Vincent Reinhart

The formation of market expectations about the economy as it relates to policy choice by the Federal Open Market Committee has the following three properties: There are a few (usually two) likely choices for the policy outcome of an FOMC meeting. The announcement of the policy action is usually fixed in time (with the rare exception of intermeeting moves). Some members of the private sector are viewed as experts in divining policy intention. These conditions are precisely the ones found in models of information cascades, in which investors defer to received wisdom about a likely action and fail to use their private information efficiently (as in Bikchandani, Hirshleifer, and Welch, 1992). Why policymakers should feel obliged to validate expectations based on the actions of traders who are taking positions because other traders who read something in the newspaper are also doing so is not obvious.

My experience in surveying you has been that if I ask the 19 of you “What is the color of an orange?” I couldn’t be sure of getting a majority on a single answer. [Laughter] This most recent survey was no exception. Almost as many of you strongly endorsed an expedited release of your projections as strongly opposed it. An equal number of you endorsed it as opposed it, and there were two lonely people who were indifferent. [Laughter] Thus, since expediting the release of the forecast is a decision that cannot be reversed, it doesn’t seem appropriate to move forward with a discussion of it today.

The Feds failure to communicate internally owes to an irony of increased openness. In 1994, under considerable Congressional pressure, the Fed became more transparent. One initiative was for historians. The FOMC decided to release lightly edited, but otherwise complete, transcripts of every meeting, five years after the fact.

What followed was a predictable social dynamic that is never factored in by economists in their theoretical reasoning that more transparency is better. Starting that year, speakers at an FOMC meeting were given a rough draft of their remarks a few weeks after each meeting. Most learned, to their surprise, that they were a lot less lucid speakers than they had imagined. Off-the-cuff responses to prior speakers looked unthoughtful in black and white. Almost immediately, some began bringing prepared remarks. This set off a readiness race that ended with virtually everyone reading from prepared texts.

Meetings got longer and less spontaneous. More problematic still, meetings became a less useful way of exchanging information and changing minds. This led to a new dynamic: When policy views are scripted, the window to influence views opens before the meeting, when scripts are being written, not during the meeting, when scripts are being read. Thus, Fed officials give more speeches and interviews before meetings to signal each other what they will read at the meeting. If a policy issue is contentious -- if it is a close call -- then the volume cranks up. It just so happens that the free investing world is listening to that conversation.