Inequality Deniers Fudge the Numbers Again

Honest businessman Photo: Aldra/iStockphoto

Since the presidential campaign is revolving in large part around the question of whether or not it’s fair to increase taxes on the richest Americans, the right is circulating its cherished pseudo-facts to press its case that the rich are already uniquely overburdened with taxation. Ubiquitous right-wing misinformation-recirculator Veronique de Rugy, in the Washington Examiner, hauls out the classic standby of citing the amount of total taxes paid by the rich and pretending this represents the tax rate paid by the rich:

Contrary to common belief, the United States already has a more progressive tax system than do the most industrialized democracies worldwide. The nearby chart uses data from a recent report by the Organization for Economic Cooperation and Development on the share of taxes (both personal income and payroll taxes combined) paid by the richest 10 percent of households in 24 industrialized countries. The bars represent the share of the total taxes collected that are paid by top earners in these 24 countries.

The richest 10 percent of U.S. households (those making $112,124 or more) contribute a greater share of taxes (45.1 percent of all income taxes) than their counterparts in any other industrialized nation.

Meanwhile, the average tax burden for the top 10 percent of households in OECD countries is 31.6 percent of the revenue collected, well below the percentage in America.

The flaw here, of course, is that de Rugy’s statistic has nothing to do with the point she attempts to make on its behalf. She claims the U.S. has a more progressive tax system than other countries. “Progressive” means the degree to which a tax system increases tax rates on higher-income earners. But her figure only shows that our rich people pay a larger share of the tax burden than the rich of other countries. But the fact that rich Americans pay a larger chunk of the total tax burden than rich people in other countries doesn’t mean rich Americans are paying a higher rate. They may be paying a higher share of the taxes because they have a higher share of the income to begin with. Which, in fact, they do.

I could not find a comprehensive comparison of the tax rate paid by the richest Americans against all other industrialized democracies. I did find a paper comparing the U.S. to the United Kingdom and France, and the U.S. tax rate for the top one percent (and the whole top 10 percent) is considerably lower.) Our rich people pay a bigger share of total taxes because, despite their low rate, they earn so much more of the income pie.

If we actually try to take de Rugy’s argument at face value, she is insisting that we can’t ask the rich to pay any more taxes because they earn so much money to begin with. It would follow, I suppose, that if something were to reverse the trend toward rising inequality, and the rich started earning a lower share of the income pie and paying a lower share of the taxes, that de Rugy’s claim would imply that we should necessarily increase their tax rate. After all, if the proper measure of a tax burden is what proportion of the total taxes a group pays, and a high proportion means they already pay too much, than a lower proportion would mean they don’t pay enough. Obviously de Rugy doesn’t believe that. Just as obviously, the role data plays in her argument is purely decorative. Of course, de Rugy is merely repeating a canard that has been floating around the right-wing misinformation chamber for years and years. I continue to be dumbfounded at the low intellectual standards of a movement which allows obvious nonsense like this to play such a major role in its intellectual case.

While we’re on the topic, the inequality denial community is up in arms over my post the other day about the Charles Murray dodge. Murray has a book arguing that declining social norms have caused working-class white people to fall behind educated elites. Inequality deniers are very excited about this argument, because it allows them to change the subject from the question of the skyrocketing share of income held by the top one percent to the smaller but also growing gap between the top 20 percent and those below.

The American Enterprise Institute’s James Pethokoukis, who considers “rising income inequality” the greatest fraud since global warming, has roused himself into a fit of uncomprehending wrath:

New York magazine’s Jonathan Chait doesn’t much like Charles Murray’s new book, Coming Apart: The State of White America, 1960-2010. In particular, he doesn’t like that Murray links the increasing social polarization of America to the abandonment by poor and working-class whites of the”Founding Virtues” of industriousness, honesty, marriage and religion. ... But that’s not what Chait wants Murray to write about. Chait doesn’t want Murray to write about culture. Chait wants Murray to write about rising income inequality as the driving force behind pretty much all of America’s troubles.

Sorry, no, that’s not what I wrote, at all. Charles Murray can write about the indolence of the white working class all he’d like. My point is that other pundits should stop pretending that an analysis of the gap between the bottom 80 percent and the top 20 percent is the same thing as an analysis of the gap between the top one percent and the bottom 99 percent. As I wrote, “Even if we grant, for the sake of argument, all the claims being made on Murray’s behalf, the basic point is that it is not a plausible response to the problem of income inequality.”

Likewise, the blog Political Calculations — whose work Pethokoukis has cited as refuting the “myth” of income inequality — is unhappy as well. My post pointed out that its supposed refutation of rising inequality is erroneous, because it relies on census data. The Census Department does not collect detailed information about rich people’s income, which is why inequality researchers look elsewhere when they want to study changing income among the very rich. Lane Kenworthy, an actual expert in this topic, helpfully explained the folly of the Political Calculations chart. I thought his explanation was too detailed to be of interest to readers here, but since they’re complaining, I’ll reprint a longer excerpt of his e-mail to me:

Over the period since 1994 the Census Bureau's standard Ginis for households and for families, which are the ones used in the Political Calculations chart, suggest little or no change in income inequality. So too do Ginis for earnings inequality among full-time year-round employed individuals (see IE-2 at

But it's well-known that the Census Bureau data miss what's going on at the top, because they "top code" very high earnings and incomes. What the Census data tell us, and what the Political Calculations post in effect simply reiterates, is that within the bottom 99% there has been little change in income inequality since 1994, whether we're looking at households, families, or individuals.

We can ask "Why begin in 1994?" But the bigger problem is ignoring what happened at the top of the distribution.

During this particular period, income inequality increased due to the top pulling away from the rest. According to the Census Bureau's calculations for households, there was essentially no change in inequality from 1994 (Gini .456) to 2007 (Gini .463). (Again, these are the data for households used in the Political Calculations chart.) The best source of income data for households is the CBO, which merges the Census household survey data with IRS tax record data. According to the report the CBO put out a few weeks ago, household income inequality including the top 1% actually increased sharply between 1994 and 2007. The CBO report is at; see Figure 11 on p. 20; the closest CBO counterpart to the Census income measure is "market income plus transfers". Eyeballing the CBO chart, it looks to me like the Gini goes from .45 in 1994 to about .52 in 2007. That's a large increase.

Pethokoukis triumphantly concludes, “Chait’s theory of the case has come apart.” In fact, “Chait’s theory of the case” also happens to be the consensus view of every expert who actually studied the issue and knows which data to examine.