Tuesday, December 19, 2017

Is the (intraracial) word gap narrowing?

Steve Sailer, via James Flynn, reports that the (credited) namesake's effect could be a thing of the past. Specifically, scores at the top aren't rising and may actually be declining.

This presents a good enough opportunity to point out what has caused me some low-level cognitive dissonance for years now--there isn't much evidence in the GSS for assortitative mating, at least when it comes to vocabulary (and thus presumably IQ).

First, the following graph shows the mean Wordsum score among whites born in the US who were between the ages of 30 and 65 when they participated in the survey, by the decade they were born in. That's a mouthful, but here comes an even bigger one it is so to avoid racial confounding, language fluency issues, and to allow respondents to have had at least three decades to build their vocabularies without pulling in people who have been beset by the cruel cognitive decline senescence brings:

The early boomers look pretty good, but there isn't much variation across cohorts. There's little to glean here with regards to the Flynn effect. That's not to suggest the effect is oversold--vocabulary is one of the areas of intelligence least effected by the, uh, effect.

What does stand out, though, is the reduction in the size of a standard deviation in scores by cohort over time:

The size of a standard deviation in scores has declined by one-quarter in little more than a generation. That is, there has been a bunching of Wordsum scores towards middling performance over time, with both especially high and especially low scores declining as a proportion of all results. Among those born in the 1940s, 1-in-7 either completely bombed or perfectly aced the test. Among those born in the 1980s, just 1-in-25 either bombed or aced it:

This is the opposite of what would be expected to happen if assortitative mating was increasingly leading to cognitively gifted men pairing up with cognitively gifted women to have cognitively gifted children and dull men procreating with dull women to have dull children.

Parenthetically, because the quiz is multiple choice with five possible answers per item, scores of two or fewer correct answers are expected from participants ignorant of any of the correct answers who merely guessed randomly on all of them.

GSS variables used: ETHNIC(2,7-15,17-19,21,23-27,32,35,36), WORDSUM, COHORT(1900-1929)(1930-1939)(1940-1949)(1950-1959)(1960-1969)(1970-1979), BORN(1), AGE(30-65)


thekrustykurmudgeon said...

this is one of those things where there isn't really a good explanation.

Jim Bowery said...

Diversity is our greatest strength! That's why equality is good!

The intellectual straight-jacket really ramped up with mass culture (ie: 3 TV networks) -- and that was probably good for the neurological/intellectual development of some people but definitely not for the gifted. Standardizing the population can be highly profitable if you are in the economic rent business -- especially if it makes potential competition manageable. That's why rent seekers are all about importing more and more people -- turning them into a standardized mass of consumers/block voters and calling it "diversity". If you're a no-brainer investor in, say, land, all you know is they aren't making anymore of it, hence, open borders is Just The Thing. Collect that rent!

Also, bear in mind that males tend to have a higher variance on a lot of dimensions than females -- and society is feminizing everyone. Some call this progress toward expressing "The Better Angels of Our Nature".

Anonymous said...

My best guess is that this is a result of the population becoming more feminine.

Women are more average than men. So men produce more homeless hobos and more Nobel Prize winners. They produce more individuals at the tail of performance. Women cluster more around the middle.

See: http://www.denisdutton.com/baumeister.htm

When testosterone levels plummet, and men become more like women, then the entire population starts to cluster around the middle.

Sid said...

I would not be surprised if the narrowing variance in Wordsum scores is a result of language becoming more standardized.

With the rise of television, local accents and dialects are becoming much less pronounced. More and more, I hear people from the South speaking with a soft affect, rather than with a thick accent. Some people who grew up in the South admit they speak with a standard American accent because of television and their own efforts to speak with that kind of accent.

People with strong verbal intelligence can easily overcome a parochial background when it comes to building a large vocabulary... But I wouldn't expect the same of people with low verbal ability.

Similarly, people with high verbal intelligence are encouraged to keep their vocabularies in check. "Sound like a normal person!" "Speak English!" If smart people don't hear their peers use challenging or obscure words, those words won't stick, making it more challenging to remember what is timorous and temerity on a vocabulary test.

But this is just my reaction. There could be another explanation.

The Crumudgeon said...

Candidate explanation for the clustering around the mean .

More TV and exposure to popular press increasing wordsum scores of dullards, less reading, and simpler English in such reading as people do, reducing the wordsum score of the cognitively gifted.

Anonymous said...

The narrowing among whites makes sense. The recording comes just as the gates of immigration which were then mostly European and at that point non-Anglophone. Subsequently there hasn't been anywhere near as much immigration from Europe. As a European I'm always shocked how novel Europeans are treated in the US, even in big cities.

