America used to be a land with great upward social mobility, but isn't anymore. America never was a land with great upward social mobility.
Which do you believe? Keep in mind that your answer will have significant implications for public policy.
Most politicians, of both right and left, favor the first statement. Conservatives say big government is stifling people's chances to move upward. Barack Obama says growing inequality of wealth is holding people down.
But Gregory Clark, British-born economist at the University of California, Davis, says they're both wrong. The second statement is correct, he argues in his new book "The Son Also Rises: Surnames and the History of Social Mobility."
The problem with previous studies of social mobility, Clark says, is that they measure differences in income, occupation or status across only one or two generations. They found considerable differences between parents and children, grandparents and grandchildren -- something that looks almost like random flux.
Clark casts his net wider. He looks at mobility not across one or two generations, but across many. And he shows by focusing on surnames -- i.e., last names -- how families overrepresented in elite institutions continue to be so, though to diminishing degrees, not just for a few generations but over centuries.
Some surname data goes back a long way. In medieval England, people with Norman surnames (from the 1066 conquest) and surnames based on place names appear in disproportionate numbers as students at Oxford and Cambridge from 1170, as members of Parliament from 1259 and in upper class probate records from 1380.
Regression to mean -- the fact that very tall people tend to have somewhat less tall children -- reduces that overrepresentation over time. But through the industrial revolution, two world wars and institution of the welfare state, overrepresentation continues. Similarly, those from underperforming families, such as England's travelers or gypsies, remain underrepresented.
Clark finds the same phenomenon in contemporary Sweden, with its generously redistributive welfare state, in Japan where those with samurai names continue to overachieve, in class-bound Chile and, even more so, in caste-bound India.
Clark's analysis touches a tender spot in American debate, for the implication is that genetics -- inherited intelligence -- tends to determine social outcomes. He doesn't quite say so, but he rules out other explanations.