“Must We Choose between Inflation and Unemployment?” by Milton Friedman Stanford Graduate School of Business Bulletin 35, Spring 1967, pp. 10-13, 40, 42 © The Board of Overseers of the Leland Stanford Junior University

We have been told repeatedly that, if we are to have full employment, the price we must pay is some inflation. On the other hand, we are told that the goal of price stability should be put above the goal of full employment. In a way the view that we must choose between unemployment and inflation, or between price stability and full employment, is something of a modern version of Karl Marx’s view about the reserve army of the unemployed. In his development, Marx believed that capitalism would always need to keep a large reserve army of the unemployed to keep the proletarians in their places. The modern version of that doctrine says that only a reserve army of unemployed will prevent workers and laborers, whether organized in unions or not, from pushing for ever higher wages. The widely expressed argument insists that if you seek to expand aggregate demand through the kind of governmental measures envisaged in the Full Employment Act, if you seek to expand aggregate demand beyond that point at which enough unemployment will keep workers from trying to get ever higher wages, if you try to push it beyond that point, you can reduce unemployment. But, the argument goes, the reduction of unemployment is possible only by paying the price of inflation. What you must do, the argument continues, is to push until you have enough demand so that prices are rising; with prices rising, you will keep unemployment low. You will get around the problem raised by the trade unions who are seeking higher wages. This notion has been described in various ways. In technical economic literature, the view is being labeled the Phillips Curve, after a New Zealander teaching at the London School of Economics. Bill Phillips wrote an article some years back in which he argued that there was a relationship between the rate at which money wages were rising on the one hand and the rate of unemployment on the other. He argued that at some level of unemployment in any community, let’s say five or six per cent, no general upward pressure on wages would exist. But, if unemployment were reduced below that point, this would tend to mean that wages would be rising. On the other side, if unemployment were above this level, wages would be falling. So the Phillips Curve is one kind of technical designation. In popular literature, the same phenomenon, called “cost-push inflation,” says that what produces inflation is the upward pressure of prices and wages. It depends on whom you are talking to. If the view is being expressed by, let us say, trade union people, they assert that it’s the grasping, monopolistic businessman trying to push up the price of his product. He’s the one producing inflation.

1

If you are talking to a businessman, he knows that the culprits are the grasping trade unionists who are trying to raise the wages that the businessman must pay, thus producing cost-push inflation. Described in another way, a trade-off exists between unemployment and inflation. Indeed, a variety of estimates is made of what the trade-off is. If you accept one percentage point more inflation, if you let prices go up at three per cent a year instead of two per cent, then it is argued you can keep unemployment maybe at four and a half per cent instead of at five per cent. The most popular judgment is that stable prices would imply a level of unemployment of five or six per cent; on the other hand, if you are going to keep the average level of unemployment to four per cent, this will require or will produce something like a three or four per cent per year price inflation. We ought to be willing, the argument goes, to bear the one for the gain in the other. Attempts have been made to devise schemes for reducing the alleged trade-off between the rate of change in prices and the level of unemployment. Actually, the favorite way for reducing the terms of trade between unemployment and price level change is the infamous “wage-price guidelines” enunciated some years back by the Council of Economic Advisers and now largely discarded. Confusion between Nominal and Real Wages This whole view that I have just described is fallacious from beginning to end, involving a confusion between nominal wages and real wages (the buying power of a worker’s wages). On the basis both of history and of economic analysis, there need be no relationship at all between the average level of unemployment on the one hand and the average behavior of prices on the other. A high level of unemployment on the average is possible while prices are rising rapidly. A low level of unemployment is possible while prices are rising rapidly. Equally, on the other side, a high level of unemployment is possible while prices are falling. A low level of unemployment is possible while prices are falling. So far as I can see, these two magnitudes, the average level of unemployment and the average secular trend of prices, are largely independent magnitudes determined by two different sets of forces. At any point of time in the year 1967, if you speed up inflation, if you increase the rate of expansion of nominal aggregate money demand more rapidly than it has been going up, you can undoubtedly lower the level of employment. If from a rate of inflation of about three per cent a year (as for the last couple of years), you advanced to a rate of five or six per cent, in the first instance you would undoubtedly make it look as if you were engaging in a boom, unemployment would go down, you would have low levels of unemployment. But suppose you hold the rate of inflation at five per cent. What would happen? We have been seeing what has been happening. Why have the wage-price guidelines gone by the board?

