The Rich Man Who Can’t be Taxed

Excellent economics puzzler from Steve Landsburg:

Stevens wants to tax the “idle rich”, her Exhibit A being Robert Kendrick, heir to the $84 million Schlage Lock Company fortune. According to Ms. Stevens, Mr. Kendrick appears to do pretty much nothing but park and re-park his four cars all day long. Taxing people like Mr. Kendrick, she says, has to be part of any solution to America’s fiscal crisis.

Here’s what Ms. Stevens misses: Assuming the facts are as she states them, it is quite literally impossible to raise revenue by taxing the likes of Mr. Kendrick. We could argue about whether it’s desirable, but because it’s impossible, the discussion is moot.

Here’s why it’s impossible: For the government to consume more goods and services, somebody else must consume fewer. But Mr. Kendrick, by Ms. Stevens’s account, consumes almost no goods or services whatsoever. He just pushes cars around all day. His consumption can’t go much lower.

Steve is quite right. The key is this sentence, “For the government to consume more goods and services, somebody else must consume fewer.” So here is a second question, if Kendrick isn’t taxed because, by assumption, he isn’t consuming any less after the tax than before, then who is being taxed in this scenario?  Steve gives one answer but there are several potentially correct solutions. One might also ask about the initial assumption, is being “idle” rich, in the sense above, such a bad thing?

Addendum: Brad DeLong chimes in. DeLong attempts to refute Landsburg by claiming incredibly (!) that no one, ever, can be taxed because “we are the government.” Try telling that to someone in prison. Lose the we, as Arnold Kling might say.

Some of Brad’s other (contradictory to the above) claims, however, are correct.  One could argue, for example, that Kendrick’s heirs are taxed. Of course, Kendrick may not have any and this in anycase supports Landsburg that Kendrick is not being taxed.

Addendum 2: In the comments I wrote: Suppose Kendrick were dead. In that case, the tax would reduce the consumption of his heirs, if he had heirs. If he had no heirs and the money were in the bank then the tax would increase interest rates and reduce the consumption of some borrowers. If the money were kept at home under his bed then the tax would increase the price level thus reducing everyone’s consumption by just a little bit. There are other possibilities but those are three main cases. Now, the premise of the problem is that Kendrick isn’t consuming anything but as far as this problem is concerned someone who doesn’t consume anything is the same as dead! Thus, exactly the same story occurs.

I see this morning that Landsburg also uses this analogy in a new post on the topic, You can’t tax a dead man. See also here.


Plant drugs in his cars, seize all his assets.

Ouch. This blog usually avoids errors this blatant.

I love this post. It is economics at its finest. By circumventing the usual discussion on taxation, it exposes the automatons who don't actually understand economics. Most of the time, their misunderstandings are hidden behind a tide of personal values... now the tide has gone out and there are no longer many correct answers depending on how you mix and match your values and economics. There's just one, and it's leaving many people flapping in the mud.

I don't understand. Surely the man eats, at least occasionally, in between parking his cars.

He doesn't produce resources, but he certain consumes them. Thus, taxing his consumption would increase revenues. But since he's not producing resources anyway, the tax would result in little deadweight loss.

On the contrary, Alex, this seems like the idealized example of a (near) perfect tax, in that it doesn't have any supply-side effects.

Wouldn't a wealth tax on his assets tax him?

There, a tax on his future consumption (even if that consumption is just in the form of passing on (less) wealth to his heirs).

I love the idea of a wealth tax debate, because then the supporters can't even pretend it's not flat-out theft anymore.

theft? you give me money, i give you a society

Joe R.

I think the Georgians might want a word with you on that one.

You give me money, I give you my society.

"Theft? you give me money, i give you a society"

Give? How is it giving when the choice is property seizure incarceration.

"Theft? you give me money, i give you a society"

Give? How is it giving when the choice is property seizure or incarceration.

“For the government to consume more goods and services, somebody else must consume fewer.” Replace "the government" with "a more productive industry" you'll see that it is "quite literally impossible" for economic growth to ever occur.

Dennis, you are exactly right. Steve's argument makes no sense.

Add that if the government taxes RK, the government has more revenue and private consumption is unchanged. The government could use the money to do something productive (build a road, hire a teacher, whatever), growing the economy.

Unless, or maybe even if, he has buried his wealth in the backyard, his wealth is invested in something much more productive than the government.

Dennis, the argument is about taxation not spending and the quote from Steve needs to be read in that context.

Wait, what?

You're saying the argument is about taxation not spending, but then the reason he cannot be taxed is because when the government spends, others cannot spend. "For the government to consume more goods and services, somebody else must consume fewer."

This assumes a zero-sum game, that goods and services always remain flat, regardless of demand. But doesn't government spending actually push up the entire demand curve, causing the price to increase, but also the supply (depending to the elasticity of the particular good or service).

Seems to me that this guy, while he may be idle, his money is not. He is making a certain amount in interest, dividends, or whatever, and taxing that income would certainly tax him.

This "puzzle" fails the Econ 101 test.


This whole thing makes no sense.

Landsburg seems to be saying that Kendrick isn't taxed if you take his money, or almost all of it, away, because he won't have to reduce his consumption.

First of all, so what? Kendrick has a preference for maintaining large financial resources. Take those away and he's worse off. That Landsburg might not share those prefernces is irrelevant. If the governmnet taxed away 90% of Landsburg's savings would he shrug and say he hadn't really been taxed, or would he complain loudly?

Second, why must others reduce their consumption? Is the total amount of goods and services in the country fixed? I don't think so.

That's the questionable assumption.

Actually that assumption is irrelevant. Landsburg's argument is that Kendrick isn't taxed because his consumption doesn't fall.Now the argument in this thread is that *no one's consumption fell* for Keynesian assumptions (Hulsey) or productive investment (Dennis). Thus, the arguments in this thread actually support Landsburg's conclusion but for different reasons than Landsburg gave.

Yes, my error was in assuming that Landsburg was making to make a less incorrect statement than the one he was making.

Anyway, as DeLong says, in English most people would view the government taking away your property as being taxed. And that is so even if you assume the property was usefully deployed by someone else who has been given temporary custody of it.

