David R. Henderson  

Reply to Scott Sumner on 80% Marginal Tax Rates

PRINT
What If Firms Could Opt Out of... The correct reason to oppose h...

On his other blog, The Money Illusion, my co-blogger Scott Sumner argues for an 80% marginal tax rate on consumption for high-consumption people. I admit my surprise at seeing Scott advocate that, which is why it has taken me a little while to process it.

What is Scott's argument? He writes:

Some conservatives questioned my assumption of very low marginal utility of consumption at very high levels of consumption. I find it completely implausible that ultra-rich get significant marginal utility from extra consumption, apart from the satisfaction of doing better than other billionaires. But note that this sort of "nah-nah, I have a bigger boat than you do" utility would not be at all affected by a tax regime that applied to all billionaires.

Here's an example. A billionaire might get a great deal of satisfaction from a 400-foot yacht if his rival billionaire has a 300-foot yacht. There is data that shows happiness increases all the way up the income scale. So I do buy that argument. But I would insist that roughly the same enjoyment would be gained from a 300-foot yacht if his rival had a 200-foot yacht. If an 80% consumption tax reduces each billionaire's consumption proportionately, then could it really impact their happiness? Perhaps I'm missing something, but that claim seems preposterous. That doesn't mean you necessarily want to go to the very top of the Laffer curve, as there are deadweight losses associated with high MTRs. But unless I see good empirical evidence that those losses are important in the very, very, very tiny labor market for the ultra-rich, I'll continue to assume that standard public finance theory plus utilitarian values implies that the MTRs on the consumption of the ultra-rich ought to be fairly high.


I don't know which conservatives he has in mind because he doesn't provide links and, as Scott well knows, I'm not a conservative.

Now let's consider his argument. Scott claims that such high marginal tax rates do not "really impact their happiness." Why? It's because, in Scott's view, very wealthy people get their utility from outdoing other very wealthy people. Scott seems very confident of that. After all, he calls the rejection of that idea "preposterous."

So if he's so confident, it follows that he would think that very wealthy people would have no problem with such a high-tax policy as long as it applied to all wealthy people. This means that if only very wealthy people were allowed to vote on such a policy, they all should vote for it.

Here's what I wrote in response to Cornell University economist Robert Frank's proposal for a very high tax rate, and it applies to Scott's argument:

Frank argues that consumption taxes on higher-income [substitute high-consumption to apply to Scott's proposal] people make them no worse off because, as noted, they care about relative income [consumption], not absolute income [consumption]. And, presumably, these people put at least a tiny positive value on the things government would spend the additional revenues on. Is Frank open to testing his assumption? Here's a test. If he's right, a majority of those high-income [high-consumption] people, indeed a supermajority, would vote in favor of higher taxes on themselves. Frank should propose a referendum on whether to raise tax rates on high-income [high-consumption] people, with only high-income [high-consumption] people able to vote. If he is right, such a referendum would pass overwhelmingly. I predict that such a referendum would go down in flaming defeat. If I'm right, then the whole empirical basis of his argument is wrong.


Comments and Sharing


CATEGORIES: Taxation



COMMENTS (25 to date)
Greg Heslop writes:
[...]completely implausible that ultra-rich get significant marginal utility from extra consumption[...]

I'm surprised at this. What happened to the impossibility of comparisons between people? One reason to pile up lots of money for consumption later could be that marginal utility for the person doing it is very high, even at "ultra-rich" margins, compared to the marginal utility of similar increments for poorer folks. Maybe that is why some piles get so high?

Jason writes:

David, you appear equally confident that the opposite is true. And you provide only an untestable hypothesis as proof. You may be right, but this post does nothing to convince me.

In addition, it's not clear to me why the proposal would pass by an overwhelming margin in the case where the super-rich are completely ambivalent as to whether they spend their marginal income themselves or if the government spends it for them.

BC writes:

DH makes a good point. Mankiw pointed out a few years ago that we already face 80-90 pct marginal tax rates. See http://www.nytimes.com/2010/10/10/business/economy/10view.html?_r=0

At least, those are the tax rates that savers pay, since our tax system penalizes savings by taxing the same income/consumption multiple times when one saves. People that spend their earnings immediately face lower rates. If we were to replace our current system with a consumption tax and made the change revenue-neutral, it's unclear to me what the top marginal rate would have to be, but at least that top rate would be determined by consumption and not by who saves the most.

Ashwin writes:

@David,

Have you considered the possibility that the ultra-rich would not vote for such a tax regime, because they incorrectly believe that their utility is derived from absolute consumption, but that after the imposition of this regime, they will realise that they were wrong and that they are just as happy as before? Presumably that is the sort of argument that Reich and his ilk would put forth in response to your proposed test.

