Paul Krugman, economist:
This insight illustrates a general principle of the economics of taxation: the incidence of a tax — who really bears the burden of the tax — is typically not a question you can answer by asking who writes the check to the government.
Paul Krugman, blogger, remarking on a straightforward application of that principle:
There are multiple things wrong with this claim, but the most fundamental, I think, is that it represents a remarkable misunderstanding of the reasons why we have taxes in the first place.
(Edited to add: My response to Krugman is here.)
What does the INTENT of a tax have to do with the EFFECT of a tax?
Krugman could as well argue that since the INTENT of tariffs is to protect jobs it makes no sense to say tariffs destroy jobs. And I expect he will …
“[T]here are a lot of people with Ph.D.s in economics who can throw around a lot of jargon, but when push comes to shove, have no coherent picture whatsoever of how the pieces fit together.”
He probably wasn’t intending to aim that remark at himself, but at least he finally gets something right!
Mr. X dies with his will declaring his wish for his $85 million fortune in $100 bills to be kept inside a vault forever.
Congress, unhappy with the “flagrantly unproductive use of wealth”, passes the “Mr. X Estate Tax Act”, which allocates $1 million to be used in the enforcement of a 100% tax rate on Mr X’s estate, not limited to breaking into the vault and seizing the cash.
What are the economic effects of Congress’ actions?
Now consider a similar scenario where to the public, everything is the same. The Act passes Congress just as in the previous scenario. However, unbeknownst to the public, the government has no plans to actually break open the vault. The $1 million was instead allocated to the United States Mint to print $85 million in $100 bills and to completely cover up the fraud.
What are the economic effects of Congress’ actions?
If your second answer is different to the first answer, I am curious as to why. In both cases the government spends $1 million of real resources and ends up with $85 million in cash. To everyone outside of the conspiracy, the two scenarios are identical.
The lesson is that confiscating money from someone who isn’t using any is equivalent to printing money. The incidence of printing money clearly isn’t on Mr. X or his family – Mr. X is dead and his family don’t know any better. Most likely, the holders of existing currency will see it devalued.
There is one significant difference between this scenario and actual cases of “idle rich” in that while the “idle rich” may not spend very much in the present, there is always the possibility that they will at some point in the future. No matter what Mr. Kendrick says now, I will always put more credence in him not spending in the future if he burnt all his money rather than just kept up a habit of not spending a lot. This isn’t irreconcilable with my analogy, though. All that needs to be changed is the printing of money being converted to some general expectation that more money may be printed in the future.
It is true that some economic theories posit that printing money can at times have positive economic effects. I’m not going to argue for or against that here. Rather, it is important to realise that arguments in favour of taxing the “idle rich” need to be recognised and evaluated for what they are: as effectively advocating printing money.
Scott Sumner chimed in on these posts (at the end of his post).
http://www.themoneyillusion.com/?p=9449
“PS. One problem with vacations is that others beat you to the story. I have been growing increasingly dismayed by progressive blog posts on taxing the rich, and was planning a post pointing out that it was impossible to tax Bill Gates and Warren Buffett more heavily, as no plausible tax regime would lower their consumption bundle (or even the consumption bundle of their heirs.) Alas, Steven Landsburg beat me to it. I guess it attracted a lot of criticism. My reaction:
1. I’m shocked that people found his argument novel or controversial.
2. I’m shocked that progressive bloggers that are so dismissive of no-nothing conservatives are themselves so ignorant of public finance 101. Tax incidence is all about expected future consumption streams, as is pretty much all of economics. I thought people knew that.”
Mr. K wakes up one morning and says, “I’ve made up my mind. I’m going to build an $84M mansion.” A few minutes later, he opens his mail and discovers an $84M tax bill. He pays the tax bill instead of building the mansion.
Has he been taxed?
Ed C: A mansion, unlike a bank account, is built from real resources. Mr K was planning to command those resources, which are now made available to others instead of to him. So 1) yes, he bears the burden of this tax (because he doesnt get his mansion) and 2) we dont have to look elsewhere for additional burdens, because we know where the resources are coming from (they are the resources that would otherwise have gone into building the mansion).
As I sometimes have to tell my students, you really would understand this better if you tried thinking about it before asking for help.
I’m disturbed by what a disingenuous creep Paul Krugman is.
LOLz: “Discussions like this really disturb me; they indicate that there are a lot of people with Ph.D.s in economics who can throw around a lot of jargon, but when push comes to shove, have no coherent picture whatsoever of how the pieces fit together.” ~ The Krugman.
