The Mellon Tax Cuts of the 1920s and What Happened to Revenue

Let me give you a puzzle, and I will label it a hypothetical so nobody mistakes it for a news story. A wealthy investor, call her Ms. A, has $100,000 to put to work in 1921. She can buy a corporate bond, issued by a company that will use the money to build a factory, which pays 6 percent. Or she can buy a bond issued by a city, which pays 4 percent, and which is exempt from the federal income tax. Which should she pick?

If the top federal tax rate is 73 percent, as it was in 1921, the answer is easy, and it is not good for the factory. The corporate bond pays $6,000 a year, but she keeps only 27 cents on each dollar of it: 6,000 times 0.27 is $1,620. The city bond pays $4,000 and she keeps all of it. She buys the city bond. The factory goes unbuilt, and the Treasury collects exactly nothing from her. A 73 percent tax rate on paper turned into a zero percent tax in practice.

Now lower the top rate to 25 percent. The corporate bond leaves her with 6,000 times 0.75, which is $4,500. The city bond still pays $4,000. Now she buys the factory bond, the factory goes up, and the Treasury collects 25 percent of $6,000, which is $1,500, from money it used to get nothing from.

That is the whole idea behind the Mellon tax cuts in one example. It is the oldest test case for the claim that lower rates can bring in revenue and growth, and both sides of the modern argument cite it. Let me explain what the law actually did, what happened, and where the honest critics have a point.

The setting: a country with a wartime tax code

The 16th Amendment, ratified in 1913, gave Congress the power to tax incomes. The first rates were modest, a top rate of 7 percent. Then came the First World War. To pay for it, Congress raised the top rate in steps, to 77 percent in 1918, and kept it high afterward. By 1920 and 1921 the top rate was 73 percent, and there was also an excess-profits tax on businesses. The federal debt stood at about $25 billion at the end of the 1919 fiscal year, more than twenty times what it had been before the war.

The voters in 1920 chose Warren Harding, a Republican, on a promise of "normalcy." Harding named as his Secretary of the Treasury a Pittsburgh banker and industrialist named Andrew Mellon. Mellon served from 1921 until 1932, under three Presidents: Harding, Coolidge and Hoover. He was a quiet man, a collector of art and a student of balance sheets, and he had firm views.

First, the budget

Before there was a tax cut, there was a budget. Until 1921, there was no such thing as a President's budget. Each department sent its spending requests straight to Congress, and nobody in the executive branch added them up. The Budget and Accounting Act of 1921, which Harding signed on June 10, required the President to submit a single budget to Congress each year. It created the Bureau of the Budget, the ancestor of today's Office of Management and Budget, and it created the General Accounting Office, now the Government Accountability Office, as an auditor answerable to Congress. The first Budget Director was Charles Dawes, a Chicago banker.

I mention this because it is the part of the story that gets left out. The tax cuts rested on a spending side. Harding and Coolidge both pushed department heads to cut. You can follow the history of that apparatus in How Washington learned to budget, from the 1921 act to the 1997 deal. The point is simple: you can cut taxes only as far as you are willing to cut or hold spending, and the Republican administrations of the 1920s were willing.

Mellon's reasoning

In 1924 Mellon published a short book, Taxation: The People's Business. His argument went like this:

1. High rates do not collect high revenue. If the top rate is 73 percent, people with large incomes will find a way to avoid it: they shift into tax-exempt bonds, they hold property that produces no taxable income, or they hold back from investments that would otherwise be taxed.
2. The money does not disappear. It goes somewhere else. It goes into tax-free securities, which the Treasury cannot touch, instead of factories, farms and new businesses, which create jobs.
3. Lower rates bring the money back. If the top rate drops enough that an investment in the economy pays better than a tax shelter, people will invest in the economy, and the government will collect on the income they report.
4. Everyone benefits, because capital creates work. Wages come from the investment of capital, and investment responds to the after-tax return.

He proposed to bring the top rate down steeply, and he said openly that he thought the government would not lose revenue by doing it. Notice what the claim is. It is not that every tax cut pays for itself. It is that a rate that high, on a particular group of taxpayers, was already past the point where it raised money.

The four acts

Congress went most of the way with him, in stages. Here is the record.

Look at the table and the history behind it. Mellon had asked for a 25 percent top rate as early as 1923. Congress, with Democrats and Progressive Republicans pushing back, gave him 46 percent in 1924, and he did not get his number until 1926. The cuts came in stages. That matters for judging the results, because it was not a single event that you can compare before and after.

Then, a worked check of the arithmetic from the opening. Take a hypothetical $1,000 of extra income for a taxpayer in the top bracket. At 73 percent she keeps 1,000 times 0.27, which is $270. At 25 percent she keeps 1,000 times 0.75, which is $750. The same extra work pays her two and three-quarters times as much (750 divided by 270 is about 2.8). That gap is the incentive Mellon was talking about.

What happened to revenue, and to the debt

Now the results. I will be careful here, because this is where both sides tend to oversell.

