Your observation is a known flaw with this line of thinking. In fact, it's such a powerful rejoinder I'm really surprised the NYT didn't address it.
If I remember correctly, movement among quintiles in the US is such that the average person moves through many. Yes, there are the permanently wealthy, and the permanently poor, but they are the exceptions rather than the rule. So when people say "the rich get richer" what they really mean is that the income segment defined by that qunintile, over all, got richer. Individual people, however, move around a lot, as anybody who has participated in the economy in the last couple of decades can attest.
I know I'm not sourcing my statement. Apologies. I'm simply relying on several rebuttals to this type of graph I've read over the years. Hopefully somebody else will be able to provide some more substance and move the discussion along.
These studies of relative mobility have produced remarkably consistent results, with regard to both the degree of mobility and the extent of changes in mobility over time.[5] Mobility in the United States is substantial according to this evidence. Large proportions of the population move into a new income quintile, with estimates ranging from about 25 to 40 percent in a single year. As one would expect, the mobility rate is even higher over longer periods—about 45 percent over a 5-year period and about 60 percent over both 9-year and 17-year periods.
Those statistics aren't actually that encouraging, though they aren't bad per se. If 60% move a quintile over 17 years, that means 40% stay put long-term. And that isn't even a measurement of large income mobility, like jumping more than one quintile. Iirc from data elsewhere, the most common quintile moves by far are churn between the 2nd/3rd/4th quintiles, with 2-or-more-quintile jumps and movement in or out of the top and bottom quintiles being less common than 2nd<->3rd or 3rd<->4th drifting.
I'd like to see more data about this, as I find it interesting.
I don't know. I feel pretty assured, however, that if 40% move every year, that the graph as drawn leaves out some very important details. As you noted, the interesting question (from a political viewpoint) is how much mobility is in and out of the top quintile as opposed to the bottom one.
And, reasoning a bit further, if people only move among the bottom three quintiles, to them it's probably evidence that one day they might be in the top quintile, whether or not that's actually likely. This might give them a much more favorable view of lenient tax rates on the wealthy. Don't know.
And from a purely political standpoint, if people feel that they are moving among the quintiles such that anybody has a chance at being in the top bunch, does it matter what the actual data is? In other words, is there a perception problem or are we trying to create one? Once again, I think you could argue this either way from the data I've seen so far.
I call b.s. In economics, the extremes are strongly self-reinforcing. I would be really surprised if in a given year more people left the top 0.05% (in net worth) than stayed there. I'd even be surprised if the ratio were less than 5 to 1.
If I remember correctly, movement among quintiles in the US is such that the average person moves through many. Yes, there are the permanently wealthy, and the permanently poor, but they are the exceptions rather than the rule. So when people say "the rich get richer" what they really mean is that the income segment defined by that qunintile, over all, got richer. Individual people, however, move around a lot, as anybody who has participated in the economy in the last couple of decades can attest.
I know I'm not sourcing my statement. Apologies. I'm simply relying on several rebuttals to this type of graph I've read over the years. Hopefully somebody else will be able to provide some more substance and move the discussion along.
EDIT: Random Google link: http://www.urban.org/publications/306775.html
These studies of relative mobility have produced remarkably consistent results, with regard to both the degree of mobility and the extent of changes in mobility over time.[5] Mobility in the United States is substantial according to this evidence. Large proportions of the population move into a new income quintile, with estimates ranging from about 25 to 40 percent in a single year. As one would expect, the mobility rate is even higher over longer periods—about 45 percent over a 5-year period and about 60 percent over both 9-year and 17-year periods.