It’s difficult to think of a more bizarre and foolish policy than the practice of taxing capital. Consider:
1. If it were appropriate to pay taxes on capital gains, why wouldn’t it be appropriate to pay negative taxes on capital losses? Economic theories tend to be symmetrical. And yet capital losses do not result in negative taxes, except in certain limited cases. And why only those cases?
2. Economic theory suggests that two people with essentially identical economic outcomes should pay identical taxes. But consider two people who both bought 1000 shares of Apple stock for $50/share at the beginning of the year. One sold the shares on November 9th at $100 and bought them back 5 minutes later at the same price. Both held 1000 Apple shares at year-end. To an economist those two outcomes are essentially identical. But one person must pay a large tax on capital gains, while the other does not. Why?
3. Long-term capital gains are taxed at a lower rate than short-term capital gains. Why?
4. Corporations can expense the cost of debt (interest payments) but not the cost of equity (dividend payments). Why? Is debt morally superior? Are we trying to encourage debt?
5. Economists tend to believe that real variables are what matter, not nominal variables. Yet, while income tax rates are indexed for inflation, taxes on capital income are not. Why not?
6. If you invest in US dollars and the dollars rise in value, you do not have to pay taxes on the gain. For instance, suppose you bought dollars for oil in March, and sold the same dollars for twice as much oil in September. Your dollars increased in value, yet that gain is not taxed. Why not?
7. On the other hand, gains in currencies such as Bitcoin are taxed. You are legally required to keep track of the current market value of any Bitcoin you acquire, and also the date you acquired them, and also the value of those same Bitcoin when you use them to purchase goods and services. That means if you buy a candy bar with Bitcoin, and the Bitcoin are worth more than when you acquired them, you must report that gain to the IRS. Hopefully you’ve held the Bitcoin for more than a year, so that it is not a short-term capital gain on the candy bar purchase. And this must be done for every single transaction.
But all this silliness pales in importance compared to the fundamental flaw in taxing capital income. The IRS assumes that the value of a future dollar is the same as a current dollar, even though that is clearly not the case.
To see how strange this is, consider the following analogy. Suppose you sold 100 ounces of gold for 4000 ounces of silver. Then suppose the IRS claimed you made a 3900 ounce “capital gain” and demanded taxes be paid on that gain. You’d probably say, “Wait a minute, there’s no gain because the market value of one ounce of gold is equal to 40 ounces of silver. So no profit was made on the transaction.” And you’d be absolutely correct. Fortunately, the IRS isn’t that clueless.
Except when it comes to money. Just as one ounce of gold might be worth 40 ounces of silver, one dollar today might be worth $1.05 a year from today. Suppose nominal interest rates are 5%, which reflects 2% inflation and 3% real interest rate. In that case $1.05 a year from today could be purchased for $1 today. If you invest $1 today and your asset rises in value to $1.05 after one year, you haven’t actually made any gain, as the money you receive is equal in value to the money you invested. A future dollar is no more equal to a current dollar than an ounce of gold is equal to an ounce of silver. Except to the IRS.
Some understand this point, but then wrongly argue that taxes should be paid on capital earnings that exceed the average return on capital. If so, then negative taxes should be paid on capital earnings that fall short of the average. However, under that regime the IRS would collect zero net revenue from capital income taxes.
A simpler and fairer solution would be to abolish all taxes on capital, and start over. Think about what the tax system is trying to achieve, and implement a tax system that achieves those goals in the fairest and most efficient way possible. In my view that would be a progressive consumption tax.
READER COMMENTS
Andrew_FL
Sep 26 2015 at 11:57am
I agree with the normative point you are making in 2, but I’m a little confused how positive theory gets you a “should.”
Kevin Erdmann
Sep 26 2015 at 12:31pm
Great point about symmetry. Consensus is that active portfolios underperform after costs, on net, but then on topics like carried interest, we implicitly assume that active management is a net gain that should be taxed.
