By Bennett T. McCallum
Monetarism is a macroeconomic school of thought that emphasizes (1) long-run monetary neutrality, (2) short-run monetary nonneutrality, (3) the distinction between real and nominal interest rates, and (4) the role of monetary aggregates in policy analysis. It is particularly associated with the writings of Milton Friedman, Anna Schwartz, Karl Brunner, and Allan Meltzer, with early contributors outside the United States including David Laidler, Michael Parkin, and Alan Walters. Some journalists—especially in the United Kingdom—have used the term to refer to doctrinal support of free-market positions more generally, but that usage is inappropriate; many free-market advocates would not dream of describing themselves as monetarists.
An economy possesses basic long-run monetary neutrality if an exogenous increase of Z percent in its stock of money would ultimately be followed, after all adjustments have taken place, by a Z percent increase in the general price level, with no effects on real variables (e.g., consumption, output, relative prices of individual commodities). While most economists believe that long-run neutrality is a feature of actual market economies, at least approximately, no other group of macroeconomists emphasizes this proposition as strongly as do monetarists. Also, some would object that, in practice, actual central banks almost never conduct policy so as to involve exogenous changes in the money supply. This objection is correct factually but irrelevant: the crucial matter is whether the supply and demand choices of households and businesses reflect concern only for the underlying quantities of goods and services that are consumed and produced. If they do, then the economy will have the property of longrun neutrality, and thus the above-described reaction to a hypothetical change in the money supply would occur.1 Other neutrality concepts, including the natural-rate hypothesis, are mentioned below.
Short-run monetary nonneutrality obtains, in an economy with long-run monetary neutrality, if the price adjustments to a change in money take place only gradually, so that there are temporary effects on real output (GDP) and employment. Most economists consider this property realistic, but an important school of macroeconomists, the so-called real business cycle proponents, denies it.
Continuing with our list, real interest rates are ordinary (“nominal”) interest rates adjusted to take account of expected inflation, as rational, optimizing people would do when they make trade-offs between present and future. As long ago as the very early 1800s, British banker and economist Henry Thornton recognized the distinction between real and nominal interest rates, and American economist Irving Fisher emphasized it in the early 1900s. However, the distinction was often neglected in macroeconomic analysis until monetarists began insisting on its importance during the 1950s. Many Keynesians did not disagree in principle, but in practice their models often did not recognize the distinction and/or they judged the “tightness” of monetary policy by the prevailing level of nominal interest rates. All monetarists emphasized the undesirability of combating inflation by nonmonetary means, such as wage and price controls or guidelines, because these would create market distortions. They stressed, in other words, that ongoing inflation is fundamentally monetary in nature, a viewpoint foreign to most Keynesians of the time.
Finally, the original monetarists all emphasized the role of monetary aggregates—such as M1, M2, and the monetary base—in monetary policy analysis, but details differed between Friedman and Schwartz, on the one hand, and Brunner and Meltzer, on the other. Friedman’s striking and famous recommendation was that, irrespective of current macroeconomic conditions, the stock of money should be made to grow “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.”2 Brunner and Meltzer also favored monetary policy rules but recognized the attractiveness of activist rules that relate money growth rates to prevailing economic conditions. Also, they typically concentrated on the monetary base, adjusted to reflect changes in reserve requirements, whereas Friedman was more concerned with M2 or M1 and, indeed, sought major changes in banking legislation, such as 100 percent reserve requirements on deposits, designed to make the chosen aggregate precisely controllable.
Friedman’s constant-money-growth rule, rather than other equally fundamental aspects of monetarism, attracted the most attention, thereby detracting from the understanding and appreciation of monetarism. In particular, this led to the comparative neglect of Friedman’s crucial “accelerationist” or “natural-rate” hypothesis, according to which there is no long-run trade-off between inflation and unemployment; that is, the long-run phillips curve is vertical. The no-trade-off view was also promoted by Brunner and Meltzer. Accordingly, it might be argued that the two fundamental monetarist propositions are (1) that cyclical movements in nominal income are primarily attributable to movements in the stock of money, and, (2) that there is no permanent trade-off between unemployment and inflation. Together, these lead to monetarist-style policy positions.
Monetarism’s rise to intellectual prominence began with writings on basic monetary theory by Friedman and other University of Chicago economists during the 1950s, writings that were influential because of their adherence to fundamental neoclassical principles. The most outstanding in this series was Friedman’s presidential address to the American Economic Association in 1967, published in 1968 as “The Role of Monetary Policy.” In this paper Friedman developed the natural-rate hypothesis (which he had clearly stated two years earlier) and used it as a pillar in the argument for a constant-growth-rate rule for monetary policy. Almost simultaneously, Edmund Phelps, who was not a monetarist, developed a similar no-trade-off theory, and, within a few years, events in the world economy apparently provided dramatic empirical support.
In the late 1970s and early 1980s, after a decade of increasing influence, monetarism’s reputation began to decline for three main reasons. One was the growing belief, based on plausible interpretations of experience, that money demand is in practice highly “unstable,” shifting significantly and unpredictably from one quarter to the next. The second was the rise of rational expectations economics, which split analysts antagonistic to Keynesian activism into distinct camps. (A majority of monetarists themselves soon embraced the rational expectations hypothesis.) The third was the Federal Reserve’s famous “monetarist experiment” of 1979–1982. The latter episode warrants an extended discussion.
