Working Paper: NBER ID: w27921
Authors: Emmanuel Saez; Gabriel Zucman
Abstract: Recent studies argue that US inequality has increased less than previously thought, in particular due to a more modest rise of wealth and capital income at the top (Smith et al., 2019; Smith, Zidar and Zwick, 2020; Auten and Splinter, 2019). We examine the claims made in these papers point by point, separating genuine improvements from arguments that do not appear to us well grounded empirically or conceptually. Taking stock of this body of work, and factoring in other improvements, we provide a comprehensive update of our estimates of US income and wealth inequality. Although some of the points raised by the revisionists are valuable, the core quantitative findings of this literature do not appear to be supported by the data. The low capital share of private business income estimated in Smith et al. (2019) is not consistent with the large capital stock of these businesses. In Smith, Zidar and Zwick (2020), the interest rate assigned to the wealthy is higher than in the datasets where both income and wealth can be observed, leading to downward biased top wealth shares; capitalizing equities using almost only dividends dramatically underestimates the wealth of billionaires relative to the Forbes 400. In Auten and Splinter (2019), business profits earned by the top 1% but not taxable (due in particular to generous depreciation rules) are classified as tax evasion; tax evasion is then allocated to the bottom 99% based on an erroneous reading of random audit data. Our revised series show a rise of inequality similar to Saez and Zucman (2016) and Piketty, Saez, and Zucman (2018) while allowing for a more granular depiction of the composition of wealth and income at the top.
Keywords: income inequality; wealth inequality; capital income; tax evasion
JEL Codes: D31; E25; H26
Edges that are evidenced by causal inference methods are in orange, and the rest are in light blue.
| Cause | Effect |
|---|---|
| misclassification of business income and capital income (E25) | underestimated rise in top wealth shares (D33) |
| low capital share of private business income (D33) | perceived modest rise in inequality (D31) |
| higher interest rates assigned to the wealthy (G51) | downward bias in top wealth shares (D31) |
| capitalization of equities based solely on dividends (G35) | underestimates wealth of billionaires (D31) |
| misallocated tax evasion among bottom 99% (H26) | inaccurate reflection of true distribution of income (D31) |
| methodological improvements (C90) | observed increase in inequality (D31) |