That is a newly published piece by Elliot M. Tucker-Drob and Timothy C. Bates, the abstract is this:
A core hypothesis in developmental theory predicts that genetic influences on intelligence and academic achievement are suppressed under conditions of socioeconomic privation and more fully realized under conditions of socioeconomic advantage: a Gene × Childhood Socioeconomic Status (SES) interaction. Tests of this hypothesis have produced apparently inconsistent results. We performed a meta-analysis of tests of Gene × SES interaction on intelligence and academic-achievement test scores, allowing for stratification by nation (United States vs. non–United States), and we conducted rigorous tests for publication bias and between-studies heterogeneity. In U.S. studies, we found clear support for moderately sized Gene × SES effects. In studies from Western Europe and Australia, where social policies ensure more uniform access to high-quality education and health care, Gene × SES effects were zero or reversed.
I would put it this way: genes matter more when you equalize environmental influences in the United States, but not in many other countries. By the way, this also holds if you control for race and the greater racial diversity of the United States. One possibility is that there are greater environmental differences to be equalized in America in the first place, compared to say the Netherlands, one country where the gradient is quite different. In any case an interesting piece.
For the pointer I thank Michelle Dawson.
Seems a little hard to believe for UK-US comparisons. There are trends where the SES linked performance gaps have increased in the UK. I don’t think that’s due to an increasing genetic divergence / assortation, so must be an environmental (in the sense of everything not-genetic) trend. I also don’t think SES gaps are as pronounced in the US (?), and it seems unlikely to me that these, slightly less pronounced gaps, have more of a signal of people underperforming their genetic potential.
The meta-analysis is about interactions between SES and behavioral genetic variance components (heredity, shared environment, nonshared environment). The main effect of the shared environment, including SES, may (or may not) be large and growing in the UK without there being any interactions between SES and the variance components.
“One possibility is that there are greater environmental differences to be equalized in America . . . .” But those environmental differences differ over time. It would be interesting to see those differences, during times of relatively low inequality and during times of relatively high inequality. Of course, trends feed on themselves, as reflected in Cowen’s recent postings about matched and mismatched marriages (mismatched marriages have been declining (and matched marriages have been rising) during this period of relatively high inequality) and educational achievement. Indeed, the “environmental differences” today cover almost every aspect of life, even travel (in private jets by the affluent). And environmental differences not only feed on themselves, but intensify them (shared genes during periods of matched marriages if you prefer the genes vs. environment explanation). Moreover, periods of relatively high inequality create class division all along the socioeconomic distribution, each quintile fearful of falling to the quintile below and less likely to rise to the quintile above. I’ve commented before about my own family: during the first gilded age, educational achievement and marriage matched their social class, but following the end of the first gilded age, not so much. Americans have an uneasy relationship with notions of equality, at once praising it (the Declaration of Independence) and opposing it (with charges of subversion for those promoting it, even presidents (Lincoln, FDR, Obama)).
One possibility is that there are greater environmental differences to be equalized in America in the first place, compared to say the Netherlands, one country where the gradient is quite different.
In related news,
White definitions of merit and admissions change when they think about Asian Americans
Asians get high standardized test marks because the family values education and hard work, they study too hard, and because they have no life. Blacks get low test marks because they are genetically inferior, not because the opposite of “Asian traits” may apply on average. Neither of them face racism, they just suffer from either social or genetic inferiority.
Oh, and people who look for cultural explanations are leftist elites who are out to lunch and ignore reality.
We should learn to listen more closely to the white nativists on these questions, because results of one-off tests that have nothing to do with anything in the real world (IQ tests) are probably the single most important CAUSAL explanation for anything having to do with outcomes.
/sarcasm
IQ tests have to do with most of the real world. And, yes, your description is correct, other than that statement.
I’m talking about the act of taking the IQ test. Why try? There’s no payoff. Even if I’m super duper smart, maybe I just don’t care. You put the test in front of me and I don’t try very hard.
Conclusion: not trying very hard on IQ tests is correlated with (not causally) not applying oneself in school, and hence divergent educational and income outcomes.
Moreover, very few questions on IQ tests are very related to the world experienced by most people. Language questions? You have cultural, language and educational bias. Math questions? You can’t tell me that’s not related to a good education. Spatial reasoning? Cultural bias.
Don’t get me wrong. It’s plain as day that some people are smarter than others. I just don’t think that some useless test that has nothing to do with anything (perspective of test taker who gets no payoff) is a very good tool for drawing any meaningful conclusions.
Comments like this demonstrate your ignorance of the psychometric literature more than anything else.
