A History of the Federal Reserve, Volume 2 (98 page)

Bagehot (1873) is famous for proposing a crisis rule: lend freely at a penalty rate against acceptable collateral. Central bankers who cite this rule often neglect another main message. Bagehot did not criticize the Bank of England for failing to lend. He insisted that the Bank announce its policy in advance and follow it. In more than ninety years, the Federal Reserve failed to announce its strategy for responding to crises. Sometimes it lends against collateral; sometimes it supports failing institutions with help from other agencies like the Federal Deposit Insurance Corporation. Past practice provides no guide to future conduct. By failing to announce a strategy, the Federal Reserve encourages failing institutions to press for assistance, urging help to avoid calamity. Absence of an explicit policy increases uncertainty. As Bagehot emphasized, market chaos continues until policy is known and implemented.

SOME PRINCIPAL ERRORS

In the years 1951 to 1986 the Federal Open Market Committee never formulated or accepted a theory relating its actions to economic outcomes. Under the convention known as the “Riefler rule,” the Board’s staff did not make explicit forecasts until the mid-1960s. The staff then began to build an econometric model of the economy and to revise it from time to time. But judgmental forecasts coexisted with econometric forecasts, and the staff adjusted the econometric forecasts using judgment and new information. By the 1980s the Board’s staff and the staffs of several reserve banks used econometric models to forecast and brief their principals. There was neither a common model nor general agreement on what it should contain. More importantly, the members of FOMC and its chairman remained appropriately skeptical about the relevance or accuracy of their staffs’ models.

In the 1960s, many in the economics profession believed, or perhaps hoped, that large econometric models would provide consistently reliable forecasts of economic variables. In the 1970s, Robert Lucas (1976) showed why these hopes would not be realized. Economic outcomes depend on expectations. Responses adjust; people learn from observation and experience. Often large changes occur after forecasts have been made.

3. The Shadow Financial Regulatory Committee and several similar groups are exceptions to this statement. A main incentive is the compensation arrangement. Investment banks pay large bonuses. This encourages purchases and sales of highly risky assets that pay large commissions.

At the start of the period, in the 1950s, the Federal Reserve relied mainly on judgments about market conditions. Chairman Martin had no interest in economic or monetary explanations of events. Winfield Riefler, his main staff adviser, used a simple rule of thumb to guide policy actions—on average keep the sustained growth of money about equal to output growth. The FOMC met every three weeks to make a decision for the next three weeks. It did not follow Riefler’s rule or any other. Members used several different policy indicators ranging from explicit values of free reserves, borrowing, or interest rates to color, tone, and feel or money market conditions. No one reconciled the members’ statements, and no one could because the members did not have an agreed framework. That left decisions to the account manager and perhaps to Martin.

The 1950s were atheoretical and procyclical. Members interpreted interest rate increases or reductions in free reserves as tighter (more restrictive) policy and increased free reserves as easier (less restrictive) policy much as they had used borrowing in the 1920s. Often the main reason for the direction of change in free reserves was a change in member bank borrowing. In a slowing economy the public borrowed less, and the member banks repaid borrowing from the Federal Reserve. The Federal Reserve interpreted the increase in free reserves as evidence of easier policy mainly because free reserves rose and short-term interest rates declined. But fewer reserves meant a decline of the monetary base and money; to a monetary economist, the decline meant that policy was more restrictive. In recovery and expansion the opposite occurred. The mistaken interpretation of borrowing and interest rates contributed to procyclicality of monetary policy.

The System’s standard explanation reasoned that banks were reluctant to borrow. When they repaid, they reduced their assets; they contracted, reducing aggregate reserves. But with fewer aggregate reserves, other banks borrowed, expanding money and credit. The Federal Reserve continued to hold this mistaken interpretation of borrowing even after it became clear that borrowing increased when it became profitable to borrow. Eventually the staff model made borrowing depend on the profitability of borrowing, but the FOMC was slow to accept that idea.