There are a lot of unspoken assumptions about smart money vs dumb money that I think inform these debates in a way that isn’t enumerated. I get the feeling from some financially unsophisticated people that have told me they would like punitive taxes on short term trading profits that they vastly overestimate the amount if short term trading that is done for retail investors, like market volumes are mostly just a fee extraction scam. This is why those taxes are politically popular, but this reason can’t be used by a respectable economist to defend them, so the popular reasons for these taxes and the putative reasons tend to diverge.
warren meyer
Sep 26 2015 at 1:00pm
I go back and forth on this and really don’t have the academic chops to argue implications on growth of taxing capital gains vs. not. As a business owner, though, let me offer this counterpoint:
Let’s consider two investors. Investor A buys a piece of land and builds a campground on it, intending to run the campground for decades. Investor A gets her return on investment from the profits each year running the campground, profits that are taxed as regular income in a pass-through from her S-corp. (Full disclosure: In my business life, I am essentially investor A).
On the other hand, Investor B buys the same piece of land and builds the same campground on it, but in about a year Investor B sells the newly developed facility, making a profit on the sale over his original investment. Investor B likely will pay taxes on this gain at reduced capital gains tax rates.
But why? When Investor B sold the property, the price he got was probably something like the present value of the expected cash flows from operating the campground. Both Investor A and B created essentially the same value, but Investor B took the value as a single lump sum rather than as a stream of income over time. In a sense, investor B took the lump sum option from his lottery ticket while investor A chose the payments over time. Why is Investor B’s approach preferred in the tax code? Or, stated another (possibly more inflammatory) way, why does the tax code favor asset flipping over long-term operations?
ThomasH
Sep 26 2015 at 1:29pm
No, the thing to do is to abolish all taxes on income and start over with a progressive tax on consumption.
Pending that, we should index capital gains and losses and tax them (and dividends) as ordinary income at the average of the marginal rates over the holding period. Yes this would “double tax” capital income, but it would keep the pressure for a change to consumption taxation.
Kevin Erdmann
Sep 26 2015 at 1:41pm
Warren, I think you would be better served to think in terms of an entrepreneur and a creditor, even if both are the same person. One is earning income through work and the application of skill and the other through deferred consumption and risk.
Locke
Sep 26 2015 at 2:33pm
Replace income taxation with land value taxation.
Andrew_M_Garland
Sep 26 2015 at 3:22pm
[Comment removed. Please consult our comment policies and check your email for explanation.–Econlib Ed.]
Scott Sumner
Sep 26 2015 at 3:22pm
Andrew, Yes, I suppose that’s a bit vague. The “should” comes from the fact that any plausible economic model would treat these two people the same, so it seems implausible that the model would yield different outcomes for taxes paid. But perhaps there is something I overlooked.
Kevin, Good points.
Warren, We should not tax people with different holding periods differently, I agree. The hard part (I think implicit in your example) is separating labor and capital income when there is an entrepreneur involved.
Thomas, You said:
“No, the thing to do is to abolish all taxes on income and start over with a progressive tax on consumption.”
That’s actually my claim too.
Locke, There’s a lot to be said for land taxes.
James
Sep 26 2015 at 3:30pm
I am curious to know how Scott’s recommended progressive consumption tax would be implemented. Would someone need to track my expenditures so that as I have spent more and more, I can be taxed at a higher rate?
This post makes many great points but ignores the more pressing problem that taxes are just too high. If taxes were reduced and offset by spending cuts or an increased reliance on consensual transactions, the specific rules that determine what to tax would be far less important.
This might seem like an odd suggestion but what is odder is how so many accept without argument that that the government must be funded by taxes. Yes, taking people’s money is easier than bargaining for it but millions of people and organizations manage to fund their spending entirely by relying on voluntary and consensual transactions. If the people in the government lack the wisdom to find ways of obtain funding without resorting to taxation, maybe that speaks to their competence in deciding how to spend whatever money they get.
Jason
Sep 26 2015 at 3:40pm
We have land value taxes in Denmark, and they are by far the most unpopular tax (and there are many to choose from). I sympathize both with the efficiency arguments and with the notion that it is unfair to be forced from your home because of rising taxes (due to rising values). I’m not sure how to reconcile those two.
In any case, there is no way it would be politically acceptable to entirely replace income taxes with land value taxes.
