Tax Notes logo

1924, 2021: Taxes of the Ultrarich, and Mark-to-Market Reforms

Posted on Jul. 26, 2021
[Editor's Note:

This article originally appeared in the July 26, 2021, issue of Tax Notes Federal.


Lawrence Zelenak is the Pamela B. Gann Professor at Duke Law School.

In this article, Zelenak tells the story of Treasury’s disclosures of the income tax payments of plutocrats almost a century ago, which mirror the recent ProPublica revelations of the income tax payments of the 25 richest Americans, and he explores why those earlier disclosures — unlike the recent ones — didn’t spark interest in mark-to-market taxation of the ultrarich.

On June 8 ProPublica revealed that it had “obtained a vast trove of Internal Revenue Service data on the tax returns of thousands of the nation’s wealthiest people covering more than 15 years.”1 The ProPublica story “compared how much in taxes the 25 richest Americans paid each year [from 2014 through 2018] to how much Forbes estimated their wealth grew in that same time period.” The story called the ratio of tax to wealth increase the billionaires’ “true tax rate.” During the five years in question, the wealth of the 25 rose by a collective $401 billion, and they paid a total of $13.6 billion in federal income tax, resulting in a true tax rate of 3.4 percent for the group. Interesting details on individuals included that Warren Buffett had the lowest true tax rate in the group (0.1 percent) and that Jeff Bezos’s tax return income was so low in 2011 (when his net worth was about $18 billion) that he claimed and received a $4,000 child tax credit.

The article explained that the primary cause of the discrepancy between huge net worth increases and much smaller tax liabilities was the tax system’s failure to treat unrealized appreciation as income, abetted by the tendency of the corporations on whose stock the wealth of the 25 was based to pay little or nothing by way of dividends and to pay relatively small (in some cases, merely nominal) salaries to their ultrawealthy founders. Although the disconnect between wealth and tax bills could be addressed outside the income tax by a wealth tax, the article explained that it could also be addressed within the income tax by mark-to-market taxation of the ultrawealthy. In that connection the article noted a 2019 discussion paper by Sen. Ron Wyden, D-Ore., now the Finance Committee chair, proposing exactly that.2

There is no possibility that the ProPublica revelations will lead Congress to enact mark-to-market taxation of billionaires anytime soon. Wyden’s 2019 paper endorses the general notion of mark-to-market taxation of the ultrawealthy, but it is far from a ready-to-be-enacted legislative proposal. Rather, it notes numerous policy choices and technical challenges involved in designing a mark-to-market system, and it solicits thoughts on those issues. Commenting shortly after the appearance of the ProPublica story, Kimberly A. Clausing, Treasury deputy assistant secretary for tax analysis, accurately said mark-to-market proposals “aren’t quite ready for prime time,” and indicated that the Biden administration would not be advocating that approach this year.3 Nevertheless, it seems likely that the ProPublica disclosures will move mark-to-market taxation (for the very wealthy, and perhaps only of their publicly traded stock) from academic pipe dream to serious legislative proposal in the next few years.

Almost a century ago, in 1924 and 1925, the United States experienced similar revelations of the precise federal income tax payments of the wealthiest Americans — albeit under express legislative directive rather than through violations of tax return privacy law. In striking contrast with the 2021 revelations, the disclosures of 1924 and 1925 generated no interest in mark-to-market income tax reforms. This article tells the story of those earlier disclosures and discusses why those revelations did not spark any interest in including unrealized appreciation in the base of the federal income tax.

I. Income Tax Payment Disclosures, 1924

The Revenue Act of 1924 directed the commissioner to prepare and make available for public inspection “in the office of the collector in each internal-revenue district . . . lists containing the name and the post-office address of each person making an income tax return in such district, together with the amount of the income tax paid by such person.”4 In late October 1924 Treasury announced that the lists were available for inspection by the general public (including, of course, journalists). Later the same day, however, Treasury warned that publication of any information from the lists would violate the general criminal prohibition against the disclosure of tax return information.5

Relying on its counsel’s advice that the criminal prohibition did not apply to information made public under the new disclosure statute, The New York Times published the names and amounts of income tax payments — payments made in 1924 on 1923 income — of hundreds of wealthy (or otherwise prominent) taxpayers from New York City and around the country.6 Although several other newspapers also published selected tax payment information, not all papers were so bold. The Wall Street Journal, perhaps advised differently by its attorneys or perhaps simply more cautious than the Times, did not publish tax payment information.7

