated Irish IQ, probably because was a Northern Irish Protestant and had an axe to grind against the Irish.
I think that "can't be trusted" should be unpacked a little bit. Of course if IQ is defined by scores on various tests then IQ scores can be trusted by definition, if your goal is to accurately estimate the IQ. Insofar as IQ is predictive of various things though, I thought it's generally the g-factor, which is why Cremieux is always adjusting for psychometric bias. I'm sure you would know better than me. So of course you are right that these scores can be "trusted", but for most useful intents and purposes one does not want to take them at face value.
Ideally, all IQ reports should reflect latent g, if possible based on bifactor CFA model. And low-effort response, anxiety, rapid guessing behavior should be also accounted for as well. But obviously, it's easier said than done. It's so much more convenient to just collect FSIQ data, especially for meta-analysis. If you meta-analyze IQ estimates based on latent g, your data is reduced to maybe 1% at best. If you require studies that account for all other external factors, you get no data. This is why you might want to use national assessments as well, as a robustness check. The interesting outcome is that some of the estimates regarded as too low, such as African IQs, seem to match closely the estimates from national assessments. As opposed to IQ tests, there have been a few studies that have tested for possible national differences in rapid guessing behavior and other external factors using national assessments.
For many international comparisons we typically don't have any recent FSIQ data right? PISA is readily available but it's not FSIQ.
Do you know if there's a non-paywalled version of this article on East/West german IQ convergance? I don't want to dismiss any potential causal confound offhand.
https://www.sciencedirect.com/science/article/abs/pii/S0160289612000864
I will say that while I'm a little puzzled by the enormous score disparity between the eurozone and much of the middle east, Saudi Arabia's national IQ does seem to be pretty consistent with their abysmal PISA scores. And on a per-capita-purchasing-parity basis they're spending roughly as much on education as Germany does.
https://gpseducation.oecd.org/CountryProfile?plotter=h5&primaryCountry=SAU&treshold=5&topic=PI
https://data.worldbank.org/indicator/NY.GDP.PCAP.PP.CD?locations=DE-SA
https://tradingeconomics.com/saudi-arabia/public-spending-on-education-total-percent-of-gdp-wb-data.html
https://tradingeconomics.com/germany/public-spending-on-education-total-percent-of-gdp-wb-data.html
Are the Saudis just lazy when it comes to taking tests, as you suggested was true for other countries? "Large performance declines during the test of between 20 and 30 score points" would only correspond to an IQ gap of 3-4 points or so. I've heard that a lot of Saudi citizens can fall back on government make-work positions or get economically-irrelevant degrees in theology, so maybe that has some effect on motivation, but is it really going to be ~20 IQ points worth of apathy?
Inbreeding depresses iq as well
Yes, I've considered that, but outside of a few isolated pockets of the muslim world with terrifying levels of consanguinuity it's not really common enough to account for more than a few IQ points.
Well Crem found a 5 point adjustment in Israeli IQ. If someone does 4 points worse on the second half than the first half, their genetic potential is probably even better than how they did on the first half, since they were somewhat lazy on the first half too. PISA also definitely doesn’t have a g loading of 1.0. Even Ravens matrices is 0.5 according to a paper Emil linked. Educational quality matters too.
The Saudi score on the PISA reading corresponds to an IQ somewhere in the 80-85 range. If you believe a few points are off due to low effort and/or bad education system and a few points are off due to inbreeding you get a genotypic IQ of 90 which makes sense.
I suspect PISA math is less g-loaded than reading. If you have a bad education system you won’t do well even if you have high g.
Inbreeding is a disadvantage for a population with genetic defects it needs to dilute, and it is an advantage for a population that needs to avoid importing genetic defects. Incest rules vary and are not based on genetic considerations. Outbreeding customs in settlements, clans, extended families, etc are based on the need for people collected in small groups to get along; and making insiders sexually off-limits depresses effects of jealousy, envy, and rejection resentment.
