What would you say if someone told you that many academics, the US government, and the media overstate income inequality, understate the real income growth of US households, overstate poverty, and understate income mobility? If someone had asked me, I would have said I believe it. I’ve followed these issues, and even written about most of them. But on reading The Myth of American Inequality: How Government Biases Policy Debate, even I was blown away by the strength of the evidence for these conclusions.
The book’s three authors are former US senator and former economics professor Phil Gramm, Auburn University economics professor Robert Ekelund, and former assistant commissioner of the Bureau of Labor Statistics John Early. The authors take a deep dive into the data and use largely government-generated data to make their case. They point out that in computing household incomes, the US Census, part of the Department of Commerce, systematically leaves out two-thirds of the transfer payments that federal, state, and local governments give to people. This dramatically understates income of people in the lowest-income two-fifths (which economists and statisticians call quintiles) because these two quintiles, and especially the lowest, receive a hugely disproportionate share of transfer payments. The Census also leaves out taxes paid to federal, state, and local governments. Because higher-income people pay most of the taxes, failure to subtract these taxes substantially overstates the income of higher-income people. Both factors cause the Census Bureau to systematically overstate income inequality. They also show that the government’s usual measure to adjust for inflation, the Consumer Price Index, systematically overstates inflation and, therefore, understates the growth of real wages and real household incomes. Adjusting the data for both transfer payments and the overstatement of inflation, the authors show that the percentage of US households in poverty, rather than being in the low teens, is actually only about 1.1 percent.
Along the way, the authors show that US income mobility is high: the vast majority of people, over their lifetimes, move from one quintile to another. They also dispel a number of myths about the rich, the top 1 percent, the top 0.1 percent, and the incredibly wealthy Forbes 400. As a disturbing bonus, they show that in recent years some federal government agencies have encouraged people to be more dependent on government welfare.
This is from David R. Henderson, “Myths of Economic Inequality,” Defining Ideas, November 3, 2022.
The results of correcting for inflation with a measure that adjusts for substitution and for quality improvements:
Gramm et al. note that while using the CPI shows a measly 8.7 percent growth in real wages between 1967 and 2017, using a measure that corrects for substitution bias and for improvements in the quality of goods and services yields the conclusion that real wage rates over those fifty years rose by a whopping 74.0 percent. Over that same period, using the CPI shows that real median household income rose by 33.5 percent, but using a price index that accounts for substitution and quality improvements shows the increase to be 93.3 percent.
That seems to accord with our observations. Think about what people have now that they didn’t and how those things have contributed to their well-being: houses with two bathrooms and widescreen TVs, mobile phones that can be direction finder, music player, calculator, and more, and medical care that improves our life expectancy and keeps us from having long stays in the hospital. Those are only three of many improvements.
Read the whole thing.
Thanks to one of the co-authors, John Early, for quickly answering a question I had about 2 of the tables in the book.
Update: Here’s a nice post on the issues by John Early.
READER COMMENTS
Daniel Kuehn
Nov 4 2022 at 11:02am
Isn’t it well known that post-transfer inequality is (almost, but not quite, by definition) lower than pre-transfer inequality? The reason why people care about pre-transfer inequality is because it tracks market income and because it’s more easily compared across policy regimes. To observe that transfers eliminate a lot of it is simply to observe that transfers are doing what they’re supposed to be doing.
The official poverty rate, as I understand it, does include some transfers although not all and it is based on pre-tax income.
David Henderson
Nov 4 2022 at 2:02pm
You write:
Yes.
You write:
That’s one reason. It’s not the only one. And of course, as I’m sure you recognize, the huge transfer payments discourage market income, especially for the lowest 2 quintiles, which is one of their points.
You write:
True. And one of their main points is that the Census Bureau should be counting this when looking at things like poverty.
You write:
Exactly. And that’s one of their major points. The Census Bureau excludes 2/3 of transfers and so the official poverty rate way overstates poverty. You did read my article, right Daniel?
Daniel Kuehn
Nov 4 2022 at 4:05pm
I read your EconLib post not your article.
