I have not read or even skimmed Mitt Romney’s 160-page economic plan; all I know is what I’ve seen in the headlines. So all of this is subject to revision. But:
1) I gather that he’s proposing to shift a greater percentage of the tax burden onto upper-income taxpayers — that is, to make the tax system more progressive — via a steeply graduated tax on capital income (with the zero bracket extending up to the $200,000 income range). Reasonable people can disagree about how progressive the tax system ought to be, and I disagree with Mr. Romney; I’d like to see it made less progressive, not more.
2) Be that as it may, we might at least agree that for a given degree of progressivity, efficient taxes are better than inefficient taxes. Continuing to tax capital income violates that principle. If you want to make the rich pay more, the most straightforward and efficient way to do it is with a steeply graduated tax on labor earnings, not by taxing capital. A tax on labor earnings distorts the incentive to work, but a tax on capital distorts both the incentive to work and the incentive to save.
3) In other words, there are two separate issues here: Should the rich pay more, and if so, how should we make them pay it? I think Romney is wrong on the first, and I’m virtually sure he’s wrong on the second. You don’t have to tax capital to tax the rich.
4) There are a number of popular misconceptions that I expect to arise in comments, for the sad reason that not everybody has read or digested all of my previous blog posts on this subject (try typing the word “capital” into the search box on this page). Let me try to pre-empt one of them: There are a lot of retired rich guys whose income consists entirely of dividends and capital gains; a tax on labor income, you might think, would not touch them at all. The reason that’s wrong is that dividends flow from savings; if you raise the tax on labor income, these guys will have save less and earn smaller dividend streams. (That doesn’t apply to the ones who have already retired today, but it does apply to all future generations, the effects on which collectively dwarf whatever happens today.) So a steeply graduated tax on labor earnings does hit more or less the same population as a tax on capital, but it does it without nearly as many destructive side effects.
Edited to add: And if it’s really really important to you to sock it to the currently retired rich, you can do it with a steeply graduated consumption tax. You still don’t need to tax capital income.
The problem with not taxing capital is that it is *in practice* hard to tell what is capital income and what is earned income.
This is especially true where entrepreneurs are concerned, as they can choose to take their income in varying combinations of dividends, loans and wages, where dividens count for tax purposes as capital income, regardless of whether the business uses any capital.
I believe (this is not an expert opinion) that the ability to arrange one’s affairs to minimise tax is the main motivation for governments’ spreading income/transaction taxes broadly across different categories of transaction.
A secondary motivation may well be that it makes it hard for the taxpayer to tell how much tax he is paying.
Ben is right, a tax on capital is usually maintained as an anti-avoidance device, so that at the margin reclassifying income as capital is less attractive. That said, I don’t think that that will have much effect.
In your tax on capital you seem to be assuming that those earning capital incomes will not work more as a result of the tax on capital and earn more labor income. In fact, making a capital tax contingent on the labor income earned previously, is a proposal forwarded by N. Kockerlakota in New Dynamic Public Finance. The reason is to make leisure more expensive and have people work more.
“That doesn’t apply to the ones who have already retired today, but it does apply to all future generations, the effects on which collectively dwarf whatever happens today”. Eventually – it seems to me that most policy has much shorter term objectives. This together with the disguising of labor income as capital mentioned above probably provides some of the incentive to tax capital.
“You don’t have to tax capital to tax the rich.”
It is not easy to tax the rich – they have ways to avoid paying. The steeply graduated consumption tax – how would that work? Do you just pay a higher rate of sales tax on higher priced goods?
What do you think the problems are with increased consumption wax compared to increased income tax?
To me, it seems there might be some negative effects because it is a sort of surprise double tax. (Eg people will work less out of fear that the government might suddenly increase the consumption tax again in the future.)
In engineering (and likely other disciplines), we have a saying: In theory, theory and practice are the same. In practice, they’re not.
I don’t think it takes an economist (or a rocket scientist) to predict the ~immediate results of a “steeply graduated tax on labor earnings” and a lower (zero) tax on capital: the rich will no longer make any labor earnings.
