Where the Monetarists Go Wrong

Henry Hazlitt, noted economist. author, editor. reviewer and columnist, is well known to readers of the New York Times, Newsweek, The Freeman. Barron’s, Human Events and many others. Best known of his books are Economics in One Lesson. The Failure of the "New Economics." The Foundations of Morality, and What You Should Know About Inflation.

In the last decade or two there has grown up in this country, principally under the leadership of Professor Milton Friedman, a school calling itself the Monetarists. The leaders sometimes sum up their doctrine in the phrase: "Money matters," and even sometimes in the phrase: "Money matters most."

They believe, broadly speaking, that the "level" of prices of commodities and services tends to vary directly and proportionately with the outstanding quantity of money and credit —that if the quantity of money (comprehensively defined) is increased 10 per cent, the prices of commodities will increase 10 per cent; that if the quantity of money is doubled, prices will double, and so on. (This of course is on the assumption that the quantity of goods remains unchanged. If this is increased also, the rise in prices due to a greater supply of money will be correspondingly less.)

This is called the Quantity Theory of Money. It is not new, but very old. It has been traced by some economic historians as far back as the French economist Jean Bodin in 1566, and by others to the Italian Davanzati in 1588. In its modern form it was most elaborately presented by the American Irving Fisher in The Purchasing Power of Money (1911) and in later books.

The monetarists have added some refinements to this theory, but principally they have devoted themselves to giving it detailed statistical support, and drawing much different conclusions than did Fisher himself regarding an appropriate monetary policy.

When Fisher began writing, the gold standard was still dominant in practice. He proposed to keep it, but with a radical modification. He would have varied its gold content according to the variations of an official price index, so that the dollar should represent, instead of a constant quantity of gold, a constant quantity of purchasing power. Milton Friedman rejects the gold standard altogether. He would substitute for it a law prescribing a precise quantitative issuance of irredeemable paper money:

"My choice at the moment would be a legislated rule instructing the monetary authority to achieve a specified rate of growth in the stock of money. For this purpose, I would define the stock of money as including currency outside commercial banks plus all deposits of commercial banks. I would specify that the Reserve System shall see to it that the total stock of money so defined rises month by month, and indeed, so far as possible, day by day, at an annual rate of X per cent, where X is some number between 3 and 5. The precise definition of money adopted, or the precise rate of growth chosen, makes far less difference than the definite choice of a particular definition and a particular rate of growth." 1

It is with considerable reluctance that I criticize the monetarists, because, though I consider their proposed monetary policy unfeasible, they are after all much more nearly right in their assumptions and prescriptions than the majority of present academic economists. The simplistic form of the quantity theory of money that they hold is not tenable; but they are overwhelmingly right in insisting on how much "money matters," and they are right in insisting that in most circumstances, and over the long run, it is the quantity of money that is most influential in determining the purchasing power of the monetary unit. Other things being equal, the more dollars that are issued, the smaller becomes the value of each individual dollar. So at the moment the monetarists are more effective opponents of further inflation than the great bulk of politicians and even putative economists who still fail to recognize this basic truth.

I must add that I also regret that ‘I take issue on this important question with Milton Friedman, with most of whose great contributions to economics, and especially to the defense of the free market, I have long been in full and admiring agreement. I hope that no reader of this article will get the impression that I fail to appreciate the extent to which Dr. Friedman’s lucidity, persuasiveness and penetration have put us all in his debt.

Let us begin by examining what is wrong with "the" quantity theory of money—which might rather be called the strict or mechanical quantity theory of money. It rests on greatly oversimplified assumptions. As formulated by Davanzati in 1588, the total existing stock of money must always buy the total existing stock of goods—no more, no less. So if you double the stock of money, and the supply of goods remains the same, you must double the average level of prices. Each monetary unit must then buy only half as much as before. As formulated by its modern exponents, the assumptions underlying the strict quantity theory of money are not much advanced from this. As "money is only wanted to buy goods and services," they argue, this proportional relationship must hold.

But this is not what happens. The truth in the quantity theory is that changes in the quantity of money are a very important factor in determining the exchange-value of a given unit of money. This is merely to say that what is true of other goods is true of money also. The market value of money, like the market value of goods in general, is determined by supply and demand. But it is determined at all times by subjective valuations, and not by purely objective, quantitative, or mechanical relationships.

