Monday, September 5, 2022

How real is the Flynn effect? Part II


Can you pass a Grade 8 exam that was given to American children in 1912? Why the Flynn effect never really happened.



My last post elicited a tweet from Elon Bachman:


Argument by incredulity (w.r.t. retardation) bedevils discussions of intelligence data. What's striking about most retards is the *congenital* speech & motor impairment, not the level of g. Height also increased 1 SD last century. There used to be a lot of malnourished peasants. @ElonBachman


In a population with an average IQ of 100, an IQ below 70 is usually due to abnormal circumstances, either freak accidents during pregnancy or rare mutations that haven’t been eliminated by natural selection. In either case, the adverse effects are wide-ranging and affect many aspects of mind and behavior. If your IQ is below 70, you’re lacking in much more than cognitive ability. You probably suffer from various speech and motor impairments. In short, you look and act “funny.”


A century ago, if we use current norming of IQ tests, average IQ was below 70 in the Western world. Yet it was not associated with looking and acting “funny.” How come?  For Elon Bachman, this is because an IQ of 70 was normal back then. It was not an outcome of freak accidents or rare mutations. The mental deficit was therefore confined to cognitive ability alone.


I understand that argument. Nonetheless, the increase in IQ between then and now is still 35 points—more than two standard deviations. Such a difference in cognitive ability should be visible in popular culture, particularly the books and newspapers that people read. It should also be visible in the academic requirements of elementary school, which was compulsory by that time in most North American jurisdictions.


The following are some questions from a Grade 8 exam that was given in 1912 to children in Bullitt County, Kentucky:


Find cost at 12½ cents per sq. yd. of kalsomining the walls of a room 20 ft. long, 16 ft. wide and 9 ft. high, deducting 1 door 8 ft. by 4 ft. 6 in. and 2 windows 5 ft. by 3 ft. 6 in. each.


A man sold a watch for $180 and lost 16 2/3%. What was the cost of the watch?


How many parts of speech are there? Define each.


What is a personal pronoun? Decline I.


Define longitude and latitude.


Tell what you know of the Gulf Stream.


Through what waters would a vessel pass in going from England through the Suez Canal to Manila?


Compare arteries and veins as to function. Where is the blood carried to be purified?


Define Cerebrum; Cerebellum.


Define the following forms of government: Democracy, Limited Monarchy, Absolute Monarchy, Republic. Give examples of each.


Name five county officers and the principal duties of each.


What is a copyright? Patent right?


Describe the Battle of Quebec.


Name the last battle of the Civil War; War of 1812; French and Indian War and the commanders in each battle.


Those are only a few of the questions. The full exam is here.


I understand why some want to believe that Western societies were much less intelligent a century ago. If we cannot use IQ to compare different populations over time, perhaps we cannot use it to compare different populations over space, i.e., in different parts of the world. And then what?


I’ve heard that sort of argument before. “If we accept genetic determinism for this, we may have to accept it for that. And then what?” If that’s your frame of mind, I have nothing more to say.


The increase in IQ and the increase in height during the 20th century


But what about the increase in height? That was real. If insufficient nutrition held back height during the 20th century, it should have also held back cognitive ability. The assumption here is that IQ and height are equally constrained by nutrition during development. Has that assumption been tested?


It has. The Dutch suffered a terrible famine during the winter of 1944-1945, as did much of continental Europe. Yet that famine had no discernable impact on the mental development of unborn children from that time:


A detailed retrospective study was made of 125,000 19-yr-old male Dutch military inductees of whom 20,000 were exposed to the Netherland's winter famine, 1944-1945, through maternal starvation. […] Findings show that (a) starvation during pregnancy had no detectable effects on adult mental performance of surviving male offspring. (b) Mental performance of surviving adult males from the entire population had no association with changing levels of mean birth weight in a hospital sample from the population. And (c) there was a strong association of social class with mental performance. (Stein et al. 1972)


It isn’t as if the Flynn effect didn’t occur in the Netherlands, As James Flynn points out:


As for nutrition, to my knowledge no one has actually shown that American or British or Western European children have a better diet today than they did in 1950, indeed, the critics of junk food argue that diets are worse. Yet, post-1950 IQ gains have been very large. Military samples tested in 1952, 1962, 1972, and 1982 show that Dutch males made a 20-point gain on a Raven’s-type test (Flynn, 1987, p. 172). Even the latest period shows a huge gain, that is, the Dutch 18-year olds tested in 1982 outscored the Dutch 18-year olds tested in 1972 by fully 8 IQ points. Did the quality of the Dutch diet really escalate that much in 10 years? The gains posted by the 1962 males over the 1952 males are interesting. (Flynn 2008)


Those 1962 males were born or still in the womb during the 1944-45 famine. Yet they “do not show up even as a blip in the pattern of Dutch IQ gains” (Flynn 2008). One could argue that the famine was too brief to affect cognitive ability. Perhaps malnutrition needs a longer time span to affect mental development. And again, what about the increase in height? What was driving that increase?