Currently, it’s not unheard of for some CEOs to take $1 in salary:
http://en.wikipedia.org/wiki/One-dollar_salary
They’re not actually earning $1/year, just $1/year in “labor”; the rest of their pay is in capital. Doing as you propose would just result in more people taking $1/year. Combined with a lower capital gains tax rate, the “rich” would pay less in taxes than they currently do, and our tax rates are already they lowest they’ve been in recent memory.
As for the “steeply graduated consumption tax” you propose in the update, as Harold asks, how would that work? Report to the government every purchase you make over the course of the year? I highly doubt that would actually fly, just for privacy concerns alone.
How about socking it to the retired rich by cutting their Social Security and Medicare benefits? I’m not seriously “rich,” but I do OK, and the government pays my wife (who never even paid Social Security taxes, if that matters) and me more in Social Security than I pay them in income taxes, which seems odd, since the money I get comes largely from people not nearly as well off as I am. (OK, I really do know the answer: Old people vote a lot.)
I don’t see how a consumption tax “socks it to the rich”. To illustrate, the burden of the late yacht tax fell disproportionately on the “poor” who lost the opportunity to build yachts. In general, it would seem that the rich would have an easier time consuming less than the poor who would no longer be able to provide consumables to the rich.
@Harold
One way to make a consumption tax progressive is to send out a check each month for, say, a few hundred dollars to everyone in the US. Then have a consumption tax of, say, twenty percent on everything.
The poor would end up getting money out of the deal, and as you moved up the income scale to people who spend more and more, you’d get closer to people paying the full twenty percent in taxes. It’s sort of like Milton Friedman’s negative income tax, but for consumption instead.
Steve Reilly – thanks. That sounds good. Has anybody worked out an example of how much you would need to send to everyone, and what rate of consumption tax would replicate current taxation levels if all income tax were abolished? Assuming that there were no change in behaviour for a first approximation.
@Steve
That certainly achieves a progressive consumption tax that targets the middle class and rich without taxing the poor. However, I don’t how see a system like that could target only the rich without turning into a massive welfare system.
I don’t think Romney is reading (reasonably!) progressive economists, nor that you will necessarily subscribe to their view, but giants of the theory of optimal taxation can argue for capital taxes, even with Atkinson-Stiglitz and the rest under our belt: http://elsa.berkeley.edu/~saez/diamond-saezJEP11opttax.pdf
It’s worth noting that Huntsman’s economic plan explicitly removes taxes on capital.
I think there are two problems with a consumption tax:
1. A consumption tax is not progressive. A person earning a low income probably spends 100% of his income, and thus all of his income is subject to the consumption tax. A person earning a middle income earns more, but spends a lower percent, saving some, and thus is taxes at a lower rate. And a person earning a high income (whether the source is labor or dividends) probably spends a low proportion of the income, reinvesting the balance.
2. Funding the Federal Government through a consumption tax would require a tax rate of somewhere between 23% and 30%. Add that to local sales taxes of 5% to 9%, and overall sales tax would be around 35%. At that level, evasion would be rampant unless the taxing authorities implement very strick enforcement.
The good thing about distributing government evenues across multiple sources is that each revenue stream imposes a small enough burden on the taxpayer to make evasion unproductive. I pay federal and state income tax, FICA, property tax, and sales tax. In aggregate, that’s a pretty hefty tax burden, but no one tax is large enough for me to be willing to chance getting caught evading it.
There are three groups/classes eligible for taxation:
1. Earned income,
2. Unearned income, and
3. Wealth.
Corporate income is not included because that is income that has not (yet!) been distributed. It is ALL unearned income to the owners/shareholders.
Earned income is taxed highest (top marginal rate) and is the widest in terms of number of people who pay it. In addition, Social Security taxes on incomes up to about $110k are added (matched by employer contributions).
Unearned income is the primary source of income for the wealthy and is used by corporations to reward management (stock options, etc) and minimize tax obligations by the wealthy. Capital gains tax rate is a fixed 15%, vs about 35% to 40% top rate for earned income. There is no Social Security tax owed or collected on capital gains.
Accumulated wealth is not taxed until after death–via the estate tax–which most do NOT pay as virtually all estates are below the minimum threshold for the estate tax ($3.5-million) and estate tax planning helps minimize those costs/taxes if the value of the estate is expected to be high enough to incur estate taxes.