Three Stages of Inflation

In a typical inflation we may roughly distinguish three stages. In the first stage prices do not rise nearly as fast as the quantity of money is being increased. For one thing, if there has been some slack in the economy, purchases made with the new money may mainly stimulate increased production. (This is the point so emphasized and overemphasized by Keynes. It can happen, however, only in the early stages of an inflation, and only in special circumstances.)

Apart from this possible early stimulative effect of an inflation, most people at first do not realize that an inflation of the currency has taken place. Some prices have risen, but many people, comparing them with the prices to which they have become accustomed, assume that these new prices are too high, and will soon fall back to "normal." They hold off buying, and increase their cash holdings. As a result, prices do not at first rise as much as the quantity of money has been increased.

If the inflation is slow and has occasional stops, prices tend to catch up with the rate of increase in the money supply, and for a while there may be a result much like what the strict quantity theory of money would predict, in which prices tend to rise roughly in proportion to the increase in the money stock.

But if the inflation (meaning the increase in the quantity of money) continues, and particularly if it accelerates, people begin to fear that it is a deliberate governmental policy, that it will go on indefinitely, and that prices will continue to soar. So they hasten to spend their money while it still has some value—i.e., before prices rise still further. The result is that prices begin to rise far faster than the quantity of money has been increased, and finally far faster than it even can be increased.

So we have the paradoxical result that, in a hyperinflation, when the government is grinding out new currency units at an astronomical rate, prices rise so fast that the existing quantity of money is not sufficient for the volume of transactions, and we have mounting complaints of a "scarcity" of money.² In the final stage of the German inflation of 1923, for example, the entire stock of paper money, though with a stamped value billions of times higher, had a gold exchange-value of only one-sixtieth of what it had before the inflation started. Of course the paper mark finally became utterly valueless, as had the French assignats in 1796 and the American Continental currency in 1781.

It is for this reason that all inflation must finally have a stop. But the point I am stressing here is that the strict quantity theory of money is not true (though it may appear to be true under certain circumstances and for limited periods). So far as quantity is concerned, it is the expected future quantity of money, rather than the immediately existing quantity, that determines the exchange value of the monetary unit.

Quality Affects Value

The value of money, however, is determined not merely by its quantity—even its expected future quantity—but also by its quality. Currency issued by a shaky government, for example, will not have as much value, other things being equal, as currency issued by a strong "legitimate" government of long standing.

In recent years we have witnessed much more familiar illustrations of the effect of qualitative deterioration in the monetary unit. Scores of nations have repeatedly announced "devaluations" of their currency. Prices have begun to rise in those countries the very next day, before there has been any chance to increase the quantity of money any further.

Still more striking is what has happened when nations on a gold standard have announced their abandonment of it. The United States went off the gold standard in March of 1933. By 1934, the average of wholesale prices had increased 14 per cent over 1933, and by 1937, 31 per cent. The U.S. formally abandoned gold convertibility again in August, 1971. Wholesale prices had actually fallen by 2 per cent from August of the year before; but by August of the year later they increased by 4.35 per cent. With all gold discipline removed, wholesale prices rose more than 13 per cent between 1972 and 1973, and more than 34 per cent between 1972 and 1974.

One of the most striking illustrations of the importance of the quality of the currency occurred in the Philippines at the late stage of World War II. The forces under General Douglas MacArthur had effected a landing at Leyte in the last week of October, 1944. From then on, they achieved an almost uninterrupted series of successes. A wild "inflation" broke out in the capital city of Manila. In November and December, 1944, prices in Manila rose to dizzy heights. Why? There was no increase in the money stock. But the inhabitants knew that as soon as the American forces were completely successful their Japanese-issued pesos would be worthless. So they hastened to get rid of them for whatever real goods they could get.3

Quantity Theory of Money

What has helped to keep the strict mathematical quantity theory of money alive, in spite of experiences of the kind just cited, is the famous Irving Fisher equation: MV = PT. In this M stands for the quantity of money, V for its "velocity of circulation," P for "the average price level" of goods and services, and T for the "volume of trade," or the quantity of goods and services against which money is exchanged.

