Latest / Razib Khan's Unsupervised Learning / Gregory Clark and Garett Jones: To Have and Have Not, genetic determinants of social status
Transcript
- 0:00SPEAKER_00: This podcast is brought to you by the Albany Public Library Main Branch
- 0:03SPEAKER_00: and the generosity of listeners like you.
- 0:09SPEAKER_00: What is a podcast?
- 0:16SPEAKER_00: God, Daddy, these people talk as much
- 0:20SPEAKER_00: as you do.
- 0:26SPEAKER_00: Razib Khan's Unsupervised Learning.
- 0:29SPEAKER_02: Hey, everybody.
- 0:29SPEAKER_02: I'm here today with Greg Clark and Garrett Jones, who have both been on this podcast.
- 0:36SPEAKER_02: And I think you guys both are familiar to the audience.
- 0:39SPEAKER_02: But just in case, let's have you guys introduce yourselves. Greg, you start.
- 0:42SPEAKER_02: And then, Garrett, you go with all your various academic affiliations
- 0:45SPEAKER_02: and what you're working on.
- 0:46SPEAKER_02: Mostly, we're here to talk about Greg's book.
- 0:49SPEAKER_02: Do what you want to do and say what you want to say.
- 0:51SPEAKER_03: Hi, so I'm Greg Clark.
- 0:53SPEAKER_03: I'm a professor at the University of Southern Denmark and also an affiliate
- 0:59SPEAKER_03: of the Danish Institute for Advanced Studies.
- 1:02SPEAKER_03: And I'm also an emeritus distinguished professor at the University of California at Davis.
- 1:07SPEAKER_03: And I'm also a visiting professor at LSE.
- 1:10SPEAKER_03: And my recent work almost exclusively has been on this tome Which is to have
- 1:17SPEAKER_03: and have not, which is an exploration of the determinants of social status
- 1:21SPEAKER_03: in Britain over the past 400 years.
- 1:25SPEAKER_01: And I'm Garrett Jones.
- 1:26SPEAKER_01: I'm a professor of economics at George Mason University.
- 1:30SPEAKER_01: I am also currently a visiting fellow at the Hoover Institution
- 1:34SPEAKER_01: at Stanford University, and I am the author of the Singapore Trilogy of Books
- 1:38SPEAKER_01: with Stanford University Press, some of which is about IQ,
- 1:42SPEAKER_01: some of which is about cultural effects of migration,
- 1:45SPEAKER_01: and some of which is about how democracy is a little overrated.
- 1:49SPEAKER_02: Yeah, so I wanted to have Garrett on because I had a little economist view,
- 1:56SPEAKER_02: maybe some critiques, but I think friendly critiques.
- 1:58SPEAKER_02: I think you guys both get along like each other.
- 2:01SPEAKER_02: And I'm interested in this topic, as from a previous podcast with Greg,
- 2:06SPEAKER_02: as listeners will know, if That was a while ago, though,
- 2:08SPEAKER_02: so a lot of listeners are new about the connection of social status to genetics
- 2:14SPEAKER_02: and heritability and whatnot.
- 2:16SPEAKER_02: I guess first we should probably talk a little bit
- 2:18SPEAKER_02: about there's some drama associated with this manuscript.
- 2:22SPEAKER_02: You haven't been able to find a publisher, an academic publisher,
- 2:26SPEAKER_02: which is frankly a little weird because there's a lot of obscure books published.
- 2:32SPEAKER_02: I don't know how to say this, but you know what I'm saying.
- 2:34SPEAKER_02: There's a lot of obscure books on abstruse things that are published.
- 2:38SPEAKER_02: Academic publishers are not like a popular press where they're taking a risk
- 2:43SPEAKER_02: that maybe you'll be one of those one out of 10 breakout hits that...
- 2:48SPEAKER_02: They just assume that there's going to be your colleagues and maybe
- 2:51SPEAKER_02: a few other people reading it.
- 2:52SPEAKER_02: And then a few cases like, say, The Bell Curve or something like that.
- 2:57SPEAKER_02: Or maybe like Farewell to Alms, Greg's first major book that I think a lot of people know about.
- 3:02SPEAKER_02: It'll break out. But usually they're like, okay, as long as it's not crazy,
- 3:07SPEAKER_02: we'll publish it if you're an academic.
- 3:09SPEAKER_02: So what happened here?
- 3:12SPEAKER_03: Well, I was hoping my publisher, Princeton, would do the book.
- 3:18SPEAKER_03: The core article had been published in Proceedings of the National Academy of Sciences.