K, I don’t wanna harp on details, but Krugman is not the best on doing his journalistic homework — and when you blog for NYT, you probably should have a basic working knowledge of how that is done. Not so hard, as it turns out: http://en.wikipedia.org/wiki/Steven_Landsburg.
Forgive me if I’ve missed anything Dr. Landsburg, but I am fairly certain your PhD is in math, not econ. That doesn’t in any way make you any less of an economist than Krugman, it just seems to illustrate Krugman’s careless blogging style.
We’re talking a lot about rich people with lots of money and no plans ever to spend it. Do such people exist in the real world? If they don’t, is this all academic?
My take on this is that Krugman did not look before he leaped on this one. He ought to know by now that, with your math background, you are not going to construct special instructive cases where you are flatly wrong.
I predict that Prof K. doesn’t bring this up anymore.
This baffles me. And saddens me, because I thought I had learned something from Landsburg – but must now confront the fact that I’ve been deluding myself.
For years, Landsburg has emphasized that the public debate about deficits — the different between government revenues and government consumption — is somewhat misdirected. Rather, we should focus on the consumption side. Government consumption imposes the burden, because it crowds out other consumption. Taxation and inflation are merely alternative means for allocating that burden.
Yet in this discussion Landsburg repeatedly emphasizes that taxation is the burden. And to illustrate his point, Landsburg creates a hypothetical of a guy who consumes basically nothing. Clearly taxing him does not crowd out his consumption. And it’s not clear how taxing him crowds out any one else’s consumption either. So where is the burden?
I despair of ever understanding this.
If Mr. K., or Mr. X in Henry’s example, hides the money somewhere (like under his mattress) and NEVER does anything with it, then all he has done is reduce the money supply, right?
But if Mr. K has the money in a bank, or in an investment fund, then it’s doing something. The bank has lent it to someone, or the investment is being used by some company to build infrastructure to produce widgets. And if we tax Mr. K., then he has to withdraw money from the bank account, or he has to sell stock, and that has to have some secondary effect, right? Perhaps the bank has to raise its interest rate to attract deposits, or it has to reduce its loans, etc.
Seems pretty straightforward to me.
Al V, here’s how I think Prof L would respond:
Tax incidence is defined in terms of the changes in consumption patterns that result both from the imposition of a tax and the subsequent spending of the tax revenue. Those persons whose consumption falls bear the burden of the tax.
If Mr. K has no plans to consume his wealth and keeps his money under the mattress, the tax burden is borne by everyone whose consumption is reduced as a result of the price changes that occur when government spends his $84M.
If Mr. K has no plans to consume his wealth and must liquidate investments to pay his tax bill, the tax burden is borne by everyone whose consumption is reduced as a result of both the price changes caused by liquidation and the price changes due to government spending.
If, however, Mr. K planned to spend that money some day, then he alone bears the tax burden.
Therefore: Only if Mr. K plans to consume his wealth is it possible to tax him.
Ed C.: Exactly!!
I know this topic has been beaten to death…
…but I just thought about what the equivalent to this blogosphere discussion would be on physics blogs.
Assume blogger L states on his physics blog: “A feather and a hammer fall at the same speed in a vacuum.” (see http://nssdc.gsfc.nasa.gov/planetary/lunar/apollo_15_feather_drop.html).
Commenter#1 writes: “Why does L assume we are in a vacuum? We are on earth. Can’t physicists think about some relevant experiments?”
Commenter#2 writes: “L hasn’t defined what he means by vacuum. I think there is a vacuum in my grandmother’s attic, and the experiment doesn’t work there!”
Commenter#3 adds: “It is likely that the person who drops the feather and the hammer releases them at ever so slightly different times. It is impossible for humans to do such an expriment. Hypotheticals! Hypotheticals!”
Commenter#4 adds: “If you do such an experiment in a room with a window, air might blow your feather away. This alone shows that this hypothetical example does not apply everywhere.”
Commenter#5 chimes in: “Who would even want to drop a feather and a hammer? What a pointless thing to do!”
I deeply admire anyone who blogs on a topic that he knows well and does not despair when he reads the comments :-)
Ed C.
Landsburg:
So how does this scenario differ from the burden everyone bears when government spends $84M without taxing Mr. K?
Again, contrary to Landsburg’s foundational claim that taxation imposes a burden, it appears to me that the burden arises exclusively from the spending. What am I missing here?
nobody.really:
Elizabeth Stevens wrote in her article that started the discussion: “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.”
Presumably she has in mind that the government will spend the revenue that it collects by “taxing the rich,” rather than simply collect the revenue and then burn the money.
