EIGHT
Monetary history reveals the fact that folly has frequently been paramount; for it describes many fateful mistakes. On the other hand, it would be too much to say that mankind has learned nothing from these mistakes.
—Wicksell 1935, 4
The Federal Reserve began operations in 1914 as a peculiar hybrid, a partly public, partly private institution, intended to be independent of political influence with principal officers of the government on its supervisory board, endowed with central banking functions, but not a central bank. Each of the twelve semiautonomous reserve banks set its own discount rates, subject to the approval of the Federal Reserve Board in Washington, made its own policy decisions, and set its own standards for what was eligible for discounting. Even branches of reserve banks initially had some independent powers.
The new system had two principal monetary powers. It could buy and sell gold, thereby changing interest rates and money, and it could set the rate at which member banks discounted eligible paper. Other activities and responsibilities included centralizing the country’s gold reserve, developing a domestic market for bills of exchange, acting as lender of last resort in a crisis, and eliminating large seasonal increases in interest rates during the autumn, when the agricultural harvest moved through the commodity markets.
The Federal Reserve had little discretion. The founders intended the gold standard to work automatically. Discounting was at the discretion of the member banks. The Federal Reserve could decide the timing of discount rate changes, but the rules of the gold standard limited the range within which it could set the discount rate. It could set the rate at which it bought acceptances, but despite its efforts, the acceptance market did not become large and active.
The original structure, organization, and methods of operation did not survive. Establishment of the Federal Reserve helped to create a national financial market that undermined the system of separate discount rates. Banks used the correspondent banking system to borrow in markets with lower rates. By the early 1920s, the System had moved toward more uniform discount rates; although differences between districts continued, they were smaller, and rate schedules became more uniform.
Wars, the growing role of the federal government, and other external forces contributed to the major changes in structure and organization in the years to 1951. Flaws in the original plan and different conceptions about the roles and responsibilities of the Board and the reserve banks combined with these external events to force changes. Different beliefs about the roles of the reserve banks and the Board, and rivalry over power and influence, worked to delay change and disperse power.
By 1951 the Federal Reserve System had become a central bank with its headquarters in Washington. The accord with the Treasury in March of that year released the Federal Reserve from Treasury control and began the evolution toward the modern Federal Reserve. Although struggles over power and influence continued, the Board of Governors had final control over decisions. The semiautonomous regional banks were now part of a unified system. The Federal Open Market Committee (FOMC) made binding portfolio decisions for the reserve banks. Open market purchases and sales of government securities replaced discounting as the principal means of implementing policy. The discount rate had a minor role.
The System’s founders would not have liked or even recognized the Federal Reserve that existed in 1951. Gold no longer had an important role. Activist policies, based on collective judgment, determined money, interest rates, and prices. A small, mostly passive institution had become the most important central bank in the world.
The Federal Reserve’s founders wanted to base currency or note issue on discounts of commercial paper to free currency from dependence on government securities. They believed that the new arrangement would permit currency and money to expand and contract with the needs of trade and the public’s demand. In the Great Depression a change, believed to be temporary, permitted the Federal Reserve once again to issue currency backed by government securities. Eventually the change became permanent. Government securities became the Federal Reserve’s principal asset.Discounts of commercial paper and bills of exchange had only a modest and inconsequential role.
Volume 1 tells how and why these changes occurred in response both to external events and to flaws in the original plan. It is also a record of the achievements and failures of the first thirty-seven years.
ACHIEVEMENTS AND FAILURES
Looking back from 1951, few would conclude that the Federal Reserve had achieved the hopes of its founders and early proponents. The Great Depression, though at the time not considered a failure of monetary policy, was the deepest and longest in United States history. The Federal Reserve had not prevented thousands of bank failures, the collapse of the financial system, and the devaluation of the dollar. The dominant view in 1951 regarded monetary policy as unimportant for economic stabilization, but it recognized that the Federal Reserve had failed to maintain financial stability. Deposit insurance, stock market regulation, and separation of commercial and investment banking, among other New Deal measures, showed that the public, through its representatives, no longer trusted the Federal Reserve alone to maintain a stable and solvent financial system.
Central bankers and most economists in the 1920s regarded the gold standard as essential for monetary stability. The Federal Reserve achieved one of its major goals when Britain and France followed Germany back to a fixed exchange rate with their currencies convertible into gold. Other countries pegged to gold also; by 1928 all major trading countries, and many others, had adopted a gold exchange standard. The Federal Reserve and other central bankers considered restoration of a type of gold standard one of the major achievements of the 1920s.
Although central bankers and governments wanted the gold standard restored, several were reluctant to accept its implications. Three problems arose. First, after restoration, exchange rates in major trading countries were incompatible with domestic price stability. The British overvalued the pound; the French undervalued the franc. Britain would not deflate after 1925; France would not inflate after 1927, so price changes could not adjust real exchange rates to remove these differences, as the gold standard required. Second, countries would not permit dynamic adjustment to work. Gold flowed to the United States through much of the decade and to France at the end of the decade. Both countries sterilized most gold flows to prevent prices from rising. With receiving countries sterilizing, the countries paying gold had to deflate or leave the standard. The gold standard had an unwelcome deflationary bias. As countries returned to the gold standard, the increased demand for monetary gold stocks added to deflationary pressure.
