October 24, 2017

Misunderstandings on Race and IQ

This is a response, at the request of a reader, to a piece written by Brink Lindsey, a senior fellow of the Cato Institute, entitled “Why People Keep Misunderstanding the ‘Connection’ Between Race and IQ“. If you would like this site to produce a response to something be sure to let us know via twitter, facebook, or the comments section of this article.

Lindsey starts of by summarizing why hereditarians believe that IQ is an important metric and that most variation in it across individuals is due to genetics:

“Without a doubt, the skills assessed on modern IQ tests are widely applicable and highly valued in contemporary American society. Accordingly, considered just as a measure of skills rather than as a proxy for underlying ability, IQ scores clearly tell us something of genuine importance. They are a reasonably good predictor not only of performance in the classroom but of income, health, and other important life outcomes.

But what about innate mental ability? Does such a thing even exist? Evidence from IQ tests provides strong support that it does. First of all, scores on the various IQ subtests are highly correlated with each other, suggesting the presence of a general underlying factor. Furthermore, IQ scores tend to stabilize around age eight and are resistant to moving around much thereafter, in keeping with a relatively fixed level of innate intellectual capacity. And studies of twins and adoptees offer substantial evidence that this capacity has a strong genetic component. The scores of twins (who are genetically identical, more or less) are much more highly correlated than those of regular siblings (who share only about half the same genes). Meanwhile, the scores of regular siblings are in turn much more highly correlated than the scores of adopted and biological children raised together.”

This is actually a pretty fair summary. One thing to add in the way of specificity is this: most hereditarians think that about 70% of variation in IQ among adults is due to genes. We are thus not “genetic determinists” in the sense that we don’t think that someone’s fate, IQ wise, is totally determined by their genes.

Other than that, adding detail would be superfluous, so let’s move on to his critique of the position.

“These studies typically assume that the similarity of twins’ shared environment is the same as that of regular siblings (highly unlikely) and that adoptive families are as diverse as families generally (in fact, parents that adopt tend to be better off and better educated). When these assumptions are relaxed, environmental factors start to loom larger.”

Let’s look at the argument that twins given up to adoption are so similar, in part, not because of their genetic similarly as adoption studies assume, but, rather, because adoptive homes are basically all upper middle class and, therefore, more similar than average, and this abnormally similar home environment is part of what makes twins raised separately so much alike.

The only study to ever directly compare adoptive and non-adoptive families from the same sample found that, yes, adoptive homes are better and more similar than average in terms of income, parenting styles, and parental mental health, but statistically correcting for this didn’t change the IQ heritability figures one iota because these variables were also seen to not effect IQ in this adoptive sample (McGue et al., 2007).

More importantly, there are studies which estimate heritability without using adoption and which have either used the military or the national school system to measure the IQ of everyone (or, for the military, every male) in an entire country, and which therefore include a sample which is nationally representative with respect to socio-economic status. These studies produce the heritability figures which are, if anything, higher that what is most often found in the literature on the heritability of IQ (Sundet et al., 1988Benyamin et al., 2005).

Of course, this response is only effective if non adoption based twin studies are valid. Lindsey argued that there were not by advancing the notion that identical twins are more similar than non twin siblings and fraternal twins because they are treated more similarly.

This argument was first empirically addressed by Loehlin and Nichols (1976) who looked at data on over 850 pairs of twins. Loehlin and Nichols measured the degree to which parents treated these twins the same way, the degree to which they were dressed alike, whether they had been put into the same classes, whether they slept in the same room, etc. They then measured the correlation between how similarly the twins were treated by their parents and how similar they were with respect to their  IQ scores. They found that increased similarity of treatment predicted almost no increased similarity in IQ.

More evidence on this question comes from studies of misperceived zygosity. “Misperceived zygosity” refers to a situation in which an MZ twin pair is thought to be a DZ twin pair or vice versa. Then, often when these twins enter into a behavioral genetics study, they find out that they are actually the other kind of twin pair. Until this reveal, these sets of DZ twins were treated as MZ twins by everyone they knew, and the MZ twins were treated as DZ twins by everyone they knew. So, if this critique of twin studies is valid, we would predict that mis-classified MZ twins will be less alike in IQ than correctly classified MZ twins and mis-classified DZ twins will be more alike in IQ than correctly classified DZ twins. However, several studies have shown that this is not the case (Scarr and Saltzman, 1979Conley et al., 2013) .

In sum then, MZ twins may be treated more similarly than average, but it doesn’t explain much about why their IQs are so similar.

