The Federal Reserve is to announce today, as it does at this time each year, its revisions of last year's money supply figures. The revisions are made because the Fed has more complete information now and because it has recalculated the seasonal adjustment factors.
Seasonal adjustment of statistics, a process that probably strikes most people as being dry as dust, is a complex, difficult procedure that can generate heated arguments among statisticians and economists because there is no "right" way to do it.
Whatever procedures are used, the seasonal adjustment of the money supply figures is critically important to the policy makers at the Federal Reserve who try to influence the economy by controlling the growth of money.
The targets are set in seasonally adjusted terms, while the Fed staff tries to hit the targets by intervening in a real world, not a seasonally adjusted one.
For the past three years the staff's job has been particularly difficult because the credit controls that the Fed slapped on the economy in 1980 at the urging of President Carter severely distorted the normal flow of money and credit that year. Since the techniques of seasonal adjustment require that patterns be estimated over a period of several years, 1980 is still giving the Fed staff headaches.
As each additional year's worth of data is plugged into the equation, a year further back in the past is dropped under the so-called X-11 procedure.
Thus, with new data for 1982, the Fed can re-estimate the seasonal factors for 1982 giving less weight to what happened in 1980. Of course, new estimates have been made for 1983 as well.
"The 1980 year has messed us up for some time," noted one staff member involved, who, like most Fed staffers, declines to speak for attribution. "It messed us up in 1981 and we will get some substantial revisions in 1982."
The Fed tried to get around the problem by using an economic model to come up with a more normal pattern for 1980 and using those figures to reduce the distortion of credit controls. That helped, but apparently not all that much.
Now there is a new distortion, the enormous accumulation of funds in the new money market deposit accounts: more than $200 billion in less than two months.
There are no historical data on which to base any seasonal adjustment for the portion of the money supply in these accounts. Nor can the Fed staff be sure that the seasonal factors estimated in the past still hold for the funds that remain in other types of accounts.
"There has been some process of selection, and the funds that remain may behave differently," the staff member said.
Other money supply revisions announced today will be based on additional information gathered from financial institutions that do not report more than quarterly to the central bank.
These changes, known as benchmark revisions, are now much smaller than they used to be as a result of more frequent reporting by many institutions under legislation passed in 1980.
Nowhere else in government is policy so keyed to such a volatile set of statistics collected on a daily basis and published on a weekly or monthly basis.
The arcane and uncertain processes of seasonal adjustment are just one more hazard for the policy makers at the Fed.