Third, under the gold exchange standard the United States and Britain held their reserves in gold. Other countries held dollar or pound sterling balances. France, especially, regarded its status as second-rate. As in the 1960s, the rules permitted the Bank of France to convert its reserves into gold. France’s gold purchases, and sales of dollars and pounds, added to the deflationary pressures imposed by the return to gold. Unwilling to follow the rules or give up the standard, countries resisted steps to restore equilibrium real exchange rates. In retrospect, the breakdown of the gold standard seems inevitable; at the time, it seemed calamitous.
As the world economy moved toward deflation and depression. The Federal Reserve’s principal concern was inflation. To contemporary economists, this concern is puzzling because the price level fell slowly from 1927 to 1929, then more rapidly. Federal Reserve officials did not base their concern about inflation on price changes or sluggish money growth. To most of them, rising stock prices and growing use of borrowing to purchase shares was all the evidence of inflation they needed. Their interpretation relied on the real bills doctrine—the belief that credit extended for common stocks, real estate, government securities, or commodity speculation created inflation because the additional credit did not give rise to additional output.
Deflationary policies contributed to the start of the 1929 recession. When the Federal Reserve raised the New York discount rate in August 1929, part of the world was in recession. Although it was not known at the time, the United States economy was at a peak. The Great Depression had started.
There is no single cause of the Great Depression or a unique monetary shock. A series of financial shocks followed—bank failures, Britain’s departure from the gold standard followed by other departures, and financial failures in the United States. Most Federal Reserve officials favored a passive policy. They viewed the depression as the inevitable consequence of excessive speculation in stocks financed by credit creation. On their view, the proper response was to purge the economic system of its excesses—excesses made more serious by credit expansion unrelated to real bills. Monetary or credit expansion to end the depression would require purchases of government securities. On the real bills interpretation, such purchases prevented the inevitable adjustment and purge of previous excesses.
If the Federal Reserve had maintained monetary growth, the country and the world would have avoided years of depression. Failure to act during the Great Depression was the Federal Reserve’s largest error, but far from its only one. Failure to expand can be explained as the result of prevailing beliefs about the inevitability of a downturn following the stock market boom. Nothing in theory or central banking practice can explain why the Federal Reserve did not respond to the failure of thousands of banks. Most of the banking failures from 1929 to 1932, and the final collapse in the winter of 1933, could have been avoided. The failing banks included many member banks. After years of recession, banks had little eligible paper to borrow against. The Federal Reserve, following the real bills doctrine, saw no reason to expand. This was a destructive and mistaken interpretation of banking theory. In Lombard Street, his classic work on banking, Walter Bagehot quotes the spokesman for the Bank of England in the 1825 panic: “We lent it ... by every possible means and in modes we had never adopted before ... in short, by every possible means consistent with the safety of this Bank, and we were not on some occasions over-nice” (Bagehot 1962, 25).
Bagehot’s work was known at the time. Senior officials referred to him, but they did not follow his advice. They tried to protect the gold reserve and, at crucial times, did not function as a system. Individual reserve banks refused to participate in open market purchases to protect their banks’ gold holdings. A design failure and a failure of leadership permitted individual banks to opt out of System purchases. There was too much autonomy built into the 1913 Federal Reserve Act, and the Board failed to use its powers to force the reserve banks to expand together.
Ideas were important too. The original Federal Reserve Act wrote the real bills doctrine into law. At the Federal Reserve Board, and at several reserve banks, officials followed this doctrine. They considered real bills— commercial credit—to be the only correct foundation for credit expansion. If banks did not borrow, they believed it was wrong to expand credit. This policy gives rise to procyclical policy action: credit and money expand when output expands and contract when output contracts. The gold standard, too, makes policy action procyclical.
The Federal Reserve’s attachment to the real bills doctrine was not peculiar. Economists, bankers, congressional leaders, and many others accepted the theory and believed the Federal Reserve was right to follow it. There were few critics at the time.
Early in the nineteenth century, one of the founders of economics, Henry Thornton (1962) recognized the principal flaw in the real bills doctrine: controlling the quality of credit did not ensure control of the quantity. At the Federal Reserve, Benjamin Strong rediscovered this proposition in the 1920s. Neither the discovery nor the rediscovery convinced real bills proponents.
Strong’s conclusion reflected experience in the postwar recession of 1920–21. After the Federal Reserve convinced the Treasury to end wartime restrictions on interest rates, the nominal discount rate rose to 7 percent in New York. Use of marginal discount rates at regional banks raised interest rates far above that level. Discounting continued to increase, in part because banks could borrow at preferential rates using Treasury securities as collateral. But that was not the lesson drawn by Strong and others.
The 1920–21 experience affected subsequent developments in two ways. First, the Federal Reserve became convinced that the traditional British central banking procedures would not work in the larger, more diverse circumstances of the United States. Second, and closely related, complaints from agricultural and commercial interests, particularly in the South and West, aroused congressional concerns. Topmost among the political concerns was the fear that the Federal Reserve would operate for the benefit of Wall Street and large banks and against the interests of farmers, ranchers, and the general public. Federal Reserve policy in the 1920–21 recession seemed to confirm these fears.
