Taxes, Revenues and the "Laffer Curve" (Part I)
Jude Wanniski
February 12, 1999


Supply-side University Lesson #2

In studying public finance, there is nothing more important than an appreciation of the Laffer Curve. Nothing. Empires have been built on the wisdom of a few men who understood the law of diminishing returns, as it applies to tax policy. Caesar Augustus understood. Napoleon understood. The architects of the Byzantine Empire understood. So did the Founding Fathers of the United States. The temporary, but sharp decline of the U.S. economy in the 1970s was the result of the failure of our political leaders to realize the law of diminishing returns was eroding the economy in a new and different way. The tax rates themselves were constant, but because of the inflation that began to take hold in 1967 when the U.S. commitment to a dollar/gold standard began to waver, real rates of taxation were rising and had passed the point of diminishing returns. We were inflating our way up the progressive tax schedules of the income tax code. Robert Mundell and Arthur Laffer were the first economists on earth to realize that basic fact and to predict the bad things that followed.

As early as 1972, Art Laffer explained to me that there always are two rates of taxation that will produce the same revenue, one at a higher level of national production and one at a lower level. The only exception is if you are at exactly the point where the rate is optimum, which means revenues will fall if you raise or lower them. When I met Robert Mundell in May 1974, he elaborated on this concept by teaching me that if a government is in the high range of the tax rate, it can lower the rate safely even though the aggregate revenue immediately declines -- as long as the economy grows faster and produces more revenue than the level predicted by the static assumption of no behavioral change.

Mundell said the deficit could increase under this condition, yet interest rates on government bonds would decline! In other words, if the rate is cut by 10%, the static prediction would be that revenues would fall by 10%. If the economy grew faster because of the incentives of the lower rate, perhaps revenues would only fall by 8% or 9%. By this analysis, he said, the revenue difference at least would have to be sufficient to pay the interest on the bonds the government would float to cover the temporary revenue shortfall. In this sense, the government would be investing in the productive potential of the people, betting that they would respond to the lower rate in a way that would produce a permanent return on that investment. It was this insight which enabled me to see, where others did not, that the Reagan tax cuts, although accompanied by an increase in the deficit, also were accompanied by falling interest rates.

The phenomenon is accepted when it is seen in the microeconomic world of the private firm. The bond market assesses a corporation's bond floatation by examining the uses to which the funds will be deployed. If it believes the issue will produce a positive return on investment (ROI), the interest rate on the bond will be favorable, and the corporation's other debt may even trade up in the secondary market. There is no reason why government bond finance should not be treated the same way, especially when the bonds are being issued explicitly to finance tax cuts. Conservatives fear such daring by governments because they worry governments routinely will go into debt for "investments" that have no chance of producing a positive return. Fed Chairman Alan Greenspan is in that position, always arguing that spending cuts take precedence over tax cuts. But even Greenspan will support a cut in the capital gains tax without offsetting spending cuts. This is because he believes that any cut in the capital gains tax will produce a positive ROI for the nation.

It is important to note that when policy makers or political leaders examine a tax structure, looking for places to raise or lower taxes, they must analyze each one individually. In the national debate now shaping up in Washington, D.C., with the federal budget throwing off big surpluses, every politician has a pet tax he or she would like to cut. The cultural conservatives would like to fix the marriage-tax penalty, for example, on the theory that it encourages mates to remain outside wedlock. Others would like to use the surplus to finance a bigger earned-income tax credit, which gives workers extra cash if they earn less than the lowest income-tax bracket. There are all sorts of tax credits offered to one group or the other, to foster bigger families, or bigger charities, or to subsidize cultural activities. Generally speaking, these have negative effects on economic growth and because they do, their enactment would cause interest rates to rise. That is, they are poor investments for the federal government to be making. In the semester ahead, we will spend a great deal of time on the themes and variations of the Laffer Curve as the debate over "How to save Social Security and Medicare" unfolds in Washington. This week, the essay is one I wrote as an adaptation of Chapter Six of my 1978 book, The Way the World Works. We have learned a lot since 1978, but there is nothing in the following treatment of the subject that I would retract. It also serves as a foundation for where we will go in future weeks, as I will be asking you to handle several variables at once in the interplay of taxation and money. This is a very long essay, so I will divide it into two parts. Part II next week. Take your time and ask questions.

