Monetary policy and the legacy of Milton Friedman.

Schwartz, Anna J.

So many well-deserved tributes have been paid to the memory of Milton Friedman that I propose to pay him tribute in a special way by talking about my long association with him. Before I do so, I note his activities before the 1950s, when we started working together. For the record, Friedman was not a monetarist when our collaboration began.

The National Bureau of Economic Research was then located at 1819 Broadway, now the site of the Time Warner Center. At that time, the research program of that organization was divided between the measurement of national income and the study of business cycles.

Friedman's Activities before the 1950s

Friedman had had an earlier connection with the NBER related to its research on the measurement of national income. This association came about in an unforeseeable manner as a result of a job he obtained in 1935 after finishing his second year of graduate study at the University of Chicago. The job at the National Resources Committee was in Washington in the heyday of the New Deal. The director of the NRC was charged with the planning of a nationwide study of consumer purchases to be funded by the Works Progress Administration that would offer jobs to unemployed individuals in selected communities in many states to collect information on their income and expenditures from a wide sample of families. Friedman was a member of a group at the NRC assigned to prepare the questionnaire, design the sample, plan the tabulation, and analyze the results. Friedman left the NRC in 1937 when the planning phase ended.

In 1936 Simon Kuznets, an NBER staff member, organized the Conference on Research in Income and Wealth to bring together academic and government agency individuals engaged in work on national income and its distribution. Friedman attended the first meeting in early 1937 as a representative of the NRC. Later that year he became Kuznets's assistant at the NBER. His first task was to edit the proceedings of the spring 1937 conference. As secretary of the conference, he edited the second and third of the proceedings meetings, and until 1945 was active in the conference, when he left to do war research.

Friedman's major accomplishment at the NBER during this period was his revision of a preliminary draft by Kuznets of a manuscript on incomes of five independent professional practitioners that the latter had collected. The NBER published the book as co-authored. The Friedmans left New York in 1940 for Madison, Wisconsin, where he had a one-year teaching appointment at the University of Wisconsin that for various reasons was a disappointment. In 1941 he collaborated with two others on a study of taxes to prevent inflation, a view he would later decry. The study led to an offer of a job with the Division of Tax Research at the Treasury Department. Here he was involved in reforming the tax system in order to win the war. He left the Treasury in 1943 to take a job as a mathematical statistician with the Statistical Research Group at Columbia to do war research. At the war's end, Friedman was offered a one-year teaching position at the University of Minnesota. In 1946, he accepted a professorship at the University of Chicago. In the fall quarter of 1950, he took a leave of absence to spend as a consultant to the Marshall Plan agency in Paris.

Collaborating with Friedman

It was in 1950 that Arthur Burns, then Wesley Mitchell's successor as the NBER director of research, approached Friedman to resume his connection with the Bureau by directing the study of the role of money in business cycles. I had joined the Bureau in 1941 and had begun some independent work on studies of banks and measures of reserves. Burns thought Friedman and I would make a good team.

Staff members were assigned the task of describing and analyzing the cyclical behavior of an economic variable--for example, hours worked, the unemployment rate, and total production. The variable, whose cyclical behavior Friedman and I were asked to study was money. We collaborated for the next three decades on three NBER books, and on journal articles before and after the books were published.

Friedman was located in Chicago and I was in New York, but our geographical separation was no impediment to working together. How did we communicate? In those days, long-distance phone calls were expensive, so we rarely used the phone, but letter postage stamps cost three cents, so we exchanged letters, discussing what needed to be done, and reporting progress, sometimes painfully slow, sometimes setting aside an initial effort to solve a problem, and trying again to achieve what we believed was a better solution.

The research effort we undertook was not limited to our two-person team. Friedman's students in the Chicago Money Workshop contributed estimates of various measures of the money stock for early dates of the series covering the period from 1867 to 1960 that was the bedrock of our study of the behavior of money and of its impact on the economy. Some members of the Workshop wrote dissertations that explored ideas that enriched our analysis. With Friedman as our leader the project developed an esprit that we were in a battle with economists who were not sympathetic to our emphasis on the importance of money. As our investigation progressed, it yielded interesting empirical evidence, so it is not surprising that it evoked a response from us that paralleled what young economists felt on first reading The General Theory. We, like them, quoting Wordsworth's words, felt "Bliss it was in that dawn to be alive, but to be young was very heaven."