Mr. Kendrick can clearly be taxed. Money is, in my understanding, a claim on future productivity. Taxing Mr. Kendrick equates to taking away some of his money, i.e. some of his claims on future productivity. Notably, whether or not he would have exercised these claims on future productivity is irrelevant. That is, the flaw in the argument seems to be equating "taxing" to "consumption of goods and services", as taxing Mr. Kendrick does not require reducing his consumption, only reducing his potential consumption.

Kendrick isn’t taxed because his consumption doesn’t fall.

Which is a very strange definition of being taxed.

Thus, the arguments in this thread actually support Landsburg’s conclusion but for different reasons than Landsburg gave.

In other words, the best defense you can offer is that Landsburg is the proverbial blind hog who has found something that, if looked at late in the evening light, just so, might possibly be an acorn.

"something much more productive than the government"

Like CDOs for instance?

I'm surprised somebody was actually impressed by this non-sequitur. We're talking about TAXATION, you know, just shifting money around, as opposed to trading. If A gives B $1000, there is no change in the overall consumption level. Landsburg's point is that because it is presumed our idle rich person consumes nothing (or nearly nothing; think carefully about why this doesn't matter), then you can't tax HIM. You are taxing someone else.

I don't think Landsburg's statement necessarily implies a zero-sum game; just that this capital taxation/spending scheme is not Pareto-improving. (Possibly granting some sort of full employment assumption.) And since we've stipulated that's not the taxee suffering (his consumption streams through his life are identical), it must be someone else who is suffering. However, I don't find this a particularly compelling question. The obvious answer is, Kendrick's heirs are suffering in the future; and probably others of us in the future as well, if the government's use of the money is directed at current consumption and not investment. Landsburg does seem to be a bit incoherent, which leads to him thinking this is a big deal.

It's not a puzzle at all: obviously selling your capital and consuming it today improves everyone's consumption today! It's only future people that are harmed.

Except you're shifting here from consumption to investment. No-one (I hope) is denying that producing more stuff means that there is more stuff that can be consumed.

Investment is the delaying of consumption today in order to increase future consumption. So investment DOES involve a short-term reduction in consumption by someone. The difference is that, by increasing the supply of goods & services, investment permits greater consumption.

So, in the long run, the government could increase net consumption by taking money from Mr. Kendrick and giving it to the founders of a new sweet factory. However, even this will necessitate the short-run reduction of consumption by somebody (those who would have bought the equipment of the new factory, for example) in order to attain this increased consumption. The difference comes from the outcomes of increased production.

The basic point that someone must consume less for the government to consume more stands. However, if the government taxes to invest in an increase in goods & services and it is successful in this goal, the overall level of consumption can increase over time. Therefore, Landsburg's point is compatible with the fact that successful investment allows for long-run economic growth.

Well his hoarding does influence the price level doesn't it? By holding his 80 million and doing nothing with it either the price level adjusts to match the money supply minus 80 million in which case there are more "claims" on work (money) then there is stuff to be claimed and the last spender get screwed OR the price level doesn't adjust in which case people are holding additional production (consumptions, goods, whatever) in hold for his theoretical purchases resulting in an inefficient outcome for everyone else when he elects not to purchase. So a wealth tax could generate some additional social surplus in either case?

If he's just hoarding it then he is paying a wealth tax - inflation. If he's invested (e.g. to keep up with inflation) then his money is being used.

Totally correct if he's invested though I think that's the crux of why we have inflation it forces non-user of money to invest it in productive ends in order to maintain purchasing power over time. Would you agree in a deflationary economy his action would be harmful? The current market works fairly well in that he's incentivized to invest his money which means there's more available to the marginal investor. I was just trying to make an argument in which an idle rich person could be actively harmful to the functioning of the rest of the economy.

I don't understand how it is impossible to tax Mr. Kendrick. If he has $84M, then that must be generating some kind of income. If not, then it is true that you can't tax him (given the current types of tax available), but you can tax his estate. A wealth tax would also be an option, but that hasn't been done in the US, as far as I know.

Off the top of my head, Muni bonds, a foundation, giving away all but a tiny fraction of the income to charities, mortgage interest, not having the investment pay dividends, especially if the profits are held by an offshore IP firm). I'm sure there are more creative ways, but all those would keep even a very rich person from paying much in taxes.

its only a puzzler if you are an economics professor

Steve is quite WRONG. The key is this sentence, “For the government to consume more goods and services, somebody else must consume fewer.”
Fixed the key part of the message for ya

This argument would apply to any tax that doesn't result in the person being taxed lowering his consumption. Even a 50% income tax might not cause some super-rich to change their behavior.

This shows how wrong-headed the current income tax structure is.

Money earned through work, "ordinary income", is taxed at a higher rate than money earned though money's work, "capital gains", "interest", "dividends", etc.

Historically, when a country has become financially dominant via greater production/innovation, that country loses it's dominance when it becomes more interested in using it's wealth to make money then to actually produce goods. See Holland and the U.K for examples of this. This is exactly the road the U.S. is going down, and I don't see Wall Street understanding this until it is far too late.

If you taxed him enough he might have to find a job and thereby become productive and since on average people do not capture all the benefit from their work society could potentially benefit, though in practice I am sure it would not.

Going to have to agree with Adam. This isn't a puzzler, it's a fallacy. This is some pretty weak sauce right here.

He does consume goods and services.

The four cars' ownership is a form on consumption. The land upon which he parks and reparks them, the fuel used to drive them from space to space, (and the income to pay for it) are all a form of consumption. The food he eats (assuming he does), the house he lives in, the utilities paid to power and heat it, the clothes he wears (and buys, when they wear out) are all forms of consumption.

Even in abstract, to own is to consume, and he could consume less; the cars, seized by the government, could displace spending to buy vehicles for police departments. The land could become a housing project. His use of these things displaces government/social use, his monopoly of the property denies others its use. Thus, he consumes.

If his estate is expropriated to become a government housing project, how is it less of a 'monopoly'? Instead of one individual 'monopolizing' it, 50 individuals will have private use of 50 small parcels. Each resident 'monopolizing' their home. You, and all other non-residents, still don't have use of the estate, just the chosen new residents.
ALL 'Government/societal' resources are still ultimately consumed by individuals.
It's just that the distribution of these resources is determined by politics (policies determined by elected representatives and non-elected bureaucrats) or chance instead by the market.

that is some impressive commenting - comments #4 and #6 point out the two main reasons why Landsburg is quite mistaken (the third one is the fact that taxing idle wealth constraints future consumption, e.g. by Kendrick's heirs).

