Is the reversal of the Flynn Effect environmental?

Maybe so, says a new paper by Bernt Bratsberg and Ole Rogeberg:

Using administrative register data with information on family relationships and cognitive ability for three decades of Norwegian male birth cohorts, we show that the increase, turning point, and decline of the Flynn effect can be recovered from within-family variation in intelligence scores. This establishes that the large changes in average cohort intelligence reflect environmental factors and not changing composition of parents, which in turn rules out several prominent hypotheses for retrograde Flynn effects.

In short, IQ relates inversely to sibling order, and the basic effect is not being generated by a changing composition of married pairs over time.

In other words, we have started building a more stupidity-inducing environment.  Or at least the Norwegians have.  But of course the retrograde Flynn Effect is starting to pop up in the data more generally, and not just in Norway.  From The Times of London:

The IQ scores of young people have begun to fall after rising steadily since the Second World War, according to the first authoritative study of the phenomenon.

The decline, which is equivalent to at least seven points per generation, is thought to have started with the cohort born in 1975, who reached adulthood in the early Nineties.

Have a nice day!

For the pointer I thank Michelle Dawson.


IQ like SAT like income inequality is not a good benchmark. "Accomplished people don’t bulk up their brains with intellectual calisthenics; they immerse themselves in their fields." Those people in 1975 also lived their childhoods with television. The decline in IQ scores of young people has seen no equivalent drop in industry or intellectual property or academic research (NBER paper right there). Also 1975 was the year the push-through tab on drink cans was invented so people had their priorities. Also the digital camera.

Another bogus study?

If only it were possible to bul up your brain with intellectual calisthenics

Strangely enough, other anonymous, I was about to post an endorsement for physical exercise at the bottom of this page. To put a random link on it:

Stop worrying about silly explanations and get out there!

"we have started building a more stupidity-inducing environment"

I'm surprised you would say that so definitively.

7 points is ridiculously huge. The Flynn effect doesn't seem to be g-loaded, which should minimize the practical effect either way. But, it does reflect some learned skill in handling abstractions, so maybe it is of some relevance- or maybe that skill is less useful now than it was recently.

We Morlocks are just increasing the supply of Eloi and fattening them via social media. It’s the nexr phase in the transition from hunter-gatherer to farmer; only now we’ll be farming our own and in the process improving the gene pool.

A 2017 paper suggests that this IQ drop is indeed due to shifting demographics, also known as immigration: "immigration predicted the decline, indicating that high levels of immigration promote the anti-Flynn effect ... immigration promotes IQ decline the most when the measure is higher in g-ness." Kind of self-evident; if a country like Norway, with an population of 5 million and an average IQ of 100, quickly imports tens of thousands of immigrants from the Middle East and north Africa, where the average IQ is about 82 (and that's being generous), how could that not drag down the mean? (the paper is called "What Causes the Anti-Flynn Effect? A Data Synthesis and Analysis of Predictors," published in Evolutionary Behavioral Sciences, September 2017)

But this study looks at within-family variation, which would seem to rule that out.

It is pretty funny how many people on this page clearly did not read or understand the article.


Even if a full fifth of the population was replaced by people drawn from a 82 IQ population, mean IQ would only fall from 100 to 96,5, i.e. _half_ of observed variation. And that is - of course - a much larger proportion of immigrants than is the case in Norway.

Bigots don't need no steenking facts.

The observed variation is a lowering by 2.5%. Where did you get the idea it was 10%? Anyway because the authors claim to have done an in-family analysis immigration should not be a factor.

The observed drop is 7-points.

3.5 x 2 = 7 ... not 10. Where did you get 10?


I'll just just leave this here and note that the trend really picked up in the early 70s and let the reader draw their own politically incorrect conclusion about the reverse Flynn effect...

Hu sez IQ is importent?

Fake news dunnit.

You only have to observe the mass media to know this is true. ;)

One word: dysgenics. When humanity broke out of the Malthusian trap with the Industrial Revolution, it also inverted the relationship between IQ, success, wealth, and fertility. This is especially the case with women - when women become educated, their fertility drops off a cliff. I wonder who has most of the next generation's kids in that scenario?

There's a small, red-pilled contingent of IQ researchers who think that man is in a race against time to create safe and effective gene editing technology (for positive eugenic interventions) before we're all too dumb to progress science any further.

'it also inverted the relationship between IQ, success, wealth, and fertility'

As demonstrated by the Hapsburgs, right?

Oh sure, it's dysgenic if you are Charles II of Spain.

But I suspect the typical "Hapsburgian" gene, dispersed as they were across thousands of individuals in the upper strata of society, did quite well for themselves. Better than a random human gene I would wage.