The legality of publication was not settled until May 1925, when the Supreme Court sustained the dismissal of an indictment against the owner-editor of several Kansas City, Missouri, newspapers for publishing information from the disclosure lists.8 Without dissent, the Court reasoned, sensibly enough, that “to the extent provided by [the disclosure statute] Congress meant to abandon the policy of secrecy altogether, and to exclude from the operation of [the criminal prohibition] all forms of publicity, including that here in question.”9

Even with the favorable interpretation by the Supreme Court, the disclosure statute was to be short-lived. After one more round of disclosure and publication — of payments made in 1925 on 1924 incomes — Congress accepted the disclosure opponents’ argument that the 1924 provision was an unwarranted invasion of privacy, and it repealed the provision in 1926.10 Less than a decade later, in 1934, Congress again legislated public disclosure of tax return information — far more extensive than the mere tax payment disclosures of the 1924 legislation — but a remarkable grassroots lobbying campaign persuaded Congress to repeal the provision in 1935, before any disclosures had been made.11

The 1924 and 1925 disclosure lists were distributed across the country in local collectors’ offices; there was no official nationwide list. Moreover, the local lists were not organized by descending order of tax payments, as a curious public might have preferred. The 1924 list of large tax payments (Table 1) is gleaned from information published by The New York Times, but it may well omit some taxpayers who paid more than other taxpayers included in the list. In any event, there is no doubt that John D. Rockefeller Jr. made by far the highest 1924 tax payment (on 1923 income), at more than $7.4 million, and that Henry Ford and his son Edsel Ford came in second and fourth, respectively (with Payne Whitney between them), with combined payments of more than $4.4 million. For 1925 payments (on 1924 income), by contrast, the Times offered what it claimed was a list of the nation’s 25 biggest individual income taxpayers (Table 2). Rockefeller Jr. again led the way (paying about $6.3 million), followed by the two Fords (together paying about $4.8 million). Treasury Secretary (and wealthy banker) Andrew W. Mellon was fourth on the list, paying almost $1.9 million. Below the names of Rockefeller, Ford, and Mellon on the list are many other surnames still familiar almost a century later, including Whitney, Dodge, Vanderbilt, Astor, Duke, and Morgan.

Table 1. Large Individual Income Tax Payments Made in 1924
(on 1923 income)a

John D. Rockefeller Jr.


Henry Ford


Payne Whitney


Edsel Ford


Anna M. Harkness


Andrew W. Mellon


Frederic W. Vanderbilt


George F. Baker Jr.


George F. Baker


S.R. Guggenheim


William W. Woodb


B.F. Holmes


aExcept for the Ford and Holmes tax payments, information in this table is from “Recapitulation of Income Tax Payments Throughout Country Previously Reported,” The New York Times, Oct. 26, 1924. The Ford information is from “Ford Total $18,902,981,” The New York Times, Oct. 24, 1924. The Holmes tax payment is illegible in “Recapitulation”; the amount in the table is from “Income Tax Returns Made Public; J.D. Rockefeller Jr. Paid $7,435,169; Ford Family and Company Pay $19,000,000,” The New York Times, Oct. 24, 1924.

bWood is best known to income tax history aficionados as the executive for whom the income tax on his salary was paid by his employer (the American Woolen Co.), leading to the Supreme Court’s ruling that his employer’s payment of his income tax liability resulted in additional taxable income to Wood. Old Colony Trust Co. v. Commissioner, 279 U.S. 716 (1929).

Table 2. 25 Largest Individual Income Tax Payments Made in 1925
(on 1924 income)a

John D. Rockefeller Jr.


Henry Ford


Edsel Ford


Andrew W. Mellon


Payne Whitney


Edward S. Harkness


R.B. Mellon


Anna R. Harkness


Mrs. H.E. Dodge


Frederic W. Vanderbilt


George F. Baker


Thomas F. Ryan


George F. Baker Jr.


Edward J. Berwind


Vincent Astor


James B. Duke


Cyrus H.K. Curtis


J.P. Morgan


Claude H. Foster


George Allen Hancock


Eldridge R. Johnson


H.H. Timken


Arthur W. Cutten


Arthur C. James


Thomas W. Lamont


aAll information is from Stuart Chase, “Ford’s Billion Breaks World Records,” The New York Times, Feb. 13, 1927. Chase obviously rounded the amounts.

How did these rankings by tax payments compare with rankings by net worth? Then as now, estimating the net worth of the wealthiest individuals was more art than science, or a very inexact science at best. Tables 3, 4, and 5 show the results of three such efforts, not too distant in time from the 1924 and 1925 income tax lists — one produced by B.C. Forbes in 1918 (the very first Forbes richest persons list),12 one from a 1923 New York Times story featuring six Americans among the world’s supposed 10 richest persons,13 and one from a 1927 Times article by Stuart Chase estimating the net worth of the richest Americans.14 Comparing the tax payment lists with the wealth lists, with particular attention to the 1925 tax payment list and the 1927 wealth list, the similarities are striking. Rockefeller, Ford, Mellon, Whitney, and Harkness are the first five surnames to appear on both the tax payment list and the wealth list; except for Whitney and Harkness trading fourth and fifth places in the two lists, even the order is the same.

