You are using an outdated browser.
Please upgrade your browser
and improve your visit to our site.
Skip Navigation

Family Wealth Lasts For Ten To Fifteen Generations

An Interview with Gregory Clark, historian of social mobility

George C. Beresford/Hulton Archive/Getty

In his State of the Union address, Obama got a standing ovation for pointing to John Boehner as proof that the American Dream is intact. But Boehner’s rise from son of a barkeep to Speaker of the House might not have anything to do with modern America or its social policies. Geoffrey Chaucer was the son of shoemakers and Charles Dickens left school at age nine, as economic historian Gregory Clark notes in his forthcoming book, The Son Also Rises: Surnames and the History of Social Mobility.

Clark uses historical records to track surnames and social status over generations, and finds that rates of social mobility are surprisingly similar—and surprisingly slow— across societies as diverse as feudal England, modern Sweden and Qing Dynasty China. Most social scientists estimate that it takes about three to five generations for a family’s wealth or poverty to dissipate, but Clark says it takes a staggering ten to fifteen generations—300 to 450 years—and there’s not much the government can do about it. According to his calculations, if you live in England and share a last name with a Norman conqueror listed in the Domesday book of 1086—think Sinclair, Percy, Beauchamp—you have a 25 percent higher chance of matriculating at Oxford or Cambridge. If you’re an American with an ancestor who graduated from an Ivy League college between 1650 and 1850, it’s twice as likely that you’re listed in the American Medical Association’s Directory of Physicians.

Alice Robb: I imagine a lot of people won’t be so excited about your findings.

Gregory Clark: I realized when I was writing this book, that some people would be very uncomfortable or very unhappy with it. There are some things here that are more speculative and absolutely, people will disagree. It’s good to have robust debate. That’s our job in trying to think about the world.

I know there will be a torrent of abuse, potentially, from certain corners, because when people find things uncomfortable, they tend to attack the messenger. I’m not worried about that. I worry about making errors, getting things wrong. The rest you can’t control. I'm not going to get upset because someone says I’m justifying the class system. I’m not a policy-maker. I’m interested in the nature of society and what controls society. I'm not running for office.

AR: How come your estimates are so different from conventional ones?

GC: When I went into this, I thought I was just going to reproduce the usual estimates. What I found was a bit of a shock.

When conventional estimates look at social mobility, they typically focus on one aspect of status—income or wealth or education or occupational status. But it turns out that for any individual, the individual aspects of status are quite weakly correlated. When you look at that kind of mobility—individual aspects of status—you tend to see a lot of random fluctuations, but if you have a more balanced account of the person’s life and status, you actually see that these things are changing much more slowly. People have a deeper, underlying overall social status that changes very slowly between generations.

Take the example of Bill Gates. If you were to measure his status relative to his father’s by education, you would think he had have moved toward the mean: He never finished his college degree. But if you measure his status by income, you get a completely different picture.

I have a friend in San Francisco who has an undergraduate degree from Northwestern and a law degree from Stanford. And his income typically is close to zero, because he spends a lot of time volunteering on political campaigns and practicing Buddhism. So if you look at income, you’d say he’s drastically declining from his father’s social status, but if you measure by education you’d say he’s close to his father’s social status. If a family sees a big decline in income from parent to child, you might predict that the child will revert toward the grandparent’s status. Conventional studies measure one aspect of status, but there’s something deeper that’s changing much more slowly.

AR: Is that underlying thing genetic?

GC: I don’t know. We know that status is changing very slowly. It is not really amenable to social intervention, and it seems very connected with families. Is it something cultural that families pass on to their children, or is it to do with biology? We don’t have the information that would distinguish those two views, but we cannot reject the idea that it’s mostly genetic.

If it were biological, then social groups with lower rates of out-marriage would be more likely to retain their status over time. We can observe historically some startling examples of such groups, and it is in fact the case that they don’t change their social status as much over time—Brahmins in India, Copts in Egypt, Jews. It doesn’t prove anything, but it does mean that we can’t reject the notion that it has something to do with biology.

AR: How did you reach the conclusion that it takes 10 to 15 generations for a family to regress to the mean?

GC: When you start looking at earnings and status over many generations, the process is very simple: Every generation gets closer to the mean by the same proportion every generation. It’s very mechanical—there’s a simple kind of underlying physics.

AR: So is there nothing governments can do to improve social mobility?

GC: There’s a lot of interest in America in things like early childhood education and how we can intervene early in a child’s life. It would be wrong to conclude that we should give up on all of this, but the amount of change that’s feasible from intervention is tiny—you’re going to be operating on the margins of a much deeper process of social mobility.

Modern Sweden, which has very high levels of these types of interventions, has not managed to increase rates of social mobility above that of medieval England, which had none of these government interventions. Or look at post-revolutionary China. Despite the fact that Mao tried to radically remake Chinese society into a new egalitarian system—killing upper-class people, sending them to Taiwan and Hong Kong—surnames that had a relatively high status before the revolution have a relatively high status today. We’re never going to have the kind of society that has the actual social mobility that people are looking for in the United States.

AR: In your conclusion, you say our best bet is to focus on improving income inequality rather than social mobility.

GC:  The evidence is that the majority of people’s outcomes are predictable at the time of birth and are not under individual control. If we understand how strong this inheritance is, we can be more sympathetic to the idea that the rewards for ending up higher or lower on the social spectrum should be made less. We don’t want a society where the winners are already determined at the point of conception.

AR: Both of your books’ titles—A Farewell to Alms and The Son Also Rises— are puns on Hemingway. Are you a big fan?

GC: In both cases, I suggested them to the press as a joke.