The new movement of geno-economics

By Jacob Ward at The New York Times.  Do read the whole thing, here is just one small bit:

The geno-economists seem confident that human genes have a measurable influence on human outcomes. But publicizing whatever predictive power does lie in our genes runs the risk of misleading the rest of us into believing that control of our genes is control of our future. They’re adamant that their motives are in forestalling the dystopian implications of the work, in fighting off misinformation and misguided policies. “The world in which we can predict all sorts of things about the future based on saliva samples — personality traits, cognitive abilities, life outcomes — is happening in the next five years,” Benjamin says. “Now is the time to prepare for that.”

Via Garett Jones.


who owns the data?

Henrietta Lacks? It's in the 13th Amendment of the US Constitution.

Bonus trivia: There's an rich man's inheritance gene, Gregory Clark says so!

As you can imagine, this will be a fairly repetitious process.
There are thousands of such requirement you possess in enterprise.
Perform not have any benefit from doing it all.

The items possess been been mentioned are about the beginning.
So why let them have essential over someone? Most sites have about four basic pages:
Index, About Us, Contact us, and what We Offer.

At some point we will learn the correlation (and uncertainty) between our personal genome and our .. potentials (plural for multiple axes).

I am not sure that is economics, but then a lot of things have to be hammered into departmental divisions that were established at universities a century ago.

I can already hear the screams of, "RACISM".

Going to be a fun ride.

No, individual genome is something else.

You might hear rebellion aagainst "determinism!"

Including from people who used to think their "race" protected them from scrutiny.

Whose race protected them from scrutiny? And why the scare quotes around race?

It is kind of the game of simple minded racism (think prison yard) that having the right skin gives you the full set of right genes.

That was always false, but how can it ever stand up in the age of personal genomics?

If you rely on skin tone that might just mean you are afraid to test.

Welcome to the new world, where skin tone and "race" is obsolete. Your genetic inheritance is your actual personal genome (and perhaps a little family microbiome).

Perhaps the Buppies of the future (picture a mass on DonGlovers) will stand amazed by their genomic educational scores, and declare both their fealty to the White Middle class (that great mass of persons with matching EduAtt scores) and the obselecence of race. However, privately, I doubt this.

Well, anybody who prefaces "middle class" with any color might have to get out of the way first.

| “The world in which we can predict all sorts of things about the future ... is happening in the next five years,” |

yeah, right. Yesterday, our very hi-tech meteorologists in New York City could not even correctly forecast the local weather 8 hours in the future.-- very low tech Geno-Economists are a century away from any predictive ability.

Yeah, but meteorologists can predict the climate a century from now. Easy-peasy.

Jacob Ward shorter: "those geno-economists are doing incredible science, revolutionary, it will change the world in five years. But it's too dangerous to let you know what it is about."

As is frequent now in NYT articles, a long read with a lot of anecdotes, a lot of Ethics/Politics level 0 discussion, and almost no science. It would have been good to see at least one example of quantitative probabilistic prediction tested on a control group to see how accurate it is. But I suppose that would be too dangerous.

"The authors calculated, for instance, that those in the top fifth of polygenic scores had a 57 percent chance of earning a four-year degree, while those in the bottom fifth had a 12 percent chance." Okay but since no two individuals (except twins) have the same genome, with the data they have it is trivial to construct a predictor based on the genome only
that tell you exactly, at 100% accuracy, which of the individuals has a four year degrees. The interesting question is how my perfect predictor, or their, perform on a group different that the one used to construct it. On this the article says nothing.

I haven't read the paper, but the natural way to deal with this would be to separate out training data (what you use to build up your model) from test data (what you use to test it)

suicide eludes hypocrisy. pain justifies life. some concrete things do defy existence. a mermaid, a cataract, a violin, a policy.

You can get a really good read on someone's intelligence by giving them a paper and pencil IQ test in under an hour. That will always give you more information about their intellectual abilities than a genetic test, in the same way that if you want to find out if I'm tall, you're better off measuring my height than looking for alleles positively correlated with height in my DNA test.

Most employers don't give IQ tests for some combination of social and legal reasons. I assume the same reasons would apply to genetic testing.

IQ test were originally developed to gauge people's abilities so human resources could be used more effectively. But their main purpose now is for use in online arguments over how much another test involving squamous epithelial cells predicts the results of the first test.

1. The main issue with stopping or slowing a scientific breakthrough is that it's unsustainable. It only takes one company or country to proceed with research for bans and policy hold-ups to break down. And there are strong incentives to deviate – just ask parents.

2. Many factors will determine which country becomes the leader in genetics, most scientific and financial, but America's multiracial society poses cultural problems that China just won't have.

China has no qualms about all the life science research they are pouring into genetics, biotech, cloning, stem cells, and other related tech that gives the US pause. That coincidentally has nothing to do with multiracialism and everything to do with the Judeo-Christian worldview that seeks to limit man's ability to "play God". China's state can create an army of genetically enhanced clones to do their bidding but such a thought would be "evil" in the US.

Cooke, it IS about race. Cloning, gene-editing, etc is seen as problematic in white countries because of the millenias-long history of white male supremacy over the human diaspora. Allowing white males to edit their genes will only further perpetuate white male supremacy over minorities like Women People, Black Men, LGBTQPN+ People, People Of Islam, etc. Imagine a reality where white males are allowed to edit their genes to be as good at basketball as Black Men are...shudder. Do you now see the problem?

"the millenias-long history of white male supremacy over the human diaspora." What a peculiar version of history.

I am amazed there are two Alistair/Alastairs here. We should check who has seniority on the name.

