In his April 13, 2011, speech on his vision for America’s fiscal future, President Barack Obama made it clear that, although he agreed to an extension of the President George W. Bush tax cuts in a compromise deal at the end of 2010, he will resume his fight to end Bush's tax cuts for people making more than $250,000 a year.
Obama justified his position, in part, on the following statistic:
"In the last decade, the average income of the bottom 90 percent of all working Americans actually declined," Obama said. "The top 1 percent saw their income rise by an average of more than a quarter of a million dollars each."
This turns out to be a controversial statistic, so we decided to check it out.
According to the White House press office, Obama's statistic is based on an analysis of IRS data on pre-tax income by economists Thomas Piketty at the Paris School of Economics and Emmanuel Saez at the University of California at Berkeley.
According to their analysis, the inflation-adjusted average income for the bottom 90 percent fell by $1,170 between 2002 and 2008 (a 4 percent decrease); while the average income for those on the top 1 percent rose by $261,930 (a 30 percent increase).
The Piketty-Saez analysis is widely cited in economic circles, but it is not without its detractors.
One of the most frequent and vocal critics has been Alan Reynolds, a senior fellow at the libertarian Cato Institute. In a phone interview, Reynolds said that the Piketty-Saez analysis looks at income before taxes, including capital gains, and excludes "transfer payments," such as social security and unemployment benefits.
"That's a fairly problematic thing," Reynolds said, as it leaves out a lot of personal income such as refundable tax credits. "If you leave out transfer payments, you leave out a big part of the story."
In an op-ed for the Wall Street Journal on Dec. 23, 2010, titled, "Taxes and the Top Percentile Myth," Reynolds argued the Piketty-Saez figures "tell us nothing about whether the top percentile pays too much or too little in income taxes."
In a published response to Reynolds' criticism, Piketty and Saez countered that nothing in Reynolds' critique invalidated their findings and that Reynolds' arguments "contain serious misunderstandings on our academic work."
We're not going to wade too deeply into the fray between groups of economists. But we thought it would be helpful to note some of the caveats and assumptions contained in Obama's claim and to point to some other independent data related to the subject of income distribution.
For one, the U.S. Census Bureau provides survey data on mean household incomes, but the data does not allow for a clean analysis of Obama's claim. The census reports on the mean household income received by each fifth of the population (which means they don't look at the "bottom 90 percent"). It also doesn't have a large enough sample to break out the top 1 percent, though it does give data on the Top 5 percent. Between 2002 and and 2008 (which we are citing for the sake of consistency), mean incomes increased in all the fifths of the population (contrary to the Piketty-Saez analysis which found incomes dropped for the bottom 90 percent). The census data also shows that the mean income for the top 5 percent rose from $251,010 in 2002 to $295,388 in 2008. That's not as dramatic a picture as the one portrayed by Obama, but again, Obama talked about the top 1 percent, and the Census figures are for the top 5 percent. Piketty and Saez argue the disparities rise considerably as you narrow the window from the top 5 percent to the top 1 percent.
In addition, the government's nonpartisan Congressional Budget Office presents average pre-tax and after-tax household income for just the top 1 percent, but only through 2007. According to the CBO report, after-tax household income for the top 1 percent went from $1,038,700 in 2000 to $730,500 in 2002, and then to $1,319,700 in 2007. The CBO does not provide 2008 data, where we would expect to see a significant dip due to the recession. In other words, the results depend greatly on which years you choose.
In an analysis of the Piketty-Saez data by the left-leaning Center on Budget Policy Priorities, Hannah Shaw and Chad Stone acknowledged that "the IRS data used by Piketty and Saez have some limitations, one of the most important of which is that they do not include people who do not file income tax returns."
In addition, they note, "neither Piketty-Saez nor Census adjusts incomes for family size, which can distort trends when average family size is changing over time."
The CBO report does adjust for family size, they wrote, and so it "provides the most comprehensive analysis of income distribution, but is not as timely as the Piketty-Saez analysis" because it only goes to 2007 (pre-recession, which hit the wealthy hard).
We'd like to point out one other issue. Obama framed his statistic as "in the last decade," yet his press office referred to analysis between 2002 and 2008. Decisions about what years to include or not include can make a big difference.
When we asked Gary Burtless, an economist with the centrist-to-liberal Brookings Institution, about the accuracy of Obama's claim, he looked at Piketty-Saez figures between 1998 and 2008 -- in other words, the latest data available for a full decade.
"The estimates imply real average income (before taxes) fell 4.2 percent between 1998-2008 in the bottom 90 percent of the income distribution," Burtless wrote in an e-mail. "In the same period they increased 17.7 percent ($136,344) in the top 1 percent of the income distribution between the same years."
If you look instead at the period between 2000 and 2008, Burtless said, real average income (before taxes) fell 7.7 percent in the bottom 90 percent of the income distribution, while it increased 3.4 percent in the top 1 percent of the income distribution between the same years.
"This justifies the President’s statement on the direction of change in pre-tax incomes in the two groups, but I do not know how his staff calculated that the average income gain per person in the top 1 percent exceeded $250,000," Burtless said.
It takes a careful selection of years (and not even a full decade at that), and a number of assumptions about income for Obama's statistic to hold up. We rate the president's claim Half True.