Thank you, Mr. Chairman, for giving me the normal role as speed bump. In that effort we are now handing out material to which we will be referring. If someone will let me know when the doughnut truck comes, I’ll pick up the pace. [Laughter] The Subcommittee on Communications began its work only fifteen months ago, [laughter] which seems like yesterday (at least in geological time). Given the considerable discussion the Committee has had since then on its communication policies, it seemed appropriate to take stock. In particular, I sent around a list of questions designed to elicit your views on the potential roles of the survey of economic projections, the minutes, and the statement in refining your dialogue with the public. The material with the cover, “FOMC Communications,” is designed to help to organize this discussion.
With regard to the enhanced economic projections process, the survey results circulated on June 15 reported general support for the key features of the process, with three notable exceptions, given at the top of exhibit 1. The subcommittee tried to address those concerns in the modified projections process used for this meeting. So the first question listed at the top of your first exhibit boils down to, how did they do? Did the changes address the concerns you expressed in answer to the May survey? In particular, some of you preferred not to share your forecast submissions, mostly on the grounds that it might make the discussion of the economic outlook at the meeting more inflexible. But some of you held that sharing was important, both to help inform your internal deliberations and as a source of material for the forecast write-up. As a compromise, the submissions were circulated on an anonymous basis, with an opt-out clause that one of you took this time. Some of you were concerned about specificity—particularly in writing down an explicit path for the federal funds rate and in quantifying uncertainty. Those questions were turned into qualitative ones that asked for comparisons with the staff policy assumption and historical uncertainty.
Another process question is item 2. In what form should the projections be released? Debbie Danker’s memo identified the three options listed in the exhibit, which basically differ by the extent to which the narrative description is tailored to suit different purposes. Option 1 is one-size-fits-all: The staff would produce a single document describing your economic projections, which would be dropped into the Committee-approved minutes and repeated verbatim in the Board-approved Monetary Policy Report. The chief advantages of this option are that the vetting process is tried and true and that there will be consistency across Federal Reserve documents. The disadvantage is familiar to anyone who has bought clothes off the rack in a big box retailer—the fit will not be perfect across documents. Option 2 opens the door to some variety. The Committee-approved minutes would include a brief description of the forecasts, as is the practice now. A separate document would give a fuller account of the projections. In that case, there would be more flexibility as to who approves the document and when it is released. This poses a tradeoff that has come up in previous discussions in that the earlier the narrative is released, the more useful it will be in explaining the policy decision but the more complicated the governance process will be. Option 3 is the bespoke alternative in which separate, complete discussions would be prepared for the minutes and the Monetary Policy Report. The fit will be better, but carrying it out will be more expensive, and even subtle differences may draw attention.
The third question asks when the projections should be finalized. The choices essentially are the day of the meeting, the end of the week of the meeting, or as late as practicable to be close to the official release of the document. But you have a more fundamental question to answer first. What is the purpose of the economic projections? If they are supposed to help explain your most recent policy choice, they should be conditioned on the same information set that the policy decision is. That is, they should be finalized as soon as possible after the meeting so that they are not contaminated by post-decision information. That is the rule we now use for the minutes. If, in contrast, the Committee’s objective is to release an up-to-date assessment that sheds light on future policy actions, then the individual projections should be nailed down only close to publication day. The middle ground of closing the books on the projections a few days after the meeting seems unsustainable to me, particularly if the narrative description is to be included in the minutes. The first time an important piece of news hits within the window between policy decision and update deadline, the explanation of the policy action will become muddled. Some immediate feedback on that score will be helpful because the schedule for this round allows for your projections to be updated through close of the week. If so, you would seem to have to incorporate Friday’s chockfull data calendar. Do you really want to do that, particularly as it may lead to confusion in the minutes about what you knew and when you knew it?
