U.S. taxes are progressive

Here is the opener, by the excellent David Splinter:

U.S. federal taxes are progressive, as shown by Congressional Budget Office and Tax Policy Center estimates, with average tax rates increasing with income. In fact, the OECD (2011) estimated that the U.S. has the most progressive household taxes among developed countries. Moreover, the 2017 tax reform is expected to have little effect on overall tax progressivity (Tax Policy Center, 2017; Joint Committee on Taxation, 2019). But Saez and Zucman (2019) argue that average tax rates are nearly equal over the income distribution. To examine this claim, this paper compares other estimates of average tax rates by income group,all of which suggest a high degree of progressivity. Three issues are found to bias the average tax rates presented in Saez and Zucman (2019). After correcting for these issues, their estimates align more closely with other estimates that show U.S. taxes are progressive.

Here is the full (short) piece.  See also this Jason Furman thread.  Here is the Kopczuk thread.  There is a great deal of Twitter malpractice on this issue circulating these days.


Twitter malpractice aside, the serious claim that I have seen is that US *marginal* tax rates are consistently ~40% across income brackets. That this paper (and the Jason Furman thread) would specifically look at *average* tax rates suggests bait-and-switch to me.

Don't economists think at the margin?

Margin speaks to the incentives, not to the burden. For that matter, average rate doesn't really capture the burden either. For that you just want to know about the amount paid. Look at it that way and you'll realize that a small fraction of the country is paying for everyone else.

U.S. income taxes are progressive for 50% of the citizens and non-existent for the rest. The free loaders should all be paying taxes too.


At one point during the Great Recession approximately 50% of Americans paid no federal income tax. It is now 44%.

In any case, taxes are too damned high and government spending is even more too damned high.

If we're truly talking burden, then it's not really just the amount paid, but the value of that money to the payer. The closer to a minimum livable wage someone makes, the more each dollar is worth to them, in terms of how that money will be spent (on necessities versus non-essentials, or investment). The burden of a dollar in taxes is higher for the poor than for the wealthy.

A relatively small fraction of the country is paying for a relatively large fraction's share of government because the small fraction has disposable income available, while the large fraction is barely making ends meet. You can debate if this is 'fair' or not, but it is the only way to make our current economic and government system functional.

"barely making ends meet": Would that include an iPhone, a car, tattoos, drugs, booze, cigarettes. As someone who grew up poor it is incomprehensible that anyone can claim to "barely making ends meet" and indulge in these luxuries.

Meh. Nobody’s starving. It may be your judgment that the $600 for a TV when you’re at the poverty line is higher value to the payer than $6000 for vacation to a teacher or $60000 for a car to a lawyer, but I don’t see it.

Besides: the poor are net recipients of government. Almost everyone is a net recipient of government as looking at amount paid makes clear. You have to be earning, sustained over your career, in at least the top decile to be paying in more than you get out. So there is no burden at the low end. There may be distortion because marginal rates mess with rational behavior, and because the government is providing you with goods and services that you would not choose to buy if you were spending your own money, but in dollars and sense there isn’t a burden.

Ha. Dollars and cents.

Language was never meant to be written down.

Obvious fake news. As I linked yesterday, the Nytimes Opinion author David Leonhardt conclusively proved this was not the case in his editorial. Every Quintile Pays the Same Average Rate. The lowest income 400 households pay more taxes than the highest income 400 households.

Twitter malpractice is just another word for “disagrees with the Kochs”


>The lowest income 400 households pay more taxes than the highest income 400 households.

This is false.

False. And ridiculous.

Here's the link to NY Times article: https://www.nytimes.com/interactive/2019/10/06/opinion/income-tax-rate-wealthy.html

Highest 400 households pay taxes at about the same rate as the lowest 400 households.

So if you want to avoid taxes you should either get very rich or very poor.