To answer your question: you are missing the context in which this discussion takes place.
nobody.really: The burden arises entirely from the spending. But taxation can transfer that burden from one person to another. Say the govt spends a borrowed dollar. You, as a future taxpayer, bear that burden. Then the govt takes a dollar from Mr K and pays off that debt. This relieves you of the burden and transfers it to someone else. Who is that someone else? Not Mr K. That’s the whole story.
Flop: You forgot Commenter #6 (the physics Nobel laureate) who says: “You obviously have no idea what the purpose of a feather is. People buy feathers to stuff pillows, not to drop them.”
nobody.really: think about the case where the govt takes a dollar from you and gives it to me, so there is no spending (only a transfer payment). ignoring any deadweight losses from this tax, there is no *net* burden. But there is still a burden on me, which is canceled out by a benefit to you.
On balance, there is no burden, but we can still talk about the burden that I bear.
Flop has made my day. At last, someone who understands!
So how does this scenario differ from the burden everyone bears when government spends $84M without taxing Mr. K?
[Y]ou are missing the context in which this discussion takes place.
Which context? The context of taxing Robert Kendrick, heir to the $84 million Schlage Lock Company fortune with which he might someday increase his consumption? Or the context of taxing a hypothetical person with a bunch of money that can never be used, by himself or any heir, to increase consumption? I sense much of the confusion in this discussion arises from people switching between these two contexts.
Yeah, I got that envelope with the dollar; thanks. It also contained an old Chipotle receipt; I guess the government agent grabbed it out of your wallet at the same time. I’d hate to burden you unduly. Should I send the receipt back? That is, do you experience the same burden losing the Chipotle receipt as you do losing the dollar?
I guess that depends on what use you might have for the dollar and the receipt. I can see the use of the dollar: you might spend it. I expect you didn’t have a lot of use for the receipt. Thus, even though the process of losing the dollar and losing the receipt were similar, you might experience a different level of burden.
Now consider the consequences for Mr. K, a guy with a whole bunch of pieces of paper under his mattress. Does Mr. K experience the loss of those pieces of paper as akin to what you experience in losing dollars? Or is it more akin to losing a stack of Chipotle receipts? Again, I guess that depends on what use might be made of those pieces of paper. If, like Robert Kendrick, K has the option to exchange his pieces of paper for goods and services (or to avoid bads and disservices), then yes, I can see that Mr. K would experience a burden by being taxed – even if it had no consequence on his current level of consumption. And, by the same token, I could see the benefit government would derive from taxing him; it could deter increases in Mr. K’s (future) consumption, freeing up resources for government consumption (without triggering inflation).
Alternatively, if Mr. K doesn’t have the option to exchange his pieces of paper for goods and services to increase his level of consumption, then it’s not clear what burden he would experience from having government agents come cart them away. By the same token, it’s not clear what benefit the government would derive from doing so.
If G’s taxing and spending of Mr. K’s savings causes capital to be consumed, and less capital means less production and therefore lower future consumption (i.e. a lower steady state), is the tax incidence infinite and borne by all future generations?
nobody.really: The context is that of taxing a person whose consumption stream does not change as a result of being taxed.
Steve, just wanted to stop by and say your comment about Krugman once teaching economics made my day. Very true! Too bad these economists who play the “liberal economist” get as much press as they do.
Ed C:
In your scenario the tax burden would be infinite only if you assume that the discount rate on the consumption of future generations is zero.
Flop: Ok. Let’s incorporate present value into the definition of tax incidence I posted earlier:
Tax incidence is defined in terms of the changes in consumption patterns that result both from the imposition of a tax and the subsequent spending of the tax revenue. The tax burden is the present value of foregone future consumption, and is borne by everybody whose consumption would otherwise have been higher.
Sound good?
Ed C:
Sounds good, even though I would prefer to define the burden in terms of lost utility rather than just consumption. Defining the burden in terms of utility permits us to record a burden for Mr K. in case he had intended to leave his wealth to his heirs. Their reduced consumption would then reduce his utility, and he would bear some of the tax burden.
But as long as we assume that people care only about their own consumption (and this is, I believe, what we have assumed for this puzzle–otherwise Ms Stevens’ focus on pushing cars doesn’t make much sense), I’d be happy to equate consumption and utility and accept your definition.
And, of course, to the extent that it is possible to measure changes in consumption while it is generally infeasible to measure general changes in utility, any actual excess burden can only be measured in terms of foregone consumption. So I am happy to accept your definition on practical terms as well.
Ed C:
On further thought, let me take back what I wrote before. I was thinking of the general cost of taxation, rather than the question of tax incidence and tax burden.
The tax burden is indeed defined in terms of changes in consumption, so I accept your definition without qualification.