- 3:23SPEAKER_03: I thought there was some validation,
- 3:26SPEAKER_03: but it turns out I'm an outsider to kind of behavioral genetics,
- 3:32SPEAKER_03: and I just have not realized how strong there's this moral feeling
- 3:39SPEAKER_03: that any attempt
- 3:40SPEAKER_03: to promote the importance of genetics in social outcomes is not just wrong,
- 3:46SPEAKER_03: But it's also a moral offense and is deeply repugnant.
- 3:52SPEAKER_03: And as I say, I was just surprised these viewers couldn't look past this and say,
- 3:57SPEAKER_03: well, look, there's lots of interesting evidence here.
- 4:03SPEAKER_03: And then again, also, I was astonished then when I tried the University
- 4:08SPEAKER_03: of Chicago Press, the report there was that they couldn't actually find anyone
- 4:13SPEAKER_03: who would agree to review the book manuscript.
- 4:18SPEAKER_03: They didn't explain why or what was happening.
- 4:21SPEAKER_03: But as I say, it's just an interesting feature.
- 4:24SPEAKER_03: And I think it must be very bad for kind of systematic scientific exploration
- 4:29SPEAKER_03: of areas to have these kind of moral fatwas stipulated in advance.
- 4:36SPEAKER_02: I will say, I'll tell you guys, I mentioned this before,
- 4:40SPEAKER_02: I did not want to review Facing Reality Two Thrus About Race in America
- 4:45SPEAKER_02: by Charles Murray because I'm friends with him.
- 4:48SPEAKER_02: OK, so I rejected attempts
- 4:52SPEAKER_02: to review or like people approach me like, could you review this book?
- 4:56SPEAKER_02: I'm like, OK, like it'd be like if I review Garrett's book or your book, even Greg,
- 5:00SPEAKER_02: it's just like if I didn't like it, I feel kind of like, OK, like what?
- 5:04SPEAKER_02: You know, you know what I'm saying?
- 5:05SPEAKER_02: So I try not to review books of people that I know personally.
- 5:09SPEAKER_02: Anyway, I ended up reviewing the book for Quillette
- 5:12SPEAKER_02: because Claire Lehman could not find anyone willing to review the book.
- 5:16SPEAKER_01: Even Claire couldn't. Wow.
- 5:18SPEAKER_02: Yeah.
- 5:19SPEAKER_02: And I had, yeah, I've had this issue before where there's a, there's like, you know,
- 5:23SPEAKER_02: controversial books.
- 5:24SPEAKER_02: People do not want to be having seen, I guess,
- 5:28SPEAKER_02: like people don't want people to know that they read them. That's really strange.
- 5:31SPEAKER_02: Now this was kind of during the awokening, during the woke period,
- 5:34SPEAKER_02: which in some ways is still going on, but it's like, definitely it's abated.
- 5:38SPEAKER_02: But like, it's like, Ritual pollution or social contagion.
- 5:43SPEAKER_02: It's like the very act of reading this forbidden book.
- 5:46SPEAKER_02: I remember one time I tweeted out that I was reading a biography of Madison Grant
- 5:51SPEAKER_02: and people were like, oh, you're clearly a Nazi.
- 5:54SPEAKER_02: I was like, okay, well, I mean, I've read Christian apologetics too.
- 5:57SPEAKER_02: That, you know, that doesn't, you know, or I read the Bible.
- 6:00SPEAKER_02: We should read widely, yeah. Yeah.
- 6:02SPEAKER_02: So it's a weird, it's a weird attitude that's out there today,
- 6:06SPEAKER_02: but I'm not entirely surprised, but I'm surprised that it's gotten so far gone that,
- 6:12SPEAKER_02: you know, just to be clear, I mean, you guys know what I'm talking about.
- 6:15SPEAKER_02: There's a lot of different types of academic books that are published out there,
- 6:17SPEAKER_02: a lot, on a lot of weird topics.
- 6:20SPEAKER_02: So this is notable.
- 6:22SPEAKER_01: And it's especially notable because academics publish books with data and hypotheses
- 6:29SPEAKER_01: so that other people can argue about them, right?
- 6:31SPEAKER_01: That is supposed to be the academic ideal, right?
- 6:33SPEAKER_01: I mean, Greg's book has already reached above that level of being just
- 6:37SPEAKER_01: an idea out there that one might or might not find valuable.
- 6:40SPEAKER_01: As you mentioned, the PNAS piece is sitting out there.
- 6:43SPEAKER_01: For those of you out there who are just dropping in,
- 6:46SPEAKER_01: this is his book To Have and Have Not And the manuscript of it is on the website
- 6:50SPEAKER_01: for Econ Journal Watch which is run by my George Mason University colleague.
- 6:55SPEAKER_01: And so you can read that book right now while you're listening to us.
- 6:58SPEAKER_02: Yeah, I'll post the link to the PDF of the show notes.
- 7:01SPEAKER_02: Hopefully it'll get somewhere.