Here's Kevin Drum:


He says: "The effect of the latest Republican tax cut shows up in the 2018 line : the very richest Americans now pay a total effective tax rate of 23 percent. The poor and the middle classes, by contrast, pay about 25 percent. As you can see, the tax rate for the rich has been dropping steadily for half a century."

Bottom line for me: US tax system used to be progressive but it isn't anymore. Or at least not for the very rich.

And those who argue otherwise have been sold a bill of goods by the rich or are rich themselves.

This stuff is so obvious why do people blind themselves to this truth?

Because it's not a "truth" and you are cherry picking data that conforms to your ideological bias?

I am not poor, but I am not very rich either. I asked my investment advisor if there was any way I could reduce my tax rate to the tax rate Mitt Romney paid.

He said: "No way."

I will waste a little more of my time on this conversation.

What is the tax on long term (> one year) capital gains income? It maxes out at 20% of the gain. Is 20% less than max payroll rate? You betcha. Max ordinary income rate is 37%.

Are capital gains subject to SS and Medicare taxes? Apparently not.

Who are best able to convert ordinary income to capital gains?

Rich people.

Wake up.

I can't stop.

Who takes advantage of off shore tax shelters?

Rich people.

I assume SS is a confounder in that its progressivity kicks in with the benefits and not the tax itself?

Oh gag, this misleading garbage yet again?!

Right, US federal taxes are by and large progressive, although not at the very top end because of fica. But overall taxes in the US taking into account state and local taxes are only barely progressive, given the regressivity of sales taxes and to some extent property taxes given that landlords in most places are able to pass those on to renters.

We have been through this many times. Why on earth are you posting this misleading nonsense yet again?

BTW, the income level that pays the highest average taxes overall is in the mid-90s in terms of income percentiles. The system is still mildly progressive overall, but it is massively not so at the top end, and it gets worse and worse as one moves further out the spectrum on that far top end.

This is just a shameful post.

Finally some sanity. Thank you Rosser.

Even with Medicaid, Food Stamps, the EITC, the free lunch program, CHIP, and Section 8 the net tax burden is barely progressive. Barely.

The average household who earns under $40,000 pays almost $20,000 in net taxes.

>The average household who earns under $40,000 pays almost $20,000 in net taxes.

CITATION NEEDED. This is a silly claim.

Even the leftwing "Citizens for Tax Justice" agrees that the tax burden is progressive and that nobody in the bottom half of the population is paying half their income in taxes. https://www.ctj.org/who-pays-taxes-in-america-in-2015/

It does not include the property tax which is completely passed to renters, nor the pollution externality tax (lead anyone) that the poor pays all of, nor the inequality imputed tax which the poor suffer under.

It adds up to over 50% in average tax.

So if you make up an arbitrary set of fake taxes you can manufacture the results you prefer. Seriously?


That was amazing satire, thank you.

"property taxes given that landlords in most places are able to pass those on to renters." I can see that the incidence might well fall on renters; has it been shown to be true empirically? In which places is it untrue?

The facts are not on your side and you are pounding the table. Where is the evidence that fica, state and local taxes make US taxes only mildly progressive? State income taxes are progressive (highest marginal rate at single digit percentage in almost all states) or if they're flat, they're 5% or lower. In some states there is no income tax at all. Sales taxes are roughly 10% in some localities, but some states exempt things like food and clothes. In some states there is no sales tax at all. As for property tax, the highest effective rate is roughly 2%. All of these state and local taxes pale against effective total federal tax rate (including fica), where the lowest 40% pay single digit percentages and the top 20% pay more than 20% of their income.

OK, I have now read all the links, and the paper does make an effort to take into account all taxes. In the main paper, see Figure 3. There are competing estimates, Saez-Zucman versus Auten-Splinter, with the latter finding substantial progressivity still for all taxes, although there are no comparisons at all with other nations in any of the links, so the main claim made in Tyler's post is not supported at all.