Beginning just a few years ago, sequencing people’s genomes became cheap enough that researchers could afford to bypass all the concerns about twin and adoption studies by just measuring the genetic similarity of pairs of unrelated individuals and seeing how well that predicted IQ similarity. This technique is called genome wide complex trait analysis (GCTA). These newer studies largely confirm the findings of early twin and adoption based studies.

Before reviewing this literature, it is important to note that GCTA gives us a lower bound estimate of the heritability of a trait. That is, it tells us an estimate of the lowest possible value of heritability. This is for three reasons.

First, in GCTA not all of the genome is measured. GCTA will only detect that proportion of IQ variation which is explained by the parts of the genomes that researchers measure (or which are correlated with parts of the genomes they measure). Since the whole genome is not measured, IQ variation caused by unmeasured parts of the genome will not be detectable by GCTA.

Secondly, IQ variation that is explained by “rare variants” may not get picked up by GCTA. A rare gene variant is one that exists at a very low frequency in the population. Because of their inherent rarity, a rare gene variant which impacts IQ may simply not exist in the sample of a given study.

Thirdly, unlike MZ twins, people who share a little more DNA than average are unlikely to share combinations of genes. This is important because sometimes the effect of a gene on a trait is dependent upon which other genes are in a person’s genome. To the extent that this is true of IQ, GCTA will under-estimate IQ’s true heritability.

With these caveats in mind, here is a chart, taken from this article, summarizing the research in this area:

10

As can be seen, in most studies that compute a classic heritable estimate and a GCTA based one, the later accounts for the vast majority of the former. It’s worth noting that these studies are mainly done on children, which is why the heritablity estimates are, in an absolute sense, low. The heritability of IQ rises with age. If the sample were older, the heritability estimates would be higher.

Thus, the mainstream science on the heritability of IQ, which pegs the figure at around .7 in adults, withstands the criticisms offered by Lindsey.

Moving on:

“consider a pair of French adoption studies that controlled for the socioeconomic status of birth and adoptive parents. They found that being raised by high-SES (socioeconomic status) parents led to an IQ boost of between 12 and 16 points – a huge improvement that testifies to the powerful influence that upbringing can have.”

This is true, adoption does have an impact on IQ. This particular finding is a bit extreme and based on a small sample size, but the general effect is real. (I cover this literature in more detail here.) However, this effect is consistent with a large heritability figure for IQ. The heritability of IQ refers to the amount of variation in IQ within a given population explained by genetic, as opposed to environmental, differences between people. Pointing out that going from the worst environment to the best environment changes IQ a lot does not imply that this explains much about why most people differ in terms of IQ. In fact, it doesn’t. Most people aren’t extremely poor or extremely rich.

To make an analogy, suppose someone pointed out that taking kids who were born into homes with illiterate parents and placing them into homes with literary professors as parents raised their reading ability dramatically and, from this, concluded that most reading variation must be due to parent quality. Would this convince you? Of course not. For the same reason, adoption studies where poor kids and placed into rich homes tell us little to nothing about what proportion of IQ variation is caused by genetics or SES. To do that, we need heritability studies that have large and nationally representative samples.

As has already been covered, studies with nationally representative samples show that IQ is highly heritable. Thus, this argument does not stand up to scrutiny either.

“A study of twins by psychologist Eric Turkheimer and colleagues that similarly tracked parents’ education, occupation, and income yielded especially striking results. Specifically, they found that the “heritability” of IQ – the degree to which IQ variations can be explained by genes – varies dramatically by socioeconomic class. Heritability among high-SES (socioeconomic status) kids was 0.72; in other words, genetic factors accounted for 72 percent of the variations in IQ, while shared environment accounted for only 15 percent. For low-SES kids, on the other hand, the relative influence of genes and environment was inverted: Estimated heritability was only 0.10, while shared environment explained 58 percent of IQ variations.”

Turkheimer’s study is not representative of the literature in general. As a recent meta-analysis showed, most studies looking for an interaction between IQ heritability and SES doesn’t find one. That being said, most studies in America, and only America, do. But the mean effect is not as powerful as the one found in Turkheimer’s study.

MEta-analysis graph.JPG

Tucker-Drobs and Bates (2015)

As can be seen, the heritability of IQ is around 25%, not 10%, even in very poor homes. At no point does shared environment account for 50% or more of IQ variation.

