Is Social Status Determined By Your Genetics? - Gregory Clark

What's this new paper of yours about the inheritance of social status?

So the paper looks at 425,000 people in England over the course of 400 years who are all linked together by descent and marriage, and just asks what describes how you inherit social status. And it ends up that there are actually three very interesting aspects. One is that there's a very strong inheritance of status, much stronger than people conventionally believe, and so there's an underlying correlation that's really strong. The second astonishing aspect is that that correlation hasn't changed over the course of 400 years. There's no more social mobility now than there was in the 17th century, (laughs) the 18th century, the 19th century. But the most surprising element of all is that if you wanna predict how correlated people will be, then that prediction is based on what's their genetic correlation. And so the data is just very consistent with a really simple model of genetic transmission, where what just matters is h- how m- how many genes do we have in common, that'll explain how much outcome we'll have in common. Uh, and, uh, that as I say is for many people very surprising and also quite troubling for a lot of people.

What is the methodology of you being able to track genetics across such a long period of time? How, how were you able to do that? Presumably Ancestry.com wasn't tracking people f- 400 years ago?

Right. (laughs) Um, eh, s- so the lineage stuff is, is fine, and so basically I was using... There, there are all these kind of interesting societies in Britain and one is the Guild of One Name Studies, and these are two or 3,000 people who've devoted themselves to following the history of particular surnames in England, and they've done fantastic jobs kind of actually tracking people's genealogy over hundreds of years. And so, eh, th- that part is, is, is straightforward, but the only thing we can do here is... We don't have... There... We have no direct genetic evidence here. What we can look at though is what would be the predictions of a genetic model of transmission? And that has a distinctive and very clear set of predictions about how correlated fourth cousins, third cousins, second cousins, first cousins will be, uh, and, and it all depends on how much assortment there is in marriage, how strong these correlations will remain. And so there's also, you know, um, predictions about siblings, about grandparents, grandchildren, and as I said the very surprising thing here is that the predictions of that model are very consistent with this data for England, uh, and it, it really is, uh, y- you know... So as I say, so there's nothing conclusive here, there's nothing direct, but there is just a very interesting empirical pattern. And then as part of the paper, I also can say, well, what about other features of inheritance of status that would be consistent with genetic transmission, do they also hold? And so here's... One example is, with genetic transmission, mothers should always have an equal influence as fathers in terms of outcomes for children, because you get half of your genetics from each parent. Now, if we go back to 19th century England, fathers play a very different social role than they do now. If you look now, mothers still actually spend much more time with children than do fathers, and so we would actually expect with social transmission that maybe sometimes fathers are more important and other times mothers are much more important. But the data for this lineage in England is very clear. If you wanna predict children's outcomes, mothers and fathers play exactly the same way. And, eh, and you know, one... So early on we have evidence on the literacy of mothers and fathers, and so that's equally predictive of what the child literacy would be. And it doesn't matter if it's a boy or a girl, in those cases it's again equally predictive. For other outcomes like occupation, w- we don't have occupations for women in the 19th century, but we can proxy women's occupations by, for example, taking their brother and then taking their husband's brother, and that'll actually give us a good proxy for what their occupational status is, and then we can say, "Which predicts child outcomes better? Is it the mother's proxy or the father's proxy?" And the answer again is no, it's exactly the same weight. And so... And that's c- true as I say all the way through the last 300 years, that mothers and fathers play exactly the same role in terms of outcomes except for one important outcome, and that outcome is wealth. For wealth, fathers are much, much more influential than mothers. And that's because wealth tended in England to flow on the patril line, that men inherited more family wealth than women did, and so what matters is what's your grandfather's wealth on the patril line as opposed to your grandfather on, on the matril line? So as I say, it's... This is one piece of interesting ancillary evidence. A second piece is what's the effect of birth order on outcome-...right? And in the social world, I mean, I don't know if you have children. I ha- had three, and the oldest child gets much, much more parental attention than the younger ones do, right? And talk to anyone who has children. Uh, this is true. I mean, the parents, a new thing when the first kid comes along, the parents have all these ideas about how they're gonna shape these kids. (laughs) And so and, and empirically, we know that older children get more attention than younger children. So you might expect that older children will do much better in terms of social outcomes. Uh, it turns out in this data, no. I- i- it's, uh, in, in almost every case, birth order doesn't matter. Your chances in life are the same whether you're first or the last. And a lot of the families in the 19th century, there are 10 children. And so you would think the 10th one is coming into this kinda crowded family. There's no more space. (laughs) My own parents were both from families of 12. And by the time, you know, my father and mother, I think both were number nine, by the time they came along, the housing's incredibly tight. You're sharing a bed with three siblings and stuff like that. And, uh, but it turns out it, it, it, it doesn't matter except, again, there's a slight exception for the top 1% of families in the 19th century, the kind of elite, the oldest son is doing better than the younger sons. But that's the only thing that matters. So the old- and the oldest son inherits more but is also more likely to be sent to university than the younger sons. So you do see slight deviations for this. But for 99% of the population, it doesn't matter what birth order it is. And then another thing you can look at is, well, what about family size? Isn't that going to influence your outcomes? Because the more children there are, the less resources there are, the less parental attention. Again, uh, in the era, uh, i- so there's an era up to marriages of by around about 1880, where in England, family size was random. People made no attempt to control fertility. And it, it is amazing that (laughs) they just apparently got married, produced children. Sometimes there's only one child, you know, just the accidents of fertility. In the sample we have, there's one guy who has 27 children (laughs) from two different wives. (laughs) And so you get this enormous variation in family size, and again, it has no effect except for richer families for wealth. And then if you're from a larger family, then your wealth actually, uh, declines, right? Uh, sorry. If you're from a wealthy family and it's larger, your wealth declines because the wealth has to get divided.