Did Incomes for Most of the Population Fall from 1923 Through 1929?

Abstract

This paper uses independent real wages, earnings, labor force data, and an examination of Kuznet's 1953 data to explore the recent contention that real per capita incomes for most of the American population declined from 1923 through 1929. It concludes that there is little evidence to support this contention.

The Great Overestimation: Tax Data and Inequality Measurements in the United States, 1913–1943

“…If these lower levels of state-level income concentration are linked with estimates from later years, it dampens (but does not invalidate) the U-curve pattern of inequality across the twentieth century. As we pointed out above, our results corroborate other findings that the level of inequality before 1945 is overstated (Geloso et al 2018;Smiley 2000). This finding indicates that movements on the left side of the U-curve of income inequality are likely too pronounced.…”

“…As a result, the reliability of tax data for measuring inequality falls as tax rates increase when income is self‐reported and enforcement mechanisms are weak. This confusion has been well documented (and partially corrected for) at the national level by Smiley (, , )…”

“…As a result, the reliability of tax data for measuring inequality falls as tax rates increase when income is self-reported and enforcement mechanisms are weak. This confusion has been well documented (and partially corrected for) at the national level by Smiley (1983Smiley ( , 1998Smiley ( , 2000. 5 The second factor is that federal income tax rates were substantially higher than state rates during the interwar period.…”

“…5. The adjustments made by Smiley (1983Smiley ( , 1998Smiley ( , 2000 were meant to remove any variation in measured inequality that stemmed only from changes in reporting behavior following tax-code changes. To do so, he ran regressions to estimate the number of tax returns in each net-income class holding tax rates constant.…”

“…Similar changes in the federal personal exemption threshold are observable, resulting in similar selectivity incentives from reporting. These tax code changes induced, as other scholars have noted (Smiley 1983(Smiley , 1998(Smiley , 2000, artificial variations in inequality measurement resulting from tax-management and tax-avoidance strategies that affect the accuracy of tax-reported income. For related reasons, the IRS's lax and inconsistent enforcement of federal income tax reporting proved to be a recurring problem until a series of tax collection reforms enacted during World War II.…”