How might one define the super-rich and how might the government tax them? Florian Scheuer tackles these questions in "Taxing the superrich: Challenges of a fair tax system" (UBS Center Public Paper #9, November 2020). Also available at the the website is a one-hour video webinar by Scheuer on the subject. Those who want a more detailed technical overview might turn to the article by Scheuer and Joel Slemrod, Taxation and the Superrich," in the 2020 Annual Review of Economics (vol. 12, pp. 189-211, subscription required).
When discussing the superrich in a US context, there are two common starting points. One is to focus on the Forbes 400, an annual list of the 400 wealthiest Americans. Another is to focus on the very top of the income distribution--that is, not just the top 1%. but the top 0.1% or even the top 0.01%.
On the subject of the Forbes 400, Scheuer writes: "The cutoff to make it into the Forbes 400 in 2018 was a net worth of $2.1 billion, and the average wealth in this group was $7.2 billion. The share of aggregate U.S. wealth owned by the Forbes 400 has increased from less than 1% in 1982 to more than 3% in 2018." It's worth pausing over that number for a moment: the share of total US wealth held by the top 400 has tripled since 1982. Scheuer also points out that one can distinguish whether those in top 400 inherited their wealth or accumulated it themselves. Back in 1982, 44% of the top 400 had accumulate it themselves, while in 2018, 69% had done so.
Of course, wealth is not the same as income. For example, when the value of your home rises, you have greater wealth, even if your annual income hasn't changed. Similarly, when the price of stock in Amazon or Microsoft changes, so does the wealth of Jeff Bezos and Bill Gates (#1 and #2 on the Forbes wealth list), even if their annual income is unchanged.
The IRS used to (up to 2014) release data on the "Fortunate 400" top income-earners in a given year; in 2014, the cutoff for making this list was $124 million in income for that year. Another approach is to looking at the top of income distribution. the top 0.01% represents the 12,000 or so households with the highest income in the previous year.
There are basically four ways to tax the super-rich: income tax, capital gains taxes, the estate tax, or a wealth tax.
In the 2020 tax code, the top income tax bracket is 37%: for example, if you are married filing jointly, you pay a tax rate of 37% on income above $622,500. (This is oversimplified, because there are phase-outs of various tax provisions and surtaxes on investment income that can lead to a marginal tax rate that is a few percentage points higher.) But one obvious possibility for taxing the superrich would be to add additional higher tax brackets that kicks in a higher income levels, like $1 million or $10 million in annual income.
The difficulty with this straightforward approach is what Scheuer refers to as the "plasticity" of income, that is, "the ease with which higher-taxed income can be converted into lower-taxed income." Scheuer writes:
Plasticity is an issue when different kinds of income are subject to different effective tax rates. By far the most important aspect of plasticity, with implications both for understanding the effective tax burden on the superrich and for measuring the extent of their income and therefore income inequality, concerns capital gains.
To put this in concrete terms, if you look at the wealthiest Americans like Jeff Bezos or Bill Gates, their wealth doesn't rise over time because they save a lot out of the high wages they are paid each year; instead, it's because the stock price of Amazon or Microsoft rises. They only pay tax on that gain if they sell stock, and receive a capital gain at that time. Thus, if you want to tax the super-rich, taxing their annual income will miss the point. You need to think about how to tax the accumulation of their wealth
In the US, taxes on capital gains have several advantages over regular income. The tax rate on capital gains is 20%, instead of the 37% (plus add-ons) top income tax rate. In addition, you can let a capital gain build up for years or decades before you realize the gain and owe the tax; thus, along with the lower tax rate there is a benefit from being able to defer the tax. Finally, if someone who has experienced a capital gain over time dies, and then leaves that asset to their heirs, the capital gain for that asset during their lifetime is not taxed at all. Instead, the heir who receives the asset can "step up": the basis, meaning that the value for purposes of calculating a capital gain for the heir starts from the value at the time the asset was received by the heir. Taken together, the "plasticity" of being able to gain wealth by a capital gain, rather than by annual income, is a core problem of taxing the superrich. Scheuer explains:
Most countries’ tax systems treat capital gains favorably relative to ordinary labor income (Switzerland being an extreme case where most capital gains are untaxed). Realized capital gains represent a very high fraction of the reported income of the superrich. For example, realized capital gains represented 60% of total gross income for the 400 highest-income Americans in tax year 2014. ... For tax year 2016, those earning more than $10 million report net capital gains corresponding to 46% of their total income, whereas capital gains are a negligible fraction of income for those earning less than $200 k.There are other ways to tax capital gains. For example, one of Joe Biden's campaign promised was to tax capital gains income at the same rate as personal income for anyone receiving more than $1 million in income in that year. Before getting into some of the reasons, it's worth noting that every high-income country taxes capital gains at a lower rate. Scheuer writes:
Five OECD countries levy no tax on shareholders based on capital gains (Switzerland being a prominent example). Of those that do, all tax is on realization rather than on accrual. Five more countries apply no tax after the end of a holding period test, while four others apply a more favorable rate afterwards. The tax rate varies widely with the highest as of 2016 being Finland, at 34%. With a few exceptions, the accrued gains on assets in a decedent’s estate escape income taxation entirely, because the heir can treat the basis for tax purposes as the value upon inheritance.