* * * * *

Taxes, Revenues, and the "Laffer Curve"

Associate Editor
The Wall Street Journal
JUNE, 1978

As Arthur Laffer has noted, "There are always two tax rates that yield the same revenues." When an aide to President Gerald Ford asked him once to elaborate, Laffer (who is Professor of Business Economics at the University of Southern California) drew a simple curve, shown on the next page, to illustrate his point. The point, too, is simple enough -- though, like so many simple points, it is also powerful in its implications.

When the tax rate is 100 percent, all production ceases in the money economy (as distinct from the barter economy, which exists largely to escape taxation). People will not work in the money economy if all the fruits of their labors are confiscated by the government. And because production ceases, there is nothing for the 100-percent rate to confiscate, so government revenues are zero.

On the other hand, if the tax rate is zero, people can keep 100 percent of what they produce in the money economy. There is no governmental "wedge" between earnings and after-tax income, and thus no governmental barrier to production. Production is therefore maximized, and the output of the money economy is limited only by the desire of workers for leisure. But because the tax rate is zero, government revenues are again zero, and there can be no government. So at a 0-percent tax rate the economy is in a state of anarchy, and at a 100-percent tax rate the economy is functioning entirely through barter.

In between lies the curve. If the government reduces its rate to something less than 100 percent, say to point A, some segment of the barter economy will be able to gain so many efficiencies by being in the money economy that, even with near-confiscatory tax rates, after-tax production would still exceed that of the barter economy. Production will start up, and revenues will flow into the government treasury. By lowering the tax rate, we find an increase in revenues.

On the bottom end of the curve, the same thing is happening. If people feel that they need a minimal government and thus institute a low tax rate, some segment of the economy, finding that the marginal loss of income exceeds the efficiencies gained in the money economy, is shifted into either barter or leisure. But with that tax rate, revenues do flow into the government treasury. This is the situation at point B. Point A represents a very high tax rate and very low production. Point B represents a very low tax rate and very high production. Yet they both yield the same revenue to the government.

The same is true of points C and D. The government finds that by a further lowering of the tax rate, say from point A to point C, revenues increase with the further expansion of output. And by raising the tax rate, say from point B to point D, revenues also increase, by the same amount.

Revenues and production are maximized at point E. If, at point E, the government lowers the tax rate again, output will increase, but revenues will fall. And if, at point E, the tax rate is raised, both output and revenue will decline. The shaded area is the prohibitive range for government, where rates are unnecessarily high and can be reduced with gains in both output and revenue.

Tax rates and tax revenues

The next important thing to observe is that, except for the 0-percent and 100-percent rates, there are no numbers along the "Laffer curve." Point E is not 50 percent, although it may be, but rather a variable number: it is the point at which the electorate desires to be taxed. At points B and D, the electorate desires more government goods and services and is willing -- without reducing its productivity -- to pay the higher rates consistent with the revenues at point E. And at points A and C, the electorate desires more private goods and services in the money economy, and wishes to pay the lower rates consistent with the revenues at point E. It is the task of the statesman to determine the location of point E, and follow its variations as closely as possible.

This is true whether the political leader heads a nation or a family. The father who disciplines his son at point A, imposing harsh penalties for violating both major and minor rules, only invites sullen rebellion, stealth, and lying (tax evasion, on the national level). The permissive father who disciplines casually at point B invites open, reckless rebellion: His son's independence and relatively unfettered growth comes at the expense of the rest of the family. The wise parent seeks point E, which will probably vary from one child to another, from son to daughter.

For the political leader on the national level, point E can represent a very low or a very high number. When the nation is at war, point E can approach 100 percent. At the siege of Leningrad in World War II, for example, the people of the city produced for 900 days at tax rates approaching 100 percent. Russian soldiers and civilians worked to their physical limits, receiving as "pay" only the barest of rations. Had the citizens of Leningrad not wished to be taxed at that high rate, which was required to hold off the Nazi army, the city would have fallen.

