A History of the Federal Reserve, Volume 1

by Allan H. Meltzer

  The first steps to ease policy came from Chicago and Atlanta. These banks reduced their discount rates to 1.5 percent on August 20. The Board approved, and Eccles urged Harrison to reduce the New York rate at the next directors’ meeting. Ever cautious, Harrison opposed the change as “too early.” He believed they should wait for the Treasury to complete its financing. Pressed by the Board, however, he agreed. New York lowered its rate to 1 percent effective August 27, with one dissent (Harrison Papers, file 2140.2, August 27, 1937, 3; Minutes, New York Directors, August 26, 1937, 11). By the end of the first week of September, all reserve banks outside New York had lowered their discount rates to 1.5 percent. These were almost the last changes in discount rates until after World War II.225

  222. “Since the last meeting of the Committee, the movement had leveled out with some reduction of prices both at home and abroad . . . [T]here seemed to be much less likelihood of a runaway movement than was the case a month or two ago” (Minutes, FOMC, May 4, 1937, 3).

  223. The Federal Advisory Council found “some recession in business activity in some districts” but “the recession was apparently temporary in character” (Board Minutes, May 18, 1937, 4–5).

  224. This was the first executive committee meeting held in the new Board of Governors building (August 18). The building opened formally on October 20, 1937.

  The FOMC voted on September 11 to undertake open market purchases of up to $300 million during the fall and to ask the Treasury to desterilize $200 million to $300 million of inactive gold. The Treasury agreed and acted promptly. The actions were taken more for seasonal than for cyclical reasons, to offset expected seasonal changes in the demand for reserves. After years of inactivity, this was a return to the 1920s policy of seasonal accommodation to prevent interest rates from firming during the harvest and Christmas seasons. Estimates presented at the meeting suggested that the banking system’s excess reserves would fall below $400 million before Christmas, and New York banks would use all of their excess reserves.226

  Only a few months earlier, the FOMC had been reluctant to let the Treasury undertake monetary action by desterilizing gold. Eccles, who appears to have felt most strongly about the issue, was not present at the September meeting. In his absence, Goldenweiser noted that “action by the Treasury also might be interpreted as violating the principle that the Federal Reserve System has primary responsibility for credit conditions and has adequate instruments for handling it” (Minutes, FOMC, September 11, 1937, 5). He urged the System to act on its own. The minutes do not record much discussion of the issue; they report that the committee recognized that “while the System could act alone. . . the most desirable action would be the suggested joint action” (12).

  The Federal Reserve had been inactive so long that it needed new criteria to guide operations. Goldenweiser (ibid., 6) recalled that in the past the rule of thumb was that borrowing by New York banks in excess of $50 million suggested tightness and less than $50 million suggested ease. That rule was no longer applicable. In its place he proposed to use excess reserves in place of borrowing; $250 million of excess reserves in New York and $700 million to $800 million for the country could be the threshold for judging ease and tightness.227 The committee did not discuss the proposal, but it reveals that the events of the 1930s had little effect on the Riefler-Burgess framework. Only the numerical magnitudes had changed.

  225. The qualification recognizes the reduction of the Boston bank’s discount rate to 1 percent in September 1939 and reduced rates on industrial loans for defense production.

  226. See table 6.6 above. At the time, excess reserves were above $3 billion. Contrary to the forecast, excess reserves rose, so the System made few purchases.

  227. The staff had estimated the level of excess reserves at which banks would begin to borrow from the reserve banks. A staff memo in February 1937 estimated that the banks wanted to hold $100 million of excess reserves, a clear recognition that not all excess reserves were redundant. At the time, excess reserves were $2.5 billion, but there is no attempt in the memo (or elsewhere that I have seen) to explain why actual excess reserves remained so far above the estimate of desired excess reserves.

