Levitt and Fryer on race and IQ

From their April 2013 AER piece:

Analyzing these data, we find extremely small racial differences in mental functioning of children age 8 to 12 months.  Absent controls, the mean white infant outscores the mean black infant by 0.055 standard deviation units — only a sliver of the one-standard-deviation racial gap typically observed at older ages.  The raw scores for blacks are indistinguishable from Hispanics and Asians, who also slightly underperform whites.  Adding interviewer fixed effects and controls for the child’s age, gender, socioeconomic status (SES) and prenatal circumstances further compresses the observed racial differences.  With these covariates, we cannot reject equality in test scores across any of the racial/ethnic groups examined.

The piece is titled “Testing for Racial Differences in the Mental Ability of Young Children.”  Versions of the piece are here, but I believe the final version is not yet in jstor.

Note that to the extent you treat parental IQ as affecting the IQ of the child through environment, these results are consistent with a wide variety of accounts of racial gaps in IQ.  Still, there is no serious evidence, from these results, against the claim that the measured racial IQ gap is due to environment and environment alone.


I imagine the mental functioning being tested of an 8-12 month old infant is pretty functional/basic and not likely to reveal the more meaningful/developed differences of an adult mind.

Right. From the article:

"To the extent that some aspects of adult intelligence only emerge at later stages of development, or that these aspects go unmeasured in the early test of mental function, the genetics story cannot be definitively rejected with these data"

If racial differences are absent at age one-year but significant at age three (as they apparently are) an alternative explanation could be that there appears some cognitive function between ages one and three which is genetically influenced and that an early test misses out on entirely.

Ha! All aspects of adult intelligence emerge at later stages of development.

One of the largest and indisputable genetic differences between races is the time to maturation. In terms of fastest to slowest time to reach maturation: Africans, Europeans, East Asians. Another large body of evidence indicates that age of maturation is highly correlated with adult IQ.

African genetics are compressing a developmental cycle into about 85% of the time as Asian genetics. The longer maturation time means that the end result is higher mental development. But at intermediate stages for the same chronological point the mental development gap could be smaller or even reverse at very early ages.

Think of the analogy of slow-cooked BBQ. Everyone knows that the best ribs are cooked low and slow for a long time. Compared to BBQ that's cooked fast you get a higher quality product. However if you insist on comparing the two types of ribs an hour after they start cooking then it's quite possible that the fast-cooked ribs will be more edible at that particular point.

Most of the development stage stuff, in terms of body size. finds that Asians are the earliest maturers, which makes sense given their small body size, even though they start out slightly smaller as well. Africans are more advanced than Europeans for sure.

"Asiatics, even those who are growing up under the best circumstances, are less tall at all ages than well-off Europeans and Africans. Chinese and Japanese mature as early or earlier than well-off Africans and considerably earlier than populations in Northwest Europe.....At puberty Asiatics have a different pattern of skeletal maturation from Europeans and Africans, increasing their rate of maturing more swiftly and reaching full skeletal maturity earlier." (Eveleth, P.B. & Tanner, J.M. (1976) Worldwide Variation in Human Growth: Cambridge Univ. Press, London p. 274)

In terms of brain development, I don't think there have been any explicit tests, but with regard to the neurocranial shape and size differences - http://www.plosone.org/article/info%3Adoi%2F10.1371%2Fjournal.pone.0035917 - Africans seem to attain slightly earlier completion of brain size development, although this may not correlate well with cortical pruning and what have you.

As the father of a pair of 25 month old twins, I'm surprised they could detect any differences in the 8-12 month old age set. There's a huge leap in intelligence from 12-24 months.

No kidding, my 24-month-old is a walking, talking little person. At 12 months she was crawling and babbling and pretty much useless (though cute).

When Usain Bolt and I were born, we ran at roughly the same speed.


Thread winner.


"Still, there is no serious evidence, from these results, against the claim that the measured racial IQ gap is due to environment and environment alone."

Isn't that... exactly what this study is attempting to demonstrate? I can't tell if this is snark, mood affiliation or a typo, honestly.

I read it as an economist compliment...this paper shows what it says (no early diff in IQ by race)...a high bar for any paper, even an AER, and especially in a large contentious literature. It's rather common with papers to accept the empirics and contest the inference (or contest both).

Less saying, more showing, plz.

(not you, personally)

(not that there's anything wrong with that)

A version of this paper has been around for a very long time, so it's nice they finally got it published. Commenting on this paper in 2006, Steve Sailer pointed out that among infants there are also no racial differences in the ability to slam dunk.

Intelligence differences in infants, to the extent they can be measured, have very low heritability within races. So perhaps this study just shows that there are no racial differences in those environmental circumstances that influence intelligence in infancy.

No one's measuring the ability to slam dunk. Relevant question is whether there are infant traits / skills that are reliably correlated with ability to slam dunk later in life.

Well, right, but there aren't...

Well, there are...but the point is to not use those ones.

In the IQ case there's a fairly good correlation between 3-year-scores and adult IQ (I think).

The interesting study may be to track Levitt's infants to age three and see if ( on an individual level ) there is or isn't any correlation between 12-month scores and 3-year scores.

No, the issue that I pointed out to Levitt a half dozen years ago is that his test of baby liveliness is a racially biased predictor: e.g., Asians average worse than whites on Levitt's baby test but average higher on IQ.

Thanks. As I pointed out 7 years ago in response to the announcement an earlier version of this Levitt-Fryer paper:

"I haven't seen Levitt and Fryer's paper, but their gimmick looks pretty obvious -- you make the IQ gap disappear, hesto presto, by pointing at something that's not an IQ test. There are no IQ tests that purport to measure IQ much below age 3, and even age 3 or 4 is pretty shaky. At "eight to twelve months," you are mostly measuring physical coordination and/or personality development, not advanced rational problem solving ability, which is the main point of IQ testing. It's been known for decades that black infants tend to be ahead of white infants on those dimensions, with Asians in third place on average."

As the late J.P. Rushton wrote:

"The lower birth weight for black babies is largely part of a genetic life-history difference with shorter gestation times. Although black babies are born a week earlier than white (or Asian) babies they are more mature as measured by amniotic fluid, bone development, and other indices. The faster maturation of black babies continues throughout life....they can do eye-hand coordination earlier, hold their necks erect earlier (by a couple of weeks), walk earlier (by about a month), enter puberty earlier (by about a year), have children earlier, and die earlier (by a few years)."

What about biracial and multiracial babies?

I don't know about maturation, but in IQ they split the difference

Just want to point out that I find both terms to be misnomers. No one has more than one race. One has one, or none. "Biracial" and "multiracial" people have mixed ancestry, and no race. When you mix a Malinois and a Border Collie, you don't get a dog with two breeds, you get a mutt.

What about American blacks? They average around 20% European ancestry. Are they technically not black? Or would they be a separate race, distinct from African blacks? What about Southern Europeans, who average a small percentage of black ancestry?

Correct. American Blacks aren't Blacks. Most other countries have a name for each racial mix. In the U.S., the "one drop rule" is responsible for our inability to understand racial complexity.

Well if American blacks aren't black because they have some European ancestry, wouldn't that be following the one-drop rule?

