I am opposed to all taxes on interest, dividends and other forms of capital income. Supporters of these taxes keep making the same fallacious argument. The purpose of this post is to shame those people out of ever making that argument again. (They are, of course, free to make other arguments.)
The fallacy I have in mind goes like this: First, economics teaches us that everything should be taxed at the same rate to avoid unnecessary distortions. Second, QED.
With appropriate caveats, the first part is true. The problem is with getting from there to the second part.
Let’s try an example. Suppose you spend your income on scones, which you buy at a dollar apiece, and suppose the interest rate is 100% per day (not very realistic, but of course the example works just as well with realistic numbers). And suppose you earn $2 a day.
The first observation is that any tax on income is effectively a tax on scones. That is, the only reason it hurts to pay taxes is that you end up with fewer scones. With that in mind, consider three scenarios:
Scenario 1 (no taxes): You earn two dollars, with which you can buy either two scones today or (after saving and earning interest) four scones tomorrow.
Scenario 2 (a 50% tax on wages): You earn two dollars, of which the government takes half. With your remaining dollar, you can buy either one scone today or (after saving and earning interest) two scones tomorrow. Either way, your consumption is cut in half. In effect, current and future scones are both taxed at 50%.
Scenario 3 (a 50% tax on both wages and interest): You earn two dollars, of which the government takes half. With your remaining dollar, you can buy either one scone today or (after saving, earning a dollar interest, and paying half that dollar in taxes) one-and-half scones tomorrow. If you spend today, your consumption is cut in half (from two scones to one); if you spend tomorrow, your consumption is cut by 62.5% (from four scones to 1 1/2). In effect, current scones are taxed at 50% and future scones are taxed at 62.5%.
Now in scenario 3, the government collects more money, and can therefore afford to lower the tax rate below 50%. But it remains the case that future scones are taxed at a higher rate than current scones.
So here’s the right argument: First, economics teaches us that everything should be taxed at the same rate to avoid unnecessary distortions. Second, it follows that current and future scones should be taxed at the same rate. Third, therefore there should be no tax on interest.
Now you might say that “taxing everything equally” is an ambiguous criterion—if you apply it to sources of income you get one answer, and if you apply it to what the income gets spent on you get another. What this overlooks is that the criterion was not just pulled out of the air; there are reasons for it—reasons you can find in any good microeconomics textbook. And if you interpret the criterion in light of the reasons we adopt that criterion in the first place then there’s no ambiguity. A scone today should be taxed at the same rate as a scone tomorrow. QED.
A not terribly illuminating quibble: if you’re converting the wage & interest taxes to consumption taxes, wouldn’t we say they’re a 100% tax on scones in the first day (or the second if there’s no interest tax) and a 167% tax on scones in the second day when there is an interest tax?
This is semantics, so of course all the reasoning still holds.
Beautiful exposition of the Chamley-Judd result!
So, do I understand correctly that you’re not really advocating the elimination of “all taxes on interest, dividends and other forms of capital income”? Not directly, anyway. Rather, you’re advocating a shift in the tax basis from a mix of income and consumption to taxing purely consumption. The elimination of taxes on interest, dividends and other forms of capital income is simply the logical outcome of making the consumption tax apply equally?
Windypundit: Yes, this is right.
Ryan Yin: Yes, fair enough.
Great article, I never thought about true nature and interest rate tax consequences…
“A scone today should be taxed at the same rate as a scone tomorrow.”
Uh, not if you’re a Keynesian who’s trying to prevent a paradox of thrift.
How do you apply this reasoning to dividend or capital gains income from assets that person didn’t pay for with income-taxed income in the first place?
Cos: The same argument works. You come by an untaxed dollar today. If you spend it immediately, you’ve paid no tax. If you save it, earn interest, and get taxed on the interest, you do get taxed. So current consumption and future consumption are being taxed at different rates (one of those rates being zero).
Taxing everything at the same rate is good IF you can tax everything, including leisure. We can’t tax leisure so it is no longer true that it is good to tax everything else the same. (Thanks to your hero, Frank Ramsey, for this result.) Leisure is consumed mainly in a household’s future, when it retires. More future scones are needed when more future leisure is consumed (longer retirement). Taxing future scones more highly than present scones is just an indirect way of taxing leisure, and that is a good thing.
Neil,
I don’t think the non-taxation of leisure breaks the Chamley-Judd result. Maybe you can make a life-cycle, OLG argument (though bequests make me rather skeptical), but I don’t think leisure per se gets you there.
chamley-judd assume infinitely lived households. retirement is not an issue for them. leisure is taken at each point of time, not in the household’s future.
Great exposition to drive home the double taxation point. In practice though aren’t things even a bit worse? Not only is my pre-investment amount is taxed when I earn it, and the return on it taxed in the future, but in between, it faces a corporate income tax.
Hmmm,
following the same logic in scenario 2, if you save a quarter dollar every day and alway re-invest the capital + interest, after 7 days you will have 8 dollars in the bank. With 8 dollars in the bank you will make every day twice as much as working and you pay 0% taxes.
Basically, every 4 dollars in the bank can be considered as a clone of yourself, that earns you two dollars per day. Clones should pay taxes as well.
T.
Neil,
Like I said, you can argue that life cycle issues overturn this result (though you have to explain away bequests to get away from the dynastic model — hard to do if one were arguing for, say, estate taxes), but it’s not leisure or its non-taxation that would drive such an argument
MW: Not only is my pre-investment amount is taxed when I earn it, and the return on it taxed in the future, but in between, it faces a corporate income tax.
It’s worse than that. If you earn a dollar and use it to buy a share of stock, you’re taxed when you earn the dollar, again when the corporation earns income, again when the stock pays a dividend, again when you sell the stock and earn a capital gain, and again when you pass it on to your heirs. I had a piece in the Wall Street Journal several years ago illustrating how this can easily add up to the equivalent of a 95% tax on future consumption.
Ryan,
My point is not which answer is right, but that the answer depends on empirical knowledge–whether or not individuals behave like finite-lived mortals who retire in old age and are not altrustic to their kids, or whether they act like Chamley-Judd infinitely-lived individuals who do not retire. Steve, with his platonist proclivities, wants to make this a matter of pure mathematical logic, so that he can put the truth out there in the platonic realm. In fact, you can’t get the right answer without knowing how people behave. Its the old Plato versus John Stuart Mill thing.
TjD: The clone *is* (implicitly) paying taxes. Because you were taxed when you earned the dollar, the clone has only half as much capital as he would otherwise, and therefore earns only half as much interest otherwise—forever. That’s effectively a tax on the clone, and it lasts forever. Any *additional* tax on the clone means he’s being taxed at a higher rate than you are.
The problem, as it were, is reality.
“First, economics teaches us that everything should be taxed at the same rate to avoid unnecessary distortions.”
What this ignores, for the mere peons in the world, is that taxing everything at the same rate will NOT avoid unnecessary distortions, because distortions come from many sources (not just taxes).
To further bring (simplified) reality into the mix, we have (at least) two groups of people:
1. The ultra rich, who gain lots (most) of their income from interest, dividends, and other forms of capital income.
2. Everyone else, who have to perform “work” to gain income.
By following your strategy of “no taxes on interest/etc.”, the first group pays ~no taxes, while the second group pays all the taxes.
(This could be used as an excellent reason to drop payroll taxes altogether and just use a sales tax instead, but sales taxes are often also regressive toward the poor as well…)
Now, while this may be “better” economics, can you really convince 90+% of the population that this is actually better for *them*?
Especially when we have stories like this:
http://taxprof.typepad.com/taxprof_blog/2007/06/warren-buffet-p.html
where Warren Buffet pays ~17.7% of his income in taxes (because most of it is capital income, which is currently taxed at a lower rate than other income sources), while his employees pay a higher rate of 32.9% because they “work”.