Failure to distinguish between real and nominal interest rates was another, no less important error. As prices fell, real interest rates rose. Federal Reserve officials, and outsiders, failed to distinguish between the two rates, a distinction recognized early in the nineteenth century by Henry Thornton and later developed more fully by Irving Fisher. Although there are occasional references to the possibility that a low nominal interest rate did not necessarily connote an easy policy, none of those making these comments offered a clear analysis of the effect of falling prices on real interest rates and exchange rates.
Failure to distinguish clearly between real and nominal interest rates is puzzling. Fisher was professionally active in the 1920s and 1930s. He warned about the high cost of deflation and urged officials to pay attention to measures of deposits and money. In the 1920s Fisher worked to get Congress to mandate price stability as the Federal Reserve’s goal. The Federal Reserve opposed the legislation, and it did not pass.
The Federal Reserve also ignored Walter Bagehot’s analysis of the role of a lender of last resort. At times Board members and governors referred to Bagehot’s Lombard Street, but they did not follow his doctrine: In a financial crisis, lend freely at a penalty interest rate; do not try to protect the gold reserve.
Theories, or beliefs, go a long way toward explaining why the Federal Reserve did not avoid crises in 1920–21, 1929–33, and 1937–38. The beliefs that officials used to interpret events, and the interpretations they reached, were conventional at the time. The Federal Reserve Act used the gold standard and the real bills doctrine as guiding principles. Faith in the gold standard and belief in its stabilizing power constituted a cornerstone of the orthodoxy of the time, an orthodoxy that was widely shared by leading members of the business, banking, and academic communities. It would have required a strong, forceful leader to recognize the need to abandon orthodox beliefs. A divided Federal Reserve could not supply that leadership. It is highly uncertain that even a strong leader could have overcome the firmly held beliefs that led to the mistakes of 1929–33 and 1937–38.
Between 1930 and 1933, the Federal Reserve did little to prevent the collapse of the United States financial system and thousands of bank failures. President Herbert Hoover and Secretary Andrew Mellon proposed a National Monetary Commission and, soon after, the Reconstruction Finance Corporation (RFC) to prevent failures from spreading. Initially the Federal Reserve was wary of these efforts, concerned that it would have to lend to insolvent banks or to institutions like the RFC that lent to insolvent banks. By June 1932, the Federal Reserve wanted the RFC to be more active. In part this change of view reflects two opposing influences. One was the System’s desire to limit or end bank failures and the large increases in the demand for currency by concerned depositors. The other was the firm belief that it could, or should, do nothing to prevent bank failures.
Financial collapse in the winter of 1933 was not inevitable. President Hoover appealed to the Federal Reserve to offer guidance. Hoover also appealed to President-Elect Roosevelt to support a bank holiday. Hoover believed he lacked authority to act, and Roosevelt was unwilling to accept responsibility when he lacked authority.
Political maneuvering and hesitancy do not explain the Federal Reserve’s failure to act. Chapter 5 offers three plausible explanations. First, some members of the Open Market Policy Conference believed that the very large open market operations in 1932 accomplished little. Additional operations would do no more. Failures, they believed, were the inevitable consequence of bad decisions and speculative excesses that had to be purged before stability could return. Second, some reserve banks, notably Boston and Chicago, refused to participate in additional purchases during the summer of 1932. They would likely have refused again, if asked, in the winter of 1933. Third, some reserve banks may have feared that open market purchases would be offset, in part, by a loss of gold. Protecting the gold reserve by refusing to lend was one of the main errors of central banking practice that Bagehot warned against.
Reliance on discounting gave monetary policy a procyclical bias. In the severe recessions of 1920–21 and 1937–38, the Federal Reserve imposed deflation. The 1920–21 recession resulted from a decision to restore the prewar dollar–British pound exchange rate by deflating prices in both countries. Britain had experienced more inflation, so it had to deflate most to restore the prewar exchange rate. Deflation by Britain alone, however, would not have removed the effects of wartime finance on the United States stocks of money and credit, contrary to the real bills doctrine. The decision delayed Britain’s return to the gold standard and raised the social cost. The two governments did not repeat this mistake after World War II.
The decision to deflate together also raised the social cost in the United States. The 1920–21 recession is the only recession in Federal Reserve history that has short-term nominal interest rates higher at the trough of the recession than at the previous peak. Severe deflation made real interest rates higher still.
The Federal Reserve took no action to end the recession. Rising real interest rates, however, attracted gold, raising the stock of base money. The counterpart of rising real interest rates was a rising stock of real balances. As prices fell and gold flowed in, real money balances rose rapidly. When the public’s real balances exceeded the amount it wished to hold, spending increased and the recession ended.
The pattern of rising real money balances and rising real interest rates contributed to ending recession in 1937–38 and 1948–49. This dynamic did not work to restore prosperity in 1929–33 because the Federal Reserve allowed the nominal stock of money to decline so much that real money balances fell despite the expansive effect of deflation on the stock of real balances. Bank failures, and fears of additional failures, contributed to the decline in real balances. Efforts to shift from deposits to currency drained reserves from the banking system. The Federal Reserve’s failure to offset the loss of reserves added to bank insolvency and brought about the result the public feared.
