The Economist Greg Clark, who you may have heard of as the author of A Farewell to Alms, had another book, The Son Also Rises, about social mobility - this time showing that it is quite slow in both egalitarian societies like Sweden and inegalitarian ones like Chile. I have heard reference to "the surname paper" a few times over the years but had never followed up on it. There is a New Republic interview with him, and there are a a few versions of the paper, of increasing length. This one is good.
Most cultures have some version of a proverb like "clogs to clogs in three generations." Three seems a little quick to those who think in terms of Old Money and land ownership, but it also sounds plausible. Clark says status, when measured by a variety of measures rather than a single one (as economists and other social scientists often do), actually last 10-15 generations. Even a lot of liberal outlets covered this when it came out. Even I can see that some of the criticism of it, some by professionals, is bogus. But I can't attest to other criticisms.
Much of his work is focused on England from 1200-1910 or so. Part of his contention is that more children of the rich survived, and thus are more common now. Hence the surnames as a proxy for that.
I've always heard it as 'shirtsleeves to shirtsleeves' in 3 generations. Americans always get it done quicker. I've seen it happen a few times too, but it always seems to involve a strong patriarch/matriarch generation,one that sets up the mechanisms for wealth transferal across generations, and makes the initial investments in the important family institutions. The few times I've seen the dynamic, trust funds were always involved and the third generation was over-indulged and carried little of the spark and determination that the first one had. Most families are not wealthy enough to support more than a couple of generations of leisure class progeny. British royalty, the Rockefeller and duPonts, and so on. I guess they would be in the running with the landed gentry for the 10 generation cycle.
Clark's data set wasn't just a few, however, it was the most accurate samples he could get from the entire population of a country. Very surprising result, but it likely means that the one generation was the aberration, not that the other two were not adequate. Or it could mean that those lines might produce more flashes like that over time.
The paper is interesting. I fear that people forget that "reversion to the mean" works in two ways. A family line may revert down to the mean from a more prominent position, but it may also revert up to the mean from a lowly position.
And when children from two different lines marry, how does one interpret that?
For example, my father descends from William Brewster, a Cambridge-educated spiritual leader of the Pilgrims. He also descends from John Billington, "America's first murderer." (https://www.mayflower400uk.org/education/who-were-the-pilgrims/2020/may/john-billington/) The lines crossed in the 19th century, if not before.
The method of using rare last names perforce chooses a small segment of a family tree. All female descendants are excluded, as are all illegitimate descendants--or people whose names were normalized through spelling changes.
Any family name can be extinguished by one generation of daughters, or a generation without issue. So there is a selection effect.
I agree. I am descended from both Brewster at least two ways and Sthephen Hopkins at least five, so I get the contrast.
But I think if you get your sample size large enough you can slowly was that out, which is the reasoning behind the genetic studies with N=3M. This isn't that, but the numbers do add up. It means little at an individual level, but in aggregate you can start to declare that the pattern has some validity. The importance is in terms of social policy, and what you can fix and what you can't. I have become more convinced over the years that we should be kind and rescue individuals, because they are the sufferers who God has placed before us, but trying to engineer long-term solutions for The World is likely to do no good and some harm.
Post a Comment