The number represented by point E will change abruptly if the nation is at war one day and at peace the next. The electorate's demand for military goods and services from the government will fall sharply; the electorate will therefore desire to be taxed at a lower rate. If rates are not lowered consistent with this new lower level of demand, output will fall to some level consistent with a point along the prohibitive side of the "Laffer curve." Following World War I, for example, the wartime tax rates were left in place and greatly contributed to the recession of 1919-20. Warren G. Harding ran for President in 1920 on a slogan promising a "return to normalcy" regarding tax rates; he was elected in a landslide. The subsequent rolling back of the rates ushered in the economic expansion of the "Roaring Twenties." After World War II, wartime tax rates were quickly reduced, and the American economy enjoyed a smooth transition to peacetime. In Japan and West Germany, however, there was no adjustment of the rates; as a result, postwar economic recovery was delayed. Germany's recovery began in 1948, when personal income-tax rates were reduced under Finance Minister Ludwig Erhard, and much of the government regulation of commerce came to an end. Japan's recovery did not begin until 1950, when wartime tax rates were finally rolled back. In each case, reduced rates produced increased revenues for the government. The political leader must fully appreciate the distinction between tax rates and tax revenues to discern the desires of the electorate.

The easiest way for a political leader to determine whether an increase in rates will produce more rather than less revenues is to put the proposition to the electorate. It is not enough for the politician to propose an increase from, say, point B to point D on the curve. He must also specify how the anticipated revenues will be spent. When voters approve a bond issue for schools, highways, or bridges, they are explicitly telling the politician that they are willing to pay the high tax rates required to finance the bonds. In rejecting a bond issue, however, the electorate is not necessarily telling the politician that taxes are already high enough, or that point E (or beyond) has been reached. The only message is that the proposed tax rates are too high a price to pay for the specific goods and services offered by the government.

Only a tiny fraction of all government expenditures are determined in this fashion, to be sure. Most judgments regarding tax rates and expenditures are made by individual politicians, Andrew Mellon became a national hero for engineering the rate reductions of the 1920s, and was called "the greatest Treasury Secretary since Alexander Hamilton." The financial policies of Ludwig Erhard were responsible for what was hailed as "an economic miracle" -- the postwar recovery of Germany. Throughout history, however, it has been the exception rather than the rule that politicians, by accident or design, have sought to increase revenues by lowering rates.

Work vs. productivity

The idea behind the "Laffer curve" is no doubt as old as civilization, but unfortunately politicians have always had trouble grasping it. In his essay, Of Taxes, written in 1756, David Hume pondered the problem:

Exorbitant taxes, like extreme necessity, destroy industry by producing despair; and even before they reach this pitch, they raise the wages of the labourer and manufacturer, and heighten the price of all commodities. An attentive disinterested legislature will observe the point when the emolument ceases, and the prejudice begins. But as the contrary character is much more common, 'tis to be feared that taxes all over Europe are multiplying to such a degree as will entirely crush all art and industry; tho' perhaps, their first increase, together with other circumstances, might have contributed to the growth of these advantages.

The chief reason politicians and economists throughout history have failed to grasp the idea behind the "Laffer curve" is their confusion of work and productivity. Through both introspection and observation, the politician understands that when tax rates are raised, there is a tendency to work harder and longer to maintain after tax income. What is not so apparent, because it requires analysis at the margin, is this: As taxes are raised, individuals in the system may indeed work harder, but their productivity declines. Hume himself had some trouble with this point:

There is a prevailing maxim, among some reasoners, that every new tax creates a new ability in the subject to bear it, and that each increase of public burdens increases proportionably the industry of the people. This maxim is of such a nature as is most likely to be abused; and is so much the more dangerous as its truth cannot be altogether denied: But it must be owned, when kept within certain bounds, to have some foundation in reason and experience.