Harold,
Google “Fair Tax” for one such plan. I’ve never read the book by Boortz and Linder, but presumably they work out the details. (And of course if you want the tax to be more or less progressive than they do, you could jiggle around the tax rate or the subsidy amount.)
There are some simple measure you could take to make even a simple consumption tax more progressive. Zero rating food, domestic energy and childrens clothes is a start.
It’s a feature, not a bug!
Al V.
1. A consumption tax is not progressive.
@ Al V.:
I don’t understand why a consumption tax is anymore inherently progressive than an income tax.
You can have an consumption tax system that says:
– The consumption tax on non-processed foods (rice, lentils, fruits, vegetables, butchered meat, etc.) are taxed at .0000001%
– Goods and services costing less than $ 20 are taxed at 1%
– Goods and services costing more than $ 1,000 are taxed at 20%
You could also have an environmentally unfriendly consumption tax policy:
– The consumption tax percentage is the same as a vehicle’s city mpg. E.g. a vehicle that gets 40 mpg city is taxed at 40%.
If the U.S. implements a VAT system you can be sure:
– It would evolve into a system with numerous levels of taxation, with some goods and services not being taxed at all.
– Conservatives would argue that we need a “flat” consumption tax on all goods and services for individuals and that businesses should be exempt for the VAT since they drive employment.
– Liberals would argue that the 30% VAT on non-processed foods unfairly hurts the poor.
Professor Landsburg:
Are the concepts of progressive vs. flat tax independent from the concept of an efficient vs. inefficient tax?
I often see these being tied together in discussion and not being an economist, I don’t know what to make of it.
E.g. Is a progressive income tax less efficient than a flat capital tax? Or is this a question that makes no sense?
If the answer is more than a paragraph, could you point me to a reference on it?
Steve: I realize you prefaced your first point with the claim that you only skimmed the proposal, but I really don’t see how you can argue that Romney’s proposal makes the system more progressive. Removing the capital gains tax for low income earners won’t make much of a difference as they don’t have that kind of income and near 200k a year is not middle class. http://www.urban.org/uploadedPDF/1000711_Tax_Fact_12-06-04.pdf is an analysis that implies that the median taxpayer doesn’t get much in the way of a reduced burden with those taxes eliminated.
By contrast Romney wants to eliminate the estate tax which only benefits the super wealthy and more than makes up for his capital gains proposal as does his proposal of lowering the corporate tax rate. In addition he talks about trying to lower marginal rates on the income tax. Since the income tax is one of the few progressive taxes most proposals to reduce rates give more benefit to the wealthy. Making regressive state and payroll taxes a larger share of total taxes is not a recipe for more progressive taxation.
Or labor income, for that matter.
Landsburg, do you prefer a (progressive) income tax to a (progressive) consumption tax?
What I can’t figure out is why folks, including Landsberg, use “rich” to refer both to persons of wealth and persons of high income.
I have never accumulated great wealth, but throughout my career as a contract engineer I have earned at an annual rate of roughly $160,000 to $200,000 (in today’s dollars). However, congenitally single and childfree, already in my 20’s I found that I would be taxed at a marginal rate of about 60% (Fed, State and FICA), starting after about six months of work. It didn’t take long for me to figure out that it made no sense for me to work for more than around 14 weeks per year average, especially when I found I could game the system by arranging to be laid off and get unemployment compensation while at the same time improving my own property, activity that would gain me tax-free income upon sale.
So, in spite of the fact that I have gained advanced degrees in physics and law (both under scholarship) in more than 26 years spent in classrooms, I could not (would not) contribute my skills to Amerika in return for a paltry 40% of their worth, opting instead to apply my talents to cabinetmaking, construction, plumbing and electrical installation in my own homes during my annual 38 weeks of “vacation.”
Now, I may not be the most talented engineer in Amerika, but it seems pretty dumb for a country to pay for 26 years of education and then incentivize the person to abandon his expensive skills. I wonder how many Zuckerbergs, Gates’s and Dells Amerika is missing because of its confiscatory tax policies.
In the end, I’m glad the gummint forced me to cut back work in my learned professions, since I have found work and great satisfaction in many years of employment and travel in Europe and South America, where I was able to gain fluency in five languages.