So when the quantity of money remains unchanged, for example, and prices start to soar (or any similar discrepancy occurs) the quantity theorists are not at all disconcerted. They are provided in advance with an easy alibi: the "velocity of circulation" of money must have changed enough to account for the apparent discrepancy. True, this requires them sometimes to assume some remarkable things. I pointed out a few pages back that in the final stage of the German inflation of 1919-1923 the entire stock of paper money had a gold value only one-sixtieth that of the far smaller nominal money stock before the inflation began. This would require us to assume that the average "velocity of circulation" had increased in the meanwhile sixty times.

This is not possible. The concept of the "velocity of circulation" of money, as held by the quantity theorists and embodied in the Fisherine equation MV = PT, is quite fallacious. Strictly speaking, money does not "circulate": it is exchanged against goods. When the turnover of money increases, the turnover of goods increases correspondingly.

(We have here an illustration of how the use of mathematical symbols may mislead an economist even in an elementary application. If MV = PT, and you double V, then it seems to follow that 2 MV = 2 PT, and that this can be read as meaning that doubling V can double P But if we spell out the equation as MxV =P x T, it can be seen that M x 2V does not necessarily equal 2P x T, but more likely P x 2T. In fact, the equation MV = PT does not mean what Irving Fisher and his disciples thought it meant. They considered MV the "money side" of the equation and PT the "goods side." But as Benjamin M. Anderson, Jr. long ago pointed out in a shrewd analysis,4 "Both sides of the equation are money sides… The equation asserts merely that what is paid is equal to what is received.")

Velocity of Circulation—Geographical Variations

There are no reliable statistics on the "velocity of circulation" of hand-to-hand currency.⁵ But we do have figures on the annual rate of turnover of demand bank deposits. As bank deposits in the United States cover about eight-ninths of the media of payment, these figures are an important index.

What is most striking, when we examine these figures, is first of all the wide discrepancy that we find between the rate of turnover of demand deposits in the big cities, especially New York, and the rate that we find in 226 other reporting centers. In December, 1975, the average annual rate of turnover of demand deposits in these 226 small centers was 71.8. In six large cities outside of New York it was 118.7. When we come to New York City itself, the rate was 351.8. This does not mean that people in New York were furiously spending their money at nearly five times the rate of people in the small centers. (We must always remember that each individual can spend his dollar income only once.) The difference is accounted for mainly by two factors. The big corporations have their headquarters or keep their banking accounts in the big cities, and these accounts are much more active than those of individuals. And New York City especially, with its stock exchanges and commodity exchanges, is the great center of speculation in the United States.

Though the velocity of circulation of money (mainly in the form of bank deposits) increases with speculation, speculation itself does not indefinitely increase. In order for speculation to increase, willingness to part with commodities must increase just as fast as eagerness to buy them. It is rapidly changing ideas of commodity values—not only differences of opinion between buyer and seller, but changing opinions on the part of individual speculators—that are necessary to increase the volume of speculation. The value of a commodity, a stock, or a house does not change in any predictable relationship to the number of times it changes hands. Nor does the value of a dollar. When 100 shares of a stock are sold, their value is not thereby necessarily depressed, because the shares are also bought. Every sale implies a purchase, and every purchase a sale. When a man buys a commodity, he "sells" money; but the seller of the commodity "buys" money. There is no necessary connection whatever between changes in the "velocity of circulation" of money and changes in the "level" of commodity prices. "Velocity of money" is merely a resultant of a complex of other factors, and not itself a cause of any important change whatever.6

Price Levels and Indexing

Still another fallacy into which many quantity theorists (and not they alone) are apt to fall is the concept of a price "level." This is the partly unconscious assumption that when prices rise during an inflation they rise uniformly—so that when the official consumer price index has risen over a given period by, say, 10 per cent, all prices in that period have risen just about 10 per cent. This assumption is never made explicitly, otherwise it would be much easier to correct. But it is latent in the discussions of most journalists and politicians. It therefore leads them greatly to underestimate the harm done by inflation. For the greater part of that harm is precisely that different people’s prices, wage-rates and income go up so unevenly and at different rates. This not only means great windfalls for some and tragedies for others, but it distorts and disrupts economic relationships. It unbalances, reduces, and misdirects production. It leads to unemployment and to malemployment. And attempts to correct this through such schemes as "indexing" only tend to increase the harm.