What about Southern Europeans? They've been found to average a small percentage of black ancestry. What does that make them?

It makes them mixed. There's no rule out of nature that says you can only be Black or White. The only reason it matters is that White has meant high status and Black low status in our country's history so there's an incentive to create a clear distinction. I know firsthand that the caribbeans, South Americans, Africans and some Europeans don't automatically label mixed White/Black people as "Black". And I only include those examples because I know them. In fact, some cultures have distinct names for the mix (Indian/African, White/African, Southeast Asian/African...) and different names depending no whether your parents are of different race or your grandparents and so on..

What about American blacks?

I suppose you could categorize "African-Americans" as a separate race of their own, or put them in the same race as sub-Saharan blacks, or call them "mixed," or go by their preference. I don't think the question or the answers impact my point.

Well if American blacks aren’t black because they have some European ancestry, wouldn’t that be following the one-drop rule?

Google "mustifee," "mulatto," "quadroon," "octoroon," etc., and you'll get an answer.

If American blacks have 20% European ancestry, then that would mean that they have mixed ancestry. If people with mixed ancestry have no race, then how could American blacks have the race of sub-Saharan blacks or be there own race?

What about Southern Europeans or other groups that average mixed ancestry? Would the same reasoning apply?

It's not that they don't have a race, it's that their race is not Black nor White. The reason we impose those classifications is better understood by studying American history.

Peter, anyone can repeat himself. Watch:

If you mix a Malinois and a Border Collie, you do not get a dog that is both a Malinois and a Border Collie. You get a dog that is neither; a mutt.


The Onion covered this long ago.


"Another test, in which the infants were placed on a mound of dirt outdoors during a torrential downpour, produced similarly bleak results.

"The chicken, dog and even worm babies that we submitted to the test as a control group all had enough sense to come in from the rain or, at least, seek shelter under a leafy clump of vegetation or outcropping of rock," test supervisor Thomas Howell said. "The human babies, on the other hand, could not grasp even this incredibly basic concept, instead merely lying on the ground and making gurgling noises."

According to Howell, almost 60 percent of the infants tested in this manner eventually drowned."

This explains Leonard from The Big Bang Theory.

Study: Dolphins Not So Intelligent On Land

GAINESVILLE, FL—Although dolphins have long been celebrated for their high intelligence and for appearing to have a complex language, a team of researchers at the University of Florida reported Monday that these traits are markedly less evident on dry land.

According to study researchers, a group of 25 bottlenose dolphins removed from their holding tanks failed 11 exercises designed to test their basic cognitive abilities and reasoning skills.

"The dolphins were incapable of recognizing and repeating simple gestures," said study co-author Dr. Scott Lindell. "Their non-verbal communications were limited to a rapid constriction and expansion of the blowhole, various incomprehensible fin motions, and heavy tremors while they lay prone on the lab table."

After capturing the dolphins from the ocean, Lindell and his colleagues tagged them and placed them under the intense, high-wattage lights of a moisture-proof lab. The researchers then administered an extensive battery of tests designed to measure everything from the dolphins' self-awareness to their aptitude for writing and reading comprehension.

"Dolphins have a popular reputation for being excellent communicators," Lindell said. "But our study group offered only three types of response to every question we posed: a nonsensical, labored wheezing, an earsplitting barrage of unintelligible high-pitched shrieks, and in extreme cases, a shrill, distressed scream." -- The Onion

What I've wondered for years is why that one is no longer on The Onion's website. The other articles published with it are still there

Aren't there also important differences across races in the predictability of adult height from birth weight and height? In that sense wouldn't intelligence be like height with differences across races in terms of development and correlation between infant indicators and late childhood/adolescent measures and then to adulthood? And of course height is also a mix of genetics and environment although as with IQ almost all environmental effects seem modest to vanishingly small after early childhood.

And of course height is also a mix of genetics and environment although as with IQ almost all environmental effects seem modest to vanishingly small after early childhood.

A caveat, environmental effects are small above a threshold level. For example, malnutrition will have significant permanent negative IQ effects, but above a certain nutritional level they are insignificant. Or high levels of lead will have a negative effect, but below a certain level they are insignificant. There is little doubt that environmental effects can be significant at extremes. But assuming minimal nutritional and health standards, environmental factors appear to play a minority role.

This kind of research, especially as interpreted in the broader public, isn't really probative regardless of what's shown.

First, there are genetic differences that don't show up at age one:
- There are no differences between male and female babies in breast size
- Family history type patterns (whether genetic or lifestyle based) in susceptibility to heart disease, hypertension, Alzheimer's or male pattern baldness would be missed

Second, there are environmental effects that do show up by age one:
- children of women who smoke or drink or use drugs, or premature babies, or those living in lead painted houses etc all probably show adverse effects at age one.
- Children in Chinese and English speaking households show systematic differences in sets of phonemes recognized and distinguished by nine months.

Shrug. We already know that IQ reverts toward what is expected due to heritability as people move toward adulthood, one reason of which is that it's difficult to measure the IQ of those who are very young.

What sort of experiment / measurement would you recommend to test whether IQ is nature or nurture? Not being snarky, just curious. You think of this as an open question or a settled one?

I place great emphasis on the speed of reaction to stimulus as being correlated to IQ, see my paper: http://journals2.scholarsportal.info.myaccess.library.utoronto.ca/tmp/2802204478791895184.pdf

Do you mean individual differences or group differences? No one doubts that individual differences are heritable to a large extent; many different methods show this. There are also methods that could settle the question of racial differences (at least as far as American whites and blacks are concerned), but this research isn't being done for obvious reasons.

I guess I meant group differences: i.e. "Is the IQ lead of adult whites over blacks a genetic or nurture issue?"

Infant studies (like this one) show no genetic group difference, but critics attack that by the "relevant skills hadn't appeared yet" argument. OTOH, testing older kids means that pesky environmental factors necessarily muddy the genetic picture.

How does one resolve this debate? A study with black kids adopted into white families? Twin studies? I don't know.

Even before looking at the data what's a clean study-design or measurement strategy?

The appropriate method is explained in the link I gave. Quoting Charles Murray:

To the extent that genes play a role, IQ will vary by racial admixture. In the past, studies that have attempted to test this hypothesis have had no accurate way to measure the degree of admixture, and the results have been accordingly muddy. The recent advances in using genetic markers solve that problem. Take a large sample of racially diverse people, give them a good IQ test, and then use genetic markers to create a variable that no longer classifies people as ‘white’ or ‘black,’ but along a continuum. Analyze the variation in IQ scores according to that continuum. The results would be close to dispositive.

I would add that it would make sense to control for things like skin color and SES. An even stronger study design would correlate white admixture and IQ between black siblings.

Real life's a pretty good test. Blacks don't create space shuttles or skyscrapers on their own, anywhere.

When are people going to realize that heritability of intelligence or any trait at the individual level necessarily implies heritability at the group level? I mean between any two groups you're starting with a given stock at a given time, with measurable differences in IQ. To the extent these groups remain intact, individual heritability will aggregate to ensure the difference remains, unless we, as Sailer has joked, start hitting the smarter kids over the head with ball-peen hammers (now there's an environmental effect we can all agree will impact, pardon the phrase, IQ).

