A History of the Federal Reserve, Volume 2

by Allan H. Meltzer

  The remaining large puzzle is to explain why this happened. Why did the Federal Reserve dismiss for years the long-run vertical Phillips curve, procyclicality of monetary policy, and the effect of inflation on nominal interest rates, wages, and anticipations more generally? Why did they ignore the finding of some of their own staff that concluded that the long-run Phillips curve was vertical? Critics pointed out these errors at the time. Propositions that attribute the Great Inflation to analytical errors of one kind or another ought to be supplemented by an explanation of why the error persisted for fifteen years before policy changed. As is well-known, policymakers began anti-inflation policies as early as 1966 and several times after—1969, 1973, 1978–79, and 1980. They were aware of the Great Inflation but, until 1979–82, they did not persist in policies to end it.

  My main objection to explanations based on persistent policy errors is that they are incomplete. Federal Reserve officials could observe inflation rates. They knew that their policies had not ended inflation. Most often inflation was above their forecast. Chart 7.9 shows that the forecast errors were too large and persistent to be ignored. Yet the System did not change course. Burns was a distinguished economist, influenced more by data and induction than by deductive theories. Yet he failed to stop the inflation and, at times, saw it rise to rates never before experienced in U.S. peacetime history. Most of the FOMC members were not ideologues or slavish adherents to a particular theory. Several regarded themselves as practical men, meaning not attached to any particular theory and willing to discard analyses that did not work.

  One additional caveat is that the Federal Reserve is not a monolith. Members of the Federal Open Market Committee (FOMC) have independent views. Particularly in the 1960s, they were mostly non-economists. They had considerable difficulty agreeing on how to implement actions, as Maisel (diary, 1973) documents fully. The staff, or part of it, had a model, but insiders who have written about the 1960s and 1970s often emphasize inconsistency in the choices made by the FOMC (Lombra and Moran, 1980; Pierce, 1980, and Maisel, various years).

  The international character of the Great Inflation is sometimes advanced as support for explanations based on errors in economic theory. The claim is that many countries made the same errors, particularly denial of the natural rate hypothesis, claiming that unemployment in the long run was independent of inflation. All experienced inflation. Once policymakers everywhere accepted the natural rate hypothesis, time inconsistency theory, understanding of the need for credibility, and rational expectations, inflation declined.

  Appealing as this argument is to economists, it fails to separate the start of inflation and its continuance. The start of inflation occurred under the Bretton Woods system of fixed exchange rates. Surplus countries experienced inflation because they would not appreciate their currencies to stop the inflation, and those that did appreciate made at most modest increases until 1971. They were fully aware of the problem; they did not want a solution that reduced their exports or slowed the growth of output and employment. They opposed dollar depreciation. Once the fixed exchange rate system ended, Japan, West Germany, Switzerland, and Austria reduced their inflation rates. Others permitted inflation to continue or increase.

  The United Kingdom was the principal deficit country besides the United States. It comes closest to support the policy errors (or preferences) explanation. Policymakers in both U.K. parties accepted and used a simple Keynesian model. The long delay of sterling devaluation from 1964 to 1967 and the policy measures chosen are evidence of the reluctance to slow growth (Nelson, 2003b).

  The Great Inflation resulted from policy choices that placed much more weight on maintaining high or full employment than on preventing or reducing inflation. For much of the period, this choice reflected both political pressures and popular opinion as expressed in polls. Many accepted James Tobin’s view that inflation would increase before the economy reached full employment, or they claimed that eliminating inflation required an unacceptable increase in unemployment. Inflation did not fall permanently until public opinion polls showed the public willing to bear the cost. Then it became acceptable politically to shift more weight to inflation control and less to unemployment when choosing monetary policy actions.

  THE 1973–74 RECESSION

  Neither the administration nor its critics and outside observers anticipated the 14 percent surge in food prices followed by a surge in oil prices in 1973.19 None predicted the steep decline in output that followed.

  The 1973–74 increase in consumer prices was the largest since 1947, following the end of wartime price controls. This time removal of price controls, devaluation of the dollar after 1971, poor harvests abroad, and increases in oil prices added to the underlying rate of measured inflation resulting from the excessive monetary and fiscal expansion of 1972. Although some of the Board’s staff recognized the distinction (Pierce and Enzler, 1974), the administration and the Federal Reserve did not distinguish between these sources of rising prices. Monetary contraction—slower monetary growth—could reduce the maintained rate of inflation driven by growth of aggregate demand by reducing growth of aggregate demand. Using monetary policy to counter the price level increase resulting from one-time reductions in the supply of food or fuel lowered the equilibrium price level by reducing aggregate demand. Reductions in both supply and demand induced reductions in output and employment. A proper policy response to the oil price increase would have recognized that it was a tax on oil users paid to foreign producers. To reduce the loss of welfare, the administration could have reduced domestic tax rates.20 Or it could do nothing about the price level shock and allow it to pass through the economy. It was not a monetary problem, as some FOMC members recognized.21

  19. Wells (1994, 111) quotes Burns as offering a political explanation of rapid money growth but not attributing the increase to the election. “Fundamentally, it is the expansion of the money supply over the long run that will be the basic cause of inflation.” To slow its growth, the Federal Reserve had to raise interest rates but, Burns said: “In the United States, there is an acute political problem regarding interest rates. Congress and the unions have become quite concerned about interest rates. Congress could unwisely legislate statutory limits on interest rates, or fail to act promptly to approve extension of the Economic Stabilization Act [controls].” This shows a willingness to sacrifice independence.

  20. The source of this error was failure to distinguish between the maintained or persistent inflation rate and other increases in the reported price level. An anti-inflation policy can have different goals. It can keep the maintained rate of inflation at zero and permit price level shocks to drive the price level to fluctuate around an expected zero long-run rate of inflation. Or it can aim for price level stability by offsetting increases and decreases in the price level. The second option probably requires larger changes in output and employment.

  21. Members of the staff correctly analyzed the oil shock and other one-time changes as affecting the price level, temporarily raising the rate of price change. See Pierce and Enzler (1974). I am grateful to David Lindsey for this reference. There is little evidence that the FOMC accepted this analysis or discussed its implication for their response.

  Secretary George Shultz and Council members Paul McCracken and Herbert Stein recognized in 1971 that a price and wage control program carried the risk that monetary and fiscal policy would be overly expansive (Hargrove and Morley, 1984, 356). Nevertheless, that was what they urged.

  In addition to policy errors, administration economists relied on faulty data. They believed that the economy had idle capacity that later data revisions removed. They believed that with idle capacity and a 6 percent unemployment rate at the start of 1972, their stimulus program would create jobs not inflation. Stein later recognized their error. “We all thought, ‘we’re a long way from full employment . . . and the inflation rate is low’” (Hargrove and Morley, 1984, 396).

  Once the election was over and memb
ers recognized the consequences of their actions, the administration reversed course. The president proposed reductions in government spending, and he tried to impound $12 billion in spending that Congress had authorized. He determined to hold fiscal 1974 spending to $269 billion.22 Money growth declined also. From a peak twelve-month rate of increase of 9.3 percent in July 1973 annual base growth fell to less than 6 percent in late 1975 after controls relaxed or ended. Chart 7.4 above shows that annual growth of the real base turned negative in the fall of 1973. It remained negative for almost two years. Chart 7.10 compares real base growth to the real interest rate. Real base growth reached its trough in January 1975, two months before the National Bureau recorded a trough to the recession. As in several earlier recessions, real ten-year interest rates (using predicted inflation) remained in a narrow range. These rates declined modestly before the recession, remained in a narrow range during the recession, again declined modestly before the trough, and rose during the early months of recovery. Chart 7.10 shows these data.

