Steve Landsburg 2, English Language 0
I want to give a disclaimer that usually when I think University of Rochester economist Steve Landsburg is wrong on something, it’s not that his conclusion is mistaken, or even that the world is a worse place for him having advanced the argument. No, it’s more like, I think Steve tries to deliberately shock his readers by putting his viewpoint into a package coated with shards of broken glass. Yeah, it gets your attention all right, but it can be dangerous.
A recent example of this occurred when Steve–who was making a very important point in the “does the national debt burden our grandkids?” debate–decided to throw me off the trail by declaring, in spite of the English language, that whether U.S. Treasury securities were owned by American citizens or the Chinese government, in either case “we owe it to ourselves.”
Today we will go over another example of Steve being too cute by half. In discussing Mitt Romney’s personal taxes, Steve writes:
Mitt Romney says his tax rate is “probably around 15%”…[T]he New York Times is quick to point out that he’s a beneficiary of the “fact” that investment income is taxed at a much lower rate than wages and salaries, leaving him with a lower percentage tax burden than the working-stiffs he employs.
For at least the eighth time on this blog, I want to point out that this widely believed “fact” is not true.
To understand Mitt Romney’s tax burden, you have to compare him to his doppelganger Timm Romney, who lives on a planet with no taxes. In the year (say) 2000, Mitt and Timm both earned (say) a million dollars. Timm invested his million dollars, saw it double over the past decade or so, and cashed out his investment this year, leaving him with two million dollars. Mitt, by contrast, paid 35% tax in 2000, leaving him with $650,000. He invested it, saw it double, and cashed out last year, paying 15% tax on the $650,000 capital gain. That leaves him $1,202,500, which is about 60% of what Timm’s got. In other words, the tax system costs Mitt almost 40% of his income.
By contrast, people on our planet without investment income collect their wages, pay 35% in taxes, and spend what’s left. The tax system costs them 35%, while it costs Mitt almost 40%. In other words, people with investment income bear a higher tax burden, as a percentage of their income, than anyone else — and that’s before you even start accounting for the taxes on dividends, interest, corporate income and inheritance.
I ultimately agree with Steve’s conclusion on all this stuff–namely that it’s ridiculous that the government levies any taxes on capital gains at all–but I have to conclude that his above discussion is simply wrong. Rather than demonstrating that “people with investment income bear a higher tax burden, as a percentage of their income, than anyone else,” what Steve has really shown is that if you add two different taxes on two different types of income together, the burden is higher than if you just look at the first tax. Yes, I will agree that (a+b)>a, where a,b>0. (That’s for Steve’s mathematical sensibilities.) But I didn’t just demonstrate that b>a.
In Steve’s example, Mitt first earns $1 million in wage income, which is taxed at 35%. Then, of the post-tax $650,000, he invests it and earns $650,000 in interest, dividends, or capital gains or however you want to define it. (Steve doesn’t actually specify.) That $650,000 of investment income is then taxed at 15%. Mitt ends up with $1.2 million, compared to the full $2 million that his doppelganger Timm enjoys.
So yes, Mitt suffers about a 40% effective average tax rate, compared to his no-tax doppelganger Timm. But how in the world does Steve blame this 40% burden on the tax on dividends? No, what Steve has shown is that the wage tax and the dividend tax act together to yield an average rate of 40%. Well, right, but I don’t think anybody doubted that. By the same token, if Mitt and Timm spent all of their wealth ($1.2m and $2m respectively) on gasoline, we’d end up concluding that motorists pay more in taxes than either wage earners or hedge fund managers. Would that make any sense at all?
This isn’t a little nit, this is the whole issue. Forget taxes for a minute. Suppose an intro to accounting professor asks his kids this question on a test:
Over a certain period, Mitt earns $1 million in salary and $1 million in interest. How much of his income is due to his investment earnings?
(A) 100%
(B) 50%
(C) 0%
(D) It depends on the point Steve Landsburg is trying to make with this example.