Let’s call it the "New York Times View of the World". It encompasses a range of views but all of them are **outside** the solid findings of behavioral genetics. You don’t have to go to polygenic scores for this, 50 years of twin studies gives you the result (simple version):
50% of a typical trait is genetic (IQ 80% in later adulthood)
About 0% is due to the shared upbringing
50% is due to the "non-shared environment" aka, no one knows what affects this

The NYTVW just doesn’t like it. It’s obvious why. Religion meets science in the 21st century.

Smart people in genetics, who probably share many of the sentiments of the NYTVW are trying to gently explain that if you keep pretending, the results in society will be worse.

Earnest writers for the NYT explain why we should keep pretending. The noble lie. Well, most of their writers probably believe the noble lie.

Can someone explain to me what shared environment means in this context? I’ve googled it but it’s confusing to me because clearly if 2 kids in a study have an environmental component that is the same, it’s intuitive that that component is not going to lead to different results. I suspect this is not what is meant by “about 0% is due to shared environment”, but I don’t know what is meant.

Is it that if I take 2 twins and separate them at birth, their different parents won’t affect outcomes? “Shared environment” seems like an odd way to describe that.

Sorry for the novice question - thanks in advance for clarifying.


"..because clearly if 2 kids in a study have an environmental component that is the same, it’s intuitive that that component is not going to lead to different results. I suspect this is not what is meant by “about 0% is due to shared environment”, but I don’t know what is meant."

Most people believe that the reason kids in the family have similarities in personality traits is due to upbringing, aka nurture, aka the shared environment. If the parents are extraverts it teaches the kids to be extrovert (using extroversion, one of the big five personality traits, as an example).

Twin studies, adoption studies and all related studies shows this to be false.

The reason kids have similarities in extroversion with their parents is genetic.

Take adoption studies - put an adopted child into a family when it is a few days or few months old. When that child grows up and leaves home how similar is it to the adoptive parents in extraversion? Zero. Basically the adopted child is as similar to the parents and adopted siblings as it is to any random stranger on the street. But it is similar to its biological mother (usually father's data is missing in these adopted studies).

Comparing identical twins with fraternal twins, comparing half siblings with siblings, comparing adopted children with biological children.. all of these family studies show that the "shared upbringing" has about zero effect on personality traits and IQ.

This is very surprising. Also unknown to most people.

I was thinking in terms of causes of differences rather than causes of similarities. So the idea that “shared upbringing is responsible for 0% of trait differences” was...obvious to the point of irrelevance.

It makes more sense when thinking in terms of trait similarities.

But now “non-shared environment explains 50% of trait similarities” seems nonsensical. Where am I going wrong here? I know I’m just misunderstanding the semantics but I can’t figure out a way it could make sense.

I grew up with my parents and you grew up with yours. There was no shared upbringing. But I get the feeling that my parents’ “nurture” doesn’t fall under non-shared environment either.


"..“shared upbringing is responsible for 0% of trait differences” was...obvious to the point of irrelevance."

Shared upbringing is responsible for 0% of trait *similarities* is what I tried to say.

What makes the biological children *similar* to their parents and to each other? Genetics.

What makes the biological children *different* from their parents and from each other? Random stuff that we don't understand.

What impact does the shared environment make (=living in the same household, going to the same church (or to no church), hearing about politics, watching tv, having lots of books around (or not)) ? None that can be detected (except in adolescence, but disappear by early adulthood).

Sorry if I've misunderstood your question.

So you're saying that knowing the genome of some baby, you can predict with 50% accuracy whether he/she will be homosexual? (Or is "homosexuality" not a "typical trait"?) Is there empirical evidence to support the claim that one can make this kind of predictions with any reasonable accuracy?

(And I think you are mixing up two different things -- that some trait is determined, for a certain part, by genetic data, and that we currently have, or will have very soon, the ability to predict that trait based on the genetic data. We can agree with the first and be very skeptical of the second.)


"So you're saying that knowing the genome of some baby, you can predict with 50% accuracy whether he/she will be homosexual? (Or is "homosexuality" not a "typical trait"?) Is there empirical evidence to support the claim that one can make this kind of predictions with any reasonable accuracy?"

No, I'm not.

Giving simple answers can lead to confusion. I recommend the textbook "Behavioral Genetics", by Knopik et al now in its 7th edition. Or Robert Plomin's new book "Blueprint" (he is also co-author of the textbook I just cited, and was the lead author for the first 6 editions). Getting a decent understanding of a complex field takes some work and can't be adequately conveyed in a paragraph or two.

Three short comments:

1. Strictly speaking, about 50% of the variance of a trait, in a given population, is explained by genetics. This doesn't mean that 50% of the explanation of one individual is necessarily genetic.

2. The big 5 personality traits (openness, conscientiousness, extroversion, agreeableness, neuroticism) are about 50% heritable (50% of the variance in the population is explained by genetics), other traits have different amounts. IQ is about 80% heritable in later adulthood. Schizophrenia is a lot less, I forget the amount. I don't know the value for homosexuality.

3. Right now the polygenic scores (as they are called) don't predict the traits particularly well. This is called the "heritability gap" - the difference between the amount known from twin/sibling/adoption studies and the value that can be calculated from GWAS (genome wide association studies).
As an aside - I've seen many detractors of behavioral genetics point to the "heritability gap" with a "you see, this shows it is massively exaggerated" but never with the relevant comment that height also has a big heritability gap from polygenic scores (is height not heritable?). Possibly Stephen Hsu's new paper has closed that gap on height but the issue is that many many points on the genome each have a very small effect. Therefore, very very large sample sizes are needed. Hsu's paper used nearly a million I think (going from memory).

Hope these comments were useful. I do recommend Plomin's book, it is very readable and up to date (published this year).

Comments for this post are closed