Isn’t it the other way around? Latent g has been shown many times to correlate fairly strongly with IQ. Because g is hard to measure, they instead tend to measure IQ, assuming that with the correlation already proven, they are also in a sense measuring g, directionally if not numerically.
All IQ tests measure g, but some measure g better than others. If nuisance factors such as lack of motivation, unclear test conditions, difficulty to holding a pen, stressed by the instructor or instructions (such as stereotype effects), etc, these factors would account for a non-trivial portion of the IQ variance/gap. This would act to diminish the variance that would be otherwise attributed to g.
"If the average IQ in Sub-Saharan Africa is 80, this would mean the massive difference in environment between Sub-Saharan Africa and the US reduces IQ by only 5 points, yet the comparatively small difference in environment between black and white Americans somehow reduces it by 15 points. "
The first and second statements of that sentence are not inconsistent with each other. One can believe that the 15 point IQ difference between black and white Americans is genetic in origin, while believing that the environment has caused only a 5 point IQ difference between black Americans and Sub-Saharan Africans. An inconsistency exists only if one believes that all population differences are environmental in origin. Where is the inconsistency for those of us who think differently?
I would also question that a "massive" difference in environment exists between black Americans and Sub-Saharan Africans. What are you talking about here? Nutrition? Some African countries suffer from malnutrition, but others don't. It's also debatable whether malnutrition has a substantial effect on IQ. The 1944-45 famine in Holland had no measurable effect on the IQ of Dutch people born during that time.
In addition, mean IQ probably differs among Sub-Saharan populations, just as it does among European populations. Cognitive ability is a product of selection by the cultural environment. Some African environments have imposed higher cognitive demands than others, so mean IQ should vary accordingly.
> An inconsistency exists only if one believes that all population differences are environmental in origin
Yes, but the specific point of the comment is to underline how the pure-environmentalist position here is implausible.
Yes. But I also thought Noah was pointing out that environmentalists are making the hereditarian argument, though they appear not to realise it.
Yes!
The argument of this post is summed up in the title "Are Richard Lynn's national IQ estimates flawed?" One can believe that Lynn's estimates are flawed, especially those for Sub-Saharan Africa, without being a "pure environmentalist."
Do you consider Heiner Rindermann to be a pure environmentalist? In a review article, he argued against the accuracy of Lynn's African data, saying that "the best guess for an African average is IQ 75."
His conclusion is worth quoting:
"The described IQ means vary between IQ 68 and 78. Averaging the given means for 2010 results in an estimated IQ of around 75 for African majority countries. This result is based on many assumptions and corrections, such as properly given tests and understood test instructions, sample representativity and school enrollment rates and their corrections, mean IQ growth per year at school and per age year, and a higher African FLynn-effect. Researchers diverge in selection and correction criteria. Given the quality of the data, it is not possible to come to a really precise result."
Rindermann, H. (2013). African cognitive ability: Research, results, divergences and recommendations. Personality and Individual Differences, 55(3), 229-233. https://doi.org/10.1016/j.paid.2012.06.022
The average IQ of the Sub-Saharan African countries in Rindermann's 2018 dataset is 70.
—NC
Rindermann's criticisms were not directed at a specific estimate. In his 2013 article, he cited a range of possible IQ estimates for Sub-Saharan Africa, varying from 68 to 78. There is no reason to accept an estimate of 70 as more credible than one of 75.
Could you please provide a reference to this 2018 study? (I don't see any likely ones for that year in Google Scholar). I doubt very much that he changed his opinion on this subject between 2013 and 2018.
