The numbers generally cited in support of this argument do not actually tell us much about what has happened to the incomes of wealthy households over time. That’s because the people who are in the top bracket today are not the people who were in the top bracket last year. There’s a good deal of socioeconomic mobility in the United States — more than you’d think. Our dear, dear friends at the IRS keep track of actual households (boy, do they ever!), and sometimes the Treasury publishes data about what has happened to them. For instance, among those who in 1996 were in the very highest income group isolated for study — the top 0.01 percent — 75 percent were in a lower income group by 2005. The median real income of super-rich households went down, not up. The rich got poorer. Among actual households, income grew proportionally more for those who started off in the low-income groups than those that began in high-income groups.
This piece appeared a day ago and I've been waiting to see a good blogospheric response to it. I'm still waiting. All the data I've looked at in recent months suggests that income mobility is as stagnant as wages in the lower quintile—and, in fact, what makes the income inequality problem a problem is that there's not much chance you're going to be able to work your way out of those lower quintiles.
But I'm not an economist: I rely on economists to make sense of the data for me. And I'd really like to know if Williamson's right or wrong about this, or if this information looks different within a larger picture. Anybody out there?