Scott Sumner  

Unpersuaded

PRINT
Why Not Capitalism?, In... Jonathan Turley Echoes Milton ...

I recently completed reading Thomas Piketty's new book entitled Capital in the Twenty-First Century. Piketty explains why the distribution of capital is becoming increasingly unequal, why we need higher tax rates on upper income individuals, and also a wealth tax on the affluent.

You probably won't be surprised to hear that I was not persuaded by his arguments. I didn't expect to be persuaded. But here's what did surprise me; the book made no real attempt to persuade me. This claim requires some explanation.

Like monetary economics, the field of public finance implies that the world is full of cognitive illusions. For example, many public finance theorists believe:

1. Economic inequality should be measured by consumption inequality, not income or wealth inequality.

2. It makes no sense to add capital and labor income. Talking about that sort of composite is like discussing the number of "fruits," when the basket includes blueberries and watermelons.

3. Relative to a progressive consumption tax, a progressive income tax is both unfair and inefficient.

4. In the long run a payroll tax is equivalent to a consumption tax.

5. Misers are more altruistic than big spenders.

My views on public finance are pretty mainstream, or at least pretty close to where the mainstream was in 2007. Like many center-left economists, I favor a progressive consumption tax, and believe that taxes on capital should be zero. Piketty clearly disagrees with this, which is not surprising as there are many other economists who also disagree with my interpretation of the public finance literature. There are good (second order) arguments against my policy views. But Piketty doesn't make them.

Going into the book I expected Piketty to try to persuade people like me that income and capital are the right variables to consider, not consumption. But he didn't do so; indeed he didn't even make an attempt to do so. The book is aimed at thoughtful non-specialists who don't know about all the cognitive illusions in the public finance literature. People who think it "obvious" that a fair and efficient tax would treat wage and capital income equally, even though a tax on interest income effectively taxes the same wage income twice. Taxes on capital income impose higher lifetime tax rates on thrifty people, as compared to less thrifty people with exactly the same lifetime resources to work with.

I didn't notice any discussion of a progressive consumption tax. The only discussion I recall of the sort of consumption tax that could easily be made progressive was the following (page 494):

In the abstract, one might imagine a direct tax on consumption, which would depend on each taxpayer's total consumption, but no such tax has ever existed.
This is an awfully dismissive remark on the tax system that many public finance experts regarded as ideal. Keep in mind that Piketty favors a global tax on wealth, but no such global tax regime has ever existed. More importantly, payroll taxes have existed, and are economically equivalent to a consumption tax, in the long run. (A more exact equivalence occurs if they are accompanied by a one-time tax on wealth at the point of implementation.)

You might wonder why this is important, given that we have other forms of consumption tax such as VATs. But the VAT is difficult to make progressive, whereas it is easy to make a payroll tax progressive. I'm left with no real understanding of why Piketty believes that taxes on capital are superior to progressive consumption taxes. Again, there are certainly arguments in favor of taxing capital instead of consumption, but he doesn't make them.

Piketty starts the book with some sweeping generalizations, such as the claim that r > g is a central contradiction of capitalism, or that only marketable wealth counts as "wealth." It is acceptable to simplify things at the beginning, as long as you return to these issues in a more realistic way later on the book. Unfortunately he doesn't do so. He never really justifies the claim about the implications of r > g, and he repeatedly uses the term 'wealth' to refer to marketable wealth, when he actually should be talking about total wealth, including the present value of future government benefits like Social Security and Medicaid, as well as human capital. To be sure, there are some purposes for which singling out marketable wealth might be appropriate, but discussing changes in economic inequality over time is not one of those purposes.

Here's another example (discussing a wealth tax):

This would replace the property tax, which in most countries is tantamount to a wealth tax on the propertied middle-class. The new system would be both more just and more efficient, because it targets all assets (not only real estate) and relies on transparent data and market values net of mortgage debt.
This seems completely wrong. A tax on residential real estate is a good way of taxing consumption. It can also be made progressive. It should not exclude mortgage debt, as people derive the same consumption from a house that is owned free and clear as they do from one on which there is a mortgage. On the other hand, wealth taxes should not apply to capital goods, for all the usual public finance reasons. Indeed even the more left-leaning countries of Western Europe exclude capital goods from VATs.

