Income Mobility

Out in the woods
Or in the city
It's all the same to me
When I'm driving free
The world's my home
When I'm mobile
--The Who

Nice little missive warning against concluding much about 'income inequality' from distribution of income studies like the one recently issued by CBO.

An important limitation of income distribution studies is that they group earners by category. For example, earners are often filed into one of five categories, top 20%, next 20%,..., bottom 20%. Thus, when CBO reports change in income for the top 20% compared to change in income for the bottom 20% over a given period, it is tempting to conclude that individual incomes in each group changed, on average, by that amount.

Such an approach is prone to faulty conclusions because it ignores the influence of income mobility. Income mobility measures the extent to which individuals move between categorical brackets.

Citing findings from a recent income mobility study by Treasury, the author demonstrates that many individuals tracked over periods of time move in directions opposite those suggested by the static distribution of income studies. In the Treasury study, for example, incomes of those in the top one and five percent in 1996 were actually lower in 2005, while those in the bottom 20% in 1996 saw their incomes increase on average by 91% in 2005.

Static income distribution studies miss the dynamic effects of income mobility as the situations of individuals change over time.