Why is capital gains taxed at a lower rate, all around the world? Why is it taxed only when those gains are realized, perhaps after years or decades, rather than taxed as the gains happen? One reason is that there is an annual corporate tax, so income earned by the corporation is already being taxed. Or if the capital gain is being realized on a gain in property values, there were also property taxes paid over time. In general, many countries want to have a substantial share of patient investors, who are willing to hold assets for a sustained time. Trying to tax capital gains as they happen, rather than when they are realized, would also raise practical questions--for example it might require people to sell some of their assets to pay their annual taxes.
Scheuer runs through a variety of different ways of ways of taxing capital gains, and you can consider the alternatives. But again, there are reasons why no country has pursued taxing capital gains as they accrue, rather than as they are realized, and why no country taxes such gains as ordinary income--and in fact why some countries don't tax them at all.
Another alternative is to tax wealth directly. I've written about a wealth tax before, and don't have a lot to add here. Scheuer offers the reminder that Donald Trump was an advocate of a large but one-time wealth tax on high net-worth individuals back in 1999, when he ran for president on the Reform Party ticket, as a way to pay off the national debt. Here, I'll just offer a reminder that a wealth tax is based on total wealth, not on gains. Thus, if there is a an annual wealth tax of, say, 3%, then if your wealth was earning a return of 3% per year, the wealth tax means you are now earning a return of zero. If there is a year where the stock market drops, and the returns for that year are negative, you still owe the wealth tax.
About 30 years ago, 12 high-income counties had wealth taxes, but the total is now down to three. The general consensus was that the troubles of trying to value wealth each year for tax purposes (and just consider for a moment how the superrich might shuffle their assets into other forms to avoid such a tax), just wasn't worth the relatively modest total amounts being collected. The one country that continues to collect a substantial amount through its wealth tax is Switzerland--but remember, Switzerland doesn't have any tax at all on capital gains. Scheuer writes:
So far, the Swiss case is the only modern example for a wealth tax in an OECD country that has been able to generate sizeable and stable revenues in the long run. It enjoys broad support, as evidenced by the fact that it keeps being reaffirmed by citizens in Switzerland’s direct democracy, where most tax decisions must be put directly to voters. However, its design and the role it plays in the overall tax system are quite different from current proposals in the United States. In particular, it is not geared towards a major redistribution of wealth, and indeed wealth concentration in Switzerland remains high in international comparison.
A final option, which is not a focus of Scheuer's discussion, would be to resuscitate the estate tax: that is, instead of taxing the superrich during their lives, tax the accumulated value of their assets at death. For an example of a proposal along these lines, William G. Gale, Christopher Pulliam, John Sabelhaus, and Isabel V. Sawhill offer a short report of "Taxing wealth transfers through an expanded estate tax" (Brookings Institution, August 4, 2020). They point out, for example, that back in 2001 estates of more than $675,000 were subject to the estate tax; now, it applies only to estates above $11.5 million. Maybe $675,000 was on the low side, but an exemption of $11.5 million is pretty high--only about 0,2% of estates are subject to the estate tax at all. They calculate that rolling back the estate tax rules to 2004--which was hardly a time of confiscatory taxation--could raise about $100 billion per year in revenue.
Taking all this together, it seems to me that a middle-of-the-road answer on how to raise taxes on the superrich would focus in part on the estate tax, and in part on the capital gains tax--and perhaps in particular on limiting the ability to pass wealth between generations in a way that avoids capital gains taxes.