The meta-analytic correlation between IQ and conscientiousness is ~0, and IQ tests have been shown time and again to be culturally unbiased. Back to the drawing board, Nathan.
msl – perhaps the psychometric literature is flawed and needs improvement?
P – one can be conscientious and still not try on an IQ test. The only reason to try hard on an IQ test is to prove to yourself how smart you are, and quite a lot of people simply don’t care. Let’s say that I’m very conscientious in day to day things. But then you present me with a test that has nothing to do with anything. Why would you suppose that my effort on the test is related to my actual conscientiousness?
Hahaha, IQ tests are culturally unbiased. This presupposes that language is not cultural, that valuing effort in learning math skills (or education in general) is not cultural, and that reasoning about spatial things has no cultural differences. Can you post a link to the best article that you know of which purports to debunk the cultural bias?
@Nathan W
Your reply to me and to P is essentially question-begging.
We already know that IQ test performance correlates with the good things in life and much of modern life resembles an IQ test. P’s point is that IQ results, for better or worse, are unlikely to be driven by the differences in conscientiousness, which is the story you’re telling above. As far as it is measurable, this is not what’s driving the IQ results.
Nathan, I’m not an expert on psychometrics, still:
1) IQ tests are mixed bags of goods; its possible for an individual to be boosted on one area due to additional experience, but on many, many different ones? That could be possible if he’s someone who systematically seeks out or retains. all kinds of information, maybe
2) this is also why factor analysis is used to distinguish out heavily uncorrelated strengths from the major pattern of correlation (g). (Information is then used iteratatively to redesign the test, increasing g loading while maintaining breadth and correct theoretical statistical properties for test populations – Gaussian distribution).
3) being perfectly culture fair is difficult; still IQ tests are probably fairer and closer to real than your or my subjective impressions. Compared to measurement on general education level, employment or any skill they are probably less influenced by individual experiences, motivations and pressures.
I’m skeptical of the results. If the true model is
Y ~ W * Genes + V * (Genes X SES)
then regressing
Y ~ X * Genes + Y * SES
should still give you a significant t-stat on Y for a sample with large enough discriminatory power. E.g. if SES only mattered for 10% of the population, then you would only need a sample size root(10) times larger than what you’d need to uncover significance if SES mattered for 100%. Yet twin and adoption studies with huge sample sizes have not found consistently significant family environment effects. It’s much more likely that there’s some hidden flaw in the meta-analysis or the data’s simply cherry-picked.
Part of this is certainly over my head, but I will remind that adoptions which seek good families produce a narrow answer. When twins are put in various “at least pretty good” families, environmental effects are small.
I think there is other data on worst case environments, that do indeed produce worst case outcomes.
Therefore, we only need bring every kid’s environment up to pretty good?
I agree which is why increasing education budgets year after year will not turn every kid into an Einstein.
I see what you’re saying for studies involving adoption. But the findings also hold in the literature comparing fraternal and identical twins. Namely all heritable variance is genetically attributable, with family environment being insignificant.
I think researchers are looking for different things, and so use different selection of subjects.
This study is about what happens when you leave very bad environments.
Are typical “environmental” studies reaching that low? Or do they start with a pool of “pretty good” homes as their base,?
Can you be more concrete than pretty good?
I would say that 97% of USA children grow up in a pretty good environment. Some people like to study and work harder at school and at earning money. Whether that is a better way to live of not I cannot say.
In the meta-analysis, the shared environment (C) component is 30% in the US versus 26% elsewhere. So, nowhere near zero. Of course, this reflects the fact that the samples are mainly of children. In adult samples C is usually not significantly different from 0, but the estimates are usually >0. Given that the moderation effect is really visible only at very low and very high levels of SES, it’s not surprising that it doesn’t much affect C when the effects are averaged across all levels of SES.
A big premise of this post is that there are IQ differences between different various ethnic groups which are at least partially genetic in nature – a huge taboo in our current political climate. (My favorite recent example is the outcry against Justice Scalia when he questioned the usefulness of NAM diversity in a physics classroom.)
I’m glad to see that Tyler is dipping his toes into red pill territory, though I wonder how he’ll reconcile his fledgling appreciation for the facts of human biodiversity with his support for untrammeled low-skill immigration. The difference between libertarianism (which accepts the idea of leftist race creationism/egalitarianism and thus sees large-scale low IQ immigration as no big deal) and paleoconservatism largely boils down to the acknowledgement of uncomfortable truths about human nature and group differences.
Actually Scalia’s comment did not differentiate between whether the differences were caused by environment or preparation; they were bothersome in that were expressed in a way that played to stereotypes.
This week’s post on Roland Fryer went over like a lead balloon because it did show that these differences are malleable, and do not confirm such priors.