Props especially for Dennis Boyle for his pithy & elegant three lines.

Giant pile-o-money tax, despite Ben Franklin's espousal of the inherent godliness of sitting on your big pile of money and not spending it on selfish consumer goods, I think social utility would be greatly increased by tossing some of that money out of the window while this guy parks his cars.

Seems to me that this is a temporal tax.. if it is a tax at all - as opposed to simply a nationalization of assets (how would you define the difference?)

Some future beneficiary will have less after the tax then before. This means that they will consume less (depending on estate taxes, this might simply be a temporal shift in tax revenue from some future date.)

In regards to the other question: How do you structure a "tax" so that it isn't an asset nationalization?

Taxing Mr Kendrick and spending his money on government consumptions becomes a joint monetary and fiscal operation. His money, since he is not using it, is not part of the money supply (or alternatively, monetary velocity is lower because of his money). Tax amount collected from him is equivalent to new money being printed and given to the government.

You can say all savers end up paying for it, bit this is a loaded monetarist debate that I am not sure OP was looking for.

Wow, really? Marginal Revolution just went down a few notches on my read-o-meter.

$84 million is a lot of good and services, owed to someone. That most certainly can be taxed. Where do the good and services come from? The future, when that money would have or could have been spent by someone. That's (part of) how money works, as store of value.

Landsburg is not saying that no one is taxed. Landsburg is saying that Kendrick isn't taxed.

But that makes no sense. Even if Kendrick chooses not to spend the money now, Kendrick *could* spend the money; after he is taxed, he no longer has the option of spending all of it, only the portion remaining to him after tax. So he has been taxed. Even if his actual behavior doesn't change, his options for behavior have changed (i.e. shrunk).

He has lost the option of spending the money, which has value (in the abstract) even if he chooses not to exercise it.

He's "idle." Any reasonable tax on him won't change his consumption levels, at least that is the assumption the puzzle is operating under.

So what? His consumption level is only his consumption level, it's not his net worth. He is taxed because he becomes poorer. The fact that he doesn't become *enough* poorer to affect his lifestyle is beside the point.

"it is quite literally impossible to raise revenue by taxing the likes of Mr. Kendrick."
If the government takes $40 million from him, that's clearly possible and a tax. The claim is that this, somehow doesn't raise revenue, because that money wouldn't have been spent on current goods and services. Clearly, that money will (if not taxed now) be spend on future goods and services. That's where the revenue is coming from. Those future goods and services are "owned" now by Kendrick, even if they're implicitly given to his heirs. So the tax is a tax on him, preventing him from giving those gifts.

Busy bodies are such irritating people. To have gone so far down the path of busy-bodiness that one like Stevens would see fit to pull out the binoculars and hunt down new people who slip through the cracks to take money from, and at all costs, has to be an icky feeling when one stops to take note of what one is doing.

It would help the discussion if some of the commentators would actually read Landsburg's post.

"Ms. Stevens’s great mistake is to confuse money with real resources."

And Mr. Landsburg's great mistake is that he believes Ms. Stevens is doing any thinking at all.

Stevens, like so many liberals, is driven by vengeance towards the wealthy. She sees a rich guy and wants him taken down. If he's idle, or he's white, that's all to the good, but the fundamental issue here is wrath against wealth.

The irony of seeking to transfer his money to a bunch of other rich guys who spend $400 million per hour is of course lost on her.

In contrast to the "liberal vengeance" that you describe, your post is full of sputtering rageful snideness. Oh wait, that's not a contrast.

Jim's right. Ms. Stevens should mind her own business. What anyone is doing with their own money, unless it's gotten illegally, is their right to do.

It's pretty bad when we are having a conversation about stealing someones assets because we can use them better.

Best comment on this thread so far.

The goods and services have been purchased already by the government and paid for by issuing debt, owned presumably by Mr. Kendrick among others. One might say this was essentially a transfer of wealth, from the government to Mr. Kendrick. Gee, in that light, the default argument takes on a whole new dimension.

The government is certainly taking consumption from Kendrick -- future consumption. If the government decides to realize that future consumption today, yes non-Kendrick people will have to consume less, but they will also be able to consume more in the future because the government has used up its option of spending Kendrick's money in the future.

Now, if the assumption is that Kendrick or his heirs will NEVER spend the money, then taxing Kendrick is equivalent to printing new money, with all its advantages/disadvantages.

Yikes!! Too many of you are responding to an argument that was not being made. Arguing about future consumption is not addressing Landsburg's point. Government's tax and consume today, so who bears the expense of that extra consumption today, if Mr. Kendrick, today, is not?

you are correct about people not understanding the point. I sort of acknowledgingly missed the point of the "here and now" part of the argument.

It seems to me that who is being taxed depends a lot on how he is holding the cash. If it's in a big pile under his mattress, the only tax is temporal. If he has investments in things like bonds, then someone will have to pay more to borrow money as the demand for bonds is changed... if it's sitting in a savings account, as the link says it's whoever the bank is loaning money to..

Forgetting about income and substitution effects (since Mr. Kendrick displays none of them) and monetary/macroeconomic effects, Mr. Landsburg is right that someone must decrease their present consumption (or investment, but we could leave investment intact and focus only on consumption) in order for the government to spend more. But the fact that I consume less (for example, by saving more) doesn't mean that I'm being *taxed* or that I'm "bearing the burden of the tax", because nobody on Earth except Mr. Landsburg defines "being taxed" as "the government takes away some of your _present_ _consumption_ to spend more"; instead, taxation is usually defined as "the government takes away some of your present _income_ to spend more".

Other people may have to reduce present private consumption in order for the government to spend more (btw, why does Mr. Lansburg suppose that the government has to increase present spending in response to increased revenue if the original piece proposes "taxing the rich" to curb fiscal deficits and public debt?), but total (private + public) current consumption stays the same (average present "burden" is 0*).

Mr. Landsburg's post concludes:

> “Taxing the rich” cannot work unless you do it in a way that induces the rich to consume less.