Responding to Krugman's Attack on Friedman

Let me conclude this recollection of the onset of my collaboration with Friedman by noting that Edward Nelson of the St. Louis Federal Reserve Bank and I recently completed a draft of an article responding to Paul Krugman's attack on Friedman, entitled "Who Was Milton Friedman?" that The New York Review of Books published in February 2007.

According to Krugman, the generally successful monetary policies observed in the United States and other countries since the 1980s amount to an unambiguous defeat for Friedman and monetarism. Krugman's discussion is confused, but he does certainly create the clear impression that monetary policy since the 1980s constitutes a return to the pre-Friedman pre-monetarism status quo. But the last 20 years have not seen a return to the wage-price guidelines and wage-price controls of the 1960s and 1970s, nor have they been characterized by anything other than wide acceptance of Friedman's position that controls and guidelines were ineffective ways to fight inflation. Replacement of those failed measures with arrangements in which central banks accept responsibility for inflation control is a major legacy of Friedman and monetarism.

Our article does two things: it corrects Krugman's mischaracterizations of Friedman's legacy and also of monetarism's legacy, and it traces the impact on modern monetary economics of Friedman's ideas on inflation beginning with his critique of the macroeconomic polities pursued in the United States in the 1960s and 1970s, in opposition to Keynesian doctrines and policies. In addition, the article demonstrates Krugman's inaccurate forays into economic history by attributing the depth and duration of the U.S. Great Depression of the 1930s and Japan's extended slump in the 1990s to a liquidity trap.

Our article also reviews Friedman's theoretical debates on inflation, and his rejection of both the cost-push and simple Phillips curve approaches that were emblematic of Keynesian 1970s inflation analysis. We describe the steps by which Friedman modified the simple Phillips curve and his criticism of Keynesian "patched up" versions of the Phillips curve. We discuss the 1970s debates on price and wage controls as a solution to inflation. Friedman rejected cost-push as a credible source of sustained inflationary pressure. Therefore, he saw no justification for incomes policy. His opposition to incomes policy, his rejection of cost-push accounts of inflation behavior, and his calls for monetary restraint were in contrast with the positions of leading Keynesians (James Tobin, Paul Samuelson, Arthur Okun, and Walter Heller).

Friedman's perspective had a more durable influence on anti-inflation policy than the cost-push and incomes-policy perspective taken by his 1970s critics. Central banks' adoption of inflation targeting in recent decades reflects an acceptance of Friedman's position that monetary restraint is both necessary and sufficient for inflation control.

Much of the discussion of monetarism in the 1970s policy debates was formulated in terms of monetary aggregates, but it was clear even in the 1970s that the distinguishing feature of monetarism was the responsibility it assigned to monetary policy for the control of inflation.

Friedman's influence on monetary policy analysis is not limited to his positions on the causes of inflation and on the need for inflation-oriented monetary policy rules. Other aspects of Friedman's work have influenced modern thinking about monetary policy. Some examples are (1) the Fisher effect and the nominal and real interest rate distinction, (2) distortions to the relative price structure produced by inflation that imposed costs in the form of a lower natural level to output and a higher natural rate of unemployment, (3) lags in monetary policy, (4) fragility of measures of potential output and the output gap, (5) benefits of flexible exchange rates, and (6) rejection of credit controls.

Krugman is not a specialist in monetary economics. His lack of appreciation of the contributions of monetarism may partly reflect the fact that many monetarist ideas were already being incorporated by moderate Keynesians into their analysis by the time Krugman became active in economies. Yet it is a puzzle that Krugman arrived at so egregious a misunderstanding of the spirit and content of modern monetary policy and of their connections to Friedman's economies. One possible explanation is that Krugman is unfamiliar with the literature of the past two decades incorporating monetarist perceptions into Keynesian and New Keynesian economies, but that he acquired superficial information about two of Friedman's positions that central banks did not embrace. In particular (1) Friedman urged the Fed to adopt a constant growth rate for a monetary aggregate as a faute de mieux policy that would reduce the volatility of money growth; (2) Friedman urged central banks to substitute as their instrument the growth rate of a monetary aggregate instead of the overnight short-term nominal interest rate. If Krugman regarded these two positions as the essence of monetarism, he could jump to the conclusion that monetarism was dead.