R Richard Schweitzer
Sep 26 2015 at 5:01pm
63 years have past but Stanley Surrey’s words still ring:
all that seemed to fit with what had been laid on 9 years before in “Naval Science:”
Discipline and revenues are still required despite the desires for, and values of, justice and logic.
Maniel
Sep 26 2015 at 7:09pm
Scott,
I would like to offer two goals and three recommendations.
Goal 1: Taxation which is less damaging to our economy.
Rec 1: A flat tax (using Friedman’s negative income tax) where all income, from work, interest, and capital gains, is treated equally. Rationale: travailler plus pour gagner plus (Sarkozy), i.e., no penalty for working more, for low earners or high; the “wealthy” (who are not identical to those whose work income is high) pay their “fair share.”
Rec 2: Zero the corporate tax. Rationale: people should pay taxes when they actually take possession of the money; re-investing in the company generally strengthens the company.
Goal 2: Federal neutrality with respect to spending or saving my own money.
Rec 3: No federal consumption tax. Rationale: it is regressive (I prefer neutral) and it is an important tool of state and local governments. For the feds to do it too would be “piling on.” The purchase of goods and services generally increases quality of life.
Of course, if we could ever shrink federal spending, that would be best, but until then…
AS
Sep 26 2015 at 8:04pm
The tax code is not designed by economic theorists to maximize social welfare or efficiency. The tax code was designed by politicians and bureaucrats for their own personal gain. To understand why the tax code is the way it is you have to analyze the political decision makers. For example, most of your proposed changes would reduce the tax burden of the wealthy, which is a political unpopular maneuver.
John Thacker
Sep 26 2015 at 10:28pm
Also for 2 note the asymmetry that if a capital loss was occurred, then under that scenario it would disallowed as a “wash sale,” though as you note there’s no wash sale equivalent for deferring gains.
For 3, the explanation I’ve long assumed is that otherwise most things currently taxed as wage income could be represented as short term capital gain through clever tax treatment. (Think of companies that allow wages to be diverted to employee stock purchase plans that purchase at a discount.) The holding period is designed to separate out those from “real” capital gains, even though it fails for other edge cases. One thing about tax treatment is that it runs into the continuum issue; it is easy to say that A is just like B, and B is just like C, and so on, but at the same time is seems clear that A is not at all like F or G. So you have to draw an arbitrary line between two seemingly quite similar things. (The treatment of carried interest is another one of these areas where people make competing comparisons to similar seeming types of income. Another boundary issue is when does fractional private jet ownership really become more like commercial flights, something that Warren Buffett’s NetJets fought the IRS for years over.)
Obviously not a problem either with a system that taxes capital gains at the same rate as income, or a progressive income tax.
Alex
Sep 26 2015 at 11:11pm
Good points, but I have some objections:
1) My understanding is that tax losses can be carried forward. So its kind of a negative tax, though not a perfect substitute.
2) Taxing “unrealized” (not cashed) capital gains would generate the following problems:
a) People may not have liquidity, so you would have to authorize to pay taxes with assets other than currency, such as shares, commodities and even jewellery.
b) Many assets don’t have a transparent market where you can check the price in real time (most notably real estate)
3) Agreed, its nonsense.
4) But debt is taxed at the hands of the lender.
5) Agreed, its nonsense.
6) See point 2
7) See point 2
bill
Sep 26 2015 at 11:34pm
Warren,
The buyer from investor B will operate the campground, earn the same income as investor A and pay taxes like investor A. So that’s the argument against taxing investor B for the gain on sale.
David W
Sep 27 2015 at 12:04am
The IRS taxes you when you have cash in hand. What they want is cash. They don’t really care about economic efficiency, they care about getting cash. Most of your puzzles have the same answer: if it were more logical, IRS agents would have to work harder to get the money. It might be less clear when you owe money. You might owe money when all you have is real capital, and they’d have to take and liquidate it.
Prakash
Sep 27 2015 at 1:22am
Jason, the standard exemption is a normal concept in income taxation. There is no reason that the same concept cannot be used in land value taxation as well. The first 100000 or whatever of land value is tax free per person. Beyond that, a rather high lvt kicks in (on the (landvalue – 100000)). I believe these are essentially solved problems.
