A number of Democratic politicians—and some economists including Paul Krugman—have recently advocated substantially higher income tax rates on high-income Americans. The current top federal tax rate on income is 37 percent for married people filing jointly, and it applies to all taxable income over $612,350. The highest state income tax rate in the United States is in California, where it is 13.3 percent on taxable income over $1 million. Thus, the highest-income people in California lose over half of their incremental income to the government. That the politicians favor higher tax rates is not surprising; that some economists do and that one particular economist is making such a bad case for higher tax rates is somewhat surprising.
Economists, whatever their ideology, tend to oppose high marginal tax rates for one very good economic reason: what they call deadweight loss. Moreover, a substantial minority of economists, including myself, opposes high marginal tax rates on philosophical grounds. Also interesting is that the majority of Americans, if they understood how high are the tax rates that high-income Americans pay, would favor substantially cutting tax rates on the top earners.
These are the opening paragraphs of my article on the Hoover Institution’s Defining Ideas site yesterday. The article is titled “The Case Against Higher Tax Rates.”
And here are my two favorite paragraphs, which I’m grateful that the editor kept:
Notice something else that Krugman seems to believe: rich people literally don’t count. He writes, “[T]he social benefit from getting high-income individuals to work a bit harder is the tax revenue generated by that extra effort.” That’s true only if rich people aren’t part of society. And in case you thought Krugman was just being careless, look at another line above: “[W]hen taxing the rich, all we should care about is how much revenue we raise.” He seems to regard “the rich” as cattle to be raised and exploited. That’s a very negative view of humanity.
It’s also a strange view. Krugman, who is quite rich whether measured by income or wealth, is excluding himself from society. Some might say that I care more about Paul Krugman than Paul Krugman does.
And finally:
Even noted activist Al Sharpton thought the rich should be taxed less. In a 2004 interview with John Stossel, Sharpton said that the rich should pay their fair share of taxes. Stossel asked him what percent of total federal income taxes the rich should pay. Sharpton’s answer: at least 15 percent. At the time, noted Stossel, they paid 34 percent of all federal income taxes. All the rich people I know, except for Paul Krugman, would take Sharpton’s deal in a New York minute.
Read the whole thing, or, as Tyler Cowen always says, “Do read the whole thing.”
READER COMMENTS
Ryan Sullivan
Jan 24 2019 at 1:27pm
Interesting article. It is my understanding that the 73% value from Diamond and Saez (2011) is the tax revenue maximizing marginal rate when combining all taxes (see the bottom of page 171 in their article) for top income earners. Therefore, when you add in state taxes, sales, etc. (particularly for Californians) the federal top marginal rate should be much lower than the 70% rate many are advocating (even putting aside the idea of treating rich people like cattle). Basically a standard 70% marginal rate at the federal level could easily put us on the right hand side of the Laffer Curve if this policy were to actually be implemented. Maybe I am wrong, but that is my interpretation of the literature.
https://www.aeaweb.org/articles?id=10.1257/jep.25.4.165
Thaomas
Jan 24 2019 at 1:40pm
Well I think that at full employment the Federal Government should be running a slight surplus or maybe a slight deficit depending on the rates of return on it’s investment opportunities. We have quite a gap to close and I don’t think lower military spending will or should make THAT big a contribution to closing it. There are no doubt some low value expenditures that can be cut, but not many of these have not been well identified so far and there are other things that the government should be spending more on such as health insurance and child care subsidies. Consequently I think that higher tax revenues should do most of the closing and that most of the additional revenues should come from the top of the incomes distribution. Ideally, this would come through taxation of the consumption of high consumption individuals rather than on their incomes per se so as maintain incentives to work, innovate and invest.
Phil
Jan 24 2019 at 1:57pm
In the design of your ideal tax, is a “high consumption individual” defined by one’s level of consumption relative to one’s income, one’s wealth, or some arbitrary dollar figure? And why?
Mike W
Jan 26 2019 at 10:29am
How about relative to the household final consumption expenditure per capita?
robc
Jan 24 2019 at 2:22pm
First cut federal spending under 15% of GDP. I think that is too high, but I will accept it*.
At that point, we can look at if we are running a surplus or not and see about adjusting the tax rate (HINT: we would be running a big surplus).
* I have said multiple times in the past that if the Feds permanently cut spending to below 15% of GDP I will shut up about spending for the rest of my life. It will be my daughter’s job to push it lower.
