≡ Menu

Relative Mobility

In my recent posts on the challenges of measuring inequality, I wrote about how hard it is to compare income distributions over time because family structure and immigration among other factors have changed dramatically in recent decades.

I provided one kind of evidence on mobility that followed the same people over time, one way of purging the data from distortion due to other factors.  I summarized the findings of one of those studies by Katharine Bradford and Jane Katz at the Boston Fed, used the
Michigan PSID survey.  PSID stands for Panel Survey of Income Dynamics, so it should be good for studying, well, income dynamics.  Here were the key finding again that the New York Times headlined, Mobility Loses Steam. 

What percentage of families during the
’70s, ’80s and ’90s stayed in the same quintile of the income

1970        36%
1980        37%
1990        40%

What percentage moved up or down one or more quintiles:

1970       38%
1980       39%
1990       39%

What percentage moved up or down two or more quintiles:

1970        26%
1980        24%
1990        21%

So it appears that there is less mobility.  Whether the change is small or large is subjective.  But my point before was that you can’t tell whether mobility has decreased because we have a more static economy or whether it’s because everyone is moving up making harder to increase one’s relative standing.

But now that I’ve thought about it a bit more, I realize that even these numbers have some of the same problems of the snapshot over time comparisons I discussed earlier.

The PSID follows one group of people over time.  New families are added to the survey as children grow up and create their own families or as people divorce and remarry.  So it’s not really the same people being followed over time.  But even if the sample is restricted to people who are in the sample from start to finish (and I don’t know what Bradford and Katz actually did in getting the results reported above), you can’t use the above numbers from the  PSID to evaluate whether relative mobility is rising or falling.  The reason is that the sample is aging over time.  I wouldn’t expect old people to have as much earnings mobility as younger people whose careers can have more dramatic increases or decreases as initial career paths succeed or fail.

What the PSID is good for is measuing absolute mobility.  Are people doing better than they did before or worse?  I’m sure people have looked at that using the PSID, I just don’t have it at my fingertips.  When I find it, I’ll post it.


Next post:

Previous post: