A History of the Federal Reserve, Volume 2

by Allan H. Meltzer

  Second, throughout the Great Inflation the Board’s staff interpreted errors in forecasting money growth as evidence of shifts in the demand for money, not as random shifts in supply or errors in their models of money supply growth. This permitted them to excuse their errors as unpredictable, random events even when they were repeated month after month as in 1973. The FOMC learned nothing. A related problem occurred in the period from 1979 to 1982, when the FOMC again tried to set a target for unborrowed reserve growth.

  Chairman Burns asked the staff to explain why the errors in June and July 1973 were shifts in the demand for money. Its explanation is not reassuring. It started by describing the demand for money as a function of a short-term interest rate and nominal GNP. In the first quarter, GNP and interest rates rose; the model predicted a 6 percent increase in M 1 demand. Demand rose much less, so the staff concluded that the demand for money shifted down. “We take the shortfall of actual M1 growth from that predicted from the model as evidence of a downward shift in the money demand function” (memo, Lyle Gramley to Burns, Burns papers, Box B_B80, July 31, 1973). This accepts the demand function as a true representation. The memo acknowledged that Federal Reserve policy could have raised the money growth rate “if the Fed had set out single-mindedly to keep M1 growing at (say) 6 percent. . . . But it would have had to drive down interest rates substantially below actual levels to accomplish this” (ibid.). The writer, Gramley, did not recognize that this is a very different argument from his claim that the demand for money shifted. It attributed the shortfall (and subsequent excess growth) to the constraint on short-term interest rates that altered the growth of RPDs and the supply of money. The same error—explaining all excess or shortfall in money growth as the result of shifts in demand—reappears repeatedly.

  A related but distinct error was Charles Partee’s argument that easing wage and price controls in January 1973 “might necessitate a somewhat faster rate of monetary growth to finance the desired growth in real output under conditions of greater cost-push inflation than would have prevailed with tighter controls” (FOMC Minutes, January 11, 1973, 70). Here the unwillingness to risk a recession or even slower real growth dominates any concern about higher inflation.76

  76. At the February 13 meeting, FOMC members learned that the agreement that month to devalue the dollar against gold from $38 to $42.22 an ounce would cause an 11 percent loss on its outstanding swap lines. The administration sent Congress a bill to devalue the dollar to $42.22 per ounce of gold. The bill also gave the president authority to raise tariffs. Burns supported the proposal as necessary (FOMC Minutes, February 13, 1973, 4). Also, the Voluntary Foreign Credit Restraint Program ended.

  Third, forecasts underestimated both growth and inflation. In February, the staff forecast that the GNP deflator would increase 4.25 percent in 1973 and that nominal GNP would rise 8.5 percent (FOMC Minutes, February 13, 1973, 144, 163). The actual price and GNP increases, fourth quarter to fourth quarter, were 8.3 and 11.9 percent. Partee did not offer a numerical estimate but forecast that “real GNP would grow at satisfactory but not ebullient rate for the remainder of the year” (ibid., 183–84).77 The Council of Economic Advisors believed that fears of inflation were exaggerated. “There is still good room for expansion of non-farm output as evidenced by figures on capacity utilization and unemployment rates of adult male workers. We expect a marked change in the food price situation after mid-year” (memo, Stein to the president, Shultz papers, Box 5, March 2, 1973).78 A few weeks later Stein wrote: “The money value of GNP rose extraordinarily in the first quarter—at an annual rate of about 13 percent. This is almost 30 percent larger than we forecast at the beginning of the year. Most of the increase is in prices, but the increase of real output also surpassed our forecast a little” (memo, Stein to the president, Shultz papers, Box 5, April 17, 1973). He recommended suspension of the investment tax credit and money growth held to 5 to 6 percent.79

  77. The staff expressed considerable uncertainty about the effectiveness of so-called phase 3 controls. Burns sent the president a copy of the staff analysis. He urged the president to speak publicly about the “enforcement of desired conduct on wages and prices” (Burns to the president, Nixon papers, Box 33, January 29, 1973). He also urged the president to “appeal to the public to reduce its buying” and suggested a “meatless day each week” to bring down meat prices (ibid., 3). His concern was that rising food prices would increase wage demands. Despite higher inflation, average hourly earnings rose less in 1973 than in 1972. For December of each year, the rate of change for that year was 8.1 and 6.3 percent. The reported 1973 inflation rate includes the start of higher oil prices that could not be predicted early in the year.