He seems to suggest that there is arithmetically possible to tax them without "inducing" them to consume less. Just by taxing them, you're _making_ them consume less, Maybe not _now_, but perhaps tomorrow.

Is this an "excellent economics puzzler"? Actually it's more like "let's mislead people twisting the conventional meaning of 'taxation' and 'burden' and then proceed to talk about 'great mistakes' and 'mistaken journalists' with thought experiments based on assumptions that have no link with reality".

* Whether public consumption is better or worse than private consumption is a different debate.

>taxation is usually defined as “the government takes away some of your present _income_ to spend more”.

By this definition, it is impossible to tax Kendrick. He has no income. What do you call it then? Presumably you would then define a land tax is a tax on potential income or production from that land (higher tax rates for commercial land vs. Residential etc)

And I think defining a tax in terms of consumption is analogous to defining it in terms of income.. because what is wealth or income other than greater options in terms of consumption. Perhaps a better approach would be defining it in terms of "option of consumption".

Since many rich people invest a lot of their income and do not necessarily change their consumption much as tax rates rise (see the book "The Millionaire Next Door") does anyone have an estimate on how much of the taxes on the rich are actually paid for by the rich reducing consumption?

Now that's nonsensical. Raise a property tax which (if he has all his fortune invested in cars) over the years will take a few out of his 840 cars. Done.

Steve is quite WRONG. The key is this sentence, “For the government to consume more goods and services, somebody else must consume fewer.”
Fixed this for ya, and the key remains the same, funnily.
By taxing this man, government decreases consumption of his ancestors in distant future, because millionaires do not consume their marginal income, they save it. But government can spend money right away and increase incomes and hence consumption of current poor people.

By taxing this man, government decreases consumption of his ancestors in distant future,

Now things have really gotten weird

)) sorry, should be "descendants" there. But you got the idea

:) sorry, should be "descendants" ofc. I got carried away a bit, but you got the idea noneverless

Isn't Ms. Stevens' argument, however implicitly, about finding politically feasible ways to finance the CURRENT size of government, without changing it? Her claim is: "If each Californian could bequeath no more than $2 million, a 100 percent tax on the surplus estate assets would wipe out the state’s entire budget deficit."

That claims that taxing Mr. Kendrick would eliminate California's current need to borrow money, meaning its debt-finance charges would drop. I realize that extracting that money from its current place in a bank would reduce the amount of money said bank had to lend out, thus raising interest rates, reducing liquidity, and so forth, but this isn't an argument about revenue in any real way, it's an argument about the size of government, what kind of taxes you could convince people to vote for, and maybe it's a little bit about maximizing multiplier effects.

For that matter, isn't the other implicit idea that Mr. Kendrick would be more economically productive if he had a job of some sort?

I'm not endorsing any of these arguments, and I'm far from an economist, but isn't this discussion about 1) which part of the Laffer Curve we're on, 2) how big the government should be, and 3) who we can get to pay more taxes?

“If each Californian could bequeath no more than $2 million, a 100 percent tax on the surplus estate assets would wipe out the state’s entire budget deficit.”

I am pretty sure that any 100% tax would move rapidly toward collecting something close to $0 over time.

If California taxed assets over $2 million at 100%, it would actually be very interesting to see the demographic changes on the state. I suspect every person who feels they will at some point have over $2 million in assets will leave pretty much immediately, and the entirety of Silicon Valley will vacate. The whole state would be in shambles.

I strongly welcome increased mobility as people can vote with their feet, creating at least some limitation to the size of government.

The next logical step is to build tall walls. This time to keep the people in. I'm sure they still have the plans from East Berlin somewhere.

"She thinks that green pieces of paper, or a series of zeroes and ones in a bank computer, can somehow help supply the government’s demand for actual goods and services. It can’t."

I don't fully accept this statement. Here is an alternative to Steve's scenario... What if the government's spending does not immediately shift consumption of goods or services but merely shifts the ownership of the "idle" money? What psychological effect could this money have on people such that we couldn't just dismiss it as mere zeros and ones on a computer? I don't have a good answer to this, but it is at least conceivable that it could change society... perhaps in terms of willingness to provide services and also the quality of those services. This would also effect goods to the extent that people improve their services that deal with goods, of course!

This is a question of how this sort of transfer could affect social capital. Would reducing the inequality in some supposedly silly zeros and ones also reduce the inequality of social capital and would that be a good thing? Possibly, and there appear to be researchers who take this question seriously.

Sorry, Alex not Tyler wrote the post. No disrespect intended.

I think we can all quibble at the margins with the statement: “For the government to consume more goods and services, somebody else must consume fewer.” But it basically holds, except in cases where the banks holding Mr. Kendrick's money are spending it on fancy mathematical bets on worthless securities. Then you have a situation where no one is consuming anything, and the government can get a free lunch.

Either we're talking about present goods, or we're talking over time.

If we're talking about present goods and services, it clearly doesn't hold. If I hire an unemployed accountant with money from my bank account, and he saves that money just like I would have, I've consumed more (present) services, and no-one has had to consume fewer. Similarly, when I purchase an audiobook, no one has to consume any less. When I buy a car that reduces the dealer's inventory, no-one is has to consume less now. It's not true of tickets to half a full concert, or any number of other examples.

If we're talking over time, it's more true, but still not actually true. My audiobook is still a good I've consumed that hasn't prevented anyone else from consuming it, ever. But now that car I bought is actually stopping someone in the future from buying that car. But that's exactly what $84mil sitting in the bank is: some pile of potential future consumption. Taking that away as a tax is shifting forward that consumption from whenever it would have happened to whenever the government wants it to happen.

I think we can all quibble at the margins with the statement: “For the government to consume more goods and services, somebody else must consume fewer.” But it basically holds, except in cases where the banks holding Mr. Kendrick’s money are spending it on fancy mathematical bets on worthless securities.

Or in cases where the economy isn't producing as much as it could presently produce, i.e., recessions. Gee, do you see one of those around here?

The government can consume more goods and services without making someone else consume less by *raising the overall quantity of goods and services produced*, if it is possible to do so, i.e. if the economy is not presently supply-constrained.