Table 3. Richest Americans in 1918
(ranked by net worth)a

John D. Rockefeller Sr.

$1.2 billion

Henry C. Frick

$225 million

Andrew Carnegie

$200 million

George F. Baker

$150 million

William Rockefeller

$150 million

Edward Harkness

$125 million

J. Ogden Armour

$125 million

Henry Ford

$100 million

W.K. Vanderbilt

$100 million

Edward H.R. Green

$100 million

aAll information is from B.C. Forbes, “America’s 30 Richest Own $3,680,000,000,” Forbes, Mar. 2, 1918.

Table 4. Richest Americans in 1923
(ranked by net worth)a

Henry Ford

$550 million

John D. Rockefeller Sr.b

$500 million

Percy Rockefeller

$100 million

James B. Duke

$100 million

T.B. Walker

Perhaps less than $100 million, perhaps as high as $300 million

George F. Baker

$100 million

aAll information from “List of Ten Richest Men in the World Includes Six Americans,” The New York Times, May 20, 1923. In addition, the article estimated the following family wealth: Guggenheim, $200 million; Vanderbilt, $75 million to $100 million; Weyerhauser, $100 million or more; Astor, $100 million to $700 million (but probably closer to the lower figure); and Mellon, $75 million (two-thirds of which belonged to Andrew).

bRockefeller Sr. paid income tax of only $128,000 on his 1923 income (paid in 1924). Chase, “Ford’s Billion Breaks World Records,” The New York Times, Feb. 13, 1927. By 1927 it was well known that Rockefeller Sr. had divested himself of most of his wealth by transfers to Rockefeller Jr. and to charity. Id. The 1923 wealth estimate was obviously based on the assumption that Rockefeller Sr. had not made major wealth transfers as of 1923, but the relative income tax liabilities for 1923 (more than $7.4 million for Rockefeller Jr., only $128,000 for Rockefeller Sr.) strongly suggest otherwise.

Table 5. Richest Americans in 1927
(ranked by net worth)a

Henry and Edsel Ford

$1.2 billion

John D. Rockefeller Jr. and Sr.

$600 million

Andrew W. and R.B. Mellon

$200 million

Edward S. and Anna Harkness

$200 million

Payne Whitney

$100 million

George F. Baker and George F. Baker Jr.

$100 million

Vincent Astor

$100 million

Frederic W. Vanderbilt

$100 million

Thomas B. Walker

$100 million

J.P. Morgan

“quite possibly . . . in the $100 million class”

aAll information from Chase, “Ford’s Billion Breaks World Records,” The New York Times, Feb. 13, 1927.

II. Explaining the Dog That Did Not Bark

The tax payment disclosures of the 1920s, unlike the recent ProPublica disclosures, did not generate any interest in mark-to-market income tax reforms. U.S. newspapers did not discuss the possibility of such reforms (not even to denounce them), no members of Congress advocated them, and no high-level Treasury officials felt it necessary to explain (anachronistically) that such reforms were “not ready for prime time.” This section discusses several differences between then and now, all of which help explain why the mark-to-market dog that is barking so loudly now did not bark in the 1920s.15

First, the true tax rates on the ultrawealthy in the 1920s, although low, were not as low as the rates calculated by ProPublica for today’s billionaires. The major reason is that dividend taxation had a greater significance in the 1920s (mostly because dividend yields were higher then, but also because dividend tax rates were higher). Second, there was something of a failure of imagination in the 1920s regarding unrealized appreciation as a component of economic income and potentially as a component of taxable income. Although the case for unrealized appreciation as income was familiar to the leading economists of the time, the concept seems to have been beyond the imagination of journalists, politicians, and the general public. Third, the disclosures of tax payment information in the two periods were framed very differently. In breaking its 2021 story, ProPublica treated as central to its true tax rate analysis the concept of unrealized appreciation as income. As framed by ProPublica, billionaires’ low true tax rates (and not the dollar amounts of taxes they paid) were the story. By contrast, journalistic coverage of the 1920s’ government disclosures was reactive, if not downright passive; the story was not how little tax Rockefellers and Fords paid as a percentage of income, but how much tax they paid in dollars.