The next exhibit focuses on the minutes. You have 2½ years of experience under your belt in publishing the minutes three weeks after the day of the policy decision. In that time, the document has received more attention from you in the editing process and from market participants and the press on its release compared with the previous regime of delaying publication until after the next meeting. How do you assess the benefits and costs of further expediting the minutes? The benefits, listed at the left, are the same as discussed in 2004. A document made public closer to the day of the decision will be more of an aid to the private sector in understanding the current outlook and the prospects for policy. In that regard, you will be able to use that material from the minutes for public statements more promptly. A speedier release may facilitate making the post-meeting statement shorter or less substantive. I point out, though, that my sense in 2004 was that a few of you supported expediting the minutes in the hope that the statement would subsequently be shortened, but that did not pan out. To appreciate the costs listed at the right, you should understand that we use the three weeks after the meeting partly to accommodate your busy schedules and partly to provide a cushion so that the drafting iterations can converge without a dissenting vote. The costs, then, would include your increased effort to ensure that your schedules align with that of the drafting and approval process for the minutes. You should also recognize that less drafting time might lead to more compromises that make the minutes less informative so as to avoid dissent, and more staff resources will be needed to prevent errors. Simply put, with fewer hours, we need more eyes looking at the document.
Of course, the decision on expediting the minutes further may interact with other potential decisions on communication policies, especially those regarding the enhanced projections process. In particular, if you decide to include a narrative description of your forecasts in the minutes, you may want to delay any speeding-up of the minutes until you are comfortable with the new process. You may also want to consider ways to make drafting and commenting easier. That is the subtext underlying question 5. How do you assess the current content of the minutes, including the staff’s description of recent data? Some time ago, it was suggested that the first eight to ten pages, which provide a backward-looking review of recent economic data, be relegated to an annex. In a similar vein, it could be described explicitly as a staff summary—with its production perhaps even linked up mechanically to a modified Part 2 of the Greenbook. That way, you would be responsible for commenting only on the forward-looking and policy portions of the document, and some of the pressures on staff time could be relieved. If you have any other views on the content of the minutes, today probably would be an appropriate time to raise them.
The last exhibit raises issues that have not been addressed for some time, namely the role of the post-meeting statement. Some of you might consider the statement as a vehicle designed only to convey coarse signals about policy, with more-nuanced information provided in subsequent communications. Others might see it as important that the statement be able to stand on its own as a reasonably complete explanation of the Committee’s monetary policy decisions and intentions. What is the appropriate role of the statement in light of the changes to the Committee’s other communication devices (question 6)? The statement’s role, which can be described in terms of a monetary policy rule, has varied over the years. At times the statement has provided hints about the odds on future action by characterizing the left-hand side of the policy rule—that is, by directly addressing the path of the federal funds rate. At times the statement has described the right-hand side of the policy rule—that is, the risks to the dual macroeconomic objectives. That is another way of stating question 7: If you believe that the statement should provide guidance about the outlook, should that guidance be couched in terms of the policy interest rate or the dual objectives? If you decide to describe the economic outlook, you will have to settle on the policy assumption underlying that projection, as in question 8. The conditioning possibilities include the appropriate path of monetary policy, some reading on market expectations, or an unchanged policy rate. This choice may interact with the role you envisage for your economic projections in explaining policy decisions.
As for question 9, your ambition regarding the complexity of the statement will determine how much time needs to be devoted to drafting. Because of constraints as to how long you can meet as a group, a complicated statement necessitates pre- meeting consultation. That may get the nuance right, but it may also force you to settle on a policy view before you have had the benefit of consulting with your colleagues on the Committee. Moreover, it shifts some of the policy discussion out of this room, for which a transcript is kept and minutes are produced, and into informal back channels. A statement that could be agreed upon mostly within the confines of these four walls is probably one that is short and relays a routinized risk assessment, given your long-held view that “nineteen people cannot edit the statement on the fly.”
Question 10 raises one last governance issue: Who owns the statement? At the inaugural of the age of statements in 1994, the Committee formally directed the Chairman to explain its policy action. Over the years, the statement has come to be viewed as a Committee product, with the words sometimes viewed to be at least as important as the immediate policy action. But you still vote only on the policy action and the risk assessment. Should that continue, or should you be asked to vote on the statement in its entirety?
These three sets of questions are interrelated, and there is no obvious place to wade into this thicket. One possibility would be to conduct two go-rounds on these issues, with the first covering your views on the enhanced forecast and the minutes and the second addressing the statement. The first two issues are closely related and perhaps closer to closure. I suspect that you are closer to the beginning of your discussion of the statement than to the end. That concludes my prepared remarks.