Let's suppose that two identical twins each make $100,000/year for 40 years. One spends the money right away, and the other saves half the money, pushing consumption far into his future, and to some extent his children's future. In both cases the present value of consumption is identical. Resources are identical. It's a simple public finance example involving two freely made choices. (The concept of "choice" rarely appears in Piketty's book.) So what does Piketty think of this example. I'm not quite sure, but here's a typical comment (p. 423):

To be sure, there is something astonishing about the notion that capital yields rent, or income that the owner of capital obtains without working. There is something in this notion that is an affront to common sense and that has in fact perturbed any number of civilizations . . .
It's not clear to me that this "rent" is actually as unearned as the Piketty seems to assume. People who live near airports are "rewarded" with cheaper home prices---so why shouldn't people who defer consumption into the future be "rewarded" with cheaper future consumption?

I find it difficult to provide an overall evaluation of this book, because I found things on nearly every page that annoyed me. (I'll cover a few of these in additional posts, some at TheMoneyIllusion.) On the other hand, I am obviously not the intended audience for this book. And if I look beyond my annoyance, I can understand why many readers found the book to be impressive, even a tour de force. If my book on the Great Depression ever comes out, I'd obviously rather someone say "it's an impressive work of scholarship, although I didn't buy the central argument about the importance of the global gold market," rather than "it's 600 pages of drivel." Piketty's book is impressive in some ways. I like his approach to methodology. He might well be correct about some of his predictions. As we saw with the General Theory, a book can contain many individual arguments that don't hold up, and still be a milestone in the intellectual debate.


Comments and Sharing





COMMENTS (27 to date)
GabbyD writes:

"1. Economic inequality should be measured by consumption inequality, not income or wealth inequality."

why not wealth?, esp in a more long-run discussion on inequality's effects, don't consumption and wealth match up in the longrun?

in the short run, YES, consumption and wealth need to track closely... but thinking about the longrun, wealth dynamics are important to inequality, yes?

TravisV writes:

Yglesias:

Greg Mankiw unpersuasive on Piketty and inherited wealth

http://www.vox.com/2014/6/23/5834034/mankiw-unpersuasive-on-inherited-wealth

Garrett writes:

It's not clear to me that this "rent" is actually as unearned as the Piketty seems to assume. People who live near airports are "rewarded" with cheaper home prices---so why shouldn't people who defer consumption into the future be "rewarded" with cheaper future consumption?

Another reason why capital (investment assets) should produce rents (income; returns above the risk-free rate) is that the asset-holder is assuming the risk of loss. Take someone who buys a house and immediately rents it out. If jobs stagnate in the local area, the owner will be impacted (asset value will decline) but the renter can just move away. It makes sense then for the renter to compensate the owner for use of the house without assuming financial risk.

Jeff writes:

I was going to say the same thing as Garrett. No, the owners of capital don't "work" in the same sense that a gardner works for an hourly wage (although evaluating different investment opportunities and managing a portfolio of investments certainly qualifies as work), but that doesn't mean the owners' income is unearned. He/she is bearing risk. Often significant risk. If your quote really is typical, the only thing "astonishing" about this is that Piketty fails to consider it.

Michael Byrnes writes:

Responding to your five points on public finance:

"1. Economic inequality should be measured by consumption inequality, not income or wealth inequality."

Agree... more generally I think your arguments about the significance of "income inequality" are convincing, even to this liberal Democrat.

"2. It makes no sense to add capital and labor income. Talking about that sort of composite is like discussing the number of "fruits," when the basket includes blueberries and watermelons."

We do, however, live in a world where this distinction is often blurred. I know that you have argued that various loopholes that allow people to pretend that their labor income is actually capital income should be closed, but in practice I think this will turn out to be easier said than done.

"3. Relative to a progressive consumption tax, a progressive income tax is both unfair and inefficient."

No argument.

"4. In the long run a payroll tax is equivalent to a consumption tax."

These don't seem equivalent, even if they really are in the long run. One seems to penalize labor. (If I come by my money through some means other than having a job, then my consumption tax rate is 0% under a payroll tax, but with a VAT I pay the same tax rate as everyone else.)

"5. Misers are more altruistic than big spenders."

Only if they don't use any of their miserly wealth to influence the political process. Is there a rich, politically active constituency that wants, say, a strong dollar and higher interest rates?

Steve Sailer writes:

My general prejudice having watched people get increasingly good at gaming systems over my lifetime is that we should have a lot of small taxes rather than one or two giant ones. That keeps clever individuals from wholly sidestepping their share of the tax burden.

Duncan Earley writes:

I don't understand how a payroll tax is like a tax on consumption. It looks like an income tax to me. Can someone help me out?

Scott Sumner writes:

Gabby, Actual wealth should be the present value of future consumption. But of course Piketty's book isn't about wealth, it's about marketable wealth.