- 7:02SPEAKER_01: Actually, and I also want to say that Chicago did not ask me to review the book.
- 7:06SPEAKER_01: I've been willing to review other controversial books.
- 7:09SPEAKER_01: Sometimes I say thumbs up and sometimes I say thumbs down.
- 7:12SPEAKER_01: So that's part of the job here.
- 7:14SPEAKER_01: They give us anonymity so that we can be honest.
- 7:16SPEAKER_01: And it's really disappointing that people won't use that anonymity.
- 7:20SPEAKER_03: Yeah, no, one thing that was actually clear in some sense
- 7:23SPEAKER_03: from the review process is had the book reached a different conclusion
- 7:27SPEAKER_03: with the same data, the same methods, it would have been accepted with Princeton.
- 7:33SPEAKER_03: There would have been no issue.
- 7:35SPEAKER_03: And so it was interesting that it's just the conclusion itself is what is causing
- 7:41SPEAKER_03: the problems, not so much the methods that the book is using.
- 7:44SPEAKER_01: Yeah. There have been multiple, just for the audience,
- 7:46SPEAKER_01: there have been multiple discussions online,
- 7:48SPEAKER_01: like Econ Journal Watch actually held a symposium on Greg's book
- 7:52SPEAKER_01: where they asked critics to come back and critique Greg's book,
- 7:56SPEAKER_01: and then Greg responded to those critiques.
- 7:58SPEAKER_01: And that earlier happened.
- 7:59SPEAKER_01: I assume this was around the time of the PNAS piece.
- 8:02SPEAKER_01: So Greg came and gave a talk at George Mason University where he presented
- 8:07SPEAKER_01: an earlier version of the same book to our economics department.
- 8:11SPEAKER_01: It was really well received.
- 8:12SPEAKER_01: People loved learning about this book early on.
- 8:15SPEAKER_01: And that gave me a first chance to read through this and read the debates about it.
- 8:20SPEAKER_01: The fact that the critiques are basically versions of, well,
- 8:23SPEAKER_01: whatabouts or versions of whack-a-mole is kind of disappointing.
- 8:28SPEAKER_01: Well, it's actually deeply disappointing.
- 8:30SPEAKER_01: That's just the point when somebody is saying, well, maybe it could be this.
- 8:33SPEAKER_01: Well, maybe it could be that.
- 8:34SPEAKER_01: When Greg has, in fact, responded in multiple ways,
- 8:37SPEAKER_01: he gives multiple lines of evidence saying, no,
- 8:39SPEAKER_01: I checked something very close to what you just said or exactly what you just said.
- 8:43SPEAKER_01: It's just shocking.
- 8:44SPEAKER_01: To me, the fact that your maternal grandfather matters just as much
- 8:48SPEAKER_01: as your paternal grandfather is absolutely destructive to the most naive sort
- 8:53SPEAKER_01: of sentimental novel views of social transmissions.
- 8:57SPEAKER_02: Yeah. Yeah.
- 8:58SPEAKER_02: So we should talk about like, like some of these findings, obviously P&S paper were there.
- 9:03SPEAKER_02: I mean, this is a much longer treatment, but how many pages is the PDF actually?
- 9:08SPEAKER_02: Like I didn't, It's 320.
- 9:10SPEAKER_01: 320, I've got it right here, yeah.
- 9:12SPEAKER_02: Okay, that's reasonable. That's a reasonable length for an academic book.
- 9:16SPEAKER_01: 1.5, probably at 1.5 line spacing, something like that?
- 9:20SPEAKER_02: I mean they're probably going to – if you get an editor, they're probably going to.
- 9:23SPEAKER_02: Just really quickly, before we get to the book because we should talk
- 9:25SPEAKER_02: about the content. Because a lot of content is interesting.
- 9:27SPEAKER_02: And I think that, actually, Garrett,
- 9:28SPEAKER_02: that's actually the main one that I bring up with normies,
- 9:32SPEAKER_02: with regular people that are not familiar with it.
- 9:34SPEAKER_02: Okay, so that's like everybody.
- 9:36SPEAKER_01: No offense, Greg.
- 9:38SPEAKER_02: But the maternal-paternal contrast.
- 9:41SPEAKER_02: But, you know, Passage Crest, Jonathan Kieberman is reaching out to you, right?
- 9:45SPEAKER_02: Okay, so you might get a publisher soon. We'll see.
- 9:47SPEAKER_03: We'll see how social media works.
- 9:49SPEAKER_03: I should update you on this.
- 9:51SPEAKER_03: Actually, I have an agreement now with Polity Press.
- 9:55SPEAKER_03: Oh, congratulations.
- 9:57SPEAKER_03: And as I say, I have an enormous debt of gratitude to Dan Klein because it's
- 10:02SPEAKER_03: as a direct result of that book symposium that they reached out and said, hey,
- 10:07SPEAKER_03: we're willing to consider it.