Now certainly, the value of his 17.7% is a higher than the value of his employees 32.9%. However, I doubt he’s living paycheck to paycheck, while I wouldn’t be surprised if at least some of his employees were (as many of the “poorer” people often do).
“Fair” depends not just on your sources of income, but how much income you actually have.
Then there’s the “market distortion” question. The market is *already* distorted in various non-taxable ways (often because of politics, monopolies, etc., etc.), very often toward those “in power” (usually meaning “those with lots of money”). Changing tax law in the manner described here won’t remove these distortions, it will in fact make them worse (as the rich will get richer, thus gaining a larger influence over the market).
Now, perhaps the market being distorted toward the rich is a good thing for everyone involved, but that should actually be argued, instead of (effectively) implied by going from the tax angle.
Steve:
Steve, the clone is not paying taxes, not ‘effectively’ nor ‘implicitly’. Paying taxes implies moving money from someone or something to a government, there is none of that with the clone. I would agree the clone is suffering from a lost opportunity, but then that happens with all forms of taxation.
The clone keeps generating money ad infinitum without contributing to society.
T.
Jonathan Pryor: As I said, you’re welcome to make other arguments about this issue, as you have. That doesn’t turn the inital fallacious argument into a correct one.
Taxing labor and capital has been mentioned, but what about taxing land? I mean “land” in the classical economic sense of naturally-occurring resources whose supply is inherently fixed (thanks, Wikipedia!). A tax on land doesn’t discourage production of land, since the quantity is fixed by nature. Moving taxes from things whose production is discouraged by higher taxes over to land would seem to be more economically efficient.
(I might as well be the token Georgist.)
Steve, it seems like Jonathan Pryor’s argument is a good one and is not really a different argument from yours. Like you, he is objecting to one policy option (not taxing any form of capital income) based on the distortions of this policy (a bias for capital income earned largely by the wealthy over income from work earned by the not wealthy).
Now, I’ll make a different argument: Why doesn’t JP’s argument suggest that capital income should be taxed at a lower rate (but greater than 0) than other forms of income in order to reduce the “biased” distortions of either taxing capital at the same rate or not taxing it at all, thereby having workers and investors share a portion of the distortions in some more equitable fashion.
And, to offer another argument, if we apply the “amnesia principle” here, wouldn’t we prefer a world with taxes on capital over one with no taxes on capital, assuming you didn’t know whether you are a member of the smaller investor class that receives most of its income from capital or the much larger working class that receives most of its income from wages?
Along the same lines, if you subscribe to the EGR, wouldn’t you prefer some level of taxation on capital over none? I’m not an economist, but is economic theory so clear that ANY taxation of capital so impedes economic growth that it overwhelms the bias against labor, and the burden placed on labor of carrying all or the lion’s share of the cost of government? How does the marginal value of a dollar lost by a larger number of lower income wage-earners weigh against the marginal dollar gained by a smaller group of investors?
Of course, I understand that endorsing either of these options would require you to abandon your support for eliminating all taxes on capital income, but do you think they also allow JP (and me)to escape the shame of continuing to support taxation of capital income?
I think John D’s comment was too easily overlooked — policy, and therefore policy discussion, is currently heavily influenced by the idea that a scone today should NOT be taxed the same as a scone tomorrow, so it would seem you have made a correct argument based on premises that no one who disagrees with you would accept. “Everything” temporally and “everything” intertemporally are two entirely different things.
Steve,
If this is what passes for logic in the economic community, it is not surprising how most of your colleagues missed the bubble. Let’s recap your scenario:
Scenario 3 (a 50% tax on both wages and interest): You earn two dollars, of which the government takes half. With your remaining dollar, you can buy either one scone today or (after saving, earning a dollar interest, and paying half that dollar in taxes) one-and-half scones tomorrow. If you spend today, your consumption is cut in half (from two scones to one); if you spend tomorrow, your consumption is cut by 62.5% (from four scones to 1 1/2). In effect, current scones are taxed at 50% and future scones are taxed at 62.5%.
You seem to conveniently ignore the dollar that it still in the savings account. Very clever. You want your person to have his scone and eat it too. If your person pulls his dollar out of savings on the second day to buy scones, he ends up consuming 2.5 scones, not 1.5, and he is in the same position at the end of the second day as in the consume immediately situation, no savings and no scones. The difference is that he actually has consumed 0.5 more scone in the second scenario despite the “double” tax on interest income.
You’ll have to forgive me if I am not shamed by your irrefutable logic.
Scott, I’m not sure where your extra dollar is coming from. A person makes two dollars. One dollar goes to taxes, leaving one dollar for scones. He buys a scone.
Another person makes two dollars. One dollar goes to taxes, leaving one dollar. He saves it, earning another dollar. Fifty cents go to taxes, leaving one dollar fifty cents for scones.
No dollars are in the savings account.
He works two days and takes home two dollars. He nets fifty cents on interest. Seems to me like it adds up to $2.50.
Scott: how come you didn’t notice the $2 that went missing in scenario 1? or the $1 in scenario 2? We’re only talking about how we spend one day’s income (presumably the second day is the Sabbath, or something).
Ah, I see. So in your initial post when you wrote “The difference is that he actually has consumed 0.5 more scone in the second scenario despite the “double” tax on interest income” you were assuming that he put in another day’s work for the extra scone. Which misses the point, which is to see what happens to a day’s income under various circumstances. If you add in a second day’s income to the saver, you should add one to the spender as well. He gets another scone at the end of the second day, and you’ll find that the percentages mentioned in scenario 3 still hold.
Scott: I cannot figure out what extra dollar you are talking about. I must have said something that confused you, but I’m not sure what it is.
Steve,
Since the man who saves his dollar spends it on the second day, Scott seems to be assuming he’ll have a second day’s income for the scones.
I wish I had what some of you guys are smoking. Fine, let’s take a one day earning scenario. First the spender. He works one day, nets one dollar, consumes one scone. The second day, since he didn’t work, he doesn’t eat. (Sucks to be him.) Total scones consumed for the spender after two days of living is one scone. Now the saver. He nets one dollar the first day but saves it. He consumes no scones. (Bummer.) On day two, he doesn’t work so he makes no money but he withdraws his dollar from the bank along with the fifty cents he netted in interest and buys one and a half scones. So the total scones consumed for the saver is one and a half. Clearly the saver consumes more scones than the spender distorting the market for scones toward later scones. So, according to your no distortion position, we should tax interest earning at 100%. As far as your percentages go, I will be glad to address them as you explain why, in your scenario, the save should have four dollars after two days.
Scott, since your numbers now agree with the scenario in the original post, I’m not really sure what you’re disagreeing with.
Your post just shows that people who earn interest on savings get more money and thus more stuff. I don’t think anyone would disagree with that.
Could you explain what it means for interest to “distort” a market to savings? That seems to me just a normal part of the market, and not in any sense a distortion. Do you think any interest on savings “distorts” a market in this way? If so, who holds the “no distortion” position against all interest?
I believe this arises as a result of undue emphasis on the concept of “wealth” rather than lifetime consumption. For instance, the possibility of “wealth condensation”, where those who are already wealthy have the means to reinvest and acquire more wealth, may initially appear unjust. If Bob has acquired a nest egg of $10 million over his working years whereas Jim has nothing, there appears to be a significant inequality. However, the consumption of the two men over their lifetimes is less obvious and thus is more easily overlooked. Hence it is easy for populist politicians to rail against the “concentration of wealth” and propose remedies against this in the form of taxes on capital income.
Sincec Jonathan Pryor is only concerned about “the rich” not paying taxes, his concerns could be allayed by combining a flat tax on income from work with a steeply progressive tax on interest/investment income. What I have in mind is a system in which I do not pay any tax on interest/investment income, but Warren Buffet does. Since Buffet sees nothing wrong with paying more tax, everybody should be happy.