The problem is how you select the the group. Has anyone ever selected a racially diverse group with a similar social status, i.e. all from educated families living in the same type of neighborhood or all uneducated living in the same kind of neighborhood? The problem with identifying people by race is that we don't actually know what genetic makeup people. We usually just base our classifications on the way they look . Someone could look Black but be 75% White and vice versa. In fact there's a phenomenon in our history called "passing" whereby Blacks passed as Whited because of their light skin color to avoid social stigma, while their siblings remain Blacks

Claude, I'll accept that objection when the gov't throws up its hands and says it can't figure out who should get the racial spoils, and who shouldn't.

Svigor, WE are the government. We've been instituting racial discrimination since the country was found. Why do you think it'll change? That's who we are. We see people as a race before we see them as people. Why do you think in an economics blog the post that gets the most attention is the one about race? The only new and exciting thing is that the race has forcefully made sure it got all the spoils (through the government) and is now fighting so it stays that way.

Twin studies of adopted twins raised together and separately, or of identical and fraternal twins. That is how you establish the range of genetic heritability in X trait. Very old method, as yet unsurpassed. And the question is pretty well settled. IQ is highly heritable. The exact estimates range from 50% of the variance to over 80%. This still leaves a portion to be explained by environment, but there are so many environmental factors as to make each of them relatively unimportant relative to genetics.

The twin method has, in some ways, been surpassed by the genomic methods developed by Peter Visscher and colleagues. There are two methods. The first compares the sharing of alleles (identity by state) between unrelated individuals and correlates the results with phenotypic similarity across those individuals. The second method correlates the percentage of shared alleles (identity by descent) between siblings (the average percentage is 50, but it varies with a standard deviation of about 4 percentage points) and their phenotypic similarity. These methods are superior to traditional family-based methods because they are not dependent on assumptions about equal environments or random placement. The downside of these methods, particularly the first one, is that they underestimate heritability somewhat.


What sort of correlation have the methods of Visscher et al. shown with heritability? Is it >.50?

The first method generally gives narrow-sense heritability estimates for IQ of 40-50 percent. The second method has not, AFAIK, yet been used to study IQ, but in a study of the heritability of height it gave an estimate of 80 percent, which is similar to twin studies.

Settled, and both, though somewhat more nature than nurture in developed countries.

Have they done any studies on this issue with twins who are raised by different-race families? It wouldn't control for everything of course, but could give us some estimate beyond 12 months.

Am I the only one who could care less about the racial IQ gap? It's sort of interesting from a purely academic standpoint, but it seems like something that's really only important to progressives who are committed to some nonsense world view in which everyone is the same and the only way to get better outcomes is with more and more funding and white nationalist HBD types who cling to "white achievement" to bolster their self esteem. And those two groups will be doing eternal battle until the end of days, over this or something else, so I'm not sure why I should care.

The appropriate unit of observation with regards to IQ is the individual. Sure, it's interesting to look at the factors that influence individual IQ, but the focus on race is just odd. You educate children, You don't educate races of children.

The problem is that "progressive" idealogies based on blank slatism result in discriminatory real world policies, particularly against Asians, and to a lesser extent, whites.

Which policies do you mean? Affirmative action? If indeed observed IQ is a somewhat even mix of nature and nurture would that bolster or weaken the case for AA?

I'm trying to imagine ideal policy responses to both extremes: IQ being pure nature versus pure nurture.

Does empiricism play a role in determining these policies or a subjective view of one considers as "fairness"? e.g. In a world where it was proven beyond doubt that blacks were indeed genetically-low-IQ would we be less disturbed to see them stay poorer / unhealthier / incarcerated etc.?

"In a world where it was proven beyond doubt that blacks were indeed genetically-low-IQ would we be less disturbed to see them stay poorer / unhealthier / incarcerated etc.?"

This is the problem: that is really a meaningless sentence. IQ is an individual trait. You can compare group averages, but the sentence "The sample mean for IQ for a population of blacks is lower than the sample mean for a population of whites" is not the same sentence as "blacks are less smart than whites."

Also, the idea of splitting all 6 billion plus people on this planet into black, white, and asian has to be one of the most absurd concepts to ever gain widespread credence. Again, you have one group saying that race is a social construct and another group saying that race is extremely meaningful and deterministic, and yet neither group has spent anytime considering a meaningful taxonomy of population groups.

The level of science in this comment section is not what I am used to on MR.

"Also, the idea of splitting all 6 billion plus people on this planet into black, white, and asian has to be one of the most absurd concepts to ever gain widespread credence."

I recommend you stop informing yourself on this subject as soon as possible or your head will explode. Here is a concept that genetically divides europe into groups, and it only gets worse from there


Here is one on world populations


I warn you, do not read them.


I think you're proving his point. If I understand him correctly, he's saying that even within the group of "white" people, there is so much diversity that to classify them all together as one group is kind of absurd. His approach is even more granular than yours, but the point is the same.

(I do understand what you're saying, about the fact that there are incredible differences between groups of people, which is also true. You can both be right, in this case.)

"Also, the idea of splitting all 6 billion plus people on this planet into black, white, and asian has to be one of the most absurd concepts to ever gain widespread credence."

Judging from the race-obsessed Census form I had to fill out in 2010, it's good enough for government work.

Oh no race-obsessed!!! How could they do that, didn't you patent that a fee years back?

The Census is only race-obsessed because we are. We have this compulsion to attribute traits to races when we can't even define what a race is. Invariably, our stereotypes and preconceived ideas precede every discussion on race and it shows. And we mix up race with ethnicity with culture with skin color and so on... Lumping all Blacks together as one race is not scientific. Lumping all Whites together as one race is not scientific. Lumping all Asians together is just stupid. In fact, there's no scientific definition of race. Race is entirely a political construct, for obvious reasons. People are tribal and they look for any visual distinctions to establish an "us" tribe and a "them" tribe.
As for IQ tests, I know this is an economics blog but please spend some time learning about IQ from psychologists rather than sociologists and economists. You'll find out the tests don't measure what you think they measure, but most importantly that IQ scores are almost entirely determined by nurture.

Fixed that for you:

"it seems like something that’s really only important to the federal and state governments of the United States which are committed to some nonsense world view in which everyone is the same and the only way to get better outcomes is with more and more funding"

Right, cause wealthy white folks don't spend all that much money on education either. They're content to sit back and rely on their good genes.

Sorry. Culture war is stupid. Being white doesn't make you a victim. You are paranoid. They are not after you.

Busing, affirmative action, disparate impact, Head Start, NCLB, and Race to the Top are some of the government policies that are explicitly or implicitly about the black-white IQ gap. You cannot explain them away with your ad hominems.

Wealthy parents spend lots of money on education for two reasons:

A) To get their children around other high-quality children
B) To get their children social connections

In that order.

An obvious addition:

c. Most parents think the quality of their kids' school really does matter for how well he does in school later. That may or may not be true (it's probably not true in terms of raising their IQ, but probably is in terms of learning good study skills and being prepared for classes they're going to see later), but it is widely believed. Just because you don't believe it, does not mean that everyone else agrees with you.

