When it comes to crystal-ball gazing, the Fed trounces the private sector.
New York Times; New York, N.Y.; Nov 21, 2002; Hal R. Varian

THE Fed gave the economy a flu shot earlier this month by cutting interest rates half a point. But it was quick to point out that, as with a real flu shot, this was intended only as a preventive measure.

''The economy's most likely projection is to come out of this soft spot and to start accelerating,'' the Federal Reserve chairman, Alan Greenspan, told Congress.

Other Fed officials are also positive about the economy. Gary H. Stern, president of the Federal Reserve Bank of Minneapolis, said, ''There is no question that an economic expansion, an economic recovery, is under way.'' And Roger W. Ferguson Jr., the Fed's vice chairman, said, ''The pieces are in place to engender a gradual strengthening in economic activity.''

Unfortunately, business leaders do not share the Fed's optimism.

According to a report by the Business Roundtable, 60 percent of executives surveyed said they planned to eliminate jobs next year, while only 11 percent thought they would add jobs. Similarly, 19 percent expect to increase their investment next year, while 24 percent expect to reduce it.

Whom should we believe: the Fed or the executives? After all, the executives are actually making economic decisions, while the Fed is only trying to forecast what they will do. What do those guys in Washington know, anyway?

Plenty, it turns out.

For evidence, turn to ''Federal Reserve Information and the Behavior of Interest Rates,'' a paper in the American Economic Review in June 2000 by Christina and David Romer, two economists at the University of California at Berkeley.

The Romers ask whether the Fed has an advantage in predicting economic variables and, if so, what the source of that advantage is. The answer to the first question is a resounding yes. The answer to the second is a bit more tentative: the Fed is better at forecasting than the private sector, but we're not sure why.

The Romers look at two variables: inflation and output growth. Both attract intense interest in Washington and on Wall Street.

Forecasts for both variables are prepared by the Fed staff before each meeting of the Federal Open Market Committee, which uses these forecasts to help set economic policy. The forecasts are closely guarded secrets and are released to the public only after five years.

How good are the Fed's forecasts? To answer this question, the Romers compare the forecasts from 1965 to 1991 with those of several private forecasting services.

The Romers first look at how well private forecasts have predicted inflation, then ask whether the forecasters could have improved their performance if they had been given access to the Fed forecasts. The researchers find that the Fed forecasts offer big improvements over the private ones. In fact, if you had access to both forecasts, a good rule of thumb would be to rely exclusively on the Fed and ignore the private predictions.

Why are the Fed's forecasts so much better? There are several possible explanations. First, the Fed may have inside information about the course of monetary policy since setting that policy is one of its most important jobs.

The problem with this explanation is that monetary policy has little effect on the economy in the short run -- it takes three to four quarters to show up in aggregate statistics. Yet the Fed's forecasts are better than the private sector's even for very short-term forecasts.

Perhaps the Fed gets official data earlier than the private sector? This is also an unconvincing explanation, as the Fed chairman gets access to current economic data just the night before it is released to the public.

The most likely explanation for the Fed's superiority in forecasting inflation, the Romers say, is that the Fed devotes ''far more resources to forecasting than even the largest commercial forecasters.''

It's not that it has better information about the state of the economy, or secret forecasting methods -- it simply looks at a lot more information, has a lot more staff and has a lot more experience. So it is not too surprising that the Fed does better.

These days the big economic uncertainty is not so much about inflation (or deflation) but about output. Everyone wants to know when the economy will resume robust growth. Does the Fed's forecasting superiority extend to G.D.P. growth?

The answer is yes: the researchers find that the Fed is also substantially better at predicting output growth than the private sector. But the Fed's advantage in this case is not so large as with predicting inflation. In the case of output growth, the optimal forecast would combine both the private and the Fed forecasts, though it would give more weight to the Fed's forecast.

In the case of output forecasting, the Romers suggest that the Fed's advantage may in fact be, at least in part, a result of better access to data since the Fed manages the collection of information used to compute the index of industrial production.

Note that the Berkeley economists' study is mostly pre-Greenspan; Mr. Greenspan joined the Fed in August 1987, while the Romers' data ends in 1991. Mr. Greenspan is well known for reading a variety of economic tea leaves, but most of the data he uses appears to come from reports available to the private sector. His forecasting success seems to come not from special ingredients, but from a finely tuned sense of how ordinary ingredients should be combined.

As he indicated last week, the economy is in a soft spot. Consumers have grown more cautious, and businesses have not yet resumed robust investment spending. The Fed's worry now is that excessive pessimism on the part of executives will dampen, or even stall, the current recovery.

This was probably part of the reasoning that led the Fed to make the unusually large cut in interest rates last week. The cost of borrowing is now at a historic low, which should encourage even the most pessimistic executives to think about investing.

Let's hope the rate cut works, for the sake of the economy -- and the Fed's forecasting record.