Steve’s formulation of the question is confusing, and possibly misleading, because he does not distinguish between the allocative and distributional effects of taxation. As I read it, he is making an allocative claim [what is the effect of current-year flows of goods and services, and of prices], but most commenters are discussing distributional effects [who is richer and who is poorer.] Stated alternatively, there is a difference between being taxed, and consuming fewer goods and services. They are not the same thing.

To illustrate the difference, even neither the government nor Mr. Kendrick change their expenditure patterns at all, Mr. Kendrick owns less capital and the government owns more. So even if the government lends out the same amount of money and it winds up being lent to the same borrowers, Kendrick’s interest income and accumulated wealth will fall [thus increasing bequests to his heirs], and the government’s net interest expenses and debt will fall.
Conversely, if there is an allocative effect so that Kendrick’s money does not get lent out for factories, vacations, etc., those people will [on the margin, because this is a very small amount relative to the overall size of the capital market] not be any worse off distributionally, because at the same time that they are deprived of their vacations, they are freed of the liability to repay their loans to Mr. K. [Of course, if this were a large enough amount of tax revenue to have second-order effects, then we’d also want to weigh these against the second-order effects from elsewhere in the economy — e.g., from reducing nonclassical unemployment, if one believes such a thing is possible.

I think the Alex's puzzle is a fallacy. If the rich person has to sell some of his (unused) cars to pay his tax bill, he reduces his consumption in the economic sense. Because the cars which have just come on the market can be used aka consumed by *somebody else*, perhaps even the government. The same goes for basically all of his other assets, even the financial ones. Even if he owns a 10,000 sq feet mansion* but does not live it, the house still represents real resources consumed by him.

*i have no idea whether that is a very big house, i'm european

A decent rule of thumb is 10 sq ft to 1 sq. m. If you need more accuracy than that google will calculate something in whatever units you wish from their search line.

Ha! I can calculate, my point was more that a 1000 sq meter house is huge here, perhaps not so in the U.S.

Delong and Tabarrok are both off here.

Of course we aren't the government, because we don;t invest in it (assuming readers aren't a part of significant money-trust operations).

Tabarrok, on the other hand, assumes perfect market knowledge and efficiency in his argument. Only within such conditions could expanded demand (assuming that was how the taxes were used, which seems likely) really be said to have no expansionary impact on economic metabolism.

As far as I can tell, however, as soon as markets reach stable or "accurate" pricing relations - growth stagnates while firms and states struggle to counteract the stagnation, which just makes things worse.

I just wanted to add that the issue is more complex than I imply in my "Tabarrok..." argument here. In any case, only some kind of immediate market response could truly counteract the growth-impact of an expanded taxation-consumption policy - and we know that these "hyper-efficient markets" are merely chimeras...

2 things:
1. "For the government to consume more goods and services, somebody else must consume fewer."
So, in your opinion we are in a full employment situation.
2. If I consume less because the relative prices have changed does not mean I was taxed.

Actually, you should read Delong´s response.
By the way, this post was pretty bad.

“For the government to consume more goods and services, somebody else must consume fewer.”

“For the [private actor] to consume more goods and services, somebody else must consume fewer.”

It is amazing the assumption made by some people that the government cannot create additional value. If the private sector can utilise resources to grow the economy, then the government can too. The argument should be IF the government reallocates resources THEN it should reallocate resources better than or equal to private actors.

Then there are the "trickle down" economics which seems to not work precisely because of the case described above. The already wealthy are more risk adverse, they are comfortable and see no motivation or drive to place this at risk. Has anybody tried "trickle up"? The poor HAVE to pass along any stimulus. The leaky sieve of inactive stimulus would apply when injected from the top than from the bottom.

Why is it so OBVIOUS that the government is worse than private? I don't know, most large organisations are, in the end, governed by boards of detached managers.

Landsburg's argument is about whether it is possible to tax Kendrick or not. The bit about government is a side issue. In fact, Keynesian assumptions or assumptions about productive investment simply add support to Landsburg's conclusion that Kendrick is not taxed!

Kendrick is not taxed

It would really be useful if you would provide a definition of what you mean by "being taxed" in this discussion.

My guess is that you mean, very idiosyncratically, "Does not have his consumption reduced," but I'm not sure. If that's what you mean, why is that the appropriate definition? If you want to argue that no one is taxed unless the payment of the tax has some negative effect on them, then do you imagine that handing his assets over would not affect Kendrick in any way?

He's referring to the incidence of taxation, which isn't idiosyncratic at all. Landsburg is not saying that no one is taxed, only that Kendrick is not being taxed because he is not bearing any of the burden of the tax.

Well he certainly won't complain when they take his money then.

I understand that. My question is why the tax doesn't affect Kendrick.

Are you, or Alex, or Landsburg, claiming that taking Kendrick's money has no effect on him? Do his indifference curves cross?

They are claiming it has no effect on his consumption. That's all.

I'm all for the day when technology or whatever allows us all to live the lives of the idle rich. I don't think the argument is that the SF scion is overconsuming; I think the idea is that we would be better off as a society if his wealth were reallocated to other projects - such as educating children, etc.

It's pointless to "re-alocate the wealth". Government can print money, i.e. "wealth" in the for of money is limitless. What taking money from K. does not produce is additional productive resources.

I'm still having trouble buying the premise that a guy with no job and $84M has no interest in consuming anything.

I found it very helpful to read Landsburg's original post. I am still not sure I'm convinced*, but it definitely made it clear why some of the responses here are superficial.

*: In the original post, Landsburg says: "So what happens if the government takes Mr. Kendrick’s $84 million away? Answer: A bunch of zeros and ones get shifted around on bank computers. Mr. Kendrick goes right on pushing his cars around. And nothing else has changed....Unless, of course, the government decides to spend some of that $84 million." But he also says: "...when Mr. Kendrick withdraws $84 million from the bank to make his tax payment, the bank makes fewer loans, interest rates rise, and someone cancels a vacations, or postpones a car purchase, or abandons a half-built factory." Which is it? Is the incidence of the tax felt when the bank no longer holds Kendrick's funds, or when the government actually commits those funds to some purpose (as a Keynesian might say, injects it back into the economy)?

yes! the answer is not clear.