No, you're not the only one. The only problematic element is that there is a certain class/type of person out there who views any racial difference in IQ as a problem that must be solved, preferably by some technocratic government intervention. I think that to the extent IQ is environmentally determined, it is beyond any capacity to "fix" by rearranging environments, which at this point would require quasi-totalitarianism anyway. As someone said above, what are we supposed to do, take black children and put them with white parents?

Policy should be directed at managing the ensuing reality. Let people raise their own children, and give up this foolish quest to equalize everything. Keep AA around if it makes people feel good, which it evidently does. I don't see that it really does anything else.

"As someone said above, what are we supposed to do, take black children and put them with white parents?"

More or less, that is the direction that the conventional wisdom of the Education Reform movement is trending. Having read many hundreds of articles in the New York Times extolling the latest Ed Reform thinking, the basic idea is to take black children away from their families during almost all of their waking hours and have Teach for America-type Ivy Leaguers cognitively stimulate them.

The logic is similar to that which drove well-meaning progressives in the early 20th Century to set up boarding schools in Canada for First Nations children and boarding schools in Australia for half-Aboriginal children. Lately, official apologies get issued to the "Stolen Generations." Perhaps in 50 years, the President will issue an apology to the "Borrowed Generations."

Forget about the apologies; it was the right thing to do then, and it's the right thing to do now.

You are equating forced schooling of Native Americans with Teach for America? Seriously?

What aspect of "compulsory education" are you missing?

He's not "equating" them; he said "the logic is similar"; it is, and the similarity is relevant, particularly in light of the failure of those early progressives. I don't mean to pick on anyone, but this is a common fallacy--"how dare you equate X with Y!"--usually when no such equating has occurred, and the objection is moral (often disingenuously so) rather than logical.

He's not doing that either, Dennis. He's talking about using Teach For America type teachers in a new program. There's no connection at all between the two programs Bill brought up

To claim that social policies like AA are about correcting differences in IQ is willfully ignorant disingenuous. There's a well documented history of institutional discrimination against Blacks in our history which meant that even if a Black person possessed extraordinary ability beyond that of anyone in the country, that would not be enough for him to .gain access to resources that most average Whites enjoyed as a right. AA is about ensuring that Blacks of talent (emphasis on talent) have at least a chance. In some cases AA policies were implemented badly but that does not mean AA in general is to be indicted. There are only two countries in the world where racial discrimination was as well written and instituted as a matter of law. The other one is South Africa. Just because the laws have changed it does not mean the people have changed with them. Just read the comments on any discussion about race.
We are using a pseudo scientific justification of our prejudices. We just don't want to let go because we are comfortable with the justification. Here are nuggets ee willfully ignore:
1. IQ tests are culturally biased and scores vary based on sociological factors around the test taker
2. IQ tests don't measure what we think they measure.
3. There is no scientific definition of race
4. The quality of resources available varies greatly based on sociological factors
5. There's no data anywhere that proves that one's rank in society is the result of his IQ level but there's plenty of data showing that we purposely stifled people's intellectual and economic growth based on their race.
The list hoes on...

"Am I the only one who could care less about the racial IQ gap?"

Yes, pretty much you are. The government, for example, spends huge amounts to count every person in the country by race and ethnicity, because it tracks race/ethnicity representation in countless fields, and devotes vast efforts to trying to equalize them, such as suing organizations for disparate impact discrimination.

And speaking of "white nationalist HBD types who cling to “white achievement” to bolster their self esteem," here comes their drum major.

Good job for not even bothering to address the point. Could you be any more intellectually dishonest?

Sailer's whole shtick is race trolling, so I have no problem trolling him.

As for points, I made several above.

@ j r

Demonstrating that you don't know what a troll is. Race is, by a magnitude, the single most suppressed and ignored factor in the overall social landscape. Sailer could easily expend ten times the amount of time he does on race and it would not be overkill.

"Race is, by a magnitude..."

That's the perfect example of why you clowns don't make any sense. By a magnitude of what? The phrase "by a magnitude" does not mean anything absent an actual unit of measure. You use it anyway cause it makes you sound like you're all empiracal-ey, but you're not. Anyone with an intermediate understanding of empirical methods can see right through you.

By that standard ancient peoples who moved from bronze to iron were kooks because they lack a metric for hardness and ability to pierce and protect. Even if you don't have a hard and fast metric ala modern physics you start to see patterns.

A hefty genetic component to intellectual ability is by far the most robust, thus best, explanation for observable differences in performance. Such a notion is routinely ignored. Can you think of any other sort of robust theory in the social sciences that is suppressed and ignored?

That’s the perfect example of why you clowns don’t make any sense. By a magnitude of what? The phrase “by a magnitude” does not mean anything absent an actual unit of measure. You use it anyway cause it makes you sound like you’re all empiracal-ey, but you’re not. Anyone with an intermediate understanding of empirical methods can see right through you.

This is a perfect example of how you dodge the issue. In this case, with a silly semantic point about magnitudes, when he obviously meant "order of magnitude," which is, of course, a specific number.

He answered your question. You didn't like the answer. And in a desire to protect your own self esteem, you began questioning his.

A lot of policies in the U.S. are framed around questions of race by both the media and the government. Racial disparities in crime, education, income, and chosen career tracks are just a few of the examples.

So don't play the naif who wonders why a lot of people still want to discuss and research race when it ought to be quite clear to every thinking adult in the room. The issue isn't going away, not when it's codified into law and custom, and no matter how much you might wish it to.

Aren't you and steve both starting from principles that state there are huge inherent differences in race that would not exist absent of culture?

j r is questioning that assumption. As a student of history, I'm with him. It wasn't millennia ago during the dark ages when an Arab would have questioned if those backwards Europeans could ever get their shit together and learn Algebra.


"Aren’t you and steve both starting from principles that state there are huge inherent differences in race that would not exist absent of culture?"

J r doesn't question that principle. He simply states he personally doesn't care about it one way or the other. He doesn't understand why race and identity should be an important matter to anyone who isn't a progressive race monger or HBDer.

But the answer is that one should care about it because questions about race and identity are now woven into the legal fabric of American society. The government cares. The media care. And they've assumed provisional answers to these questions in the way they frame policies that perhaps aren't justified and, in any case, clearly deserve more than the apathy j r demonstrates for them.

j r is questioning that assumption. As a student of history, I’m with him. It wasn’t millennia ago during the dark ages when an Arab would have questioned if those backwards Europeans could ever get their shit together and learn Algebra.

History has shown pretty clearly that the backward Europeans did get their shit together. And did so to the extent that roughly 95% of Human Accomplishment to date has come from Europeans.

Despite seeing the trails blazed, paved over, lit, and lined with rest stops selling gas and giving out free maps, blacks still haven't demonstrated the ability to maintain advanced civilization, let alone create or advance same. Until they do, I think it's reasonable to at least consider the possibility that they're incapable. It's certainly reasonable to question taxes justified on the assumption that they're as capable as anyone else.

It's important because the world is being ruled by the egalitarian regime. The delusion of racial equality is baked into the cake, and has more downrange effects than I'd like to belabor here. The "progressives" you refer to have blamed the white race for the "failure" of blacks to measure up to their delusion.