Enial Cattesi writes:

I can not help but to note the emphasis on "the bigger yacht." Why not other spending? Larry Ellison recently made a donation equal with the total of the rest. It was deliberate: he asked how much money the organizer collected, it was about 4.5 million, and then he said "OK, I'd like to equal that." Clearly he took great pleasure in this and he should be taxed for it.

@Ashwin:
In other words the billionaires should realize there is a bigger pleasure in paying taxes then in buy stuff for yourself. But only them ...

Sean writes:

That wasn't really my take-away from Sumners' posts.
I think his real point is that for any given level of public spending then the optimal tax regime to maximise social welfare is that tax code that serves to reduce peoples' consumption the least. Overall, according to some utilitarian definition. I think he's right - most consumption above say £10m a year is probably mostly on positional goods, and we all then pay more for land, cars, and everything else.

Tom writes:

This surely doesn't follow David. Sumner argues that marginal utility is *very low*, not negative or zero. So the referendum would be defeated, because although the MU of cash for rich people is very low, it's positive.

Jeff writes:

That's interesting, because Scott just posted something here about how the line between investment and consumption is blurry for the very rich; e.g., owning an NBA team or piece of beachfront property can be profitable and really fun at the same time. I think there's a lot of utility there that has nothing to so with status. Sure, NBA ownership is an exclusive club, but look at how much money wealthy boosters throw down on college sports for which they get no financial return at all. Likewise, look at how much people will pay to rent beachfront property for a week or two.

Also, if space tourism ever takes off, a seven figure lunar vacation I suspect will have a lot of value well beyond just bragging to your friends about it. Expensive medical procedures, too, like Peyton Manning's stem cell injections to fix his messed up neck.

Handle writes:

Is my home a consumption good on which I should pay a hefty annual MRT, or is is a capital good, or is the imputed income actually income, and should thus be taxed at the income tax rate like I believe they do in Switzerland? Or should I just pay the ad valorum property (i.e. wealth) tax, like most folks do in the US? It's not always so easy to tell what's going on.

At any rate, in terms of 'optimal form of taxation', I agree with Sumner in abstract principle that 'tax the relative-status rat race amongst the rich' seems to be the most efficient if you buy the assumption that utility of the rich is hardly affected by the imposition of those high rates. The utility of the people who make the things that the rich buy, however, is a different matter.

But the problem is distinguishing between rat-race purchases and others.

So, for example, it is funny that Sumner uses the example of yacht. I have a neighbor who is one of the most thrifty and status-indifferent people I know and whose dream it is, upon retirement, to cash in his decent (but by no means 'ultra-rich') lifetime savings and sell his home and buy a large boat so he can pursue him dream of living in it while sailing around the world.

As with the question above, is this now a durable, capital good which is more like a real property 'investment' because he can live in it, resell it, and give it to his kids as an inheritance, or is it consumption? Yes, the depreciation and maintenance, are high (we all know the boat jokes), but they are still a small fraction of the purchase price.

Does it matter that had he consumed slowly and steadily over his life, instead of saving, that he would have been taxed at a much lower rate, but that because he saved and chose to enjoy his consumption in the form of one large purchase, he's going to get hit a huge MRT, even though it had nothing to do with 'status'?

It seems to me that such a tax plan would seriously harm the utility of people who had preferences and status indifference levels similar to my neighbor. I think where Sumner errs is that he is arguing for a case in which high consumption MRT's wouldn't matter that much, but not admitting that not all large purchases are instances of that case.

Ken P writes:

@Enial:
Great point. How could a bigger yacht provide more utility?

joshua writes:

I argue this is not a good test because for the same reason the rich want bigger boats, they are opposed to being singled out for high taxes. A better test would be are they willing to apply high MTRs to those with, say, 10x their income. Also the debate about high MTRs can be about whether the rich "deserve" their income (see Elizabeth Warren or Barack Obama) which is unrelated to marginal utility.

Bostonian writes:

All people, including the rich, have the moral right to spend most of the money they have earned. If you don't believe this you are not a libertarian but a collectivist.

David R. Henderson writes:

@Ashwin,
Have you considered the possibility that the ultra-rich would not vote for such a tax regime, because they incorrectly believe that their utility is derived from absolute consumption, but that after the imposition of this regime, they will realise that they were wrong and that they are just as happy as before? Presumably that is the sort of argument that Reich and his ilk would put forth in response to your proposed test.
No, I hadn’t. That could be. So then have a provision that, 5 years after the tax is imposed, only the very wealthy get to vote in a referendum about whether to repeal. If they are just as happy as before, they’ll vote NOT to repeal. I’m not trying to convince “Reich and his ilk,” unless by “his ilk” you mean Scott Sumner.
@Tom,
This surely doesn't follow David. Sumner argues that marginal utility is *very low*, not negative or zero. So the referendum would be defeated, because although the MU of cash for rich people is very low, it's positive.
Good point. But I took seriously, in a way that you apparently didn’t, this statement by Scott that I quoted above:
If an 80% consumption tax reduces each billionaire's consumption proportionately, then could it really impact their happiness? Perhaps I'm missing something, but that claim seems preposterous.
@Jason,
And you provide only an untestable hypothesis as proof.
One can certainly imagine such a referendum. It’s unlikely but one can imagine it. But you’re incorrect in saying that I provide it as proof. I don’t claim to have proved anything. It’s a thought experiment to get Scott Sumner and others who agree with him to think through their proposal a little more.