Where I get tripped up is how "increased taxation" automatically means the government is "consuming more goods and services". After the increased taxes couldn't the government consume the exact same amount of goods and services, pay cash for more of them, and then have the Chinese consume fewer US debt instruments?

Or is one of the possible answer the buyers of debt are paying the tax because under this scenario there would be less government borrowing?

It seems to me the only way this makes sense is if "taxing Kendrick" means exactly the same thing as "cause Kendrick's short term consumption of goods and services to fall," which is a very strange way of parsing what it means to "tax Kendrick."

Further, even if we make the bizarre assumption for the sake of the exercise that his consumption is basically zero (this is a real person!), he has lost his option to consume more (by being taxed). I am not sure why that somehow does not count as a tax on him.

The puzzle would have been easier to understand had Steve Landsburg (and Alex and Brad) just pointed out that the tax on the money holdings of this non-consumer is equivalent to an inflation tax. (Although then it wouldn't have been a puzzle anymore.)

If I accummulate $84mn without any intention of ever spending this money (which is an important assumption), then I have effectively reduced the money supply. As a consequence, everybody else's dollar bills have become slightly more valuable as everybody else can now consume the goods and services that I have decided not to consume.

If the government takes the $84mn from me and spends them, the effect is the same as if the government had printed an additional $84mn. Everybody ele's dollar bills are worth a bit less, everybody else consumes a bit less, and the government consumes more.

There is as little reason to debate this "puzzle" as there is to debate the Monty Hall problem.

This is actually a brilliant argument for the wealth tax.

So, you take Kendrick's money which is sitting in the bank. Which really means the bank is using the money to own assets and make loans. Therefore, to give the money back to Kendrick, who will then give it to the government, those assets must be sold. Thereby increasing the supply of those assets on the market, reducing the price of those assets.

The people who have to reduce their consumption are the other owners of those assets who are consuming based on asset ownership (i.e. people consuming more based on the wealth effect or people borrowing against their assets). Which means a tax on wealth is a tax on the consumption of the wealthy who do consume.

This tax seems to work as one would want. It would have the least impact on the actual lifestyles of the most thrifty wealthy and most impact on the least thrifty.

Wealth = the net sum of future consumption

No way to avoid that. Kendrick won't live forever, so he has to either spend his money (consumption) or give it away (consumption). If you seize Kendrick's wealth, you automatically reduce his future consumption. And you definitely can seize his wealth, so you definitely can reduce his future consumption.

I'm lost. A tax is defined ( "a sum of money demanded by a government for its support or for specific facilities or services, levied upon incomes, property, sales, etc. " Whether or not that tax influences Kendrick's consumption seems like a side issue to me. I'll buy that it doesn't affect his consumption, and that it is more likely to raise bank interest rates thereby affecting the consumption of others. But the money came from Kendrick and went to the government - I believe that is a tax is it not? To me a major factor in the fiscal crisis is the debt that exists today. The government has previously consumed the good and services that it already denied to others. By taxing Kendrick we could pay off a portion of that debt, yes? Is there a judgment hidden somewhere in Steve's argument that reduced private cash reserves is worse than government debt?

The implication is that if an event occurs which has no impact on your future patterns of consumption, then you weren't actually taxed, i.e. examining the actual incidence of the tax rather than the nominal application of the tax. The premise of the post is that while we've taken money that was labelled as Mr. Kendrick's we haven't actually changed his future consumption stream, raising the question of where the tax incidence has fallen, since evidently it fell someplace. The obvious answer is, his descendents.

To my way of looking at it, we're actually retroactively taxing his parents. They were the ones who earned the assets, and presumably participated in consumption at some level. The government also participated in consumption at that time. This argument assumes the definition of taxes as an equivalent chunk of goods and services. If you just look at who owns the money - it would come from Kendrick and go to the government.

Steve clarified this for me. I believe he's considering taxes as an equivalent chunk of goods and services. Since Kendrick will hypothetically not be using his assets for consumption, then we won't be taxing him. I'm not sure I've paraphrased the argument correctly... he's been great to respond to many comments on his page. I recommend checking it out.

This makes sense to me if what we mean exactly is: even if the government seized half of Kendrick's assets, it wouldn't make any difference because it wouldn't affect Kendrick. By the same token, it wouldn't help the government, because money taken from Kendrick may as well have been money created by the Fed. Is that what we mean here?

This post is a reference to 4/20, right?

Sort of like saying that global warming is not really a problem because its been contained to the vicinity of the earth.

I think the Soviet Union also was simply taxing itself. Well, after bleeding its satellites dry anyway.

In modest support of the post, the guy's money is not likely sitting in piles of dollar bills in the guy's closet, as many of the comments seem premised on. It is likely sitting in a bank, or in treasuries, or some ultra safe money market substitute. The deposits support the bank's lending, or the Treasuries fund the government or the money market fund is in CP that funds a private company, etc. So when a government entity taxes / takes that, those investments have to be liquidated and the cash turned over to the government. Which means the bank has to call in loans, or the CP issuer has to find someone else to hold its paper, or the US government is simply applies the money it takes from him to the debt it owes him. It's the current borrowers of the guy's cash that wind up with less, not the rich guy. I think that is the answer to the disagreement between the comments and the post. The author should have expanded his argument beyond goods and services to include capital transfers.

This doesn't work either. Suppose the government takes Kendrick's money and doesn't increase its spending. Then it effectively is paying down debt, which means it is buying back Treasury debt, and making the $84 million available to the capital markets.

What's changed? The sellers of treasuries now have Kendrick's money available for investment purposes, just as Kendrick did before.

The point is buried when you refer to taking "kendrick's money". Unless it is sitting in a bag in his house, the money is somewhere else. That is the point I think Landsberg is making and I am trying to refine.

If Kendrick's money is really a bank deposit, or similar extension of credit or investment, he has to go get "his money" back from the bank, which means the bank has to adjust its loans to compensate for the lost funding, or its shareholders' returns, which means the bank's borrowers or shareholders are the ones whose consumption is affected, which is the point Landsberg is making.

Your point that Treasury repayment fuels someone else'c consumption does not conflict with Landsberg's point. It is consistent - he acknowledges that government is fueling someone else's consumption, he just points out it is not coming at the rich guy's expense.

