Have I mentioned that I have a paper on this subject?
Update: Vero has a nice take on this subject here.
Back in December I did a post on income mobility. I used data from a 2007 Treasury Department study that followed actual individuals over time (most press treatments of the subject fail to do that). At the time, I could not find the link. My colleague, Seth Goldin, however, recently dug it up. So I’m reposting the entire post below, now with a working link.
Before I get to that, though, here is an illuminating video by Professor Steven Horwitz of St. Lawrence University. He addresses the same subject with a different dataset. Yet he arrives at the same conclusion:
Here is my updated December post:
Every so often, the Census releases new data on income inequality. Typically, the numbers look something like this:
Pretty bad, huh? The chart, derived from data in a Census report, seems to show that the rich are getting richer while the poor are basically stagnating.
Unfortunately, these numbers are not particularly helpful in describing what is actually happening. The problem is that they show a snapshot of groups over time. They do not, however, tell us how individuals’ incomes have changed over time. Maybe people who were in the bottom 20% in 1996 were no longer in the bottom 20% by 2005?
To see the problem, imagine the story were not about income inequality but about mathematical proficiency. Imagine that you saw the following data (which I completely made up).
Here we see that 12th graders grew more proficient in trigonometry over time. But, sadly, 6th graders seem to have been left behind. Few of them knew trig in 1996 and only a slightly larger percentage knew it 9 years later. But wait. Those people who were in 6th grade in 1996 grew up. By 2005, they were out of high school. Wouldn’t it make sense to follow individuals and see whether they knew more trig 9 years later?
We can apply this same line of thinking to the often-cited numbers on income inequality. And when we do, we see that the first chart above really isn’t helpful at all.
Is there a better way? Yes. Consider a Treasury study from a few years ago. It is called “Income Mobility in the U.S. from 1996 to 2005.” Instead of checking in with groups over time, it used IRS data to follow the same people over the course of nine years. The results were quite different than those presented by the first chart.
Among those who were in the bottom 20 percent of earners in 1996, their average incomes had increased by 233 percent just nine years later. Using the IRS data, it appears that low-income people experienced pretty significant income growth from 1996-2005. A totally different picture, huh?
As far as I am concerned, the important thing is to have a dynamic and growing economy with plenty of opportunities for those who are the least well-off among us. We can debate about whether or not ours is that sort of economy—especially given the current economic climate—but a simplistic analysis of income groups gives us very little information on this score.