- 10:09SPEAKER_03: And they had a few conditions in terms of disclaimers they wanted added,
- 10:15SPEAKER_03: but essentially the book manuscript will come out.
- 10:19SPEAKER_03: And actually it's been revised since the symposium.
- 10:23SPEAKER_03: And so there will be this free online version available,
- 10:27SPEAKER_03: but the actual published version will have some additional material as well.
- 10:31SPEAKER_03: All right.
- 10:32SPEAKER_01: I look forward to buying it. That's great.
- 10:33SPEAKER_01: And yeah, Dan just did enormous work overseeing the whole symposium
- 10:37SPEAKER_01: and he gets credible critics.
- 10:39SPEAKER_01: And so we can see what his critics are actually saying and have been saying
- 10:42SPEAKER_01: over the last couple of years.
- 10:44SPEAKER_02: Yeah. All right.
- 10:46SPEAKER_02: So I guess, you know, before we go into like the details of the books
- 10:49SPEAKER_02: and the browse tags, I just want to like tell people, I think the biggest, you know,
- 10:53SPEAKER_02: Greg, I think I had coffee with you like a decade ago in Davis.
- 10:57SPEAKER_02: That's correct, yeah.
- 10:58SPEAKER_02: And I think some of these data, some of these results you've known for a long time.
- 11:04SPEAKER_02: So the 0.8 persistence, that came up.
- 11:08SPEAKER_02: Like a decade ago with me in Davis social status, right?
- 11:13SPEAKER_02: Generation to known I think the other thing is
- 11:17SPEAKER_02: with the PNF paper there was huge controversy a lot population genesis
- 11:23SPEAKER_02: and behavior genesis got involved
- 11:25SPEAKER_02: in this our favorite Gadfly,
- 11:31SPEAKER_02: Sasha Gusev devoted a lot of his mental energy creating alternative models.
- 11:38SPEAKER_02: But basically it shows that the patterns
- 11:43SPEAKER_02: of decay and persistence of social status
- 11:47SPEAKER_02: across society, at least in your data set, are entirely compatible with a Well,
- 11:53SPEAKER_02: can I say overwhelming genetic inheritance pattern of the social status.
- 11:59SPEAKER_02: So it's like social status is basically in this model, which, again,
- 12:05SPEAKER_02: may not be correct or may be correct, but it's consistent with the data.
- 12:08SPEAKER_02: Social status can be almost entirely a matter of some sort of heritable component,
- 12:13SPEAKER_02: like it's a heritable trait.
- 12:15SPEAKER_02: And the correlation of social status
- 12:18SPEAKER_02: across individuals is directly a function in these data.
- 12:23SPEAKER_02: It could be directly a function of the genetic relatedness, right? Yes.
- 12:28SPEAKER_02: And so basically you're arguing here, well, it could be something else.
- 12:32SPEAKER_02: It could be some less parsimonious model,
- 12:34SPEAKER_02: but this is a pretty good model and we should consider it, right?
- 12:38SPEAKER_03: Yeah, no, I mean,
- 12:39SPEAKER_03: the whole argument of the book in some ways is Occam's Razor.
- 12:46SPEAKER_03: If you have a genetic model which correctly predicts a whole bunch of things
- 12:51SPEAKER_03: and predicts things like, for example, gender equality across the lineage,
- 12:57SPEAKER_03: In terms of outcomes, and you can only get such predictions
- 13:02SPEAKER_03: in a cultural transmission model or a social transmission model
- 13:07SPEAKER_03: by cobbling something on that's not inherent to that model,
- 13:11SPEAKER_03: that there's an argument then for saying, well, look,
- 13:13SPEAKER_03: the genetic transmission story is in some sense the simplest and most powerful story
- 13:19SPEAKER_03: in the case of England.
- 13:21SPEAKER_01: Yeah. And that just, to me, that debunks or at least strongly critiques the way most
- 13:28SPEAKER_01: of us learn about British social life,
- 13:29SPEAKER_01: which is through novels where something tragic happens when someone's young.
- 13:33SPEAKER_01: When someone is young, they're orphaned and that ruins their life. Right.
- 13:37SPEAKER_01: They inherit a lot of wealth and that totally improves their life.
- 13:41SPEAKER_01: You actually test these theories in multiple ways.
- 13:43SPEAKER_01: So these ideas that we have floating around in our head
- 13:45SPEAKER_01: about how life outcomes are determined here, you're able to test them.
- 13:49SPEAKER_01: Like, does it matter if your dad dies when you're really when you're a little kid?
- 13:52SPEAKER_01: Does it matter if you're there are 10 kids in your family instead of two?
- 13:55SPEAKER_01: So the wealth has to get divided up among more people.