  78. These changes occurred at a time that the dollar was devalued by 10 percent (midFebruary). Burns considered the devaluation necessary, but he did not like it. He continued to oppose a floating exchange rate (Wells, 1994, 107). In March the dollar floated. That ended the Smithsonian agreement and the effort to fix the dollar exchange rate. Volcker wrote that “despite his enthusiastic support of fixed exchange rates, he [Burns] seemed . . . to have a kind of blind spot when it came to supporting them with concrete policies” (Volcker and Gyohten, 1992, 104). The Council used a 4 percent unemployment rate as the natural rate, so it overestimated excess capacity (Orphanides, 2002). Pierce (1998, 7) pointed out that the Board’s model used to estimate inflation and output did not allow for any effect of money growth until later in the decade.

  79. The GNP data showed 8 percent inflation and 6 percent real growth. Consumer expenditure rose at 16 percent annual rate in these early data.

  Taking account of the first quarter results, the Council and private forecasters revised their forecasts. Table 6.8 shows that the forecasts by both government and private forecasters were as much as nine percentage points lower than reported inflation. As noted, part of the difference reflects one-time increases in oil and food prices.

  Late in March, Partee summarized the staff position. It expected real growth to fall to 4 percent by year-end, and it anticipated that inflation would rise despite slower increases in food prices. He cautioned against a tough anti-inflation policy on the grounds that the FOMC would not persist in lowering inflation. “As unemployment rose, there would be strong social and political pressure for expansive actions, so that the policy would very likely have to be reversed before it succeeded in tempering either the rate of inflation or the underlying sources of inflation” (FOMC Minutes, March 19–20, 1973, 6). He favored 5.5 percent M 1 growth.80

  Robert Black (Richmond) asked a critical question: Should they raise the accepted level of the unemployment rate to 5 from 4 percent? Partee avoided the issue by responding that inflation would continue, according to the model, even at 5 percent unemployment rate (ibid., 28).81

  In testimony to Congress, Burns gave a different explanation. He explained that “fundamentally, it is the expansion of the money supply over the long run that will be the basic cause of inflation” (quoted in Wells, 1994, 111). Money growth (M1) reached 8.2 percent in first quarter 1973, but Burns did not urge a much more restrictive policy. He told the BIS members that “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 an extension of the Economic Stabilization Act” [wage and price controls] (ibid.).

  80. The FOMC voted at the March 19–20 meeting to authorize negotiations to increase swap lines with other central banks by up to $6 billion in the aggregate. The swap lines totaled $11.7 billion before the increase.

  81. The model used a Phillips curve relating unemployment and inflation. Partee acknowledged that the model had not worked well until they adjusted for “special factors” (FOMC Minutes, March 19–20, 1973, 30). The government did not change its goal for unemployment to 5.1 percent until 1976.

  At the FOMC meeting, Burns expressed conc
ern about a possible recession by the end of the year. He warned the members that “the Federal Reserve had a history of going to extremes” (FOMC Minutes, March 19–20, 1973, 108). Itwas now trying to do something it had not done ever before— slow inflation without bringing on recession. He accepted the higher inflation rate and urged the FOMC to accept a pause in restraint. Seven members agreed, but five disagreed. Robertson, Brimmer, and Hayes suggested changes in the ranges that would make Burns’s proposal acceptable. The compromise called for the federal funds rate to remain between 6.75 and 7.5 percent, M1 growth at 4 to 7 percent near term, and RPD growth of 12 to 16 percent. This proposal passed unanimously. During the period between meetings, deposits grew more slowly than expected, so the manager let the federal funds rate decline to 7 percent.

