The US is Less Unequal Than We Think But Inequality Has Risen No Less Rapidly

• In Event: Inequality and Measurement

Abstract

We propose alternative estimates of the level and trend of U.S. pretax and posttax income inequality to reconcile the debate between Piketty, Saez and Zucman (2018) and Auten and Splinter (2023). To do so, we use our approach in Pinkovskiy, Sala-i-Martin, Chatterji-Len and Nober (2024) and amend Census / ACS reported income using an elasticity of underreporting that we estimate by comparing survey and national accounts local area incomes. This approach provides an empirical method of testing assumptions about the allocation of income concepts that are not featured in the tax records. We find that the rates of increase of the top 10% share of post-tax income are closer to those found in Piketty et al. 2018 than in Auten and Splinter 2023. In contrast, the levels of top 10% income shares tend to be closer to those found in Auten and Splinter 2023, with the Piketty et al. 2018 estimates closely approximated by the top 10% shares of GDP rather than by the top 10% shares of personal income. We exploit the richness of the Census / ACS data to show that our results are robust to allowing differential underreporting in urban vs. rural areas.

Authors

• Maxim Pinkovskiy, Federal Reserve Bank of New York

• Xavier Sala-i-Martin, Columbia University

• Beckett Pierce, Federal Reserve Bank of New York