Otoh, I do not mind if Buffet keeps all his income from investments, since he earned it: my only concern would be if all this money were to remain in his family for generations. So I would be happy with some sort of steeply progressive tax on inheritance, if it can be made to work.
A Georgist tax on land also sounds good to me.
Snorri,
Suppose two people make exactly $X (wage income plus any initial assets) over the course of their lifetime. Person A spends all the money immediately, while person B consistently saves most of the money. By construction, the two people’s lifetime wealth are identical (in the sense that the set of all possible combinations [including timing] of goods and services over the course of their lifetimes are identical). On the basis of what moral principle are you claiming it’s obvious that the latter should pay more tax?
Now consider two families of N>2 generations with exactly the same flows of wage income. Again, their total options are exactly identical. Suppose one family tends to save more. By what argument should we de facto transfer resources from one to the other (by taxing one more than the other) on grounds that one family is “richer” than the other? (As I suggested earlier, if there were ever a case where the dynastic model is clearly more appropriate, it’s the inheritance tax.)
Oops, that fourth line should read “…the set of all possible combinations [including timing] of goods and services that they could choose to buy over the course of their lifetimes…”.
Ryan: I did not claim that the morality a steeply progressive tax on interest/investment is “obvious”. What I said is that such a tax would allay Jonathan Pryor’s concerns. I also said that, since Warren Buffet apparently does not mind paying such a tax and I do, the threshold should be set well over my interest/investment income.
In the case of inheritance tax, it seems to me that there are at least 2 ways of looking at it. One is, by what right Buffet’s children inherit more money than I do? the other is, doesn’t Buffet have a moral right to dispose of his wealth any way he likes? But let’s leave both of those aside. The fact remains that, IF we accept Jonathan’s concerns, then an inheritance tax seems to me a better solution than a tax on investment income.
Snorri,
I still disagree. Suppose we accept Jonathan’s concerns and decide that the rich should pay more than the poor. But that’s not what the interest tax does. It taxes different people/families with the same total PDV wealth at different rates.
If Buffet wishes to pay more, he’s entirely capable of writing a check. Surely the discussion is about making people who don’t wish to pay more do so anyway, right?
can you put a link to the WSJ article you wrote you referenced?
Suppose we accept Jonathan’s concerns and decide that the rich should pay more than the poor. But that’s not what the interest tax does.
Hmm… it depends on the definition of “the rich”. If “the rich” are those who have a huge interest income, huge enough that they can live the high life without working, then I don’t see any way to tax them higher than “the poor” except with an interest tax.
What’s PDV?
PS:
If Buffet wishes to pay more, he’s entirely capable of writing a check. Surely the discussion is about making people who don’t wish to pay more do so anyway, right?
Actually, I am much more concerned about not paying interest tax myself than I am about those richer than me paying it. Anyway, they can go to live in a tax haven.
Present Discounted Value.
In what sense could you say one of the two people in my example (or to make it simpler, two people in Landsburg’s example who work exactly one day) is richer than the other? I mean, their options are precisely the same — if W is their wage and R the gross interest rate, then either of them can buy $W worth of stuff on the first day or $WR on the second day or any convex combination of the two. If two people have the exact same real purchase options, then they must be equally rich, right? But the only way the tax code taxes them identically is if either (a) they save precisely the same amount, or (b) there is zero tax on interest.
bpc: The Journal article was March 5, 2001; I’m not sure their online archives go back that far. I will dig out the original text and post it in the next week or two; remind me if I forget.
Ryan: your definition of wealth is acceptable, but I have already defined “the rich” at 1:24 pm (blog time) and I think that my definition is acceptable too, at least in the sense that it is [a] unambiguous and [b] a useful concept.
Meanwhile, walking to the supermarket, it occurred to me that there is a good reason why we should take Jonathan’s concerns seriously: if even Warren Buffet thinks that there should be a tax on interest income, what hope is there to convince “the poor” that only income from work must be taxed? Our only hope is to persuade, if not the poor, then the middle class, that only “the rich” should pay interest tax.
Snorri,
Fair enough, and I’m sure you’re right to focus on the importance of persuasion. But I can’t help wondering, how useful is it, particularly in the long run? For one thing, it obscures the “where exactly did this capital come from” question. For another, it reinforces the notion that two families that have identical, very large total budget possibilities should be taxed at different rates on grounds that one is “richer”. Shouldn’t we also focus on making good and accurate arguments, not making misleading ones because they have the advantage of appealing to people’s mistaken intuitions?
Well… Uhhhhh…. Hmmmmm….. It would appear that I have a bit of egg on my face. I am obliged to chow down a healthy helping of crow. It would seem that you guys are not smoking anything. Your understanding of logic is not wanting. I misunderstood the argument. My original critique of the argument presented here was completely off base. Wrong. Wrong. Wrong. My bad. That being said, the argument presented here is still lacking.
Simply put, there is no reason given to accept the premise, and thus any conclusion that logically follows is irrelevant. According to Landsberg, “economics teaches us that everything should be taxed at the same rate to avoid unnecessary distortions”. Having not studied economics, I’ll have to take him at his word. He says “there are reasons for it—reasons you can find in any good microeconomics textbook.” To this, I would say fine. Again, having not studied economics or had the argument presented here, I am guessing as to what these reasons might be. From my lay person’s understanding, I am assuming that the argument is something to the effect of: We don’t want to tax product A more than product B because that would cause more Bs to be produced than the society would ideally demand. It biases people toward purchasing B when they would otherwise purchase A. The end result is that people don’t actually get what they really want. The makers of A would probably be a bit happier, but that would come at the expense of everybody else. And that is not a choice rational and moral people would make. I think it’s important, though, to note the nature of that reason. It is at root a moral decision. We believe that one group of people should not arbitrarily benefit at the expense of another. The policy that follows is just the logical extension of the moral position.
Let’s say we all agree with the principle stated above and that all products should be taxed equally. I would argue that within that agreement, within that moral sense that compels us to accept that position, is the notion of a given point in time. It’s clear to me why one would choose not to benefit one group at the expense of others. It is not clear which moral principle I am violating if the tax changes over time as long as it is applied equally to everybody and everybody understands that it is going to change and roughly how much it is going to change so they can plan appropriately. What exactly is this distorting? Landsberg just expects us to accept that it is the same thing, but it’s just not. Or maybe I’m missing something obvious. It’s possible, but the burden of argument in on Landsberg. He hasn’t made it.
I would assert that if that argument can be made then he’s pretty much done. All that nonsense about tax rates across time is just window dressing, it’s economese, dressing up a moral position with logic and numbers to make appear otherwise. Don’t buy it? Then why is an income tax even discussed. It’s irrelevant. All he really needs is two conditions: no taxes whatsoever and an interest tax. The difference would be obvious. The person who earns interest has his scones taxed, the person who doesn’t earns interest pays no tax on his scones. Zero versus something. The case is made. But the case is also stripped bare. We don’t have the confusion of x percent tax versus y percent tax. We simply have the question of whether taxing one type of income and not another unnecessarily distorts the market, i.e. creates something that we think is morally wrong. I would argue, not necessarily.
I think it’s now appropriate to revisit the first statement that Landsberg makes in this post. He is “opposed to all taxes on interest, dividends and other forms of capital income.” He wants for us to see that this is not just some arbitrary opinion, it is the only reasonable one, the only logical one. Well, I would argue, Not so much. Everything depends on the premise. The premise was simply asserted, never proven. The position, therefore, is simply asserted, not proven, despite the foray into unnecessary logical games.
Ryan:
Shouldn’t we also focus on making good and accurate arguments, not making misleading ones because they have the advantage of appealing to people’s mistaken intuitions?
In short, yes.
However, having made such arguments, we should also keep in mind that perhaps there is something missing in our arguments.
In this specific argument, I have nothing more to add, except that a tax on wealth seems to me preferable to a tax on interest income; though I still think that an inheritance tax might be best — or least bad, if you prefer.