"But I suspect the typical "Hapsburgian" gene, dispersed as they were across thousands of individuals in the upper strata of society, did quite well for themselves."

Good for it.

"I suspect" is not evidence. Conjecture doesn't make it so.

And WTF is a "Hapsburgian gene?"

Well, not a single gene per se, but the Hapsburgs were noted for their, umm, genetic closeness - 'Charles II of Spain and the Habsburg Jaw

Also called the Habsburg Lip and the Austrian Lip, the Habsburg Jaw is a physical condition known by the modern term mandibular prognathism. It is characterized by a jutting lower jaw that is often accompanied by an abnormally thick lower lip and sometimes a tongue that is abnormally large.

It is believed that the Habsburg Jaw originated in a family of Polish royals, and the first person to have been known to have it was Maximilian I, a Holy Roman Emperor who ruled from 1486 to 1519. Many portraits of this monarch feature a pronounced underbite.

Through generations of royals marrying each other and therefore closing the ranks of their gene pools, the Habsburg Jaw manifested itself just about everywhere in Medieval Europe. The House of Habsburg, which is named after the Habsburg Castle in Switzerland, is associated with the Habsburg Jaw because so many of its members had it.'

Basically, at least from one way of looking at the subject, the Hapsburgs historically demonstrate what happens when a group of people with wealth and power decide to keep their purity of essence by not mixing with their inferiors.

It also used to be considered one of the cases used to demonstrate the importance of genetic mixing to increase IQ and fertility, as the Hapsburgs tended to show a marked decline in both over the centuries. But apparently these days, we don't care as much about history, and instead some dream of a superior group of people, using the Hapsburgs as a model.


Kings get a lot of tail. It's good to be king. Even minor nobility have more surviving children than the peasants. Evo Psych 101.

The average gene carried by a member of the Hapsburgs nobility would have done well enough, outside the inbred upper echelons.

'Even minor nobility have more surviving children than the peasants. Evo Psych 101. '

Which makes it a just so story, basically. The example of kings and their fellow aristocrats getting lots of tail pretty much is not at the level of fellow nobility, but at the level of the servant class. And there is basically no evidence that the servant class had fewer children than the nobility in 1500 (in aggregate, not necessarily in every individual case). Just ask which group continued to grow over the centuries, to the point that in more recent times, the concern has not been about the vast number of nobility dominating the lower classes through increased reproduction, but exactly the reverse.

At best, the picture is muddy - does the child of a kitchen wench and a lower level aristocrat count as nobility or not? Certainly not in the eyes of the aristocracy in 1500, but their opinion need not be definitive, of course.

"And there is basically no evidence that the servant class had fewer children than the nobility in 1500 (in aggregate"

Clearly you need to read Greg Clark, immediately. There is excellent evidence of this.

The fact that the nobility did not grow until they outnumbered the plebs... is a statement about the definition of nobility. The marginal aristocrat had kids who fell outside the definition.

What's fascinating about Prior is he can write lucid, if meandering, thoughts, that seem ever-so-slightly incongruous. Then, a few posts down the line the paint peels off and it's revealed that it all sits on top of a mountain of empirically false belief.

Just a

[Apologies for the incomplete post - accidentally submitted]

What's fascinating about Prior is he can write lucid, if meandering, thoughts, that seem ever-so-slightly incongruous. Then, a few posts down the line the paint peels off and it's revealed that it all sits on top of a mountain of empirically false belief with no logical coherence.

Take this ridiculous Hapsburg strawman, for instance. He's driven to deny some very basic observational science and history because he doesn't like evo psych or selfish gene stuff (yes, Prior, there's absolutely, certainly, no reproductive advantage to having wealth and power, whatever you say....) . Ladies and Gentlemen; I give you motivated reasoning 101!

But but but ... you can’t quibble with Prior on this, he used the devastating phrase Just-So-Story!

Thor get your act together. No wonder Loki gets all the chicks.

Kings get a lot of tail.

Yes. I'm aware of that. Though why we should assume that kings, Hapsburg or other, were particularly intelligent I don't know.

If a certain breed of animal stopped reproducing, and was supplanted by another, wouldn't Darwinian theory hold that the second one was in fact, superior?

It would be nice if people would read the post before writing their own posts about it. No, dysgenics does not explain younger full siblings being less intelligent than older.

In short, IQ relates inversely to sibling order, and the basic effect is not being generated by a changing composition of married pairs over time.

Inversely related to sibling order? So the basic problem is that there are too many only children and the ratio of eldest child to all the others is too high?

Oh please God let it be true. And please let me live for another couple of Thanksgiving Reunions to come!