So, it's a moral duty for all righteous men to smash that delusion. Only an immoral man can stand by silently while these "progressives" line up whites in swaddling to pay for black "failure."

Like other commenters, I'd be concerned that the measurement error is very high when measuring cognitive abilities in babies.

This is important if we are using standard deviation as our unit of comparison. In this case, I'd guess that a one-standard deviation difference in 1-year-olds would predict a much larger than one standard deviation difference in 5-year-olds, simply because the standard-deviation-unit for 1-year-olds is effectively larger because it includes so much more noise. (This is the same issue as whether one should use partial-R^2 or beta in interpreting results of a regression. Adding noise will lower R^2, but not create bias in beta-hat.)

Has anybody read the paper closely enough to know if the authors considered this point? In other words, is the .claimed "055 standard deviation" difference in units of the 1-year-old scores, or in units of the predicted later scores?

Since the most linguistically fertile period of the critical period is probably (probably) over by 30-36 months, I would suggest that the lack of a lexically-rich environment contributes substantially to observed academic handicaps between demographics. We know that the average child from a lexically-rich environment enjoys a vocabulary that is bigger than that of the average child from a lexically-deficient environment by tens of millions of words. Language deficiency is self-replicating; students who lack language skills are unable to effectively study or communicate with educators, increasing their relative disadvantage compared to their peers. That's especially true because we socially promote students with their age cohort despite a lack of requisite skills.

Of course, if language development is conditioned as substantially as early as we think, then you guys can no longer blame public school teachers for poorly-performing students, and use that blame as a cudgel with which to beat teachers and unions.

Did you mean "tens of thousands"? There are barely over a million words in the English language and a standard working vocabulary is only 10,000 words. The entire French language has under 100,000.

Anyway, plenty of studies address the issue you mention (adoption studies for one).

I don't blame teachers for poorly-performing students, U.S. students do about as well as anywhere after adjusting for demography. Still, results could be better at significantly lower cost.

The old "black parents aren't loquacious enough" theory.

Just come right out and say it, Freddie: you think black people talk funny.

This assumption, based on an old Hart and Risley paper, that black people don't talk enough is one of the pillars of the current conventional wisdom.

I guess that explains why you never see a black rapper.

I'm being overly snarky toward Freddie. There's a very real problem here, which is that that having a large vocabulary is not only a product of being smart, but is, to some extent, an aid in being smart. And the family members of black children tend to have smaller Standard English vocabularies than the family members of white children. Thus, the logic of many of the Ed Reform articles in the New York Times is pointed toward taking black children away from their families for as much of their waking hours as possible and depositing them in the care of college graduate professional cognitive stimulators. This strikes me as cruel and expensive relative to the fairly small IQ benefits, but if you don't want to wind up demonized as horrible person like Jensen or Murray, what else are you going to say?

Lol @ "tens of millions of words."

Perhaps FdB failed to understand this:


"The disparity was staggering. Children whose families were on welfare heard about 600 words per hour. Working-class children heard 1,200 words per hour, and children from professional families heard 2,100 words. By age 3, a poor child would have heard 30 million fewer words in his home environment than a child from a professional family. And the disparity mattered: the greater the number of words children heard from their parents or caregivers before they were 3, the higher their IQ and the better they did in school. TV talk not only didn’t help, it was detrimental."

You mean, middle class mothers should stay home and talk to their children rather than hiring a Spanish-speaking Guatemalan to mind their children?

ya do realize there was no mention of what language the words were in, right? When was the last time a non-bilingual-Indian person won the national spelling bee?

Just saying, in my Ivy League undergrad there were many people raised by non-English-speaking nannies that turned out just fine.

Also, middle class mothers can't afford nannies. They do day care.

"in my Ivy League undergrad there were many people raised by non-English-speaking nannies that turned out just fine."

You mean, the reason that blacks don't have as big vocabularies as whites on average isn't mainly because they were raised by their loved ones rather than by college graduate professionals?

> Also, middle class mothers can’t afford nannies. They do day care.

I don't think you've priced these two things.

No, I don't mean that, but thanks for the bizarre, irrelevant aside.

This is the sort of paper that explains why I would never make it in academia. Were I their professor, I'd tell my students to go back and try again if they handed me this paper. If I subscribed to the AER, I'd cancel my subscription.

Were I Tyler Cowan, I'd not find it blogworthy, except as a cautionary tale.

This paper has all the trappings of sophistication and meaning mounted atop a foundation of total bullshit (the idea of IQ testing an 8-12 month old children).

A few months back, I was searching for IQ tests for infants.

The only test that I could find was a test developed by the toy company Fisher-Price. The "test" (more of a checklist) was scored by giving the baby, say, two points, for ability to roll over, etc.

Once a baby reached a certain maximum score, the test didn't allow for him to exhibit anymore intelligence.


Yeah, why wasn't Tyler Cowen responding to this with some like, "how young do you have to dig to find intelligence at parity?" Kinda like how he does with economic policy and "there's always a higher derivative!"

Just a quick comment. In many ways, I agree with those who said that the group averages are pretty uninteresting. Although, as The Bell Curve pointed out, there *are* public policy questions that will ultimately require addressing any major differences (e.g. if there are no more well-paying manufacturing jobs, and all the jobs that are left require a certain IQ or higher).

Here is why I see that the individual is still the key. I work in high-tech, and used to work in greater Boston with an abundance of MIT and Harvard grads. While they were all smart, they did not all work equally hard or effectively. As an interview manual I once read said: figure out two questions, 1. Can they do the job, and 2. Will they do the job?

Finally, although being smart can certainly give you a pretty good job with no heavy lifting, it doesn't always keep you employed past your 50th birthday, whereas the same IQ and effort put into a print shop or car repair business would make you much more money in the long run.

I agree with those who said that the group averages are pretty uninteresting

The difference between Europeans discovering industrial age and Tasmanians losing an ability to make tools is uninteresting?

Sorry if I was being too flippant. I just meant I am not inclined to try and use group differences (which I believe are real) as a weapon in any kind of policy debate, as it seems to me some people want to do.

There has to be some subtle point they're trying to make that I'm not catching because I can't for the life of me understand the need for a paper that confirms that it's hard to distinguish the intellectual capabilities of kids not yet a year old.

If by "subtle" you mean that they're talking about averages across races and not individual kids, then yes, I guess so.

I think they're specifically addressing a literature that has found differences in IQ across races/ethnicities. If such a difference is not found among people in a particular age range, that's relevant to the discussion.

Let me lead off by saying I'm completely innumerate.

That said, I'm aware of the identical twins studies and other studies showing that the heritability of IQ is something like .50 to .80.

Let's say it's on the lower end of that spectrum and is .50. Someone like Bryan Caplan would tell us that such a finding should just lead us to have as many kids as we please and not stress about giving the kids all these resources to enrich them.

But what does that .50 variability represent in terms of IQ points?

Let's say a kid's genetics indicate an average IQ of 100. Then let's say that parent provides the kid with optimal environment of intellectual stimulation.

If the heritability of IQ is .50 what is the maximum amount that the parent could increase the kid's IQ? Does the .50 correlation mean that potentially environmental intervention could raise the IQ to 150?