In this Fig. 3, there are two Saez-Zucman estimates for all taxes, and one of them finds the average tax rate falling for the top group. Their estimates show only mild overall progressivity, with one showing regressivity at the top. A-A find more progressivity, although it seems very unclear how these are differing, plenty confusing.

There are differences over the impact of property taxes, but there is no question that in general together state and local taxes are regressive, although in five states they are progressive. Sales taxes are the main reason, which is the largest source of revenue for states.

A source on this is itep.org/whopays .

BTW, going back to the original post by Tyler, the OECD report does not include state and local taxes. It is almost certainly not the case that once those are taken into account that the US has "the most progressive tax system," my original complaint.

" given the regressivity of sales taxes"

All dollars of wealth will be hit by sales tax. If not, why would you care that they exist? They are earnings which have not been converted into consumption and been donated to society.

That you pose as an intellectual is shameful.

so what is the ethical & constitutional basis for taxing an $80K income person more than a $40K person... for the same level of government services?

+1 Average rate isn't the right thing to look at. Amount paid is the right thing to look at.

"so what is the ethical & constitutional basis for taxing an $80K income person more than a $40K person"

Because more votes are for sale among those who earn $40K than among those who earn $80K. There are other motives possible but I suspect that's the one that matters to the political class.

Stunning that people do not understand, or claim not to understand, the purpose of tax progressivity. Civilization is regressing. Title that comes to mind? Barzun’s “From Dawn to Decadence: 500 Years of Western Cultural Life”.

That was a fantastic book, +5 internet points

Same level of service? Will fire fighters in your area save an entire $200,000 home, but when a $10 million dollar mansion catches fire they let $9.8 million of it burn away before putting the flames out?

Thread winner, +5 internet points

There's an important assumption that Auten/Splinter make differently from Piketty/Saez/Zucman that I have a question about:

"First, wage and business income in national income data exceed amounts reported on tax returns by over a trillion dollars. Both PSZ and AS include these untaxed underreported amounts; the PSZ estimates, however, over-allocate underreported income to the top. This is because PSZ allocate underreported income using an ad hoc assumption that such income is proportional to reported source-specific income, whereas AS rely on representative IRS audit studies that are the basis for including these amounts in national income. These studies consistently show that the ratio of underreported income to reported income declines for higher levels of reported income, while the PSZ approach effectively assumes the opposite. Difference sin the allocation of underreported income have large consequences for top incomes."6

Does it really seem plausible that richer people are LESS likely to underreport their income than poorer people? I've seen many articles about how declining IRS funding means that rich people are less likely to be audited, and even if audited, also less likely to have to pay more (e.g. https://www.foxbusiness.com/money/irs-audits-poor-taxpayers-easier).

They justify this with a cite to Johns, Andrew, and Joel Slemrod. 2010. “The Distribution of Income Tax Noncompliance.” National Tax Journal 63(3): 397–418. https://www.ntanet.org/NTJ/63/3/ntj-v63n03p397-418-distribution-income-tax-noncompliance.pdf?v=α&r=8924300134538536

The Johns/Slemrod article does back this assumption up, but it seems like a lot depends on how they bin results (e.g. Table 3). Their earlier work (e.g. https://www.forbes.com/2008/10/21/taxes-irs-wealth-biz-beltway-cz_jn_1021beltway.html#495cd2487dae) also doesn't seem consistent with the "rich people underreport their income less" hypothesis.

Does anyone else find this odd?

Being in the bottom of the distribution doesn't make you "not rich". It just means you don't have a lot of taxable income on your tax return. If you have a very successful cash-only food truck, restaurant, liquor store, etc. you could be both "cash rich" and "income tax poor" due to under-reporting.

"Does it really seem plausible that richer people are LESS likely to underreport their income than poorer people?"

Yes it does. When you are high income, your income is from corporate entities that all file with the IRS. There is little opportunity for white collar folks to work for cash like the poorer people can. You'll give you a cash discount, your accountant, or your landscaper?

