Who Pays What in Taxes? The Answer Isn’t So Simple.


WASHINGTON—It is a tax policy question with a seemingly simple answer: Who pays what?

In reality, how taxes are distributed across income groups is hotly debated by economists—with enormous policy consequences. New findings by economists

Emmanuel Saez


Gabriel Zucman

of the University of California, Berkeley, have added fuel to this debate, challenging the conventional view that the poor pay relatively little and that effective tax rates rise with income.

In their new book, “The Triumph of Injustice: How the Rich Dodge Taxes and How to Make Them Pay,” Messrs. Saez and Zucman find that tax rates, including state and local taxes, are roughly flat, ranging between 24% and 30% across most income groups. Then, they focus on the 400 highest-income adults and find a 23% tax rate, lower than any other group for the first time.

That conclusion challenges the view espoused by lawmakers and tax analysts affiliated with both parties that the U.S. has a progressive tax system. According to the Congressional Budget Office, the top 1% of households paid an average federal tax rate of 33% while the bottom one-fifth has a federal tax rate below 2%. Estimates including state and local taxes show somewhat higher taxes at the bottom.

Messrs. Saez and Zucman use their findings to prescribe sharply higher taxes on the rich, including an annual wealth tax, global coordination to raise corporate taxes and tougher enforcement. Democratic presidential candidates

Bernie Sanders


Elizabeth Warren

have endorsed a wealth tax, and both rely on Saez-Zucman estimates.

But the conclusions of Messrs. Saez and Zucman have been met with skepticism from some other economists who differ on how to assess and describe tax burdens. They split on what should count as income and taxes and on how varying techniques fill gaps in publicly available data.

“There is a very wide range where the truth may lie,” said Wojciech Kopczuk, a Columbia University economist who has been critical of the Saez-Zucman approach.

For those at the top of the income scale, Messrs. Saez and Zucman assume all corporate taxes are paid by shareholders. Because corporate ownership is concentrated among the wealthy, the 2017 tax law, with its lower rates, caused tax burdens for the top 400 adults to plummet in 2018. Some of those effects may be temporary as the law’s effects play out.

Government models from the Treasury Department and Joint Committee on Taxation, by contrast, assume that some corporate taxes are in effect passed on to workers through lower wages. Mr. Zucman said that approach works for measuring effects of potential tax-law changes but not for describing who pays what in taxes.

Messrs. Saez and Zucman also differ in how they estimate who receives income, such as retained corporate earnings, that shows up in national statistics but not on tax returns or other measures. They attribute more of that to high-income households than others do, increasing that group’s income and thus lowering its tax rate.

Some of the bigger disagreements among economists are over tax rates—measured as taxes paid divided by income—for people at the bottom end of the economic spectrum.

The lowest-income people pay sales and payroll taxes but generally don’t pay federal income taxes. They receive refundable tax credits—cash payments delivered through the income-tax system.

Many analysts subtract those payments, such as the earned-income tax credit, from total taxes when calculating tax burdens. They are sometimes counted as income, reflecting the idea that they are monetary resources available to households.

Messrs. Saez and Zucman, however, look at taxes as a share of pretax income. They consider those credits government transfer payments outside the tax system. They don’t count them as income, even though households may use the funds to purchase goods and pay sales taxes that show up as part of their tax burden.


Should the U.S. tax system move toward a flat tax or a more progressive tax? Why? Join the conversation below.

Consider someone who paid $2,000 of payroll taxes and received a refundable income-tax credit of $1,500. In more conventional analysis, that person has a $500 net tax burden. In the Saez-Zucman framework, that person has a $2,000 tax burden and a $1,500 transfer that doesn’t show up in the tax-rate calculation.

As a result, the Saez-Zucman method makes the effective tax rates on lower-income households look higher than in more conventional approaches.

Including those credits in income would make total income across the population higher than 100% of national income, Mr. Zucman said. And subtracting them from taxes is arbitrary, he said, because the same analysis could apply to other government payments, such as food stamps.

David Splinter, an economist with Congress’s Joint Committee on Taxation, counts the earned-income tax credit as income but doesn’t use the sum to reduce the net tax burden in his estimates. In a public response last week to Messrs. Saez and Zucman, he estimated the tax rate on lower-income households at 13%, compared with the Saez-Zucman method’s 24%.

The eye-catching Saez-Zucman finding of relatively flat tax rates across the income spectrum doesn’t present the full picture of how the U.S. redistributes income through taxes and spending. Lower income households benefit directly from food stamps, refundable tax credits and other assistance, though the wealthiest reap more intangible gains from patent enforcement and a legal system that overall protects holders of property and assets.

Messrs. Saez and Zucman’s “book is best appreciated as a window into the burdens imposed at the very top end and there, of course, the book is devastating” in showing low tax rates, said Ed Kleinbard, a law professor at the University of Southern California. “It’s simply not asking or answering the separate question, which is: What is the net well-being of America’s lowest-income” people?

Write to Richard Rubin at [email protected]

Copyright ©2019 Dow Jones & Company, Inc. All Rights Reserved. 87990cbe856818d5eddac44c7b1cdeb8

Source link