164. Of course, it is possible that the System maintained secrecy for reasons it was unwilling to state publicly. Cukierman and Meltzer (1986) and Cukierman (1992) argued in a rational expectation framework that the System used secrecy to surprise the public when it wanted to increase output or slow inflation.
An internal staff memo gave the same arguments about speculative dangers and precipitate, disruptive responses. It claimed that uncertainty increased policy effectiveness because it made market participants cautious and prevented “precipitous” responses. And it ended with a warning. “If the FOMC is required to publish its Domestic Policy Directive at the beginning of the period to which it applies, its ability to discharge effectively its statutory responsibilities to carry out monetary policy in the national interest may be significantly impaired (memo, Board staff to FOMC, Board Records, April 10, 1978). A subsequent study by Donald Kohn of the Board’s staff summarized the empirical evidence. It concluded that a 10 percent increase in interest rate volatility would increase the difference between bid and asked yields on Treasury bills by one basis point (0.01 percentage points) (memo, Kohn to Arthur Broida, Board Records, July 13, 1978).
When the Board appealed to the Supreme Court, it added that when acting on its account or as agent for the Treasury, transparency would cause commercial harm because the directive authorized buying and selling transactions and so did the instructions from the Treasury. Commercial harm was explicitly mentioned as a reason for secrecy under FOIA, so the Supreme Court in 1981 reversed the decision and found for Federal Reserve secrecy.
Burns’s initial reaction to the district court’s order favored releasing the directive information after a delay. The court order held unlawful the fortyfive-day lag between a decision and its publication. It ordered prompt publication of the directive in the Federal Register. Also, it ordered prompt release of the “reasonably segregable positions” of the memorandum of discussion (minutes). Burns opened the February 17 meeting by recommending that, if the court order applied only to the domestic policy directive, the Committee could “comply with the spirit as well as the letter of the order and make public at the same time the short-run ranges of tolerance adopted for the aggregates and for the Federal funds rate” (FOMC Minutes, February 17, 1976, 2). He favored asking Congress for authority to delay release of the directive for up to forty-five days, and proposed that a subcommittee, chaired by Governor Coldwell, reconsider the information included in the minutes. The members agreed.
165. I list the arguments in the text; Goodfriend (1986, 69–83) presented the Board’s case more fully and offered a careful critique. He, too, found little merit in the Board’s case.
Burns did not seek to release the targets for federal funds and money growth promptly. Instead, the Board’s appeal asked the Circuit Court to reinstate the forty-five-day delay. Thomas O’Connell, the Board’s counsel, opined that the district court had “simply pronounced what the law said” and would be upheld on appeal (FOMC Minutes, March 15–16, 1976, 5). O’Connell believed that annual money growth rates were “understandings, objectives, or goals” and did not have to be published on the same basis as the directive (ibid., 7). The Board’s affidavit also asked to exempt confidential information received from foreign governments.
The “reasonably segregable positions” referred to “facts.” The Board’s appeal concerned mainly matters other than this material. Disagreement arose about what was “fact,” but Burns’s main concern was that released material would get to Congress. They would ask for supporting material, memos, and additional reports.166
Discussion of alternatives to releasing the directive promptly continued in 1976 and 1977. As O’Connell predicted, the Federal Reserve lost on its appeal at the Circuit Court. Much of the December 1977 meeting considered available options if the Merrill opinion prevailed at the Supreme Court and if Congress put the Federal Reserve under the Government in the Sunshine Act. Burns proposed asking Congress to legislate relief by excluding bank examination reports from coverage.
A subcommittee chaired by Governor Partee discussed changing the directive that the Committee wrote for the manager by giving the manager more discretion, replacing quantitative with qualitative information, and widening the ranges for the federal funds rate and money growth rate. The range for the funds rate would be one percentage point, but the subcommittee proposed an inside range of one-half percentage point for desk operations. This was a modest change. The transcript would omit the names and dissents of individual members. These would remain in the Annual Report, as before.
The subcommittee proposed to make few changes in the directive. In part, they did not want to avoid the finding of the court (ibid., tape 1, 17). The main proposal called for instructing the manager to adjust to changing financial conditions following early release of the FOMC record. This would “establish a degree of looseness between the specification of the manager’s actions” and his performance. The purpose was to “provide an uncertainty in the markets mind” (ibid., tape 1, 18).
166. Burns did not hold Congress members in high regard. “It’s a talking committee and that would cause some mischief” (FOMC Minutes, Executive Session, March 29, 1976, 15).
Concern about the release of information led to a discussion of procedures, one of the very few ever held. President Roos (St. Louis) called this a “revolutionary change in procedure” (Burns papers, FOMC, December 19, 1977, tape 2, 17). Henry Wallich proposed that the Committee should test its procedures to see if the range set for the federal funds rate was consistent with control of the monetary aggregates (ibid., tape 3, 1–2). Wallich doubted they would agree to widen the proposed range enough to make the two compatible. President Balles (San Francisco) spoke in favor of a target for the aggregates; Burns agreed that this was technically right but politically wrong. Congress would respond to increases in interest rates “just the way it had been doing” (ibid., tape 2, 16).