A. Taxing Dividends, Then and Now

The megarich of the 1920s benefited enormously from the nontaxation of unrealized appreciation in their stock portfolios and other assets. Consider the case of Henry and Edsel Ford. In the first decade of the 20th century, Henry (and a few others) invested $100,000 in closely held Ford Motor Co. (FMC). (FMC did not go public until 1956, nine years after Henry’s death.) From 1904 to 1926, FMC earned net profits of $925 million. A well-informed observer estimated in 1927 that FMC had paid dividends of only about $175 million over that period, and that Henry had probably not taken much above $100 million from FMC for his personal use during that time.16 Despite the FMC shares not having been publicly traded, we have a good lower bound for the value of the stock as of early 1927: In February 1927 The New York Times reported that Henry and Edsel (then the only two FMC shareholders) had recently received a serious offer (from J.W. Prentiss, on behalf of Hornblower & Weeks) of $1 billion for all their shares.17 Thus, over roughly a quarter-century (including about a decade preceding the enactment of the individual income tax in 1913), Henry and Edsel had enjoyed at least $1 billion of untaxed appreciation in their stock while receiving at most $175 million in taxable dividends (less than that, to the extent that the dividends were paid before 1913).

With the FMC stock not having been publicly traded, it is impossible to determine the amount by which it appreciated in 1923 and 1924 and thus calculate the Fords’ true tax rates for those years using the ProPublica approach of putting the stock appreciation in the denominator of the tax rate fraction. But suppose, not implausibly, that their combined 1924 income under the ProPublica approach was $100 million. Their combined tax on 1924 income (paid in 1925) was a bit under $4.8 million, giving them a ProPublica 1924 true tax rate of 4.8 percent. To be sure, that is not a very high tax rate. It was, nevertheless, high enough to make them the second- and third-biggest individual income tax payers in the country. It was also considerably higher (about 41 percent higher) than ProPublica’s 3.4 percent true tax rate for its 25 billionaires as a group (for 2014 through 2018), and much higher than the true tax rates for the four billionaires analyzed in detail in the ProPublica story (Buffett, 0.1 percent; Bezos, 0.98 percent; Michael Bloomberg, 1.3 percent; and Elon Musk, 3.27 percent).

Moving from the Fords to the ultrawealthy owners of publicly traded stock in the 1920s, the crucial question in determining their true tax rates is the relative significance of taxable dividends and untaxed stock appreciation. The low true tax rates calculated by ProPublica for 21st century billionaires are driven, in significant part, by the fact that today’s billionaires are receiving very little in the way of taxable dividends relative to their wealth. In fact, of the four billionaires spotlighted by ProPublica, three (Buffett, Bezos, and Musk) owe their wealth to unrealized appreciation in the stock of corporations that have never paid dividends (Berkshire Hathaway, Amazon, and Tesla, respectively). By contrast, the Fords and Rockefellers owed their wealth to appreciation in the stock of corporations that may have retained most of their earnings but also paid significant taxable dividends (FMC and the various Standard Oil companies, respectively).18 Of course, the significance of dividend taxation of the ultrawealthy was reduced by the 2003 innovation of taxing dividends at the same low rates as long-term capital gains,19 but the decline in the significance of dividend taxes paid by billionaires owes more to the decline in dividends than to the decline in dividend tax rates.

The above comparison may be anecdotal, in that it is based on a tiny number of wealthy taxpayers in each era, but when the question is the taxation of the very richest of the rich, there is no sharp dividing line between anecdote and data. Moreover, a less anecdotal approach also suggests a substantial decrease in the significance of taxable dividends relative to untaxed appreciation, from the 1920s to now.

The S&P 500 stock index was not introduced until 1957, but economist Robert Shiller has constructed S&P 500-equivalent indices for earlier years, including the 1920s. Consider a hypothetical member of the ultrawealthy in 1923 and 1924, with a net worth of $100 million, all of which was invested in the (Shiller-constructed) S&P 500. The dividend yield on the S&P 500 was 6.2 percent in 1923 and 5.4 percent in 1924.20 As for stock price changes, the S&P 500 index declined by 0.9 percent in 1923 and increased by 19.8 percent in 1924.21 For the two years combined, our investor would have received dividend income of about $12 million and would have enjoyed unrealized appreciation of about $19 million, for total income of about $31 million. If he had no other income and paid tax on the dividends at an average rate of 40 percent in both years,22 his two-year tax of $4.8 million would have produced a ProPublica true tax rate of about 15.5 percent — not terribly high, but perhaps not shocking low, and certainly much higher than the rates calculated by ProPublica for today’s billionaires.