Kevin Dick
Jan 24 2019 at 2:02pm
Good article. Here are two papers that I refer my progressive friends to:
IMF estimate showing that revenue maximizing marginal rate in advanced economies is 44% _even if you give no weight to the utility of top earners_:
Page 13 of http://www.imf.org/…/2017/10/05/fiscal-monitor-october-2017…
This calculation is based on Saez 2001, so hardly a conservative/libertarian slant.
Then Tyler recently linked to this preliminary paper showing a welfare maximizing top marginal rate of ~30% if you combine the literature on optimal taxation with the literature on idea-driven economic growth.
https://www8.gsb.columbia.edu/faculty-research/sites/faculty-research/files/finance/Macro%20Workshop/toptax.pdf
Kevin Dick
Jan 24 2019 at 2:37pm
Nice article.
I have been pointing my progressive friends to the following two papers on this topic:
(1) IMF paper showing revenue maximizing marginal rate of ~44%, even if you put _zero weight_ on the utility of top earners.
https://www.imf.org/~/media/Files/Publications/fiscal-monitor/2017/October/pdf/fmc1.ashx (page 13).
This is based on the methodology in Saez (2001), so hardly a conservative/libertarian slant.
(2) Preliminary paper linked by Tyler combining the literature on optimal taxation with the literature on idea-driven economic growth to show the social welfare maximizing top marginal tax rate is ~30% under baseline assumptions.
https://www8.gsb.columbia.edu/faculty-research/sites/faculty-research/files/finance/Macro%20Workshop/toptax.pdf
David Seltzer
Jan 24 2019 at 3:25pm
Let’s talk about the elephant inn Krugman’s room. What he really suborns, in his misguided moral purpose, is the use of force to implement this travesty. What he may not understand, in his naivete, some of us know how to use greater force to countervail this attack on property and liberty. The reader is free to infer what he chooses.
Vivian Darkbloom
Jan 24 2019 at 3:36pm
“The current top federal tax rate on income is 37 percent for married people filing jointly, and it applies to all taxable income over $612,350”.
Careful here. There is no single “top federal tax rate on income” and that 37 percent rate does not apply to “all taxable income over $612,350 for married couples”. You can argue whether qualified dividends and long-term capital gains are properly classified as “income” and/or whether we should tax it as such, and you can argue endlessly over the incidence of corporate level taxes; however, the Code treats these items as “taxable income”, albeit not taxable *ordinary income*. The top federal income tax rate on these types of income is not 37 percent. My guess is that for those with total taxable income over $612,350, these types of “income” are not insignificant.
This isn’t an argument for higher marginal federal tax rates on any type of income. I’m simply clarifying the record for the sake of the discussion.
RPLong
Jan 24 2019 at 4:19pm
Doesn’t this fact support David’s claim that x-percent tax rate increases result in less-than-x-percent tax revenue increases?
If I were someone who had high ordinary income and high wealth and faced a steep increase to my marginal income tax rate, I’d certainly consider turning some of my employment income into capital gains income instead.
Vivian Darkbloom
Jan 24 2019 at 4:56pm
RP Long,
Thank you for agreeing that what I pointed out in my comment was “factual”. My only objective was to counter the non-factual assertion in the original post that “The current top federal tax rate on income is 37 percent for married people filing jointly, and it applies to all taxable income over $612,350.”. My point was that no, that rate does not apply to *all* taxable income. Also, in partial response to Zeke, below, here’s a recent breakdown of the sources of income by income group as of 2016:
https://taxfoundation.org/sources-personal-income/
As you can see, those who would be affected by the proposed 70 percent marginal rate garner more income from dividends and capital gains than from wages. But, who knows what would happen between $612,350 and $10,000,000 and whether the top rate would apply to *all* income or not. This whole debate does not strike me as very well informed…
I have no doubt that there are diminishing returns to higher marginal rates on whatever type of income is taxed in the sense stated, that is, an X percent increase in the tax rate will not result in an X percent increase in tax revenue, but I’m not sure how what I wrote either supports or detracts from that assertion.