- 13:57SPEAKER_01: None of these things matter.
- 13:59SPEAKER_01: At least to a first approximation. Right.
- 14:01SPEAKER_01: In the long run. In the long run.
- 14:02SPEAKER_03: For wealth, it matters for a while. Yeah.
- 14:05SPEAKER_03: For wealth, it'll matter for two or three generations.
- 14:08SPEAKER_03: But eventually, it washes out.
- 14:11SPEAKER_01: And even then it has minimal effect on occupational status, if I understand correctly.
- 14:16SPEAKER_01: So wealth matters for wealth, but wealth doesn't matter for other things.
- 14:19SPEAKER_01: And so much of modern progressive discourse is about saying that wealth matters for everything.
- 14:25SPEAKER_01: Yes. And you're able to test that hypothesis repeatedly and in multiple ways
- 14:30SPEAKER_01: across multiple centuries. Right.
- 14:32SPEAKER_03: So this is another weird feature of the book is that some of the other
- 14:39SPEAKER_03: Chapters are actually now coming out as journal publications.
- 14:43SPEAKER_03: And so the piece on wealth is coming out and review of income and wealth or just came out.
- 14:48SPEAKER_03: It's a good journal, yes.
- 14:49SPEAKER_03: You know, and then, so that's the, as I see,
- 14:53SPEAKER_03: it's another strange thing is there's not a lot of books that get rejected
- 14:57SPEAKER_03: where something like six journal articles have actually been published based
- 15:01SPEAKER_03: on the material in the book.
- 15:03SPEAKER_01: Yeah.
- 15:04SPEAKER_01: The way the publishers have treated you is just embarrassing to those publishers.
- 15:08SPEAKER_01: Right.
- 15:09SPEAKER_02: Yeah. You know, one thing that I'll say is many people are pretty fascinated.
- 15:14SPEAKER_02: It's like drama.
- 15:15SPEAKER_02: Many people are very fascinated about the way you've been treated, the controversy,
- 15:20SPEAKER_02: you know, the fact that it wasn't published.
- 15:24SPEAKER_02: But I think actually the results are...
- 15:27SPEAKER_02: More interesting, unfortunately,
- 15:28SPEAKER_02: but maybe this is a way to get the results out there to the people.
- 15:33SPEAKER_02: So, for example, the maternal-paternal thing, patriarchy is a fact of life.
- 15:38SPEAKER_02: It's not just made up.
- 15:40SPEAKER_02: Patriarchal societies and inheritance patterns,
- 15:42SPEAKER_02: they're baked into the cake of many human societies,
- 15:45SPEAKER_02: including into Western societies until very recently.
- 15:48SPEAKER_02: Even in England, there was primogeniture,
- 15:51SPEAKER_02: inheritance by the oldest son for aristocratic or gentry families.
- 15:57SPEAKER_02: So the result of maternal and paternal effects are similar.
- 16:01SPEAKER_02: I think that should be shocking.
- 16:03SPEAKER_02: It should be groundbreaking in terms of our understanding of how strong social effects are.
- 16:09SPEAKER_02: And by analogy, I will use a pattern, something.
- 16:14SPEAKER_02: So in a lot of Indo-European mythology, Indo-European societies are all...
- 16:18SPEAKER_02: Pretty much all of them originally are patriarchal.
- 16:21SPEAKER_02: There's a few matrilineal patterns you see among the pigs or something.
- 16:25SPEAKER_02: But the common core mythology is patrilineal
- 16:30SPEAKER_02: and all Eurasian step nomads are patrilineal.
- 16:33SPEAKER_02: Alice Evans has done the survey on this.
- 16:35SPEAKER_02: There's some reasons for this about herding and whatever and how you split up
- 16:39SPEAKER_02: the herds among sons, whatever.
- 16:41SPEAKER_02: But there is also a pattern in a lot of these societies
- 16:44SPEAKER_02: where maternal uncles become very important.
- 16:48SPEAKER_02: And this is in contradiction to the official norms.
- 16:52SPEAKER_02: And, you know, my family's from Bangladesh.
- 16:54SPEAKER_02: And in Bangladesh, it's like basically your family is officially your just paternal family.
- 16:59SPEAKER_02: Right.
- 17:00SPEAKER_02: Like in terms of like where you're supposed to be from and what your affiliations are.
- 17:04SPEAKER_02: But basically everyone just ignores that in reality.
- 17:07SPEAKER_02: And the maternal family is also super important.
- 17:08SPEAKER_02: And often, you know, as I think people know from, you know, just experience,
- 17:14SPEAKER_02: you know, your mother's relationship to her parents like have a huge effect on you.