We don’t know for sure. We do know that height increased independently of the increase in IQ.


Norway was cited above as a nation in which the nutrition hypothesis is viable thanks to greater gains in the lower half of the IQ distribution. Actually, it counts against the posited connection between height gains and IQ gains. Height gains have been larger in the upper half of the height distribution than in the lower half (Sundet et al., 2004). This combination, greater height gains in the upper half of the distribution, greater IQ gains in the lower, poses a serious problem. Are there two kinds of enhanced nutrition, one of which raises height more than it does IQ, the other of which raises IQ more than it does height? (Flynn 2008)


There is no common cause behind the increase in IQ and the increase in height. The latter seems to be due to people eating a more varied diet after the Second World War. In particular, fewer and fewer people were subsisting largely on bread. That food item is problematic not only because it lacks many nutrients but also because it has a high content of phytic acid, which binds to calcium and other essential minerals in the food we eat (Harrison and Mellanby 1939; McCance and Widdowson 1942a; McCance and Widdowson 1942b; Sandstead 1992). With the postwar boom, and with efforts to educate people about the benefits of a balanced diet, bread made up a smaller and smaller proportion of total food intake. Bones could now fully develop.


Excessive consumption of bread, and the resulting lack of essential minerals, constrained skeletal development during the early 20th century. But it did not constrain cognitive ability.




Flynn, J.R. (2008). Requiem for nutrition as the cause of IQ gains: Raven's gains in Britain 1938-2008. Economics and Human Biology 7(1):18-27.


Harrison, D.C., and E. Mellanby. (1939). Phytic acid and the rickets-producing action of cereals. Biochemical Journal 33: 1660-1680.


McCance, R.A., and E.M. Widdowson. (1942a). Mineral metabolism of healthy adults on white and brown bread dietaries. The Journal of Physiology 101: 44-85.


McCance, R.A., and E.M. Widdowson. (1942b). Mineral metabolism on dephytinized bread. The Journal of Physiology 101: 304-313.


Sandstead, H.H. (1992). Fiber, phytates, and mineral nutrition. Nutrition Reviews 50: 30-31.


Sieczkowksi, C. (2013). 1912 Eighth-Grade Exam Stumps 21st-Century Test Takers. Could You Pass This Eighth-Grade Exam from 1912? Huffpost, August 12.


Stein, Z., M. Susser, G. Saenger, and F. Marolla. (1972). Nutrition and mental performance. Science 178(4062): 708–713.


Sundet, J.M., D.G. Barlaug, and T.M. Torjussen. (2004). The end of the Flynn effect? A study of secular trends in mean intelligence test scores of Norwegian conscripts during half a century. Intelligence 32: 349– 362.






siriusactuary said...

Do we have data on the changes over time in the distribution of measured IQs? Some of the questions I wonder about are:

Has the variance decreased, especially at the lower end?
Is the standard deviation lower?
Have the highest IQs moved higher over time? I don't think we see any evidence of this, but it would be good to know what the data say.

Appreciate all of your insightful blogging on this and other topics. Fascinating!

Santocool said...

''or rare mutations that haven’t been eliminated by natural selection.''

How to eliminate something that usually happens spontaneously?

Truth Seeker said...

When you read the writings or speeches of people prior to, let's say, 1970, you're struck by how sophisticated and high-brow their vocabulary was. These days, people can't even distinguish between "who" and "whom," or "less" and "fewer," or be bothered to insert "of" in "because of" (they increasingly say "because ." But people's verbal abilities were much more developed several decades ago than now. And the farther back in time you go, the more elaborate and sophisticated their language was. To me, it seems like people's verbal IQ has definitely declined, for unclear reasons.

Santocool said...

Wait! So dutch famine epigenetic effects in subsequent generations was made up???

yeong said...

I went through the exam and the questions posed are recall-heavy, and do not require much in the way of understanding and application. The arithmetic questions require some application but if the children were exposed to similarly phrased questions, then those could be recall questions as well.

Compare to the the proficiencies expected on the PISA test, and the difference is striking.

Recall skills are not particularly useful in a world with search engines and unlimited access to data.

Anonymous said...

Those are very easy questions, though some of them ask for information I don't possess.

I was pretty bothered by the sub-question of "where is the blood carried to be purified", since different aspects of blood purification occur in different places. The liver removes some toxins. The kidneys remove other toxins. The spleen removes blood cells that have gone bad. "Purifying blood" is not a single activity performed in a centralized location.

Anonymous said...

Yes a lot of schooling during those time periods is what i assume to heavy in recall but recall was necessary and is still today the core of intelligence and practicality! if you can recall things well you become a productive member of society not only because you benefit from actually accomplishing things like building something engineering whatever but also you expose yourself to different ideas and can learn new things from people or just have a frank discussion with them on important topics like politics etc.

this is not written in jest but in truth. novel problem solving is the only thing higher in the hierarchy than recall i think and recall was a lot more important back in the day for various reasons already mentioned by other commenters so i believe that for the time period these type of tests would be effective in judging who is competent and who is not!