See Table 2 here: https://gwern.net/doc/iq/2022-warne.pdf
—NC
That's a secondary reference. If we go to the primary reference, we read the same caveats that Heiner Rindermann made in his 2013 article:
"The African data base is quite unsatisfactory: Samples are rather small and
frequently not representative, there are only very few internationally comparable SAS. Therefore data for African countries were added from three studies ... The average of the three African studies SACMEQ, MLA and PASEC is given for 29 countries. These are all student assessment studies, but the usually information given by SAS publications, especially on age, participation and school attendancy rates, is missing. The representativity of data is unclear; an international norm was not given. Therefore, we did not include them in our international SAS measure. However, conventional SAS are not given for the majority of African countries and psychometric intelligence test samples are not always convincing. Thus we used this data set for all African countries and added them to our cognitive ability grand mean"
https://www.tu-chemnitz.de/hsw/psychologie/professuren/entwpsy/team/rindermann/pdfs/RindermannCogCapAppendix.pdf
Nothing in this document indicates that he changed his mind after writing his 2013 article.
Why don't you simply say that mean SSA IQ falls somewhere between 68 and 78? Why do you want to create an impression of certainty that is not justified by the data?
Did you actually read the article, or my own reply for that matter?
The excerpt you quoted is immediately surrounded by two other sentences:
"Indeed, it really doesn’t make sense for [environmentalists] to argue that the average IQ in Sub-Saharan Africa is as high as 80... ...This point has been made a number of times, including by Lynn himself, but environmentalists still don’t seem to get it."
Noah is going out of his way to point out how the pure-environmentalist perspective is incoherent on it's face. Why are you sperging out over an article that agrees with your own position?
I did read the article, and I'm not "sperging." It's important to point out that disagreement with Lynn's estimates is not confined to pure environmentalists. I gave the example of Heiner Rindermann, and I could give others (including myself). That isn't a trivial point.
If there is one thing that draws curious onlookers to "our side", it's that we act like academics who are willing to debate the facts without getting angry, without trying to impose a uniform position, and without trying to throw the other side in jail.
> It's important to point out that disagreement with Lynn's estimates is not confined to pure environmentalists.
The post never said otherwise. That part of the post was specifically addressing a problem for the critics who happen to be environmentalists.
> If there is one thing that draws curious onlookers to "our side", it's that we act like academics who are willing to debate the facts without getting angry, without trying to impose a uniform position, and without trying to throw the other side in jail.
The only impositions I see are in your comments. The article never says that every critic of Lynn's figures are environmentalists.
The article implies that Jelte Wicherts attributes the black-white IQ difference to environmental causes. Admittedly, the article doesn’t actually say that, but it wouldn’t make much sense otherwise.
I’m familiar with Wicherts’ work. He is interested in genetic influences on intelligence and has argued against the environmentalist position, notably the theory of “stereotype threat.” In one paper, he concluded that the literature on stereotype threat might be seriously distorted by publication bias. (https://doi.org/10.1016/j.jsp.2014.10.002).
In another paper, he expressed skepticism about the Flynn effect:
“The Flynn effect raises many questions: How can IQ be substantially heritable, yet show such strong gains that appear to be due to environmental factors? Were Dutch males in 1982 so much smarter than Dutch males in 1952?”
Yes, he has also criticized the hereditarian position. It doesn’t follow, however, that he rejects the possibility of group differences in intelligence:
“This does not mean, however, that the group differences were not genetic in origin. Rather, it means that Rushton et al.'s analyses shed no light on the question of the origins of the group differences.
… at present, the methods available to address the question of the extent of involvement of genetic determinism in group differences in intelligence are not sufficient to resolve it. This is because we do not have sufficient understanding of how genes are involved in intelligence to interpret the heritability statistics we obtain, regardless of the groups within which we obtain them. https://doi.org/10.1098/rspb.2009.0238
This is not the position of an activist or an ideologue. It's the position of an honest academic. You characterize him as a “pure environmentalist” when he is nothing of the sort.
The Reddit comments on a related post exemplify the worst of woke academia. This is what people are learning at university!
https://www.reddit.com/r/sociology/comments/1b14xm1/regarding_average_iq_for_a_country_data_from_iq/
I followed the link; I found the content on that page troubling.