  The Council of Economic Advisers forecast for 1973 that inflation would fall to 2.5 percent and that real growth would approach full employment, still considered a 4 percent unemployment rate. Instead, the four-quarter average increase in the GNP deflator reached more than 8.25 percent and rose 10 percent (annual rate) in the fourth quarter. Real GNP rose 3.5 percent, but in the last three quarters, the average rate fell to 1.4 percent. The unemployment rate remained about 5 percent. Despite the commitment to reduce inflation, the Society of Professional Forecasters raised the expected four-quarter inflation rate from 3.7 percent to 5.4 percent between fourth quarter 1972 and 1973. A year later, it reached 7.7 percent. Hourly wage growth increased from 6.3 to 8.1 percent, and ten-year interest rates rose 0.5 percentage points to 7.4 percent.

  22. He achieved his spending limit, but Congress resisted his unprecedented use of impoundment. They removed this power. Earlier in October 1972, Congress increased social security payments by 20 percent just before the election and for the first time indexed social security benefits so they would increase with prices automatically.

  Food price increases that began with large grain purchases by the Soviet Union ended with the new harvest. Oil price increases began at about the same time. At first oil prices more than doubled to $5 a barrel. By year-end 1973, the oil price reached $11.65 a barrel. Instead of allowing domestic gas and oil prices to reflect market prices, the administration maintained price controls. These controls began in August 1971, so they froze prices when heating oil was relatively cheap seasonally and gasoline relatively expensive. Subsequent supply problems reflected this policy. Heating oil became scarce in the winter, when its price usually rose.23

  The National Bureau dates the peak of the expansion in November 1973 and the trough of the recession sixteen months later in March 1975. The recession was the longest since the Great Depression and the deepest since 1937–38. Real output fell 4.9 percent, more than in any postwar recession to date (including 1981–82). The unemployment rate reached 9 percent in May 1975 after the recession ended.

  23. On October 23, 1973, most members of OPEC, led by Saudi Arabia, imposed a boycott on exports to the United States and the Netherlands, demanding a change in their policy of supporting Israel and to punish the United States for supplying weapons to Israel in the 1973 war. The boycott had little effect on supply. Oil shipments shifted. The United States received its oil from non-Arab producers. The boycott soon ended.

  Prior to the recession, the FOMC could not agree on appropriate policy or how to conduct it. Policy called for controlling growth of monetary aggregates by controlling growth of RPDs (reserves against private deposits) to reduce inflation; the members never discussed how much unemployment they would accept to reduce inflation. In practice, differences persisted.24

  At its August 20 meeting, the FOMC recognized that part of the recent surge in wholesale prices reflected the early end of the June 13 price freeze on July 18. Interest rates had increased during the summer as money growth slowed. In July, the Board removed interest rate ceilings on consumer time deposits of $1000 or more and four years maturity. It raised ceiling rates on shorter maturities. At the end of July, the System decided to rescue a Treasury offering by purchasing $350 million of a $500 million issue. The Treasury purchased almost half of the twenty-year bond issue for its trust accounts.

  To offset growth in RPDs that was more rapid than expected, the FOMC raised the maximum federal funds rate from 10.25 to 11 percent. As in its later experience from 1979 to 1982, the spread between the federal funds and discount rates induced a surge of borrowing to more than $2 billion in August.25