Do lawn crews have to file tax returns? I mean, "effectively"?

For obscure reasons I once spent a couple weeks' worth of days in a dreary Fiesta store, and besides the surprise (to me) of how many people came in to gaze at, or pay layaway on, the gold chains in the gold chain store-within-a-store, was how many people came by each day to dump a wagon's worth of coins in the coin changer machine.

The man who runs a coin-operated washateria, say, and faithfully records his take for tax purposes, would, I think, be the man Diogenes searched for.

And how significant is the amount that illegal immigrants with fake IDs pay into social security, that they are unable to recover?

Is this a well-known figure?

If anyone knows of a good book on the shadow economy, I'd be interested.

Posted tax rate vs effective tax rate. I doubt the OECD would rate America as the most progressive by taxes actually paid. America is a great place if you have ambition and obsession with a career/business goal. If you want to live a fair and normal life, you are better off in any other developed country.

Looking at inequality as measured by apparently recent Gini Coefficients, I see Australia and Canada are now at around .34 while the US is .42. That's a hell of a jump. Australia's is actually .35 which is extremely high for us. I don't remember it being above .33.

The elephant in the room is always capital income. If you add capital income to labor income the wealthiest pay much lower marginal tax rates, hence Warren Buffett paying a lower rate than his no-doubt well-compensated secretary. There is an argument to be made that capital income is taxed more than once, because of taxes on business profits, but nevertheless the low taxes on capital income and the absence of taxes on capital (other than some limited estate taxes) make it much easier to stay rich in America than it is to become rich, which is fundamentally illiberal.

I would heartily support making dividends a deductible business expense, and taxing all capital income (dividends and capital gains) at the same progressive rates as wage income.

Further, or alternatively, I support a tax on capital of 1-3% that would strongly encourage the investment of wealth in productive assets. Too much is parked in low-productivity investments.

"Too much is parked in low-productivity investments."

What do you have in m ind?

"Too much is parked in low-productivity investments."

Wow! I had no idea.

How do I put money into your Mutual Fund? Also, I have a lot of friends that are looking for higher returns. I'm sure that they would be very interested in investing as well.

Without having any data to back my claim, I honestly swear that if I was rich I'd under report to everyone... except my wife, of course...

Does anyone find this odd???

It's hard to under report with W2s and capital gains that are recorded on the other side by a regulated financial institution. It's easy to under report cash income.

Yes, I don’t see an avenue for much under reporting of normal non-cash income. I’d be curious to see some common examples from those who think otherwise.

I could see opportunities for those with complex financial affairs to perhaps change the character of income, in ways that defer or reduce taxes. Hollywood accounting seems to be notorious for inventive schemes, for example.

Very clear and well written. The fact that people can't even swallow the simple fact that American taxes are in fact progressive is amusing.


no, but several prominent US mega billionaires complain of being under-taxed.
(of course, they could easily donate extra millions direct to US Treasury to soothe their shame -- but never do so)

Of course, the tax debate started with the Reagan revolution, in which income taxes paid by high income earners were cut while payroll taxes paid by everybody else were increased. Besides making taxes less progressive, the Reagan revolution had the bonus of turning ordinary folks into tax protesters because their "taxes" were going up; to most folks, "taxes" are what's withheld from their paychecks, and the increase in payroll taxes greatly increased the amount withheld. Another part of the Reagan revolution, the "strong" dollar, facilitated the shifting of production overseas; thus, while working Americans were paying higher "taxes", their jobs were being shifted overseas.

I'm not opposed to Cowen's wealthy friends paying less in "taxes", but I do oppose an economy that is prone to financial and economic instability, which is what we have today because of the high level of inequality: the concentration of wealth producing a depressed level of return on productive capital producing investment chasing higher yield through greater risk. I'm with Cowen's Austrian friends on this, who would not send the Fed to the rescue in the next crisis to inflate asset prices, but instead would lett asset prices (and inequality) collapse. Cowen has indicated that he has been influenced by the Austrians, yet he supported the "bailout", the euphemism for inflating asset prices and restoring the level of inequality to the level in 2008. I suppose I am grateful that the great depression was avoided. For a time. Inequality means another crisis ahead. Will the Austrians prevail next time? Or the wealthy?