Flop:
Thanks for sharing your thoughts. Actually, that’s not how I think of tax incidence. I’m just doing my best to understand how Prof L is treating it here.
When I think of tax incidence I have in mind the textbook example of a sales tax in a single good market. The tax acts as a price wedge, and from the shapes of the supply and demand curves we can calculate the new prices paid by sellers and buyers as well as the total taxes collected. From there we can calculate the tax burden and how it is distributed between sellers and buyers. We don’t talk about how the total consumption of these buyers and sellers and everyone else changes and define tax incidence in those terms. That’s an interesting question, of course. I just see it as being separate from tax incidence.
So my take on Ms. Stevens’ article is that, while Mr. K bears 100% of the tax burden, what’s interesting is what happens when government spends the money. I would talk about how the prices of C goods will rise because Mr. K wasn’t consuming much, and how this affects everybody but Mr. K in the short term (Prof L’s main point). Then I would move on to consider how the tax would discourage saving, which would lead to less capital accumulation, economic regression, and ultimately less consumption.
I just wouldn’t go so far as to define the tax burden in terms of all of these secondary consumption effects. I understand the reason for wanting to do so (consumption is the goal of economic action so it makes sense to define a burden in terms of lost consumption) but it seems to me to unnecessarily complicate the more narrow and straightforward concept of tax incidence as I understand it. Retaining the narrow definition would also let us avoid the peculiar notion that Mr. K’s tax burden is a function of his state of mind. (I.e., if he plans on consuming all or even part of his wealth some day, he bears the burden; otherwise, he doesn’t. And the burden is shifted from him to everybody else whenever he changes his mind.)
I should have written in my previous post:
We don’t talk about how the total consumption of these buyers and sellers and everyone else changes *when G spends the tax revenue* and define tax incidence in those terms.
Ed C:
Also thanks for sharing your thoughts!
I believe the difference between our viewpoints is that you look at incidence from a partial equilibrium perspective while I look at it from a general equilibrium perspective.
In my mind, Mr K sits at home, maximizing his utility by pushing his cars around. Whether or not he has $84mn does not affect his behavior–he does not get any utility from having them. I then look at the tax burden as lost consumer surplus–all the things Mr K wants to buy but cannot buy anymore because he has lost his $84mn. Because he didn’t want to buy anything, there is no loss in consumer surplus and he therefore does not bear any burden.
Thus if taxing Mr K does nothing but transferring unused funds from him to the government vault, then the tax has no burden–simply because, by assumption, the government took away something that has no value to Mr K. If the government took away my trash and called it a tax, then this “tax” wouldn’t have a burden either.
But if the government spends the tax revenue, then it does something that Mr K wouldn’t have done–it uses resources that are now unavailable to someone else. Thus this “someone else” must bear the burden. In other words, what Mr K planned to do with his $84mn is essential.
Consider Prof Landsburg’s reference to Mr Scrooge. Assume we are in a two-person economy. If I work for you for free, then you are better off. If I work for you only if you give me green pieces of paper but then I never present these pieces of paper to collect goods, then you are as well of as you would have been had I worked for you for free. Mr K has collected many green pieces which he does not plan to spend. So Mr K essentially worked for free. By taxing him and demanding goods in exchange for the green pieces of paper, the government undoes his generosity.
So collecting the tax does not cause any burden in this scenario. The burden arises only when the government spends the money.
Economists are quite aware that the way in which the government spends the tax revenue has price effects in many markets. These price effects must be considered in a general equilibrium analysis when we want to determine the total effect of a tax–simply because in general equilibrium, we must explain what the government does with the tax revenue that it collects.
‘So collecting the tax does not cause any burden in this scenario. The burden arises only when the government spends the money.’
The problem here is that if the Govt. doesn’t spend money it does not mean that no extra spending will occur anyway. Indeed, if a Govt stops doing Govt. type things, it doesn’t mean those activities will not arise. On the other hand you could have a shadow Govt which steps in. One can’t call a model a truly general eqbm. one if it ignores the fact that Govt. is a fuzzy concept.
There’s an example from India. A guy called Vinobha Bhave decided to turn the state of Bihar into a Utopia by getting all the wealthy landowners to give their extra land to poor people. It was a voluntary tax or redistribution of ‘entitlements’. Bhave vowed not to leave Bihar till he’d finally redistributed all the land. This was speedily done by the voluntary gift of the entire state of Bihar to his cause. Nothing changed on the ground. The old Gandhian nutjob went back home happy. But, the result was not what Landsberg would predict- viz. no general eqbm effect because the Govt. isn’t spending any money. On the contrary, spending on land disputes went up because now the State was out of the picture various sorts of shadow-states and parallel Govts. arose.