  In his biography, Eccles reports that he expressed concern to the president about the economy as early as March. His concern at that time was that rising prices and wages would prevent the economy from reaching full use of capacity (Eccles 1951, 296–97). In August, he urged the president to encourage housing construction.228 None of these concerns appear in the September 11 minutes. Williams again recognized that the economy had slowed, but he was uncertain whether a recession had started (Minutes, FOMC, September 11, 1937, 6–8). Harrison noted that bank credit was available, so the “causes of the present situation were not in the monetary field” (9) The implication was that there was no need for policy action.

  Once again, the monetary base and the money stock tell a different story. Chart 6.4 shows the growth rates of the adjusted monetary base and the money stock from 1936 to 1939.229 Growth of the monetary base turned negative after gold sterilization in December 1936. The base fell throughout 1937, much of the time at a 7 to 11 percent rate. The money stock lagged behind; although its growth rate fell throughout 1937, the money stock began to fall only in August, two months after the start of the recession. Money growth remained negative until the early months of the recovery. With inflation (deflator) rising at a 6 percent average for the first three quarters, real money balances fell. In fourth quarter 1937 the situation changed; the price level fell and the base rose, so the real value of base money increased as real rates of interest rose. Again, as in 1920–21, the rise in the real value of the base and money dominated the effect on economic activity of rising real rates of interest.

  228. Eccles (1951, 361) defends himself against the charge at the time that the increases in reserve requirements and the undistributed profits tax had caused the depression. These beliefs appear to have been held most vigorously by advocates of a balanced budget who may have wished to avoid criticism of fiscal tightening in 1937.

  229. The adjustment corrects the base for changes in reserve requirements by reducing (or increasing) reserves by the dollar value of reserve requirement changes.

  Although the recession began in June, the FOMC made no purchases until November. Even a sharp stock market break, reducing the stock price index by 26 percent from late August to mid-October, did not induce a response. At last, on November 9, the FOMC executive committee voted to begin purchases at once and to purchase $50 million by the end of the month, using the authority of the September meeting. Eccles again wanted purchases because the Treasury threatened to act on its own by desterilizing gold. Harrison opposed purchases but voted in favor (Harrison Papers, file 2140.2, November 6, 1937). Perhaps because excess reserves rose and there was no evidence of seasonal tightening, the system bought only $38 million in November.230 On November 16, Harrison and Eccles agreed to stop purchases after checking with Morgenthau. The System made no further purchases until March 1938.

  Williams explained why the Federal Reserve purchased so little and stopped so soon. He described the period as a small depression but with “continued monetary ease,” and “for that reason, a policy of monetary ease could not be counted on as a major corrective” (Minutes, FOMC, November 29, 1997, 3). “Those in authority should not sit back and do nothing. . . . steps should be taken to devise a means of encouraging private investment” (4). But he made few suggestions about what should be done, opposed any increase in government spending, and stressed the importance of a balanced budget and other fiscal measures: reduction of the undistributed profits tax, the capital gains tax, and the surtax.

  230. Unlike earlier recessions, the Federal Reserve learned about the severity of the recession slowly. At the October 8 meeting of the Federal Advisory Council, participants talked about a tendency toward decline and suggested that only steel, textiles, and construction were below September 1936 levels. The council expected that
the fourth quarter “would be satisfactory” although below earlier anticipations (Board Minutes, October 8, 1937, 4). By mid-December the Advisory Council recognized that there had been a sharp business recession. Some plants had been closed. Others produced for inventory only. The tone remained relatively optimistic about recovery in the near future (Board Minutes, December 14, 1937, 6). This contrasts with statistics on industrial production. The Board’s index shows a decline from 106 in September to 83 in December (22 percent). The Miron-Romer (1989) index shows a modest decline in this period (1.2 percent) and a very large decline in January 1938 (23 percent), with further declines cumulating to 44 percent by July 1938. Kindleberger (1986, 271) reports that on several measures the recession destroyed half the recovery from the 1932 lows.

  Goldenweiser agreed that monetary policy had been easy since early 1932. The increases in reserve requirements had not reversed the easy money policy; the recession was due to (unspecified) nonmonetary causes. He agreed with Williams that the third increase in reserve requirements should have come earlier (ibid., 6). He saw no reason for “any major monetary action at this time” (9). The System remained inactive.