In 1923 and 1924, dividends constituted nearly 40 percent of the total return (dividends plus appreciation) on the S&P 500. The picture is very different for 2011 through 2020. The S&P 500 index nearly tripled over that 10-year period, going from 1,282.62 to 3,793.75 — an increase of more than 195 percent, and an annually compounded rate of return of more than 11 percent.23 Over that same period, annual dividend yields hovered in a narrow range, between a low of 1.58 percent (2020) and a high of 2.2 percent (2012).24 Thus, dividends fell from almost 40 percent of total returns in 1923 and 1924 to below 20 percent over the past decade.

Given the volatility of S&P 500 index prices and the resulting sensitivity of dividends as a fraction of total return to the examined periods, it may be more instructive to focus simply on the dividend yields themselves. By that standard, too, dividends are much less significant today than they were in the 1920s. While annual dividend yields in the 2010s hovered around 2 percent, from 1921 to 1930 S&P 500 dividend yields ranged from a low of 3.67 percent (1928) to a high of 6.32 percent (1930).25

To sum up, in the 1920s taxing the dividends of the ultrarich was sufficient to produce effective tax burdens high enough to be politically acceptable, but the same approach today may not be sufficient for the same purpose. Although the change from taxing dividends at ordinary income rates to taxing them at long-term capital gains rates is part of the story, the more important part of the story is the declining significance of dividends to stock market investors in general, and especially to the ultrawealthy owners of stock of corporations that never pay dividends.

B. A Failure of Imagination?

In its June article, ProPublica assumed that the concept of unrealized appreciation as income was familiar enough — or at least plausible enough — to its readership that an analysis of effective tax rates premised on that concept would strike a chord. Time will tell whether ProPublica was correct.

A century earlier, the same assumption would clearly have been unfounded. At the dawn of the modern federal income tax, the great question was whether realized capital gains were taxable. When Congress enacted the individual income tax in 1913, the income tax of the United Kingdom — the leading model for Congress in designing the income tax — did not tax capital gains, even upon realization.26 The 1913 legislation featured language broad enough to include realized capital gains in the tax base; the act defined income as including “gains, profits, and income derived from . . . dealings in property, whether real or personal.”27 On the other hand, the statutory language said nothing about investment assets (as contrasted with inventory) in particular, and arguably gains from occasional sales of investments were not derived from “dealings.” Moreover, during the 1913 House debate on the legislation, Cordell Hull — then a young Democratic representative from Tennessee, and the primary drafter of the income tax provisions in the Ways and Means Committee bill — had told his colleagues that capital gains were not taxable under the bill (except in the unusual case of a taxpayer buying and selling an investment asset in the same year).28

After enactment of the 1913 income tax, Treasury consistently took the position that realized capital gains were taxable, although it did not promulgate a formal regulation to that effect until 1919.29 Some taxpayers strenuously disagreed with Treasury’s interpretation, and it was not until 1921 that the Supreme Court settled the issue by declaring that realized capital gains qualified as income for purposes of both the 16th Amendment and the income tax statute.30

In 1924, when the tax payments of the Rockefellers, Fords, Mellons, and their fellow plutocrats became public information, the taxability of realized capital gains had been settled for only three years. And, of course, the Supreme Court’s 1921 pronouncement did not convince everyone that realized capital gains should be understood and taxed as income. In a world in which treating realized capital gains as income was intensely controversial, treating unrealized capital gains as income was simply outside the bounds of argument and analysis — outside the Overton window, to use a 21st century term to describe the early 20th century situation.31 The idea that unrealized capital gains were income in an economic sense, and might reasonably be treated as income for tax purposes as well, would have been seen as too far-fetched to serve as the basis for a mid-1920s ProPublica-style analysis of the true tax rates of the Fords, Rockefellers, and Mellons.32

To be sure, sophisticated economists of a century ago understood that income could — and, some argued, should — be defined broadly enough to include unrealized appreciation. Writing in 1921, Robert M. Haig proposed an economic definition of income inclusive of unrealized appreciation: “Income is the money value of the net accretion to one’s economic power over two points of time.”33 Although Haig did not insist that the statutory definition of income follow the economic definition in this respect, he argued that any divergence of the legal definition from the economic definition would require a convincing policy justification. “The scientific economist in advising the legislator,” wrote Haig, might accept “that no tax be placed on a gain arising from the appreciation of a fixed asset until it is actually sold,” but only as a “concession . . . to the exigencies of a given situation.”34

As Haig’s essay indicates, a ProPublica-style analysis of Rockefeller and Ford true tax rates (with unrealized appreciation in the denominator) would not have been beyond the imagination of everyone in the 1920s. If some proto-ProPublica of the 1920s had published such an analysis, no doubt Haig would have read it with great interest. The point remains, however, that in terms of popular understandings of income, such an analysis would have been far ahead of its time — too far to have had any political impact.