Anonymous said...

the pisa and what not are heavy on problem solving. this is great because we face new challenges every day in terms of our modern society.

on another note i think the Flynn Effect is only in specialized abilities as noted by scholars on the topic and not on general intelligence! so there is that.

Anonymous said...

btw in my first comment when i said "tests" i dont mean IQ tests i mean the ones they administered in school. secondly i would like to say that recall helps with facilitating problem solving! and thirdly recall i would add that recall would build peoples interests in things in general allowing room for creative growth etc. even though the phenomenon is rarely mentioned and people usually think of those two things (recall and creativity) as being orthogonal to one another!

Peter Frost said...


Variance in IQ has decreased over time. The increase in the bottom half of the distribution has been greater than in the top half:

"There is some evidence for decreasing variability of IQ task performance and a narrowing of IQ distributions in several countries, although decreasing variability has not been found in all accounts (e.g., Dickens & Flynn, 2002; Pietschnig et al., 2010). Indeed, the available evidence indicates that decreases of IQ variability may have played a role in only some, but not all, countries (for an over view, see Flynn, 2012, pp. 41–42). However, a considerable number of recent investigations reported direct evidence for decreasing IQ variability over time in Austria, Denmark, Norway, Spain, and the United States (Colom et al., 2005; Pietschnig et al., 2013; Rindermann & Thompson, 2013; Sundet, Barlaug, & Torjussen, 2004; Teasdale & Owen, 2005)"

This reduction in IQ variability can be explained in several ways. Perhaps the bottom half were malnourished people who are now better nourished and can thus better express their IQ potential. Or perhaps the bottom half were people outside the "modern paradigm" of asking questions and seeking answers. They have now become "modernized." The first explanation is plausible for Western societies before the 1950s. For me, at least, it cannot explain the IQ gains since the 1950s.

Pietschnig, J., and M. Voracek. (2015). One Century of Global IQ Gains: A Formal Meta-Analysis of the Flynn Effect (1909-2013). Perspectives on Psychological Science 10(3): 282-306.


Mutations appear spontaneously but they don't disappear spontaneously.

Truth Seeker,

Yes, when I read popular magazines from a century ago, I notice a difference not only in vocabulary but also in sentence length and structure. Sentences were longer and had more subordinate clauses.


Please educate me by providing a reference.


There has been a cultural shift over the past century. In the past, recall was all important. Students were expected to memorize huge quantities of information: historical data, spelling, names and places, etc. Original thinking was less important, and often discouraged (particularly in areas like religion and national mythology).

Lately I've become "recall lazy." Why try to remember a name or date if I can find it on the Internet?


We don't have the original answer sheet. The probable answer is "the liver and the kidneys." That's what I learned in school.


IQ tests have been designed to be "culture-free" and this is a big reason why recall questions are less important on them. This may also be why IQ scores have risen over time: people have been reallocating cognitive ability away from recall and toward abstract reasoning.

Anonymous said...

whatever you say Peter....when i was saying recall questions are not evident in IQ tests i didnt mean it disparagingly just as a testament to the priorities of our society.

also yes abstract reasoning is the form of intelligence most directly affected by the Flynn Effect if we look at the rate of gains for Raven's versus other tests.

i do agree that the reallocation might be sensible. culture free is also a way of totally ridding the tests of recall altogether because a culture free test is defined as something you cant use previous information from!

Santocool said...

Show me a reliable source about it. I searched on google and all sources i found are about confirming that dutch famine intergenerational or "epigenetic" effects are real.

About spontaneous mutations. How something that cannot be properly predicted can be properly predicted and eliminated??

Pumpkin Person said...

Yet that famine had no discernable impact on the mental development of unborn children from that time:

Did it have any discernable impact on their height? If not, what's the relevance of the study?

Fin said...

I think the estimate on bread depressing height might be overstated. Most bread during this time period consumed was made with white flour which lacks the phytic acid of whole wheat. Also the wheat growing regions of China are significantly taller than the rice growing areas. This is interesting despite white rice being devoid of phytic acid. Wheat has over twice the protein of rice and many times more calcium. My theory as to increase in height has more to due with the spread of refrigeration and fresh foods like fruits, milk, meat being available. Also a Dutch study found alleles for height had increased in Dutch population supposedly due to sexual selection by women for taller men. The same had not occured in all countries studied.

Fin said...

Also it seems that wheat may depress IQ in Japanese school children when compared to rice. This seems to affect verbal IQ less so.

Fin said...

Here is the study regarding Dutch I believe.

Anonymous said...

Obviously, the difference stems from lack of education in the past. Regardless of the education, people who can read, assimilate data, etc., develop higher IQ. You are making a fetish of this malleable and flawed tool. Also, why are you using google accounts for a messaging service, given censorship and tracking?