- 17:18SPEAKER_02: And there's data from, you know, Virpi Luma's data in Finland that, you know,
- 17:23SPEAKER_02: maternal, you know, The survival of a maternal grandmother is much more important
- 17:28SPEAKER_02: than the survival of a paternal grandmother.
- 17:30SPEAKER_02: And I'm bringing that up because we have all these theories based on what people say
- 17:36SPEAKER_02: and the laws.
- 17:37SPEAKER_02: And then when you look at the data and the practice, obviously sometimes it's in contravention.
- 17:41SPEAKER_02: So I don't think that these results should be entirely shocking,
- 17:45SPEAKER_02: although they are surprising, I think, right?
- 17:47SPEAKER_03: Yes. No, no.
- 17:51SPEAKER_03: Now there's enough data that that gender equality is really quite strong.
- 17:56SPEAKER_03: I mean, it's within 10% or something in terms of the error bounds.
- 18:01SPEAKER_03: And it's the same for 18th century England, 19th century England, 20th century England.
- 18:07SPEAKER_03: You've got all of these changes in the role of women and men in these periods.
- 18:12SPEAKER_03: And we could also split it up by poor households and rich households.
- 18:17SPEAKER_03: And they have very different social structures in Britain.
- 18:20SPEAKER_03: But they have this kind of same equality in genders in terms
- 18:24SPEAKER_03: of predicting child outcomes.
- 18:28SPEAKER_03: And so that has to say,
- 18:29SPEAKER_03: I think it's a genuinely surprising and kind of powerful result.
- 18:36SPEAKER_03: And as I say,
- 18:37SPEAKER_03: inexplicable on any theory of social transmission of status.
- 18:45SPEAKER_02: Yeah. Yeah.
- 18:47SPEAKER_02: So this is like underlying, you know,
- 18:50SPEAKER_02: these underlying parameters
- 18:53SPEAKER_02: that are not like enshrined into norms and customs, obviously.
- 18:57SPEAKER_02: In fact, they contradict what you would expect. Right.
- 19:02SPEAKER_02: And, you know, you're talking about like the persistence of this statistic across societies.
- 19:08SPEAKER_02: Actually, like I haven't like read the whole book.
- 19:09SPEAKER_02: Like, is there Do you have enough data to see if this pattern persists outside of England?
- 19:17SPEAKER_03: Oh, yeah. So the reason I took this job in Denmark was actually
- 19:23SPEAKER_03: to get access to Danish registered data.
- 19:27SPEAKER_03: And so I've worked with...
- 19:29SPEAKER_02: Okay, that's just really funny.
- 19:32SPEAKER_02: The reason I took this job is to get access to Danish registered data.
- 19:36SPEAKER_02: Just let that sink in for the listener.
- 19:38SPEAKER_02: That signals commitment, yes.
- 19:40SPEAKER_03: But go on, Greg.
- 19:42SPEAKER_03: Anyway, and it's actually, it's been great, I mean,
- 19:44SPEAKER_03: because the Danes actually then funded a couple of PhD students.
- 19:49SPEAKER_03: And what we're able to show is that because people think of Denmark and Sweden
- 19:56SPEAKER_03: as paradigms for societies where the welfare state has successfully intervened
- 20:02SPEAKER_03: and successfully raised social mobility rates.
- 20:06SPEAKER_03: And so just the empirical claim is that Danish
- 20:11SPEAKER_03: or Swedish modern social mobility rates are only very slightly higher
- 20:17SPEAKER_03: than somewhere like England.
- 20:19SPEAKER_03: That essentially they have the same social structure.
- 20:22SPEAKER_03: And then the other thing is you can fit the PNAS genetic model to modern Swedish
- 20:29SPEAKER_03: and Danish data and it'll actually fit pretty well and it will generate
- 20:34SPEAKER_03: the same parameters.
- 20:36SPEAKER_03: And so you do observe very clearly
- 20:39SPEAKER_03: in these societies surprising long-term persistence.
- 20:45SPEAKER_03: So you can look across, you know,
- 20:47SPEAKER_03: three marriages and uncles and aunts and stuff like that and get to people
- 20:51SPEAKER_03: who are very distantly related,
- 20:53SPEAKER_03: but they're still significantly correlated with the target individual.
- 20:59SPEAKER_03: And as I say, so part of the book is actually saying
- 21:02SPEAKER_03: that the same social structure seems to actually also occur in these societies.
- 21:07SPEAKER_03: And one of the important things in the book is that these slow rates
- 21:11SPEAKER_03: of social mobility with genetic transmission are only possible
- 21:14SPEAKER_03: if there's very high degree of assortment in marriage.
- 21:18SPEAKER_03: And that shows up very, very clearly in Danish and Swedish data is the same number.