Twenty years later, in The Wealth of Nations, Adam Smith had no such problem: In his hypothetical pin factory, what is important to a nation is not the effort of individuals but the productivity of individuals working together. When the tax rates are raised, the workers themselves may work harder in an effort to maintain their income level. But if the pin-making entrepreneur is a marginal manufacturer, the increased tax rate will cause him to shift into the leisure sphere or into a lower level of economic activity, and the system will lose all the production of the pin factory. The politician who stands in the midst of this situation may correctly conclude that the increase in tax rates causes people to work harder. But it is not so easy for him to realize that they are now less efficient in their work and are producing less.

To see this in another way, imagine that there are three men who are skilled at building houses. If they work together, one works on the foundation, one on the frame, and the third on the roof. Together they can build three houses in three months. If they work separately, each building his own home, they need six months to build the three houses. If the tax rate on homebuilding is 49 percent, they will work together, since the government leaves them a small gain from their division of labor. But if the tax rate goes to 51 percent, they suffer a net loss because of their teamwork, and so they will work separately. When they were pooling their efforts, since they could produce six houses in the same time it would take them to build three houses working alone, the government was collecting revenues almost equivalent to the value of three completed homes. At the 51-percent tax rate, however, the government loses all the revenue, and the economy loses the production of the three extra homes that could have been built by their joint effort.

The worst mistakes in history are made by political leaders who, instead of realizing that revenues could be gained by lowering tax rates, become alarmed at the fall in revenues that results when citizens seek to escape high tax rates through barter and do-it-yourself labor. Their impulse is to impose taxes that cannot be escaped, the most onerous of which is a poll tax or head tax, which must be paid annually for the mere privilege of living. Hume had no difficulty in pointing out the fallacy of that line of thinking:

Historians inform us that one of the chief causes of the destruction of the Roman state was the alteration which Constantine introduced into the finances, by substituting a universal poll tax in lieu of almost all the tithes, customs, and excises which formerly composed the revenue of the empire. The people, in all the provinces, were so grinded and oppressed by the publicans [tax collectors] that they were glad to take refuge under the conquering arms of the barbarians, whose dominion, as they had fewer necessities and less art, was found preferable to the refined tyranny of the Romans.

The trouble with a poll tax, as Hume noted, is that it can be escaped -- one method being not to defend your country against an aggressor who promises to remove the tax as soon as he has gained power. Montesquieu made a similar observation in Book XIII of The Spirit of the Laws:

Because a moderate government has been productive of admirable effects, this moderation has been laid aside; because great taxes have been raised, they wanted to carry them to excess; and ungrateful to the hand of liberty, of whom they received this present, they addressed themselves to slavery, who never grants the least favor.

Liberty produces excessive taxes; the effect of excessive taxes is slavery; and slavery produces diminution of tribute. . . .

It was this excess of taxes that occasioned the prodigious facility with which the Mahommedans carried on their conquests. Instead of a continual series of extortions devised by the subtle avarices of the Greek emperors, the people were subjected to a simple tribute which was paid and collected with ease. Thus they were far happier in obeying a barbarous nation than a corrupt government, in which they suffered every inconvenience of lost liberty, with all the horror of present slavery.

Modern governments have at least abandoned the notion of using a poll tax to generate revenues. Instead, they often go directly to the barter economy in search of revenues. Activities previously not admitted to the money economy and public marketplace because of public disapproval -- e.g., gambling and pornography -- are welcomed because of the promise of revenues. But this process tends to lower the quality of the marketplace itself, hastening the exodus or discouraging the entry of enterprises that have earned public approbation. (End of Part I )

* * * * *

If I were to write this essay today from scratch, I would write it a bit differently, stressing the difference between taxing labor and taxing capital. I've learned new things about the way the world works since 1978 and the subtle differences between taxing labor and capital is one of them. It led me, for example, to see the proper tax rate on capital gains is zero, where I had originally thought it might be closer to 10% or 15%. A zero tax would seem to be inconsistent with the Laffer Curve, but capital gains have to be seen as something which should not be subject to taxation at all, and that any tax on capgains causes overall government revenues to decline. We will take this up in asubsequent lesson. In next week's Part II of the 1978 article, the emphasis is on how the principle of the Laffer Curve has worked through history. Again, take your time on this lesson and ask questions.