The first rule of forecasting should be don’t do it. Nothing good comes from it. The second rule, is, if you give a number, don’t give a date; or, if you give a date, don’t give a number. My rule, the third rule, is, if you have to do it, do it often.
Much has been made of Wednesday’s FOMC “forecast” of the economy being downgraded. Many critics noted that they are just now, with a long lag, catching up to private sector forecasts. Well, duh! The problem is they don’t do it often enough.
Cable TV and the blogosphere is rife with forecasts—multiple forecasts every day from all sorts of people and points of view. As economic conditions evolve, so will the consensus forecast. Even individual forecasters have frequent opportunities to tweak their forecast so that it never strays too far from the consensus, or the obvious. The official FOMC forecasts don’t have that luxury since they come out less frequently, quarterly, I believe. I know my gentle readers know all this already, but the nature of some of the commentary out there suggested to me that a reminder was in order.
I retired from the Dallas Fed and the FOMC in November 2004, so the “forecasting” procedures may have changed since then, but I doubt it. I put quotation marks around the word because it really isn’t a single forecast. The Presidents and Governors, in separate consultations with their respective research staffs, bring their individual guesses to the designated FOMC meeting. They have each made their own assumptions about policy and other relevant factors are likely to be during the forecast period. Efforts to impose some common assumptions on the process were always abandoned and difficult and subject to unintended consequences.
FOMC members probably all assume that appropriate monetary policies will be followed, but there will likely be different views of what is likely to be appropriate. But what about fiscal policy? Will they still assume that fiscal policy will be appropriate? Then, there is the oddity that the forecasters may have some small ability to make their forecast a self-fulfilling prophesy. That always raised the question of whether we were submitting a forecast or a goal, or some combination. Actually, they are calling them “projections” these days.
You are made aware at the meeting of the range of forecasts and where the central tendency seems to be, and are given a couple of days to revise your forecast if you are swayed by where you stood relative to the others. I took pride in never changing mine—foolish pride perhaps. Other than pride, a more practical reason for not changing is to avoid reopening the debate bank home on what the revision should be. You’ve already disappointed some of your economist in choosing your initial submission. A change would run the risk of disappointing the rest of them.
Anyway, after we had a chance to amend the initial forecasts, the Board staff recomputed the ranges and central tendencies, and often omitted outliers before they become part of the official FOMC record and, these days, announced at a press conference.
Like sausage, huh?