Scott, I’m afraid you’re completely confused about what is meant by economic distortions. Here’s an article from EconLog on a consumption tax with some details about the distortions created by a tax on capital income. http://www.econlib.org/library/Enc/ConsumptionTax.html
Though if you’re interested in arguing with Steve Landsburg, you might try his book The Armchair Economist. There’s a chapter in there that discusses the distortions caused by taxes, and I think you’ll see that your last post is making some unfounded assumptions.
@Steve Reilly Thanks for the link. It helped me to realize that economists define common words differently than the rest of us which I’m guessing adds confusion to any discussion. For instance the definition of “efficient: : “A tax is neutral (or “efficient”) if it does not alter spending habits or behavior patterns from what they would be in a tax-free world, and thus does not distort the allocation of resources.” I would argue that most people think of “efficient” as a good thing yet the economist’s definition makes no such claim. In fact it could easily be argued that some taxes, the tax on tabacco for instance, are deliberately inefficient and the distortion they produce is their goal rather than something to be avoided.
So, going all the way back to Steve Landsburg’s post that started this he wrote “First, economics teaches us that everything should be taxed at the same rate to avoid unnecessary distortions”. Most of the posts here seem to make the implicit assumption that any distortion is a bad thing but that doesn’t seem to be the case. And if the distortions aren’t necessarily bad then the argument that taxes that cause distortions are bad also falls apart.
Stephen Coy: There is a separate (standard) argument (not provided here, but in all the textbooks) to the effect that distortions are generally bad. (With important exceptions, however.) But that’s neither here nor there: My point here was not to argue against a tax on capital income, but to argue against a particular argument for the tax on capital income. That argument, which I am hearing increasingly often, takes it as given that distortions are bad, quotes the textbook result that distortions are minimized when everything is taxed equally, and then misinterprets what it means to tax everything equally. My only goal in this short blog post was to correct the final misinterpretation (and thereby nullify the entire argument, while acknowledging that other good arguments are still possible.)
Steve Reilly,
Thanks for the link. I see where we are speaking different languages. I should have known better than to try to enter an argument without being versed in the language and history. You sound like you know a bit about this economic stuff. Maybe you could clear up some stuff that was asserted in the page you linked.
Why do they assume that the resource allocation in a tax free environment is optimal? What are the criteria for judging it to be optimal? This would seem necessary for judging one tax to be superior to another.
Why do they assume the people’s “natural” spending habits are the best spending habits? Again, what is the criteria? For example, my dad (true story) spent a good deal of his money gambling on race horses, lost his house, doesn’t have a dime to his name and is living off Social Security. In what way was his saving versus consumption choices optimal?
Are there any data to back up the claim that when subjected to income taxes, “people work less—and choose more leisure—than they would in a world with no taxes”? Or is that just an article of faith. It seems equally likely to me that people would want to work more so they have more to spend. I know for sure of a case where a factory in Central America (don’t recall which one) decided to pay their people piece work to encourage output. It worked. The people produced more, made more, but they made as much in the first couple weeks as they used to make in the month so they didn’t show up for work the second half of the month. It would seem that if economists are so certain about this fact of human behavior, they probably have some very strong data to back it up.
When they say “that the tax wedge created by taxing capital income does enormous long-term damage to the economy”, just how enormous is the damage? Are there percentages available?
Are there any data to show that “[t]axing interest, dividends, and capital gains” results in “less saving than society would choose in the absence of any taxes”? The tax rate doesn’t affect my decision to save, but I’m not everybody.
Do they have defined “optimal” here as nothing more than freely chosen? “If each potential saver could collect the market interest rate, the result would be an optimal amount of saving—that is, an optimal division of resources between current consumption and future consumption. “Optimal” in this sense refers to the amount of saving that individuals, deciding freely on the basis of market prices, would choose to do on their own.”
If “less saving results in less investment, less innovation, slower growth, and lower future living standards than would be enjoyed without a tax on saving” and “[f]uture consumption is reduced by both the extra current consumption and the forgone returns that greater saving would otherwise have produced,” why not just tax consumption more heavily to encourage savings? I’m guessing they would say that is because we couldn’t guess the “optimal” savings rate, which seems nothing more than what people would choose if there were no taxes. But is there any data to show that without taxes people will choose savings rates that maximize investment, innovation and growth, and maximize the standard of living? Or is that too just an article of faith? And what if the data show that something other than the “optimal” maximized these things? Would economists no longer desire “optimal”?
I won’t even touch the question of whether a society should desire a maximized standard of living without consideration of the actual distribution of resources. That is a quite different topic.
Consider this
In exchange for your 1.00 worth of labor you are given a “bagel option” which was purchased or set aside by the bageler at $0.25.
Scenario 1 – Your Wages then would be taxed at 50% (12.5 cents) and when you exercised the option and took possession of the bagel you would pay nothing on the spread since that is capital gains. Meaning that your bagel was taxed at 12.5% but your neighbor’s was taxed at 50%
Scenario 2 – Your wages would be taxed at 50% (12.5 cents) and then when you exercised your option and received your bagel you would pay a tax of 50% on the spread (75 cents * 50% = 37.5) which when added to the tax you paid on your wages (37.5 + 12.5 cents = 50 cents) means that your bagel is taxed the same as your neighbor.
So what is the bigger distortion? That a bagel bought today may cost more in relation to your income than if you wait a day? Or that the bagel you buy today may be taxed 4 times as much as your neighbor?
Scott said “Why do they assume that the resource allocation in a tax free environment is optimal? ”
This is a great question. Historically speaking, tax free environments have lead to rent-seeking behavior. People’s China for example, for a time taxes were illegal. But less connected paid heavily and more connected “guang xi” recieved them. The average person ended up paying less for the same goods under a tax structure than without it, because the taxes funded the local efforts to disrupt corruption.
Steve Reilly:
Why do they assume that the resource allocation in a tax free environment is optimal?
This is not assumed; it is proved (for an appropriate and carefully defined definition of “optimal”). The result is called the Fundamental Theorem of Welfare Economics. Of course, like all theorems, this one requires assumptions, and the assumptions might not always hold.
So what is the bigger distortion? That a bagel bought today may cost more in relation to your income than if you wait a day? Or that the bagel you buy today may be taxed 4 times as much as your neighbor?
Benkyou: A distortion occurs when a bagel (to you) today is taxed at a different rate than a bagel (to you) tomorrow.
A distortion in one market may “occur when a bagel (to you) today is taxed at a different rate than a bagel (to you) tomorrow”, but isn’t the utility gained from social harmony bought and sold on a market?
If you and I work the same job and make the same income, but I do my work for an industry able to pay me in “ISO”s which are taxed entirely at the capital gains rate, which under your scenario means not at all, how does that not distort the value of the utility people enjoy from our financial system?
Benkyou,
If we both get the same income, and that income comes from working, then why do you say that the wage tax only applies to me and not you? If the wage tax is 50% and the employer pays you $100 worth of dollars, you pay $50; if he instead pays you the $100 in shares of stock, or in bonds, or in euros, or in precious jewels, or in gift certificates to Target, you still pay $50. IANAL, and it’s possible there are subtleties in the law that I don’t understand, but the principle here seems pretty clear: you pay the wage tax for income you get for working, and you pay the capital gains/dividend/interest/savings tax when, after receiving your wage income, you decide not to consume it all today but to invest some.
Ryan Yin,
I’m referring to a situation in which a person is compensated with Incentive Stock Options. Which if exercised and liquidated correctly make the whole of that compensations taxed at the capital gains rate. Which in Stevens scenario should be set to 0%
Benkyou: The form of compensation is quite irrelevant. The question is what it’s compensation *for*. You are talking about compensation for labor, which of course, should be taxed at the same rate as any other compensation for labor.
I was actually giving an incomplete picture of a person being compensated with stock options. In the case of Insentive Stock Options, the worker is not taxed at all when he recieves the option or when he excercises them (buys the shares of stock at the option price) and if done right the entire value of the compensation is taxed at the capital gains rate.