Yes I would like that statement to put into plain English as well (I didn't see it in the abstract and I can't access the paper). Does this really mean that older children are less intelligent than younger ones? Or that only children, deprived of opportunity to sharpen their wits against their siblings, are more dumb than children from larger families?

Either way I suspect they are seeing an artifact where there is some hidden variable.

I am going with elder children being as dumb as a bag of rocks. You know you might think at my age I wouldn't care. And I don't. Honestly. But I may just feel compelled to drop that casually into conversation from time to time.

Except that in the context of the article, it cannot only be the opposite of what you suggest, else it wouldn't be cited as evidence of within-family decline over time. I'm a third-born and even I could figure this shit out.

*can only, pardon

Tsk tsk. Third born? You spell like a fourth born. (Joking.)

Yes, it doesn't make much sense unless it means that older children in the same family are less intelligent. Hence control for genes and family environment.

However, the IQ and birth-order effect has been known for a long time, so this must be an ADDITIONAL effect on top of that. I wonder at the controls...

Sorry, SMFS, it's the other way around. Older siblings have a +1.5 advantage per rank

A phenomenon, that I, as a firstborn (born at the height of human intellectual achievement apparently, circa 1973), feel compelled to point out. :-)

I hate to brag about it.

OK, big problem with this study: How the heck do they control for households with 0 or 1 child?

The study as reported seems entirely consistent with the "smart people aren't having kids" and/or "dumb people are breeding faster". Even if (k) child is smarter than (k+1) child, having large families could be correlated with low IQ itself.

Basically, this doesn't seem to rule out genetic or population dysgenic effects, regardless of the claims above.

Ladies and gentlemen: the good old nature Vs nurture debate. Nature = genes, nurture = environment where children grow.

The study from Norway is interesting because it shows the IQ variation occurs inside families. In simpler terms: parents with high IQ have kids with lower IQ.

Assuming IQ tests are a good measuring tool, these results show the environment matters more than genes. The big question is WHAT part of the environment? School, food, pollution, hobbies, culture?

Could it be the infamous resource curse kicking after 40+ years?

Of course reversion to the mean says that high IQ parents are likely to have children lower than theirs. But overall average intelligence of the population should not be affected by this. I can't read the paper so I don't know, but is the effect really material (even if it is significant in statistical terms). It could well be less children in a household, consistent with the theory that intelligence developed via an arms race against other people. The smarter your kids the more chance, in a zero sum world, that they will get more than the average share. Nothing is more zero sum than being a kid in a large family, you need every tool in the book.

Why shouldn't regression to the mean count for overall IQ as well?

Then it wouldn't be regression to the mean.

The effect is significant. Sample size 700k males, average IQ drop ~2.5 over 12 years, confidence interval ~0.1

More children creating stronger competition among them is interesting. But, that requires a new study. Even another dataset =/

There's a reason today's helicopter parents will do almost anything to get their children in high achieving schools, starting with preschool: Rene Girard. IC (intellectual curiosity) is likely as important as IQ in intellectual development. I would point out that the decline in IQ coincides with the development and pervasive influence of computer science in education. Not everyone has the propensity to excel in computer science; smart young people in the past built something tangible whereas smart young people today build something intangible.

"I would point out that the decline in IQ coincides with the development and pervasive influence of computer science in education. "

There has been no 'pervasive influence of computer science in education'. Not in the U.S. anyway. A very small fraction of students ever take any kind of programming class. Or if you mean ubiquitous use of things like iPads and Chromebooks, that doesn't work either. It has been roughly only in the past half decade that prices of devices have fallen enough for one-to-one usage (where every child has a device in class and to take home). Much too late to account for a trend that started more than 40 years ago.

We didn't have computers or computer science in school in the 1990s.
Well I guess we had *one* computer in *one* classroom, but that hardly counts as pervasive.

What makes you think they had some sort of pervasive computer infection of the schools of the 1980s and 70s?

If there's any technology to blame it would be television ... with sets first becoming common in most homes *starting* in the 1970s.

There's an old science fiction short story called "The Marching Morons". In it intelligent people realize that the Earth is becoming over populated and tend to have fewer children than the rest of the population. Over time the average IQ drops disastrously.

Not entirely fanciful as a source of Flynn, though I doubt enough time has passed for that idea to have much effect. If you combine it with the tendency of wealthier, more secure families (and nations) to have lower birth rates you might have something. Easy enough to test anyway.


Why come you got no tattoo?

My initial thought was that older parents will have accumulated more mutations in their gametes, especially fathers, so within family variation could actually be possible. By it's not credible that that could cause a 7 point decline over one generation.