If I still haven’t convinced you–maybe you’re getting hung up on the “double taxation” thing–then I’ll go the other way: That clown Landsburg isn’t going far enough. Sure, in order for you to have any income to buy investments with in the first place, you need to have earlier earned some labor income. But Landsburg forgot that workers don’t operate on solar power–it takes food to keep them going. So in order for Mitt to have even had the ability to earn that $1 million salary in the first place, in the previous period he had to earn $50,000 in salary, out of which the government took 35%. Then Mitt had to buy groceries and pay his mortgage with it, subject to an 8% sales tax (depending on his state of residence) and property taxes. I mean, you can’t be a worker if you don’t have food or shelter! So really, when you do the math correctly, you see that compared to the no-tax Timm, Mitt the Investor is actually paying something like 60% of his income in taxes.
Does everybody see what I mean? You can play this framing games all day long. The only non-arbitrary approach is to say, “What portion of my income was actually due to my ownership of investments? Now how much on the margin was that taxed?” And the answer is 15%, which means it is perfectly correct to say, “Investment income is taxed at a lower rate than labor income.”
To conclude, I am not saying that issues of fairness or whatever dictate that investment and labor income should be taxed at the same rate. (I happen to advocate that they be so, at 0%.) All I’m saying is that it is needlessly confusing, as well as just plain wrong, to have someone earn two different types of income, add those taxes together, and declare that the total is due solely to the tax on investments.
Essentially, in the example it only matters where the starting point and ending point lie, as well as what taxes apply to the different characterizations of revenue, and how these interact.
Oh gosh, it must have been at least 6 years ago, but I did a 3 month audit of my own finances to figure out what percentage of my revenues went to taxation. I included every tax applied no matter the circumstance and found that my average was just under 40%. Of course, I couldn’t identify which tax was ultimately the most cumbersome, because there was a multiplicative effect depending upon my course of action (as noticed by the variance between the weeks and months).
What I ultimately concluded is that all forms of taxation represent a burden, and that taxes of different sorts can have a multiplicative effect when found in combination. It was only later, after having learned Rothbard’s theory of production, that I realized that not only do taxes have a multiplicative effect for individuals, but they also have a similar effect upon the original factors of production (which ultimately bear the ultimate burden of taxation).
I don’t get this debate. Is everyone missing the fact that interest income and dividends are taxed as income? The only way to pay the 15% tax rate is to buy a stock and sell it at a profit after a year or more. Profits on sales of stocks held less than a year also face the income tax rate.
John, I might be muddying the waters a bit. Aren’t there some weird things like with “carried interest” and the like?
I’m not that advanced. I do know that for the average American investing in the stock market in any form, what I just said holds true (that’s part of my job). I’m not sure if Mitt Romney is any different. There are a lot of forms of tax deferred and tax qualified investment, but payouts are almost always taxed at ordinary income.
In any case, it is worth getting the facts straight on how someone manages to pay the preferential 15% rate. Krugman is flat out wrong that any money put into stocks or bonds gets taxed at 15%.
It seems to me that it’s perfectly reasonable to define your total tax burden by looking at the ratio of your consumption under the present tax code to your consumption in a world with no taxes. By that measure, if you earn a dollar, pay 35 cents in taxes and consume the rest, your tax burden is 35%. If you earn a dollar, pay 35 cents in taxes, invest the rest until it doubles, then pay 15% on the investment income and consume what’s left, your tax burden is about 40%. I claim that 40 is greater than 35.
Steve Landsburg wrote:
It seems to me that it’s perfectly reasonable to define your total tax burden by looking at the ratio of your consumption under the present tax code to your consumption in a world with no taxes. By that measure, if you earn a dollar, pay 35 cents in taxes and consume the rest, your tax burden is 35%. If you earn a dollar, pay 35 cents in taxes, invest the rest until it doubles, then pay 15% on the investment income and consume what’s left, your tax burden is about 40%.
Right, which is what your original post was. Should I just paste in my own post, which explains why I think that’s not the right way to do it? 🙂
Look, even in your own summary here, you casually say “if you earn a dollar” as if that’s a neutral starting point. But no, you mean “if you earn a dollar through wage income, and then invest it…” If Timm / Mitt started out life as the heirs to some stocks, and never worked a day in their lives as an employee, then the initial $1 million would be due to investment income. Then you’d find that Mitt would have a lower tax rate than a wage earner, even if you looked at consumption
I claim that 40 is greater than 35.
I agree with that sentence.