Old people may have to move to cheaper places to live on a retirement income, or choose more modest accommodation. Neither of these are killer objections.
Don Boudreaux
Sep 27 2015 at 5:58am
John Stuart Mill – in his Principles of Political Economy (Book V, Chapter II, paragraphs 22 & 23) – addressed this issue clearly and in a way that I believe Scott would endorse:
B Cole
Sep 27 2015 at 6:49am
I agree; cut the federal government in half and go to a national sales tax, and taxes on fossil fuel consumption. Unlike Sumner I am open to vice taxes. But no more income taxes. On funding Social Security and Medicare I concede I am not sure; but perhaps we can finance those systems through QE for many years and worry about it later .
ThomasH
Sep 27 2015 at 7:35am
@James
Progressive income taxes do not require tracking people’s expenditures. Basically it’s a (more) progressive income tax with a deduction for savings and re-investments. The tracking is movement into and out of savings/investment vehicles. Most of us without much “capital” income and with fairly low annual saving out of “ordinary” income would not notice any difference except (depending on the progression) slightly lower rates.
Although a lot of the complexity in income taxation comes from the “need” to make distinctions between ordinary and capital income and and ordinary income/capital income and capital gains, a rel life consumption tax will not be free of complexity.
There will probably still be some expenditures that will not be treated exactly like either “consumption” or saving. Will “consumption” on health care be taxed? Charitable gifts? “Investments” in self improvement? Education of children? Do we/how do we maintain negative rates for low income/consumption people? Does Silas Marner get a tax break? And the biggest conundrum at all is how to treat inheritances. Is what you pass on taxed like your last definitive blast of “consumption?” Or is it taxed only when the heirs blow it?
And although business taxation per se would be eliminated, business would still have to report and account for income that would be imputed to the owners [and whether imputed profits or just dividends would go into personal income is another issue] and this would leave plenty of room for mischief treating different economic activities (percentage depletion, credits for “green” investment) differently as the present corporate tax code does. The political desire to hide subsidies will not disappear.
Notwithstanding the difficulties, I’d prefer to make hard choices starting from a position that makes more sense to me intellectually. An additional point is that we could make incremental approaches to a progressive consumption tax by raising (eliminating?) caps on 401 K contributions and raising top rates.
ThomasH
Sep 27 2015 at 7:53am
@ James (2)
I don’t start from an assumption that tax collections are too high and that total federal expenditures should decrease. While it is easy to identify our favorite wasteful or unnecessary expenditure (and probably a lot of the most wasteful/unnecessary ones are NOT easy to identify) I can also think of things I’d like to spend more on which includes substituting explicit expenditures and transparent tax subsidies for hidden ones. To the extent that we have a tax system that has fewer dead-weight losses (as I’d argue that a consumption tax would) this argues on that margin for greater expenditures.
What we should spend taxpayer money on and what to regulate and how needs to be looked at line by line and regulation by regulation using cost benefit analysis. Skepticism about expenditures (or regulation) is OK as a stance as long as it is not just clunked in as a policy (like our so-called “sequester.”)
Nathan W
Sep 27 2015 at 8:08am
1) capital losses – you can just carry them over to the next year. Almost no one realizes capital losses in the long run.
2) I guess the non-symmetry is strange. But there needs to be SOME rule of when the capital gain is realized. The rule is that the capital gain is realized at the time of sale. The only way I can imagine to get rid of the non-symmetry while maintaining the essential character of capital gains taxation would be to tax the capital gain every single year at its December 31 market value regardless of whether you hold or sell. (I’m not American, so I’m not quite sure if capital gains are applied quite the same there.) This might be fine for shares, where you can sell a small amount to cover the tax, but for other classes of illiquid assets such as real estate, calculating the capital gain on current market value would force many people to divest at inopportune times.
Progressive consumption tax: consider that income taxes (which include capital gains, since a capital gain is income) penalize American labour, whereas taxing consumption instead would apply a greater share of taxes to foreign products relative to the status quo of low consumption taxes.
On a practical note, though. How would you make a consumption tax progressive? Do you could how much money people have left over (didn’t consume) and calculate based on reported income? Do you create different classes of goods, and have various levels of luxury goods?