We don't know all the factors that affect IQ, but we can be reasonably sure that parenting (above some low cutoff) is not one of them, because studies have looked for an effect and have not found one.

Parenting matters, but that's hard to see in most adoption studies because 20th Century adoption agencies generally did a good job of not letting potentially bad parents adopt: thus, restriction of range.

A French study searched hard for unusual Oliver Twist-type cases in which children were born to parents near the top or bottom of society and adopted by people at the other end. They tried to find 10 children for each of 4 cells (born poor / raised poor, born poor / raised rich, born rich / raised rich, born rich / raised poor), but could only find 8 participants for the last cell. I believe they estimated, based on this sample size of 38, that in these extreme circumstances, IQ at age 14 was something like 59% nature, 41% nurture.

The sample size is small, but the results seem plausible to me.

59-41 seems entirely plausible to me as well. What I don't get is the policy implications. What if it had been 90-10 or 10-90? Would be doing anything different?

What are the policy implications of understanding better how the world really works? Well-informed people will come to different conclusions about policy implications, but my assumption has always been that the well-informed will, on the whole, tend to reason their way to better conclusions than those reached by misinformed people.

However, many influential figures disagree with my view, and actively strive to misinform themselves and their fellow citizens.

Couple points:

1: Environment doesn't mean child-rearing, or rather, does not mean only child rearing. That is a very small part of IQ variability. Fetal nutrition has a much bigger chunk, as well as proper nutrition in childhood. Add in personality traits, random accidents (head injuries tend to reduce IQ), social peer group, schooling, and you start to get the mass that is "environment". Parenting, in terms of how you raise your child, isn't a large part of it.

2: IQ is clustered at 100, the definitional mean. Humans don't get down to zero, so the range of scores is much smaller, moreso because the test is normalized into a bell curve. The functional IQ range within which 95%+ of humans fall is from 70 to 130. I'm not sure where the math actually leaves us, as IQ is largely stable over the lifespan, but no, 50% of the variance would not get you to 150. This is especially hard since we by definition cannot test the impact of environment without genetics.

Providing children with nourishing food is a well-known child rearing task that parents work to fulfill. Keeping children from being hit on the head is another one, though not quite as iconic.

Yeah, I wonder if Tarrou thinks parenting has little influence on your schooling or social peer group either.

Is there any reason to believe that the relationship between genes and environment is linear? And if it's not linear, then percentages don't really mean much. More likely that genes provide an upper and lower bound to IQ along with a probability distribution of every point in between. Where one ends up on that curve is probably the result of an infinite number of environmental and developmental influences that affect how genes express themselves.

The whole nature vs. nurture argument is based on a rather suspect binary opposition that upon deep reflections seems rather arbitrary, but it serves to enforce people's pre-existing world view so we're stuck with it.

We know what "nature" is and where it comes from: evolution. What is "nurture" and where does it come from? I've been challenging nurturists to give me a physical-causal mechanism by what they mean "nurture" and have never had that challenge answered.

The fact is, though, that we all know exactly what nurture is and where it comes from ...evolution. See, nurture is just a function of nature and what the nurturists really mean by "nurture" is "free will", just another manifestation of what has been called spirit, the supernatural.

The social environment that does shape humans is no less a product of evolution than is an individual organisms genome.

Agreed. People create nurtures that suit them. At the racial/ethnic/group level we call this "culture."

I believe you could get a cleaner result if you tested the children at 2-4 months, so as to remove the muddying effect of the subsequent 6-8 months of child-rearing.

Oh no! You must test IQ in utero.

Research will eventually push the crucial date of environmental difference back to 8 months and 29 days per birth.

But not a day sooner!

I pointed this out to Levitt years ago: The tests he's looking at involving babies less than a year old are _not_ IQ tests in the normal sense of the word. They have some use as predictors of future IQ, but ... they are racial biased predictors. For example, on average, Asians outperform whites on IQ tests, but on Levitt's markers, whites outperform Asians. Why? Because what he is using is racially biased.

Begging the question. If racial gaps in adult IQ are caused by genes then Levitt and Fryer's test is "racial[ly] biased" but that's assuming what you set out to prove. If, on the other hand, Flynn is right and high Asian IQ is due to environmental factors unique to certain Northeast Asian environments, Levitt and Fryer's data is also consistent with that theory. 8-12 months is too soon to observe the effects of the sort of environmental factors Flynn has in mind.

I think this paper was written by scholars with African-American and Jewish backgrounds, was it not? In general, what are those major groups trying to prove? I say "In general" because there are always dissidents (Herrnstein, for example). But, in general, Blacks will want to prove the hereditary hypothesis, right? They'll want to "fix" the problem with massive amounts of taxpayer funds. In respect to Fryer: "he does admit to having a mission: 'I basically want to figure out where blacks went wrong.'" This statement in itself disqualifies him as a scientist and scholar, because it is the driving force of all his research and presupposes that an answer to the spurious question exists

The Jews -- in general -- also want to prove the hereditary thesis,because they want the goyim to believe in it and thus not be privy totheir "secret": the fact that they have been practicing a form of eugenics for centuries that has raised their IQ to the highest level of any group on earth. Perhaps this paper is just another link in the chain that binds Jews and Blacks in an already decades-long alliance.

Dr. Fryer's personal motivations for his research were the subject of a lengthy 2005 article by Stephen Dubner (Dr. Levitt's Freakonomics partner) in the New York Times Magazine:

Toward a Unified Theory of Black America

Roland G. Fryer Jr. is 27 years old and he is an assistant professor of economics at Harvard and he is black. Yes, 27 is young to be any kind of professor anywhere. But after what might charitably be called a slow start in the scholarly life, Fryer has been in a big hurry to catch up. He was in fact only 25 when he went on the job market, gaining offers from -- well, just about everywhere. ...

... he does admit to having a mission: ''I basically want to figure out where blacks went wrong. One could rattle off all the statistics about blacks not doing so well. You can look at the black-white differential in out-of-wedlock births or infant mortality or life expectancy. Blacks are the worst-performing ethnic group on SAT's. Blacks earn less than whites. They are still just not doing well, period.''

To Fryer, the language of economics, a field proud of its coldblooded rationalism, is ideally suited for otherwise volatile conversations. ''I want to have an honest discussion about race in a time and a place where I don't think we can,'' he says. ''Blacks and whites are both to blame. As soon as you say something like, 'Well, could the black-white test-score gap be genetics?' everybody gets tensed up. But why shouldn't that be on the table?''

Fryer said this several months ago, which was well before Lawrence H. Summers, the president of Harvard, wondered aloud if genetics might help explain why women are so underrepresented in the sciences. Summers -- who is also an economist and a fan of Fryer's work -- is still being punished for his musings. There is a key difference, of course: Summers is not a woman; Fryer is black.

Dubner's article about Fryer went on:

In person, Fryer gives the appearance of coming from a middle-class background, some kind of Cosby kid all grown up. But as I spent more time with him, it became obvious that that wasn't remotely the case. He began to tell me stories about his past that -- although I didn't know it then -- he didn't share with people in his ''new life,'' as he called it. ...