Anonymous said...

I'd also say that assortive mating hadn't really taken off until the Gen-Xers and mass enrolment in universities and women working out of the home seeking men of equal or greater financial standing. So the effects may not be as pronounced yet.

Jig Bohnson said...

Assortive mating by IQ (or more precisely by proxies for IQ such as income, education, and social class) is so obvious in our society now that I would look for alternate explanations here.

First of all, the Flynn effect is almost entirely manifest on the abstract reasoning portions of IQ tests such as Raven's Matrices, not on vocabulary. Perhaps vocabulary is the least linked to raw brain power (as would result from e.g. improved nutrition and lower disease load) and the most linked to societal conditions?

Second of all, we have an obvious change in societal conditions in this whole period which would depress vocabulary scores among those most potentially able to accumulate a large vocabulary - the rise of television and other mass media and the decline of reading, paralleled with the decline of educational rigor.

Quantitatively, there are statistics on e.g. the average number of hours of reading and writing assigned in college courses over the decades, and the trend is stark and exactly what you would expect. The trend holds for the elite universities and is far from being just an effect of so many non-college-ready people going to shitty colleges now.

Probably the explanation for the lack of exploding Wordsum scores at the high end is there.

Dan said...

Some possible explanations:


High school graduation rates went from under 10 percent in 1900 to 50 percent in 1940 to 60% in 1960 up around 77% in 1970. It has remained steady around that level.



Reading just isn't what it was. Now we have TV, cell phones, cars, Internet, video games and tons of other things to occupy us. News once came from newspapers. Now it is TV and twitter.

Back in the day, books were the only form of entertainment. Since my kids aren't allowed much TV or video games, I see that the amount of reading that goes on is tremendous. Most of it is crap like Harry Potter but back in the day books were more difficult.


Back in the day, there was emphasis on Latin and foreign languages among the most educated, which helps with the difficult words. Almost none of that is in education now.


Communication styles have gotten simpler in America. If you read a newspaper or political speech from 100 years ago, it is not so easy to read. Now there is an emphasis on avoiding difficult words. When was the last time anyone needed to check a dictionary for a news report or magazine article? Never! Big words are avoided as bad style when a smaller one will do. If an uncommon word has a simpler synonym, the uncommon word is left to die on the vine.

I think this explains a lot. Literacy is basically 100% and huge amounts are spent ensuring that ever last person is able to read. On the other hand, reading is not a common activity anymore and when it is, style dictates understandability at the expense of big words.

DissidentRight said...

This is really interesting and seems to defy explanation.

DissidentRight said...

Related: how can we judge between the competing effects of assortative mating and regression to the mean? I recall various articles calling attention to regression to the mean in black families. Naturally the effect is much stronger for blacks, because blacks at +115 are rocking 2 SDs. If their child is in the high 90s, that’s still a huge improvement over the black mean, though of course it won’t look that way to the parents.

To me it seems highly unlikely that we would see any measurable effects of assortative mating within only two or three generations, but what do I know.

And exactly how assortative is the mating, really? It’s not like the +2 SD white collar chads didn’t have plenty of access to girls in average-IQ professions, even in their own companies. Given the worthless nature of college degrees these days, they still do. Besides that, go back to regression to the mean. If +2 SD parents have a +1 SD child (or even a +0.5 SD child), that child is going to benefit from the parent’s social status, and that should have a significant effect on who the child marries (especially for girls). For assortative mating to have the full effect, you’d have to unpack that knapsack of white privilege.

Compare to miscegenation, which is “assortative” mating of a sort. If an 85 black knocks up an 85 white, regression to the mean suggests that the child will be more intelligent than either of the parents, because the child’s mean isn’t the black mean, it’s the black-white average. Here the one-drop rule makes ”assortative” mating appear clearly (from the black perspective, at least), since blacks are pulling themselves up by white bootstraps.

Halvorson said...

Those extremely difficult 1920's high math exams you see on the Internet aren't fake. I find it plausible that those stricter educational standards, particularly in writing and grammar, were better at producing people with perfect wordsum scores. Of course there were far more rural idiots 50 years than there are today.

A couple of weeks ago I read a history of my tiny Midwestern town written by an ex-principal in the 70's and was stunned by the quality of the writing. It was so far beyond what anyone of the same status would be capable of today that I felt like a barbarian looking at a Roman ruin.

Audacious Epigone said...

Really interesting speculation across the board, thanks.


Will disaggregate by sex, thanks.

Dissident Right,

The one-drop rule (for becoming anything other than white) means that miscegenation with non-Asians will increase the mean IQs of both blacks and whites (by increasing the share of the former and decreasing the share of the latter). Just an artifact of measurement but still kind of funny.