Another problem with these studies is that most of them are on children. It is well known that the heritability of IQ rises with age, so it is not clear what general heritability figure we should compare this to. Given this, the data may actually be consistent with the heritability of IQ being more than 50%, in adulthood, even among the poor and even within America.

And again, most people are not poor and nationally representative samples show a high heritability for IQ.

“Comparisons of IQ scores across ethnic groups, cultures, countries, or time periods founder on this basic problem: The cognitive skills that IQ tests assess are not used or valued to the same extent in all times and places.”

This is the most disappointing part of the article. As reviewed in this article, nearly all experts in intelligence research agree that IQ tests are not biased against non Whites, IQ tests predict life outcomes equally well for all races, and the hardest IQ problems for Whites and non-Whites are the same. IQ tests measure intelligence equally well for all races.

“The mass development of highly abstract thinking skills represents a cultural adaptation to the mind-boggling complexity of modern technological society. But the complexity of contemporary life is not evenly distributed, and neither is the demand for written language fluency or analytical dexterity. Such skills are used more intensively in the most advanced economies than they are in the rest of the world.”

Even in poor nations such as India and China, areas that are smarter are also richer:

1.PNG

Lynn and Yadav (2015)

So it looks like IQ is valuable in unadvanced nations too.

“Among the strongest evidence that IQ tests are testing not just innate ability, but the extent to which that innate ability has been put to work developing specific skills, is the remarkable “Flynn effect”: In the United States and many other countries, raw IQ scores have been rising about three points a decade. This rise is far too rapid to have a genetic cause. The best explanation for what’s going on is that increasing social complexity is expanding the use of the cognitive skills in question – and thus improving the opportunities for honing those skills. The Flynn effect is acutely embarrassing to those who leap from IQ score differences to claims of genetic differences in intelligence.”

The Flynn Effect has no bearing on the heritability of IQ. .

Consider a plant garden. If every plant is exposed to the same amount of sunlight, grown in identical soil, and given the same amount of water, height differences are going to be almost entirely due to genes. Thus, the heritability of height might be something like .95. However, if we introduce a new fertilized the mean height might rise by a considerable degree. This will not change the fact that basically all height differences within the garden are due to genes.

Similarly, the fact that the environment has changed in a way that raised the population’s IQ does not imply that the heritability of IQ is not high.

It’s worth noting that over the time that the Flynn effect has occurred the racial IQ gap has basically remained the same (Chuck 2013).

This would be like if, in the previously referenced garden, one type of plant was consistently taller than another type before and after the new fertilizer. This is what has happened with race, IQ, and the flynn effect, and somehow egalitarians think that this is evidence that the gap is not caused by genes!

Finally, I’d like to note that there is a lot of evidence on race and IQ that Lindesy’s piece did not mention:

  • The more white a non white person is the higher their IQ tends to be
  • The more white the mean person in an area is the smarter, and richer, the mean person in that area is
  • The more heritable a subtest is the bigger the racial gap on it tends to be
  • Racial IQ gaps are present even among infants.
  • Consistent racial gaps are found all over the world.
  • Racial differences in brain size are present at birth, have been constant for a long time, and mirror racial differences in intelligence

This evidence, and more, can be found here.

There are many positive things that can be said about Lindsey’s piece. It is calm and fair in its tone, a rarity in this subject matter, and it makes objections which are plausible at first glance. However, I think a closer look at the evidence reveals that none of them offer us sufficient reason to reject the hereditarian view of race and IQ.

Facebook Comments
  • Mark Martinson

    “Comparisons of IQ scores across ethnic groups, cultures, countries, or
    time periods founder on this basic problem: The cognitive skills that IQ
    tests assess are not used or valued to the same extent in all times and
    places.”

    I don’t think he’s saying they are culturally biased in the way we normally use that term. He’s saying that the skills IQ tests for are those skills needed in the western world. So, for example, an IQ test for Boliviian farmers or Esquimaux might focus on skills needed in their culture. Perhaps an IQ test for Esquimaux would use questions indicating whether a person could navigate the unusual topography.

    Even so, I don’t think the point is that significant. I assume and Esquimaux with an IQ of 100 has an easier time of it than one with 90.

