Tuesday, June 25, 2013

On Income Mobility

There's been a great deal of discussion about income mobility lately, largely stemming from Greg Mankiw's excellent essay. Many have responded by noting that most measures of relative income mobility put the USA below most other developed nations.

Somebody once explained income mobility to me by comparing it to a race. Relative mobility would measure how the ranking of the race participants changes over time. Absolute mobility (another way to look at mobility) would measure the absolute progress in the race that each of the racers make over time (ignoring the ranks).

I would caution against reading too much into relative income mobility data. It would make sense that societies with more equal income distributions would have higher relative income mobilities. To go back to the race analogy, the various racers are much closer in proximity to each other in a more "equal" race than in an "unequal" race. This means that much smaller changes could lead to larger changes in the various ranks in the race.

The point isn't to discount relative income mobility data. It is to say that data purporting to show that other countries have higher income mobility than the USA is really just more data showing that other countries have more equal income distributions than the USA. The supposedly higher mobility is really just a statistical artifact of greater equality. Of course, absolute mobility data has drawbacks as well, so I am not recommending looking at that in isolation either (namely the fact that it mostly reflects average income growth).

Income mobility is hard to measure. Instead of using very flawed methods of measuring, we should focus on institutional barriers to upward mobility. I would categorize licensing, high implicit marginal tax rates, and monopolies as such.
Post a Comment