What is clear. The federal government ran a budget surplus in each year of the 1920s. Total federal receipts were lower in the late 1920s than at the wartime peak, which is no surprise, because the wartime taxes were repealed and the war was over. The federal debt came down. It stood at about $25 billion in 1919, and by 1930 it was about $16 billion. Do the subtraction: that is a decline of roughly $9 billion, or a little more than a third. Harding and Coolidge paid down debt with surpluses while cutting rates.

What Mellon argued. He pointed to Treasury data showing that after the cuts, more income was reported in the highest brackets and the very wealthy paid a growing share of the income tax collected. His argument was that the rate cuts had brought income out of shelters and into the tax base. You can check the annual volumes of the Treasury's Statistics of Income for the numbers, and I would encourage you to. Some modern economists accept this finding for the top brackets in this period. Others point out that part of the increase in reported income came from the rising stock market and a booming economy, not only the rate change.

What is not clear. Whether the tax cuts caused the boom. That is the trap. The decade's economic expansion was real, but nearly everything was happening at once.

The strongest critique

The critics deserve a hearing, and some of what they say is right.

One: the boom had other causes. The 1920s were the decade of the automobile, electrification of homes and factories, radio, the assembly line and a recovery from the sharp recession of 1920 and 1921, which was already ending when the first cut came. New industries were being born. A country that gets electricity and cars will grow whatever its tax code looks like.

Two: inequality grew. The top rate fell, the highest incomes grew rapidly, and the share of income going to the wealthiest households rose during the decade. Critics argued that the cuts helped those who needed help least. The reply is that the same data show the wealthy paying a larger share of the income tax, but an honest person admits that is a statement about taxes, not about who gained from the boom.

Three: farmers and some regions lagged. Agriculture suffered a long slump through much of the decade, and many rural areas did not share in the prosperity. A tax cut on the highest incomes did nothing for a farmer with a mortgage and falling crop prices.

Four: the crash came anyway. The decade ended in the market collapse of 1929. Mellon's cuts did not prevent a speculative bubble or a depression. His defenders say they were not meant to, and that the later disaster owed more to the monetary policy of the Federal Reserve and to what happened abroad. But it does mean nobody should point to the 1920s and say "tax cuts bring lasting prosperity" with a straight face.

What the 1920s show, and what they do not

Here is how I would put the verdict, and I want to separate the fact from the claim.

Fact: Cutting the top rate from 73 percent to 25 percent did not collapse revenue, and the debt was paid down while it happened. The tax base at the top did not shrink the way a naive prediction would have suggested.

Fact: Mellon's logic about avoidance was sound. When the rate on the last dollar is that high, people spend money and effort avoiding the tax instead of earning more. Modern economists agree that behavior responds to rates.

Claim, mine, and I'll flag it as a judgment: The Mellon experience supports a narrow proposition, that rates far above the revenue-maximizing point lose money, and cutting them can raise or hold revenue. It does not support the broad proposition that any tax cut pays for itself. A cut from 73 percent to 25 percent is a different animal from a cut from 35 percent to 30. The Laffer curve, which you can read about in The Laffer curve explained, and the claim it never actually made, makes exactly this distinction, and many people who invoke it forget it.

That is the correct way to hold the argument. Later chapters bear it out. President Kennedy's rate cuts, enacted in 1964, are the subject of kennedy cut tax rates too, and what the 1964 law proves. Ronald Reagan's cuts of 1981 and the reform of 1986 are in reagan's 1981 tax cut and the quiet end of bracket creep and the 1986 tax reform that lowered rates by closing loopholes. For the long run from Harding to the present, see Republican tax cuts side by side, from Harding all the way to Trump. Each time, economists have argued over how much revenue was gained or lost, and each time the answer has depended on how high the starting rate was and what happened to spending.

What this means for your family

You will hear tax plans described in the next few years with a lot of confidence on both sides. Here is a short test for any of them, drawn from the 1920s.

• Find the rate that matters. The rate on the next dollar earned, the marginal rate, drives behavior. The average rate, which is the total tax divided by total income, is what you feel in your budget. Ask which one the plan changes.
• Ask what is being avoided. If a rate is high enough that people pay lawyers to avoid it, look at what the avoidance costs the economy.
• Look at the spending side. A tax cut that comes with a spending cut is a different thing from one that comes with a bigger deficit. Check the budget office's published analysis, which is public.
• Do not trust a single decade. The 1920s are a rich example, and they are one example.

The bottom line

• The Revenue Acts of 1921, 1924, 1926 and 1928 cut the top federal income tax rate from 73 percent to 25 percent, and the Budget and Accounting Act of 1921 gave the government its first President's budget.
• Mellon's central idea, that very high rates drive money into shelters and lose revenue, was sound for the top brackets. The debt fell from about $25 billion to about $16 billion over the decade.
• The honest critique stands: the boom had other causes, inequality grew, farmers lagged and the decade ended in a crash. Do not claim more than the evidence supports.
• When you hear that every tax cut pays for itself, or that no tax cut ever does, remember the difference between cutting a 73 percent rate and cutting a 35 percent one.
• Read the actual tax law and the budget office's analysis for any plan, and ask your own representatives which spending they would cut to match.