Daniel Kuehn
Nov 4 2022 at 4:08pm
In any case there’s obviously nothing wrong with post-transfer distributions, they’re really valuable to see how things look on net. And they’re produced particularly in discussions of the effect of the safety net. But by the same token there’s nothing wrong with pre-transfer distributions either, and I would not want to see them replaced by post-transfer distributions. Pre-transfer distributions I think give us the best understanding of the level of inequality generated by labor and capital markets.
David Henderson
Nov 4 2022 at 6:33pm
You write:
It gives us part of the understanding. Particularly for the bottom 2 quintiles, the low labor income reflects the huge amount of transfer payments.
Jose Pablo
Nov 6 2022 at 1:55am
Pre-transfer distributions I think give us the best understanding of the level of inequality generated by labor and capital markets.
No, pre-transfer distributions give us an understanding of the level of inequality generated by labor and capital markets affected/modified by the existing transfer distributions.
Without the existing redistributions labor and capital markets could be significantly different.
Daniel Kuehn
Nov 6 2022 at 11:02am
David you write: “It gives us part of the understanding. Particularly for the bottom 2 quintiles, the low labor income reflects the huge amount of transfer payments.”
Well naturally – that’s why I had said both measures are valuable. But your initial post said that the pre-transfer measures “overstate” inequality and poverty, and that’s just wrong. The fact that a pre-transfer measure and a post-transfer are different does not mean that the former overstates inequality any more than it means the latter understates it. They’re measuring different things. Nor did the Census “fail” to include anything. If your goal is to generate a pre-transfer measure then you didn’t “fail” to account for transfers you correctly excluded them (otherwise it wouldn’t be a pre-transfer measure).
My point is that your description here treats the pre-transfer measures as if they’re inherently wrong rather than simply measuring a different thing from the post-transfer measures.
Post-transfer measures only give us “part” of the understanding as well because they don’t pull out the different components.
Thomas Lee Hutcheson
Nov 5 2022 at 9:33am
How certain are you that “huge” transfer payments discourage market income among the lower x/yth. That would seem to be a really tough relation to identify given that transfers are given because of low market incomes.
Jose Pablo
Nov 6 2022 at 1:58am
The statement “recipients of transfer payments are discouraged to participate in the labor market” is very plausible.
What’s the part of the statement that you don’t get?
vince
Nov 4 2022 at 1:46pm
From the original article: ” These federal transfer payments were Social Security Old Age and Survivors Insurance ($10,958), Medicare ($7,986), Social Security Disability ($2,575), Medicaid and the Children’s Health Insurance Program ($9,634), food stamps ($1,504), and miscellaneous programs ($9,126). ”
A recipient wouldn’t get both Medicare and Medicaid. Are these amount averages over the entire quintile? Does miscellaneous programs include ACA subsidies?
David Henderson
Nov 4 2022 at 2:05pm
You write:
I’m pretty sure that’s false. Many low-income seniors get both, mainly in the form of subsidies to be in nursing homes.
You write:
Yes.
You write:
I’m not sure.
vince
Nov 4 2022 at 2:33pm
ACA is a huge benefit. A 40-year-old couple making the average AGI of $45,000 in 2022 would pay $12624 for the standard silver plan with no subsidy. With the subsidy, they pay $1800. That’s a subsidy of $10824.
About 25 percent of the population is enrolled in Medicaid. We virtually have national health care.
vince
Nov 4 2022 at 2:37pm
From the main article: “In 2017, the top 1 percent paid about 40 percent of their income in taxes.”
According to IRS data, in 2017 the top 1 percent paid federal income tax of less than 27 percent on their Adjusted Gross Income. How does that get to 40 percent?
David Henderson
Nov 4 2022 at 2:45pm
That undercounts their taxes. It leaves out federal excise taxes, their share of corporate income taxes, state sales and excise taxes, and, most important, state income taxes.
Kevin Corcoran
Nov 4 2022 at 4:07pm
I think you’re conflating the claim “they paid 40% of their income in taxes” with the claim “they paid a 40% income tax rate.”
Vivian Darkbloom
Nov 4 2022 at 5:41pm
Another difference is that AGI and taxable income are not the same thing. For the top 1 percent a big difference would be that charitable contributions are subtracted from AGI. The actual income tax paid as a percentage of taxable income is necessarily higher than against AGI for that and other reasons.