Theory or beliefs also contributed to the Federal Reserve’s reluctance to end pegged interest rates after World War II. Many economists and businessmen claimed that a large outstanding government debt limited the size of permissible interest rate changes. Marriner Eccles, Federal Reserve chairman at the time, repeated frequently that, to be effective, interest rate increases had to be large. Large increases, however, imposed large losses on debt owners (with gains to the Treasury). Unwilling to impose large losses, Eccles sought other ways to reduce spending growth.
His proposals show the absence of careful analysis at the time. Eccles often favored higher reserve requirement ratios, secondary reserve requirements to force banks to hold more (low yield) Treasury bills, and controls requiring higher down payments and shorter duration of consumer loans.
With interest rates fixed (or pegged), increases in reserve requirement ratios transferred incomes from banks to the government. Banks sold securities to meet the additional requirement. To keep interest rates unchanged, the Federal Reserve supplied the additional reserves by buying the securities that banks sold.
Congress never agreed to secondary reserve requirements. Such requirements would force banks to hold more government securities, reducing their profits. With unchanged growth of base money and government debt, the total supply of credit would remain unchanged. Portfolio composition of the principal institutions would differ. Banks would own more Treasury bills; other lenders would acquire loans that the banks would forgo.
The Federal Reserve was not alone in these errors. Many in the academic profession, and other economists, made similar statements.
The Federal Reserve had some notable successes during its first four decades. Evidence of success and acceptance was the agreement in 1927 to replace the Federal Reserve’s twenty-year charter with a permanent one. The new charter evoked little of the passion and attention so much in evidence in 1913. The relatively stable price level and stable interest rates from 1922 to 1929 lay behind acceptance of the Federal Reserve and its increased congressional support. Strict adherents to the real bills doctrine criticized the use of open market operations to supplement discounting of real bills. They saw open market operations as a departure from the letter and spirit of the law. These criticisms found little congressional support as long as the System avoided major recessions or a return of financial crises accompanied by failures and surging interest rates.
Before 1914, United States interest rates rose sharply during the scramble for liquidity that became a standard feature of a financial panic. The Federal Reserve avoided financial panics between 1914 and 1928. Interest rates rose much less in the 1920–21, 1923–24, and 1926–27 recessions than in the 1890s or in 1907–8. Also, before 1914 interest rates had a large seasonal element. The Federal Reserve removed the seasonal swing using discount policy and acceptance and open market purchases. This fulfilled one of the founders’ main reasons for creating the institution.
The Federal Reserve helped to finance both world wars; it provided credit and money by lending to commercial banks at fixed interest rates or by open market purchases. In addition, the System acted as the principal bond salesman for the Treasury, using its network of regional and branch banks, and its relations with the leading commercial banks, to place the bonds. The Federal Reserve’s decision to allow banks to profit from bond sales to the nonbank public gave banks a powerful incentive to cooperate in the financing.
During the 1920s, the System undertook pathbreaking research and the development of new statistical series to support its work. The absence of an operative gold standard immediately after World War I, and widespread criticism of discount policy and discount rates in the 1920–21 recession, encouraged consideration of operating procedures and market signals about the need for policy action. Concern for market signals, in turn, required the development of new data series and fostered the use of new analytical techniques. By the mid-1920s, System economists had constructed measures of production, inventories, department store sales, and other variables. These are the forerunners of the data series that markets and policymakers rely on to this day. Developing these series and combining them required skillful use of index number theory.
In its 1923 annual report, the System discussed a general framework that sought to reconcile the passive stance implied by the real bills doctrine with more active use of open market operations. The new framework tried to achieve the Bank of England’s control of discounting without relying very much on the discount rate. Also, it tried to satisfy both advocates of the real bills doctrine and their opponents. Subsequently, economists at the Board and the New York bank developed a more explicit framework to guide policy decisions. This framework, though based on observations by many people, was mainly the work of Winfield Riefler and W. Randolph Burgess. Their work implied that the Federal Reserve could control the volume of member bank borrowing with fewer and smaller changes in interest rates. Open market purchases supplied reserves and encouraged banks to repay borrowing, offer more loans, and reduce interest rates; open market sales drove banks to borrow, restrict lending, and raise interest rates. The emphasis satisfied real bills advocates. Quantitative control through the use of open market operations satisfied Strong and others who no longer believed that the quality of credit restricted the quantity.
The new framework brought together open market operations, discounting, discount policy, and credit expansion as part of a theory of central banking. The theory required the strong proposition that banks did not borrow to profit from higher market rates. This proposition removed the need for an unpopular penalty rate, set above the rate on prime commercial paper. Experience in 1928–29, when the Federal Reserve tried to control the volume of discounts without increasing the discount rate, rejected the proposition but failed to change it. Federal Reserve officials continued to claim that banks did not borrow for profit. They found it necessary, however, to inform bankers that borrowing was a privilege and not a right of membership and to impose administrative restrictions to limit the amount and duration of borrowing. This was a long step away from the original idea that the Federal Reserve’s main function was to discount for member banks.
The Riefler-Burgess framework combined banks’ reluctance to borrow with another proposition that did not distinguish between individual banks and the banking system: “Banks disliked being continuously in debt and hence tended to contract credit when the level of indebtedness was increased and to expand credit when the level of indebtedness was reduced. Because of the tradition against continuous borrowing, when the Federal Reserve System sold securities, the resulting increase in indebtedness tended to cause banks to control credit” (Subcommittee on General Credit Control and Debt Management 1951, 283).