- 21:25SPEAKER_03: In terms of the underlying correlation of partners and marriage that you get
- 21:30SPEAKER_03: in modern England And this is
- 21:35SPEAKER_01: something that someone would actually learn accurately
- 21:37SPEAKER_01: if they just read a bunch of old novels,
- 21:39SPEAKER_01: because people back then really cared about the social status
- 21:41SPEAKER_01: of the person you married. Yes.
- 21:43SPEAKER_01: So there are eras, there are things, social processes in old novels might be wrong,
- 21:48SPEAKER_01: but descriptions of social institutions are probably more accurate.
- 21:52SPEAKER_01: And is it, do you think that weakened dramatically in, say, the mid-20th century
- 21:57SPEAKER_03: in Britain?
- 21:58SPEAKER_03: Or did it not really change much?
- 22:01SPEAKER_03: So one of the very nice data sets we have that was generated
- 22:05SPEAKER_03: with my colleague Neil Cummins is We have something like 1.7 million marriages
- 22:11SPEAKER_03: in England from 1835 till now,
- 22:14SPEAKER_03: where the English marriage records record the occupation of both fathers
- 22:21SPEAKER_03: and the groom and the bride as well.
- 22:23SPEAKER_03: And so we can actually derive consistent measures of what
- 22:27SPEAKER_03: the underlying assortment is in marriage across this long interval.
- 22:32SPEAKER_03: Where, as I say, society has changed in such fundamental ways that exactly
- 22:37SPEAKER_03: the same number comes out.
- 22:39SPEAKER_03: That underlying correlation between Brighton and Room seems to be about 0.8 in terms
- 22:45SPEAKER_03: of their kind of occupational, implied occupational status.
- 22:50SPEAKER_03: And as I say, it's that stability that's then kind of consistent with the stability and how slow
- 23:00SPEAKER_03: How strong is persistence across generations?
- 23:04SPEAKER_03: And that also matches with how strong persistence is across siblings.
- 23:09SPEAKER_03: So the nice thing about the English data is, as I say,
- 23:11SPEAKER_03: we can look at all of these different aspects of family life.
- 23:15SPEAKER_03: And I'd say it took a long time to assemble all of this data.
- 23:19SPEAKER_03: Another thing I should mention is that...
- 23:23SPEAKER_03: I tried, you know, my previous books, I was actually able to get some NSF funding.
- 23:29SPEAKER_03: I tried both in England and in the U.S. to get funding to build up this database.
- 23:34SPEAKER_03: And I think it's a very interesting database.
- 23:37SPEAKER_03: Everyone says that, even your critics.
- 23:39SPEAKER_03: But no one would fund this
- 23:44SPEAKER_03: and I'm sorry,
- 23:48SPEAKER_03: no official funder would fund this.
- 23:50SPEAKER_03: And I even got rejection notices that said things like,
- 23:55SPEAKER_03: we should not fund anything that seeks to establish why it's not possible
- 24:00SPEAKER_03: to change rates of social mobility.
- 24:04SPEAKER_03: We should only fund projects which show us how to increase rates of social mobility.
- 24:09SPEAKER_01: Yeah, I've run across the same kind of statements myself, yeah.
- 24:12SPEAKER_02: We know how to increase rates of social mobility.
- 24:15SPEAKER_02: The Chinese did it for a little while.
- 24:16SPEAKER_03: Yes.
- 24:18SPEAKER_03: And then maybe the Russians also.
- 24:22SPEAKER_02: So, you know, you're talking about Scandinavian societies, which, you know,
- 24:28SPEAKER_02: for the listeners who have, you know,
- 24:30SPEAKER_02: like a lot of my listeners will know about the Hajdan Line and all these things.
- 24:34SPEAKER_02: These are Northern European societies.
- 24:35SPEAKER_02: These are the weirdest of the weird societies. Right.
- 24:38SPEAKER_02: Do you think, you know, I mean, I know the persistence rate is generalizable,
- 24:44SPEAKER_02: but the gender symmetry, the sex symmetry, is that, do you see any evidence of that,
- 24:50SPEAKER_02: say, in China or elsewhere?
- 24:52SPEAKER_03: Oh, that's a good question.
- 24:54SPEAKER_03: And it's actually, that's a completely open question.
- 24:57SPEAKER_03: It would be very interesting to kind of try and go to Chinese society and figure,
- 25:03SPEAKER_03: is that also the case within a society like China
- 25:07SPEAKER_03: or various other Middle Eastern societies, various other societies?
- 25:12SPEAKER_03: Is this the case?
- 25:14SPEAKER_03: And I honestly have no Great expectations there.
- 25:20SPEAKER_03: I know that, you know, Northern Europe has very similar social structures.
- 25:27SPEAKER_03: But whether this also extends to societies like China, I mean,
- 25:32SPEAKER_03: my guess would be yes, that fundamentally there's not that much difference in kind
- 25:37SPEAKER_02: of social structure between
- 25:39SPEAKER_03: China and Europe.
