A Rehabilitation of Monetary Policy in the 1950s

Policymakers in the 1950s raised nominal interest rates more than one-for-one with increases in expected inflation, meaning that the FOMC increased the real federal funds rate. That is similar to the 1980s and 1990s.

Would William McChesney Martin, who was chairman of the Federal Reserve Open Market Committee (FOMC) during the 1950s, feel at home at a modern day FOMC meeting? If Alan Greenspan, or his predecessor Paul Volcker, were transported back to the 1950s' Federal Reserve, would they fit in? Based on the evidence presented by Christina Romer and David Romer in a new NBER Working Paper, the answer is they probably would. In A Rehabilitation of Monetary Policy in the 1950s (NBER Working Paper No. 8800), Romer and Romer show that the FOMC of the 1950s had similar priorities and a similar approach to setting interest rates as the FOMC of the last 20 years.

These findings contrast with the standard characterization of monetary policymakers in the 1950s as unsophisticated, inept, or both. Economists tend to portray the 1950s' Fed in uncomplimentary terms, or to ignore it in studies of post-war monetary policy. The starting point of Romer and Romer's analysis is the macroeconomic evidence: between 1952 and 1960, average annual inflation was less than 2 percent, the U.S. economy expanded at an average annual rate of 2.9 percent, and unemployment averaged 4.7 percent. Such good macroeconomic performance is not proof that macroeconomic policymaking was similarly good, but it certainly suggests that the conventional wisdom should be revisited.

Romer and Romer proceed along two paths. First, they examine the Federal Reserve records from the 1950s, chiefly FOMC minutes and Federal Reserve officials' Congressional testimony. They show that FOMC deliberations were actually quite sophisticated. According to Romer and Romer, many statements by FOMC members from the 1950s could be inserted into the narrative record of the 1980s and 1990s without anyone noticing. Second, they conduct statistical analysis of the 1950s data. Estimates of the Taylor rule, which relates changes in interest rates to changes in output and inflation, confirm the similarities between Fed policymaking under chairman Martin and under chairmen Volcker and Greenspan.

The narrative evidence suggests that policymakers in the 1950s can be best characterized as having a deep-seated dislike of inflation, which they preemptively acted to control. They emphasized the costs of inflation and the absence of a long-run trade-off between output and inflation. This is very similar to the rhetoric of the 1980s and 1990s. It is the 1960s and 1970s, by contrast, when the FOMC's framework looks old-fashioned, first in its use of a naive Keynesian model with an exploitable trade-off between output and inflation, and subsequently with a natural rate model with an unrealistically low estimate of the natural rate. The result was the high inflation of the late 1960s and 1970s.

The FOMC of the 1950s was not exclusively concerned with inflation - it also expressed concern about unemployment and growth frequently - but its deliberations suggest it was prepared to overlook these concerns if it thought that inflation was about to rise. For example, following a mild recession in 1958 the records show that the FOMC started worrying about inflation in the spring, as soon as it thought the recession had reached its trough. By September 1958 rates were back to their peak level of the previous year. Chairman Martin summed up the attitude at one FOMC meeting: "Inflation is a thief in the night and if we don't act promptly and decisively we will always be behind."

In the statistical analysis, Romer and Romer look at how the Federal funds rate responded to developments in the macroeconomy in the 1950s, and compare that relationship with later periods. They show that policymakers in the 1950s raised nominal interest rates more than one-for-one with increases in expected inflation, meaning that the FOMC increased the real federal funds rate. That is similar to the 1980s and 1990s. During the 1960s and 1970s, by contrast, policymakers responded less aggressively to rising inflation expectations. Then the coefficient was less than one, indicating that the real federal funds rate fell as inflation rose.

Policymakers in the 1950s acted similarly to their modern day counterparts, based on the researchers' Taylor rule estimates. The main difference is that the members of the FOMC of the 1950s thought that the negative effects of rising inflation were felt very quickly. Few modern day economists take such an approach. Yet, unlike their successors in the 1960s and 1970s, policymakers did not believe in a long-run positive trade-off between output and inflation.

-- Andrew Balls