The discussion did not reach a consensus, in part because it concerned the FOMC’s response if it lost the Merrill decision in the Supreme Court. It showed, however, that members were aware both of the reasons they did not succeed in controlling the aggregates and the importance of control of aggregates for inflation control. And Burns gave the reason; either he was not willing or did not feel able to assert Federal Reserve independence as a response to congressional complaints.167
After much further discussion, the Federal Reserve in 1994 began to publicly announce its federal funds target and other information at the end of each meeting. Some central banks began earlier. None of the alleged harmful consequences occurred. Some evidence suggests that markets are less volatile. For example, long-term interest rates changed very little in 2004–5, when the Federal Reserve increased the federal funds rate from 1 to more than 4 percent. A main reason was that its actions and talk convinced the market that the target funds rate would likely rise by 0.25 percentage points at each FOMC meeting. Long-term rates appear to have incorporated the anticipated path, so they did not react to each separate announcement.
New Procedures
Aroused by the Merrill case, general congressional concern about excessive secrecy by the administration and the agencies, or specific concerns about
167. Unemployment was a political concern also. Later in the December meeting, Roos described the goals of monetary policy as maintaining economic growth and keeping unemployment at a reasonably low level. Burns responded that it was not the overall unemployment total but concentration among certain groups in the population. He then spoke about race and cultures (Burns papers, FOMC, December 19, 1977, tape 5, 19).
economic policy, the Federal Reserve renewed attention to the directive and release of information in 1976. Coldwell’s subcommittee expressed concern about the ability to conduct monetary policy, if the court required prompt release of the memoranda of discussion (minutes). It proposed to use summaries of the discussion in place of actual statements and eliminate exchanges of views.
&n
bsp; Burns moved to change procedures more radically. Both the Board and the FOMC had kept minutes of each meeting; for the FOMC the minutes started when the FOMC became an official entity under the 1935 Banking Act. Burns proposed to end the transcripts or minutes after the March 15, 1976, FOMC meeting and substantially increase the length of the policy record released after forty-five days and included in the Annual Report. The report would include views of individual members, but they would no longer be identified. They would continue to record dissents, and members would retain the right to give their reasons and identify themselves (FOMC Minutes, Executive Session, March 29, 1976, tape 1).168
Both banking committee chairmen objected strongly. Burns insisted that the minutes were used very little and that the only objections to the proposed change came from the two banking committee chairmen. Burns’s major concern was that a future court might expand the information required to be released and thus become available to Congress.
The FOMC supported Burns and agreed in April to discontinue the memo of discussion (called Minutes here), expand the policy record, and release the record including the directive shortly after the next FOMC meeting instead of after forty-five days (FOMC Minutes, May 1976 tape 1). The vote in May to make this change was ten to one; Coldwell abstained because the decision reduced available information, and Holland did not participate. But Burns argued that the longer policy record increased the information available to historians.169
Under the new procedure, the FOMC secretariat would continue to record and circulate the discussion transcript taken at the meeting as before. Members could correct the record. The intention was that the revised minutes would furnish the basis of the expanded record of policy action and then be discarded. They were not discarded, however. At a subsequent congressional hearing in 1993, the existence of the stored archives became known. Federal Reserve officials admitted the truth, although a congressional staff member, Robert Auerbach, claimed the opposite (Auerbach, 2006); Board staff cited the record, showing that no denials occurred. The FOMC resumed publication after 1993, with a five-year lag, and began editing the material for the years (working backward) from the 1980s to 1976 that had been stored but not published. Historians owe a debt to the House Banking Committee for opening the archives for these historically important years and restoring the prior procedure.
168. We relied on material from the Burns papers at the University of Michigan, hence the references to tapes for 1976 to February 1978. The Burns papers include transcripts for 1976 to February 1978 that did not become available until after there was a draft of this chapter. We appreciate the help we received from the Board staff in later providing the redacted transcripts.
169. Burns reopened discussion of the decision to eliminate the memo of discussion and took a new vote. The result did not change. Governor Wallich suggested that the policy record include names, but Burns objected and did not take a vote. The meeting also discussed a suit by Congressman Reuss claiming damages for losses he incurred because reserve bank presidents voted on FOMC policies but were not appointed by the president and confirmed by the Senate. Legal counsel said he expected the court would dismiss the suit because Reuss lacked standing. He was correct. During the discussion, Burns claimed Reuss should thank the FOMC for keeping inflation from rising more rapidly.
Discussion of the proposal at the May 1976 meeting kept sunshine legislation in the background though it was clearly the main force for change. Thomas O’Connell, the Board’s counsel, expected that the sunshine legislation would not apply to the Federal Reserve. Burns urged members not to discuss applicability of the legislation to reduce the risk that Congress would remove the “ambiguity” in the legislation, on which the Board believed its exemption rested (Burns papers, May 18, 1976, tape 1, 4).