This reasoning does not explain why open market sales would contract total bank credit. Why didn’t other banks borrow when an individual bank repaid its indebtedness? The proper answer would have required the Federal Reserve to develop a framework linking its operations to market interest rates and the supply of bank reserves or monetary base.
The tenth annual report and the Riefler-Burgess framework covered over, but did not resolve, differences between opponents and proponents of the real bills doctrine. The conflict emerged first in the 1924 and 1927 recessions when, under the leadership of Benjamin Strong of the New York bank, the System expanded credit and the monetary base both to help the British and to encourage recovery from domestic recessions. The conflict became more open in 1929, when the Board wanted to control borrowing by discouraging speculative credit and New York and some other reserve banks wanted to raise the discount rate.
The Riefler-Burgess framework retained a central role in the Federal Reserve’s analysis of monetary developments until the 1950s. The staff adjusted the framework to reflect new developments, notably the increase in excess reserves during the 1930s.
The 1923 annual report, books by Riefler and Burgess, speeches by Strong and Adolph C. Miller (a prominent Board member from 1914 to 1936), and other statements and publications moved toward greater openness about procedures and analysis. Nineteenth-century central banks were secretive about what they did and why they did it. Gold standard rules were known, of course, but central banks often did not follow the rules automatically. One of Bagehot’s (1962) main criticisms of the Bank of England in the nineteenth century is that it failed to preannounce its policy response to financial panics. The movement toward transparency was slow, but by the end of the twentieth century, all leading central banks had moved decisively toward greater openness.
Other major accomplishments included extension of the par collection system, development of the payments system, and a national money market. Interest rates and discount rates became more uniform within the country as banks’ size increased and new money market instruments were developed. The founders failed in their attempts to create a broad national acceptance market to replace reliance on stock market call loans as a money market instrument. By the 1930s, Treasury bills and certificates served this function. Wartime increases in government debt made the government securities market the market of choice for short-term reserve adjustment. By the 1950s, the government securities market and the market for federal funds (bank reserves) achieved one of the founders’ goals in a way they did not envisage. These markets replaced the call money market as the market in which banks adjusted reserve positions. Monetary operations and bank adjustment were freed from dependence on stock market activity.
The lasting achievements of the early years include the development of a high-quality professional staff. Although research on central banking lagged in the 1930s and 1940s, Federal Reserve staff pioneered in research on topics such as the measurement of government deficits and the effects of budget deficits on the economy. In areas such as supervision, regulation, and banking law, Federal Reserve staff made important contributions. Two notable examples are legislation closing the banking system for the 1933 bank holiday and the Banking Act of 1935.
In the early years, international monetary policy was a central bank responsibility. Central bankers dealt with their counterparts abroad. In the 1920s the New York reserve bank and its governor, Benjamin Strong, negotiated and granted loans to foreign central banks to help restore the gold standard and to coordinate actions. Governments borrowed in the marketplace, assisted by investment bankers.
Although central banks attempted policy coordination, the Federal Reserve was explicit that it would not change its course for the benefit of another country if the change required inflation or deflation at home. This restricted the role of coordination. Some economists assign a large role to insufficient policy coordination. They claim that governments could have maintained the gold standard and prevented worldwide deflation and depression by acting together in the 1920s and 1930s.
This claim neglects exchange rate misalignment, particularly the misalignment of real exchange rates. Lending and borrowing or simultaneous intervention in exchange markets had a limited role at best. In the 1920s, countries on the gold standard had to accept inflation or deflation to adjust real exchange rates. Surplus countries would not inflate; deficit countries were reluctant to deflate after the mid-1920s. The remaining solution was to devalue or revalue against gold and other currencies. Britain left the gold standard in 1931. Other countries followed.
In the 1930s, the Treasury replaced the Federal Reserve as the principal negotiator of international financial agreements. Secretary Henry Morgenthau signed the Tripartite Agreement with Britain and France. The agreement sought to stabilize exchange rates between the three countries, but again real exchange rates were misaligned, and countries followed independent policies. French policy, especially, was inconsistent with the agreement, necessitating devaluations of the franc that violated the spirit of the agreement.
Again in the 1940s, the Treasury negotiated an international monetary agreement. The Bretton Woods Agreement attempted to formalize international policy coordination. Member countries agreed to fix exchange rates but retained the right to devalue (with international approval) to correct structural imbalances. The agreement tried to reconcile domestic and international stability and provide a means by which surplus countries could lend to deficit countries.
The agreement divided the Federal Reserve. The Board sided with the Treasury, favoring the agreement. The leaders of the New York bank opposed. They preferred a return to the gold standard, not adjustable exchange rates. Neither side had much influence on the agreement. The Treasury took control and retained it.
INDEPENDENCE AND CONTROL
The Federal Reserve’s independence was so well established in the first twenty years of its existence that President Hoover was reluctant to even ask its advice during the financial crisis at the end of his administration. Within a few years, this independence was lost. From 1934 to 1951, the Treasury Department severely restricted Federal Reserve actions. When William McChesney Martin Jr. became chairman of the Board of Governors in 1951, one of his tasks was to reestablish the independence of the Federal Reserve System from the executive branch, particularly the Treasury.