Also, and for Zeke’s benefit as well, I don’t think it is as easy as you assume to “convert” ordinary income to dividend or capital gain income. I suppose the most common way to garner more of the latter and less of the former (not a “conversion”) would be to forego salary in exchange for stock options which is not at all a sure bet. And, there are usually tax winners and losers there, losers being corporate deductions and corporate shareholders. Finally, “carried interest” does not entail “converting” ordinary income to capital gain income. This is a complete misnomer suggesting that somehow King Midas is involved in the transaction. It entails managing partners negotiating with other partners over who gets what out of the profit pie in an arm’s length transaction called a partnership agreement. The problem is that the counterparty giving up their share of (potential) capital gain is quite often a not-for-profit or foreign taxpayer not subject to US tax. For them, the classification of income is not relevant. For the rest, I’m not worked up that taxable Partner A cedes part of his potential capital gain to taxable Partner B (with also a corresponding loss of an ordinary income deduction for ordinary compensation).
zeke5123
Jan 24 2019 at 5:51pm
I think that is an incorrect view of carried interest. The people who generally get carried interest (i.e., PE or Hedge Funds) aren’t relying on allocations of cap gain to one party and away from another party. Instead, they are relying on the general fact that the income generated by PE and Hedge funds tend to be capital in nature.
So, when PE fund manager is paid his or her performance bonus (i.e., the 20% in the 2 & 20 framework) based on the profits of the PE fund, well, of course the profits will be capital gain because PEs generally only invest in capital gain property. The real magic of the carried interest for PE funds is that profits interests can be issued for free (as they have zero liquidation value even though they very likely have FMV) and the profits allocated to these interests are going to be capital in nature. This is King Midas. It is turning income generated by performing services into income taxed as capital.
What you are suggesting is Party A and Party B form a partnership. Under the partnership agreement, Party A is allocated capital gain and Party B is allocated ordinary income. Depending on the arrangement, if Party B is indifferent to the nature of the income, then both parties can be in a better after-tax fact pattern. But I’d have strong reservations that the above economics reflect substantial economic effect. Thus, it might not stand up on audit.
Vivian Darkbloom
Jan 24 2019 at 6:26pm
Zeke, I am not at all mistaken. The hedge fund or partnership agreement decides which partners get what percent of the ultimate capital gain produced by partnership investments (even if not proportionate to capital contributed, which is what typically happens here). The total amount of CG reported to all partners is not affected by how the allocation is agreed to. Thus, if we agree that the total CG produced by the partnership is 100, that is the amount to be allocated among partners. The agreement does not magically allow the partnership to earn, say, the CG of 100 to one of 110. Thus, as I stated, it is merely a question of how the CG is allocated. While the allocation to the managing partner of CG non-pro rata to his capital interest is not taxed when the agreement is entered into, the partner’s giving up potential CG are also not claiming any up-front loss or deduction. One partner’s potential and ultimate gain is another’s potential and ultimate loss and there is no King Midas effect. King Midas touched something base and turned it into gold. He didn’t take gold from someone else’s pocket and put it into his own. Of course, as I stated quite clearly, if one partner is a tax-exempt, he is indifferent to whether his allocation of profit is ordinary or CG (and also indifferent to the nature of his deductions). That is the tax play. I wish that this would not be so often misrepresented in part because tax exempt investors (such as Harvard Endowment, etc) engaging in this tax status arbitrage should be getting as much criticism as hedge fund managers. They are simply not such easy targets for political rhetoric.
Viv
RPLong
Jan 25 2019 at 9:48am
Vivian,
I like your writing style, and I always appreciate your comments when I see them, including the two above.
You call David’s assertion “non-factual,” but I am not sure that’s fair. We all know he’s talking about the federal income tax rate. Your point about other forms of income is not lost on me, though. All I was trying to say is that it’s consistent with David’s main point in his article, ie. that raising the federal income tax rate tends to reduce tax collections as people shift from more onerous income declarations to less onerous ones.
Note that in the above paragraph my language may be imprecise in that you still may feel inclined to differentiate between various federal taxes on various forms of income. But, I believe I have been precise enough to convey my meaning. I hope so anyway.
In short, you make a good technical clarification that is still consistent with what David wrote, and which describes the mechanism by which it occurs.