I tend to prefer consumption taxes over income taxes, but I think that any tax system must be progressive to be fair. This question of how a progressive consumption tax could be implemented always stumps me.
I strong agree with AS. Tax codes are not generally set by economists. They are set by politicians. Just consider the vast number of boutique tax credits on offer in the current Canadian election: while they are always marketed as having some vague economic rationale (generally under the tax reduction mantra), it is an obvious fact that each one of these tax credits is a sales pitch for votes which targets specific segments of the electorate. Similarly, in the US, I think the complicated tax code in many ways reflects the aggregate effect of diverse lobbying efforts over many decades, where lobbyists are also political contributors, which gives them sway over politicians (or rather, in a less corruption-oriented form of the statement, politicians compete for donation dollars by offering tax credits which benefit certain groups which are likely to make political contributions to the candidate who can deliver those benefits).
Scott Sumner
Sep 27 2015 at 11:46am
James. One option is an income tax with unlimited 401k privileges. Another is a progressive wage tax. Another is to tax luxury good like expensive homes, cars, yacht, airplanes, etc., at a higher rate. Another is a VAT that rebates to everyone the tax paid on poverty level consumption.
Maniel, You haven’t responded to the arguments in this post. And I don’t agree that consumption taxes are regressive.
AS, My proposed changes might help the wealthy, but they would also help the poor and middle class.
Yes, politicians design our tax code, but what bearing does that have on this post? Did I say politicians don’t design the tax code? Don’t politicians ask for advice from tax experts?
John, Lots of those problems also occur with income taxes. What is consumption and what is a business cost?
Alex, Point one doesn’t conflict with what I said.
I’m not in favor of taxing capital gains, realized or unrealized.
#4, Dividends are also taxed at the level of the stockholder.
David, I don’t agree, a consumption tax is easier to administer.
Prakash, I’d add that America does tax land.
Don, Thanks, I wonder what proportion of modern economists understand this concept as well as Mill. I’d guess maybe 20%
Ben, People with vices suffer enough, no need to punish them with higher taxes.
Nathan, You said:
“But there needs to be SOME rule of when the capital gain is realized. ”
No there doesn’t, we could stop taxing capital gains.
See my earlier responses on consumption taxes.
And I thought a person’s cap. loss write-offs for a given year could be no more than $3000 above gains? Is that right?
ThomasH
Sep 27 2015 at 12:24pm
Nice to agree with Don for a change; it reinforces my view that liberals and libertarians are natural allies, notwithstanding disagreement on many parameter values and what to do in some second-best cases.
michael pettengill
Sep 27 2015 at 2:34pm
2 is factually wrong. It’s a wash sale that does not result in a change in basis for a capital asset.
The rule is in law to both prevent tax avoidance without actual economic activity, and also to allow things like swaps of land under force from triggering taxes, such as a merger resulting stock A being replaced by some number of stock B or a parcel of land needed for a road being replaced by an equivalent parcel.
The IRS has an administrative court system to handle the execution of the law consistently, but ultimately the specific rules used to define a wash sale for an instant can be appealed to SCOTUS just as all case law can.
You are basing part of your argument on a strawman. Legislators are not idiots but pay a lot of attention to economic reasoning. The law long ago took into account your argument in 2 and there after.
Michael Byrnes
Sep 27 2015 at 3:52pm
I like the idea of VAT with rebates for poverty-level consumption.
If we did an income or payroll tax (while abolishing taxes on capital – does that include corporate income?), I imagine the financial engineers who gave use synthetic CDOs and the like would start figuring out ways to recharacterize their labor income as untaxable capital gains. Under VAT, that distinction (labor vs capital income) wouldn’t matter at all as none of it would be taxed.
Andrew_M_Garland
Sep 27 2015 at 4:25pm
About (2) Capital Gains. The explanation is grasping government.
Government would tax capital gains every year if it could (mark to market). But, the investors would have to take other money to pay the taxes. The pushback is large enough to force the government to wait until cash is available from the sale of the assets.
