All Commentary
Tuesday, October 1, 1996

Cutting Marginal Tax Rates: Evidence from the 1920s

Tax Cuts Lead to Economic Growth

Dr. Smiley teaches at Marquette University.

Recent political debates have raised the issue of adopting a flat marginal rate federal income tax. Though the marginal rate would be flat, the addition of a generous personal exemption would make the average personal income tax rate rise as it approached the fixed marginal rate of, say, 17 or 20 percent. This issue has generated considerable controversy in political debates and in the press. Among the criticisms leveled at a flat marginal rate tax system are that, contrary to proponents’ claims, a flat marginal tax rate will provide a windfall of after-tax income for the already wealthy, worsen the distribution of income, and exacerbate the already swollen federal government deficits. Supporters have usually concentrated on extolling the virtues of reducing the distortions caused by rising marginal tax rates and of encouraging greater entrepreneurial activity.

Ideally, there would be no personal income tax. The history of the debates over an income tax in the 1890-1911 era makes it clear that an income tax was viewed by its advocates as a means to redistribute income and wealth. It has remained this way as indicated by the vestiges of the progressive marginal rate structure which remain in the code. Such a system leads to an emphasis on obtaining more through political redistribution rather than the expansion of economic activity. And by separating the perceived benefits of a governmental activity from any taxes dedicated to supporting that activity, the income tax made it easier to expand government and increase taxes.[1] The creation of a federal income tax system aimed at the redistribution of income as much as creating a new source of federal tax revenues was one of the worst mistakes in American history.

The Tax Cuts of the 1920s

There are three periods where there were significant tax rate cuts which moved toward a flatter tax rate structure: the 1920s, the 1960s, and the 1980s. All exhibit some of the same characteristics, but the tax cuts of the 1960s were smaller than those of the 1920s, and in the 1980s the sharp increases in tax rates for the Social Security system partially offset the cuts in the federal income tax rates.

The first permanent federal income tax was enacted in 1913, and during the First World War there were dramatic increases in the rates in an attempt to generate increased tax revenues. At $4,000 net income, the marginal rates rose from 1 percent in 1915 to 6 percent in 1918; at $25,000 net income from 2 percent to 23 percent; at $100,000 net income from 5 percent to 60 percent; and, at $750,000 net income from 7 percent to 76 percent. The rates were reduced in 1922, 1924, and 1925. By 1925 the highest marginal rate was 25 percent for $100,000 and more net income. By the late 1920s only about the top 7 to 8 percent of Americans were subject to federal personal income taxes.[2] Though the marginal rate was not constant, the changes were close enough to that which would occur with a flat rate tax that the results of the tax cuts of the 1920s can suggest what would happen with the adoption of a flat rate federal income tax.

Tax Cuts for the Wealthy?

A common criticism of the proposal for a flat marginal rate tax is that it would generate a windfall for the wealthy and create greater inequalities in income distribution. Such charges were also made in the 1920s, 1960s, and 1980s. In the 1920s, tax rates were reduced much more for the higher-income taxpayers because, obviously, they had much higher marginal tax rates in 1918. For example, the marginal income tax rate was reduced 51 percentage points (76 percent to 25 percent) between 1918 and 1925 for taxpayers with at least $750,000 of net income, while the reduction for a taxpayer with $6,000 net income over that period was only 10 percentage points (13 percent to 3 percent).[3] However, the relative reduction (decrease as a percent of the 1918 marginal tax rate) was somewhat larger for the lower-income taxpayers than for the higher-income taxpayers.

More importantly, the reduction in tax rates shifted the effective burden of taxation. When rates had been increased between 1915 and 1918 the higher-income taxpayers had found various ways to shelter their income from taxes. At the same time as the number of returns in the lower net-income brackets rose as exemptions were reduced, the number of returns in the higher-income brackets fell. As examples, for the $500,000 to $1,000,000 net income class, the number of returns fell from 376 in 1916 to 178 in 1918, and for the $250,000 to $500,000 net-income class the number of returns fell from 1,141 to 629 over the same period. The result was that the share of income taxes paid by the higher net income tax classes fell as tax rates were raised. With the reduction in rates in the twenties, higher-income taxpayers reduced their sheltering of income and the number of returns and share of income taxes paid by higher-income taxpayers rose. For example, the share of total personal income taxes paid by taxpayers with net incomes of $1,000,000 or more rose from 5.75 percent in 1923 to 15.9 percent in 1927. For taxpayers with net incomes of $250,000 to $500,000 their share of total personal income taxes rose from 6.82 percent in 1923 to 12.40 percent in 1927. The share for taxpayers with net incomes of $100,000 to $250,000 rose from 15.7 percent in 1923 to 21.91 percent in 1927. However, taxpayers with net incomes of $25,000 or less paid 36.22 percent of all personal income taxes in 1923 but only 12.83 percent in 1927. Thus, cutting tax rates effectively shifted the tax burden from the lower-income taxpayers toward the higher income taxpayers.