Why can't the Treasury simply claim control of the exact same investments? Kendrick has, let's say, an $84 million CD. Treasury in effect tells the bank to move the CD from Kendrick's account to the Treasury's. None of the things you describe happen. There is no "getting the money back." There is only changing the name of the owner.

Of course, if Treasury puts the money in a different bank then you might argue that the loans are redistributed, etc. But if we're assuming an absurdly abstract, frictionless system, then the borrowers can simply shift to the new bank. They must be the best borrowers at the margin, right, or how did they get their lona to begin with?

OK, but the claim is that "it is quite literally impossible to raise revenue by taxing the likes of Mr. Kendrick." I can see how the effects of the tax could fall on other people, but I can't see how that automatically implies no net increase in revenue.

I think the broader issue with the attitude of Stevens that Landsberg is responding to is the arrogance, busybodyness, presumption and inequality behind it. Here you have someone who does nothing, yes, but seems to ask for nothing from his fellow citizens. So he hurts no one by his indolence. Why is Stevens outraged by that? Who is she to intervene in his life? If the guy was living off of government benefits and he was otherwise as indolent, would Stevens call for government intervention to remedy the matter?

The pie, she never changes size, she just sits there being subdivided and re-subdivided.
I was taxed once, involuntarily, and all I got to show for it was a version of a society I would never willingly pay for.
All taxation is theft. Vote for the thieves of your choice.

It appears that this agrument persists not because of illogical conclusions reached by various persons, but because of illogical reasoning. The parameters of the thought experiment have not been clearly defined--this system has too many variables and too few defining equations. No agreement can be reached on the correct conclusion to the thought experiment because a single solution simply does not exist. Strings of commenters will continue in perpetuity, making small changes to particular parameters of the thought experiment and creating different results.

Sounds like this guy should have a meeting with Elmer Fudd, investment banker:

Here's something thing to ponder over. Assume Kenrick goes out Viking-style. He has no heirs and withdraws his money, converting to paper cash, and then goes out burning it on his funeral pyre. Also assume the government hasn't taxed him. No assets or productive capacity are destroyed, apart from some green paper. Nor is consumption impacted much, as Kenrick was already postulated to be pretty much economically inert (leaving aside a small amount of food, gasoline, etc.. small relative to his paper wealth.)

"Steve is quite right. The key is this sentence, “For the government to consume more goods and services, somebody else must consume fewer.”

This statement is only true if the economy is at potential output. If the economy is depressed below potential output, as it is now, using stimulative economic policies (whether monetary or fiscal) to move the economy to potential output makes it possible for both the government and other to consume more at the same time. In other words, moving the economy to potential output when it is below potential output provides society with a free lunch.

Leaving the economy below potential output means that a large unexploited gain is left on the table. And that free lunch that is being left on the table does not stick around to be enjoyed in future periods. It is forever lost and can never be recovered.

Sales tax, ad valorem tax, cell phone taxes...errr.... I know, it's not a tax, fishing license, auto tags, tolls, inheritance tax, etc.

I don't think Kendrick gets off scott-free. Plus, what sort of burden does this guy place on society? None, I bet.

This is classic Landsburg in that he waives away all deadweight loss considerations. If the idle rich man's money is taxed away from him, then the money is paid out of transfer of spending power to the government, rather than by distorting labor or capital supply decisions. This means, all else equal, that the government can tax someone else less who has greater elasticities, lowering the deadweight loss from that person. In other words, if the entire government could be funded by taxing such satiated, nonworking people, then we could achieve a much less tax-distorted economy.

We don't tax assets, only income (except at death).
His interest income and dividends are taxed.

More proof that DeLong believes that the government owns our labor. A 23% effective income tax rate means the government is giving you 77% of the income your employer paid to the government.

Not much difference between the Southern Plantation model pre Civil War and the DeLong model.

Couldn't you tax "lazy capital"?

In one of Jesus' parables, the master gives three servants some talents. The servant that just kept his talent and didn't invest it was severely reprimanded and has his talen taken away and given to the others.

In common law, ownership of unoccupied property can be lost via "adverse possession" to a squatter who uses the property.

If Mr Kendrick seems happy to just push cars around all day, why not distribute his wealth to others who would use it more productively?

Consider the possibility that his money is in fact being so used because (a) it is in Treasuries meaning the government is already spending it, or (b) it is in the bank and being loaned out to consumers and businesses. Unless the money is in a bag in his house, or a gold bar, it is already in use.

This is the point of capitalism. Merely because he is not spending it does not mean it is not being used productively.

Re-allocating property rights changes the incentives for production, as well as consumption. Picturing the economy as zero-sum consumption of a fixed pie is arguably more of a layman's error than the money illusion which SL claims (without providing any evidence that I can find) to be calling out. And a more OMG LIBRUL fallacy at that :-)
Over-ruling Kendrick's seemingly irrational preference for future instead of current consumption (possibly by his heirs), which I'm guessing isn't sensitive to the price of current consumption goods, could benefit workers and businesses that are currently "idle" due to lack of demand for their products, without reducing anyone's current consumption (and in theory without even reducing K's heirs' future consumption, depending on what they had planned on doing with the purchasing power they just lost).
The downside of confiscating K's wealth doesn't work through some sort of spurious accounting identity, it works through the incentives to other (more active) economic agents. Weaker property rights mean weaker incentives to accumulate wealth (which may be a bad thing, if it discourages productive activities as well as rent-seeking), but it creates much stronger incentives to get friendly with whoever is doing the confiscating. The more discretionary the confiscation, the more effort will go into corrupting the confiscators.
Weighing, or even just forecasting, the overall costs and benefits is difficult, so it's easier to identify emotionally with the characters in whichever morality play your political party spins this as, and just pick the same option each time.

Say's Law dies a slow death.

If his money is invested in government bonds, he gets interest and so his wealth grows and so does the government debt. Using this logic you can not give someone a tax cut either if they do not spend the money, and since someone is buying the treasury bonds that the government issues to pay for the cut the net effect is also zero

I find Landsburg's statement grossly over simplified hyperbole.

1) Why do we assume a zero sum game with the wealth transfer from rich guy to government?
2) Even if we assume Kendrick only moves his cars around, at minimum he needs food and, probably gasoline, for that activity. When his bank account is taken where do the funds for his future consumption come from?

