David R. Henderson  

The Uneasy Case for Progressive Taxation

Mail Delivery: Drop Saturday o... The Long-term Debt Issue...

I posted on David Cay Johnston's article on taxation recently. I appreciate his acceptance of my apology and his willingness to discuss the issues. In his comment on another comment by Joe Barnett, Mr. Johnston defends "progressive" taxation, that is imposing a higher marginal tax rate on income as income rises. He writes:

perhaps reading upon marginal utility will help you understand why your analysis is not supported by standard economic theory.

I assume by "upon" in the quote above, Mr. Johnston means "up on." I also assume that what he is getting at is the principle of "equal sacrifice," the idea that the tax system should impose equal sacrifices across people. With diminishing utility of income, this principle does imply higher taxes for the high-income person, but it does not imply a higher tax rate for the high-income person. Indeed, it is consistent with a lower marginal tax rate on the high-income person because even a lower marginal tax rate above some income level will take more money from the high-income person than from the low-income person. [I'm assuming away another problem, which is that we can't do interpersonal utility comparisons.] You need a much stronger assumption, specifically, a very steep decline in the marginal utility of income, to get Johnston's conclusion.

Here's how Arthur Pigou put the issue in his 1951 book, A Study in Public Finance:

All that the law of diminishing utility asserts is that the last ₤1 of a ₤1000 income carries less satisfaction than the last ₤1 of a ₤100 income does. From this datum it cannot be inferred that, in order to secure equal sacrifice . . . taxation must be progressive. In order to prove that the principle of equal sacrifice necessarily involves progression we should need to know that the last ₤10 of a ₤1000 income carries less satisfaction than the last ₤1 of a ₤100 income; and this the law of diminishing utility does not assert.

Walter J. Blum and Harry Kalven Jr. quote this passage from Pigou in their 1954 classic, The Uneasy Case for Progressive Taxation.

Of course, it's possible that I have not stated Mr. Johnston's argument at all.

Comments and Sharing


COMMENTS (25 to date)
Tom West writes:

I think you are quite correct that the law of diminishing returns suggests that $1 of a $1,000 is not as bad as $1 off a $100.

However, I think most people would assert that losing 10% of your income when you are trying to live on $10,000 is a *lot* worse than losing 10% of your income when your income is $100K, and I suspect that that is the assumption that progressive taxation is based upon.

Whether that assertion is correct or not is pretty much irrelevant. Enough people believe it to be so (including myself) and thus it is part of our tax policy.

kebko writes:

If someone making $10,000 qualifies for a $2,000 subsidy that they do not qualify for at $20,000, this is a 20% net marginal tax rate. Considering that most low income citizens receive significant benefits, I think it is useful to note that marginal net tax rates are very high on low income households. The only ways to fix this are to reduce the level of poor-subsidies or to extend them to higher income households.
If diminishing utility is important, then this should be a high priority dilemma. But, I think there are Hansonian issues here.

Curt writes:

David states: "Indeed, it is consistent with a lower marginal tax rate on the high-income person because even a lower marginal tax rate above some income level will take more money from the high-income person than from the low-income person."

Not sure I follow that logic. If diminishing returns on income hold all the way up, then all we know is that the last $1 of $100,000 income is of less value than the last $1 of $1,000,000. We don't really know whether the value of the last $1 of $100,000 is 'the same' as the last $10 or $100 or $1000 or $10,000 out of a $1,000,000 income. So I don't see what kind of conclusion we can draw about what marginal tax rates would inflict the same 'pain' at different income levels.

But from studies I've seen, it seems as if for the most part once people get above a certain 'comfort level' of income, the marginal value of extra income isn't real high. Which is not to say that there's no pain involved with paying those marginal taxes!

R Richard Schweitzer writes:


Where are the arguments for progressive rates of taxation applied to corporations and business enterprise on the basis of the amount of "income" of the enterprise.

Business makes $100,000 rate is 25%

If it earns another 100,000 it's rate on that margin is 30%

and so on

The more it makes, the more and higher rates it should pay as its fair share. You know, like people!

