The Genes Debate: Assortative Mating and Embryo Selection

Came across these two interesting articles, which cover the same subject matter, so I have lumped them together in a single analysis.

Why Resist Blank Slate Thinking? For One, Look to No Child Left Behind

Genes Believe in You

He’s correct regarding how biological determinism calls into question the prevailing orthodoxy of the blank slate. Kids who are ‘slow’ in elementary school tend to remain slow their whole lives, and that consequently efforts to raise scholastic achievement have failed, due to the inherent and immutable limitations imposed by biology. There should be nothing controversial about this, only that the media, in a collective willful denial of reality, have decided that genes do not matter.

The blank slate is not just limited to the left: many mainstream conservatives, such as on talk radio or in op-ed pieces, continue to cling to this antiquated mode of thinking in which poverty and wealth inequality are downstream from effort and laziness, when ‘bad genes’ that give rise to low IQs may also be to blame. America’s persistently low labor force participation rate, its cause, and implications, will be discussed in a forthcoming article.

I think however Fredie overstates the HBD case in some instances, for example he writes:

Within 50 years, perhaps within 30, rich parents will routinely pay to have children whose genomes have been manipulated or selected for higher intelligence and other attractive qualities. I do not know what specific technologies will enable this to happen, but I do know that it will happen. And when the monetary elite uses genetic science to further strengthen the unearned dynastic advantages of their progeny, locking in the privileges that they already enjoy and pass down through inheritance, DNA, and our rotten system … what will the people attacking Paige have to say about it? What arguments will they be able to muster, against genetic engineering for those who can afford it, after decades of denying that genes matter in human behavior at all? What smarmy little jokes will the liberal gene denialists tell then? Saying “eugenics” won’t ward off that future. Saying “Gattaca” won’t ward off that future. Saying “Charles Murray” won’t ward off that future. Nothing can prevent a future in which our technological capacity to manipulate the genome has ever-increasing social consequences, almost certainly very bad ones.

In theory this should happen, but progress has been excruciatingly slow (as in non-existent as far as IQ is concerned). Some of this is due to political resistance, but also due to the inherent complexity of traits such as IQ. Rather creating smart babies, much of human genomic research seems to be focused on screening for certain diseases, raising life expectancy, and creating custom therapies.

There are a couple issues one runs into:

1. A trait can be innate but not fully heritable or not heritable at all. An obvious example is Down’s Syndrome. So in utero development may contribute to IQ as well. Parents may contribute 50-80% of offspring’s IQ (based on twins studies) and the rest occurs in utero as the brain develops.

2. The polygenic nature of IQ is probably the biggest barrier to the viability of embryo selection for IQ. Imagine there are thousands of genes that play a role in IQ, analogous to a giant switchboard. The default state corresponding to an IQ of 100 is half the switches ‘activated’ and the other half turned off. Each switch activated may correspond to an extra few fractions of a point of IQ, and the probability of having an excess number of on or off switches decreases exponentially, hence the bell curve distribution. So there is inherent randomness, and one is born with a certain number of switches flipped up or down. Indeed, such randomness is how it’s occasionally possible for average IQ parents to have genius-IQ children.

3. The geonotype does not always equal the desired phenotype. For the geonotype to result in the desired phenotype , it requires that the genes be expressed. It also requires that counterproductive factors be suppressed. So having genes predictive of endurance ability (like running, if this is the desired phenotype) would not help if paired with genes which give rise to obesity. Having a high IQ does not help much if coupled with genes that give rise to poor work ethic or mental illness.

Regarding assortative mating, he writes:

Also, to return to Quiggin’s tweet, we are already changing the gene pool. Assortative mating, which has massively increased in recent decades, is among other things an effort in genetic engineering. Mate selection among humans is a very complicated thing, but there’s no doubt that we are in part selecting for reproductive fitness, broadly defined. If someone decides that they want to partner up with someone else because that person will help provide financial stability – a very common concern in marriage and a perfectly legitimate one – that person is, to some degree, selecting based on genes. Physical attraction is also, among other things, related to our perceptions of the desirability of the genes that potential partner might pass on to our children. But of course it is; we are the products of evolution, and evolution forces us to want to produce offspring who are more likely to produce lots of offspring. Those professional class liberals who are delaying marriage and kids until later and later in life are practicing excruciatingly exacting mate selection, looking for just the right person to make some babies with. That is genetic engineering; the fact that it’s the polite kind does not change the fact that, if such trends continue, on a long enough timescale we will have a rigidly stratified species based on genetic parentage. I do not need to share the extremely durable research showing that more highly-educated parents have more highly-educated children, which has serious consequences even if you suppose that influence is entirely environmental. If it’s even partially genetic, the consequences are civilization-altering. But how can we think through that condition if we must pretend genes and behavior are totally disconnected?

I think he also overestimates the efficacy of assortative mating. As discussed above, the genotype does not equal the desired phenotype. Smart parents not uncommonly have mediocre children. Look at the Kennedy family, as an example of how even under the most optimal of conditions that individual outcomes can vary enormously. The Kennedys and Rockefellers were major forces in US politics in the 20th century; not so much anymore, as the bloodline became diluted and mediocrity set in. Robert F. Kennedy, Jr. is probably the only remaining Kennedy who still somewhat relevant in the public today. The last living Rockefeller who was involved in politics is Jay Rockefeller, aged 84.

However, raising the national mean IQ by even just a few points can make a huge difference at the tail-end of ability. So if through aggressive embryonic selection and assortative mating, over many generations, it may be possible to raise the biological mean IQ of humans from 100 to 105 or so, which may not seem like much, but it means that the predicted number of people with IQs above 150 is tripled [1]. However, this is a long way from creating designer babies…It’s more like a rising tide. Then the question also is, do we want more smart people. Having more Elon Musks also means more Jack Dorseys , Mark Zuckerbergs, or critical race theorists.

[1] original calculation with a mean IQ of 100

Increaing the mean IQ to 105


  1. One factor not discussed is age of mother. For example, down’s syndrome is 40 per 100k births at age 20 and an incredible 855 per 100k births at age 40. So that 20yo mother’s children might be doing a lot better in terms of genetic fitness than the higher iq mother having a child at 35, assertive mating and societal expectation notwithstanding.

    1. As a matter of fact. Men can afford to be older. Women cannot without damage to their eggs and the DNA within the eggs.

      Whilst mutations do occur within spermatozaa. The female reproductive system is designed to filter out defective sperm with defective DNA.

Leave a comment

Your email address will not be published. Required fields are marked *