Average income for the bottom 20 percent of all households with children was 35 percent higher in 2005 than it had been in 1991.
I am a bit surprised by their findings, because they seem inconsistent with the failure of the poverty rate to decline.
It is important to remember that the bottom quintile in 1991 and the bottom quintile in 2005 are not the same people. The Census followed a specific panel of households, but only for three years. The CBO analyzed this panel, looking at those who were low-income in 2001.
For those households, average inflation-adjusted income increased by about 45 percent, from $16,700 in 2001 to $24,100 in 2003.
The impression that I get is that a household that is in the bottom quintile in any given year has a fairly high probability of being in a higher quintile in subsequent years. That does not surprise me. First of all, there is bound to be some mean reversion in income--some of the people in the bottom quintile in any given year are merely having a bad year that year. Second, for many households, particularly young households and new immigrants, the trend of income is upward.
Again, for me the most surprising result of the study is the fact that the average income in the bottom quintile was so much higher in real terms in 2005 than in 1991. I would have thought that new immigrants would have held down that figure.
I wish that more would be done to explore the dark continent of income dynamics.