RPLong writes:

This is a pattern that appears from time to time: Assume the worst about a population of people, and then use that fact to argue for an otherwise morally objectionable outcome. I don't know why Scott Sumner believes that all the rich care about is one-upping each other, but it seems like an unsavory assumption to me.

Even if true, it still would not justify imposing an 80% marginal tax rate on anything. That it is possible to extract more tax money from the population without compromising their utility doesn't mean you should do it. That you can lie and not ever get caught and never hurt anyone doesn't mean you should do it. That you might (and I concede that Sumner does not) believe in an expansive social safety net does not mean that the money to fund it must come from steep taxes on people who are assumed to have no better motivation than keeping up with the Joneses.

That, to me, is the real problem with the idea.

Ken P writes:

@Joshua
Being singled out for higher taxes would signal status. If high consumption is truly only about status and not utility then it would meet that goal and they should be fine with the additional taxes

JWJ writes:

Jobs would be lost by builders of 400 ft yachts and other providers of high-end luxury items.

The govt would collect more money and spend it wisely?

Ashwin writes:

@David,

Sorry, I meant to say "Frank and his ilk", not "Reich".

Christopher Renner writes:

I can see 2 major things wrong with Sumner's proposal. First, the actual behavior of the very wealthy (in demonstrating a preference for avoiding taxes and giving to charity) suggests that they wouldn't find that an 80% tax rate increases their utility.

Second, given the history of other taxes, it's not likely that the 80% MTR would be restricted to billionaires for very long.

MingoV writes:
If an 80% consumption tax reduces each billionaire's consumption proportionately, then could it really impact their happiness?
I'd be extremely unhappy if a government penalized me for buying whatever I want.

I don't understand the purpose of a high consumption tax. Are billionaires only supposed to invest in stocks and bonds? Is it bad for billionaires to support the boat building industry by purchasing big yachts?

This is yet another double tax. Tax earnings and then tax spending. Tax earnings and then tax interest, dividends, and capital gains even when the rates of return are negative when corrected for inflation.

Why don't we make it simple and confiscate earnings above five million per year? Who needs billionaires?

MikeP writes:

A lot of commenters appear not to have followed the threads to their base.

First, there is no tax except this consumption tax. That's the point of it.

Second, it is a feature, not a bug, that this may disemploy the yacht building industry.

The key notion behind this thinking is that people who get paid for services but don't consume the equivalent of their pay are simply in trade deficit with the rest of the economy: a good thing for the rest of the economy. The rest of the economy got the services of the savers without using up scarce resources for the consumption of the savers. Those resources were then cheaper and better able to fill the needs those less well off in the economy. Instead of one yacht, we get 20,000 cars.

I am by no means sold on this idea. But it is certainly a hell of a lot better than what we have today, where a typical middle-income saver faces a 90% tax by the time his earnings are consumed.

JVM writes:

Scott is a pure utility theorist. According to utility theory, consumption has diminishing returns. Therefore marginal utility of addiitonal consumption is least for those who consume the most.

It follows that taxing high-consumers diminishes global utility the least of any tax, since any other tax would destroy more utility for the same gain.

You don't have to agree with utility theory or value global utility, but Scott does and his reasoning is extremely easy to follow. Whether high consumers would themselves prefer this outcome is neither here nor there.

RPLong writes:
Whether high consumers would themselves prefer this outcome is neither here nor there.
I think what you mean to say is, "Whether high consumers would themselves prefer this outcome only matters if their utility decreases more than anyone else's increases."

It's not a silly misunderstanding, it's a question of whether Sumner's utility calculus is accurate enough to warrant what most moderate-utilitarians see as an immoral outcome.

Scott Freelander writes:

Here's a potential problem for your perspective:

"CNBC survey shows millionaires want higher taxes to fix inequality"

http://www.cnbc.com/id/101634240

David R. Henderson writes:

@Scott Freelander,
Well said. It’s a potential problem.
Solution: Let millionaires, and only millionaires, vote on it--a secret ballot, of course. And insist on a large supermajority.

David R. Henderson writes:

@JVM,
Scott is a pure utility theorist. According to utility theory, consumption has diminishing returns. Therefore marginal utility of addiitonal consumption is least for those who consume the most.
Your third line does not follow from the middle one. The middle one is about individuals. Your third line is across individuals. One of the strongest findings of economics is that you can’t make interpersonal comparisons of utility.

Comments for this entry have been closed
Return to top