How many of his close family members, I asked him, had either died young or spent time in prison? He did a quick count: 8 of 10. ''Suppose you can separate people into two camps: geneticists and environmentalists,'' he said. ''Coming up where I came up, it's hard not to be an environmentalist.'' ...

His most ambitious project, which grew out of his belief in the power of environment, is an experiment designed to see if incentives can inspire minority students to improve their grades. ...

His full name is Roland Gehrard Fryer Jr. Two years ago, as he was entering the job market, the name suddenly led him to panic. He worried that some university dean might Google it and see that Roland Gerard Fryer was convicted of a 1993 sexual assault in Lewisville, Tex. But that wasn't Roland; that was his father.

The more Fryer told me about himself, the more it became clear that his research is directly, even painfully, inspired by his own past.

And here's the twist in Dubner's NYT Mag article on Dr. Fryer:

The final stop on our tour of Fryer's past was Tulsa, Okla. His mother, Rita, lives there with her second husband, Harold, in a black working-class neighborhood.

On this day, however, Roland's mother explained that things weren't as simple as Roland had assumed. ... When he mentioned that he used to play the saxophone, his mother brightened. ''My whole family was musical, you know,'' she said. Her mother, it turned out, attended Juilliard and played eight instruments. An uncle was a saxophonist with Duke Ellington. Her family, she said, had been a real force in Tulsa, running restaurants and a variety of other businesses.

''They really were the Talented Tenth,'' Harold said.

Fryer smiled. The concept of the Talented Tenth was promoted by none other than W.E.B. DuBois. It referred to the need for an educated black elite -- the top 10 percent -- that would serve as example and inspiration to their brethren.

Later that night, over Scotch and soda at an airport hotel in Tulsa, Fryer sifted through the discoveries of his trip. He hadn't known that his father was a math teacher. He hadn't known that so much accomplishment ran in his mother's family. ''I used to consider myself a genetic aberration or maybe an impostor,'' he said. ''But I actually have some pretty good genes.''

Interesting. I have seen Fryer described as basically coming straight from the ghetto, so it's surprising to read that his father was a math teacher and mother went to Juilliard. It makes his success much less remarkable.

Dubner's anecdote about Fryer ought to be famous -- it's a brilliant piece by a high profile bestselling journalist about a high profile academic with tons of high profile friends -- but when I type into Google:

"But I actually have some pretty good genes"

I only get two hits: Dubner's original article in the NYT Mag and my immediate blog post about it!

Fryer's relatives in prison also had good genes. I think your notion of remarkable success is somewhat incoherent.

Who would ruin good scotch with soda?

Maybe it was bad scotch?

In this thread: people who are certain that the infant test does not reflect true intelligence and are certain IQ tests do.

My problem with IQ tests (or any standardized tests of talent) is that they are not accurate for affirming the positive, only rejecting the negative. If you were looking for a genius and had 1000 IQ scores, would you choose the top one? Probably not. There are a dozen or so pathological ways I can think of for a non-genius could top that list. However, could you reject the bottom 50%? That sounds a lot more reasonable.

Same thing goes for athletic talent. Is the next LeBron out there the best at the lay-up drill? I doubt it. I think he would be wiser than to waste too much time trying to perfect it. But I'm fairly confident he's not in the bottom 80%.

Hence, I'd say IQ and other standardized tests say more in the right tail than the left tail, that is, they can rule out prodigies, but can't identify them.

Sounds like baseless conjecture

Its not baseless: just look up the list of the world's highest scoring persons on IQ tests...they are not a very impressive bunch. They don't seem to have come up with any cures for diseases, brilliant inventions, great works of literature etc. Their forte seems to be scoring high on IQ tests but not much else.

I grew up in a majority black neighborhood. Ya know, there's a "blackened sepulcher" joke here, somewhere.

There may be a joke, but certainly not a good one.

The important thing is that Cowen acknowledges the racial IQ superiority of east asians.
China already has a higher avg IQ than America. And since white people don't have babies anymore that gap is just going to keep getting larger.
Get used to it suckers.

And since white people don’t have babies anymore that gap is just going to keep getting larger.

East Asians ain't exactly crankin' em out these days either.

"East Asians ain’t exactly crankin’ em out these days either."

That's the understatement of the decade.

Almost every single East Asian political entity I can think of - Macau, Hong Kong, Taiwan, Singapore, South Korea, Japan, China - has a considerably lower fertility rate than that of white America.

The two exceptions? North Korea and Mongolia.

A couple of these East Asian political entities rank as having the lowest fertility rates in the world and about half of them have lower fertility rates than any place in Europe.

Indeed. The true future superpowers of the world are in Africa since by 2100, extrapolating from present trends, Africa will have a very large percentage of the world's population, maybe around half of the world's population and given that this region is beginning to develop, as it is gaining political and economic stability, we can expect, over the very long term, that Nigeria and Ethiopia will join China and India as the new great powers of the 21st century.

Extrapolating by current trends is useless, since fertility rates have a close relationship to economics. Poor countries, and people from poor countries, have lots of kids (unless prevented, as in China). This spikes quickly during the shift to industry, then dies as the standard of living reaches a modern level. Any nation which becomes rich enough to vie for the title of superpower will have a low birthrate by that time. Incidentally, this is also why the hispanic boom in the US is reliant on continuous immigration, because second-generation immigrants have birthrates more like natives than their parents. As long as Africa remains poor and backward, they will have high birthrates. A generation after they get to some reasonable level of stability and prosperity, that trend will end quickly.

Do tell! I love the part where they are gaining political and economic stability lol!

When I first glanced at the post assumed that it said 8-12 years old. That would be a very interesting finding. I mean they couldn't mean 8-12 months old, that would make it worthless.

Then as I started reading the comments I realize it was a useless blog post.

I believe (and want to believe) these findings. We need these findings to be true. And that's the problem. Would the opposite findings ever be published in a top journal by top academics? No.

My point is that if opposite results would never be published, it is not clear to me we should bother studying it. We should simply acknowledge that we believe it to be true, and we don't want to find out (honestly) if it is not true.

Politically incorrect social science findings do get published in academic journals, but you have to read the papers carefully. They don't get publicized much, however. And when they do get publicized in the mainstream media, an awful lot of people don't notice. For example, the Science section of the New York Times published dozens of articles over the last dozen years thoroughly dismantling the popular shibboleth that "Scientifically, race does not exist." How many NYT subscribers noticed?

What we need are for academics to issue public apologies. For example, Tyler could write in his New York Times column, "Like many economists who haven't actually studied the economics of immigration, throughout the 2000s, I consistently derided critics of illegal immigration as ignorant. Sadly, subsequent history has shown clearly that I, the representative of the conventional wisdom, was the ignorant one." Things like that would get people's attention.

This sounds like confirmation bias, to me. It's easier to understand and remember and accept data that agrees with your existing model of the world than data that contradicts it. So, people who are strongly ideologically drawn to the idea that race doesn't exist will tend to not remember that sort of article, or not read it, or discount it as meaningless or something.

There's probably a point at which confirmation bias starts to make sense. Eventually, most folks have to make a decision and move on.

This study tells us little to nothing about observed racial differences in IQ scores. The impact of genes increase with age. The fact that the racial difference is smaller among children than adults actually provides evidence for a genetic component behind racial differences. The studies that look at adults find evidence for genetic differences as well.