The more sophisticated financial journalists of the 1920s were well aware that a ranking of the ultrawealthy by net worth would not perfectly coincide with a ranking by income tax payment amounts, and that much of the economic income of the wealthy was not subject to the income tax. For example, Chase, writing in The New York Times in 1927, acknowledged that “income taxes do not accurately reflect total wealth.”35 Two stories that had appeared in late 1924, one in the Times and one in the Journal, had delved deeper into the discrepancies between income tax payments, on one hand, and economic income and wealth on the other. The Times story quoted an unnamed banker:

One significant thing . . . is that the big income taxes are being paid by the “comers,” by the men who are today making America. The men with fortunes vastly greater, which were inherited, are paying only small taxes. They have their money in tax-exempt bonds, but the men of industry are getting theirs in cash and are giving about half of it back to the Government.36

The Journal story offered a more detailed analysis.37 Anticipating ProPublica by almost a century, the story noted that the release of the tax payment lists had “given rise to the assumption, and in some cases the outright assertion, that many people enjoying large incomes are tax dodgers.” The story went on, however, to offer four reasons why, without cheating on their taxes, “persons supposed, or known to be among the very rich, in some cases paid a smaller income tax than persons supposed only to be in comfortable circumstances”: (1) great wealth does not necessarily result in great income — economic or taxable — in any given year; (2) “a large number of chances to claim exemptions of various kinds”; (3) tax-exempt securities (municipal bonds); and (4) the special low (12.5 percent) top rate on capital gains. Of these four reasons, the story identified tax-free interest income from municipal bonds as by far the most important “means of holding down the amount paid to the government.” The story emphatically concluded that “the problem that worries students of the income tax is the tax-exempt security. Other avenues of escape can be closed, as the experience with the income tax shows the way. They will be decreased while the tax-exempt securities grow.”

With municipal bonds identified as the culprit — by both the unnamed banker in the Times article and the uncredited author of the Journal article — the nontaxation of unrealized appreciation enjoyed a free pass. Municipal bonds were a more appealing target of criticism than stock appreciation, for several reasons. Bond interest was received in cash, which the recipient was free to spend on any form of consumption; even if it could not constitutionally be taxed,38 it fit squarely within widely held understandings of the nature of income. Thinking of municipal bond interest as income involved none of the conceptual challenges posed by unrealized appreciation. Also, as the banker’s comment to the Times indicated, the paradigmatic recipient of tax-exempt interest was a member of the idle rich, of a younger generation than those whose labors had produced the family wealth. The paradigmatic owners of appreciated stock, by contrast, were either the founders of immensely successful corporations (Henry Ford) or, if of a younger generation (Edsel Ford and John D. Rockefeller Jr.), were themselves titans of industry rather than indolent coupon clippers. Finally — as discussed in more detail in the next section — the absolute size of the multimillion-dollar tax payments of the Fords and Rockefellers distracted observers from thinking of those payments in percentage-of-income terms. Second- and third-generation municipal bond coupon clippers, by contrast, might enjoy great wealth (even if not quite at the Ford-Rockefeller level) and large cash incomes while paying no income tax whatsoever.

C. Different Frames

Despite their low true tax rates, ProPublica’s 25 billionaires paid a total of $13.6 billion in income tax from 2014 through 2018 — an amount that ProPublica acknowledged was a “staggering sum.”39 Over those five years, the four billionaires in the ProPublica spotlight paid $23.7 million (Buffett), $292 million (Bloomberg), $455 million (Musk), and $973 million (Bezos).

Suppose ProPublica had not been so enterprising as to calculate true tax rates for a mark-to-market definition of income and instead had simply published a list comparable to the lists of 1924 and 1925 — that is, a list of the individuals making the largest tax payments from 2014 through 2018. According to the IRS Statistics of Income division, the 400 individual income tax returns with the largest adjusted gross incomes in 2014 (with an average AGI of $317,818,000) showed an average income tax liability of $73,513,000.40 Although the SOI’s data suggest that Buffett might have been embarrassed by his relatively small tax payments, Bloomberg and Musk would have made respectable showings among the top 400, and Bezos might have been at or near the top.

Because $973 million is a lot of money in absolute terms, absent ProPublica’s framing, the story would have been how much tax Bezos paid, rather than how little — just as the stories in the 1920s had been about how many dollars the government had received in Ford and Rockefeller tax payments, rather than how little the tax payments were as a percentage of wealth or economic income. For what it’s worth, the $7,435,189 tax paid by Rockefeller Jr. in 1924 (on 1923 income) translates to about $116 million in 2021 dollars, and the $6,278,000 paid by him in 1925 (on 1924 income) translates to about $98 million.41 In inflation-adjusted dollars, Bezos’s average annual tax payments from 2014 through 2018 were roughly twice Rockefeller Jr.’s average tax payments in 1924 and 1925. Yet the story in the 1920s was the enormousness of the Rockefeller payments, while the story in 2021 is the paltry size of the Bezos payments as a percentage of economic income. Framing is everything.