Practice changed less than the discussion suggests. The manager continued to give primary attention to control of the federal funds rate. This was explicit when the manager described his operations. “Incoming data on, and projections of the aggregates are compared with the [FOMC’s] ranges each week to determine the Desk’s posture with respect to reserve provision and the federal funds rate. The Manager’s response to undesired behavior is constrained by a range of permissible variation in the weekly average federal funds rate” (Holmes, 1976, 413).
Burns proposed setting 4 percent bands for money growth rates and an inside (say) 2 percent band that he called a “zone of indifference.” The manager could ignore money growth rates in the zone of indifference but was to respond outside the band. Burns hoped in this way to allow for the imprecision of the money growth estimates and their short-run variability (Burns papers, FOMC, February 17, 1976, 9–13).170 The FOMC referred to the “range of indifference” at a few subsequent meetings. Within a few months, it disappeared from the discussion. FOMC members did not agree on the appropriate size of the range or how to use it.
170. The FOMC based the wide range chosen for the money growth rates on staff work reporting on thirty different seasonal adjustment techniques. They suggested a range of uncertainty about the “true” seasonally adjusted value.
Burns commented that the imprecision of the two-month ranges led him to question their usefulness. He proposed lengthening the period to up to six months. Others responded that his proposal put more weight on uncertain projections (ibid., 15–16). Burns’s proposal suggested more attention to the medium-term and sustained changes. The members did not reach consensus or agree to change.
Operating Procedures
As inflation rose, the FOMC recognized more fully than earlier that targeting the federal funds rate did not achieve its objectives. They experimented over the decade with different ways of reducing or controlling inflation by including in the directive targets for measures of reserves and money. Dissatisfaction with outcomes and problems in reconciling the interest rate and aggregate targets brought the topic back to their agenda several times.
The staff concluded that responding to monthly data was unwise. It reached a correct and long-delayed conclusion that to control inflation and improve outcomes the FOMC had to take a longer view and pay more attention to persistent evidence showing money growth rates above or below target (memo, staff to FOMC, Board Records, February 10, 1976). This was an opportunity to shift more emphasis to longer-term consequences. There is no evidence that the FOMC agreed.
A separate staff report also considered the effect of legislative changes and market innovations on the composition of financial balances. Some of the changes relaxed regulation to help thrift institutions adjust to inflation and regulation Q ceilings. The report concluded that the introduction of NOW accounts and other changes in saving accounts to make them more like transaction balances had reduced demand deposits. The staff proposed that the FOMC should give more weight to the broader aggregates without abandoning M1 (memo, staff to FOMC, Board Records, February 10, 1976).
At a special meeting in March 1976, Governor Robert Holland, as chairman of the Subcommittee on the Directive, proposed replacing the target for RPDs (reserves against private deposits) with nonborrowed reserves (total reserves minus member bank borrowing). Since required reserves depended on lagged deposits, this proposal would renew the role of free reserves (excess reserves minus borrowing) for a short period. The federal funds rate would be a constraint; the manager would target nonborrowed reserves as long as the funds rate remained within the target range. The FOMC would resolve conflicts or inconsistencies by revising instructions.
Following his report, an exchange with Chairman Burns suggests that members of the FOMC did not agree on the operating target. Holland said it was the funds rate.
Burns objected. “That is not our operating target—we use the federal funds rate with a view to reaching certain monetary aggregates.” Holland replied, “I think it is in practice our only operating target and that we need to have a structure that has a reserve measure in it, that is a workable one that the committee can work towards” (FOMC Minutes, special meeting,
March 29, 1976, tape 2; emphasis added). Volcker agreed with Holland. “As long as the Federal Reserve used it [the funds rate] as its mechanism for implementing policy, the market would continue to interpret any move in the funds rate as an indication of a change in policy” (Burns papers, FOMC, March 15–16, 1976, 82).
Surprisingly, the FOMC did not explore the range on the target federal funds rate required to keep the monetary aggregates within their bounds most of the time. Nor did most members accept that to control monetary growth they would have to permit more frequent changes in the funds rate. Instead, Burns asked the staff to report on the forecast error in the aggregates under prevailing conditions. The reports showed large shortterm errors, two-thirds of the time as much as 4 percentage points around the two-month money growth rate (ibid). Governor Wallich concluded the discussion again by pointing out that the System could improve control of the aggregates by permitting greater variability in the funds rate. The FOMC was not ready to accept this proposal.
Holland explained that the main reason for rejecting RPDs was that the manager could not control them effectively. Governor Partee pointed out that the data did not show that a nonborrowed reserves (NBR) target would be better. The staff agreed, but Axilrod said that NBR could be controlled better. Governor Holland said that politics, more than economics, was the reason “for not naming federal funds as the FOMC exclusive operating target” (ibid., tapes 3 and 4). Setting an interest rate “is a kind of lightning rod in terms of some of the public discussion” (ibid.). He might have added that the problem of choosing a nominal short rate was more acute during an inflation when nominal interest rates reached unprecedented levels. He favored allowing the funds rate to move within a one percentage point band adjusted up or down by 0.25 percentage points as growth of monetary aggregates changed. Contrary to the evidence presented earlier, this proposal put excessive weight on short-term, often random changes in the aggregates.
A History of the Federal Reserve, Volume 2 Page 42