The idea that all races are exactly equal in average intelligence is fairly absurd. We know evolution occurred in the brain just as it did in the body, so it stands to reason there would be at least some differences in intelligence (a trait that has a fairly high genetic component).

So the AER now accepts papers about "the measurement of racial differences of intelligence". May I ask why? This isn't economics and instead should have been published in another type of journal.

But when you win the John Bates Clark medal, you get to publish crazier and dumber papers.

Edmund Phelps spend ~5 years on trying to make sense of the Austrian business cycle theory from 2005 onward, in part because he knew if he could rigorize it (he couldn't), he'd get it published in a top journal.

The Economics Department at University of Chicago is supposed to be one of the world's leading centers of academic excellence, yet two very famous professors from the U Chicago Econ Dept (apparently befuddled by the need to align their work with the dictates of Political Correctness) have now published papers with patently ridiculous assertions in regard to the Black-White IQ gap.

So yes, now many of us are chuckling over Professor Steven Levitt's silly claim that data obtained by supposedly testing the mental ability of 8-12 month old infants can provide important insights into understanding racial group differences in IQ. But perhaps many of you have overlooked the even more preposterous claim recently published under the signature of the esteemed Nobel prize winning Professor James Heckman who now opines that Blacks are plenty smart enough to perform well on IQ tests but it is just that they are holding back on doing well because they are not sufficiently motivated. Professor Heckman claims that the answer is to offer the Black students a reward of some M&M candies, then the Black students will score just as high as the White students.

I am really not kidding, check it out yourself-- on page 14 of the paper HARD EVIDENCE ON SOFT SKILLS, our brilliant Professor Heckman baldly claims "The Black-White gap in IQ can be completely eliminated by incentivizing students with M&M candies."


So in today's politically correct world of academia it is de rigueur to vilify, ignore, and suppress the work of the bad "scientific racists" (Hans Eysenck, Arthur Jensen, Phil Rushton, Richard Lynn, Tatu Vanhanen, Linda Gottfredson, Charles Murray, Richard Herrnstein) but when the good "antiracists" publish utter crap (like the two instances discussed above), everyone is supposed to pretend like they have made a worthy contribution to our understanding of racial differences in IQ.

I hear the Department of Education has already placed huge orders for M&M's.

To be fair, Heckman also wrote a paper (can't remember if he published) around the 2000s that definitively showed a black-white IQ gap adjusting for *everything*.

After reading the paper, one can no longer believe that various subgroups have the same mean IQ (or would were it not for discrimination, low income, nutrition, whatever an equalitarian comes up with).

I have been reading this blog for many years at this point and this post is the only one over that entire span (and thousands of posts) that I consider worthless (other than as a piece of mood affiliation).

Several reasons why:
1) There are no IQ tests that can reliably test an infant. The concept does not even make any sense given that if infants were having sophisticated rational thought, they couldn't express it in any way (gurgling doesn't count). The paper's description of its methodology is a joke for anyone with even a rudimentary grasp of the field (and the existing literature).

Anyone who pretends that there is such an IQ test either a) does not understand the concepts behind IQ testing, b) has no real knowledge of infants/children or c) has a strong pre-existing agenda that they are trying to prove by making s**t up.

2) I am shocked that this paper by Levitt and Fryar was published in any peer-reviewed journal (or that Tyler would non-incredulously blog this). I was tempted to just bash the journal editors as idiots but Tyler re-posting this with those comments is highly discouraging.

3) Even if one accepts the paper as legitimate (rather than an elaborate Sokol-like hoax) and that it raises a valuable argument, the fact that few, if any, other piece of evidences in more than a century of testing (across all formats and including all current tests) show that small a racial gap would tend to make one believe that this is a random false negative.

It is a troll post. Just look at the number of comments. Tyler posts this kind of crap occasionally, perhaps to attract readership to his blog. Highly discouraging indeed.

Infants differ significantly by age 1. Granted, my daughter is an extreme outlier, but by one year of age she knew over a hundred words. (I personally did not say my first word until I was over two years old)

As a believer in racial difference in IQ driven partly by hereditary I see this actually as a potentially valuable paper. The Null Hypothesis is always actually worth confirming. It doesn't prove anything more than it says of course as both groups (environmentalists and hereditarians) all would have been surprised if there were differences. If there had of been differences it would have been powerful evidence for the hereditarians. I don't know the quality of the tests of course, it may just be an artifact of the testing methodology.

Why is this in an economics journal? Because they use statistics?

At times like this, I can't help wondering whether Prof. Cowen isn't engaged in some sort of triple bank shot thingy. He's obviously sufficiently smart & knowledgeable to realize that on-average racial differences in intelligence & behavior are real & probably genetically based, but he also knows he can't say that in public without endangering his increasingly prominent & lucrative career as a public intellectual. So, from time to time, he links with approval to a bit of feeble environmentalist tripe - knowing that Steve Sailer will show up in comments and demolish his (i.e., TC's) own ostensible position with his (i.e., SS's) signature combination of un-flappable geniality & hate-facts. So the truth gets out - but without Prof. Cowen's fingerprints. At least, it gets out to those who read the comments - which, let's face it, is the only place where anything interesting ever happens on this blog.

I was pretty harsh above too (the post is essentially a troll post and should be called out as one) but marginalrevolution as a blog does provide a lot of value. Tyler's thoughts in particular on a wide range of issues are pretty valuable (I think) and his recent books have been amazing...

".. genetic influence increases with age. The heritability of general cognitive ability increases significantly and linearly from 41% in childhood (9 years) to 55% in adolescence (12 years) and to 66% in young adulthood (17 years) in a sample of 11 000 pairs of twins from four countries.."

The heritability of general cognitive ability increases linearly from childhood to young adulthood
Molecular Psychiatry November 2010, C M A Haworth et al.
http://www.nature.com/mp/journal/v15/n11/full/mp200955a.html ..

I'm not reading through the whole thread to see if someone made this same point, so just in case no one did . . .

"Analyzing these data, we find extremely small SIZE differences in CHEST SIZE of children age 8 to 12 months. Absent controls, the mean GIRL infant outscores the mean BOY infant by 0.055 standard deviation units — only a sliver of the one-standard-deviation CHEST SIZE gap typically observed at older ages . . ."

So, clearly, if differences in phenotype aren't present at age 1 but present at age 3 or 10 or 15, then environment is most assuredly the cause of the differences.

Couldn't the cross-racial validity of "IQ" testing 8-12 month old infants be determined by seeing if those babies of a certain race who scored better at 8-12 months also turned out as adults to have higher IQs than their racial brethren? If there was no correlation within a race, it would not seem you are measuring anything real.

Maybe Levitt and Fryer can start looking at Apgar scores and calling them a measure of IQ.

The heritability of IQ in infancy is <0.2 and the infant-child genetic correlation for IQ is <0.3 -- meaning that genes contribute little to IQ in infancy and that the genes that do are largely not the same ones as those that contribute later on. So...

Hasn't it been proven that 5-month old pigs and dogs are smarter than 5-month old humans?

Shows how worthless this study is...

Comments for this post are closed