Which Steven is suggesting we tax at 0%.
My earlier analysis was based on my understanding of non-qualified stock options, the ISO is a more recent device.
Benkyou:
Which Steven is suggesting we tax at 0%
In other words, you just ignored my previous comment.
Steve,
I’m interested in how you’ve overcome the “amnesia principle” you advocate in The Big Question. If you didn’t know which class you would be born into, wouldn’t you prefer a world in which a small minority who pay most of their taxes through taxes on capital pay all or most of the tax burden, rather than a world in which a large majority who pay most of their taxes through taxes on wages pay all or most of the tax burden?
Now for the sake of simplicity, let’s assume there are two groups in the world: those who receive all their income for wages and those who receive all their income from capital.
Eliminating all taxes on capital income imposes all costs of government on those who receive all their income from wages (a large majority) while allowing the tax exempt owners of capital (a minority) to share in the benefits of government services without bearing any of those costs. (This argument holds whether you think the costs of government are too high or too low.)
Moreover, since those who receive all of their income from capital also have significantly higher incomes (on average) than those who earn most of their income on wages, the marginal value of a dollar lost by the latter is higher than the marginal value of a dollar gained by the former. How is this different from the example (again, from your book) of the $2 and hour Mexican immigrant and the $10 an hour American worker?
How can this be justified under the EGR?
Maybe you’ll respond that the tax exempt capital owners should make a judgment of what those government services are worth to them and make a charitable contribution of that amount to the government. As you point out, they could lie about the value of those services, and considering how much money we’re talking about here, I suspect a lot of them would.
As you also say, we could impose a “mandatory charitable” contribution on the tax exempt capital owners. (An aside: how can charity be “mandatory”? “Charity” strongly connotes a voluntary act reflecting “benevolent good will toward or love for humanity”–Wedster’s.) I assume you mean this would be done by law.
Aren’t we now back around to imposing a mandatory contribution on capital owners–in other words, a tax? How is this mandatory contribution imposed by the government different from a tax?
I also anticipate that you may say, as in a response above, that “other good reasons are still possible” for taxing capital.
But your opening line in this blog post says “I am opposed to all taxes…on capital income.” How has your argument based on economic distortions trumped your advocacy of the amnesia principle and the EGR?
Philip: I am reasonably well convinced that in the long run, the elimination of capital taxation would make pretty much everyone richer. Clearly it would enrich those people who earn most of their income from capital. And what about the people who earn most of their income from wages? Well, in the long run wages are determined by the amount of capital (there is ample theory and evidence for this), and if you stop taxing capital there will be far more of it. So I am fairly confident that the EGR would tell you that any capital tax is a bad policy.
Scott,
Very good questions. You ask, “Are there any data to back up the claim that when subjected to income taxes, “people work less—and choose more leisure—than they would in a world with no taxes”?” There is economic theory backing this up. If something costs more, people want less of it. The data (which,admittedly isn’t definitive–economic data will never be definitive) can be found in Spin-Free Economics by Nariman Behravesh, specifically in his chapter Are Americans Crazy or Europeans Lazy? Different countries have different income taxes. And those countries have changed their income taxes over time. The data, I think, really does back up the case that a higher rate on marginal income discourages work.
One other thing involving resource allocation. Steve Landsburg mentioned the Fundamental Theorem of Welfare Economics. In case you find that too theoretical, or you dislike any of the assumptions, here’s the best way I can answer your question with no economic jargon. I think that “Maximizing” and “optimizing” are different. I hope I’m not nitpicking here, but “maximizing” an investment sounds as though you want the highest dollar value out of it. The best way to get that, most likely, is to put all your money into a current investment. Optimizing, I think, is diiferent, and involves a trade-off between the easy gratification of a current purchase and the better stuff you can get if you invest your money wisely. How to optimize this is in part a matter of personal choice. (This takes us pretty far from the original post, but even if we grant that people are bad at making their own decisions about spending versus saving, and we wish that the government would help people make those decisions better, it’s still not easy to say what sort of taxes the government should levy to do so. Maybe the government should have taxed your father’s spending in order to help him save more. On the other hand, I know people who died rich, which might indicate that the government should have taxed their savings in order to have made them spend more. Is it better to die poor, or to work a whole lot your whole life and not spend much? Beats me. And if you put me in charge of the tax system, I hope I’d admit to that ignorance, and accept that, even if a tax system could, theoretically, help people save and spend just as they should, it’s probably best over all to let people make their own decisions.)
Steven-
You said:
“In other words, you just ignored my previous comment.”
and before that:
“You are talking about compensation for labor, which of course, should be taxed at the same rate as any other compensation for labor.”
I wasn’t ignoring your previous comment, it hadn’t shown up on my list. When I made my 3:01pm post it showed up on my screen directly underneath my 2:37pm post.
But that still doesn’t answer the question. If you and I are both heavy machine mechanics. And I work for Jabil Circuit (as I was and did in 1996 when they were a Fortune 50 company and Forbes stock of the year). And you work for an small machine shop. And 100 percent of your compensation is paid as wages. And about 80% of my compensation that year was paid in stock options. Then even if you and I make the same pay for the same hour of the same work, we will be taxed much differently because most or all of my pay can be taxed at the capital gains tax rate. Which in your scenario, you suggest should be 0%.
In 1996 I was paid enough in real wages at the end of my contract to exercise the option and sell enough stock to pay 130% of my bid price for that years worth of labor. A year later I sold the stock. If I had negotiated for ISO’s as some of the other machine specialists had I would have paid my entire years income at a maximum federal rate of 20%. You would have paid 39.6% on the same pay that year.
It seems like your argument is based on the premise that any capital gains or interest income a person may earn originated from income that person earned as wages.
But some people earn 0% of their income from wages, 100% of it from interest. Their income is based on the value of companies which is generally determined by the quality of that companies products which is generally determined by the people working for it, who almost always work for wages.
As Warren Buffet pointed out, in 2006 he paid about 16% tax on his income. His employees paid about 34%.
Reilly- You said “The data, I think, really does back up the case that a higher rate on marginal income discourages work.”
And I think you are right, but with some qualification. You are right that when much of the economy is depressed and people are strapped, increasing taxes will move people away.
But there is plenty of empirical evidence to demonstrate cases where taxes are raised yet productivity remains the same or increases. As well as the inverse.
The 90s say a huge lift in GDP along with its raise in the marginal tax rate. In fact most of this countries most productive years were under a 70-91% marginal income tax rate.
I use an analogy I call “The Diamond Mine”. Say you are granted mineral rites to a diamond mine but must pay 30% of your revenue in taxes. It’s a one man operation not too strenuous working banker’s hours. Your revenue is $1,000,000 per year so you net $700k. Pretty good gig. Next year they raise the tax to 60%. Now your 40 hours a week only nets you $400k. Do you quit? Do you cut back your hours?
Real life example is the oil companies in Venezuela. When Chavez nationalized 3 out of the 5 major oil companies working in his nation’s waters the remaining 2 out of 5 were the ones that agreed to pay a 90% tax on their revenue.
Benkyou:
Then even if you and I make the same pay for the same hour of the same work, we will be taxed much differently because most or all of my pay can be taxed at the capital gains tax rate. Which in your scenario, you suggest should be 0%.
In other words, you’re still ignoring my comment.
You’re being dismissive, which I appreciate is easy to do with lesser souls such as myself. But you say I’m ignoring your comment and hold my scenario there as evidence of this. So I have to ask, which comment?
I would think that pointing out that it’s possible for two people to do the same job for the same pay and yet for arbitrary reasons or those of privileged opportunity one of them gets to keep nearly all of his earnings while the other pays nearly half of it in taxes, would have some relevance to your comment where you say “You are talking about compensation for labor, which of course, should be taxed at the same rate as any other compensation for labor.”