John writes:

Seems to me that this attempt to equalize the sacrifice based on diminishing marginal returns is missing something else as well. This is only looking at the supply side of the case, where is the demand side? The assumption appears to be that the benefit enjoyed is equal for all to whatever the cost that is presumed to have been equalized by the progressive taxation.

I think if one wants to make a case for progressive taxation a more solid basis would be on benefit derived from being part of the society. Clearly those at the bottom of the socio-economic ladder are getting the minimum benefit while those at the top are getting the maximum.

Josh W. writes:


Do you think those at the bottom of the socio-economic ladder don't benefit immensely from living in a rich nation?

How much of a one-time transfer do you think you would you have to pay someone who is poor by American standards to move to a dirt poor country?

Pandaemoni writes:

@ Tom West

I think you are quite correct that the law of diminishing returns suggests that $1 of a $1,000 is not as bad as $1 off a $100.

To take more realistic numbers, imagine a tax burden of 25% on someone earning $50,000 vs. $5 million. The man who earns $50,000 will be left with $37,500...which is livable, but that can easily be woefully inadequate if he has a family and there is not a lot of surplus there. I can imagine that man struggling to save for retirement and other high cost items (like, say, college tuition for his kids). The man earning $5 million takes home $3.75 million each year. It's a little hard to for me to imagine him weeping over his inability to add a second pool to his mansion.

It seems to me that if we were to equalize the sacrifice in those two cases, for most people, The millionaire would pay a greater percentage than 25% (or the man earning $50,000 would pay a smaller percentage).

One way to do that is to introduce a series of escalating marginal tax rates.

Getting the schedule of this tax rates "right" is a pie dream, of course. You'll never be able to actually be confident that you are equalizing the subjective sacrifice...but for me personally, the marginal utility of $1 above about $1 million, is pretty close to zero.

Emerson White writes:

It wouldn't be at all surprising if the last $10 of a $1000 income carried a smaller marginal benefit than the last $1 of a $100 income. However, with payroll taxes and sales taxes and property taxes and medicare contributions do we even have a particularly progressive system of taxation? It was my impression that someone who brought home $100,000 a year paid a significantly higher portion of their paycheck to taxes than someone who brought home $5,000,000. Admittedly I have no data for this, only real basis is what I read in a Warren Buffet interview a couple of years ago.

noiselull writes:

Wieser notably first used this argument in Social Economics. Mises then addressed the argument in Human Action by pointing out the negative effects of decreased capital accumulation. Also see Bob Murphy's study guide to Human Action where he argues that the argument falls flat because of the impossibility of interpersonal comparisons of utility. (All 3 books are available as free pdfs from the MI)

kurt writes:

I agree, focusing on social sacrifice is nice, but if the productivity losses (in the broadest sense) from the additional taxation of someone earning $5 million are even greater, society as a whole seems to be worse off in these models.

ajb writes:

I thought that Mirrlees addressed some of these questions in his theoretical work and that the upshot was that high tax rates (about 25% or so) were sufficiently bad for the economy that even if one wanted redistribution that aggregate welfare went down at rates much above that. So any assumption of high diminishing marginal utility must also be coupled to low negative effects on productivity of higher rates. Plus all this ignores rent-seeking and the transactions costs of monitoring, complying with, evading, and collecting complicated taxes. And I have never seen any serious post Mirrlees work which even begins to address the rent-seeking/TC issues.

David R. Henderson writes:

You're right on all counts. My point is simply that even without Mirrlees insights, Pigou, much earlier, put the kibosh on arguments from declining marginal utility to progressive income taxes.

mark writes:

Thanks to all above for interesting thread. Esp to noiselull for concise summary of a lot of thought.

I have always been astonished that anyone could make the "equal sacrifice" argument with a straight face. The point of interpersonal comparions of MU being impossible is obvious. U is not even measurable on a single person basis. Then there is the intertemporal or life cycle issue - there are poor people who have lived to be 100 and rich people who died at 45 - how do you know everyone's future and factor that into the "equal sacrifice"? Consider that the UK's high income tax rates (and drug policies) influenced John Lennon's relocation to NY - might he be alive today if their tax rates had been less progressive. Yes I know that's hyper-speculative but how do you know what everyone's future is when you purport to identify their present relative sacrifices? I have a friend who is a really highly paid lawyer (>1 million a year) who just adopted two school age children when their mother was found unfit. I think the amount of disposable income they had played a big role in enabling them to reach that decision. How does an "equal sacrifice" scheme factor that in? Even at its most extreme, say Bill Gates vs a really poor person, who can say that Bill Gates is sacrificing less by paying taxes and surrendering control over the spending of his money. Less material consumption, maybe, but it's clearly at the heart of Bill Gates's life to spend the wealth he has built up in a way that he believes with his whole heart is better for others than the way the government spends its money.