Travis, That post doesn't actually address Mankiw's points.

Garrett and Jeff, I agree.

Michael, In the long run, all income is either wage income, or capital income that indirectly derives from wage income.

A 10% consumption tax is equivalent to a 10% tax on labor.

Steve, I agree.

Duncan, A payroll tax excludes income from capital, which an income tax includes.

GabbyD writes:

"Gabby, Actual wealth should be the present value of future consumption. But of course Piketty's book isn't about wealth, it's about marketable wealth."

yes, wealth is PV value of future consumption. this is what i was trying to get at about the "long run"...

what is unmarketable wealth? isnt wealth convertible in a "market" for consumption, investment goods, etc?

BC writes:

It's astonishing to me that Piketty's astonishment that capital yields "rent" and that he finds this to be an "affront to common sense" has not been noted by many, if any, other reviewers.

To Scott's point, it's not even clear to me that deferred consumption is "cheaper" than immediate consumption. A lender exchanges immediate consumption for future consumption with a borrower at some agreed upon exchange rate called the interest rate. By definition, the present value of both consumption baskets are equal, so in what sense is one party gaining "cheaper" consumption? The exchange is no different from two parties exchanging consumption in Canada for consumption in the US at the USD-CAD exchange rate. Whether the interest rate exceeds some calculated inflation rate is no more relevant than whether the USD-CAD exchange rate equals some calculated purchasing-power-parity rate. The interest rate and foreign exchange rate are the market-determined "fair" relative prices of consumption at different times and in different countries, respectively. Both of these cases are like exchanging apples for oranges of equal market value.

AbsoluteZero writes:

As Garrett and Jeff said, the owners of capital bear risk, and gathering and analyzing information, and so on, is work. But it's more than that. Often there's indirect work. Consider a person who's in a position to know certain things, and this information makes it easy for him to make money. Many would say it's just because he's in that position. Anybody in that position can do the same. They don't consider all the work the person did to get to that position. Another way to look at it is sacrifice, just like in Scott's example of people who buy houses near airports.

And this is what I find strange about people who hold views exemplified by that Piketty quote. Consider a person who's highly skilled in doing something. Because of this he's paid a lot. People don't look at that and say, all he does is something very easy (for him), yet he's paid so much. They assume it's not easy to acquire that level of skill. The person must have worked very hard and sacrificed a lot to get there. But if they see a person with money, it's as if they assume the person just magically have that money. Unless it's inherited, he probably worked hard and sacrificed a lot to have what he has today, so why shouldn't his capital generate profit for him now. To me it's the same.

BC writes:

GabbyD, I believe that "marketable wealth" does not include human capital (and also present value of future government benefits like Social Security and Medicaid).

Consider two people of equal human capital, one a spendthrift that consumes every paycheck immediately and one a saver that saves part of each paycheck. Pre-tax, each will have the same lifetime consumption. The spendthrift will never accumulate any "marketable wealth": his bank account balance will always be zero. The saver, however, will accumulate savings and, thus, will be subject to Piketty's wealth taxes. That's why wealth inequality is not economic inequality.

(Marketable) wealth results from giving up past consumption. Regarding Michael Byrnes's point about misers using "wealth to influence the political process", what about big spenders' political influence in the past, when they were outspending the misers?

Shane L writes:

Economist Ronan Lyons weighs in interestingly too:
http://www.irisheconomy.ie/index.php/2014/06/18/thomas-piketty-and-the-subsidy-of-leverage/

Piketty recommended that property taxes should be lower for those with unpaid mortgages:

"I think if you have a house that’s worth €400,000 but you have a mortgage of €390,000, you know you’re not really rich. Your net wealth is €10,000 and you are paying back in interest payments as much as a tenant will pay in rent. So there’s no reason why you should pay as much property tax as someone who inherited his €400,000 house or who has finished reimbursing his mortgage 20 years ago."

Lyons argued:

"For example, in the case of Ireland, there are (very roughly) a third of households owning without a mortgage (the richest third), another third with mortgages and the final third living in rented accommodation (by and large the poorest third).

"If you give a tax rebate to the middle group - who, remember, have wealth that the poorest third do not have - then by definition the other two groups have to pay more in tax to compensate (assuming that there is some fixed target for government revenues)...

"Taxing the value of property (or ideally the value of land) is simple. Introducing tax rebates for debt - however well-intentioned - turns it into a game where everyone wants to minimise their tax liability (in this case by increasing their debt liabilities)."

