I have another working paper out (it was out a few weeks ago, I just forgot to share it here). Its with two of my graduate students (Jacob Hall and Patrick Fitzsimmons) at George Mason University. We recollected the state-level IRS SOI data in order to create corrections for state-level disparities in price levels from 1921 to 1941. The idea was to better measure “real income” inequality across the United States during the first half of the 20th century. We find that adjusting for state-level price differences attenuates the level of inequality in the United States but also leads to a faster collapse from 1921 to 1941. The abstract is below and the link to the SSRN paper is here.
The U-Curve of income inequality in the United States is a longstanding stylized fact in economic history. The “Great Levelling” that led to the trough that lasted from the 1940s to the early 1980s is argued by scholars like Piketty and Saez (2003) to have happened precipitously during the 1940s whereas others like Geloso et al. (2022) argue that it was a gradual levelling that began with the Great Depression. In this paper, we argue that large regional price level differences make it hard to measure “real” inequality levels. More importantly, as these price differences collapsed in the first half of the 20th century, the trends in “real” income inequality could be far different than those using “nominal” income. Adjusting income levels from 1921 to 1941 for regional price levels shows a faster decline in inequality during the period. We argue that this indicates that the Great Levelling was a gradual process than began far earlier than the 1940s.
Vincent Geloso 