Bill
Nov 4 2022 at 8:11pm
On tax “fairness” in the book review, you ask, “Can anyone seriously argue that the top 1 percent got commensurate benefits from governments in the United States?”
The Tax Foundation explores this issue here:
https://taxfoundation.org/distribution-tax-and-spending-policies-united-states/
vince
Nov 5 2022 at 1:43pm
Who is protecting the wealth of that top 1 percent? Word searches for FBI and military in the article turned up nothing. Don’t get me wrong–both articles provide some shocking information about redistribution.
nobody.really
Nov 6 2022 at 7:59am
Maybe rely on the argument of revealed preference? If anyone has the power to emigrate from the US, surely the top 1 percent do. The fact that they don’t seems suggestive.
Kevin Corcoran
Nov 7 2022 at 2:54pm
The fact that they don’t emigrate suggests that they gain more benefit from living in the US than not, but that doesn’t really do any work to demonstrate the top 1 percent therefore have obtained “commensurate benefits from governments” specifically. (One could reach that conclusion if you add in the auxiliary premise that all benefits from living in the US are exclusively flow through the US government, but one would need to make that premise explicit to make a valid (though still unsound) argument.) There are lots of other reasons other than commensurate benefits from government which would influence ones decision to emigrate – family attachments, a sense of community, business connections to maintain, or just sheer inertia. It could very well be the case that the lack of emigration is despite a lack of commensurate benefits from government, but this net negative in government benefit is outweighed by other factors. Or not! There’s an interesting discussion to be had there, but just saying “revealed preference!” doesn’t really contribute much to that potential discussion.
As an analogy, it’s almost a trope at this point that various high profile leftists in America (usually Hollywood types) vow to leave the country if a Republican is elected to the presidency. This was particularly true in the lead up to the 2016 election. And yet, almost none of them actually followed through. Maybe you’d say this shows by “revealed preference” that they actually gained “commensurate benefits” from Trump being president – after all, millionaire Hollywood actors and singers can certainly afford to emigrate, and almost none of them did! But to me, that just shows an impoverished understanding of what the whole idea of revealed preference actually is, and how it works.
Mark Brady
Nov 5 2022 at 7:23pm
1. To what extent do these sort of analyses take tax and subsidy incidence into account?
Certainly a tax on the site value of land is borne by the landowners, but to what extent are income taxes borne by those who write the checks to the government?
2. To what extent do these sort of analyses take the degree of legitimacy of property rights into account?
For example, to what extent do (exclusive) patent rights that prevent other independent inventors from earning rewards for their inventions (for the term of the patents) confer illegitimate incomes on the patent holders? And to what extent does occupational licensing confer illegitimate incomes on the license holders?
3. To what extent do government tax revenues finance truly public goods (nonexcludable and nonrival) and to what extent do those revenues finance private and club goods (excludable)?
And to what extent do different categories of taxpayer benefit from government-funded public goods?
Daniel B
Nov 6 2022 at 7:01pm
Another book for the reading list 🙂
Although I have a question about “the missing workers” section. I’ve heard it said that the Census Bureau’s statistics on the amount of low-income people working don’t account for unreported income, so those working under-the-table are not counted. I’m just wondering if the book deals with this issue?
For what it’s worth, I think this issue doesn’t really change anything. Even if more low-income people are working then we thought – they’re just working secretly – why are those low-income people working secretly and falsely reporting their income to the government? Because otherwise they would lose welfare benefits.
Evasion is not free. For example, a book called The Human Cost of Welfare tells how welfare recipients have to declare the amount of money in their bank accounts to the government (p. 12). Not keeping cash in your bank means you have to keep it in cash. But that risks theft and the loss of the money due to fire, makes it more difficult to build credit (the book never explains why but one reason is that it will make credit cards more inconvenient and difficult to use), and of course prevents the earning of interest (conventional banks like Wells Fargo give pitiful amounts of interest, but HYSAs offer noticeably more interest).
So even if there is a lot of evasion (which I doubt), that still wouldn’t change the fact that low-income people face an implicit tax for working due to the welfare system. It would only change the issue of whether they are working much if at all (which is an issue I and probably many other people are interested in).
Comments are closed.