  Burns’s concern about recession was misplaced. Industrial production declined in March, but the decline did not continue. Inflation, not recession, became the problem. Although the federal funds rate was at the highest level ever reached, after adjusting for current or anticipated inflation, the real funds rate was about 3 percent.

  The unemployment rate remained in a narrow range around 4.9 percent during the first six months of 1973. Guided by the Phillips curve, the staff described the rise in inflation as “an aberration” resulting from the end of phase 2 controls and dollar devaluation (FOMC Minutes, April 17, 1973, 4–5). They did not anticipate continued inflation at the first quarter rate. The model predicted 4 to 4.5 percent for the next two to three years, based on a 4.5 to 5 percent unemployment rate. In response to a question, Partee agreed that the “tradeoff between unemployment and the rate of advance in prices had become less favorable recently” (ibid., 9). The model forecast could be reconciled with the data only by assuming that the Phillips curve had shifted. He blamed the shift mainly on phase 3.82

  Brimmer and Balles (San Francisco) said that the 1973 surge in the inflation rate was a lagged response to the monetary and fiscal stimulus in 1972. But M 1 and M2 growth slowed in the first quarter and were below the FOMC’s objective, so Daane opposed any increase in the federal funds rate.83 Hayes agreed with Brimmer and Balles that the System should tighten by lowering the money growth rate, raising the discount rate and the funds rate, and removing remaining regulation Q ceilings. To prevent higher inflation, “firmer wage and price controls were needed . . . [but] controls were not of much use in the absence of appropriate fundamental policies and in the present situation monetary policy was one of the fundamentals” (ibid., 72).

  82. In a statement reminiscent of the 1920s, Robert Black (Richmond) thought that prices were “beyond the control of the Federal Reserve.” Partee agreed (FOMC Minutes, April 17, 1973, 10). Statements of this kind remain ambiguous. Was inflation independent of monetary policy, or did it mean that the System would not attempt adequate restriction? Or did Black refer to one-time price level changes?

  Balles urged the FOMC to consider ways to improve control of money. He suggested controlling nonborrowed reserves and the monetary base (ibid., 75). The FOMC made no changes. After much discussion, it agreed to seek moderate growth in the monetary aggregates. In April, the funds rate remained unchanged, but M1 and M2 growth rose more than the FOMC’s desired path of 5 to 5.5 percent.

  As market interest rates rose, the prime rate and rates on consumer credit began to increase. Banks raised the prime rate to 6 percent (from 5.25) in December 1972. Early in February, four New York banks raised the rate to 6.25 percent. As head of the Committee on Interest and Dividends (CID), Burns spoke to the bankers and persuaded them to rescind their announcement. Congress had renewal of authority to control prices and wages under consideration, and neither Burns nor the bankers wanted Congress to include interest rates in the bill. By the end of February, however, the 6.25 percent prime rate had become general. CID issued guidelines that permitted banks to increase their lending rates as money markets rates rose, but they were supposed to adjust for the zero rate of interest paid on demand deposits.

  Pressed by Congress, especially Congressman Wright Patman, Burns and the CID urged the banks to limit their rate increases. But markets imposed higher marginal costs as open market rates rose. By September 1973, the federal funds rate averaged 10.8 percent, an increase of 5.5 percentage points since December 1972. The prime rate reached 10 percent, an increase of 4 percentage points in the same period.84 These data suggest a possible modest effect of CID efforts, since the lending rate rose less than the borrowing rate. We do not know how other terms and conditions of the loans changed, whether banks restricted prime rate loans, or whether borrowers shifted to other markets.