III. Conclusion

In early 1942 Irving Berlin contributed to the war effort by composing a song, “I Paid My Income Tax Today,” and donating the copyright to the federal government.42 The lyrics included the lines, “You see those bombers in the sky / Rockefeller helped to build them / So did I.” Twelve years earlier, “On the Sunny Side of the Street” (music by Jimmy McHugh, lyrics by Dorothy Fields), among the biggest popular song hits of the early Depression, featured the lines, “If I never have a cent / I’ll be rich as Rockefeller.” In the 1920s, and continuing through the Depression and World War II, Rockefeller was the songwriters’ obvious choice to represent both the biggest income taxpayers and the ultrawealthy. Today, names such as Bezos, Bloomberg, Buffett, and Musk could well replace Rockefeller in an updated version of “On the Sunny Side of the Street” (at least if the names happened to scan), but the same names might not work nearly as well as Rockefeller replacements in “I Paid My Income Tax Today.”

Rockefeller, Ford, and their fellow plutocrats of a century ago had taxable sources of income — especially dividends — in amounts at least somewhat reflective of their net worth and of the unrealized appreciation in their stock. That is not true of the American plutocrats of the 21st century. Absent a most-unlikely sea change in the dividend policies of the corporations whose shares are the source of the wealth and economic income of today’s multibillionaires, mark-to-market taxation may be the only way to achieve even the very rough congruence of the 1920s between lists of the payers of the highest income taxes and lists of the wealthiest Americans.


1 Jesse Eisinger, Jeff Ernsthausen, and Paul Kiel, “The Secret IRS Files: Troves of Never-Before-Seen Records Reveal How the Wealthiest Avoid Income Tax,” ProPublica, June 8, 2021. As to the source of the information, ProPublica said only that it was “not disclosing how it obtained the data, which was given to us in raw form, with no conditions or conclusions.” If an IRS employee supplied the information to ProPublica, that employee would have committed a felony under section 7213. Moreover, section 7213(a)(3) makes publishing that information a felony, although that prohibition may conflict with the First Amendment.

2 Wyden, “Treat Wealth Like Wages” (2019).

3 Jonathan Curry, “Treasury Official Doubles Down on Limited Capital Gains Reform,” Tax Notes Federal, June 14, 2021, p. 1798.

4 Revenue Act of 1924, section 257(b). For an excellent discussion of the political background of the provision, see Marjorie E. Kornhauser, “Shaping Public Opinion and the Law: How a ‘Common Man’ Campaign Ended a Rich Man’s Law,” 73 Law & Cont. Prob. 123, 126-129 (2010).

5 “Income Tax Returns Made Public; J.D. Rockefeller Jr. Paid $7,435,169; Ford Family and Company Pay $19,000,000,” The New York Times, Oct. 24, 1924.

6 “Recapitulation of Income Tax Payments Throughout Country Previously Reported,” The New York Times, Oct. 26, 1924.

7 There was one exception: The Journal published, perhaps inadvertently, a very short article noting that Henry Ford had paid $2,467,946 in tax in 1924 on his 1923 income. “Ford’s Income Tax,” The Wall Street Journal, Oct. 24, 1924.

8 United States v. Dickey, 268 U.S. 378 (1925).

9 Id. at 388.

10 Revenue Act of 1926, section 257(e) (providing for disclosure of the names and addresses of taxpayers, but without tax payment amounts or any other information). For the political background of the 1926 repeal, see Kornhauser, supra note 4, at 127-129.

11 For the details of that intriguing story, see id. at 129-145.

12 B.C. Forbes, “America’s 30 Richest Own $3,680,000,000,” Forbes, Mar. 2, 1918.

13 “List of Ten Richest Men in the World Includes Six Americans,” The New York Times, May 20, 1923. The story has no byline, but similarities of presentation and method with the 1927 Times article on the richest Americans suggests the 1923 article may also have been the work of Chase.

14 Chase, “Ford’s Billion Breaks World Records,” The New York Times, Feb. 13, 1927.

15 Inspector Gregory: “Is there any other point to which you would wish to draw my attention?” Sherlock Holmes: “To the curious incident of the dog in the night-time.” Gregory: “The dog did nothing in the night-time.” Holmes: “That was the curious incident.” Arthur Conan Doyle, “The Adventure of Silver Blaze,” The Complete Sherlock Holmes, Vol. 1, at 413, 415 (2003).