The "7 point decline over a generation" is NOT on the article. It's an extrapolation made by journalists.

Figure 1A from the article show the Average IQ for children born between 1992 and 1991. The max is ~102.2 (1976) and the min is ~99.5 (1988) the difference is 2.7 over 12 years. Some clever editor took the trend from 1976 to 1988, extrapolate it over 25 years and.......a great attention-grabbing headline was born.

A really nice guy shared this figure on twitter and also was interviewed on TV to explain the article

Thanks Axa - the graph is very interesting - it seems like for some reason there was a spike in IQ in the mid 1970's and now it has return back to previous levels. Since we don't know the causes of the positive Flynn effect it is probably useless to speculate on why there is a negative Flynn effect - but I wonder if it is our old friend tetra-ethyl lead again? But my guess is that it will be some quirk in Norwegian army selection.

This is an effect but super small. Dysgenics is a bigger effect but still much smaller than claimed

Agree. Seems correlated with parental age at birth.

A decrease in IQ over time is easily observed in the general increase in tattoos. Of course, the tattoos are stylistically better so maybe the tattoo artist cohort is becoming more intelligent.

A decrease in IQ over time is easily observed in the general increase in tattoos.

Excellent point. But I think the increased quality is probably driven by increased demand, which induces more talented artists to enter the field.

Vaccines, kids take more and more vaccines, my kid got vaccinated against like 30 diseases, we need to eliminate all these diseases so we can get rid of the vaccines and stop destroying our children's brains.

In this specific study, they ruled out demographic changes but that doesn't mean broader changes in other studies such as the earlier German study I link to here can't be driven by changing demographics as well.

I was just thinking about this last night.
To throw out another possible theory - since gestational length is correlated with intelligence, the trend towards older childbearing and consequently early labor induction might be a cause. Basically in women over 39, doctors will tend to induce at 39 weeks, rather than wait for labor to occur spontaneously. Since the average is 40 weeks, and the standard deviation is 1 week, that shortens gestation for over half of all babies born to women over 39.

Link on brain development benefitting from longer gestation.

Link on reasons why labor induction is often used to reduce risk of fetal death:

And link on the inceasing use of labor induction to shorten a pregnancy:

It would be interesting to control for that variable in some manner other than birth order.

Some text from that first link that should horrify any parents who were convinced to induce early by their OBGYN:

Our findings indicate that, among healthy, low risk populations, shortened gestation, even within the boundaries of normal term delivery, has persisting influences on neurological development.


Interestingly, the benefits of longer gestation for development of the temporal cortex were apparent even for children born at term (37–41 gestational weeks).

The gestational age when the transition from intrauterine to extrauterine life occurs in normal healthy births can profoundly alter the course of brain development with long-lasting consequences. The significant linear association between gestational age at birth and brain development in young children challenges the commonly held assumption of a “non-linear” developmental course that defines fetal maturity as occurring at 37 gestational weeks. Our findings emphasize that there is a benefit for the developing brain of increased gestational length throughout the course of fetal development. In addition to providing new information about the importance of longer gestation beyond 37 gestational weeks for brain development, this finding has implications for medical decisions involving assisted deliveries. The decision about when to deliver a fetus, especially after 37 weeks, rarely involves concerns about the fetal nervous system. The findings reported here suggest that the neurological maturity of the fetus should enter the decision algorithm because even modest increases in gestational length have significant and long-lasting influences on the structure and function of the nervous system.

Looks like going over 40 isn't a great idea either.

Well, what if normal gestational length varies naturally and is hereditary, but has a 40 mean averaged across the whole population. So for some people, the normal length is 43 weeks, for others 37 weeks. In the cases where the natural gestation is 43 weeks, having a rule of never going past 40 weeks would reduce intelligence. For the people who naturally have 37 week pregnancies gestation would not be shortened because they would tend to spontaneously labor before they got to 40 weeks. Even if the optimum time to give birth is 40 weeks on average across both groups, you would still have an effect of reducing intelligence because you would be shortening gestation unnaturally for some people.

Side notes: Boys tend to gestate longer than girls so problems caused by early labor induction would be more likely to affect boys. And problems like hyperactivity and autism (which have been rising) seem to be more prevalent in boys. Coincidence?

But children born through IVF have similar or higher test scores than their classmates

This is probably just because smarter women tend to have professional careers that delay childbearing.

Actually they tend to be white, wealthy, well-educated, and in stable marriages. Fathers the same. But the babies tend to be born early without negative impact on IQ.

Also the relationship between IQ and sibling order might also be related to gestational length. First pregnancies have tendency to run a bit longer than later pregnancies.

