[iguy]

Clark finds this pattern in most societies, IIRC Sweden and China have (in his data) almost identical rates of status persistence. It's not a quirk of English manners.

What varies more is the degree to which ordinary people today descend from the nobility in (say) 1100. In some societies they had many more surviving children than average, e.g. it's easy for them to double every generation, within a basically static total population, implying that their offspring make up a high proportion of people after a few centuries. But in other societies, they did not.

His books are pretty readable, BTW, interesting data.

[JoeAltmaier]

Interesting choice - those two societies also have 1000 years of respect for nobility? It seems to be a strong factor then.

[iguy]

I don't know about respect. The data is on persistence of status. They can do this in many countries, those are just two I remember (besides England).

Direct records of ancestry are too scattered to piece together long timescales. What he (and collaborators) do is to find very rare surnames, in records at some distant time (e.g. Oxford graduation in 1600, high-status, or common criminals executed then, low-status) and then trace look for the same name in later data (e.g. Victorian wills, or today's tax data). Rare names give you a fairly targeted marker. One which the carriers are often unaware of.