Vivian Darkbloom
Jan 25 2019 at 12:12pm
RP Long,
Thanks for your comment. I certainly get the gist of your message; however, I still need to insist that there is not one federal marginal rate that applies to *all taxable income*, in particular the 37 percent rate. To say the 37 percent rate applies to “all taxable income” is not factual and I don’t think that is even debatable. And, I know from reading David’s comments here that he holds not only himself but others to a high standard of accuracy. Despite that, I also get David’s general point and I happen to agree with it. I tried to make the correction in a straightforward way and in the spirit of not playing “gotcha”, but I sense that message is difficult to get across even in the spirit of friendly cooperation. This is not a trivial point given the proportion of overall income earned by those subject to the highest thresholds that is classified as qualified dividends and LTCG’s. See, e.g., the Tax Foundation site I linked to elsewhere.
Not to belabor the point (and I wouldn’t but for the pushback); however, David links below to an IRS page referring to the 3.8 percent Medicare surcharge on *investment income* to propose that the marginal rate becomes 37 percent plus 3.8 percent to equal a federal marginal rate of 40.8 percent. This is true only with respect to ordinary investment income such as net rental income, ordinary passive partnership income and short term capital gains. Again, it is *not* true with respect to the large category of qualified dividends and LTCG’s because those are not subject to the 37 percent federal income tax rate in the first instance. It is true with respect to self-employment income and wage income, but only with respect to the latter if one imputes the employer-paid Medicare contribution of 1.45 percent to the employee. The category of wage income is subject to a separate surcharge of only 0.9 percent (not the investment income surcharge) because Medicare on wages is not subject to a ceiling and therefore wages are subject to a 2.9 percent levy before the surcharge (employer and employee each 1.45 percent) and so with the surcharge the total is 3.8 percent.
When the debate inevitably gets more detailed as to whether high income taxpayers should be subject to a higher marginal rate on qualified dividends and LTCG’s, David certainly won’t be able to state that, well, “all taxable income” is already subject to a federal marginal rate of 37 (or 40.8 percent) as he’s done here. As Warren Buffet might say, that’s when you will know who’s bathing without shorts!
Viv
zeke5123
Jan 24 2019 at 3:49pm
Well, that wouldn’t be true Vivian for certain IB, lawyers, accountants, doctors, etc. Basically, anyone who is (i) making a large sum of money, (ii) cannot (depending on the kind of IB) convert wage income into capital gain via carried interest, and (iii) do not qualify for the 199A deduction.
My guess is there actually is a not insignificant group of tax payers paying tax on OI above 600K.
Vivian Darkbloom
Jan 24 2019 at 4:13pm
Zeke,
Where did I write that there is not a significant group of taxpayers paying tax on ordinary income over that threshold? That same group does not also typically have investment income in the form of dividends and long-term capital gains? What, then, exactly, “wouldn’t be true” with regard to what I did write?
Viv
zeke5123
Jan 24 2019 at 5:54pm
I misread your above post. I thought you said OI would not be insignificant, when instead you said income taxed at preferential rates would not be insignificant.
Mark Z
Jan 24 2019 at 4:50pm
David, Krugman is also ignoring the productive saving and investment of rich people that is deterred by a higher marginal tax rate, no? To the extent that they consume little, rich must necessarily save a lot.
Which leads me to the following tangent/question (for which I apologize). It’s often claimed that the empirical evidence suggests no association between marginal tax rate and economic growth rate, which might suggest higher tax rates don’t deter saving/investment. I was wondering what you thought of this claim? It occurred to me that people – or particularly rich people – may have strong preferences regarding future consumption, which may render their saving habits highly invariant to the marginal tax rate. That is, they may respond to an increase the tax rate by increasing their saving rate and cutting back consumption, because they’re committed to saving at least a certain amount per year to guarantee their preferred standard of living in retirement. This might explain private investment and growth rates being fairly invariant to marginal tax rates, assuming that’s what the data show.
David Henderson
Jan 25 2019 at 9:37am
Mark Z,
No apology necessary. It’s not that tangential.
I’m not familiar enough with the literature to know that, as you say, there is “no association between marginal tax rate and economic growth rate.” What I do know is that Krugman and some other economists have essentially dropped virtually all their training in econometrics and basic economic reasoning in making that claim. They point to the high growth of the 1950s along with a top marginal tax rate of 91% and say, essentially, “Look, we had both, so high marginal tax rates don’t hurt growth.” This reasoning makes sense only if the only factor that differed between then and now is the marginal tax rate on very high income people.
Dan Culley
Jan 28 2019 at 12:06pm
But note that, if that is the desired behavior, you can achieve the same end by only taxing their consumption and not their savings. (And you can do it progressively if you want.) Then you don’t have to care how invariant their savings rates are to tax rates and there is one less failure mode.