Although recessions occurred in 1953–54, in 1957–58, and in 1960–61, the Federal Reserve successfully maintained prosperity and low inflation to the mid-1960s. Despite the deficiencies and errors of its framework, policy was more successful than in the late 1960s and 1970s. Under its 1951 Accord with the Treasury, the Federal Reserve regained much of its independence, but it retained responsibility with the Treasury for successful debt management. Its role was to maintain unchanged money market conditions, called even keel, when the Treasury sold securities. During the
Eisenhower years, the administration favored balanced budgets except during recessions. Consequently, fiscal actions and deficit finance put much less pressure on interest rates than in later years, so money growth and interest rates remained generally moderate, and inflation was generally low. President Eisenhower expressed concern about unfavorable long-term consequences of anti-recession policy. After the deep 1957–58 recession, the Eisenhower administration promptly reduced its large budget deficit and rejected tax reduction because of its long-term budget consequences. The Federal Reserve moved to reduce inflation, reestablishing price stability by 1961. This was one of the rare occasions when policy acted to achieve long-term stability. This policy decision did not prevent tax reduction. The new administration reduced tax rates in 1964. More than thirty years later, President Clinton rejected proposals for tax reduction. The second Bush administration implemented them soon after taking office.

President Eisenhower started holding regular meetings to discuss economic policy that included Chairman Martin, but he did not pressure Martin and respected central bank independence. Principal members of Congress criticized the Federal Reserve’s “bills only” policy. With support from many academic economists, they blamed the increase in long-term interest rates for slow growth. This was an analytic error; it claimed that the Federal Reserve could change the shape of the yield curve.

This claim became the basis of policy in 1961. The Federal Reserve accepted administration urging to buy long-term and sell short-term debt. Evidence suggests that the policy failed to achieve its objective of twisting the yield curve.

A major change occurred in the 1960s. The new administration gradually introduced activist Keynesian policies.
⁴
In addition to “twisting the yield curve,” it urged the Federal Reserve to coordinate its operations with administration fiscal actions, called policy coordination. Many academic economists favored coordination. In practice, this meant financing more of the budget deficit by monetary expansion. This reduced central bank independence but did little harm as long as deficits remained small.

Deficits to finance the Vietnam War spending and for the so-called Great Society increased the size of deficits. Chairman Martin often said that the Federal Reserve was independent within government. In practice, this meant that the Federal Reserve was free to raise interest rates enough to stop a private spending boom. But Congress approved government spending without increasing taxes, so the Federal Reserve was obligated
to help finance the deficit. Martin’s beliefs about independence and policy coordination meant that money growth and inflation began to rise. Chairman Martin was strongly opposed to inflation, but his decisions started the Great Inflation. The Federal Reserve made the mistake of sacrificing its independence to coordinate. Many years passed before it restored independence. It learned that coordination worked only one way. The Federal Reserve supported administration actions, but there was no reciprocity.

4. Congress took a strong Keynesian approach in the report called
Employment,
Growth
and
Price
Levels
(Joint Economic Committee, 1960b, 6). The Federal Reserve demurred.

The simple Keynesian models used at the time encouraged the belief that economists could control short-term changes in such principal economic variables as economic growth, inflation, employment, and investment. This was hubris. The models at the time did not recognize anticipations and uncertainty, did not distinguish temporary and persistent changes, or distinguish real output and inflation, or real and nominal values of interest rates and exchange rates. As the text notes, some FOMC members pointed to these errors or omissions at times, but the comments did not influence either theory or policy.

Perhaps the two most costly errors in the 1960s were the sacrifice of Federal Reserve independence and reliance on belief that economists could improve welfare by trading a little more inflation for a lower unemployment rate. At his meetings with Presidents Kennedy and Johnson, Chairman Martin rarely made explicit commitments, but he was slow to raise interest rates despite recognizing the start of inflation after 1964.
⁵
In 1968, he agreed that the administration’s tax surcharge would lower interest rates. He did not promise to lower them by Federal Reserve action, but he responded to pressure to do so. By year-end he acknowledged that he had made a mistake.