Are you implying that financial devices that defer taxation, grant tax incentives, or otherwise encourage investment in fledgling companies should be abolished?
Benkyou: I don’t know how many ways to say this. If wages are taxed, then all wages are taxed, regardless of what form they’re paid in.
Benkyou,
Let’s try this: what do you think his comments meant? (Specifically, the ones at 2:46pm 1/30 and 5:54am 2/1. Or my comment at 10:26 1/30?) If you’re reading them as not responsive, you’re misunderstanding.
(Hint: ISOs are wages too.)
Steve, thanks for your response.
I’ll resist citing Keynes on “the long run” and just note that wage earners will suffer economic losses (through shifts in the tax burden to wage earners) until wages finally “catch up” and allow them to enjoy a “net benefit” from the economic growth generated by eliminating taxes on capital. And how confident can we be that wages will rise beyond this point of “net benefit” to eventually compensate wage earners for the losses they incur until the long term arrives. Maybe they do in the “long, long run.”
Second, how do the different marginal values of a dollar (between a high-income capital owner and a lower income wage earner) play into these calculations?
Third, does the amnesia principle allow us to assume one has an advanced degree in economics in order to make our choice?
I understood you to mean for the EGR to be applied by those with a modest knowledge of economics. If so, it seems reasonable, given the complexity and uncertainty involved in these calculations, that a naive economist, using the EGR, could reasonably come to the opposite conclusion about taxing capital.
I don’t doubt that reductions in taxes on capital will increase aggregate wealth. As you say, capital owners will clrealy be enriched. What I’m questioning is whether wage earners ultimately benefit from that increase in aggregate wealth, and whether they benefit more than they would from alternate policies, for example, eliminating taxes on wages. If taxes on wages were eliminated, wouldn’t they work harder and benefit more directly and immediately than if taxes on capital were eliminated? Why wouldn’t aggregate wealth increase in the same way it would by eliminating taxes on capital? It seems that when one introduces the complexity of differential benefits between alternate policies, which is the real world question, our certainty about the net benefits for the majority of taxpayers (wage earners) from eliminating taxes on capital is further eroded.
Finally, and I’m not saying this is a factor in your conclusion, but I can’t help suspecting that many seek zero taxation on capital in hopes of creating an incentive for wage earners to join them in a political agenda to cut government spending, irrespective of its merits or whether wage earners benefit from that spending, as an alternative to wage earners supporting a more equitable division of the tax burden between capital owners and wage earners.
Ryan- “(Hint: ISOs are wages too.)” They are wages that are not taxed until the shares are sold and 100% of the value of the ISO optioned stocks are taxed at the capital gains rate (max of 15%)
Steven- And if your labor is compensated by something other than a wage? After Graduation I worked almost a year for free + medical insurance. A value of about $1400 per month to me because I have 2 kids and my wife has a pre-existing condition plus another $700 per month for my wife’s medication. I “made” about $25,000 that year paid no taxes, and still netted a refund.
I get your point. If there were zero tax on interest and I earned all of my income from interest, then the bagel I buy today costs the same as the one I buy in 2 years. My criticism is that your model presents a “Begs the question” fallacy.
By limiting the criteria of the model (simplifying) you’ve shaped the possible answers yielded by it. Like when Goldline says “gold has doubled in value since 2001” if they had use “since 2000” or “since 1999” the performance of gold wouldn’t seem so spectacular.
To be valuable and not just valid, your model would need to prove that eliminating tax on interest eliminates a market distortion without creating a distortion in some other market.
Benkyou,
This is rather frustrating. When you say, “And if your labor is compensated by something other than a wage?” you are very clearly making no effort to understand what is being said. That sentence literally makes no sense in this context — it literally means “what if your labor is compensated in something other than compensation for labor?” I am not sure what possible answer you expect for that. You aren’t understanding what is meant by “wage” here.
Benkyou: I want to be careful about putting words in Steve’s mouth, and I could be mistaken about this, but I believe he is suggesting that in the scenario you put forward the value of the options should not be taxed under the capital income rates but should instead be taxed under the rates that apply to wages. In other words it is definitionally impossible for your labor to be compensated by something other than a wage- if it is compensation for your labor it is a wage and should be taxed accordingly. If the tax code allows you to treat something received as a wage as capital gains then that is a deficiency in the tax code.
Of course avoiding loopholes here would be complicated, but one could hardly accuse the current tax code of being simple either.
Tagore Smith: You are quite welcome to put these particular words in my mouth, but I have little hope that Benkyou will pay them any attention.
ryan- Okay, by definition a wage is taxed as income. By definition compensation is something given in exchange for something else, in this example labor. What you are suggesting is that these two things must always be the same thing, but that is not always the case.
My point is that a tax code that creates or allows a tax-free form of compensation is going to distort something, even if at the same time it corrects a distortion in the future cost of bagels. The tax exclusion on employer provided health insurance, for example, created a form of untaxed compensation. The resulting distortion in healthcare costs have priced care out of reach for many.
Tagore,(and Steve by proxy)- In my post at February 1, 2010 at 3:23 am, I asked directly and plainly if that was what he was implying. I could not find anything in Steven’s posts to establish this in his model. And when I ask questions, instead of answers I get told I’m not reading him right. After buying and reading Steven’s book on Price Theory (required by the most Keynesian Professor at Oakland University) and then his two most recent books, I had always considered the Author a paragon of explication.
Benkyou: And in his response to that post Steve indicated that he had already answered your question. And he had.
I’m basically predisposed to be friendly to you just because of your name: I like study and burritos and I speak reasonable Japanese. But I have to say that I understand why Steve thought you were being intentionally obtuse, or at least negligently so. If the penny finally dropped for you because of my tremendously explicative post, I’m glad. That was my intention. But I think you ought to go back and look at earlier posts by Steve and Ryan. I think you’ll find that there’s nothing in my comment that can’t be easily inferred from their comments.
Anyway, even if the Author is a Master of Explication when writing a textbook that doesn’t mean that he must explicate everything to you in the comments section of his blog.
At any rate, I mean no offense to you Benkyou-san. But I think if you go back and look at the comments you’ll see that your questions were answered immediately, but that you didn’t realize that they had been.
By the way, the middle paragraph of your comment suggests that you still haven’t quite gotten the point…
Benkyou: Landsburg’s first reply to you after you started talking about different types of compensation for labor was this:
“The form of compensation is quite irrelevant. The question is what it’s compensation *for*. You are talking about compensation for labor, which of course, should be taxed at the same rate as any other compensation for labor.”
What he has said is explicit. ISO, under Landsburg’s argument, should be taxed exactly the same as a cash money paycheck. The ISO should be taxed at the income tax rate because it -is- compensation for labor. The fact that the tax code permits other behavior is entirely besides the argument presented and in no way invalidates the argument.
Landsburg is making a normative, prescriptive argument. Attempting to attack it based on a flawed, positive implementation (the US tax code) of some other ideal is itself a flawed argument.
Okay, I got it. And I do apologize. I read “You are talking about compensation for labor, which of course, should be taxed at the same rate as any other compensation for labor.” as a positive assertion. More like “go outside tonight and you should see a full moon.” or “there should be a bus along any moment”.
It was plainly stated and I missed it. And for my obvious predilection for explaining a thought in as many ways possible I took Steven’s refusal to restate his position as one of obstinate dismissal.
That said…”that doesn’t mean that he must explicate everything to you in the comments section of his blog.” is a very true statement on obligation (and obligation to me in particular). But the blog is an obvious value-added to my (our?) purchase of his book.
I’m not suggesting that he owes anybody anything. But if there wasn’t value in having his writing challenged, there wouldn’t be comment boxes.
I appreciate that the initial argument was to demonstrate the fallacy of one defence of tax on capital, and acknowledges that there may be other arguments. However, some thoughts on the matter. Steven acknowleges that the economic theories underlying the arguments are based on assumptions which may not be absolute. It reminds me a little of the physicist who can predict absolutely which horse will win any given race. Trouble is, it only works for a spherical horse in a vacuum.
