|
Is the tax code sufficiently progressive? The answer depends not only on the values of the person answering the question, but it also is a surprisingly tricky empirical exercise. What counts as income and who pays the corporate tax? How should one measure households?
These details are at the core of an ongoing debate about economic inequality and the tax system. One camp relies on research by economists Thomas Piketty, Emmanuel Saez, and Gabriel Zucman, which indicates that the share of national income attributable to households at the top has surged, and that tax policy changes have magnified these gains. Careful research by economists Gerald Auten and David Splinter tells a very different story. Using largely the same data sources but with seemingly minor adjustments, Auten and Splinter’s results show that the long-term trend in the share of national income going to very high earners is actually remarkably flat.
Some of the adjustments are common sense, like adjusting the data for children who file tax returns, changes in marriage rates among different income populations, and variations in household size. For example, Auten and Splinter recognize that a college student with a part-time job who is claimed as a dependent by her parents is not a poor household by herself, but a part of her parents’ household. Other adjustments concern how unreported income is distributed across the income distribution, how health insurance and retirement savings and distributions are measured, and what fraction of taxes paid by corporations is actually borne by workers rather than by capital owners. Other adjustments are more technical, such as replacing realized capital gains data with corporate retained earnings as a better way of distributing corporate profits among households.
Taken together, these adjustments to the measurement of national income matter a great deal. The Picketty-Saez-Zucman distribution of national income makes it seem that average tax rates on high-income earners are falling. We find the opposite result when we analyze the Auten-Splinter data.
Each frame in Figure 1 below illustrates taxes — federal, state, and local — as a share of income for the highest-earning households from 1960 to 2015. Auten and Splinter’s more comprehensive measure of income and careful measure of the tax burden illustrate that effective tax rates on the top 10 percent, top 1 percent, and even top 0.01 percent of households slowly increased from 1960 to 2015. (The data do not show the tax burdens of the top 400 households, which is the focus of recent research by Saez and Zucman, but the top 0.01 percent of households in the last frame below represents roughly 12 thousand such units).
Figure 1. Average Effective Tax Rates, 1960-2015
This debate over the progressivity of the tax code is of great importance for our country right now. Progressive and populist rhetoric attack a perceived increase in the power and wealth of the very few at the expense of the well-being of the middle class. And it would be a significant concern if the tax code were becoming more biased in favor of the super–rich. US democratic political institutions are based on the principle that all citizens should be able to have economic opportunity and political influence.
We think that the best available data show that the tax code has not become less progressive. To the contrary, it has become somewhat more progressive, at least through 2015, the last year for which complete data is available.
Did the Republican tax cuts in 2017 undo this result? Hardly. Recent research by Splinter applies a forecast of the impact of the Tax Cuts and Jobs Act to historical income data and concludes that the effects of the tax cut on the progressivity of the tax code are quite modest. Tax rates do decline for those at and near the top, but the long-run trend is still toward an increasingly progressive tax system.
Rejecting the conclusions of Picketty, Saez, and Zucman does not imply that inequality is not a problem, or that the tax code is sufficiently progressive. That’s a separate matter and one that deserves careful analysis. Policymakers should know, however, that significantly increasing taxes on high-income Americans — Senator Warren’s wealth tax, for example — in order to alter the degree of inequality is a radical departure from policy norms, and not a return to the progressivity of a tax system a half-century ago. And such steps will likely come at a high economic cost as such policies would yield high effective marginal tax rates that distort and discourage growth.
The radical changes to the tax code being proposed by progressive politicians are supported by the misconception that the tax code has grown significantly less progressive. On a topic as important as this, we think that politicians, journalists, and public intellectuals would do well to moderate their rhetoric and take a hard look at the data.
Is the tax code sufficiently progressive? The answer depends not only on the values of the person answering the question, but also the tricky evidence.
|