The irony is that now the CIA has courted me, but every year I see less and less in this country worthy of defending and more and more, like Kissinger’s Operation Condor, worthy of doing penance.
JImbino:
That is a choice YOU make–but you also HAVE the option. Others may not. You prefer to receive a lower income in exchange for more “free time” to do other things. What would happen if your employer said “no more” to that arrangement–and they hired someone else?
Others might (and do) choose to get paid for more of their time by working and thus having more cash for whatever they choose.
Jimbino:
What state and year is the basis of your 60% marginal rate calculation.
In 2000 (before the Bush Tax Cuts), a person with an income of $ 200,000 and no tax deductions would pay $ 60,051 of their salary in taxes.
In California (highest tax rate in nation), a person with a $ 200,000 income and no deductions would pay $ 16,480.
Being self employed, your FICA and medicare would have been about $ 15,000 (15% of first 100,000 of income).
So your tax burden would have been about 90k or 45%.
I think you either made an error in your tax calculation (unlikely for a genius like you) or you are not being honest.
Yo Jerry,
You seem to miss the point that I shed no tears for myself; the point is that Amerika is the big loser, incentivizing well educated folks to slack off or emigrate.
Hell, my skills in nuclear physics alone could (still) settle me well in many places, including Brazil, where I do have a home.
And Will A, I think you confuse “average” with “marginal”:
1. Marginal federal income tax rate: 70% then, 40% now.
2. Marginal state income tax rate: 6% to 7%
3. FICA tax rate (high phase-out for singles): 15.3%
4. Unemployment compensation rate: 3%
5. Workers compensation rate: 2%
Now, normal folks are confused in thinking that the employer pays half the FICA and all the unemployment and workers’ compensation insurance, but economists know different. So do I, as I have compared and bid on the same job on a W-2 and on a 1099 basis.
Result: a single earning $100,000 now will pay over 60% on the margin. A single plumber who does $40 worth of work for me charges $100 before taxes. To pay him, I have to earn $250, so I do the plumbing myself, and I can drink beer while working, which I can’t in my “high-tech” cubicle.
Jimbino:
The Marginal Rate only applies to income above a certain level. Currently the marginal rate is only applies on income above $ 379,150.
So if you make $ 379,160, only $ 10 is taxed at the marginal rate. So again, what year were you using to come up with your 60% total taxation on $ 200,000 in today’s dollars.
Also, a person making $100,000 will pay a higher percentage because FICA is capped just above $ 100,000. This would not apply to you since you were making $ 200,000.
Here is my argument for deporting all illegal immigrants:
Each illegal immigrant costs us $ 2,500,123 on average per month. We simply can’t afford it.
“Result: a single earning $100,000 now will pay over 60% on the margin.”
Bull. The only way you can get close to 60% is if you include the estate tax (which almost nobody pays since the exclusion is so high).
Steve,
You had one blog post (I tried to find it by searching for ‘capital’ but couldn’t) where you claim to have proven (with a diagram) that poor and rich alike should prefer zero tax on capital gains.
You assumed that a person would modify their present and future consumption to maximise utility, and the govt would take that into account, and choose one of the many tax regimes that gave a certain fixed present value of govt revenue. Then you showed that the taxpayer would prefer zero tax on capital gains.
There is a fallacy in the argument. You assumed there was only one taxpayer (or that all taxpayers had identical utility functions).
If you throw in more than one taxpayer and give them different utility functions, it is easy to construct examples where some (those who prefer immediate consumption) prefer nonzero taxes on capital gains in exchange for lower tax rates on income.
Therefore, it is not true that all people always prefer zero tax on capital gains.
Nor is it true that zero tax on capital gains is always preferred by a majority of the population. Nor that total utility is always maximised by having zero tax on capital gains. This all comes out of the maths, it’s not just waffle.
The only way you can still insist that having zero tax on capital gains is, I suspect, to add to the model the possibility that individuals could bid to have the tax regime changed. Then, I suspect, economic efficiency would demand that the “saver” should pay a bit extra to the “spender” for the sake of enjoying a 0% tax on capital gains.
I’m not sure if you want to advocate that votes should be put up for auction though.
I haven’t had a chance to write up the maths properly yet.