16 Chase, supra note 14.

17 Id.

18 Within the relevant time frame, the practice of the various Standard Oils (of California, New Jersey, New York, Ohio, etc.) was to pay “small regular dividends.” “Standard Oil Dividends,” The Wall Street Journal, Feb. 7, 1923.

20 “S&P 500 Dividend Yield by Year” (based on Shiller data).

21 “S&P 500 Historical Prices”(based on Shiller data).

22 The top marginal rate on ordinary income (including dividends) was 43.5 percent for 1923 and 46 percent for 1924. Urban-Brookings Tax Policy Center, “Historical Highest Marginal Income Tax Rates, 1913 to 2020” (Feb. 4, 2020).

23 “Historical Prices,” supra note 21.

24 “Dividend Yield by Year,” supra note 20.

25 Id.

26 The income tax of the United Kingdom did not feature a general tax on capital gains until 1965. Hugh J. Ault and Brian J. Arnold, Comparative Income Taxation: A Structural Analysis 123 (2004).

27 Revenue Act of 1913, Section II.B.

28 “As to an occasional purchase of real estate not by a dealer . . . this bill would only apply to profits on sales where the land was purchased and sold during the same year.” 50 Cong. Rec. 506 (Apr. 26, 1913). For a detailed discussion of Hull’s several inconsistent statements during the 1913 House debate concerning the income tax treatment of capital gains, see Lawrence Zelenak, Figuring Out the Tax: Congress, Treasury, and the Design of the Early Modern Income Tax 13-16 (2018).

29 Reg. 45, art. 21 (T.D. 2831).

30 Merchants Loan & Trust Co. v. Smietanka, 255 U.S. 509 (1921). For much more on the early disputes over the taxability of capital gains, see Kornhauser, “The Origins of Capital Gains Taxation: What’s Law Got to Do With It?” 39 Sw. L.J. 869 (1985). As Kornhauser explains in detail, at the conceptual level, the dispute was between proponents of the res theory of capital (under which even realized capital gains were viewed as belonging to capital rather than as income) and proponents of the quantum theory of capital (under which “capital consisted of a money value equal to the original cost” of an investment, with the implication that the sale of an asset for more than its cost produced income). Id. at 888.

31 The Overton window describes the range of policy options that are politically possible at any given time. The concept was first described by, and is now named after, Joseph P. Overton, an executive of a conservative think tank in Michigan, who introduced the idea in the 1990s. Maggie Astor, “How the Politically Unthinkable Can Become Mainstream,” The New York Times, Feb. 26, 2019.

32 In Eisner v. Macomber, 252 U.S. 189 (1920), the Supreme Court famously rejected Congress’s attempt to include stock dividends in the base of the income tax. The Court interpreted the term “income” in the 16th Amendment to include a realization requirement, and it held that stock dividends did not constitute realization events. As long as the Court adhered to that position, and barring another constitutional amendment, taxing unrealized appreciation of any sort would not have been a legally available policy option for Congress. The crucial point for present purposes, however, is that even without a constitutional restriction, no Congress of the 1920s would have had the remotest interest in taxing simple unrealized appreciation, such as the increase in value of stock during a taxpayer’s continued ownership.

33 Haig, “The Concept of Income — Economic and Legal Aspects,” The Federal Income Tax 1, 7 (1921).

34 Id. at 14.

35 Chase, supra note 14. Not every financial journalist seems to have recognized this point. In 1918, when Forbes published his first estimates of the net worths of the 30 richest Americans, he also estimated the incomes of each. Forbes, supra note 12. Although the article is not explicit on this point, it appears that each income estimate is simply 5 percent of each net worth estimate. Forbes did not describe these as estimates of taxable income, but even taking them as estimates of economic income, the approach was remarkably crude.

36 “Wall Street Buzzes Over Tax Publicity,” The New York Times, Oct. 26, 1924.

37 “Tax Evasion Cry Misleading: Public Assumption of Wealth Generally Wrong; Evil of Tax-Exempt Securities Proved,” The Wall Street Journal, Nov. 10, 2014.

38 At least that was the conventional wisdom of the time. The Supreme Court ruled otherwise, many years later, in South Carolina v. Baker, 485 U.S. 505 (1988) (income taxation of municipal bond interest income was not constitutionally prohibited; taxation of interest income from bonds not in registered form upheld).

39 Eisinger, Ernsthausen, and Kiel, supra note 1.

41 U.S. Bureau of Labor Statistics, “CPI Inflation Calculator.”

42 “Berlin Writes Song for Treasury, ‘I Paid My Income Tax Today,’” The New York Times, Jan. 26, 1942.


Copy RID