Still much confusion. This is a problem in tax incidence theory. Think about it this way. Suppose Kendrick were dead. In that case, the tax would reduce the consumption of his heirs, if he had heirs. If he had no heirs and the money were in the bank then the tax would increase interest rates and reduce the consumption of some borrowers. If the money were kept at home under his bed then the tax would increase the price level thus reducing everyone's consumption by just a little bit. There are other possibilities but those are three main cases. Now, the premise of the problem is that Kendrick isn't consuming anything but as far as this problem is concerned someone who doesn't consume anything is the same as dead! Thus, exactly the same story occurs.

All of these cases assume a tight equilibrium, which is something of a variation on the zero-sum game observation. Outside of equilibrium it's not clear the "someone has to get crowded out" claim holds.

John - That's what I've been trying to say.

Alex is exactly right. It is tax incidence point, seeking to dispel the populist myth-perception of Stevens that the nation is taxing "him". If the person or entity being taxed is economically inert, s/he is a pass through, as far as taxes go. This is just a variation on the tax incidence of a corporate tax. The corporation is not a person who eats or drinks etc. so the corporate income tax is not borne by "the corporation" but by one or more of its constituencies.

This has nothing to do with whether such a tax is better or worse in a utilitarian sense. It simply says that if you evaluate its merit based on a belief that it is being borne by the inert person or entity you are not correctly assessing its incidence or its merit.

Ms. Stevens writes:
"Arguably, it is wealthy people like Mr. Kendrick whom President Obama had in mind last week when he declared that taxing the rich had to be part of any solution to America’s fiscal crisis.
Source: The Bay Citizen ("
Unfortunately, that's not what the President is proposong. He's proposing taxing couples with combined annual incomes of $250,000, who I assure you are not "idle".

Yes, Alex, that's right! And we don't need intrusive heirs of the failed Carter Mondale administration installing crosswalks in streets because thanks to Zeno's paradox it's impossible for cars to hit pedestrians.

Sweet mother of pearl!


Dear Brad Delong: I suggest to you, try to consume, produce and save using their own resources and not stealing the resources that other people were able to accumulate. You should know that I am not the government but I am myself: my powers, my freedom and my property. And I always try to defend everything I am with all my strength. If You are part of the government, if you are appropriating the resources of others, then you are a thief, and you are an enemy of human life...

It is nice to read [i]latinos[/i] defending american values.

I see that Langsburgh is doubling down on the original error (or I should say original unstated assumption), as does Alex with this: "If he had no heirs and the money were in the bank then the tax would increase interest rates and reduce the consumption of some borrowers."

You are assuming Kendrick doesn't have the money in a mattress and that it is otherwise productively employed. If its sitting on a bank's balance sheet as excess reserves, it may as well be in the mattress, and it is invalid to assume that someone else must consume less as a direct result of the tax (assuming here that government consumption isn't increasing).

Of course none of this makes "to be taxed" to mean "to reduce current consumption." Today Landsburgh shifts to "feel the burden of tax" but I don't think thats right either. Taking some of Kendrick's money deprives him of future potential consumption which is most definitely a burden of being taxed.

In case anyone's still reading, I thought I'd post an analogy here (I have enough on Landsburg's blog as it is).

Let's say I have a ticket to a concert. We can all agree that the ticket, like money, is not itself wealth, but a claim on wealth. Let's say I hate the performers and don't plan to go. (I don't plan to exercise the option to go to the concert that the ticket gives me.)

Then, the government takes my ticket.

Landsburg is essentially saying that, because I wasn't planning to use the ticket, I wasn't taxed -- and indeed, can't be taxed via ticket seizures.

That is wrong. I have lost the OPTION to go to the concert. Maybe I wasn't exercising it before. Maybe it's not worth a lot _to me_. But I could have sold it and bought something I did want.

The government has reduced my potential to consume, just like it has with any other tax.

Taking a ticket from me is taxing me. Emptying my bank account is taxing me. Neither conclusion is affected by economic incidence arguments.

So... if I am currently saving 10% of my income, then if the government increases my taxes by 10%, and I in response consume exactly the same amount but instead save no money, the government has not "raised revenue"? What?

I find the entire discussion rather irrelevant.

The focus should not be on whether the government taking people's property is economically justifiable because of its effects, or even what the economic ramifications of such an action are.

The focus should be on the fact that the government taking people's property is immoral.

One of the problems is that you folks are sitting here and arguing the finer points of something obscure and hypothetical while here in the real world POTUS is threatening, once again, to tax the hell out of us, refuses to yield on spending, and meanwhile, there's not a single opposition candidate who's not a joke.

If people who actually care about the contunuing assault on freedom - in a real, gut-wrenching, "I can't get treated for my prostate cancer because some idiot bean counter set a quota on the drugs for white males" sort of way - if they do nothing but argue in the abstract about whether Schrodinger's cat is dead while the world self-immolates around them, then it simply won't matter one iota.

All of this brilliant reasoning perhaps should be directed at how we're going to unseat the outlaw government we currently have.

Just a thought.

I shall prove that black is white.
(copyright Alex T.)

Specific quotes from Buchanan and Tullock refuting the romantic "we are the government" nonsense here:

This statement "For the government to consume more goods and services, somebody else must consume fewer" mistakenly assumes the economy is a zero sum game. If the government gets someone to produce who would otherwise be idle, then the nobody needs to reduce consumption by an amount equal to the government's increase. When the economy is at less than full capacity this is possible.

For example if the government taxed Kendrick and paid for a few more potholes to get filled, then an otherwise underemployed (or unemployed) construction worker might have enough money in the bank to pay a dentist to fill a cavity in his son's teeth. The construction worker calls the dentist office, which adds the son to the patient list. The dentist office will remain open an hour later this Friday to fill the cavity, and some employees will have some more money to spend....

No inflation need be created here, the movement of money just told the construction worker that we have resources to fill the cavity and the dentist that he should work another hour to fill it.

Furthermore this argument is rather academic; the government cannot separate who consumes with their money and who lets it sit idle or in poor investments. A graduated income tax or wealth tax directs more of the factors of production to go towards the desires and needs of the lower income group; e.g. produce more food and basic medical care and fewer face lifts and Mercedes.

Comments for this post are closed