The mere fact that so many nonscientists, non-geneticists are so eager to leap onto an academic paper about genetics, and draw vast conclusions about public policy, is evidence that IQ is the new phrenology, the mystic woo that validates prejudice.
If IQ were such a determining Darwinian factor, there wouldn't be any need to do anything about it at all. The higher IQ people would just naturally rise to the top. Alpha wolves, after all, don't need a referee to disqualify the beta.

And if the low IQ people somehow gain power over public policy, then maybe IQ isn't the determining factor after all.


Well, IQ is indeed the new phrenology, less intelligent people, as measured by esoteric testing, drag down the standard of living of the "normals" and the elites, except when they're babysitting or mowing the lawn. If there's too many of them, however, they supposedly tend to vote for populist ideological magicians and mess up the faux liberal societal paradigm. There's been a belief that lower IQs tend to engage in criminal activity more than their better endowed brethren. Have there been publicized studies of the prison population to see not what the average IQ among the convicts is but what percentage of them are well above normal in intelligence? Is Martha Stewart of below normal IQ? How about Bernie Madoff? Or Ivan Boesky?

I would like to see the numbers for a bunch of other factors such as child survival rates, height, what Hazel referred to with gestation duration, change in dietary composition, who looked after the kids during infancy and preschool. Mom had to chase them away with a stick to not get a shot to dry up her breasts so she could breast feed in those years, so the post war gains from better nutrition could have been turned around.

Could it be as simple as the post war generation were born from survivors of a pretty rough few years, were possibly the first generation that didn't miss a meal, then a reversion to the mean?

If you actually had reliable numbers, what use could be made of them?

What other massive change occurred since the 60's? Widespread availability of birth control.

As well as the prevalence of divorce.

Widespread consumption of rock (Beatlemania, etc.) and RAP.

Don't forget the other kind of "rock".

... crack rock cocaine became an *incredibly* popular thing starting in the 1970s.

"In short, IQ relates inversely to sibling order"

Paper is gated so I can't read it. Does it control for age of the mother? Later siblings are born to older mothers...

This is a well-known phenomenon and is independent of maternal age. Of course the effect is very small.

Could there be a biological effect in the womb that makes later children have lower IQ that is through off the results?

As I brought up above, shorter gestation. First pregnancies tend to deliver later.

I always assumed the Flynn effect was the best way to increase IQ scores is to feed and maintain health of child and give them a good basic education. So the big increases of IQ from 1900 to 1970 was mostly developed nations solved a lot of these so IQ went up. In terms of lower births have a very small decrease in IQ is probably more of parental interaction with younger birth decreases. (So I believe this benefited Japan from 1975 - 1995 in economic growth of Japan Inc. Unfortunately in the long run, that means you have falling populations which is another problem.)

In terms of Norway, can we all accept a small petro-state of 5M+ and nearly $1M for every new born is an exception to all other countries and higher birth rates will probably benefit the nation more in 20+ labor supply than .01% decrease in IQ scores.

One issue that isn't related to the elephant of immigration: Licorice root. Apparently it screws with the placenta's ability to shield the fetus from the mother's cortisol levels. Pretty much any dose, at any time in the pregnancy, will result in lower IQ baby. And its use is spreading (especially in herbal teas) because they can make stuff taste sweet. We imagine we would know, because licorice candies taste so strongly of licorice, but as ingredient in something else those licorice flavors are often masked. I was surprised at how many of my teas had licorice root in them. Licorice also raises cortisol and blood pressure levels, so it probably isn't good for indiscriminate use anyway.

Norway has seen a big jump in immigrants in the last 30 years.
"Beginning in 1970, immigrants and Norwegian born with immigrant parents constituted 1.5 per cent of the population. In 2010 this group had increased to 13.1 per cent of the population"

54% of immigrants come from Europe, 29% Asia, 11% Africa, 5% Americas. There are significant differences between the gender of immigrants and the area they come from.

Children born in Norway to immigrant parents are heavily from Pakistan, Somalia, and Vietnam.

Since the 1980's gender equality programs have been a growing force in Norway. Children spend more time in daycare less with parents. Girls have passed boys in academic success during mandatory schooling and are dominant in higher education.

Lastly BMI and IQ have a negative relationship.

This explains the number of Trump voters these days.

N general, people who are nearsighted do better on intelligence tests and achieve higher educational levels than those who are not,

The effect is via modeling the environment. The myopic youngster keeps a more complex map of where things are, for survival. The brain can be exercised, it is a muscle.

There seems to be a reversal of the Flynn effect going on in the comments. I've been a long timer reader but man everything has gone downhill from where it used to be.

Comments for this post are closed