I do not mean to suggest that all those who call themselves monetarists make this unconscious assumption that an inflation involves this uniform rise of prices. But we may distinguish two schools of monetarism. The first would prescribe a monthly or annual increase in the stock of money just sufficient, in their judgment, to keep prices stable. The second school (which the first might dismiss as mere inflationists) wants a continuous increase in the stock of money sufficient to raise prices steadily by a "small" amount-2 or 3 per cent a year.

These are the advocates of a "creeping" inflation. The late Professor Sumner H. Slichter of Harvard was the most prominent of these. He thought that a planned price rise of 2 or 3 per cent a year would be about right. He forgot that even if the government could hold an inflationary price rise to a rate of only 2 per cent a year it would mean an erosion of the purchasing power of the dollar by about one-half in each generation.

And it would not produce the results that Slichter expected of it. For inflation is always a swindle. It cannot be candidly and openly planned. People everywhere will take compensatory actions. If a price rise of 2 per cent a year is announced as the official goal, lenders will immediately add 2 per cent to the interest rate they would otherwise have asked, union leaders will add 2 per cent to the wage increase they would otherwise have demanded, and so around the circle. Not only will the "creeping" inflation begin to race, but its effect on production and employment will be disruptive rather than stimulative.

But our concern here is not with the advocates of creeping inflation (in the sense of creeping price rises, at no matter how low an annual rate) but with the monetarists strictly so called—that is, with those who recommend instructing government monetary authorities to increase the monetary stock every year only enough to keep prices from falling. What increase do the monetarists think sufficient to accomplish their purpose?

Problems of Determining How Much Inflation Is Enough

Let us return to the prescriptions of the acknowledged leader of the school, Professor Milton Friedman. We have seen that, in 1962, in his Capitalism and Freedom, he recommended that the Federal Reserve authorities be instructed to increase the total stock of money (including "all deposits of commercial banks") at an annual rate of.somewhere between 3 and 5 per cent. But three years later, in a memorandum prepared for a consultant’s meeting with the Board of Governors of the Federal Reserve Board on Oct. 7, 1965, we find him recommending "as the new target a rate close to the top of the desirable range of 4 to 6 per cent for M-2" (currency plus demand and time deposits).⁷

Still later, in 1969, we find him scaling down this rate considerably, though with misgivings and vacillations:

"A policy fairly close to the optimum would probably be to hold the absolute quantity of money constant… However, this policy, too, seems to me too drastic to be desirable in the near future although it might very well serve as a long-term objective."

He then discusses the relative advantages of a 1 per cent and of a 2 per cent rate, and then goes on:

"I do not want to gloss over the real contradiction between these two policies, between what for simplicity I shall call the 5 per cent and the 2 per cent rules. There are two reasons for this contradiction. One is that the 5 per cent rule was constructed with an eye primarily to short-run considerations, whereas the 2 per cent rule puts more emphasis on long-run considerations. The more basic reason is that I had not worked out in full the analysis presented in this paper when I came out for the 5 per cent rule. I simply took it for granted, in line with a long tradition and a near-consensus in the profession, that a stable level of prices of final products was a desirable policy objective. Had I been fully aware then of the analysis of this paper, I suspect that I would have come out for the 2 per cent rule…

"Either a 5 per cent rule or a 2 per cent rule would be far superior to the monetary policy we have actually followed. The gain from shifting to the 5 per cent rule would, I believe, dwarf the further gain from going to the 2 per cent rule, even though that gain may well be substantial enough to be worth pursuing. Hence I shall continue to support the 5 per cent rule as an intermediate objective greatly superior to the present practice."⁸

One hardly knows whether to twit Dr. Friedman for tergiversation or praise him for remarkable candor. But his hesitations, as I hope to show, really point to the inherent difficulties in the monetarists’ proposals.

I made a distinction earlier between the monetarists strictly so called and the "creeping inflationists." This distinction applies to the intent of their recommended policies rather than to the result. The intent of the monetarists is not to keep raising the price "level" but simply to keep it from falling, i.e., simply to keep it "stable?’ But it is impossible to know in advance precisely what uniform rate of money-supply increase would in fact do this. The monetarists are right in assuming that in a prospering economy, if the stock of money were not increased, there would probably be a mild long-run tendency for prices to decline. But they are wrong in assuming that this would necessarily threaten employment or production. For in a free and flexible economy prices would be falling because productivity was increasing, that is, because costs of production were falling. There would be no necessary reduction in real profit margins. The American economy has often been prosperous in the past over periods when prices were declining. Though money wage-rates may not increase in such periods, their purchasing power does increase. So there is no need to keep increasing the stock of money to prevent prices from declining. A fixed arbitrary annual increase in the money stock "to keep prices stable" could easily lead to a "creeping inflation" of prices.