In scenario 2 (50% tax on wages), all one need to do is go hungry for one day (no scones) then you can have 1 scone a day for the rest of your life and a exponentially increasing amount of money. In this world, no-one would work more than one day in their lives, and the economy would not exist. Of course, in this world the returns would not exist either, but the principle remains from the figures given. Using this simplistic assumption it is required to tax capital to keep the economy going.
I appreciate that the initial argument was to demonstrate the fallacy of one defence of tax on capital, and acknowledges that there may be other arguments. However, some thoughts on the matter. Steven acknowleges that the economic theories underlying the arguments are based on assumptions which may not be absolute. It reminds me a little of the physicist who can predict absolutely which horse will win any given race. Trouble is, it only works for a spherical horse in a vacuum.
In scenario 2 (50% tax on wages), all one need to do is go hungry for one day (no scones) then you can have 1 scone a day for the rest of your life and a exponentially increasing amount of money. In this world, no-one would work more than one day in their lives, and the economy would not exist. Of course, in this world the returns would not exist either, but the principle remains from the figures given. Using this simplistic assumption it is required to tax capital to keep the economy going. Not taxing interest is a very real dis-incentive to work.
I thought my post was so good I posted it twice! Still, maybe nobody will notice, lurking down here at the bottom of an old(ish) thread.
Harold,
I don’t understand why this is a fundamental problem. First, as you noted, it might seem unrealistic to suppose interest rates are invariant to labor supply. So why doesn’t this imply to you that you would expect labor supply to NOT fall to zero and interest rates to NOT remain constant? Second, you seem to suppose that if you can get any positive amount of income from capital, this implies that you will never work at all. (As a corollary, suppose we do assume that if someone can get one scone per day w/o working they will never choose to work. As long as you let interest rates net of taxes to be positive, this will always be possible. So why don’t you conclude something much stronger — that you need a 100% tax on capital income in order for anyone to work in the long run?)
My third question is perhaps most fundamental: suppose for the sake of argument that real interest rates are constant at 100% (independent of labor supply) and that no one would ever want more than one scone per day. I agree that in such a world, you’d eventually have zero work and everyone living off of interest (though I think if we wrote this down mathematically, “eventually” would be “after an infinite amount of time”). But so what? Why is that bad? It seems to me, given everyone’s implied preferences, that would be the best of all possible worlds — once we get to that point, everyone really is as happy as they possibly can be.
In other words, I agree that wealth effects make people work less, all else equal, but I don’t see why this is a bad thing per se. Being richer is a good thing, and one reason why it’s a good thing is that it allows us to have more leisure when we want it.
Hi ryan yin, thanks for pointing out a few inconsistencies – on further thought I agree in many ways. Ultimately, in order to prevent everyone living off interest you would need 100% tax on interest int he long run. In this case, the long run could be effectively infinite. It is the old story of investing $1 in 1800 and now living off the income. Were I to invent a time machine I would probably do just this. However, your point 3 I disagree with. I could be wrong, it has happened before. If you go 2 days working without eating, you generate enough interest to grow exponentially. Day 1 – earn 2, tax 1, spend 0, save 1. Day 2 – earn 2, tax 1 capital plus interest 2, spend 0, save 3. Day 3 – earn 0, tax 0, capital plus interest 6, spend 1, save 5. After 2 weeks you have enough income to buy 500 scones a day, and still grow your capital exponentially. In this world, consumption is enormous and work is zero, and this is not approaching infinate time, this is after 2 days.
None of this undermines the point of the scenario – to demonstrate that taxing interest breaks the tax everything equally “rule”. My concern is that it is easy to go from a proven rule in theory to assuming that it applies in the real world. This thought addresses the “they are free to make other arguments” part. Of course, given the scenario of the made up world above, it would be great if everyone could have almost infinate consumption with no work, and this is what would rapidly happen. This is, of course, impossible, so therefore the world as set up cannot exist. By creating a paradox, we are alerted to the need to question a bit deeper.
Keeping the “tax everything equally” rule bangs up against other “rules” or desired outcomes. The world above is designed to be simplistic to illustrate a point, and I am now using the same simplistic set-up to illustrate another point.
Lets say in this world there are two types of people, those who can go 2 days working without eating, and those who cannot or will not – this perhaps includes the feckless and those with small children. Those who consume all they earn will remain with a consumption of 1 scone a day. Those who managed to go without at the beginning soon have an almost infinate amount of money. You can imagine the scene: “Jeeves – throw some more money on the fire, I am a bit cold” “Cirtainly sir. How about a pay rise so me kiddies can eat more than 1/4 a scone each a day, sir?” “Cirtainly not, you have already shown that you will work for $2 a day or you would be as rich as me!” At which point Jeeves hits the master over the head with a poker.
So by keeping to this rule, we run up against other rules and rapidly end up with revolution. We increase differences in income to a point where any eventual benefit in total output is undermined by the cost of the perceived unfairness. We would rather be a bit poorer than treated unfairly (as studies of giving money to be shared have shown). This is the same point as made above several times, so I am really just repeating others. Ultimately, I suppose it is the old argument that markets are already distorted, so you need to introduce new distortions to correct for them.
Harold,
First, when I suggested that it would probably take quite a long time to hit the “no one ever works” steady state, I wasn’t saying that it was impossible to hit this state very early — I was saying that people might not want to. This is kind of getting far afield, but the idea is that people like to smooth consumption, so they wouldn’t want to save 100% for several periods — they’d be more likely to buy part of a scone each day and slowly build up capital and work less and less over time, asymptotically approaching the zero work steady state.
Second, and much more importantly, I don’t see why you disagree with my claim that if you supposed that people had such preferences where they really would ever choose to stop working entirely, then such a scenario would be ideal. You say you’re creating a paradox, but I don’t see this at all. It seems like your model implies people value consumption and leisure, and that the utility from consumption maxes out around 1 scone: ok, so if in that world after 2 days they literally get everything they ever wanted, so much better for that world!
Third, I agree that some people might prefer to save more than others, and that this can generate the result that over time, stocks of wealth diverge. What I don’t agree with is why you think this is obviously bad. Think of it this way: suppose you and I both like apples and bananas, but I am always willing to trade more apples for a banana than you are. Surely you would agree that we can arrange a Pareto improvement where I eat more bananas and you eat more apples. This is what we want markets to do. So why if I say “by ‘bananas’ I mean ‘scones today’ and by ‘apples’ I mean ‘scones tomorrow'” would you suddenly say this is NOT optimal?
But suppose you also think there should be some redistribution from the high-savers to the low-savers. That’s possible. Ok, but we already established that the low-savers prefer (or need) consumption more in earlier periods than later periods, and vice versa for the savers (we know this from the revealed preference — by definition, a saver is more willing to forgo consumption today to get a given amount of consumption in the future). So if we’re going to transfer between them, we should do it as early as possible — that’s the consumption that “costs” the savers the least (in utility terms) and is most valued by the non-savers. That is, make a single lump sum tax at time zero on those people who would tend to save. This has the added bonus of, again, minimizing distortion — true, we would be taxing the people who save more than the people who don’t save, but for any given person we would be taxing all their goods equally. So again, even under the assumption that we want to tax savers more, it seems to turn out that we really shouldn’t tax savings.
ryan yin. I will have to think more about the final paragraph. It takes a while for me to get my head around it – and a pen and paper! I will get back to that. Re Point 1 – I was working on the assumption that scones were not for sale in bits – it was one or none. This is a simple model world here.
Second point – the consumption does not max at 1 – it carries on rising. After 2 days they start to get more, and they keep on getting more. It assumes they value higher rates of income and leisure. The paradox arises in that after 2 weeks the consumption has risen to 500 scones a day per person, and nobody is making scones any more because the baker has retired to live on his interest.