Early versions of the Phillips curve implied that increased inflation would permanently lower the unemployment rate. This was based on an error—failure to distinguish nominal and real responses. Friedman (1968b) pointed out the error; he explained that persistent inflation would increase expected inflation and restore equilibrium employment. Any reduction in unemployment would be temporary, but the increase in inflation would persist.

Policy actions in the 1960s were based on two mistaken beliefs that persisted until disinflation policy began in 1979–80. First was the claim that in a modern market economy like the United States, inflation started to increase before the economy reached full employment. The policy solution called for government intervention in the wage and price process to
encourage non-inflationary settlements. In practice, the policy failed in the United States and abroad. It did not distinguish between price levels and rates of price change.

5. In 1962, he committed to keeping free reserves within a range. He did not tell the FOMC.

Second was dismissal of the role of money in monetary policy. For many years, the Federal Reserve gave most attention to money market conditions
⁶
and to credit and interest rates. Later, it introduced a “proviso clause” that called on the manager to raise market rates if some measure of bank credit, reserves, or money grew faster or slower than a specified rate. The manager rarely acted on the proviso. The Board staff often explained the failures by claiming that the demand for money shifted. They never provided evidence, not even when challenged to do so. Another explanation was that responding to money growth required large changes and much greater variability of market rates. A solution adopted by the German Bundesbank and later the European Central Bank ignored short-term variations but responded to medium-term growth of a principal monetary variable.

Nelson (2003a, 1054) argues that money “serves as a proxy for a variety of yields that matter for aggregate demand.” As he notes, this is the implication of studies of the demand for money by Friedman, Schwartz, Brunner, and Meltzer. In periods when markets function well, short-term rates may summarize financial information. In periods when expectations are changing, other relative prices remain important.

Additional support comes from an exhaustive study of 1,000 demand functions for money. Knell and Stix (2004, 20) conclude in part, “In cases where a short rate is included with a long rate the short rate acts as an own rate (positive sign) for broad money but not for narrow money (negative sign). Concerning the size of the impact it is found that the sensitivity of money demand with respect to long rates is higher than with respect to short rates.”

A reason often given for neglecting money in policy analyses is that monetary velocity is highly variable. Chart 10.1 relates base velocity to a long-term interest rate. Two notable features of these data are: (1) when interest rates in the 1960s returned to the levels reached in the 1920s, velocity returned to those levels also; (2) during the disinflation of the 1980s, velocity and the interest rate declined along the path on which it rose during the peak years of the Great Inflation.

A possible third error mentioned in the text was the often stated idea
that the Federal Reserve should create uncertainty. Rational expectations implies the opposite. It took time for the Federal Reserve to recognize that it depends on market responses, so it has a strong interest in informing markets about its policy and its intentions, beliefs, and interpretations. There are, of course, limits to what it can say because it is necessarily uncertain about the future. By letting the public know about the rule or quasi-rule that it follows, it permits the market to interpret incoming data more accurately.

6. In Chapter 4, I quote Stephen Axilrod, a principal staff member, expressing frustration about money market conditions. He noted that emphasis shifted from borrowing to free reserves, to color, tone, and feel, and other indicators. Money market conditions were often vague and loose and not closely related to inflation and growth.

Chapter 3 shows that several members of the FOMC, led by Sherman Maisel, recognized some of the principal problems with operating procedures in the 1960s. Members mentioned uncertainty, the problem of distinguishing permanent and temporary changes, and absence of agreement on how Federal Reserve actions should be judged and how they affected the economy. Vice Chairman Balderston at times commented on the absence of procedures to achieve long-term objectives. Chairman Martin showed little interest and made few changes.

The new Nixon administration in 1969 brought a different team of economists to give advice and, after a year, a new chairman of the Board of Governors. The new team accepted that excessive money growth was the principal cause of inflation and that the long-run Phillips curve was vertical. Higher inflation could not permanently reduce the unemployment rate.