Nick Bradley writes:

Instead of marginal tax brackets, you could have a progressive consumption tax that burdens those that can pay more without any of the negative side-effects of progressive taxes: decreased capital accumulation, disincentives to produce, etc.

David Cay Johnston writes:

Mr. Henderson,

Yes, "upon" was a typo and should have been "up on." Thanks for catching that. With age I have acquired some flaw in my copy reading eye that spellcheckers do not help.

The Blum and Klaven law review is a crucial essay to understand tax. I use it in my course on the history of tax and property from ancient Athens to today.

It is also important to read Blum & Klaven in the context of its time. They wrote in 1952 when the top marginal rate was astronomical compared to today.

Still, the authors came out every so slightly on the side of progressive taxation, as has every other classic worldly philosopher including (in the mildest possible way) John Locke.

It us also important to understand the moral basis of progressive taxation, which was crucial to the invention of democracy and is one of the really profound insights of all time.

The Athenians, during the two centuries we call the tyranny that came with the invention of coinage and the massive social disruption coinage caused, concluded that all men are equal.

They choose leaders by lot for every job except strategoi (generals). And those in power were held to strict accountability.

The Athenians reasoned that having more money did not give you any more say in government, which is not at all our practice today.

They also reasoned that without Athens there was no legitimate wealth because in his natural state, in the jungle, there is no wealth. Building wealth requires civilization's rules to define and protect property and adjudicate disputes.

Thus, they reasoned, the greater the economic gain you manage (or inherit) because of Athens the greater your moral duty to sustain Athens. Their system of liturgies (an early form of progressive tax) did not have audits as we think of them, but it did have a clever enforcement mechanism called antidosis, which I encourage readers here to study. My students always find it fascinating, especially when I tell them we used to have a similar concept in our tax code.

Blum & Klaven looked at progressive taxation largely as a distributional issue to deal with inequality and not as who should bear the burden of civilization.

I tend to look at this the way the Athenians raised the issue. I have also shown that our existing system redistributes upward to a very tiny but very wealthy slice through subtle mechanisms explained in my bestselling books on hidden subsidies, FREE LUNCH, and on tax avoidance and evasion, PERFECTLY LEGAL.

In my work I have challenged the current concept of redistribution, suggesting we should view much of what we call redistribution as timing transfers in which you benefit, then pay, then benefit and (unless you ended up in very bad circumstances)continue to pay. The benefits I refer to here are education and health care, which are age related.

A vigorous and civil debate on this would, I believe, tell us whether we are spending our tax dollars wisely to build a solid foundation for private wealth and making sure the liberties of the people endure or are undermining private wealth at all levels from the lowliest hovel to the mightiest skyscraper and also undermining this, the second American republic.

Tax (and how taxes are spent) is a subject which ideological marketers simplify and package with the same disregard for fact that real Mad Men did when magazine ads featured doctors declaring that smoking is good for you and that telling men concerned about their hair and on the prowl that "a little dab I'll do ya."

David R. Henderson writes:

Well done. Your lawyer story reminded me of my friend in Detroit, whom I blogged about a year or so ago, who used some of his disposable income to adopt a black teenager from the Detroit ghetto and send her to one of the premier colleges in the Northeast.

biL. writes:

Others have touched upon this, but I would note that the "diminishing marginal utility of income" argument for progressive taxation is based upon the same logic that supports the supply-side arguments for low tax rates.

A talented surgeon making $400K a year, facing high marginal tax rates would likely opt for longer vacations and performing fewer life-saving operations, since the marginal value of the after-tax income he or she forgoes by not working is lower.

