Tax Notes logo

Economic Analysis: Risking the Wrath of 900 Billionaires

Posted on Oct. 21, 2019

You can safely ignore the possibility of a U.S. wealth tax. But you should not ignore the political and economic developments that are putting it in the news.

About the impossibility of a wealth tax: First, as is common knowledge, there may be at least even odds of Democrats winning the White House and keeping the House, but their taking the Senate is a long shot. And even in the unlikely event of a Democratic hat trick, the Dems would have to use the budget reconciliation process or invoke the nuclear option (repealing the Senate rule requiring 60 votes to proceed) to enact any tax increase over Republican objections. Those are some high hurdles. (Perhaps the Democrats’ chances improve in 2022, but we’ll wait on that.)

Second, and more fundamentally, despite the wealth tax being the shiny object for those wanting to tax the rich, why should Dems put all their chips on a tax that could easily fall short of hyped expectations? Reforms of existing taxes can achieve similar objectives. On the estate tax: Raise the rates, get rid of the valuation allowance for family limited partnerships, and adopt a half-dozen other estate tax reforms drafted at Treasury during the Obama administration. On the income tax: Raise the capital gains and dividend tax rates, raise the regular income tax rates at the top end, and — to effectively tax the superrich — tax capital gains on an accrual basis or repeal the step-up in basis at death. Lawmakers have a record of starting with a bold, complicated reform like a wealth tax and settling for something less grandiose. For example, in 1993 what started as a Btu tax ended up as a simple hike in the gas tax. What started in 2004 as an intricate corporate integration plan ended up as just a cut in the dividend and capital gains rate.

Part 1: A Provocative Book

Now let’s move on to why the current conversation about a wealth tax is prevalent, even if it has about as much chance as Tulsi Gabbard becoming president. At the center of the debate is a new book by professors Emmanuel Saez and Gabriel Zucman, both at the University of California at Berkeley. This 200-pager reminds us why they call it “political economy.” Saez and Zucman are economists who are advising presidential candidates. And now with their book, they are targeting the general public on a white-hot topic (“tax injustice”) as we move into what promises to be a nasty election year. Their message: We need wealth taxation, and it is doable (Saez and Zucman, The Triumph of Injustice (2019)). (To clarify, because it is the book’s subject matter, the word “Tax” should have been inserted before “Injustice,” but that probably wouldn’t have helped sales.)

The authors deserve much praise for driving home what are (as far as we can tell) two indisputable facts. Number one: Over the past few decades, the superrich are getting richer at the same time working- and middle-class incomes have stagnated. Number two: The U.S. tax system, so proficient at taxing workers’ wages, does a lousy job of taxing the privileged few. These are developments that we need to let sink into our policy and political consciousness. And as you contemplate our fiscal future, it would probably be wise to expect these 600-pound shortcomings of our tax system to no longer be swept under the rug.

Saez and Zucman argue that the United States used to excel — perhaps was the best in the world — at taxing the wealthy. Rooseveltian ideals and policies that evolved during the New Deal made this possible. But there has been a decline in corporate and individual tax rates, which the authors attribute to politicians who have defunded and defamed rather than defended the IRS. In their lawmaking, politicians have bowed to pressure to provide loopholes at the top that are overtly justified by economic reasoning. That reasoning includes one prominent economic idea that maximum economic growth requires zero taxation of capital and the belief that excess taxation is fruitless from a revenue perspective because wealth is inherently too slippery for any tax collector.

Making numerous debatable assumptions, Saez and Zucman estimate, contrary to most presentations from official and private sources, that the U.S. tax system is not progressive but essentially flat, with lower rates for the superrich. Take a good look at Figure 1. In official estimates from Treasury and the Joint Committee on Taxation, the extra-low rates paid by the superrich are hidden because they are lumped together with the not-so-superrich who do pay high rates.

