As I understand it, the big difference between Auten and Splinter's inequality estimates and those of other researchers are that Auten and Splinter (a) have not harmonized their estimates with the components of the System of National Accounts, (b) assume low-income business owners evade a greater share of the taxes they owe than do high-income business owners, (c) do not assign undistributed pension earnings to their ultimate owners, (d) assume that the corporate tax is largely borne by low-income current retirees, and (e) define the top 1% by summing the incomes of all earners in the family but calculate their earning by dividing income by the number of adults in the family (see PSZ: "Distributional National Accounts: Methods and Estimates for the United States Data Appendix").
I do not, given the random audit studies, understand why they do (b). And I think that (e) is simply wrong. (e) has a big effect on inequality trends because of increasing female paid employment and decreasing marriage rates among the non-rich. Thus at the moment at least I find myself strongly on the side of Piketty, Saez, and Zucman in this disagreement—and not just because two of those three work down the hall from me.
But I do strongly share Auten and Splinter's dissatisfaction with the concept being measured here by the standard estimates. I want to see inequality defined as the lifetime distribution fo economic and social power, and I would very much want to see Piketty, Saez, and Zucman—and others—try to lay out how they think of getting from the income estimates they report to an assessment of the inequality concept we really want to see: Austen Clemens: Progress Toward Consensus on Measuring U.S. Income Inequality: "The Organisation for Economic Co-operation and Development... the team from the U.S. Bureau of Economic Analysis... Gerald Auten... and David Splinter at the congressional Joint Committee on Taxation... Thomas Piketty... Emmanuel Saez and Gabriel Zucman.... All... attempt to quantify changes in income inequality after taxes and transfers...
..The series produced by each of these two teams and includes the after-tax-and-transfer series—transfers being government programs such as Supplemental Nutrition Assistance and the Earned Income Tax Credit—produced by the Congressional Budget Office for comparison....
CBO and... Piketty, Saez, and Zucman find very similar trends in the rise of inequality.... The BEA estimates... [are] close to the CBO’s estimate... and Piketty, Saez, and Zucman’s estimate.... In contrast, Auten and Splinter’s data series shows that after accounting for taxes and transfers, incomes fluctuated, but the distribution of that income has been largely consistent since 1960. Their findings are an outlier....
#noted
