Economic Inquiry has just accepted my article with @PhilWMagness on income inequality in the United States before the 1940s. We show that the IRS data is of very poor quality and overstates the actual level of inequality #econhist #econtwitter
LINK: papers.ssrn.com/sol3/papers.cf…
LINK: papers.ssrn.com/sol3/papers.cf…
We point out that the IRS income data relied on self-reporting in an era of rapidly changing rates and (at high levels) with lax enforcement. This makes the errors of estimation unstable and the levels inaccurate.
So we used state-level income tax systems that were differently enforced than the IRS (Wisconsin, Delaware, North Carolina, Utah and Minnesota with the 1st two being the most important) to compare IRS-level estimates of inequality.
We find that these states had much high levels of compliance and also saw much larger shares of total income being reported than with the IRS in those states (especially Delaware). (see graph)
We find systematically lower levels of income inequality in Wisconsin than those who relied directly on the IRS data using Piketty-Saez's methodology
In Delaware we find the same thing with the caveat that we also discover that the smoothing methodology used by Piketty-Saez underestimates inequality in 1929-1931
In other states, we find the exact same pattern: the IRS data heavily overstates inequality levels.
When you pile all these together, you can find that the data has a systemic upward bias (and the few years Delaware is understating is a methodological error in the the transformation made by PS and others)
Finally, we also point out that our contribution fits with the great leveling storyline of my personal idols (Peter Lindert and Jeffrey Williamson who have taught me so much). Its just that in our case, the leveling of inequality is from falling K-Gains
For those who might be interested: @stephenfgordon @wwwojtekk @gabriel_zucman @AlexNowrasteh @m_clem (because of our shared Williamson connection) @MarkKoyama @adamgmartin @AlanReynoldsEcn @CorneliusEcon @gabriel_mathy @GimenezRoche
@stephenfgordon @wwwojtekk @gabriel_zucman @AlexNowrasteh @m_clem @MarkKoyama @adamgmartin @AlanReynoldsEcn @CorneliusEcon @gabriel_mathy @GimenezRoche @D_Kuehn will probably also like this (I know him and my co-author are not the greatest of friends, but he is going to be interested and worth engaging with).
@stephenfgordon @wwwojtekk @gabriel_zucman @AlexNowrasteh @m_clem @MarkKoyama @adamgmartin @AlanReynoldsEcn @CorneliusEcon @gabriel_mathy @GimenezRoche @swinshi - you will probably love this
