A decrease in the marginal tax rate that raised the after-tax share of income by 1 percent raised reported taxable income by 0.2 percent.
Marginal income tax rates in the United States changed frequently and substantially in the 1920s and 1930s, and those changes varied greatly across income groups at the top of the income distribution. In The Incentive Effects of Marginal Tax Rates: Evidence from the Interwar Era (NBER Working Paper No. 17860), Christina Romer and David Romer use this variation to get a clearer sense of the economic effects of changes in marginal tax rates.
Interwar tax changes typically had small effects on revenues (because tax rates were low for most households) and even smaller effects on budget deficits (because taxes and spending usually changed in the same direction). Romer and Romer therefore argue that to the extent the changes mattered, it was likely through the incentive effects of changes in marginal tax rates. They analyze the short-run effects of marginal rates by examining the response of reported taxable income, and the long-run effects by studying the behavior of investment.
Romer and Romer find that changes in marginal rates had a statistically significant but economically modest effect on reported taxable income. A decrease in the marginal tax rate that raised the after-tax share of income by 1 percent raised reported taxable income by 0.2 percent. This elasticity is lower than what most comparable studies using postwar data find, particularly for high-income taxpayers. Furthermore, because of the extreme variation in marginal rates in the interwar era, the elasticity is measured with greater precision than in most postwar studies.
Although these results suggest that the effects of changes in tax rates on short-run labor supply and the tendency to shield income from taxation were of limited economic significance in this period, it is possible that tax rate changes skewed investment in ways that affected longer-run growth. Romer and Romer therefore examine indicators of investment and entrepreneurial activity. They find no evidence that the large swings in marginal rates had an important impact on investment in new machinery, commercial and industrial construction, or the costs of financing government versus private investment, but suggestive evidence that they may have affected business formation.