Michael Byrnes writes:

Steve Sailer wrote:

"My general prejudice having watched people get increasingly good at gaming systems over my lifetime is that we should have a lot of small taxes rather than one or two giant ones. That keeps clever individuals from wholly sidestepping their share of the tax burden."

I think this is exactly right and can be extended beyond the realm of taxes. People are going to find a way to game any system or no system.

J.V. Dubois writes:

I wanted to write a little bit critical comment on payroll tax being equivalent to consumption tax but I will at least postpone it here. Because the debate truly is about something else.

For some reason Piketty wants global tax on "wealthy people". The key point is that his definition of wealth is very narrow. I recently read that for anybody to belong to global 1% you just need to have an income of $34,000 a year. I just wonder how many people would defend a special wealth tax for the whole developed world in the name of true social justice.

For some imaginary fellow in a poor country with official income of $1 and who has to earn food by barter and self-production these discussion about how to rearrange wealth among global top 1% have to sound outrageous. For her most people born in developed world are already lucky heirs to incredible riches that are virtually unattainable to her.

I Yam who I Yam writes:

[Comment removed for supplying false email address. Email the webmaster@econlib.org to request restoring your comment. You can be whoever you want to be, but you still have to provide a valid email address to post comments on EconLog and EconTalk.--Econlib Ed.]

ThomasH writes:

Granted my first hand knowledge of what public finance theory was mainstream dates from ~1963, it did not include any view that "economic" inequality WAS consumption inequality rather than income inequality. Both seem like equally valid concepts to describe the way the world is.

Yes, a labor-income only tax is the same as a total income minus saving tax in the long run, but the former suggests more ways to disguise labor income as non-labor income than the later suggest ways to disguise consumption as saving.

Should anyone who favors a progressive consumption tax feel especially disappointed that one more proposal for collecting more tax revenue from people who receive high incomes/have high consumption targets their incomes rather than their consumption as the way to do it?

Politically I think a route to consumption taxation would be to tax all income (included imputed business income) at the same rates and then when people point out that they are being "double taxed" fix it by abolishing the corporate income tax and making savings deductible.

Michael Moran writes:

My comment is not exactly on your post, which in general I agree with. My comment is as a tax lawyer, who is in the trenches of the tax system. The tax system cannot distinguish between a tax on capital and a tax on income. Most real income is a combination of both.

And a tax on wealth is a joke. Cannot be enforced. Compare resources IRS (when they were concerned about enforcing tax system) employed on estate tax, a one time wealth tax, with income tax, and compare about of money raised by each.

And a progressive consumption tax. Do incentives matter. At what point does taking money from those that work and giving it to those who do not slow growth (wait, Europe, where work is a lifestyle choice discouraged by the government through high taxes). Man, they have been burning it up with growth.

mico writes:

"I like his approach to methodology. He might well be correct about some of his predictions. As we saw with the General Theory, a book can contain many individual arguments that don't hold up, and still be a milestone in the intellectual debate."

Honest question - do you think this book is going to be a milestone in any intellectual debate?

It might be a milestone in a public debate, but it sounds like this book doesn't even address most of the intellectual issues, rather becoming a best seller for packaging stale left wing policy conclusions for "thoughtful non-specialists who don't know... the public finance literature".

I want to know because I am still wondering whether to buy this book. It seems to be becoming important just because others think it is, but I'm not interested jumping aboard a bandwagon - and indirectly financing a propaganda machine for a cause that goes against my conscience - if that is all it is.

Ray Lopez writes:

Essentially Sumner is complaining that Piketty did not use his pet arguments. Lame. This is telling: "He never really justifies the claim about the implications of r > g, and he repeatedly uses the term 'wealth' to refer to marketable wealth, when he actually should be talking about total wealth, including the present value of future government benefits like Social Security and Medicaid, as well as human capital. To be sure, there are some purposes for which singling out marketable wealth might be appropriate, but discussing changes in economic inequality over time is not one of those purposes"

(1) Perhaps marketable wealth is an illusion, as Piketty understands Social Security can be changed but actual wealth cannot?

(2) "Mystery Writing" is improperly used by Sumner, a technique where you imply some sort of mysterious factors exist but don't mention them (Sumner's last sentence above).

Jordan writes:

[Comment removed pending confirmation of email address. Email the webmaster@econlib.org to request restoring this comment. A valid email address is required to post comments on EconLog and EconTalk.--Econlib Ed.]

Scott Sumner writes:

ThomasH, You asked:

"Should anyone who favors a progressive consumption tax feel especially disappointed that one more proposal for collecting more tax revenue from people who receive high incomes/have high consumption targets their incomes rather than their consumption as the way to do it?"