Point 3 – the result is not optimal because it results in revolution! Bananas = scones today, apples = scones tomorrow. In the above example, if I prefer bananas, (i.e my kids have eaten the scone), I have no way of getting any apples. If you prefer apples (you can forego your scone today), then in two weeks you can have 500 bananas a day and just about an infinate number of apples. The system in this obviously too simple world results in diverging wealth, which is itself a bad result. We know it is a bad result because people prefer to have less and be equal than to have more and be relatively poor.
I have now considered the last paragraph, and yes, I see it now. In the above example, it would be much better if we could arrange for you to pay me some of your money from day 1, so I could save a bit and still feed my kids. Then we can all enjoy the benefits of the amazingly good 100% interest rate. However this still leaves us the original problem of the paradox. And quite how to arrange that initial payment.
Harold,
The reason why it’s a paradox is that you’re assuming (A) the real interest rate is independent of the labor supply (so interest rates are always 100%), and (B) the real interest rate depends on the baker working exponentially more (or certainly not quitting). But A implies (Not B). If you assume A, you can’t start assuming B halfway through. Alternately, if you assume B, and real interest rates are positively related to the labor supply, then when people work less and less the interest rate should fall. In the first case, there’s no problem and no paradox. In the second case, people don’t stop working.
I don’t understand why “how to arrange that initial payment” is a puzzle — you seem to be assuming taxes on wealth are always possible, except at time zero. For that matter, “time zero” can just mean “just right now”, so that would seem to imply you should only tax capital today but never in the future — which, again, is the “don’t tax different goods at different rates” result.
If people really do prefer equality to money, then I’m not clear why you think we need taxation to achieve this — presumably people should give away their money when they have more than mean income, right? No fair inferring revealed preference or a “this is optimal” conclusion from the results of a political process!
I’m not sure what to take from the “revolution” argument. On the one hand, it seems like we were discussing what is optimal and now we’re discussing what is politically necessary — again the “is vs. ought” distinction seems important. And it also makes me wonder whether it’s literally true that ongoing capital taxation is really the easiest and most efficient way to avoid violence. But even that seems to be missing the point. If someone is making an argument to the effect of “we would all be better off if we had policy X instead of policy Y”, how can the argument be “well, yes, but then there is violent resistance”? Or rather, how can that be the only argument? If Y actually isn’t better than X, or if there isn’t a mutually improving alternative to X, then that seems like the argument that should be made against the policy proposal. But if the argument is purely the violence question, how can that make sense — i.e., isn’t it like saying, “yes, I would prefer this new policy, but we shouldn’t do it because eventually it will mean I kill you”? Is this like a commitment problem — e.g., the reason we should tax savings indefinitely is because those who would choose to save less than most cannot commit to refraining from confiscating savings in the future and/or violence?
Harold-
I think the source of your paradox is the unrealisticly high 100% interest rate, not the capital tax rate.
A) With a capital gains tax rate of 50%, none of the situations you described are eliminated; they’re just pushed back a day or two.
B) The situation you described does actually happen in the real world. People do work until they can collect enough savings to live comfortably on the interest, then they retire. It just takes thirty or fourty years of work to get there. :)
Thanks to you both for your patient discussion. I now see it is not a paradox. If you assume 100% interest, you do not need to specify where it comes from. It is merely an impossible assumption. (This does not detract from its purpose of demonstrating the “taxing everything equally” rule, for which it was created.)
My revolution talk was meant to be tongue in cheek – a result of some having nothing and the rest being infinately rich may lead to revolution. It is only an illustration that the “bad” outcome of extreme inequality has a bad outcome for everybody. This is in the model world, which exists in our heads only to exagerate outcomes to illustrate points.
People tend to have less problem with inequality itself than with being on the “wrong” end of it. Thus people do not give their money away when they have more than average, but would willingly take it if offered.
In our imaginary world, we could see that the inequalities could be very rapidly amplified. We could also see that these inequalities could be addressed by transfering some money from the saver to the spender at the start. I could see a problem arranging that payment. This is because to “desired” outcome is to ensure that everybody saves. However, why should you pay me just because my outgoings are higher? At what rate? What if you pay me enough to save, but I go and spend it anyway? These are not un-solvable issues, but not trivial either.
However, I like the idea. If it were offered to everyone now, a tax on all capital, to be distributed equally, then no further tax on capital intertest ever, would that be popular?. In the imaginary world, you would have to be completly stupid not to invest the extra you recived, and rapidly your pot would grow.
My ponderings were I suppose about what outcomes do you get from the application of this economic theory, and do you need to do anything about it. It seems that you might possibly want to do something about capital income to prevent rampant inequality. Only one way to do this is to tax interest, but this may not be the best way.
However, I have thought of a completely different point – not thought it through fully yet, but bear with me. Is there not an economic “rule” that future is valued differently from today? In effect, we care less about our future selves than our current self. I came accross this in particular about the Stern Report on the economic costs of global warming. There was quite a discussion about the appropriate rate of discounting. If the future scone is not worth a scone, but less than a scone, perhaps we need to tax it more to tax it equally?
I have answered my own question – this discounting and also inflation (assuming a real interest rate of 100% – i.e. inflation 100% and interest 200%) seem to increase the effect of taxing the interest. At least as far as I can work out.
Harold,
Discounting is actually just a preference issue — yes, most (all?) people seem to value the future less, all else equal, than the present, and how much they do so varies quite a bit, but that’s a lot like how some people are indifferent between apples and bananas and others like apples much more. It’s really just like saying there are some goods (“apples today”) that you like a certain amount more than others (“apples tomorrow”). The fact that apples might be more valued than bananas doesn’t tell us we should tax apples more, so neither does discounting tell us anything about capital taxation.
I still don’t see why the changing inequality is a huge issue per se. If I always think apples are better than bananas, and you the opposite, then we should trade. This would lead to huge apple inequality (I would have them all) and huge banana inequality (you have them all). Does this mean I should be mad that you have all the bananas? That seems like more of a feature than a bug — we are better off precisely because I traded my bananas for apples. So if I prefer stuff today way more than you, then why shouldn’t I also trade some of my future stuff for some of your current stuff? Isn’t that trade also the whole point?
Or to put it another way, suppose we start with the same amount of money and we face the same prices and we get the same return on investing. Then in a lifetime sense, we have exactly the same wealth — in other words, we have the exact same lifetime options. If you choose to invest more than I, you will have more stuff in the future, but this isn’t actually inequality of wealth — it’s just that you were more willing to give up future stuff than I was. In that case, both of us made the right decisions and we were equally wealthy.
Having thought some more, I think the discounting thing is a distraction, but I am not totally sure! Anyway, it doesn’t seem to lead in any way to the conclusion that you should tax interest.
The inequality thing, I am thinking that the issue is boiling down to the “rational” actors one, or possibly a “free market”, which is a bit of a can of worms. If everyone has a choice to spend or save, then I am coming round to your view that it is not logical to tax interest. In our simple example world, almost no-one would choose to forgo all the future benefits of saving. If you did choose to do so, one assume you must have a good reason: perhaps you knew you would die very soon. If you turned out not to die, then you could get on the saving bandwagon, just a bit later than everyone else. This only becomes a problem if some people are unable to get on the bandwagon- either because they must spend all their income, or are unable to understand the benefits. I think this is a totally different point, and all economic discussions could end up here, if we let them. That is not to dismiss these arguments, just that they are for another time.
I am still unsure quite why this works. If we take the real world, then somoeone who has pots of money and derives $1000 per month from it – it looks like income. Why should they not pay tax on theirs, when I have to pay income tax on mine! This is, of course, a rhetorical question, or we will be back where we started, and was the point from where I started. It takes a bit of working through to get from there to here, so thanks particularly to ryan yin for bearing with me. I have satisfied myself that there are no inconsistencies within the argument. If I wanted to challenge it, I would have to go further back into the theory.