The US tax code is highly progressive. Official US Treasury estimates show that the top 10 percent of income earners pay more than 60 percent of all federal taxes and 72 percent of income taxes. Yet some political rhetoric may lead many Americans to believe that the wealthiest taxpayers pay tax rates in the single digits or otherwise sidestep tax laws that apply to everyone else. New research shows that billionaires pay tax rates at least 13 percentage points higher than those of ordinary households.
However, measuring how much tax the wealthiest Americans actually pay is not as simple as it may seem. Depending on the definitions and assumptions, estimates of the effective tax rate for the Forbes 400 range from 38 percent to as high as 73 percent, when including charitable giving. Recent estimates that claim billionaires pay lower tax rates than the average American rest on flawed methods that, when corrected, show the US tax system remains highly progressive, which means as income rises, so do tax rates.
The wide range of reported effective tax rates for America’s wealthiest taxpayers illustrates how sensitive the analysis is to the researchers’ assumptions. Understanding why these estimates diverge is essential to understanding the broader debate over tax fairness.
Correcting Recent Estimates
A new National Bureau of Economic Research working paper by Akcan Balkir, Emmanuel Saez, Danny Yagan, and Gabriel Zucman (BSYZ) matches the Forbes 400 list of wealthiest Americans with individual, business, estate, and gift tax returns, including federal, state, local, and international taxes. The benchmark estimate finds that between 2018 and 2020, the wealthiest 0.0002 percent of Americans (367 taxpayers) paid an effective tax rate of about 24 percent of their annual economic income, six points lower than the estimated 30 percent average for the overall population.
In a detailed critique, economist David Splinter argues the BSYZ estimate is still likely 13 percentage points too low. The researchers double-count income, fail to account for dynastic family wealth spread across multiple tax returns, and miss important taxes. After making adjustments, Splinter finds the effective tax rate for the wealthiest 0.0002 percent (hereafter Forbes 400) is 38 percent.
His corrections include three major steps:
The original estimate double-counts some corporate profits by imputing a measure of unrealized “pure” capital gains in addition to dividends and realized gains. Excluding “pure” gains raises the Forbes 400 effective tax rate by 3 percentage points.
Many Forbes 400 fortunes are spread across multiple tax returns (siblings, children, etc.). Correctly distributing dynastic wealth to a more accurate subset of tax returns reduces the denominator estimate of wealth and income for the individual tax units and pushes effective rates up by 8 points.
Following tax law changes in 2017, federal administrative tax data appear to underreport state and local taxes paid by 57 percent. Adding these likely missing taxes raises the effective tax rate by 3 points.
Splinter also argues that the reported average tax rate for the entire population is overstated. Following standard practice, Splinter counts transfers (e.g., Social Security and unemployment benefits) as income and subtracts refundable credits from taxes. With these adjustments, the average effective tax rate for the population falls from 30 percent to 25 percent.
Defining the Tax Base
Any estimate of an effective tax rate depends on two choices: what counts as taxes paid (the numerator) and what counts as income earned (the denominator). Splinter’s critique highlights how alternative definitions substantially change the results. But other judgment calls remain, such as whether to count philanthropy, how to assign corporate tax incidence, and whether to measure income annually or over a lifetime. Each choice leads to different answers.
BSYZ adopt a broad measure of “economic income” in their denominator, imputing business profits to owners each year, regardless of whether they are distributed. They also allocate 100 percent of corporate taxes to shareholders to keep the underlying accounting identities balanced, but this decision ignores the unsettled question of who really bears the economic burden of the corporate tax. Empirical studies generally find that workers bear at least half, and possibly all, of the corporate tax burden (government sources assume workers bear 25 percent). Dropping corporate taxes (but not capital gains and dividends) and the associated imputed income yields an uncorrected effective tax rate of 30 percent (using data from an earlier BSYZ study). Applying Splinter-style corrections of about 11 percentage points implies a rate closer to 41 percent for the Forbes 400.
Including philanthropy in the numerator also dramatically raises billionaire contribution rates. The US tax code incentivizes charitable giving in lieu of taxpaying on the theory that private philanthropy is often better at deploying resources to help people than the government, and Americans seem to agree, given their pattern of voluntary giving to private charity over the US Treasury. The Forbes 400 give about 11 percent of their annual economic income to charity and donate 47 percent of their estate wealth at death. Splinter estimates that adding annual charitable giving raises the billionaire “tax-and-giving rate” to 59 percent, and including end-of-life bequests pushes it as high as 73 percent.
Lastly, point-in-time measures can be misleading for ultra-wealthy households whose income, wealth, and expenses fluctuate significantly, as illustrated in the difference between annual and lifetime giving. BSYZ find that top effective rates fell by six points after the 2017 tax cuts. Some of this decline is expected due to tax rate reductions. But some policies, like full expensing for businesses, only temporarily depress business taxes, and over time, these timing effects reverse, and future tax payments rise (the BSYZ definition of economic income in the denominator is not affected by the same timing issues). A lifetime perspective that more accurately accounts for realization, estate taxes, and end-of-life giving likely paints a more accurate picture of high-income tax burdens.
Conclusion
The debate over high income tax rates is less about arithmetic and more about the definitions researchers choose and how they interpret imperfect data. By just about any measure, the US tax system is highly progressive, with rates rising from single digits for the lowest-income Americans to between 38 and 73 percent for the wealthiest taxpayers. The headline from a recent study that billionaires pay lower tax rates than average Americans makes several errors. After appropriate adjustments, the data show that billionaires pay tax rates significantly higher than those of most, if not all, other Americans.