The assertion that the tax cuts would primarily benefit higher-income taxpayers was tied to the contention that this would create more income inequality. It has always seemed contradictory to me to argue that allowing a person to retain more of the income he or she generated would create more income inequality, but that has been the common contention. The conventional measures did show significant increases in income inequality during the twenties but there were problems with these measures. They were developed from the income reported on income tax returns and separate estimates of total income in the economy. However, as tax rates fell during the twenties, higher-income individuals began shifting wealth so that less of their income was sheltered from taxes. A portion of the greater income gains of the higher-income individuals represented not additional income but income from wealth which was shifted from tax shelters to assets subject to taxation. Correcting for this significantly reduces the rise in income inequality during the twenties.

What of the rise in income inequality that did occur? Individuals receive earnings from the productivity of their capital investments and land as well as their labor. They also receive income in the form of the realized gains in the values of their assets. The values of financial assets, particularly stocks, began to rise by the mid-twenties and this culminated in the great stock market boom of the late twenties. To see what effect this had, I calculated income shares which excluded realized capital gains, and when this was done, essentially all of the rise in income inequality in the twenties disappeared.

Thus, this evidence suggests that the dramatic tax cuts associated with moving toward a flatter rate tax structure would not provide windfalls of income for the wealthier taxpayers. It would encourage them to shift wealth from tax-sheltering investments to taxable investments to receive larger after-tax returns. The movement of economic activity out of lower return tax sheltering into higher return taxable assets will create more efficiency and make people in the society better off.

Larger Government Budget Deficits?

Another argument frequently thrown at the supporters of a flat marginal rate income tax is that it would worsen the annual deficits of the federal government. This would occur because expenditures would continue at the same level while revenues would decline. Once more we can examine evidence from the twenties which is related to this. With the end of the First World War the federal government’s expenditures dropped sharply, though not to the prewar levels, and budget surpluses were created. There were calls to reduce the income tax rates to direct investment into more appropriate channels rather than into activities which were primarily directed to tax avoidance, and to reduce the widespread legal tax avoidance by the upper-income taxpayers. For example, Andrew Mellon, Secretary of the Treasury, reported that when William Rockefeller (John D.’s brother) died in 1922 he held less than $7,000,000 in Standard Oil bonds but over $44,000,000 of wholly tax-exempt securities. The inability of Congress to find legislation to effectively reduce this tax avoidance was one force leading to the twenties’ tax cuts.

The first of the major tax cuts was passed in November of 1921. On average it reduced marginal personal income tax rates by 13.8 percent, and this led to a decline in real total federal personal income tax revenues of 4.3 percent. The second major tax cut was approved in June of 1924 and it reduced marginal income tax rates by an average of 7.5 percent. This tax cut lead to an increase in real total federal personal income tax revenues of 5.9 percent. The final major tax cut was introduced in December 1925 and enacted in February 1926. It applied retroactively to 1925. On average marginal personal income tax rates were reduced 33.6 percent by these changes. Rather than falling, real federal personal income tax revenues increased by 0.5 percent with this large tax cut.

The evidence clearly indicates that, in general, tax revenues rose with the tax cuts of the twenties. The federal government’s budget surpluses were not reduced with the final two tax cuts and, over the course of the twenties, these budget surpluses allowed the federal debt to be reduced by 25 percent.


The flat marginal rate income tax may never be enacted. Many people, and this certainly includes many politicians, believe that it is only fair that higher-income individuals face higher marginal rates of income taxation. The tenacity with which supporters of progressive tax rates cling to this idea is indicative of their redistributionist philosophy. It also indicates their refusal to face reality. The tax cuts of the twenties as well as every major income tax cut has resulted in an effective shift of the tax burden from lower- to higher-income taxpayers. As the twenties show, it does not have to worsen the government’s deficit. Economic growth in the twenties surged with the tax cuts, and prices were nearly stable while unemployment rates averaged around 4 percent.[4] The government ran surpluses which allowed it to reduce the federal debt by 25 percent. The decreases in marginal tax rates led individuals to pull their investments out of ones designed to avoid taxes—investments such as tax-exempt municipal bonds, personal service corporations, and other avenues to avoid distributing corporate profits. The result was a rising tide of investment in new, growing, and sometimes risky businesses and industries such as radio, consumer household electric appliances, electric utilities, airplane manufacturers, rubber tire manufacturers, supermarket chains, and so forth. The 1920s were a vibrant, growing decade, and the tax cuts of the 1920s certainly were an important part of what brought this about.

1.   See Robert Higgs, Crisis and Leviathan: Critical Episodes in the Growth of American Government (New York: Oxford University Press, 1987).

2.   Personal exemptions were also increased during the decade.

3.   Much of the following discussion relies upon two sources. Gene Smiley and Richard H. Keehn, Federal Personal Income Tax Policy in the 1920s, The Journal of Economic History, Vol. 55 (June 1995), pp. 285-303; and, Gene Smiley, New Estimates of Income Shares During the 1920s presented at Calvin Coolidge and the Coolidge Era, a library of Congress Symposium on the Politics, Economics, Social, and Cultural History of the United States in the 1920s, October 6, 1995, and forthcoming in a conference proceedings volume.

4.   Between 1919 and 1929 real per capita GNP grew 2.61 percent per year. (1920 was the first year of the 1920-21 depression and is not an appropriate starting point.) For comparison, real GNP per capita grew 1.48 percent per year from 1950 to 1959, 3.26 percent per year from 1960 to 1969 (with significant tax rate cuts), 2.68 percent per year from 1970 to 1979, and 2.09 percent per year from 1980 to 1988.