If Timm / Mitt started out life as the heirs to some stocks, and never worked a day in their lives as an employee, then the initial $1 million would be due to investment income.
Which in turn would be due to the wage income of some ancestor, so that an analogue of my calculation will still work.
*sigh* You’re going to make us do this, I see. OK Steve, right, an analog of your calculation would work, at which point I would invoke my point from my post: Why stop at two steps? If you have to go back to Mitt Romney’s dad to make your point, why stop with his dad? Mitt Romney’s dad, after all, didn’t just appear out of Zeus’ mouth, with an infinite capacity to labor and generate wage income.
I think if there were a blogger running around saying the actual tax rate on investment income were 99.884%, because he went back to the time when humans began walking upright, you’d think that was a bit weird. So I’m saying the only non-arbitrary point is to go back one step, when the income we are considering is earned. You haven’t told me why two steps is better than one, you just keep reiterating that if we do it your way, the tax rate is 40%.
(BTW I can’t convey body language over the internet; I hope the above doesn’t sound irritated.)
Let me try it this way, Steve: It is not true that all investment income can be traced back through time, to labor. There were natural resources that people homesteaded. We can get really crazy talking about the proper treatment of nonrenewable ones, but what about renewable resources? Suppose there’s a guy who owns a piece of farmland that his ancestors captured from the state of nature. The rents it throws off are just as “primitive” as labor income. If you want to say they would be taxed at 35% too, we can come up with some crazy scenario where they don’t rent it out, but let the rabbits breed on it for 100 years and then sell everything for a big capital gain. Then they use the proceeds to buy stocks.
Look, the big picture thing here, is I don’t like how you and Scott Sumner think income is “really” consumption, and/or that all income is “really” derived from labor. I don’t care about the tax consequences, I care about defining terms like “income” properly. This is probably an idiosyncrasy of mine due to my Austrian background, but nonetheless I get very uncomfortable when Scott Sumner says “income is a meaningless concept,” etc. (And sorry if I’m unfairly lumping the two of you together, but you seemed to be joined at the hip when Sumner was uttering such things.)
I’m not sure what the strict economic definition of “captured from the state of nature” is, but it’s easy to construct a model where this farmland capital comes from labor: the ancestors’ homestead labor was exchanged for a wage of real estate from its prior owner (“The US government”, “Earth”, or something else).
Two legs are better than one. Or something.
It’s not true that the only non-arbitrary point is to go back to when income was earned. In fact, “earned” is totally subjective. The only non-arbitrary point is to go forward to when resources are consumed.
Steven E Landburg wrote:
“Which in turn would be due to the wage income of some ancestor, so that an analogue of my calculation will still work.”
The plug and chug math is right, but this is one place libertarians are not consistent. Typically, as it was here, it’s argued that the income was earned by the ancestor, so the lower tax rate for the person who inherits it is fine. So the benefit passes down from one generation to the next.
But if the ancestor dies, and owes more than the estate is worth, all debts stop right there and they are not passed to the next generation. The typical libertarian argument is the next generation is not responsible for the excesses/sins of the previous generations. I’ve seen Landsburg write things like reparations for income earned through the use of slavery are not his responsibility (Not intending to imply there is any slaveholding in Landsburg family background, just know Landsburg has publically written that current generations are not responsible for sins of generations past).
If benefits are passed down from generation to generation, why wouldn’t burdens?
I think this is where the “stock vs. flow” dichotomy comes in handy – if you look at an investor’s tax rate for any given
year, you’ll get a much lower figure than if you calculate their “lifetime average” tax rate. This is because, as Landsburg pointed out, the funds used to finance aninvestment must have originated from somewhere, and wherever those funds came from, be it wages, inheritence, etc… chances are the government taxed them.
I feel the gasoline/groceries analogy doesn’t quite work, since all people have to consume energy and food (regardless of whether they’re a laborer or capitalist), whereas the decision to invest or consume the remainder of your income is a personal choice (which is subject to influence through the tax code)
Oops delete the previous (it says “awaiting moderation), something messed up my name field somehow. Here’s the body of the comment I tried to make:
Somewhat related: Krugman has been claiming that the rich pay 15%, and people have been asking, “What if you include the corporate taxes they pay?” He recently put up a post where he reluctantly admits its much higher, but tries to change the topic, I called him on it, and a bunch of commenters (though of course not Krugman) replied … with more evasion.