Jimbino:
You are simply overpaid. But, that is the employer’s choice. I am familiar with Brazil and am well-acquainted with physics and physicists. Worked for a company that employed a LOT of PhD physicists to develop products. They called themselves a “nano tech” company, but it was more fancy tools than true nano-tech (IMO). Just because you are pushing atoms/molecules vs larger masses does not make it “nano”. I fixed the problems they had created and was done. Their equipment used radioative materials as a radiation source (as well as lasers) operating in UHV. Plus they had internally-developed proprietary/unique processes that were being abandoned in favor of copying everyone else. The reason? They would have been required to invest in the business to keep those processes going. Instead, they chose to not make the investment (which substantially reduced costs in the long term compared to alternative outsourcing) and chose to contract out the alternative processes–which added substantially to the total cost of future products.
Another omission is the deduction of *costs* that are permitted under tax law. Thus, any required classes (for example) to maintain your job are tax-deductible at your top-marginal rate. So, you get something for nothing–and that is not reflected in your analysis.
@Will A “The consumption tax percentage is the same as a vehicle’s city mpg. E.g. a vehicle that gets 40 mpg city is taxed at 40%”
Can I get an old lemon that will never move again, plant corn in the back seat, a diesel distiller in the front seat, claim it gets negative infinity mpg, sell it to my wife and both of us retire rich?
I think that the consumption/VAT tax is accepted as being best for the economy. It catches Wages, Profits, Inheritance… when they are spent for consumables rather than re-invested in the economy.
As a Federal government we currently spend about 20% of the GDP and as I want to also spend for health adjusted medical insurance vouchers and rebate LOTS to make the taxes regressive, I think that we would be targeting to $1,000/month citizenship rebates to every American to provide basic living costs, and a Health Adjusted voucher to allow the purchase of basic medical insurance, I assume the sales/VAT would be in the range of about 40%.
In return for this most, or all, other federal taxes and entitlements would be eliminated.
I realize that even Bill Gates would get the $1,000/month, but I effectively he would be paying for it by a slightly higher tax percentage and the advantage is that paying all Americans is simpler, ensures that there are not incentive glitches in the system, and that we are all treated equally and share in the system.
Obviously the exact numbers have to be worked out and there may be some tuning to ensure that taxes are payed, but as a starter, does this make sense.
Of course the real value of the VAT, is the business reclaim system, (where businesses pay tax on what they sell, but reclaim the tax on what they buy).
This makes businesses into tax collectors for the government constantly checking up on each other…
Bottom line is VAT ends up getting levvied equally on labour, capital and imports.
László Sándor: Than you for the link to the paper. I had a quick look at the capital taxation part. The authors conclude that the key papers are not “policy relevant”. I think it means that you can’t just take the paper’s conclusions that zero rate capital tax is optimum and stick it into a policy.
There is a questioning of the assumptions underlying the models that conclude zero rate optimum – specifically rationality:
“Thus, the strong asymptotic zero tax result of Chamley and Judd requires that rational intertemporal decision making not only holds for entire lifetimes, but extends across dynasties. Both assumptions have been heavily challenged in the empirical literature.”
Apparently, Atkinson Stiglitz assumes a certain homogeneity among people which is also questionable – including that savings propensity does not vary with worker skill level. If more skilled workers tend to save a higher proprtion of income, then a zero capital tax rate is not optimum.
The inability to distinguish labor from capital income is also a major factor, as has been discussed above.
These are just the points I happened to grasp – there are others.
This is a CESifo working paper – I think not not published in a peer reviewed journal. However, the authors appear to know that of which they speak.
They do seem to have some quite strong arguments for a non-zero rate of capital taxation, but I am not familiar at all with the subject, so should not really comment.
@ Mike H:
I think you misunderstood the point I was making. I was coming up with an example of an environmentally unfriendly policy.
A Hummer 10 mpg would have a VAT of 10%, but a Prius would have a VAT of 40%.
I don’t see how mileage would ever be negative (going in reverse is still moving in some direction) so car that doesn’t run on gas would have a mileage of infinity and therefore the VAT would be infinity.
The advantage of this tax would that it would make the job killing Nissan Leaf unsaleable.
Why is the Leaf job killing? I suppose the subsidies it attracts can be accused of that, but not the car itself. I think in the US you can get a maximum of $7500 tax credit on an all electric vehicle.