In this example (with relatively inelastic demand), the supply-side effects of high taxes upon our most productive leads to higher medical costs as well - in order to induce the most productive to provide their valuable services, we have to pay them a premium to cover the higher marginal tax rates they face.

ajb writes:

I also object to the Athenian arguments of Johnston because they excessively penalize inequality based on material wealth over non-monetary inequalities. If you say that someone with more money should not have more political influence than those with less, this obviously shifts the locus of power to those who are more rhetorically persuasive, or beautiful, or intelligent, or politically savvy, or physically intimidating.

I don't understand why those inequalities are acceptable, but monetary inequality is not. The other winners also benefit from society but they choose to direct their opprobrium at those who have monetized their talents. In my view, this is the great worm in the apple of even the most idealistic redistributive structures.

The older I get, the more I feel safer in a world in which the bourgeoisie are well protected than one in which attempts to establish wealth equality are not matched by sensible attempts to establish power equality.

Randy writes:

The advocates for higher taxes are always preaching about progressive taxation as if we don't understand the concept. We do understand it - and it doesn't justify unrestrained plunder.

David Cay Johnston writes:

@biL, you are quite right so the question is at what level of tax would behavior be significantly influenced. At 91% for sure. At 35%?

And not all taxpayers respond the same way. When my taxes go up I work more (or smarter), which studies have shown is a common response.

There is a lot of work by behavioral economists showing that some of the assumptions pushed by the Chicago School (where I studied for two quarters in 1973) do not stand up in the real world.

Take someone who has almost purely money motivated, like a hedge fund manager. Their propensity to earns rise with higher tax rates, much as they like everyone else would like lower rates.

We also know from studies done some time ago that the estate tax tends to increase savings, which is intuitive since many people pick a target they want to leave heirs and thus reduce consumption to meet the pre-tax cost (including strategies to reduce the tax cost). The upside is more investment, the downside is less current consumption, but then those subject to the estate tax even when the threshold was $600,000 were not in need of added work to eat.

We also know a large portion of inheritances are gone within 18 months, raising questions about whether the estate tax encourages strivers and encourages expectant heirs to work instead of clip coupons. (See my piece on entrepreneur Eric Zoller at nytimes.com.)

There are also studies showing that a corporate tax rate does not affect investment decisions (which should be based on seeking profit) until one gets to a 60% statutory (which therefore is not actual) rate. I was astounded by that and want to see more research in this area, not the least because the incidence of that tax depends on market conditions.

@ ajb, you make good points. The world is not black-and-white., but it is the law of the land in America that you get one man, one-vote (an idea a lot of people died for).

I have never suggested that monetary inequality is not acceptable. In a market economy we expect people to sort themselves out based on their talent, skill, drive and to some degree factors beyond their control such as luck.

I would argue that we need to give all children a fair shot. A lot of smart and productive people who have created jobs and pay the top tax rate (like me) started out poor (my dad was a 100% disabled veteran and I have worked full-time since I was 13). If I were turning 18 this year in my 1967 circumstances there is no way I would have gone to college.

CBO shows the comprehensive income of the bottom half of taxpayers in 2005 was $15,800. Average state school tuition is now $7,600 so with books and whatnot make that more than $10,000. How many smart minds are we wasting compared to our competitors where tuition is free to modest (max 500 Euros in Germany, for example)? Our competitors have relatively more bright students effectively paid to study full-time instead of working and studying.

Is turning higher education into a heavily government protected financial business ($1 trillion in student debt, consistent annual ROE for biggest lender about 50% compared to 17% for big banks over the long haul)?

The most valuable asset we have is the gray matter between young ears that needs public investments to become rigorous and valuable.

The goal, I think, should be to raise enough tax to create a solid foundation of commonwealth on which private wealth can be built. Superstructures of private wealth built on sand collapse.

Thus we need civil law and its enforcement, education, public health and infrastructure, which in the modern world is not just physical plant like roads and bridges, but a great deal of intangible infrastructure.

A complex society has complex costs that will be paid whether or not we account for them. For example our food borne illness rates have risen as budgets for food safety inspection have been cut and by one measure are now 21 times those in France and 8 times those in Britain. We need to measure whether the tax dollars saved exceed the cost of those illnesses, which I highly doubt.