    • Why Tea

      This type of argument is usually presented by those who don’t understand the body of research around “g”, and who aren’t familiar with culture-fair tests lik Raven’s Progressive Matrices.
      This evasion comes up again and again as speculation, with no empirical data or concrete findings to support it.
      The converse of their assertion would *likely* be true–that these members of other cultures would be more cognitively adept at some other set of tasks than someone from our culture, when both of the subjects were trained on how to perform those tasks that were “more important” to the culture in question.
      That’s not what we find. If a high IQ subject is shown how to solve problems or perform tasks that are “important” in this other culture they perform better than someone from that culture with a lower IQ when measured by our best culture-fair tests.
      More simply stated:
      The idea that somehow my ability to operate a fucking iPhone or fill out a department of transportation form makes me better able to line up shapes dots and lines and predict the next figure in a sequence when compared to someone from a culture that uses primarily manual labor and hunting in the bush and then finding their way home is pretty silly and without strong experimental evidence that would somehow contradict decades of effort in making such tests “culture-neutral”. Those sorts of assertions can be dismissed as the speculations of someone uninformed about the state of the art in the field of psychometrics.
      https://en.wikipedia.org/wiki/Raven's_Progressive_Matrices

  • Mark Martinson

    “Among the strongest evidence that IQ tests are testing not just innate ability, but the extent to which that innate ability has been put to work developing specific skills, is the remarkable “Flynn effect”: In the United States and many other countries, raw IQ scores have been rising about three points a decade. This rise is far too rapid to have a
    genetic cause. The best explanation for what’s going on is that increasing social complexity is expanding the use of the cognitive skills in question – and thus improving the opportunities for honing those skills. The Flynn effect is acutely embarrassing to those who leap from IQ score differences to claims of genetic differences in
    intelligence.”

    But the rise cannot be entirely an increase in intelligence as well. We aren’t geniuses and our grandparents weren’t morons. Actually, the fact that the gap stays the same between blacks and whites is devestating for environmentalists. You would have expected blacks to “catch up” as their situation improved.

    • blackacidlizzard

      It’s almost like egalitarians don’t care about evidence, only about making assertions. It’s almost like they follow Saul Alinsky’s advice: “say whatever sounds good, even if it doesn’t make sense.”

      • Why Tea

        “Almost.”

    • Ghost of Virtve

      This analysis indicates that intelligence has actually been decreasing since Victorian times: https://media.8ch.net/pdf/src/1437855837967.pdf

      These [Flynn Effect] gains are thought to be due almost entirely to environmental improvements stemming from factors such as improved education, nutrition, hygiene, and exposure to cognitive complexity (Neisser, 1997)

      • MUltan

        The samples weren’t random, the testing methods differed, and there is no overall trend in the data. Further, such big differences in the means (~200ms vs 300ms 19th c. vs 21st. c.) would imply not just big differences in mean IQs, but over an order of magnitude decrease in IQs over 140, which hasn’t happened.

    • Xor

      Except the gap has actually narrowed over time:
      http://phys.org/news/2006-09-black-white-iq-gap-narrowing.html

  • Mark Martinson

    Has anyone been following Karlsson’s responses? They are pretty lame. Basically: (1) correlation isn’t causation (no I’m not joking); and (2) because there could be some environmental explanation for differences there must be.

    • I’ve seen them. I will probably create another response to him.

  • Mark Martinson

    Does anyone have an explanation for the high IQ scores of the Bahamas? It’s been reported to be in the 90-93 range which would seem inconsistent with hereditarianism given that the nation is mostly black and the race admixture is the less than black Americans.

    • Ryan Faulk

      If their median IQ is 93, and the Bahamas have an IQ-relevant environment similar to the United States, well, what’s the problem?

      They would then be black people with an IQ of 93. They exist.

      • Mark Martinson

        Then their IQ is higher than Black americans by a fair amount. I don’t see any environmental changes that could increase black IQ from 85 to 93.

        • Ryan Faulk

          I agree, there is genetic variance within races. I’m glad we have settled this non-issue.

    • Jeff
  • max

    Here is Lindsay on the Caribbean Blacks:

    https://robertlindsay.wordpress.com/2016/04/05/the-flynn-effect-in-blacks-is-ongoing/

    He seems to think that the Flynn Affect is raising Black IQs relative to whites. But of course he is a Marxist after all.

    • Jeff

      He seems to have gotten that from this:

      http://humanvarieties.org/2013/03/12/hvgiq-the-bahamas/

      It’s mostly done on children who have malleable IQs. It was weighted upward by white students who took up most of the top schools and the scores weren’t separated by race. This is far from proving that blacks in the Bahamas have an IQ of 90-93.

  • Mark Martinson

    You mentioned regression to the mean in your most recent video. I have always found this perhaps the strongest evidence for race realism. I’ve never heard a good explanation of this by an egalitarian.