Harold:
Anything considered liberal or that liberals like is job killing. European countries have job killing nationalized health programs. The U.S. has job killing environmentalists.
The Leaf only gets 100 miles per charge. What good is that? And the subsidy is just another way of the government dictating what you can and can’t do.
The government wants to keep me in a 100 mile radius of my home to help some spotted wooded sap sucking flea when I could be pumping money into the economy buying Gas, Pepsi, Cheetos, and Snickers at my locally run and maintained gas station.
You just watch, once everyone is in their 100 mile radius limited cars, that’s when the government will strip us of our liberties because it will be impossible for us to mobilize as a nation against the new world order.
So sorry I should have said job killing liberty destroying Nissan Leaf.
Laszlo thanks for the link to the Diamond-Saez paper. I certainly hope, to paraphrase Harold, that one wouldn’t just take its conclusions – e.g. that the optimal top marginal rate is between 48 and 76 percent – and stick it into policy either. Interesting quote from the paper: “The optimal top tax rate… does not depend on the total revenue needs of the government.”
Here’s the Landsburg post someone was looking fors:
http://www.thebigquestions.com/2010/09/14/getting-it-right/
It’s about the double taxation issue, making the point that when investments are made from income that was already taxed as compensation, any future stream of returns is lower than it otherwise would have been by the same rate…so any capital gains tax is a “surtax”. David Friedman’s blog comments on Landsburg’s post (you have to scroll down a few), arguing that some investments might involve labor (but many of us invest pretty passively)…but even in that case he says only the real (inflation-adjusted) return should be taxed as income. (So should it be considered a positive return on your labor if you don’t beat the index?) Martin Feldstein has written for decades about the perverse combined impact of inflation and taxes on real investment returns. I had assumed that all economists recognized the disincentive to saving as a negative aspect of taxing capital (possibly offset by equity considerations), however Diamond-Saez in fact suggest that some people save more than an “optimal” level, e.g. “discouraging savings enhances the ability to provide insurance against future poor labor market possibilities” (one might’ve thought that’s what people who save are trying to do for themselves). Their main argument for taxing capital does seem to be about people shifting income (to avoid taxes in the first instance), but to me that seems like an accounting issue that shouldn’t necessarily drive policy. It should also be recognized that some of this “shifting”, like exchanging salary for options, does involve a change in substance (e.g. risk) and not just form, and therefore perhaps should be treated differently. Some like Larry Lindsey have suggested that a cash-based system (e.g. if/when options are actually exercised) is better: “cash is a fact, income is an opinion”.
The optimum taxation approach includes an equity as well as an efficiency part that may not be universally accepted -is this correct?. However, it seems to me that there are enough problems with the separation of capital and labor income and other assumptions that a non zero rate of capital income tax is desirable. Steve has said before that one may have other reasons for taxing capital, but the common argument that it is otherwise untaxed is false. This is correct, but here we have the other reasons.
The Diamond Saez paper claims:
“Academic arguments against capital income taxation typically draw on one or both of two theoretical analyses: (1) the theorem that the optimum has no asymptotic long-run taxation of capital income in Chamley (1986) and Judd (1985); and (2) the theorem that the optimum has no taxation of capital income in Atkinson and Stiglitz
(1976).14”
Footnote 14 is unfortunately all Greek to me, as is the above passage, but is it essesntially true?. I have assumed that Steves arguments for zero capital taxation is more or less because of the above, but I am not quite sure.
Harold: Yes, my argument is essentially the Chamley/Judd argument.
Iceman: thanks for this well researched and thoughtful comment.
I think the post someone was looking for is this one, about the least painful tax. It shows that even if the overall tax rate is the same, some taxes are prefered over others.
http://www.thebigquestions.com/2010/09/23/the-least-painful-tax/
The following may be useful to some re: the regressivity (or not) of consumption taxes. Readers will likely find the citation of Metcalf therein useful as well. It argues that over one’s lifetime, cons. taxes are pretty flat. btw–the question is studied in a model in which the capital income tax is not guaranteed to be zero by the usual Chamley/Judd logic.
http://www.richmondfed.org/publications/research/economic_quarterly/2009/winter/pdf/athreya.pdf