I still have to look at that post…
I think your criticism is fair and the argument really boils down to whether investment income should be taxed like income. If it should be taxed like income, then it’s unfair that Mitt pays only 15%. If it shouldn’t be taxed, then it’s unfair that Mitt pays more than 0%.
To the extent that investment income is soley intertemporal substitution with some risk premium, fairness and economic prudence would seem to dictate that it shouldn’t be taxed, hence Steve’s argument. But some people who disagree with Steve think that investment income should be treated like labor income. And it seems like there’s at least some substance here; there isn’t just one savings rate of return, so investment income is the result of both invested capital wealth and labor to identify the right investment opportunity. Plus, carried interests sometimes look like labor income.
Ben wrote: I think your criticism is fair and the argument really boils down to whether investment income should be taxed like income.
STEVE LANDSBURG, do you see what you’ve done? You just made poor Ben write, “whether investment income should be taxed like income.”
Ben, do you think poodles should be taxed like dogs?
Clearly the intent was for the word ‘labor’ to be inserted between ‘like’ and ‘income’.
Ben wrote:
Clearly the intent was for the word ‘labor’ to be inserted between ‘like’ and ‘income’.
Maybe so, but I don’t think it’s a freak coincidence that you made that Freudian slip. Guys like Sumner really do think that labor income is more “real” than investment income. Read my link if you’re curious.
Perhaps the key question is what thing we’re measuring. Steve is measuring taxation as it affects consumption while you’re measuring taxation as it affects things that can change the numbers in a bank account balance (called “income”). I think the former is the better way to view the problem:
http://www.themoneyillusion.com/?p=7091
The implicit assumption in that view, however, is still that people can’t use labor to affect their rate of return.
No, Ben, Steve does not have a monopoly on looking at the effects on consumption. He is saying, “If a guy earns labor income that gets taxed at 35%, then takes the remainder and invests it where the income from that gets taxed at 15%, then the total impact on consumption is 40%. Hence, I conclude that the investment tax is higher than the wage tax.”
In my (convoluted) scenario of a family that lives off the farmland it inherited from its ancestors, I could talk about the after-tax consumption path and it would be 15%.
Pretty close, except that last sentence should read “Hence, I conclude that his investment income is taxed more than his wage income.”
Bob: Let’s try a different question so that I can make sure I understand the locus of our disagreement.
You and I each build a house, and we each get paid 12 apples, of which the govt takes 6. You eat your 6 apples. I trade my 6 apples for 6 oranges (the going rate of exchange being one for one), pay a 3 orange transaction tax, and eat my 3 oranges.
Is it or is it not fair and accurate to say that my tax burden is bigger than yours?
Steve: Yes, absolutely, there is a very useful sense in which your tax burden (as the orange lover) is higher than mine (as the apple lover). I’m with you so far.
Bob: Okay. Next question:
You and I each build a house, and we each get paid 12 apples, of which the govt takes 6. You eat your 6 apples. I trade my 6 apples for 6 bonds, each of which promises 2 future apples (the going rate of exchange between apples and these bonds being one for one). The government collects a 3-bond transaction tax, leaving me with 3 bonds.
Is it or is it not fair and accurate to say that my tax burden is bigger than yours?
After I get your answer, I have one more followup question.
Yep. You may proceed.
Actually, before following up, I want to change the numbers slightly to make things clearer. I don’t think this will change your “Yep”, but tell me if it does.
Reworded question:
You and I each build a house, and we each get paid 12 apples, of which the govt takes 6. You eat your 6 apples. I trade my 6 apples for 6 bonds, each of which promises 2 future apples (the going rate of exchange between apples and these bonds being one for one). The govt collects a 2-bond transaction tax, leaving me with 4 bonds.
Is it or is it not fair and accurate to say that my tax burden is bigger than yours?
Here’s why I presume that I can carry over your “Yep”.
Followup question: Same story, but instead of collecting a 2-bond transaction tax, the govt lets me keep my 6 bonds until I cash them in for 12 apples next period, then collects a 4-apple tax on my capital gain. Is there any significant difference between this and the question to which you said “Yep”?