Independence
One of the anomalies of the 1930s and 1940s is that the Treasury had more influence over the Federal Reserve after the secretary left the Board. Secretary Morgenthau permitted Congress to eliminate his statutory position as chairman of the Federal Reserve Board, but he acquired another means of influencing the Federal Reserve. He held most of the profit from devaluing the dollar against gold in the Exchange Stabilization Fund. He used the fund, and other Treasury trust funds, to buy and sell gold or foreign exchange, and he could threaten the Federal Reserve with his power to supply reserves and lower interest rates. Occasionally he did just that.
Morgenthau’s threats and influence were not the only reason the System failed to resist Treasury control. Eccles believed that monetary policy was powerless, since interest rates were at historically low levels. His greater interest was fiscal policy. He wanted to advise the president and participate in budget and legislative decisions. His principal interest in Federal Reserve independence in the 1930s surfaced when Morgenthau threatened to act in place of the Federal Reserve.
Wartime Treasury influence or control had a different origin. The Federal Reserve agreed in 1942 to finance the war at low nominal interest rates, as central banks traditionally have done. Regaining independent authority to set interest rates after World War II proved difficult, just as it had after World War I. Regaining independence of decisions and actions required political support from the administration, the Congress, or the public. Political support began to form in 1949 under the leadership of Senator Paul Douglas. Support strengthened after the Korean War started in 1950. Heavy-handed action by Treasury Secretary John W. Snyder and support for an anti-inflation policy in Congress helped the Federal Reserve get an agreement that allowed interest rates to rise provided they rose slowly during the transition to greater independence.
Independence was never thought to be absolute. Independence prevented an administration from deciding unilaterally to use monetary expansion to gain temporary political advantage or to finance too much of the budget at the central bank. Allan Sproul, president of the New York reserve bank from 1941 to 1956, recognized the nuances hiding in the term “independence”:
I don’t suppose that anyone would still argue that the central banking system should be independent of the Government of the country.1 The control, which such a system exercises, over the volume and value of money is a right of Government, and is exercised on behalf of Government, with powers delegated by the Government. But there is a distinction between independence from Government and independence from political influence in a narrower sense. The powers of the central banking system should not be the pawn of any group or faction or party, or even any particular administration, subject to political pressures and its own passing fiscal necessities. It is clear that in war or in any other great emergency, the policy of the central banking system must support the national plan of action. It seems to me equally clear that in less emergent circumstances it is wise for government to set-up barriers or buffers of protection of the central banking system from narrow political influence. (Letter to Robert R. Bowie, Sproul Papers, Memorandums and Drafts, September 1, 1948, 2)
This statement of general principles seems well crafted. However, it does not say what happens if the government and the Federal Reserve disagree about the importance of the emergency. Secretary Snyder argued that “the President has the right, and the duty, to discuss disputes without attempting to dictate to the Board of Governors but by full and complete consultation with the Board” (Subcommittee on General Credit Control and Debt Management, Answers to Questions 1951, 31).
The secretary also favored creating a “discussion group” consisting of the secretary of the treasury, the chairman of the Board of Governors, the director of the budget, the chairman of the Council of Economic Advisers, and the chairman of the Securities and Exchange Commission (ibid., 31).2 The Federal Reserve’s statement did not mention a coordinating body. It favored a more independent role. When conflicts arise “each agency involved shares the responsibility for finding ways to resolve the conflict” (264).
The meaning assigned to “independence” did not progress much subsequently. Resolution of its conflict with the Treasury did not settle what a central bank should do if the government ran large or regular deficits in peacetime. FOMC members recognized that Congress approved the spending plan and deficit finance. A central bank could not, and they believed should not try to, reverse congressional decisions. But that appeal to democratic rule did not answer the question, How much should the central bank raise interest rates, or permit them to increase? It took years of sustained inflation to force attention to that question. In the 1950s the Federal Reserve hoped it could avoid the issue by joining a coordinating body of the four leading economic agencies known as the quadriad during the Kennedy administration. The quadriad continued through the early 1970s until replaced by less formal arrangements.
In both world wars, the Federal Reserve surrendered its independence to assist in war finance. Each time it found that regaining independence was difficult and long delayed. It did not learn from its experience after World War I to negotiate an end to pegged interest rates before it made a commitment in 1942. It did not foresee that raising interest rates would be unpopular after the war. It worked hard to gain public support for independence among journalists, academics, bankers, and the public, and within the government by undertaking unpopular duties that Congress and the executive branch did not want to do. Only after the Korean War started and concern about inflation rose did the Federal Reserve muster the popular and congressional support necessary to sustain an independent policy.
Control
President Wilson’s compromise, establishing semiautonomous reserve banks and a supervisory Federal Reserve Board, did not resolve the issue of control. Conflicts arose not only because the act dispersed control but because, from the start, officials had different ideas about how the new System should function. New York bankers especially wanted a central bank, under their leadership. The Board often tried to stretch the term “supervise” until it meant “decide.”
Benjamin Strong avoided the Board’s control by responding to the interests of other reserve banks. Several of the governors thought of their activity as banking, and they wanted their banks to profit. The act granted a dividend on the shares held by member banks, so earnings had to be sufficient to pay the dividend. In the early years some reserve banks—particularly the smaller banks in predominantly agricultural regions—did not have enough discounted paper to pay expenses and the dividend. Strong offered to pool the income on acceptances and then on government securities. By adjusting the allocation formula, he helped the smaller banks solve their problem. In return, they supported his decisions.
