We really don’t have much intergenerational mobility, and it’s been this way forever. There are a bunch of studies, obviously — Greg Clark and Neil Cummins have made a whole career out of showing this. Their 2015 paper on social mobility in England between 1858 and 2012 gives us a generation to generation correlation of wealth of .7 — quite high. They emphasize that the intergenerational correlation gets stronger as the number of generations increases, as random noise gets averaged out. The transmission of social status in England, from 1754 to 2023, is just as strong through the mother and the father. The intergenerational correlation in educational attainment is the same today as it was in the middle ages, as their paper covering the years 1170 to 2012 shows. In Sweden, the same holds true, with social mobility rates similar to England and the US. Even in America, the land of opportunity, intergenerational mobility isn’t very high. Chetty, Hendren, Kline, Saez and Turner (2014) looked specifically at America from 1971 to 1993, and found that intergenerational mobility remained extremely stable and low. This is a finding consistently replicated by other papers, and is consistent throughout time.
One wonders, though, why we should care. That is not to say that we shouldn’t care about unfairness, about a society which does not allocate on the good one does for others, and which accords special privileges on birth alone. A lack of class mobility could be an indicator of that, but it is not a bad thing in and of itself. A society undergoing a communist revolution could have extraordinary class mobility, and yet have everyone be worse off — I would not trade that for a society where class mobility is lacking but everyone, top to bottom, is getting better off! Class mobility is only the shuffling of people’s rank-order, not an improvement for all.
After all, there are many meritocratic reasons why intergenerational mobility might be low. All behavioral traits have a substantial genetic component, which will cause outcomes of parent and child to be correlated with each other. Clark and Cummins 2023 show how social status transmission fits perfectly onto an additive genetic model with assortative marriage. This explains why including multiple generations increases the correlation of social status. It doesn’t need to be specifically genetic – it can be very hard to tease apart cultural transmission and genetic transmission, to say nothing of gene-environment interactions – but there is some factor increasing productivity which remains, generation after generation.
Even if the advantages of birth are due to upbringing, and these advantages are transferable, changing the allocation of jobs would make us worse off. It’s too late. They’re due to genuine differences in talent, and hiring less talented people — regardless of the unfairness of how that difference came about — would simply make us poorer. It’s not clear that you could ever plausibly transfer the advantages of upbringing. It is not wealth alone, but the underlying traits which led to that difference in wealth. In China, an extreme example of this occurred when the Communist Party expropriated the “black elements”, killed many, and barred them both from party membership and from college. And yet, guess who came out back on top?
Calling these elite families does not mean that they are the top .001%, or anything like that. These were overwhelmingly just particularly successful peasants — those who, by dint of planning and hard work, had worked their way up to owning property. If mass expropriation is insufficient to cause a total shakeup, then our softer, gentler policies in the developed world aren’t going to change much.
Obviously, nepotism is bad. Bloom and van Reenen 2007, in their survey of management practices in the US and Europe, show that firms which are owned by a first born son have considerably worse management practices, and that differences in primogeniture rates are a substantial contributor to regional disparities in firm productivity. In the public service, Felipe Riaño 2023 documents nepotism in Colombia’s bureaucracy. People related to those in power received higher salaries, and were more likely to be promoted. After their family left office, they were subject to meritocratic pressures, and were substantially less likely to be promoted. Thus, appointing family members undermined the quality of government services.
We must vigorously stamp out nepotism, for the same reason we should stamp out corruption. But we oughtn’t stamp out children of successful people doing well. They are doing well because they are simply better, not because they have gotten a leg up which makes society worse off.
Is assortive mating by class or merit? And why aren't all the good genes already at the top?