  The committee voted unanimously to continue the authorization to purchase for seasonal adjustment agreed on in September. Not a single dissenting voice suggested that the committee should purchase for expansion. Once again, the level of money market interest rates misled the FOMC. The members failed to see that falling prices meant that real rates of interest had increased as deflation and recession took hold.231 Using nominal interest rates instead of monetary growth as an indicator of the policy stance gave the wrong signal in 1937 just as it had in 1929–33. Even growth of bank loans would have told the Federal Reserve that policy was restrictive; in the year ending June 1938, total bank loans fell 7.5 percent, reflecting restrictive monetary policy and the recession.232

  Kindleberger (1986), Roose (1954), and Eccles (1951) describe the recession as principally an inventory recession. Eccles is representative of this view. He denied any monetary influence and attributed the 1937–38 cycle to four causes: (1) the buildup of business inventories at a time when (2) government spending declined; (3) the introduction of Social Security tax payments ($2 billion); and (4) labor disturbances that threatened to raise future production costs (Eccles 1951, 294–95). Instead of the $4 billion deficit in 1936, the Treasury had a cash surplus of $66 million in the first nine months of 1937.233

  Chart 6.5 shows two measures of the change in inventories during these years. The sharp peaks in fourth quarter 1936 and second or third quarter 1937 are clearly visible. The change in inventories, however, is small relative to the change in GNP or final sales.

  231. The meeting made a small adjustment in the allocation formula to increase earnings at banks that might have difficulty covering expenses and dividends. After 1933, the reserve banks did not pay franchise tax to the Treasury. Earnings above dividends and expense increased earned surplus.

  232. Investments fell also, so total bank earning assets declined.

  233. Morgenthau’s figures are slightly different. He has a cash deficit (excluding gold and silver purchases) of $288 million for the first nine months compared with a $2.8 billion deficit in the same period of 1936 (Blum 1959, 383). Morgenthau worked tirelessly to get the budget balanced and, to Eccles’s consternation, made a speech in New York on November 10, 1937, promising to balance the 1939 budget. Among the cabinet, James A. Farley and Henry Wallace endorsed his view, and Roosevelt also adhered to it until late in the recession.

  A more plausible interpretation is that the very large decline in final sales made inventories seem excessive. Chart 6.6 shows that rapid money growth in 1935 preceded the increased growth of final sales in 1936. The deceleration of money in 1937 preceded the sharp decline in final sales, and the resumption of money growth preceded the resumption of growth in final sales. On this interpretation, gold inflows caused an acceleration of the monetary base followed, as in chart 6.4, by an acceleration of money and, as in chart 6.6, by faster growth of final sales. Fiscal changes, especially bonus payments, reinforced these effects. At peak deceleration in the summer of 1937, the monetary base declined at an 11 percent annual rate. Final sales (and real GDP) reached their trough at the end of 1937.234

  Contemporary observers within and outside the administration gave considerable weight to the president’s “antibusiness” rhetoric and actions. Although the undistributed profits tax did not produce more than about $400 million in revenue in fiscal years 1936 to 1938 (much below projections), the revenue aspect seems to have been less important than the president’s message (March 3, 1936) citing the large growth of corporate profits in the 1920s as a source of disturbance. Regulation of foreign exchange and the capital markets by the Securities and Exchange Commission also gave rise to concerns.235

  234. Eccles (1951, 299) appears to change his argument without noticing: “Soon thereafter the inflated price bubble burst for want of purchasing power to sustain it, and the slump started in earnest” (emphasis added). This recognizes a monetary effect. Consistent with his belief that monetary policy was impotent in recessions, Eccles’s proposals for responding to the recession never mention Federal Reserve actions. He urged the president to lower mortgage down payments and interest rates on loans from the Federal Housing Administration (302).