During the 1970s, inflation rose in the United States, as well as in many other industrial nations, to levels unprecedented on a multiyear basis during periods of relative peace. This occurred as a consequence of various “shocks”—oil price increases, the Vietnam War, and especially the 1971–1973 demise of the Bretton Woods system of fixed exchange rates (itself caused largely by the failure of the United States to maintain the gold value of the dollar). This demise left central bankers with a major new responsibility; namely, to provide a nominal anchor for national fiat currencies to replace the gold standard. The Federal Reserve announced several times during the 1970s that it intended to bring inflation under control, but various attempts were unsuccessful. Then, on October 6, 1979, the Fed, under Paul Volcker’s chairmanship, announced and put into effect a new attempt involving drastically revised operating procedures that had some prominent features in common with monetarist recommendations. In particular, the Fed would try to hit specified monthly targets for the growth rate of M1, with operating procedures that emphasized control over a narrow and controllable monetary aggregate, nonborrowed reserves (i.e., bank reserves minus borrowings from the Fed). The M1 targets were intended to bring inflation down from double-digit levels to unspecified but much lower values.
In retrospect, the events that occurred from October 1979 to September 1982 are widely viewed as the crucial beginning of a necessary and successful attack on inflation that led, eventually, to the worldwide low-inflation environment of the 1990s. At the time, however, the “experiment” seemed anything but successful to many Americans. Short-term interest rates jumped dramatically in late 1979 under the tightened conditions, and 1980 witnessed a major fall in output in one quarter followed by a major jump in the next, due primarily to the imposition, and then removal, of credit controls. Finally, in 1981 and into the middle of 1982, a sustained period of monetary stringency brought about the deepest recession since the Great Depression of the 1930s and began to bring inflation down, more rapidly than many economists anticipated, toward acceptable values. Some relevant statistics, designed to put the episode in perspective, are reported in Table 1. Throughout the entire episode, moreover, both interest rates and money growth rates were highly variable. To be specific, the standard deviation of monthly percentage growth rates of M1 increased from 3.73 (for the previous three years) to 8.22 for the period from October 1979 to September 1982, while the standard deviation of monthly percentage changes in the federal funds rate jumped from 2.86 to 23.1 (annualized units).
|Year or Average Over||CPI Inflation Dec.–Dec.||Fed Funds Rate||“Real” Funds Rate||M1 Growth Rate||Adjusted M1B Growth||Unemployment Rate|
|Sources: Economic Report of the President, 2003, except: Adjusted M1B Growth: Alfred Broaddus and Marvin Goodfriend, “Base Drift and the Longer Run M1 Growth Rate: Experience from a Decade of Monetary Targeting,” Federal Reserve Bank of Richmond Economic Review 70 (November–December 1984), pp. 3–14.|
|Note: Adjustments to M1 growth rates were made and used by the Fed, at the time, to take account of estimated distortions brought about by the introduction of NOW (negotiated order of withdrawal) accounts.|
Many critics characterized the “experiment” as a macroeconomic disaster. Some believed, moreover, that it provided strong and definitive evidence invalidating monetarism—partly by showing how undesirable it was to have money growth targets and partly in showing how poor are operating procedures for controlling M1 money growth by means of tight control of a narrow monetary aggregate. Monetarists argued that the episode was actually not monetarist in its design because growth rates of M1 fluctuated very widely on a month-to-month basis; the operating procedures in place were, because of lagged reserve requirements, extremely poorly designed for the control of M1; and the Fed never forswore discretionary responses to current cyclical conditions. It now seems clear that the Fed’s use of a narrow monetary aggregate for week-to-week control was highly effective in terms of public relations. The reason was that it permitted the Fed to escape political responsibility for the resulting high and, therefore, unpopular interest rate levels by claiming that these were simply the consequence of market forces.3 At the same time, by adopting a putatively monetarist approach, the Fed could at least, even if the episode was a failure, discredit monetarism and the Fed’s annoying monetarist critics. As matters played out, the episode was soon seen as a strategic success, despite temporary unhappiness, and monetarism was discredited as well!
What is left today of monetarism? While some disagreement remains, certain things are clear. Interestingly, most of the changes to Keynesian thinking that early monetarists proposed are accepted today as part of standard macro/monetary analysis. After all, the main proposed changes were to distinguish carefully between real and nominal variables, to distinguish between real and nominal interest rates, and to deny the existence of a long-run trade-off between inflation and unemployment. Also, most research economists today accept, at least tacitly, the proposition that monetary policy is more potent and useful than fiscal policy for stabilizing the economy. There is some academic support, and a bit in central bank circles, for the real-business-cycle suggestion that monetary policy has no important effect on real variables, but this idea probably has marginal significance. It is hard to believe that the major recession of 1981–1983 in the United States was not caused largely by the Fed’s deliberate tightening of 1981—a tightening that shows up in ex-post real interest rates and in M1B growth rates as adjusted by the Fed at the time (Table 1, column 6) to take account of major institutional changes.
In 2005, most academic specialists in monetary economics would probably describe their orientation as new keynesian. Also, monetary aggregates currently play a small or nonexistent role in the monetary policy analysis of academic and central-bank economists. In terms of its underlying scientific rationale, however, today’s mainstream analysis is much closer to that of the monetarist than the Keynesian position of, for example, 1956–1978. In addition to the points noted above, current thinking clearly favors policy rules in contrast to “discretion,” however defined, and stresses the central importance of maintaining inflation at quite low rates. It is only in its emphasis on monetary aggregates that monetarism is not being widely espoused and practiced today.
Milton Friedman, Capitalism and Freedom (Chicago: University of Chicago Press, 1962), p. 54.
This claim was somewhat deceptive in the following way. Each period’s interest rate level was indeed market determined once the quantity of nonborrowed reserves for the period was determined, but this quantity was itself set by the Fed.