Assortive mating by IQ (or more precisely by proxies for IQ such as income, education, and social class) is so obvious in our society now that I would look for alternate explanations here.

I don't disagree with the sentiment, though I recall a Pew survey from not too long ago showing little change in the likelihood of people of similar levels of educational attainment marrying than in the past. May have just been an artifact of fewer people spending as much time in higher education a generation and two generations ago, though. I can't find the study now, unfortunately.


A couple years ago I leafed through a sample of letters written by WWI soldiers to their spouses at home and it is striking. There are some spelling and grammatical errors, but eloquence and sophistication of the expression is way beyond what you'd get from the modal American today.

Feryl said...

The further you go back (adjusting for different ethnic groups/civ. epochs), the more likely it is that written material was produced by a gifted and well-educated elite. So it's definitely not representative of what the low or possibly even middle strata was capable of (not much better than what you see today).

That being said, the (evident) peak of getting the best out of everyone with any potential occurred in the I-Fucking-Love-Education....And Science!.......And the classic Western ideal of education culture of the 1950's-early 60's. That's why 1940's births ace so many tests;

Later generations have been hurt by several things:

-dumbing down of material, most likely to help non-whites, which seems to be getting worse with each passing decade as the overall IQ of the population drops. If schools were still segregated, this wouldn't be as much of a problem.

- Speaking of which, segregating schools has been a disaster. Elite parents increasingly house their offspring in a handful of districts, and many other districts have to desperately shovel tons of money at POC. The end result is that many whites who don't have elite parents are underserved. And some of these prole white kids are pretty smart (as mentioned above, assortative mating hasn't put all the high IQ genes into the offspring of elites).

There were two pernicious trends to emerge in the late 60's (somewhat related to the above factors but still distinct):

- hatred of traditional Western civ./culture, which now accounts for a lot teaching curricula. We ought to be spending less time on guilt tripping the declining white population and making the growing POC population feel resentful. But given the demographics, the die has been cast (by the early 1990's, the then quite white child population of America was already being hammered by cult-Marxism which had grown even more gnarled since the late 60's and 70's; why reverse course when, almost 30 years later, the POC population is now much bigger?).

- Secondly, hostility to intellectuals that peaked in the 1970's. This definitely affected later Boomers and all Gen X-ers, and the jury is out to see if Millennials can finally get this out of our system.

Dan said...

Interesting things afoot in Europe. Eastern Europe (plus Austria, alone among top GDP countries, bitchez) are totally rejecting migrants and this is the entire goal of their respective leaderships.

Meanwhile Germany keeps doubling down. They want to use all the power of the EU to attack Poland, which is retarded because:

1 - They don't have the votes. It has to be unanimous. Hungary under Orban has already vowed to side with Poland.

2 - Germany attacking Poland again? Really? The political speeches write themselves. If I were Poland, I'd make an hour long speech at the EU or UN, **in German** about how tyrannical and imperial Germany is on the march and trying to conquer Poland again. I'd spend at least half of that time telling stories of horrific atrocities committed by the Germans who are clearly trying to destroy the world with terrorism again, God help us all.

Really, I'd have great fun with it. If I were Poland, I would make it my whole mission to see to it that all the good will Germany has tried to build up in the last seventy years is like a fart in the wind.

Honestly I have no idea how the German people can at once be so clever engineering-wise and politically so dumb. Perhaps the same wiring that prevents them from connected properly with humor prevents them from changing course politically. Humor involves abrupt mental changes in direction (i.e. the punch line). That's something missing in German wiring.

The world has changed, the old paradigm is wrecked, and those Germans are the last to get it.


Feryl said...

To clarify RE:my segregation comment. There's racial segregation, and then there's economic segregation. When we had racial segregation, very few whites balked at sending their kids to public schools. School district shopping was virtually unheard of in the 50's and early 60's. Most white kids in the 50's and early 60's went to similar quality schools, and teachers didn't have to worry about Tyrones getting in the way of their ability to teach white kids. Once schools were fully desegregated by the early 1970's, white parents in the South, the Mid-Atlantic, and the urban Midwest had no choice but to move to an area with "good schools".

As the non-white population has exploded over the last 40-50 years, with each passing year it becomes more expensive to find an area with "good schools". If you can believe it, as late as the mid 1970's many white parents in the Los Angeles area sent their kids to public schools. As each generation has fewer white kids, each generation of parents must expend more resources to keep their kids near as many white kids as possible.

Prole white parents can't be as selective, and their kids usually end dealing with at least a few black kids causing trouble in the class room. For all intents and purposes, we now have economic segregation.

Dan said...