Landsberg’s ‘you can’t tax a dead man’ paradox is purely verbal. It assumes that unlike Laffer’s Monopolist who has to compete with entities in parallel dimensions which might suddenly just phase shift into our dimension, Landsberg’s Govt. has no similar competition.
Why not? I watch Fringe. It could happen. I’m pretty certain I really did see Maria Shriver grin so wide her reptilian teeth shot out and she bit the head of her husband the Governator. It was all over the Huffington Post.
Bottom line, true, under certain circumstances you get 100% crowding out for Fiscal Policy. But Govt. cuts or failure to spend might equally cause a more than 100% rise in private sector spending- i.e. cuts, too, have an incidence and an excess burden.
There is another point that might be made about how the math goes haywire w.r.t catastrophic risks and how from what be called a ‘Cliodynamic’ p.o.v Govt. life-cycles have to do with perceived solvency and capacity to face such risks.
Thank you Dan from above!
I am happy to concede the point that a man who consumes nothing can’t be taxed…
Now find one
@Tony Cohen: We already did, the subject of the newspaper article. The *assumption* was we’d be taxing money he would never use.
Look, most of the time taking money is taking wealth. But money and wealth are not quite the same thing so *logically* there is a difference. The case in hand is a nice one to highlight this logical difference, and Steve did just that. But for many — the types were adumbrated beautifully by Flop — logic is harder than denial. So we see lots of denial.
I’ll add an observation. If you do not see why Steve is right about this point then fundamentally you do not understand what MONEY really is.
Hahaha
Cheers to flop!
I’m annoyed that in these posts, Steve says “when the govt taxes, it imposes a burden”, but in other posts he says “taxing is irrelevant, it’s when the govt spends, it imposes a burden”
I just hope I don’t get totally confused by this inconsistency.
Mike H:
I’m annoyed that in these posts, Steve says “when the govt taxes, it imposes a burden”, but in other posts he says “taxing is irrelevant, it’s when the govt spends, it imposes a burden
If I take a dollar from you and give it to someone else, then (ignoring disincentive effects for purposes of this discussion), I have imposed a burden of $1 on you and a burden of *minus* one dollar on the recipient, for a net burden on society of zero.
Disincentive effects aside, only spending can create a *net* burden. But any tax creates a positive burden for *someone*. This post was about identifying that someone.
‘Disincentive effects aside, only spending can create a *net* burden.’and this happens through crowding out. But crowding out could be zero or negative.
Does any tax with non-zero yield, ‘create a positive burden for someone?’ Not necessarily if a new hybrid sort of agent is created by the tax such that the non zero yield for the hybrid agent is more than offset by extra revenue or utility from the activity the Govt. makes possible. Suppose there is some new type of resource, say bandwidth, or the right to refer to Judges as ‘yo, bitch!’ during your trial, which would otherwise not be an economic good absent the Govts. action in taxing it- where is the ‘positive burden for someone’ which has not arisen through the market?
One might say, it was only after the Govt. made this activity possible, only after it made it an economic good, that agents had agency in this respect- but, if paying the tax is by itself an entitlement to that new economic good, in a sense a new hybrid sort of agency has been created. The agent is extended in a hybrid way, something of the distinction between the State and the individual has blurred.
Yet, at bottom, is not every economic activity, in its genealogy, precisely the result of this sort of hybrid agency where the boundary between State and individual gets fuzzy?
The lesson of the crowding out effect was worth spelling out in the late 70’s. Its inverse, the situation where Govt. spending cuts cause private spending to increase more than equally with a much larger dead weight Welfare loss, also needs to be considered.
You can’t tax the guy whose happy just parking cars- true, but what happens when the roads decay to a point where he can’t get his jollies the only way he knows how? Cuts in Govt. spending impose a burden- a very large one in failed States- compared to which the crowding out effect, at least under conditions of large scale unemployment, of Govt. spending is a mere bagatelle.
Taxes, ultimately, are price one pays for Civilization- or, if you live in Texas, the price you pay to prevent Civilization from taking a baseball bat to your Criminal Justice system.
” Suppose there is some new type of resource, say bandwidth … which would otherwise not be an economic good absent the Govts. action in taxing it ”
Suppose there was a new kind of absurd claim, say the claim that ‘communications technology has no economic effect absent the govt taxing it’ …
The telegraph was invented in about 1840. Remarkably it had no economic effect until it was taxed. Bet you didn’t know that!
Of course I did. You lose your bet Ken B.
You say ‘remarkably it had no economic effect till it was taxed’
You are kidding right?