2

Because the workers and the unions have gotten the message. They are now asking for ten or twelve per cent instead of their formerly modest five and eight per cent. So that if you speeded up the rate of inflation and you kept it at, let’s say, five or six per cent, unemployment would gradually creep up and return to its present level as trade unions and employers got the idea and started adjusting their behavior in anticipation. The truth: Workers bargain for real wages, not for nominal wages. No worker cares how many pieces of paper he takes home. He cares how much those pieces of paper will buy. No businessman sells at bottom for nominal prices. No businessman cares fundamentally how many pieces of paper he gets in return for his product. He wants so much in the form of goods and services he can buy with the proceeds from his product. No worker is made wealthier by getting twice as many dollars and having to pay twice as high a price, and no businessman is better off selling his product for twice the price if he has to pay twice the wages and twice the cost. Result: Workers and businessmen ultimately bargain for real wages and real prices, and not for nominal wages and nominal prices. There is, in my opinion, no trade-off between inflation and unemployment. We do not have to choose between inflation and unemployment. From American experience we find periods of rapidly rising prices and high unemployment, falling prices and low unemployment. Inflation in Brazil and Japan An extreme and illustrative example was offered by Brazil a few years back. Brazil was having an inflation, not one of the mild kinds of inflations we have in this country. Prices were rising at the rate of something like 75 per cent a year, a fairly mild inflation as these historical episodes go, but sizeable by our experience. Brazil undertook measures to cut down the rate of inflation. They succeeded. They cut the rate of inflation down to something like 40 per cent, and the short-run immediate result was the emergence of immediate unemployment. If you were talking about a trade-off in Brazil, you would have had to say that the trade-off was between the 75 per cent a year rate of inflation and a zero unemployment. But what Brazil illustrates is that what matters is not the level of inflation, but whether it is higher or lower than before. A more sophisticated example is provided by the experience of Japan which has now gone through three business cycles in the last ten years. Their first recession in this period was associated with declining prices. So they stepped on the monetary accelerator and expanded the quantity of money very rapidly, resulting in price inflation and also economic expansion. Then they ran into balance of payments difficulties. They had to taper off. They did taper off. They went into another recession, but this time, because of the background of rising prices prior to recession, prices were either stable or rose a little bit. But as the recession proceeded, with a little unemployment (by their standard), and with their balance of payments back into equilibrium, they stepped on the accelerator again.

3

They went off on a still steeper course. Once again they ran into balance of payments problems and had to step on the brake. This time, prices rose at something like six or seven per cent a year in the midst of a slowdown in output and the emergence of some unemployment. Now, I cite Japan’s experience because I believe that is our experience, only we are in the early stages of that process. I believe there is no possible way at the moment for the United States to get on a stable price path without having a significant recession, which may continue longer than recent recessions. We must unfortunately go through another period like 1956–60 to break the pattern of price anticipations that has been formed and to get people to look forward to a world of favorable prices. They must be willing to engage in their bargaining and their activities on the basis of stable price anticipations. The way to achieve that would, in my opinion, be for the Federal Reserve Board to move toward a constant rate of growth in the quantity of money and keep it; to pick a rate of growth in the quantity of money which can be maintained indefinitely, not jiggled up or down. We have had, unfortunately, the practice of going too far first in one direction and then in the other. I would like to see them hold it there. To avoid misunderstanding, let me emphasize that I am not predicting that this suggestion will be followed. On the contrary, what will almost surely happen is this: As we get a recession developing, along with some unemployment, a widespread public clamor will arise to the effect that the Employment Act of 1946 requires that we expand the quantity of money, that we have government deficit spending, that we take any measures necessary to start output going up and avoid unemployment. We can do so, but only at the price of speeding up of inflation, and that speeding up of inflation will only give a temporary answer. We will be launched into a kind of Japanese sequence with a succession of expansions, each with more rapid inflation, and each acceleration of inflation produced by over-reaction to the recession which precedes it. But if we will, we can have stable prices by accepting the temporary cost of a period of relatively higher rates of unemployment—not a major depression, nothing like that, but something comparable to 1956–60. What about unemployment? The problem with unemployment is not a question of government’s policy on demand. Rather, it concerns the characteristics of the labor market. What is the unemployment situation today? We have very low unemployment among large sectors of the community, but we have very high unemployment among certain sectors, particularly among the unskilled, among the Negroes, among the young. Why? Because government has decided to produce unemployment among these groups. Inflation and unemployment of the existing kind are both made in Washington. The major reason for such a high level of unemployment among the Negroes, among the teenagers, and among the unskilled is because we have boosted minimum wage rates.

4

There is no mystery about the matter. When fully effective, the recent rise in the minimum wage rate will raise the unemployment level among Negro male teenagers to well above 30 per cent—now up to about 25–28 per cent. We also have the Walsh-Healey Act and the Davis-Bacon Act, other measures designed to produce unemployment. Nobody says that, of course. You never saw the preamble of any act saying, “This act is designed to produce unemployment”; but we must not look only at what people say. We should have a twin policy. To keep the average level of unemployment low and to eliminate unnecessary unemployment, we should have a policy of freeing the labor market, of trying to eliminate governmental interferences with wage rate determination, Walsh-Healy, Davis-Bacon, and so on. We should also try to reduce as far as possible private monopolistic interferences in the labor market, whether on the side of the labor market or on the side of enterprise. To produce relatively stable prices, we should look to the people who are really responsible for inflation. They are the people who print the pieces of paper we call our money. The appropriate way to control inflation is to hold down the rate at which the quantity of money increases and to keep it rising at a rate which will provide enough money to match our growing output, but not enough to produce price inflation. In that way we could reconcile stable prices with full employment. 10/5/12

5