Yes, because a progressive consumption tax is good public policy and a progressive income tax is not.

Ray, Social Security can be changed but actual wealth cannot? You'll have to explain that one to me.

mico, I was a bit disappointed that reading the book didn't add much to what I gathered from reading the many reviews. I expected stronger defenses of the r > g issue, and also something on why capital taxation is superior to progressive consumption taxes.

Jordan, It's a healthy mix of history, data and basic theoretical principles. I think he has the wrong data and the wrong principles, but that's another issue.

Luke Lea writes:

Speaking of progressive consumption taxes, I've long been enamored of the idea of a graduated expenditure tax as described in Kaldor's book on that subject. Recently I had a happy thought for how to construct a one parameter version of such a tax, by which I mean that instead of defining marginal tax rates and the associated tax brackets to which they apply by a set of arbitrary numbers, one would adopt a single adjustable paramater, call it d, which together with the unit of currency (in our case the dollar) is sufficient to define the tax.

How so? Well, the parameter d is just the parameter used to construct an arithemtic sequence. In this case it would be a very small number on the order of ten to the minus eight. The first dollar spent is taxed at a marginal rate of d, the second at a marginal rate of 2d, etc. The total tax owed on any amount of spending is then easily calculated using the formula we all learned in high school to sum an arithmetic sequence.

I believe I have got that right: a way to raise a target amount of revenue that is transparent, straightforward, fairly simple, and not easily monkeyed around with. Graph it out and you will see it has a very nice shape.

TaxHistorian writes:

There are plenty of good arguments for a progressive consumption tax, but this notion that taxing capital income is taxing "the same wage income twice" is nonsense and needs to be dropped.

Only the gain is taxed. That is "new" income, not the wage income that was saved. In fact, capital gains, in that sense, are essentially the only wage income in our current system not "taxed twice."

For example, the property tax (which is strangely defended in this post) actually taxes an unrealized capital good (both the gain and the, in effect, principal) every single year on the basis of a value affixed by a bureaucrat. And, of course, the property tax is also paid out of wage income (effectively taxing that income twice). As an aside, once could also argue that the property tax represents a steep sales tax on homes. If one stays in one's home for decades, one could easily pay a 50 or 100 percent sales tax on the home, in effect. And, speaking of sales taxes, that's another clear example of the same already taxed wage income being taxed "twice."

TaxHistorian writes:

To expand a little more, I think the larger issue with differentiating between labor income and capital income, or even consumption and saving/investing, is that both dichotomies are ultimately arbitrary lines drawn by government policy. This is where the capital/labor income division fails and where the consumption/saving division (even though better in some ways than the former) would fail, too.

Let's use the now omnipresent "human capital" concept as an example. Let's say I'm earning 25k with my HS diploma. Then let's say I invest in my human capital by going to college. Then I graduate and immediately begin making 50k. Why shouldn't that additional 25k be taxed at lower capital gains rates? It's ultimately the same philosophical principle, thinking broadly, as our current definition of a capital gain.

To turn to consumption versus saving/investing, there are a few issues. How is purchasing a stock different than purchasing a computer? The common argument is that the former is deferring utility and/or pleasure for later, why the latter is taking them now. But who's to say that one doesn't gain pleasure or utility from purchasing and owning stocks? I certainly know plenty of people who get more pleasure out of playing the market than buying tech equipment. There's also the argument that buying the stock somehow improves future growth and the economic well being of society in a way that buying a computer now does not. But who's to say that such an argument is true in all circumstances? More pointedly, what if the person buying the computer uses it to learn to code, gaining human capital and perhaps inventing some app we all come to use and love? How is that purchase of the computer (subject to sales tax) not simply a capital investment for the individual who buys it, someone who will -- like the person who goes to college and gains human capital and earning power -- be paying income tax on their "gain," unlike the person who bought the stock and will pay preferential rates on their gain.

Ultimately, no matter how you slice it, the only philosophically defensible form of taxation is income taxation that treats all income neutrally. Any differentiation between types of income and/or expenditures is simply allowing the government to arbitrarily pick winners and losers. (Corporate income proves to be the most difficult issue, but that's an even longer discussion.)

LD Bottorff writes:

Should I spend my time reading a book by an economist who doesn't understand why capital must be rewarded?

Floccina writes:

Many people are for higher taxes on people wealthier than themselves, what is Piketty's position on raising his own taxes. Surely he is among the 1% world wide. I wonder how much money he gives away to not relatives.

Comments for this entry have been closed
Return to top