Of course, a lot of the other compliance and recharacterization issues are still there. But it is hard to think of a reason why one would prefer to tax high earners progressively on income rather than on consumption.
Fred Foldvary
Jan 24 2019 at 5:13pm
The most efficient tax is one that has a zero marginal tax rate, so that even a high average tax rate would have no deadweight loss. The most equitable tax is on subsidies, which offset the subsidy. Therefore, the tax that maximizes both efficiency and equity is on gains which would otherwise be a subsidy, and which falls on potential rather than actual values. Why this is not more widely recognized is a mystery.
Ken P
Jan 24 2019 at 9:39pm
I agree with your article overall, but in the 2004 interview, Sharpton was likely referring to long term capital gains tax which is the majority of income for a lot of wealthy people. This is also the basis for Warren Buffet’s claim about paying lower taxes than his secretary. In 2004, the maximum long term capital gains rate was 15%.
Brandon Berg
Jan 24 2019 at 10:11pm
The top federal tax rate on wage income is actually 37% + 3.8%. Don’t forget Medicare.
David Henderson
Jan 25 2019 at 9:32am
Right you are. Thanks for the correction.
Here are the details:
https://www.irs.gov/newsroom/questions-and-answers-on-the-net-investment-income-tax
Frank
Jan 25 2019 at 10:24am
1. Are understating income and overstating expenses supposed to be examples of deadweight loss?
2. I assume you are in favor of some taxation of income, so the mere existence of deadweight loss does not settle the matter. How do you decide when to stop?
Fred in PA
Jan 26 2019 at 1:33am
Mark Z;
I assume you are questioning the “often [made] claim … that the empirical evidence suggests no association between marginal tax rate and economic growth rate, which might suggest higher tax rates don’t deter saving/investment.” And that you are offering a hypothesized explanation for how that might be possible.
(Thus your “rich people [–] may have strong preferences regarding future consumption, which may render their saving habits highly invariant to the marginal tax rate. That is, they may respond to an increase the tax rate by increasing their saving rate and cutting back consumption.” Yes?)
As a merely dilettante economist, I should think the likely situation would be the reverse. That the truly rich can already provided for any reasonable level of future consumption. And that cutting back one’s lifestyle / consumption is fairly painful for most of us (and for them).
Thus the social-justice types won’t get their desired reduction in the wealthy’s “offensively luxuriant” lifestyle — the wealthy will pretty much continue in the style to which they’ve become accustomed. Rather, it is their savings rate that will take the hit. “After paying Junior’s tuition at Princeton and replacing the rotting dock, there is just less left over this year for padding our (already substantial) portfolio.”
I would guess that if “the empirical evidence suggests no association between marginal tax rate and economic growth rate,” it might be that the tax rate goes up (and the investment level goes down) in year Y but the lost growth (from that lost investment) won’t be manifest until two or three years later. (And teasing it out from the statistical noise will be the devil!)
Mark Z
Jan 26 2019 at 11:12am
That’s what I was indeed attempting (I haven’t read any of the literature so I’m ignorantly agnostic on the relationship between tax rate and growth, I’ve just seen people make the claim).
I can see a narrative going either way. Of course, a responsible poor person who wants to guarantee his retired self a sufficient standard of living may be more likely to increase his savings rate in response to an increase in his tax rate than a rich person will, since the latter already makes more than enough to finance a decent retirement. But it may also be the case that the rich person already makes more than enough to finance his desired present lifestyle, and perhaps values the prospect of having a yacht when he’s retired more than he values having one today, so he’s willing to reduce present consumption today by X in order to increase future consumption by .9X.
That may be a highly unusual preference, but I think many middle class and wealthy people do have a mentality of “I have to save at least $X per year, no matter what,” and might reduce present consumption to make that goal in the face of a tax rate. This may be a big reason for why they’re rich: they habitually have a high savings rate and so accrue a lot of capital, while poor people may be more likely to have a habitually low savings rate, which may also be invariant to the tax rate, as they may simply consume more in response to a reduction in their tax rate.
And you may be right that there just isn’t solid enough empirical evidence to clarify whether there is or isn’t a clear association. An association (or lack thereof) between marginal tax rate and growth would be a difficult relationship to distill from all the potential confounders.
Comments are closed.