But woke indoctrination of that sort is not necessarily confined to universities nor even community colleges. My first psychology course was in high school.
Frankly the most interesting Black national IQ data wouldn't be from SSA, but from the West Indies (where data quality would likely be much higher). Has anyone remarked on these?
These have been done too. https://www.researchgate.net/publication/298214364_Admixture_in_the_Americas_Regional_and_National_Differences
Wouldn't the West Indies have too high an admixture of white genes to confound the data?
My question is always to be the same, how can individuals coming from a population with an IQ below 80 manage to live in Western society? And their offspring to partake in the western educational system? Aren’t individuals with a IQ below 80 not considered “legally” retarded?
Apologies for referencing something practical , but this Lynn nuff nuff wouldn’t last 5 minutes in sub saharan africa as his “IQ” and utter uselessness would mean he would die in a very short time. I suppose he could eat his thesis, not much nutrition in that though (physical or mental)
It is ludicrous to believe that the Japanese have an average IQ of 105, most of the population is very old and primitive, we are talking feudal people who are very short due to deficits in nutrition growing up, imagine some Babushka from 300 years ago in a village of Russia, that's the level.
I came to Japan in part thinking they were gonna be intelligent, and after 2 years here and having learned the language and met hundreds of them I can tell you they are very unintelligent, they are extremely superficial, gossipy and never ever do they speak about anything intellectual, they have no interest in sciences or engineering or religion, they just play their social role in the feudal primitive system that they have, then use the money for vices. They can't speak languages, they have never visited foreign countries, this people are medieval.
Lynn's estimates must be very unscientific, my guess is he assumed sophistication because they are obedient cowardly slaves as opposed to problematic warring Africans.
There is also the possibility of foul play by the Asian groups performing the testing, in the cases where there is any testing at all. Their cultures value saving face and pretending to be better than they actually are above all else, plus their societies and languages are some of the most impenetrable in the world, and they lack freedom, so having foreigners do the testing seems unlikely.
Sure , no fake it till you make it types faking their IQ in advanced western countries they are all to busy trying to figure out how many sexes there are. For each year of exposure to woke, their IQ drops by 5 points according to a recent study by the University of the Inconsequential’s Department of Rat Studies based in Brussels
What are the two outlier blue countries with high educational achievement but IQ in the mid 70s and low 80s?
Noah Carl and his like accept IQ estimates derived from academic achievement when they confirm the racial IQ gap, but not when they disconfirm it. Eric Turkheimer collects a lot of academic achievement evidence that the racial IQ gap is diminishing, but no “hereditarian” that I’ve come across has ever been willing to accept Turkheimer’s argument. https://www.vox.com/the-big-idea/2017/6/15/15797120/race-black-white-iq-response-critics
It's hard to believe that rural Chinese are all that bright . . . they don't act intelligently. They are being dragged out of the Middle Ages by government programs and formal education. I also don't think people living four hours off the grid are being assessed on IQ scores. If the results indicate an IQ of 105 as the national average, it's probably safe to knock 5 or 10 points off it.
This article has an excellent point: that you can take the results for American ethnic groups and shave some points off for very poor countries of origin (in terms of ancestry). A huge portion of Chinese live in isolated rural villages with bad nutrition (esp. for anyone over 25), terrible educational opportunities and very little intellectual life. That ABC's (American-Born Chinese) would have IQ's of 105, on average, and China itself would be just as high, seems highly implausible.
Noah, why was Irish IQ once thought to be 87, and now it is effectively the same as the UK? See Keith Woods substack on this. Could purely environmental factors or unfamiliarity with testing have led to such a low result in earlier tests?
Russ Warne wrote a detailed post on this: https://russellwarne.com/2022/12/17/irish-iq-the-massive-rise-that-never-happened/
–NC
I just read the Warne article. What a clear, well-argued piece of writing! I might have to buy his (expensive) book.
I'll check it out. Thank you.