Finally, these Census data tend to imply that people are stuck in the same income group year after year, when in reality there is substantial upward mobility. Almost everyone spends some time in the bottom quintile when they are starting out in life or because they have suffered a temporary loss. But they move up the income ladder and rebound from setbacks.

 In previous years, the Census Bureau has simultaneously released income mobility data, but it has not done so this year. Earlier data have shown that a large fraction of those who were in poverty or a low income quintile one year were out of poverty or in a higher quintile within a year or two.

 Most people know this even if the Census Bureau doesn't tell them. A 2003 Gallup poll found that 31 percent of the population believes it is very likely or somewhat likely that they will become rich. No doubt, an even higher fraction believes that their children have a good shot at it.

 Indeed, almost every generation of families is better off than the earlier one. Not only have real incomes risen sharply over time, but most people are well aware that they are better off than their parents. A 1996 Reader's Digest poll found that while 18 percent of people considered themselves to be in the lower class as children, only 6 percent put themselves in that class as adults.

 These Census data will probably be grist for the campaign trail, as Democrats charge that Republican tax cuts have worsened the distribution of income. But voters should be aware that these data tell us nothing whatsoever about how taxes affect incomes because taxes are another item excluded from the calculations. All the figures are before taxes.

 Thus not only does the Census Bureau exclude most of what we do to help the poor, it also excludes everything we do to bring down the rich and equalize incomes. In truth, it tells us very little about wealth and poverty in America.