Even the most hardened hereditarian (I have Steve Sailer in mind when I say this) believes that achievement gaps be be shrunk somewhat if you throw enough cash and specialized interaction at the students.
The question, which Roland Fryer hit on near the end of his talk, is whether or not any *reasonably priced* intervention can scale. The evidence, thus far, suggests that they cannot.
msl December 19, 2015 at 4:43 pm
Even the most hardened hereditarian (I have Steve Sailer in mind when I say this) believes that achievement gaps be be shrunk somewhat if you throw enough cash and specialized interaction at the students.
I am not sure I am even a soft hereditarian, but we have tried this. We have thrown a lot of money and specialized teaching at minorities. And by “we” I mean the West as a whole. America. Britain. The Netherlands. Everyone.
Nothing moves the needle an inch. You can have an impact on young children but it does not last past puberty (suggesting it is not genetic). There is nothing that works. We have tried everything and nothing can shift those outcomes. I mean we have tried everything too. It doesn’t matter what you think up, it has been tried. Moving minorities to White schools. Enriched early education. Black Nationalist-focused class rooms. Ebonics. Everything.
There may be an intervention that works, but no one has managed to discover it. Jamaican can choose any education system they like free of racism and colonialism. They have tried too. Their results are not good.
This is simply false.
1. Please watch the Roland Fryer video for examples of interventions that change outcomes past puberty.
2. http://slatestarcodex.com/blog_images/early_intervention.pdf
3. Differences in lead likely explain significant portions of the achievement (Janet Curie has a project on this at Princeton) and there exists very suggestive time series evidence on this. “Equality” in lead exposure will likely further shrink academic achievement gaps.
Are we even looking for efficient allocation of educational resources now? And to what goal? It certainly looks like more AP classes are a win for future GDP, but I don’t know if longer school hours are better, if anyone is framing it that way.
I suspect resources are more split between pleasing parents and pleasing teachers. Too much IQ reducing football.
“Are we even looking for efficient allocation of educational resources now? And to what goal? It certainly looks like more AP classes are a win for future GDP, but I don’t know if longer school hours are better, if anyone is framing it that way.”
Uh…I am an economist, posting on an economics purportedly about economics…so…yes?
“I suspect resources are more split between pleasing parents and pleasing teachers.”
…OK?
“Too much IQ reducing football.
I agree.
*posting on a blog purportedly about economics.
I guess I am using a different, or more expansive, “we.”
Roland Fryer talked about the resistance to documented improvements, and that certainly fits with many other stories of resistance to change.
And yet you did thumbnail it above as an argument about more money. Possibly sometimes, but that does presume efficient allocation as a starting point.
Get back to us in 100 years when we figure out which genes are responsible for whatever we determine at that time to be “intelligence”.
Saying that you need to sequence specific genes to identify genetic hereditatism of some trait, is like saying you have to measure individual photons to discover the basic principles of optics. For example Mendel was able to model genetic inheritance of many traits to a high degree of accuracy a century before DNA was even discovered.
I agree with what you said.
So I reiterate, get back to us in 100 years if you want to say something concrete about heritability of intelligence. It must exist, but given the existence of cultural differences, existing socioeconomic discrimination etc., it is absurd to think that a one-off test is evidence of systematic racial differences in intelligence.
The tests we’re using to measure intelligence today are essentially the same that we used a hundred years ago. This is because of the complete failure of find credible alternatives to those tests. In 100 years, we may no longer use such tests as more straightforward ways to measure the same thing will have come available.
Slow but steady progress is being identifying genetic variants underlying intelligence differences. We don’t need 100 years. It’s more like 10 years. Here’s the abstract of a recent conference presentation:
Genome-wide association studies (GWAS) have revealed much about the biological pathways responsible for phenotypic variation in many anthropometric traits and diseases. Such studies also have the potential to shed light on the developmental and mechanistic bases of behavioral traits. Toward this end we have undertaken a GWAS of educational attainment (EA), an outcome that shows phenotypic and genetic correlations with cognitive performance, personality traits, and other psychological phenotypes.
We performed a GWAS meta-analysis of ~293,000 individuals, applying a variety of
methods to address quality control and potential confounding. We estimated the genetic correlations of several different traits with EA, in essence by determining whether single nucleotide polymorphisms (SNPs) showing large statistical signals in a GWAS meta-analysis of one trait also tend to show such signals in a meta-analysis of another. We used a variety of bio-informatic tools to shed light on the biological mechanisms giving rise to variation in EA and the mediating traits affecting this outcome.