What we are doing is letting our infrastructure crumble -- we are consuming it to our long-run detriment -- and failing a large portion of our population in terms of education, including many who get higher education.

Our public budgeting does not capture the value consumed or the value added by taxes, giving us misleading signals.

One of my main points in my book Free Lunch is that a large share of top wealth does NOT come from the market, but from hidden subsidies.

My next book will document rules that in some cases literally no one else has written about as a journalist or academic that thwart competition, artificially inflate prices, violate constructive trusts granted to monopolies and other policies that damage the economy, damage social stability (one of your concerns) and provide unjust desserts.

Of course if no one knows about this stuff but the beneficiaries -- in many cases even consumer groups missed the ball -- then the beneficiaries can gorge themselves on these unjust desserts to our overall detriment.

@ Randy, our Constitution places no practical limits on the taxing power of Congress, making it a political issue. Dislike our taxes, vote the incumbents out. It takes only six years for a complete turnover in the House and White House and a dozen to replace a majority in the Senate.

Did you know that 75% or more of capital gains are never taxed? Beware of reports like Simpson-Bowles that ignore this in their data, creating false impressions.

Given the fact that we have seen lower tax rates did not produce the promised growth or jobs -- but plummeting income tax revenue since 2001 (see chart in my column "Awful" at tax.com) -- it is time we get away from simplistic politician claims about tax rates and focus on what the data tells us.

I also recommend to you the charts at the presimetrics blog in tax rates and economic growth.

My next book will document how you can actually have much more money in your pocket and we can increase efficiency by raising your taxes. It will include an actual case where for each buck of added tax I have $1.61 more in my pocket.

How? The focus is on the other side of the equation -- what we use our tax dollars to buy.....

Noah Yetter writes:

"[I'm assuming away another problem, which is that we can't do interpersonal utility comparisons.]"

Don't do this.

You wouldn't "assume away" dividing by zero, would you? Some things are just impossible.

Jim Glass writes:

Mr. Johnston in a previous thread also takes the strange (if common enough in some quarters) position...

@ Robert Simmons, you seem not to grasp the context of my point on Social Security, which was simply that the burden of that tax is born overwhelmingly by people down the income ladder, in contrast to the federal individual income tax.
The logic seeming to be: start with a progressive income tax, add a separate progressive Social Security system (paying no benefits on wages over $108,600), and you've made the net less progressive.

Progressive + progressive = increasing the tax burden "on people down the income ladder".

Well, if SS is viewed as a separate progressive social insurance system, as FDR and every head of the SSA since Altmeyer has presented it, this argument is just nonsensical. Progressive + progressive = more progressive.

OTOH, if he wishes to argue that SS really is just another tax-and-spend program, I could have a good deal of sympathy for him. But then he doesn't protest nearly enough. In that view the payroll tax truly *is* regressive, and he should be protesting that it should be repealed outright and entirely, with SS financed by income tax, with correspondingly increased income tax rates across the board, which as the income tax is seriously progressive would drop most of the cost on the rich.

And he certainly should be protesting the extremely regressive practice of having Warren Buffett's employees at Dairy Queen being charged double-digit payroll tax from their first dollar of wages to pay Warren's Social Security and Medicare benefits. If one can't get rid of the payroll tax dropped "on people down the income ladder", at least stop using it to make direct transfers from them to millionaires.

Yet these ideas have pretty much zilch support on the progressive side of the political spectrum, and while I haven't read everything Mr Johnston has written, I don't recall him making the case for these changes. So I find his observations about the cost of SS taxation, well, inconsistent.

On the larger scale, it's a bit odd to see someone who so clearly detests tax shelters (legal and otherwise) lobbying so strongly for higher tax rates, when they are the driving engine of tax avoidance. As the deadweight cost of taxes rises not with the tax rate but by the square of the increase in the tax rate, increasing tax rates drives up the incentives for taxpayers to buy tax shelters -- and for politicians to sell them -- with literally *exponential* force.

(See the tax shelter industry before and after TRA '86, which dropped the top rate from 50% to 28% -- yet also increased the effective tax rate paid by the very richest relative to the less rich.)

Feldstein estimates we are at a point where increasing personal tax rates "involves a deadweight loss of 76 cents per dollar of revenue" -- and having it rise exponentially from there is surely a non-trivial matter.