Interesting pieces of Austria's new policy:

* Asylum is only temporary
* Asylum seekers must hand over their money and their cell phone. Cell phone is used to determine where they are from for when it is time to go back home.
* "Far right" Freedom Party is charge of interior ministry / homeland security (That is, Sebastian Kurz's Austrian People's Party, already right wing, handed over control of immigration to a party further to the right)
* Austria, while giving not an inch, wants to have a close and tight relationship with Germany. I read that as, they want to guide Germany out of the darkness and into the light. Austria and Germany, with a common language and more than a thousand years of historical ties, are naturally close.

Are Mark Steyn's obituaries of Europe premature?

Audacious Epigone said...


Wrt to the Germany-Poland relationship, yes, I can't believe the theme isn't worked harder--much harder--than it is. Steve Sailer's mentioned it a few times, but I've not heard/read much in the way of anyone else doing so.

Sid said...

Poland has some of the world's most bitter Tragic Dirt. Maybe the absolute worst. The Nazis built their worst death camps there (Auschwitz, Treblinka, Sobibor, among others). Needless to say, being in between Germany and Russia is bad, bad news geographically, and Poland faced partitions and foreign occupations time and again.

Usually that would elicit sympathy from progressives, but that's not the case here. In this case, the Tragic Dirt taints the Polish people and ensures they'll victimize people.

Look at this illustration: https://twitter.com/SidPolitics/status/885976541355798533?s=09

Orban's fascist shadow covers the whole of Hungary, envelops Slovakia, and slices through Poland, touching on Czechia and even Austria.

But Germany? No, Merkel took in the refugees so Germany was let off the hook. But if the illustration were redrawn today, I imagine the SPD's success in East Germany would ensure the shadow would creep over there.

Alliumnsk said...

There are two factors which complicate this. There is decline of intelligence with age and correlation of intelligence with longevity. So I suspect your finding is an artifact of this.

2Sid: in WW2, Belarus was quite harder hit than Poland.

Sid said...


Exact numbers are not easy to determine, but Belarus lost roughly 1/4 of its population from the war, while Poland lost around 1/5. (Many more people died in Poland than Belarus but this a product of Poland's greater prewar population.)

So yes, Belarus was marginally harder hit, but the devastation in both places was comparable.

In Western historical memory, the war crimes and human rights abuses in Belarus aren't as vividly recalled as what happened in Poland. The death camps in Poland are part of our lexicon, but the initial "Holocaust by bullets" (in which the SS shot Jews with firing squads) is not something that's as widely known about. Similarly, most everyone who isn't completely stupid knows Germany invaded and occupied Poland, whereas the partisan warfare in Belarus isn't well-known outside of the former USSR.

As such, in the Western mind, Poland has perhaps the most superlative "Tragic Dirt" in the world, but what's striking is that Western liberals are increasingly associating the modern Polish people with the darker parts of their history. So if Poles don't want to accept migrants by German edict, it's the Poles who are associated with fascism and atrocity, and not the Germans.

Audacious Epigone said...


I restricted the age range of participants at the time of participation from 30-65 in an effort to control for this, and I'm not sure why it would make a difference systematically over time. Maybe it should be pulled back to 30-50 or so to eliminate sampling anyone who has already begun to show signs of cognitive decline with age. On the other hand, vocabulary scores (as measured by wordsum) top out in the 50s and remain strong through the mid-60s. They don't really start declining until the late 60s.

legateofjudea said...

All the proposed explanations that have been written here are incorrect. The explanation is very simple. Race is not a constant independent variable over the time period measured. The racial makeup of people is changing faster than the increase in self reported multiracial population. People tend to report the race of one of their parents as theirs, usually the mother's race, rather than saying they are multiracial. This reduces the predictive power of the self reported race variable over time. Race is simply a ~15% correlation of genes within groups who were separated by continental distance over time. Take away the distance, add intermarriage and interdating, and the descriptive validity of the variable decreases as the number of generations increases.

legateofjudea said...

So why is the SD of the white group decreasing if the variable should have less predictive validity between groups? If we assume that assortive mating is taking place, it implies that people at the ends of the distribution, both top and bottom, are the more likely to have interracial children. If these children self report at multiracial or non-white, it will reduce the SD of the white distribution. While it has been reported that children are more likely to self identify with the race of their mother when they are biracial, I suspect this is because more non-white women are having multiracial children than white women. From personal experience, mixed white / hispanic children and mixed white / asian children tend to self identify as non-white unless they have very light skin.

Audacious Epigone said...


Interesting. The GSS doesn't ask respondents about the race/ethnicity of spouses, unfortunately. That would be tremendous data to have access to.