We identified 74 independent SNPs associated with EA (p < 5E-8). The ability of the
polygenic score to predict within-family differences suggests that very little of this signal is due to confounding. We found that both cognitive performance (0.82) and intracranial volume (0.39) show substantial genetic correlations with EA. Many of the biological pathways significantly enriched by our signals are active in early development, affecting the proliferation of neural progenitors, neuron migration, axonogenesis, dendrite growth, and synaptic communication.
We already have.
Please see work by Pinker et al. available here: http://www.pnas.org/content/111/38/13790.abstract
If you’d like to comment intelligently, please familiarize yourself with the basic results of the field you’re attempting to comment on. This is not a forum for liberal pieties. For that, go to Oprah.com, DailyKos, or DemocraticUnderground or something.
I never said the level of existing knowledge was zero.
You are assuming that I know nothing because you don’t like what I have to say. But I am the one with the ideological filter on.
As devil’s advocate, I suggest that they misidentified the genes as being explanatory for intelligence. Perhaps the identified genes may explain being servile and following orders, which may be useful for things like doing what the teacher says and trying hard on a test when you’re told to. This is a perfectly plausible explanation of the data.
Given the correlation between IQ and creativity/academic output in a variety of settings (some involving rule-following behavior and some not), it is not especially likely that IQ tests are simply measuring obedience/rule-following behavior.
IQ test performance has all the features you would expect it to have if it were actually measuring intelligence. Especially persuasive for me personally was a series of experiments performed in the 1980s in which students were asked to rank the intelligence of their classmates. On average, your ranking by your classmates coincided with the ranking that would be imposed by measured IQ.
In addition, even if that is the case, this is a distinction without a difference: IQ tests forecast much of what is good in life: healthy marriages, high income, high impulse control, and better encounters with the police are all forecasted pretty well by IQ tests at age 14. If you’re suggesting that this is simply rule- or try-hard behavior, then so be it (!), but it is unlikely that this has any meaningful policy implication vis-a-vis personal or public policy.
msl – re: last sentence. That’s exactly the position I’m taking. I am skeptical that it has any meaningful implication vis-a-vis personal or public policy.
Again, rule-following behavior is not a plausible explanation of the stylized facts involving IQ (although I am aware Chomsky made this argument in his review of The Bell Curve, so perhaps you’re merely echoing this).
I believe you have misunderstood me. My point was that your “distinction without a difference” should not impact personal or public policy. Knowledge of IQ and what it means has obvious public policy and personal impact. For instance, people who do better on ASFVAB should get positions with more responsibility/I should hire tutors who do better on the SAT/Colleges should be able to discriminate on the basis of IQ.
None of these conclusions reverses or even changes if the test is only measuring “rule-following behavior”.
I don’t have a subscription to this journal so I can’t read the article but I find the abstract to be bothersome. They state that Western Europe has better healthcare and education but neglect to mention that Western Europe has lots of Western Europeans. Surely the higher IQ of these people matters.
Another point is why not include other phenotypes besides IQ such as running ability. This is even easier to measure than IQ. How have running times changed over the years in different countries based on SES differences?
It might be worthwhile to simplify the methodology to focus upon something certain of measurement, such as height, and focus on a diaspora population in multiple countries, such as Sikhs, to provide more apples to apples comparison.
In the U.S., it’s now not uncommon for young Sikhs to grow quite tall. My son’s Sikh high school friend is about 6’7″.
Open access version: https://drive.google.com/file/d/0B3c4TxciNeJZUE9kQ2NMSnBmVGs/view?pref=2&pli=1
Cross National IQeq according to international scrabble ranking WESPA http://drjamesthompson.blogspot.com.au/2015/12/chanda-chisala-and-african-heredity.html#comment-form
Note only 0.4 IQeq point btw US and Nigeria.
n = 10 PoolAv = 959.424 PoolSd = 381.602 date=20151215
WESPA IQeq Region
2066.5 143.5 United Kingdom
2049.4 142.8 United States
2047.9 142.8 England
2039.3 142.4 Nigeria
2037.2 142.4 Australia
2012.0 141.4 Thailand
1873.2 135.9 Canada
1861.1 135.4 New Zealand
1859.8 135.4 Singapore
1847.8 134.9 Scotland
1839.7 134.6 Kenya
Number of players in the top 100 in WESPA system:
23 Nigeria
17 United States
12 Australia
11 England
10 Thailand
6 Scotland
4 Canada
3 Singapore
2 New Zealand
2 Kenya
What leads you to believe that the US is more meritocratic than Western Europe, Canada or Australia?
It’s an axiom; what sort of punk asks for evidence?
@Nathan
-The U.S. has fewer taxes and transfers.
OK, 200 years. But I think things are moving faster now.