If economics teaches anything, it is that people -- including, and especially, politicians -- respond to such incentives. In my reading of tax history there hasn't been any government yet, ever, anywhere, that has found any way to defeat them, any more than anyone has ever found a way to stop the sea from rising, for all one may rage against it.

David Cay Johnston writes:

@ Jim Glass, FICA has been vastly OVERcollected for the last 27 years. At the peak it was half again as high as needed to cover benefits on a pay as you go basis, explaining close to half of the decline in the savings rate of the vast majority.

The program has also fallen from a tax on 90% of wages to 83% while Congress has expanded opportunities to declare compensation for services as capital, as I have documented exhaustively and I do not see any tax economists disputing.

Even so SS is the most soundly financed program in government, with a huge surplus and a future shortfall that would vanish with modest tweaks and better economic performance.

Its future would be sound if returned to the long-term growth in labor income. The SS trustees project future growth rates 15% lower than in the long-term past. That may be overly optimistic, however, as our current policies resulted in FICA taxes in real terms increasing just 4 percent from 2000 to 2010. The median wage was flat throughout the aughts and job creation was smaller in absolute number than during Ike's administration.

Illegal tax shelters flourish when the law is not enforced, as we saw with the severe decline in enforcement at the top of the economy, which reached a nadir during the Clinton administration and then worsened during the GWBush administration, though yeomen efforts went into trying to make the awful statistics look better than reality. For fine details subscribe to tracfed or see its free website at trac.syr,edu or my books and column, which use data backed up by extensive interviewing of IRS field auditors, specialists, supervisors and corporate tax officials.

Finally, there is some research and much empirical evidence cheating increases significantly when tax rates fall (see above).

For sure that is counterintuitive. When I found my first example of this I was stunned and spent two days trying to figure out why what the numbers in front of me showed had to be wrong.

In Florida the intangibles tax was halved. Six years later (during boom years in the 90s) the revenue was unchanged despite overall wealth and population growth.

Again, the behaviorists are making findings that raise questions about the assumptions used in neoclassical models. And heterodox economists who (like me, though I am not an economist) warned of the Internet and housing bubbles have pointed to problems with neoclassical theory that blinds adherents to facts that do not fit the model.

The incentives you cite are widely taught and believed, but a growing amount of evidence shows that they do not work in the real world with the diagram-like efficiency of textbook economics.

We should never cling to assumptions (as with geocentrism) when evidence points to flaws in economic models.

At a minimum we should explore so we don't miss new ideas (like heliocentrism) that may show us how to raise the necessary funds to provide the commonwealth on which private wealth is built, including closely examining what we buy with tax dollars.

Lance writes:

David Clay Johnston,

I would certainly like to see your support for the statement that corporate tax rates do not affect investment decisions until the 60% statutory rate. A simple Google search reveals plenty of articles arguing the exact opposite, not to mention the findings of Charles Jones or Stokey, et., al, given that physical assets depreciate and need to be replaced, and their rate of replacement is determined by factor returns.

The difference between 'Neoclassical' economics (parenthesis because I think you use this term as a political pejorative) and behavioral economics only exists in the assumptions, not the modeling choice. Both theorists use "as-if" modeling where a certain set of axiomatic parameters are used to fit theoretical propositions with empirical data.

Both 'sides' rely upon the Positivist model sketched in Friedman's seminal article on economic methodology. So, rejecting policies you disagree with because they may receive support from economists who use modeling assumptions associated with the Neoclassical school does not seem too robust given your affinity for Behavioral economics.

David Cay Johnston writes:

@ Lance, Its Cay, not Clay.

Nothing pejorative about neoclassical, it is just not the only way to look at the world. And what the behaviorists are showing lacks real worldliness is not neoclassical, but some teachings of that subset known as the Chicago School. Neoclassicals have their well documented blind spots, but theirs is the standard theory.

I am busy finishing my next book right now, but do some research and you will find studies, articles and official government reports on matters such as lower tax rates and higher avoidance/evasion, the 60 percent theory (which I did not say was per se fact, but put out as a counterintuitive idea worth pondering and which I also labeled astounding).

Comments for this entry have been closed
Return to top