Capital losses are not fully deductible even against capital gains. The reason comes from the movie Goodfellas: (insult) – Pay Me [Bad language at the link]
( https://www.youtube.com/watch?v=3XGAmPRxV48 )
Capital gains shouldn’t be taxed at all, and in that case losses wouldn’t be a deduction. Say I buy a stock for $10,000 and my friend buys part of a house for $10,000. Both of these come from after-tax savings. We both have paid tax on income and this $10,000 remains. My investment increases income for those around me and might result in some net profit to me in the future, which I would then enjoy spending. Why is that profit taxed when the enjoyment of my friend in his house is not taxed?
I bought part of a business and he bought part of a house. What we get from that should not be taxed in either case. As Warren Meyer points out, I will be taxed on the income stream from the investment, and if not me, then the buyer will be taxed on the income he receives from purchasing the stock from me. The income stream is taxed and currently the capitalized value of that expected income is taxed, which is double taxation.
This is also why long-term capital gain tax rates are lower than income rates. This double taxation at full rates is just too much to encourage enough investment, so the government relents on the capital gains tax as a practical matter.
Maniel
Sep 27 2015 at 4:45pm
@Scott: “Maniel, You haven’t responded to the arguments in this post.”
My responses to your arguments, by the numbers, follow:
1. My proposed flat tax would aggregate capital gains and losses.
2. I would treat “essentially identical outcomes” (incomes) the same.
3. I would tax capital gains and losses in the year they are realized regardless of the holding periods.
4. I would eliminate corporate income taxes in favor of taxing the incomes of people.
5. I would treat all income the same, whether “passive” or from work. I would index the tax rate and zero-tax point to the costs of government spending.
6. Short sales of stocks and resales at a profit currently result in reported income (yes, I know, it’s more complicated than that). I believe that this is as it should be and that such gains (and losses) should be treated as any other.
7. Concur: Bitcoin taxation is petty.
Re items 1-4: I concur with your premises; I prefer to base investment decisions on market factors rather than tax implications.
@Scott: “And I don’t agree that consumption taxes are regressive.”
I believe that they are regressive because, in general, the less discretionary funds I have, the higher the percentage of those funds that is likely to go to essentials (food, shelter, clothing, etc.); thus, I am likely to find it necessary to make a higher percentage of taxable transactions than if I had more money. I make this argument without consideration of government assistance such as food stamps, which I would replace with a negative income tax.
J.V. Dubois
Sep 28 2015 at 6:10am
I think that capital gains tax is a necessary evil of taxing any income (including wage income). Since it is almost impossible to differentiate between labor and capital income in modern economy, you have to tax all income otherwise you will end up with large tax evasion.
That said, abolishing all income taxes and taxing only consumption solves this problem.
John Thacker
Sep 28 2015 at 7:34am
Sorry michael pettengill, but you’re the one who is factually wrong. The laws on wash sales only apply when a security is sold at a loss. You cannot defer gains via claiming something is a wash sale. It is, as I said above, asymmetrical. (It is asymmetrical because the person doing the selling is able to time things, so behaviors are asymmetrical as well.)
John Thacker
Sep 28 2015 at 7:51am
Agreed Scott, of course. It’s certainly an answer to the “why,” though. The IRS and others who work on the tax code consider it to be too easy to make wage income look like capital income, so they treat capital income as wage income unless it you can prove it has been deferred for at least a year.
As noted by ThomasH, the same sort of issues would come up with a progressive consumption tax, even though I too favor it. Certainly educational spending and probably health spending, as mentioned, would likely succeed in lobbying to be considered investments “in human capital” rather than consumption. (That is even though modern university tuition contains a fair amount of consumption spending mixed in, having most students live at an upper middle class life.)
My understanding is that there are several ways to achieve a progressive consumption tax. Some ways involve individuals having to report financial transactions, which carries with it lots of complexity and worry about evasion. Other ways, like the X tax, involve looking at all wage income, but have some difficult cases like distinguishing between wage income and business cash flow for the small business owner (and any potential issue with disguising wage income to look like capital gains.) The difficult issues seem to me to be no worse than that with the income tax, and the progressive consumption tax has other advantages, so I certainly strongly support it.
Comments are closed.