  End of the Recession

  The Federal Reserve did little to correct the mistakes that contributed to the recession. Since short-term market interest rates remained low, it regarded its policy as easy. It continued to express more concern about future inflation than about current deflation. It took expansive actions when prodded by the administration and to avoid criticism for hindering the expansion program that the administration finally adopted.

  Although the recession started in May 1937, policy did not change until 1938, when Morgenthau and the Treasury pressed for an end to gold sterilization and reductions in the reserve requirement ratios. Eccles continued to urge increased government spending and a larger deficit. Repeal of the undistributed profits tax relieved some of the real or psychological effects its passage had generated.

  In November 1937, when Morgenthau first proposed to end gold sterilization, Eccles opposed on the grounds that “Roosevelt might grab the idea as a panacea for solving all economic problems. He considered excess reserves plentiful and contended that neither desterilization nor loosening of reserve requirements would actually ease credit” (Blum 1959, 393).

  235. Harrison reports, as an example, a conversation with George Whitney, a partner in J. P. Morgan. Whitney attributed the recession “largely to the Government’s attitude about taxes and business regulation and the rapidly growing fear of business that it will not be allowed to make a profit” (Harrison Papers, file 2610.1, November 12, 1937). Jacob Viner partly endorsed these views (Blum 1959, 384).

  Morgenthau persevered. On February 8 he met with Roosevelt, deeply concerned about the continuing recession and the effect of the recession on other democratic countries. Despite his concerns about deficits, he wanted to spend an additional $250 million in the remaining months of the fiscal year to increase WPA employment by 650,000. He believed falling prices encouraged delays in private spending, so he proposed to end gold sterilization. Roosevelt accepted both suggestions (Blum 1959, 400). Morgenthau’s diary explains that Morgenthau ended sterilization, in part, to prevent more government spending.236

  Eccles hesitated to approve unlimited gold purchases. The gold purchase program announced on February 14 limited the monetary base increase to $100 million a quarter, retroactive to January 1.237 Chart 6.4 (above) shows the immediate large change in the growth rate of the monetary base. In the second quarter, final sales rose modestly but inventories declined, so real GDP did not rise until the third quarter.

  Eccles regarded the gold program as Morgenthau’s plan. He was at best cool to the idea (Blum 1959, 406). Gold purchases expanded money without increasing Federal Reserve earnings, but h
is principal concern was the size of excess reserves. At the March 1 FOMC meeting, he favored continuing the program of selling long-term bonds and buying bills. The committee discussed open market sales to reduce excess reserves. This action would have neutralized Morgenthau’s program, in effect resterilizing gold. Eccles concluded: “No useful purpose would be achieved by reducing the total amount of securities held in the System open market account.” He did not propose purchases (Minutes, FOMC, March 1, 1938, 6–7).

  Again in April, the Roosevelt administration, not the Federal Reserve, acted to spur the recovery that by then was under way. Morgenthau leaves no doubt about the reason for action. On March 25, while Roosevelt was on vacation in Georgia, the stock market fell sharply. The proponents of spending within the administration seized the opportunity to convince the president that more spending would help the economy and the Democratic Party.238 After much internal wrangling, and Morgenthau’s threat to resign over the budgetary consequences, Roosevelt announced a new recovery plan on April 18.239 Spending for construction and welfare increased by about $2 billion. On the monetary side, the Federal Reserve reduced reserve requirement ratios, and the Treasury desterilized $1.4 billion, all the remaining gold sterilized since December 1936. The reduction in reserve requirement ratios released an estimated $750 million, reversing the May 1, 1937, increase for central reserve and reserve city banks (to 22.75 percent and 17.5 percent) and lowering to 12 percent and 5 percent the requirement for country banks and all time deposits. The Board’s minutes refer to the change as part of the president’s program (Board Minutes, April 15, 1938, 1–2).240 The very rapid expansion of the monetary base is apparent in chart 6.4 (above). Estimated excess reserves rose to $3.9 billion when desterilization was complete. Within a few months, the money stock began a sustained increase.