Figure 1. Average Tax Rate by Income Category in 2018

There is a lot of unglamorous work that goes into putting together a figure like this. Unlike most other estimators of the tax burden distribution, Saez and Zucman include state and local taxes (which are more regressive than federal taxes), include a significantly broader measure of income in the denominator of their tax rates, and are highly transparent about their assumptions and methods (which includes providing much of the underlying data).

Of course, the boatload of assumptions they had to make will be reflexively disputed by the political right. But the authors can expect a lot of nonpolitical academic criticism as well. For example, on the incidence of the corporate tax, one of the most difficult and politically sensitive in tax economics, they assume all the burden falls on shareholders while most other economists assume the burden is shared by all capital (not just shareholders) and labor.

Why tax wealth? If you believe in ability-to-pay taxation, wealth is a far better measure of economic well-being than income. For example, assume two recent law school graduates — both making $100,000 — who are more or less taxed the same under the income tax. But suppose one has $250,000 of debt and the other just inherited $10 million from Grandpa. Does not the trust fund baby have considerably more ability to pay than the debtor?

And while income is unevenly distributed across income classes, wealth is even more unevenly distributed across income. The top 1 percent of wealth holders account for about 40 percent of national wealth, but the top 1 percent of income earners account for about only 20 percent of national income. So if you really want to address progressivity, an income tax — even a theoretically perfect version — is a highly imprecise instrument. Most importantly, a wealth tax imposes burden on inherited wealth, while an income tax can leave it untouched. (Ironically, a consumption tax — generally considered highly regressive — is better at taxing inherited wealth than an income tax.)

Now let’s delve more deeply into the political side of political economy. Not unreasonably, Saez and Zucman argue that concentrations of wealth give rise to concentrations of power. To what degree this is a problem is certainly open to debate. Saez and Zucman assert it is an existential threat to democracy. The effect of tax on political power is something that rarely, if ever, enters into modern tax policy discussions. Given that members of Congress will jump for four-figure political contributions like puppies jump for biscuits, the lopsided political influence of individuals with 10-figure wealth certainly deserves attention. But is taking a modest bite from the wealth of America’s 900 billionaires the best way to address the issue?

And, of course, another reason for a wealth tax is that the government can use the money. Presidential candidate Sen. Elizabeth Warren, D-Mass., is proposing an “ultra-millionaire tax” on household net worth in excess of $50 million. The rate, applied annually, is a 2 percent tax on every dollar of net worth above $50 million and a 3 percent tax on every dollar of net worth above $1 billion. Saez and Zucman estimate the tax would apply to 75,000 households and raise $2.75 trillion over a 10-year period. Not to be outdone, Sen. Bernie Sanders, I-Vt., is proposing a “tax on extreme wealth.” It has a graduated rate structure that starts at 1 percent for a married couple with $32.5 million and rises stepwise to 8 percent for wealth that exceeds $10 billion. It is estimated to apply to 180,000 households and would raise $4.35 trillion over the next decade.

“Assume a can opener” is the punchline to one of the oldest jokes about economists. In a similar vein, when it comes to revenue estimating, Saez and Zucman assume compliance with this entirely new tax will be a minor issue. They assume Congress will write legislation that will tax wealth comprehensively (as we might hope Congress would do if it were proposing an income tax for the first time). They assume the IRS will be well funded and supported by a “public protection bureau” that will regulate the tax-dodging industry in which, according to the authors, the Big Four figure prominently. They assume international cooperation among developed nations and sanctions against tax havens. They assume non-traded private businesses and unique works of art can be valued. Unfortunately for proponents of wealth taxation, some recent developments are inhospitable to wealth tax enforcement. Rising nationalism erodes the institutions that foster international cooperation. Low and negative interest rates encourage wealth to leave financial institutions and increase demand for cash, precious metals, and other hard-to-police stores of wealth.

There is no doubt that more can and should be done to improve enforcement against evasion and aggressive tax avoidance. Saez and Zucman’s thoughtful suggestions deserve consideration. But ivory-tower optimism should not be built into revenue estimates of legislation made in a sausage factory. Former Treasury Secretary Lawrence H. Summers and coauthor Natasha Sarin doubt the size of Saez and Zucman’s estimates. Their own estimates, based on U.S. experience with the estate tax, are between one-half and one-tenth of those by Saez and Zucman (Summers and Sarin, “Be Very Skeptical About How Much Revenue Elizabeth Warren’s Wealth Tax Could Generate,” The Washington Post, June 28, 2019).