  83. The FOMC voted to release the 1967 Minutes after deleting some passages in the discussions of foreign currency.

  84. Interest rates on three-month certificates of deposits rose from 5.67 to 10.71 percent in about the same period. In April, Burns asked the president to let him resign from the CID. The conflict of interest that some had foreseen was now apparent to all. The Federal Reserve’s standing was hurt by the conflict. The president asked him to stay on, and he remained.

  Burns recognized that any success he had in reducing lending rates increased spending and borrowing. He felt powerless to let rates rise, reluctant to hold them down. In April, a Senate bill called for lower interest rates, a rollback of rate increases. The bill failed by only four votes in the Senate (Wells, 1994, 113). The close vote was more than enough to frighten Federal Reserve officials, who were rarely comfortable about relations with Congress. Burns responded by ruling that banks had to split their lending rates. On April 16, the CID established voluntary guidelines for bank interest rates. Farmers and small business with less than $1 million in assets received a lower rate than other borrowers. Two days later, banks raised the prime rate to 6.75 percent.

  In the next few months, as inflation rose, the System allowed market interest rates to rise to levels never experienced in the previous sixty years of Federal Reserve history. Although the Board was often hesitant to raise the discount rate, by July it had reached 7 percent, the highest level in Federal Reserve history to that time.85

  From May through August–September, the System worked to control money growth and inflation. President Nixon worked to control spending growth. The administration called this “the old time religion”—a program to control inflation by traditional means. Nominal budget outlays in 1973 rose 3.2 percent, a real reduction of at least 5 percent. The budget deficit fell from $20 to $15 to $8 billion between 1972 and 1974, and annual M1 growth declined from 8.8 percent in December 1972 to 5.3 percent a year later. It continued to fall. By the end of 1973, industrial production started to fall.

  Supplementing fiscal and monetary restraint was the president’s effort to end phase 3 by freezing prices, discussed above. The attempt failed. Phase 4 of the controls program replaced it. This was a reworking of phase 2. That also had little effect; the explanation at the time was that the economy was closer to full employment. In fact, the rate of increase in hourly earnings, which Burns considered central to inflation control, rose during the period of controls and phase 4.

  At the May FOMC meeting, Partee put the issue squarely. After forecasting that the rate of expansion would slow, he told the committee:

  85. One of many examples of hesitation and reluctance to raise the discount rate came in June 1973. On June 25 and 26, the Board disapproved requests for a 7 percent discount rate ostensibly because the request came too soon after the June 8 increase to 6.5 percent. Three days later, June 29, it approved the 7 percent rate. Consumer prices rose 8.2 percent at annual rate that month.

  Inflationary pressures are likely to remain substantial. . . . I do not have much hope that these underlying inflationary forces can be dampened appreciably without profoundly adverse consequences for the economy later on. (FOMC Minutes, May 15–16, 1973, 22)86

  The Council of Economic Advisers finally recognized that it had to give up the idea that full employment meant a constant 4 percent unemployment rate (H
argrove and Morley, 1984, 399). Part of “the old time religion” was a willingness to accept larger increases in the unemployment rate as a cost of reducing inflation. Some members of the FOMC did not accept that reasoning. At the May meeting, Eastburn (Philadelphia), Black (Richmond), Coldwell (Dallas), Winn (Cleveland), and Daane expressed concern about too much restraint.87 The GNP deflator rose at an 8.6 percent annual rate that quarter, about the same rate as the CPI.88

  Instead of protecting the value of money, the May meeting concerned itself with trivia. It considered a proposal to increase the cost to banks of issuing large CDs by increasing the marginal cost of these deposits. The proposal increased reserve requirement ratios for time deposits to 8 percent. This combined the standard 5 percent and a marginal increase of 3 percent for increases in large-denomination CDs and commercial paper above the average amount outstanding for the week ending May 16. The new requirement became effective on June 7. The Board approved the requirement on May 16. At the same meeting, it reduced the reserve requirement for euro-dollars from 20 to 8 percent, the same as large time deposits. In September, the Board increased the marginal reserve requirement ratio to 8 percent, making the effective requirement on new time deposits 11 percent.