The Chicanery Behind Inequality Data

[MJ]

[ARTICLE LINK]

Aug 29, 2025
TGIF:
The Chicanery Behind Inequality Data
Sheldon Richman
 

If self-described progressives decry anything more fiercely than poverty, it is income and wealth inequality. Some have even suggested that they would prefer low-income equality to inequality, regardless of how affluent the lowest level was. What counts is the gap.

The terms poor and low-income are relative, of course. We’d be better off talking about the poorer and lower-income. Also, it’s better to be poor in America than anywhere else if we factor in immeasurables such as good prospects. However, some people don’t understand the point or perhaps don’t want to understand it. Reasonable people ask, “How am I doing and how can I do better?” not, “How much less am I making than Jeff Bezos and Elon Musk [but not Taylor Swift or Juan Soto]?”

So let’s talk about inequality­not in the legal and political sense but in terms of income and wealth. You can’t go a day without hearing politicians and commentators complain about the top 1, 10, or 20 percent. Those complaints seem to be backed up by government statisticians and parts of the economics and sociology professions. Dissenters are rarely invited on television and podcasts. The impression given, to which compassionate laypeople will be vulnerable, is that America is riddled with extreme, even obscene (so Bernie Sanders says) inequality. Is it true?

Economists Phil Gramm and Donald Boudreaux make an overwhelming case against it in their book, The Triumph of Economic Freedom: Debunking the Seven Great Myths of American Capitalism . Gross inequality is one of those myths. (Last week I discussed their chapter on poverty.)

“[T]he claim that income inequality in America is high and rising on a secular basis is almost universally accepted as true,” Gramm and Boudreaux report. But: “Census numbers overstate the difference between the top and bottom quintile household incomes by over 300 percent.” Can they back up that claim? Let’s see.

The U.S. Census Bureau tells us that in 2017 the average income of households in the richest quintile (the top 20 percent) was 16.7 times greater than the average in the lowest quintile. No one can say, without other information, whether that number is appropriate or not. But is it accurate?

“The official Census data also show,” Gramm and Boudreaux continue, “that income inequality has grown on a secular basis and, by 2017, was 22.9 percent higher than in 1947.” That’s not all. According to the Organization for Economic Co-operation and Development, the United States has the worst record on this count among the wealthy countries­and it’s been getting worse.

Let’s pause for a word about the morality of income and wealth inequality. Individuals contribute unequally to the production of wealth, which improves living standards even for those who contribute little or nothing. So why would anyone expect their incomes and wealth to be equal? Now back to our regularly scheduled program.

Gramm and Boudreaux disclose a puzzle about the government’s numbers: “According to the official statistics of the nation’s two leading statistical agencies, the bottom 20 percent of American households had an average income of $13,258 in 2017 yet, in that same year, consumed $26,091 of goods and services.”

This fact raises the obvious question of how the bottom 20 percent of households can consume twice their income. This extraordinary gap between the official measure of income and the official measure of consumption has grown more or less steadily since 1967, when funding for the War on Poverty began to ramp up.”

That indeed is a puzzle. Could it be that the government agencies do not count everything that’s relevant? Write Gramm and Boudreaux:

[T]he Census Bureau does not count two-thirds of all transfer payments to the recipients as income [88 percent for the lowest quintile], instead counting only $0.9 trillion of $2.8 trillion of government transfer payments. In addition, the Census Bureau neither adjusts household income for taxes paid nor counts tax credits as income received by the recipients, even though they receive checks from the Treasury. Census does not count food stamps as income, despite beneficiaries receiving debit cards to pay for groceries. Also not counted as income are benefits received from Medicaid, under which the government pays for each beneficiary’s health care. And also uncounted as income are the transfer payments dispensed through more than one hundred other federal, state, and local programs.

In other words, the government understates the incomes of the poorest, while overstating the incomes of the richest, ignoring that America has a generous welfare state (coercively paid for) and the most progressive income tax in the world. That strikes me as a rather shoddy way of estimating income inequality.

“Because the Census Bureau excludes $1.9 trillion of transfer payments as income received and fails to count $4.4 trillion of taxes paid as income lost to taxpayers,” Gramm and Boudreaux write, “the Census measure of household income ignores some 40 percent of national income, which is either gained in transfer payments or lost in taxes.”

The advocates of even more confiscation and distribution do not want to acknowledge what is really going on. Why not? The resulting distortion is scandalous. The authors show that instead of the officially estimated top-versus-bottom income ratio of 16.7 to 1, the real ratio is 4 to 1, a fourth of the often-lamented official estimate.

“But even these numbers for household income,” Gramm and Boudreaux write, “overstate income inequality by failing to account for differences in the number of individuals living in the average-sized household of each income quintile.” The average top-level household contains more people than the bottom-level household (3.10 versus 1.69). (See the book for details.)

What has happened to inequality over time, considering that transfer payments and taxes have increased? Gramm and Boudreaux report:

Over the seventy years from 1947 to 2017, after adjusting for inflation, the real value of all transfer payments grew 212.2 percent, faster than earned personal income had grown. Taxes grew less dramatically, rising only 7 percent faster than earned income over these seventy years. Income and payroll taxes rose 21 percent faster than income. Sales, excise, and property taxes rose 8.3 percent slower than income. The net result was that the US income-tax system became significantly more progressive in the seventy years leading up to the COVID-19 pandemic as an ever-larger share of the tax burden has been shifted from low- and middle-income households onto higher-income households, reducing income inequality. [Emphasis added.]

They also show that while the standard international method for measuring inequality, the Gini coefficient, indicates an increase, “much of this increase was due simply to two very significant changes made in the way the Census Bureau collects and records data.” The Bureau acknowledges that those changes distort the picture, but it does not adjust accordingly.

In fact, “the [adjusted] Gini coefficient is actually slightly lower today than it was in 1947…,” Gramm and Boudreaux write. “America’s Gini coefficient falls to a level roughly in the middle of the seven largest developed countries.”

I began by denying the importance of income equality to human wellbeing. As Gramm and Boudreaux explain, more equal does not mean richer: “Major developed nations that have more equal distributions of income than the United States have significantly lower incomes overall….  [They] also have larger portions of their populations that are poor.” Beware a fixation on equality. Better to agitate for the repeal of government obstacles to the creation of wealth, starting with taxes on savings and investment.

A closing note for envious readers who despise the 1 percent. Gramm and Boureaux write that according to a study titled “Income Inequality in the U.S.: Using Tax Data to Measure Long-term Trends” (2024), “when all transfer payments and taxes are counted, the share of national income going to the top 1 percent of American households is about the same as it was in the mid-1960s.” [Emphasis added.]

Of course, there is no such thing as “national income,” as Gramm and Boufdreaux would agree. That’s a statistical category. In reality, there is only your, my, and their income.