It is worth noting the reasons Friedman adopted these positions. Historically, major mistakes in monetary policy occurred in periods when central banks associated low nominal interest rates with easy policy, ignoring the signal--coming from a stagnant or declining money stock--that monetary policy was tight, or when they interpreted a high nominal rate as implying tight monetary policy, even when rapid money growth and inflation indicated the opposite. In addition, the use of the nominal interest rate as the instrument could lead to situations where the rate was pegged, even though macroeconomic stabilization required a changing interest rate. For these reasons, Friedman regarded a monetary aggregate instrument--with the aim of a constant money growth rate--as superior to an interest rate instrument.

These policy positions, however, were never the central core of monetarist doctrine. Instead, monetarist propositions were about the structure of the economy and the effects of monetary policy. Friedman (1983: 4) noted that while he favored a constant money growth rate, some monetarists favored a different rule for monetary growth. A particular prescription was slow, steady money growth. Friedman's policy prescriptions of constant money growth and money aggregate instrument have been rejected, but many core theoretical and empirical propositions have been incorporated into the mainstream. Two of the most important monetarist propositions--the nominal rate/real rate distinction and the need for inflation control to be assigned to monetary policy--now guide the formulation of interest-rate policy by central banks in a way that they did not in the 1970s. Consequently, the recommendation that central banks move to a money aggregate instrument has fallen by the wayside.

Friedman understood that interest-rate instrument rules could in principle deliver stable inflation, and that the choice between the interest rate and the money aggregate as an instrument was a tactical, not a strategic matter. But actual experience with interest rate rules in most countries up to the late 1970s was discouraging. Policymakers apparently were unwilling or unable to make interest-rate decisions needed to restore price stability. A money aggregate rule had merit as an automatic means of delivering the needed movements in interest rates. The switch in many countries after the late 1970s to more stabilizing interest-rate rules did not come about by accident, but arose from acceptance of core monetarist propositions. the foundation of that regime change and of interest-rate decisions today, is acceptance of Friedman's position that monetary restraint is a sufficient condition for controlling inflation.

Financial innovations, such as sweeps programs and interest payments on money, have weakened the relationship between many monetary aggregates and nominal GDP. Innovations affected M1 in the United States more seriously than others with the result that defining money has become a more difficult empirical task. In light of these developments, the most durable aspects of monetarist theory are those that hold even in environments where there are not reliable money data. These aspects of monetarism, with particular reference to Friedman's work, have been listed above.

Yet financial change in itself is not a legitimate reason for not devoting resources to the careful measurement and study of money, nor is it a basis for ignoring monetary aggregates when making policy decisions. What is more, the usefulness of money (both as an indicator and as a policy instrument) is likely to increase when short-term nominal interest rates reach very low values, as in the case of Japan in the 1990s and the United States during the Great Depression. Monetary aggregates may 'also be valuable data for central banks that have reason to intervene in the foreign exchange market. Information on base money becomes useful because central-bank sterilization of the exchange transaction--that is, operations in domestic securities that offset the impact of the foreign exchange operation on the aggregate level of base money--may be the most reliable means of ensuring that the intervention does not produce an unintended change in aggregate demand.

The judgment on Krugman's mischaracterization of the economics of Friedman and of monetarism is that Krugman does not speak authoritatively on subjects on which he has no expertise. His scholarly work on trade does not qualify him as an authority on Friedman's work, which he rarely mentioned in his publications. When he did, he acknowledged the contributions of Friedman and monetarism in a way that contradicts his New York Review of Books essay on Friedman. Friedman's reputation is intact despite Krugman's efforts to denigrate him and his contributions.

References

Friedman, M. (1983) "Monetarism in Rhetoric and in Practice." Bank of Japan Monetary and Economic Studies 1 (2): 1-14.

Friedman, M., and Friedman, B. D. (1998) Two Lucky People: Memoirs. Chicago: University of Chicago Press.

Krugman, R (2007) "Who Was Milton Friedman?" New York Review of Books 54 (2) (15 February): 27-30.

Nelson, E., and Schwartz, A. J. (2007) "The Impact of Milton Friedman on Modern Monetary Economies. Setting the Record Straight on Paul Krugman's "Who Was Milton Friedman?'" NBER Working Paper No. 11546 (October). Cambridge, Mass.: National Bureau of Economic Research.

Anna J. Schwartz is a Research Associate at the National Bureau of Economic Research.