I’m skimming this but it sounds right to me. I think you see what my position is, in the follow-up post. I am not disputing that the particular guy in your thought experiment has a higher tax burden (as a % of lifetime income), I’m disputing that his existence proves the point you derive from it.
Murphy, I agree with you! One of us must be wrong.
You disappoint me, MamMoTh. Surely you could have said that coconuts are better than apples.
Well, coconuts do have essential fatty acids…. And, the amino acid profile is almost complete (a little lean on the lysine). Of course coconuts are better than apples. Ok, so coconuts are a little short on the micronutrient side, so what? Eat a few greens and you’re in the pocket.
Thinking purely about consumption makes this easier IMO. Consumption, after all, is what we really care about.
Let’s say that there were a flat consumption tax of 50%. It doesn’t matter how Mitt Romney earns his keep, he still gets 50% less consumption than if he didn’t have to pay taxes at all. In this scenario there are no “framing games,” and nothing is arbitrary. Mitt’s tax burden is 50%.
If we taxed wage income at 50% instead, then the result would be exactly the same, as I believe Landsburg has repeatedly made clear. With zero capital gains tax, Mitt consumes exactly 50% less than if he didn’t pay taxes at all. Mitt’s tax burden is, therefore, 50%.
You say that “The only non-arbitrary approach” is to declare that Mitt’s tax burden in the second example is whatever he gives to the federal government this year (which would be zero if he’s retired). I say that a better, less arbitrary approach is to see how much consumption Mitt forgoes because of taxation, which is 50% in either case.
Mark wrote:
You say that “The only non-arbitrary approach” is to declare that Mitt’s tax burden in the second example is whatever he gives to the federal government this year
I don’t think I said that. What I said was, the only non-arbitrary way to determine the relative taxation of labor versus investment income, is to see how much of your labor income the government takes, versus how much of your investment income. They take 35% of your wages, and 15% of your capital gains. ergo, capital income is taxed more lightly than labor income.
Mark I strongly encourage you to read my article on this. I think Sumner (and possibly Landsburg) do a disservice to the reader by focusing on stylized examples and thinking it’s all about consumption.
Mark wrote:
If we taxed wage income at 50% instead, then the result would be exactly the same, as I believe Landsburg has repeatedly made clear. With zero capital gains tax, Mitt consumes exactly 50% less than if he didn’t pay taxes at all. Mitt’s tax burden is, therefore, 50%.
Just to make sure you’re getting my problem with this, Mark: What you wrote above is not true in general. It is only true if all investment income is ultimately derived from labor income that was saved and invested, and moreover if that labor income was taxed at a rate at least as high as the current rate.
But that’s not true theoretically, and it’s not even true in practice. So I don’t know why Landsburg, Sumner, and lots of other free-market economists assume that capital gains is “really” just deferred consumption from your prior labor paycheck.
“… It is only true if all investment income is ultimately derived from labor income that was saved and invested…”
Yes, it works because the tax on wages ripples through to make the investment smaller and hence make the dividends smaller.
Tel, for the love of all that is decent and good, I understand Landsburg’s specific example. You don’t have to convince me. My point is, there are other examples of people who earn an income from some other source (like from owning a forest), and then invest it. It’s like Landsburg said, “Suppose Mitt buys a dog. Then we can say that all people who earn investment income shoulder a higher tax burden, and have to pick up poop every day.”
I’m trying to agree with you here.
The effect is a ripple-down from income tax, so if your money came from somewhere that did not get income tax applied, then no ripple-down effect either.
The wage tax reduces the investment, therefore it also reduces the dividends so the dividends are effectively taxed by the wage tax… then they are taxed again by the capital gains tax and hence reduce even more. That’s why such a big difference cranks out (and I note that Landsburg uses a hefty multiplier by presuming the investment doubles in size).
For this reason capital gains tax is a double tax, and worse is discourages saving and investment so it is also a downright stupid tax if you want your economy to grow.
I don’t believe it discourages savings, but it can distort investment, especially when the tax on capital gains is independent of the timeframe in which those gains were realized. Plenty of countries tax capital gains independently of whether you make them in 1 year or 10, which is a stupid incentive for short term speculation.