In 1919 the Board was able to get the acting attorney general to interpret its power to include changing discount rates even if a reserve bank opposed the change. The Board used the power again in 1927 when it ordered Chicago to reduce its rate.
By the mid-1920s, discounting had a much-reduced role compared with the original plan. Open market operations became the instrument of choice for affecting interest rates and member bank borrowing. Board members could reject the reserve banks’ decision, but they could not order the banks to buy or sell. That decision remained with the directors until changed by the 1935 act. The 1935 act not only placed all Board members on the Federal Open Market Committee, for the first time it gave the Board a majority of the votes.
During the years of depression and war, the Board was slow to use its powers. Regular open market operations did not begin until the Federal Reserve was again independent.
In 1927 Strong decided to help Britain remain on the gold standard by lowering interest rates, without first consulting the Board or other governors. The Board and some of the governors later concluded that Strong erred. They blamed the decision for the stock market boom and blamed Strong for the mistake. The Banking Act of 1933 stripped New York of its dominant role. After devaluation of the dollar, control shifted to the Treasury.
WHAT REMAINED IN 1951?
Much of the original plan and organization did not survive to 1951. Gold remained part of reserves, but the dollar, not gold, became the world currency. The Federal Reserve neither thought nor acted as if interest rates and money creation depended on capital flows. Monetary policy became discretionary. In the 1920s the Federal Reserve sterilized part of the gold inflows. In the 1950s it ignored them as a reason for policy action. Increasingly, domestic objectives became the main guide to action.
Vestiges of the real bills doctrine remained part of Federal Reserve thinking. Credit controls such as regulation of down payment requirements and length of loan reflected the mistaken idea that the Federal Reserve could control inflation and the quantity of money by controlling the type or quality of credit. Later these ideas faded away, encouraged both by the difficulty of administering controls and by their ineffectiveness as an anti-inflation policy.
Open market operations in government securities had much earlier replaced the discounting of eligible commercial paper as the principal means of intervening. These operations were more efficient. They did not require decisions about what was eligible, and they did not require the Federal Reserve to accept credit risk. The Federal Reserve determined the size and timing of purchases and sales.
One of the Federal Reserve Act’s major innovations removed government securities as collateral behind Federal Reserve notes. The intent was to make note issues more “elastic,” capable of expanding and contracting with commerce, agriculture, and trade. When borrowing declined in the 1930s, the Federal Reserve had to use more than the required percentage of gold as backing for its notes. The Glass-Steagall Act of 1932 reversed the original innovation by permitting the Federal Reserve to use government securities in place of eligible paper as backing for its note issue. Originally a temporary measure, after several renewals the use of government securities as collateral became permanent. Later, Congress removed the required gold backing.
The change in collateral behind notes symbolizes the decline in the real bills doctrine as a guiding principle. The doctrine required procyclical monetary expansion: the Federal Reserve provided additional currency and reserves as the economy expanded and withdrew currency and reserves in economic contractions. The revised Federal Reserve Act, in 1935, retained “the needs of commerce” as a policy objective but added “the general credit situation.” The Employment Act of 1946 did not impose a clear objective on the Federal Reserve, but it emphasized employment and production. Maintaining production and employment required countercyclical policies.
The 1946 legislation suggests the change in public attitudes about the role of government. The change affected the Federal Reserve by endorsing its transformation from a largely passive authority to an activist policymaker. The 1913 Federal Reserve Act gave little scope for discretionary action. By the 1950s, a generation trained in Keynesian analysis rose to prominence at the Federal Reserve and elsewhere in society. Its members believed that budget policy would have the senior role. The role of monetary policy was secondary, supportive of fiscal actions, but useful as a means of keeping interest rates from rising. The emphasis on interest rates fit well with traditional practices.
POLICY LESSONS FROM THE EARLY YEARS
The wide range of monetary experience—wartime inflation, deflation, economic expansion in the 1920s, depression in the 1930s—provides evidence of the relative roles of money and interest rates in the transmission of central bank actions. In some cases money growth falls as interest rates rise or money growth rises as interest rates fall. Since changes in money growth change interest rates, binary comparisons cannot distinguish in these cases whether the transmission of monetary impulses operates principally through changes in interest rates or through changes in money operating through other relative prices and real wealth.
Previous chapters showed that at times interest rates and money growth moved in opposite directions. In 1937–38 and 1947–48, deflation occurred with the short-term interest rate near zero. In both cases the economy recovered without much expansive action by the Federal Reserve. Deflation increased real money balances and real interest rates. The increase in real money balances dominated the effect of the higher real interest rate; output and economic activity increased. These experiences contradict the belief that monetary policy becomes ineffective when the short-term interest rate remains close to zero.
The 1920–22 experience was similar. The short-term interest rate was not zero in this case, but the economy experienced severe deflation. As prices fell, real balances and real interest rates rose. Falling prices also attracted gold from abroad, increasing the monetary base. The ex post real interest rate on government bonds reached 37 percent at its peak. Nevertheless, economic activity and output recovered, consistent with the increase in real balances but contrary to the rise in the real interest rate.