Part 2: Do the Rich Pay 23 Percent or 1 Percent?

One conclusion of The Triumph of Injustice that is highlighted by the authors and is attracting a lot of attention is that “for the first time, billionaires pay lower tax rates than the working-class.” That specific result depends on the last data point in both figures 1 and 2, which show that the top 400 wealthiest Americans on average paid 23 percent of their income in taxes in 2018. As explained below, there is good reason to suspect the Saez-Zucman estimate of 23 percent is too high and therefore that the U.S. tax system is now and has been in the past less progressive than they report. (Yes, you read that right: We are saying these left-leaning economists’ estimates are too conservative.)

Figure 2. Average Tax Rate of 400 Richest Americans vs. the Bottom 50% of Income Earners, 1960-2018

The text of the book tells us that the income measure used by the authors is a broad measure from the Department of Commerce National Income and Product Accounts known as “national income.” (In the aggregate, national income is the same order of magnitude as GDP. But national income, unlike GDP, excludes depreciation. Another difference is that it includes net income inflows from foreign sources.) The fact that is critical for evaluating the authors’ measured tax burden of the superrich is that national income (like GDP) excludes income in the form of capital gains (both realized and unrealized). It does, however, include undistributed corporate profits that account for a portion of capital gains on equities.

We all know taxable income and even adjusted gross income can be far less than true economic income. Although it gives economists no pleasure to discomfit lawyers, we must insist that true economic income includes unrealized capital gains. (See, for example, Henry C. Simons, Personal Income Taxation (1938); and Treasury, “Blueprints for Basic Tax Reform” (1977).) To best portray any taxpayer’s income tax burden, the denominator of an effective tax rate should as much as possible approximate economic income. Without adjusting national income upward to include unrealized capital gains, the denominator of an effective tax rate would be too small, and so the estimated effective tax rates would be biased upward. Because the vast bulk of Americans own small amounts of stock, this bias is trivial for most. But for the wealthy, it could be large.

The online technical appendix to their book explains how Saez and Zucman adjust national income to account for unrealized capital gains. Because national income already includes corporate retained earnings, they need no adjustment for gains attributable to that component of unrealized gains. For the critical remaining component, they assume that in any year in which realized capital gains exceed 3 percent of national income (their proxy for corporate retained earnings), that excess is economic income that should be added to national income, which is the denominator in the calculation of effective tax rates.

This estimate of unrealized capital gains seems low. In years that gains exceed 3 percent of national income, unrealized gains are in effect assumed to be zero because total gains are assumed to be equal to realized gains. In years that the 3 percent threshold is not reached, total gains are assumed to be less than 3 percent of national income. Given that national income is roughly two-thirds of the value of corporate stock, this would be the equivalent of assuming average stock market gains of about 2 percent (and 0 percent for all capital gains on assets that are not equities).

But you don’t have to follow all this mumbo-jumbo. Saez and Zucman themselves provide an excellent example that makes the same point as the prior paragraph. As reported in the text of the book (at 129), Warren Buffett reportedly paid federal income taxes of $1.8 million in 2015. In the same year, Buffett’s reported wealth was $65.3 billion. Using a 5 percent rate of return, Saez and Zucman estimate Buffett’s true economic income was $3.2 billion, and then use this figure in the denominator to calculate an effective tax rate for Buffett of one-twentieth of 1 percent.

If the experience of billionaires with surnames like Walton, Bezos, Gates, Zuckerberg, and other members of the top 400 is similar to Buffett’s, it is likely that the 23 percent figure estimated by Saez and Zucman is too high and the tax system is less progressive at the high end than they report. Income giving rise to astronomical accumulations of individual wealth is routinely taxed at rates in the single digits.

Copy RID