Your arc here has been fascinating. I never woulda guessed full on "2008 redux sandwich board raving dude", but here we are.

There were quite a few deductions made non-deductible under Reagan, like deducting auto interest and sales tax. Tax shelters were closed, too.

There weren’t just cuts in the rates.

You never spent any time in my working-class household during tax time.

My mother would have at least 1 grocery bag full of receipts from the previous year to add up and deduct. This was pre-calculator. We were lucky, we had an adding machine.

I was curious so I looked it up and found this:


It has some nice simple graphs that even an Australian could follow. It looks like US taxes are progressive until you get to around the highest income 5% and then goes very regressive.

This must be very embarrassing for the top income 5%. People might think they are only rich because they're rich. Hope it gets fixed soon.

Most of the dip at the very highest income levels are due to capital gains taxes being below top other income taxes, and so when people realize very large long term gains on assets, they tend to max out their incomes for that year, and also get a lower average tax rate.
There are good economic reasons for treating capital gains differently, associated with incentives to savings and investment. So why not strip out that effect from the tax analysis and debate capital gains tax treatment separately?
The capital gains tax issue also illustrates what a willful liar Warren Buffet is when he says that he's never heard of business decisions affected by tax rates.
"I have worked with investors for 60 years and I have yet to see anyone — not even when capital gains rates were 39.9 percent in 1976-77 — shy away from a sensible investment because of the tax rate on the potential gain."
Oft repeated by brainless leftists, but I'm in the investment business and constantly sees investment decisions distorted by capital gains tax considerations.

Indeed: by what miracle could income taxes affect investment decisions yet capital gains taxes don't? Has money ceased to be fungible?

I have done more digging and also seen the latest WaPo story coming out on this. Furman and ITEP do show the top 1 percent paying more in total taxes than other groups in average rates, although not by all that much. Overall system still best characterized as "modestly progressive," as I initially claimed. I was wrong however that the highest payers are in the mid-90s levels. It is higher, but it does not keep going up.

Eventually as one moves up to higher levels within the top one percent there is a peak, and it beings to drop due to a variety of things ranging from fica and sales taxes to capital gains taxes and a variety of other loopholes only available to the super rich.

This is emphasized by story in WaPo yesterday drawing on Saez and Zucman data showing that while the bottom 50 percent pay an average overall tax rate of 24.2 percent, the top 400 families, roughly the billionnaires, pay an average rate of only 23 percent. The main criticism of Saez and Zucman is them not counting EITC, which certainly helps people at the much lower end, although not all of them are eligible. Taking that into account, it may turn out that the billionnaires are paying a higher overal rate than the bottom 50 percent. But it is also certain that they are paying a lower overall rate than high income people who are not at their super rich level. It remains true, as I initially posted that at the top end the US tax system is regressive, even if it is unclear exactly at what point that average rate of payment stops rising and starts falling to the substantially lower rates paid by the billionnaires.

The link for the WaPo story is washingtonpost.com/business/2018/10/08/first-time-in-history-us-billionnaires-paid-lower-tax-rate-than-working-class-last-year .

The story notes, which I am sure is correct, that there has been a noticeable decline in the average rate of payment by the billionnaires as a result of Trump's tax bill, lowering their rate by more than 2 percent.

Hey, are gambling taxes included when working out the tax burden? They are a big chunk of government revenue here in a Australia and lower income half of the country contributes a greater portion of their income to them than the higher income half.

The Australian state and federal governments collect close to $200 US per person per year in gambling taxation. That includes babies and stuff. So either our infants are very precocious or there are adults making up the difference.

Comments for this post are closed