Chart 8.1 shows changes of the real monetary base and the real interest rate during most of the early Federal Reserve history. Growth of the base is measured year to year. The year-to-year change in the GDP deflator measures the rate of price change subtracted from the Treasury long-term rate to convert nominal rates to real rates. The very high real interest rates in 1921 and 1931–32 reflect the severe deflation at these times.
Real base growth is negative before the Great Depression and in its early years. Ex post real interest rates were comparatively high, above the average for the period shown in chart 8.1 but consistent with cyclical peaks in the 1920s. Both measures suggest that monetary policy was restrictive in 1928–29, contrary to the interpretation made at the time. The data for the late 1920s suggest that a productivity-based expansion, as industry adopted new technologies, was ended at least partly by a deflationary monetary policy.
At the start of the depression, real base growth remained low and ex post real interest rates rose. Base growth rose in 1931 and remained high under the impact of the currency drain. The real interest rate is a better predictor than real base growth for this exceptional period.3
Notable also is the collapse of real base growth in 1937 and renewed expansion in 1938. The real interest rate and real base growth moved together in the early postwar years. The common movement reflects the rate of price change, highly positive in 1946, modestly negative in 1948–49, briefly positive at the start of the Korean War in 1950–51. After each of these periods, economic activity moved in the direction implied by base growth.
We can summarize these data in three propositions:
Proposition 1: when growth of real balances rises sharply, expansion follows whatever happens to the real interest rate. Some examples are 1921, 1934–36, 1939–41, and 1943–45. An exception is 1931–33.
Proposition 2: when real balances decline, or their growth is comparatively slow, the economy goes into recession even if the real interest rate is comparatively low or negative. Examples are 1920, 1923, 1926, 1929, 1933, 1937, and 1947. An exception is 1941.
Proposition 3: if the real interest rate is comparatively high, the economy expands if real balances rise and does not expand if they fall. Examples are 1921, 1925, 1927, and 1938–39. Again, 1931–33 is an exception.
These comparisons suggest that the Federal Reserve erred by ignoring the information in the growth rates of real and nominal balances. For short periods, changes in real balances may have little information. The data suggest, however, that attention to money growth would have enabled the Federal Reserve to avoid its largest errors.
The errors the Federal Reserve made in the years 1913 to 1951 were not unique to the System. The few critics of the real bills doctrine and the gold standard were out of step with the dominant views of the period. Many shared the belief that the Federal Reserve could not have prevented the Great Depression or reduced its duration. Historically low nominal interest rates were considered relevant evidence. The view that monetary policy was akin to “pushing on a string” antedates Keynes’s liquidity trap.
Similarly, many bankers and economists as well as ordinary citizens believed that the gold standard was the correct way to harmonize international monetary policy. Efforts to restore the gold standard in the 1920s, and to fix exchange rates within a gold-based system, met little opposition. Many opponents of the Bretton Woods Agreement criticized its differences from a gold standard.
The gradual dissemination of Keynesian ideas in the 1940s slowly transformed the consensus view. Keynes’s emphasis on the role of interest rates and neglect of money fit well with the views widely held by central bankers and in time displaced them. The change to activist, discretionary monetary policy that produced the Great Inflation of the 1970s had not yet occurred by 1951, but important changes had been made. The Federal Reserve gained scope for a more independent, discretionary policy. The United States had an ample supply of gold and, like other parts of the government, a mandate to maintain a high level of employment.
Increasingly, the public looked to government to manage the economy. Within a few years, governments would look to their central bankers to take a leading role in making the macroeconomic policies that first produced the Great Inflation and then learned how to control it.
The shift toward government responsibility required a change in the intellectual consensus on two issues: the roles of gold and government budget deficits. Although some populists opposed the gold standard in the nineteenth century, by 1900 most contemporary opinion in the industrial countries, and many others, viewed the gold standard as the proper way to restrict monetary policy and prevent long-term inflation. The gold standard was a main issue in several presidential elections in the United States. Each time, the gold standard candidate won.
This consensus no longer existed in the 1950s. The population had become more urban and more educated, the country more industrialized, and the workforce more unionized. The public in many countries favored policies that stabilized output, even if the currency value changed.
The belief that balanced budgets should be the norm except in wartime gave way to a loose commitment to cyclically balanced budgets. When private spending declined, government deficits could replace private spending until employment rose.
Weak attachment to the old standards of financial rectitude left the financial system without a belief system that central bankers could appeal to. The new consensus eliminated what had gone before without offering a clear set of rules. At the next stage in the evolution of central banks and governments, the major problem was to learn how to operate in the new, more discretionary environment.
1. The European Monetary System suggests that this statement is no longer true.
2. The Treasury also pointed out that section 10 of the 1913 Federal Reserve Act gave the Treasury power to override the Board in the event of conflict (Subcommittee on General Credit Control and Debt Management 1951, 28). The wording is: “Wherever any power vested by this Act in the Federal Reserve Board or the Federal Reserve agent appears to conflict with the powers of the Secretary of the Treasury, such powers shall be exercised subject to the supervision and control of the Secretary” (Krooss 1969, 4:2450). The section protects the Treasury against any interpretation of the Federal Reserve Act that limited the Treasury’s authority. The Treasury’s interpretation seems extreme.
3. The real money stock, M1/p, fell.