  The staff forecast called for much slower growth, 1.2 percent average real growth for the four quarters of 1974. “The dilemma that the Committee faces . . . is in deciding how to weigh the very real prospect that growth in the economy will be slowing to minimal levels against the very real risk that there may be enough remaining strength to keep unacceptable pressure on critical resources and hence on our structure of costs and prices” (FOMC Minutes, August 21, 1973, 21). In response to a question, Charles Partee, a senior staff member, said he expected a recession in late 1974 (ibid., 30). The staff report urged the FOMC to choose between higher employment and lower inflation. The member’s responses show their thinking. President Hayes (New York) wanted to slow inflation before it became embedded in wages. “He did not see much evidence that high interest rates were in themselves bringing demand pressures under control thus far.” And he described the increase in nominal rates as “severe restraint” (ibid., 14–15). However, he failed to mention that real rates of interest had changed very little and were modestly lower (Chart 7.10 above). Governor Brimmer thought a recession was likely to come sooner than most expected. “The issue before the Committee would be not whether but how soon monetary policy should be eased” (ibid., 32–33). Burns, Hayes, and Francis disagreed. As usual before the recession started, they gave priority to inflation control. Burns said that “reduction in the rate of inflation would require an environment of tighter budgets and a relatively restrictive monetary policy” (ibid., 39).

  24. The Board also increased marginal reserve requirements from 5 to 8 percent on certificates of deposit and commercial paper on May 16. The new requirements became effective June 7 but applied to the increase after May 16. At the same time, the Board removed the ceiling rate on all ninety-day single-maturity time deposits of $100,000 or more. On May 21, the Board sent a letter to large banks asking that “the rate of credit extension be appropriately disciplined” (Annual Report, 1973, 88–89). The federal funds rate rose, but annual monetary base growth also rose. The Board increased reserve requirements by 0.5 percentage points on June 29. It exempted the first $2 million of a bank’s deposits to pacify members of Congress who wanted to protect small banks and to reduce the loss of members. With bank credit continuing to grow rapidly, on September 7, the Board increased the marginal reserve requirement from 8 to 11 percent. Three months later, the Board restored the 8 percent marginal requirement.

  25. The System intervened in the foreign exchange market in July to stop dollar depreciation. The August rise in market rates reversed the decline. The System repaid the $273 million swap in August with an $8.5 million profit. Coombs was elated and urged more frequent, more visible, and larger interventions.

  Governor Bucher believed “the Fed had overreacted in the past and had created undesirable shocks in the economy” (ibid., 61). President Coldwell (Dallas) wanted to lower growth of RPDs and ignore M1, but Governor Holland thought the Committee should focus on interest rates. And Governor Daane wanted to remain restrictive, but he didn’t believe the System could achieve a precise target.

  President Balles brought expectations into the discussion. He quoted a newspaper article saying that “the Federal government doesn’t want to see the economy suffer either a recession or more inflation. If there is a choice, however, federal officials lean heavily toward more inflation
” (ibid., 67). He thought that some measures of money had to slow. Part of the problem was the narrow range kept on the federal funds rate. Governor Brimmer responded that “policymakers faced with a choice between more or less inflation and more or less unemployment were inclined to accept a little more inflation” (ibid., 73).

  Divided on several dimensions, the FOMC was unlikely to agree on decisive action. They made no effort to discuss the reasons for disagreement; they agreed to make no change. Chairman Burns summarized the comments as favoring M1 growth of 1 to 4 percent, RPD growth of 11 to 15 percent, and the funds rate between 10 and 11 percent. The Committee changed RPD growth to 11 to 13 percent. It voted ten to one, with one absence. Darryl Francis dissented because of the narrow range for the federal funds rate. He did not believe the desk would hit the money target. Burns had recognized that “failure to bring the monetary aggregates under control in recent months fundamentally resulted from a failure to control RPDs” (ibid., 53). Yet he did not set a wider range for the funds rate or support Francis’s effort to do so.

  Soon after the meeting, data showed that RPD growth remained above the target. The desk responded by raising the federal funds rate by 0.25 to 10.75 percent, and the Board on September 7 increased reserve requirements against large-denomination CDs.26

  Industrial production followed a very variable path during the fall, then fell 19 percent in December. Stock prices generally declined. The average annual increase in the CPI rose above 8 percent in November. The federal funds rate reached a peak in September and declined gradually, suggesting less restraint. As usual, policy moved procyclically. Monetary base growth declined from 9 percent in July to 7 percent in December, suggesting increased restraint.