Playing with Statistics

Bill Moyers appeared on The Daily Show tonight and fretted over the rising gap between "the rich" and "the poor." That these terms are arbitrary and vague didn't seem to bother him. Nor did he seem to mind that the numbers are useless. Comparing a gap now and a gap twenty, ten, or even five years ago assumes nothing else has changed.

But lots of things have changed, including the people the numbers are looking at. People are moving higher up in incomes and starting at lower incomes (because of immigration, more schooling, etc). This is a great scenario but Moyers' naive examination of it would suggest disaster. When we follow the individuals (which we don't do enough of so the data is a little old), "....the bottom 20% in 1975 were also in the top 40% at some time in the 16 years follow." (Sowell, p135) Here's a table from Steve Horwitz's page that gives us a more complete picture:

Income Mobility 1975 to 1991 (UM Data)
	Bottom 20% (1991)	Fourth 20%	Middle 20%	Second 20%	Top 20%
Bottom 20% (1975)	5.1	14.6	21.0	30.3	29.0
Fourth 20%	4.2	23.5	20.3	25.2	26.8
Middle 20%	3.3	19.3	28.3	30.1	19.0
Second 20%	1.9	9.3	18.8	32.6	37.4
Top 20%	0.9	2.8	10.2	23.6	62.5