Economic Inequality: Facts, Theory and Significance

Policy Reports | Economy | Government | Social

No. 312
Tuesday, June 10, 2008
by David R. Henderson

Are the Rich Getting Richer while the Poor Are Getting Poorer?

If we cannot infer from the data on increasing inequality that the rich are getting richer while the poor are getting poorer, what is true?

Those who get their information from New York Times columnist and economist Paul Krugman can be forgiven for concluding that the poor — and everyone else below the top 10 percent — are worse off.  But Krugman cannot be forgiven because it is not true.  In a 2004 Nation article, Krugman wrote:

According to estimates by the economists Thomas Piketty and Emmanuel Saez — confirmed by data from the Congressional Budget Office —between 1973 and 2000 the average real income of the bottom 90 percent of American taxpayers actually fell by 7 percent.9

But as economist Alan Reynolds has shown, Krugman’s statement is wrong for two reasons.10 First, Congressional Budget Office (CBO) estimates go back only to 1979.  Second, the CBO data show that from 1979 to 2000, average after-tax income in each quintile of the household income distribution rose. 11 In 2003 dollars, for the lowest quintile it rose from $13,500 to $14,600; for the second-lowest quintile it rose from $27,300 to $30,900; for the middle quintile it rose from $38,900 to $44,700; for the second-highest quintile it rose from $50,900 to $63,300; and for the top quintile it rose from $89,700 to $151,100. 

“The after-tax income of every income group rose from 1979 to 2000.”

Thus, in order for Krugman’s claim to hold true (the CBO reference aside), the average income of the bottom 90 percent would have had to have fallen drastically between 1973 and 1979 to more than offset the later increase.  Reynolds uses U.S. Census data to show that no such thing happened.

As for the Piketty-Saez study that Krugman and many others have cited, Reynolds points out just how implausible their data are as a measure of family income.  Piketty and Saez write that in 2000, “the median income, as well as the average income for the bottom 90 percent of tax units is quite low, around $25,000.”12 Note the use of the term “tax units.”  “Tax units” are not the same as families.  A single family, for example, could have two tax units:  a husband and wife filing jointly, and a child filing on his own.  But that has not stopped Krugman and others from writing as if “tax unit” and “family” are synonymous.  Reynolds points out that if tax units were the same as families, highly implausible implications would follow.  For instance, given the meaning of the word “median,” for Krugman’s claim to be true, 45 percent of families (half of 90 percent) would have had to make less than $25,000 in 2000.  However:

  • U.S. Census data show that the median family income for 2000 was $50,732, which means that 50 percent of U.S. families made more than $50,732.
  • If half of all families made more than $50,000, and half of the “bottom 90 percent” made less than $25,000, the remaining 5 percent of the families not accounted for (100 percent minus 50 percent minus 45 percent) would have had incomes above $25,000 but below $50,732.

It is implausible that the income distribution would be so skewed.  The problem is the association of "tax unit" with family.

Furthermore, the way Krugman and others have relied on the Piketty-Saez data on family welfare is completely at odds with their criticism of using tax data to estimate income mobility.  In 1992, the U.S. Treasury published a study which showed that taxpayers in one income quintile in 1979 were highly likely to have moved to another income quintile by 1988.  In other words, there has been substantial income mobility.  But, in 2002, Krugman wrote:

The restriction to individuals who paid taxes in all years immediately introduced a strong bias toward including only the economically successful; only about half of families paid income taxes in all ten years.  This bias toward the successful was apparent in the fact that by the end of the sample period the group contained very few poor people and a lot of affluent ones: indeed, only 7 percent of the sample were in the bottom quintile by the sample's end, while 28 percent were in the top quintile.13

This is an important criticism.  But caution applies to Krugman’s use of data on taxpayers over time.  Just as one cannot use taxpayer data to infer families’ income mobility, one cannot use “tax unit” data to estimate how well families in various income classes have done.

One other point is important also.  Truth be told, we really have no idea what the true income is of the highest-income people because so much of it is in the form of capital gains.  An increase in the price of an asset is income to the owner.  Capital gains are reported only when taxpayers sell their assets.  If they don’t sell their assets, they could have a huge income and we would not know it by looking at published government data.  Someone could have a zero reported income but a true income of $1 million if his assets appreciated by $1 million in a particular year.

Economic Growth.  One factor noted by Reynolds that causes increasing inequality is simply economic growth.  When the rate of economic growth is high, workers see large real increases in income.  But in 2001, 72.6 percent of the money income received by the bottom quintile was in the form of government transfer payments, excluding all noncash transfers such as Medicaid.14 In boom times, people in the bottom quintile do not get large increases in real income because much of their income is transfers from government.  The bottom quintile stay in place or move up slightly while all other quintiles do better due to higher pay and more work.  But this simply reflects a booming economy, not that the lowest quintile is doing worse.

Of course, if high economic growth is accompanied by inflation, even moderate inflation, many people in the bottom quintile could do worse because welfare payments are not indexed for inflation but adjusted only with a lag.  (This group would not include Social Security recipients because their benefits are indexed to inflation.)  But it is important to attribute the worsening condition to its real cause — inflation — not to economic growth, and certainly not to the “ill-gotten” gains of those with higher incomes.

“The top fifth has a lot of income from capital gains, while the bottom fifth has government transfers.”

Ignoring Consumption.  The issue of inflation brings us to another point:  The Consumer Price Index (CPI) has exaggerated inflation for at least a few decades.  Economist Michael Boskin estimates that the CPI overstates inflation by 0.8 to 0.9 percentage points annually due to three factors:  1) failure to account for substitution among goods, 2) failure to account for substitution among retail outlets — the so-called “Wal-Mart effect,” and 3) failure to account for new products.15 This overestimate does not “add up” over time — it “compounds up.”  Thus, over 20 years, a constant 0.9-percentage-point bias overstates the increase in the cost of living by 19.6 percent.  Over 32 years, it overstates the increase in the cost of living by 33.2 percent.

Moreover, before the Bureau of Labor Statistics made adjustments in the late 1990s, the CPI exaggerated inflation by more than 0.9 percentage points annually.  This means that even when “inflation-adjusted” incomes tend to be stagnant, they have actually risen.  One way to see this is to look at what people have — the kinds of clothing and shoes they have, the kinds of appliances they have, the kinds of cars they drive, and so forth.  Michael Cox and Richard Alm found that on virtually all of these dimensions, the life of virtually everyone in the United States improved between the early 1970s and the mid-1990s.16

Read Article as PDF