The Laffer curve explained, and the claim it never actually made

The legend goes like this. December 1974, a restaurant in Washington, a table, a drink or three. An economist named Arthur Laffer, a newspaper man named Jude Wanniski, and two young officials of the Ford White House, Dick Cheney and Donald Rumsfeld, are arguing about taxes. The administration is debating a tax increase. Laffer grabs a napkin, draws a curve on it, and explains to the room why raising a tax rate does not always raise the tax collected. The napkin goes into history, and a political movement gets its logo.

That is the story as Wanniski told it, and I'd file it under lore. Laffer himself has been cheerfully vague about the napkin. The restaurant, the date, and the cast are what Wanniski recorded, and you can believe as much of the dinner as you like. What matters is not the napkin. What matters is the idea on it, because that idea has been misquoted by friends and enemies alike for fifty years. Conservatives claim it says more than it does. Liberals say it says something it never said, and then say it was debunked. Both are wrong, and I'd like to help you be right.

What the curve actually says

Take a tax on income. Ask what revenue the government collects at every possible rate from zero to 100 percent.

1. At a rate of zero percent, revenue is zero. Nobody pays a tax that is not charged.
2. At a rate of 100 percent, revenue is zero too. If the government takes every dollar you earn, there is no reason to earn it officially. You work less, you work for cash, you work somewhere else, or you don't work at all.
3. Revenue is greater than zero between those two points. That is simple arithmetic, since a positive rate on a positive income yields a positive sum.
4. So the revenue curve rises, peaks somewhere, and falls. A curve that starts at zero, ends at zero and rises in between must have a highest point.

That is the Laffer curve, in four premises. It is not controversial. Economists of every political stripe accept it, because the logic is airtight. The only things in dispute are where the peak is and what it means for a tax cut today. Notice what the four steps do not say. They do not say the peak sits at 25 percent or at 70. They do not say that any given country is on the downward side. And they do not say a tax cut brings in more money. They say a peak exists.

An invented example, with the arithmetic

Let me make the shape concrete with a toy economy. These numbers are made up to show the logic, not measured from anywhere. Suppose that when the tax rate is zero, the economy generates $100 of taxable income. As the rate rises, people shift income out of the tax base: they work a little less, take more of their pay in untaxed forms, or move income to places it is taxed less.

Check the multiplications: 0.20 times 95 is 19; 0.40 times 85 is 34; 0.50 times 75 is 37.5; 0.60 times 60 is 36; 0.80 times 35 is 28. In this toy world the peak is near 50 percent.

Now watch what a tax cut does in two different places.

• Cut the rate from 80 percent to 60 percent. Revenue goes from $28 to $36. The government collects $8 more by charging less. This is the Laffer claim at its best: on the wrong side of the peak, a cut raises revenue.
• Cut the rate from 40 percent to 20 percent. Revenue falls from $34 to $19. The government loses $15. Even though the economy is a bit bigger (the base rose from $85 to $95), the cut does not come close to paying for itself.

Those two lines contain the whole debate. A tax cut that starts on the downward side of the curve pays for itself. A tax cut that starts on the upward side does not. Which side of the peak you are on depends on the tax, the economy, and how easy it is to avoid the tax. For a tax with a top rate of 91 percent, like the one that applied in 1963, the downward side is plausible. For a lower rate, it generally is not. A serious conservative should say so out loud.

The older versions

The idea did not begin on a napkin. Laffer himself pointed to the fourteenth-century North African scholar Ibn Khaldun, whose Muqaddimah, from the 1370s, observed that at the start of a dynasty taxes are light and bring in large revenue, and at the end they are heavy and bring in little. A historian in a tent in the 1300s had the shape of the curve.

The British economist John Maynard Keynes, no friend of the supply-side school, wrote in 1933 that taxation may be so high as to defeat its own object. He was making the same point in plainer words. That is useful ammunition when a liberal tells you the idea is a right-wing invention.

And there was practice. In the 1920s Treasury Secretary Andrew Mellon, who served under Harding, Coolidge and Hoover, argued that the very high wartime rates were driving the rich to shelter their money in tax-exempt bonds rather than put it to work. Congress cut the top rate in steps through the decade: from 73 percent at the start of the 1920s, to 58 percent under the Revenue Act of 1921, to 46 percent in 1924, and to 25 percent under the 1926 law. The Treasury's figures showed that the highest earners paid a growing share of the income tax as the rates fell. I want to be careful about what that proves. Total federal revenue was lower in the late 1920s than at the wartime peak, because wartime taxes were wound down, so this is a story about the top brackets and their share, not about total collections. Critics add that the decade's boom, not the rate cuts, explains the numbers. Both points are fair. For the full story, see The Mellon Tax Cuts of the 1920s and What Happened to Revenue. And for the 1960s, when a Democratic President made the same argument for cutting the 91 percent top rate, see kennedy cut tax rates too, and what the 1964 law proves.

The claim it never made

Now the part that needs saying to our own side. The Laffer curve does not say every tax cut pays for itself. It says that somewhere, at a high enough rate, a cut can raise revenue. A cut from a moderate rate cannot be defended with that argument, and a politician who tries should be asked for the numbers.