- 25:41SPEAKER_03: But I would be open, you know, just to see what does that look like.
- 25:46SPEAKER_03: And I think now we will be able to develop data
- 25:49SPEAKER_03: that will show even historically what these patterns are.
- 25:54SPEAKER_02: Well, I mean, so talk us, I mean, you know, we've talked about this before.
- 26:00SPEAKER_02: And just like review, again, like there's been a lot of new people listening,
- 26:05SPEAKER_02: you know, It makes sense to me.
- 26:09SPEAKER_02: Okay, so I read Fisher's paper from 1918 about the correlation between relatives
- 26:15SPEAKER_02: and stuff like that. Most of the people have not read that.
- 26:18SPEAKER_02: So when I was reading your original piano, I was like, okay, this makes sense to me.
- 26:22SPEAKER_02: Can you talk us through this,
- 26:25SPEAKER_02: why it seems like it's genetic or like it's, you know,
- 26:30SPEAKER_03: Well, as I say, I approached this topic with great naivety.
- 26:35SPEAKER_03: And so I was developing this lineage and I was already interested in, you know,
- 26:42SPEAKER_03: to what extent is genetics underpinning status transmission
- 26:47SPEAKER_03: and to what extent is it actually socially determined.
- 26:51SPEAKER_03: And then I read the Fisher paper and that was kind of a revelation
- 26:56SPEAKER_03: because it specifies for everyone in the lineage.
- 26:59SPEAKER_03: What the expected correlation is going to be,
- 27:02SPEAKER_03: where the only parameter it depends on is what the underlying genetic correlation
- 27:09SPEAKER_03: in partnering is.
- 27:13SPEAKER_03: And so it's a very simple model.
- 27:15SPEAKER_03: I mean, essentially when you fit it,
- 27:18SPEAKER_03: it only has three free parameters
- 27:23SPEAKER_03: and And what I was then amazed to discover was
- 27:27SPEAKER_03: that we could take this data for England.
- 27:30SPEAKER_03: And so one set of data, for example, was people's literacy in the 18th and early 19th century.
- 27:36SPEAKER_03: And another set of data was people's house values
- 27:41SPEAKER_03: in England from the year 1999
- 27:44SPEAKER_03: until the present.
- 27:46SPEAKER_03: And so these are very different periods, very different times.
- 27:50SPEAKER_03: And the astonishing thing is That if you fit that Fisher model to the data,
- 27:56SPEAKER_03: it generates the same underlying parameters for modern England and for England
- 28:02SPEAKER_03: in the 18th century.
- 28:03SPEAKER_03: And in fact, because it's looking at relatives as distant as fourth cousins,
- 28:08SPEAKER_03: you're actually tracing also this back all the way till 1600.
- 28:13SPEAKER_03: Because, you know, when you're looking at fourth cousins,
- 28:16SPEAKER_03: they're only connected by a common ancestor or something like, I can't remember now,
- 28:20SPEAKER_03: five generations back.
- 28:22SPEAKER_03: And so you're actually looking at this long structure of English society
- 28:28SPEAKER_03: and with nine different social outcomes, you get essentially the same answer.
- 28:35SPEAKER_03: And as I say, it's a very interesting answer.
- 28:38SPEAKER_03: It has this one kind of disturbing element,
- 28:40SPEAKER_03: which it requires a genetic correlation of spouses and relevant genetics
- 28:46SPEAKER_03: of something like 0.57.
- 28:49SPEAKER_03: And that is, as I say, unexpectedly high genetic correlation.
- 28:53SPEAKER_03: But the interesting thing is that the phenotype correlation from the data I have,
- 28:58SPEAKER_03: it's very clear, is somewhere around 0.8.
- 29:01SPEAKER_03: So 0.8 actually will allow for a genotype correlation of as high as 0.57 with quite
- 29:08SPEAKER_03: a reasonable level of heritability of the trait.
- 29:14SPEAKER_03: And so the fact that people are correlating so strongly in terms
- 29:17SPEAKER_03: of the phenotype It's really going to imply that there's likely to be also
- 29:23SPEAKER_03: a quite strong genetic connection between the parties.
- 29:28SPEAKER_03: And so as you see, it's a very surprising discovery.
- 29:34SPEAKER_03: I mean, and that's not disputed.
- 29:37SPEAKER_03: I mean, some people have gone in and said, well, look,
- 29:41SPEAKER_03: when you counted up fourth
- 29:43SPEAKER_01: cousins, you overweighted
- 29:46SPEAKER_03: You know, particular sets for cousins.
- 29:49SPEAKER_03: But honestly, if you were to do something like that,
- 29:51SPEAKER_03: if you're really overweighting whatever way you're doing, then it should just mess up things.