A Political Football

But this brings us to what I consider the fatal flaw in the monetarist prescriptions. If the leader of the school cannot make up his own mind regarding what the most desirable rate of monetary increase should be, what does he expect to happen when the decision is put in the hands of the politicians?

We do not need to allow our fancies to roam very far. We already know the answer from what has been happening in the United States since we left the gold standard 43 years ago—and from what has been happening, for that matter, in nearly every country in the world since the gold standard was abandoned. The decision regarding the national money-supply has already been in the hands of the politicians everywhere. And this situation has led practically everywhere to continuous and usually accelerating inflation.

Dr. Friedman would take the decision out of the discretion of appointed monetary authorities and make it a "legislative rule." But what rate would a popularly-elected legislature set? We may be sure that it would pick a "safe" rate of monetary expansion—at least 6 per cent a year to begin with—to make sure that there would be no depression or unemployment. But at the first feeble sign of unemployment or "recession"—brought about by excessive union wage demands or any other of a score of factors—politicians seeking election or reelection would demand that the legislative monetary-increase rule be raised to 8 per cent, 10 per cent, or whatever rate the political scramble for office might suggest.

The prescribed rate would become a political football. The tendency nearly always would be for the highest bidder to win. For the belief in inflation as the master solution of every economic problem is not new in this generation. Throughout recorded history it has always been latent. It did not need the Keynesian rationalization. Whenever there has been depression and unemployment it has always been popularly blamed on or identified with "not enough money." In 1776, in his Wealth of Nations, Adam Smith was pointing out that "No complaint is more common than that of a scarcity of money."

The fatal flaw in the monetarist prescription, in brief, is that it postulates that money should consist of irredeemable paper notes and that the final power of determining how many of these are issued should be placed in the hands of the government—that is, in the hands of the politicians in office. The assumption that these politicians could be trusted to act responsibly, particularly for any prolonged period, seems incredibly naive. The real problem today is the opposite of what the monetarists suggest. It is how to get the arbitrary power over the stock of money out of the hands of the government, out of the hands of the politicians.

The solution to that problem cannot be offered in a few lines. The present writer has attempted to deal with it in his article, "The Search for an Ideal Money," in The Freeman of November, 1975.

¹ Capitalism and Freedom (University of Chicago Press. 1962), p. 54.

² See, e.g., on the French assignats, Andrew Dickson White, Fiat Money Inflation in France (Irvington-on-Hudson, New York: Foundation for Economic Education. 1959), p. 83 and passim, and on the German inflation of 1923, Costantino Bresciani-Turroni, The Economics of Inflation (London: George Allen & Unwin. 1931). pp. 80-1.

³I have never seen a reference to this striking event in any textbook on money. See e.g., The New York Times Jan. 30, 1945.

⁴ The Value of Money (New York: Richard R. Smith. 1917 and 1936), p. 161.

⁵ Milton Friedman and Anna Jacobson Schwartz, in their Monetary History of the United States.- 1867-1960 (Princeton University Press, 1963), do offer annual estimates and tables of "velocity of money- based on worksheets of Simon Kuznets made for another study. But they define this velocity as "the ratio of money income to the stock of money.- This hardly makes it a transactions velocity. Moreover, they appear to attach very little commodity-price determining importance to it: "Velocity is a relatively stable magnitude that has declined secularly as real income has risen.^" (p. 34).

⁶1 have treated this subject at greater length in an essay, "Velocity of Circulation," in Money, the Market and the State: Economic Essays in honor of -James Muir Waller, edited by N. A. Beadles and L. A. Drewry (University of Georgia Press, 1968.)

7 Reprinted in Dollars and Deficits (Englewood Cliffs, N.J.: Prentice-Hall, 1968), p. 152.

⁸ The Optimum Quantity of Money (Chicago: Aldine Publishing Co., l969), pp. 46-48.