Why does this matter? Because a movement that overclaims gives its opponents an easy win. When a conservative says "tax cuts always pay for themselves," a liberal can point to a deficit and say "debunked." He has not debunked the curve. He has debunked a slogan that the curve never supported. But the slogan is the movement's fault, so own it.

Our side has a record here, and part of it is not flattering. In 2012 Kansas, under Governor Sam Brownback, passed one of the boldest state income tax cuts in the country, including an exemption of business pass-through income from the state income tax. Laffer was an adviser. Revenue fell short of what its backers predicted, the state had budget trouble, and in 2017 the Republican-controlled legislature repealed most of it over the governor's veto. A fair conservative reading is that the experiment shows what the curve implies: at those rates, the cut did not pay for itself, and the lesson is to pair tax relief with spending discipline. The other side's version is that the whole idea failed. The truth is in the first reading. If you want the state-level view of income taxes in general, see Income Tax or No Income Tax, and What Each Choice Costs a State.

The honest fight: where is the peak?

Strip away the slogans and the real argument is empirical. Where does the revenue-maximizing rate sit? And does it sit in the same place for every kind of tax?

The answer is that it depends on what you tax and how easy it is to avoid.

1. A tax on labor income is hard to avoid for a wage earner whose employer withholds it, so the response is mostly in hours and effort, and the effect is often small. The peak is probably high.
2. A tax on capital gains is easy to avoid by simply not selling. A gain is taxed only when you realize it, so a high rate on gains can freeze investors in place. The peak is probably lower.
3. A tax on corporate profits can be shifted across borders, and a high rate can send investment and reported profit abroad. The peak is lower where the borders are open.
4. A tax on a single item such as a luxury good or a cigarette can drive buyers elsewhere in a way a broad tax cannot.

Economists disagree on the numbers, and a conservative should not claim the peak sits below today's rates without showing the work. That is the right place to fight: not "does the curve exist," but "what is the elasticity of this particular tax base at this particular rate." It's an honest question, and it can be settled with evidence. The people best placed to answer are the staff at the Congressional Budget Office, the Joint Committee on Taxation and the academic economists who study taxable income. Their findings are published, and you can read them.

There's another honest point. Cutting a rate often has a growth effect that recovers part of the cost, even on the upward side of the curve. How much comes back is the subject of "dynamic scoring," the practice of estimating how a tax law changes behavior. Everyone who scores bills now does some of it. The dispute is about how much, not whether.

How to answer "Laffer was debunked"

This is the line you will hear at the family table. Here is a method.

1. Ask what exactly was debunked. If the answer is "the idea that every tax cut pays for itself," agree. Nobody serious made that claim, and Laffer himself did not argue it as a universal law. Say so and take the point off the table.
2. State what remains. "The claim that rates can be so high that cutting them raises revenue is just arithmetic. At 100 percent, nobody earns a taxed dollar. So a peak exists, and the question is where."
3. Offer the shared ground. Keynes thought taxation could defeat its own object, a Democratic President cut the 91 percent rate on that logic, and the idea is far older than any party. It isn't a Republican trick.
4. Move to the real question. "Which tax, at which rate, and how much does the base move?" That is a debate you can win with facts, and it is one your opponent has to join on level ground.
5. Concede what you must. If someone brings up Kansas, do not flinch. Say yes, that cut cost more than its backers hoped, and that is why a tax cut should come with a spending plan. A conservative who can admit that is believed on everything else.

The joke here writes itself: a politician who says every tax cut pays for itself is like a man who says every diet works if you eat enough pie. The curve has a peak, and he has missed it. But the joke is on our side as much as theirs, and that is the point.

Why it matters to your family

A family does not live on a curve. It lives on a paycheck. The curve matters because it is the logic behind a simple thing: at some point, a higher tax stops being a way to raise money and becomes a way to discourage the work that makes the money. The overtime shift you turn down because the extra pay is mostly taxed away, the second job a spouse declines, the shop owner who does not hire a sixth employee: those are the margins on which the curve operates, and they are decisions made at kitchen tables, not in think tanks.

To see how the idea shaped actual law, read reagan's 1981 tax cut and the quiet end of bracket creep and the survey in Republican tax cuts side by side, from Harding all the way to Trump. And to keep your arguments tight, test them against seven tax myths that scare working families and cost them money.

The bottom line

• The Laffer curve makes one claim: revenue is zero at a 0 percent rate and at a 100 percent rate, so a peak lies in between. It does not claim that any particular cut pays for itself.
• A tax cut raises revenue only if it starts on the high side of the peak. From a moderate rate, a cut loses revenue, and a conservative should say so before the other side does.
• Use the idea where it belongs: the argument over how far marginal rates discourage work, saving and investment, which is a fight about evidence, not a slogan.
• When someone says Laffer was debunked, ask what claim they mean. Agree that "every cut pays for itself" is false, and hold your ground on the part that is just arithmetic.
• Pair every tax cut you support with a spending plan. Kansas is the cautionary tale, and our bills page lets you see what your own representatives are voting on.