What Is Happening in the U.S. Economy?

William Poole*
President, Federal Reserve Bank of St. Louis

AAIM Management Association
St. Louis
Jan. 11, 2002

*Robert H. Rasche, Senior Vice President and Director of Research at the Federal Reserve Bank of St. Louis, provided extensive assistance, but I take full responsibility for errors. The views expressed are mine and do not necessarily reflect official positions of the Federal Reserve System.

Everyone wants to know what is happening to the economy. So do I.

I'm presumed to know, which is why I am here this morning. But the fact is that I am as curious as you to find out the answer to the question. Although I do not have a working crystal ball, I can offer a perspective on where we are right now and on how we got here. I can also offer a perspective on why economic forecasts are as hazy as they are.

Before proceeding, I want to emphasize that the views I express here are mine and do not necessarily reflect official positions of the Federal Reserve System. I thank my colleagues at the Federal Reserve Bank of St. Louis for their comments, and especially Bob Rasche, Director of Research, for his extensive assistance, but I retain full responsibility for errors.

I'll begin this morning by reviewing how the U.S. economy got to its current state. Then, I'll discuss the range of possible outcomes for the economy over the next year or so. Finally, I'll make a few general comments on forecast uncertainty and its relevance for monetary policy.

History

As the year 2000 came to a close, the consensus outlook was that the economy would slow in the first few months of the new year, but that the slowdown would be a brief "V" shaped event that would not end a decade of economic expansion. The Blue Chip consensus forecast released on January 10, 2001 projected 1.9 percent real growth for 2001:Q1, increasing to 3.1 percent by 2001:Q3. As the year progressed, the anticipated slowdown did materialize, but it also persisted, as had not been anticipated. The bottom of the projected "V" came to look more like a "U" as forecasters adjusted their projections of a recovery further into the future. As late as August, forecasters continued to project near-term recovery without a recession. The Blue Chip consensus forecast released on September 10 projected 2.9 percent real growth for 2001:Q3 and 3.6 percent real growth for 2001:Q4.

All bets were off as of September 11. The tragic events of that day were unprecedented and created an environment where forecasters had no relevant history on which to base projections of the future course of the economy. Forecasters quickly concluded that the losses of lives, property and jobs in New York City and the disruption to transportation networks, particularly air travel, would trip the economy into a recession. Considerable uncertainty prevailed about the depth and duration of such a recession. The dispersion of forecasts for the 2001:Q4 and 2002:Q1 was much wider than typically occurs.

In a press release dated November 26, 2001, the Business Cycle Dating Committee of the National Bureau of Economic Research announced that it had determined that a recession was in progress, and called the cycle peak at March 2001. It is impossible to know whether we would have had an official recession without the terrorist attacks; clearly, the economy was not performing well. We might still have had a recession, or perhaps just a pronounced flat spot in a continuing expansion. In the event, the terrorist attacks did occur and tipped the economy over rather decisively into a recession.

The current recession has some fairly standard features, and some that are decidedly atypical. It is not unusual for special events to tip a soft economy into recession. Iraq's invasion of Kuwait in August 1990 is an example. In that case, the cycle peak is dated July 1990. Another example is the General Motors strike that ran from September 1970 to January 1971; without the strike, we might not have had a recession that is today dated as having a cycle peak in December 1969 and cycle trough in November 1970.

Failure to predict a turning point in the economy, or even realize for several months that a peak has occurred, is typical. Missing the onset of a recession is sometimes due to unforecastable events, such as the terrorist attacks, but more generally to incomplete scientific knowledge of how the economy works. Consider the minutes of the August 21, 2001 FOMC meetings, which stated in part that:

The staff forecast prepared for this meeting suggested that, after a period of very slow growth associated in large part with very weak business fixed investment and to some extent with an inventory correction, the economic expansion would gradually regain strength over the forecast horizon and move to a rate around the staff's estimate of the growth of the economy's potential output. …

In the Committee's discussion of current and prospective economic developments, many of the members commented that the anticipated strengthening in economic expansion had not yet occurred and, indeed, that the economy and near-term economic prospects appeared to have deteriorated marginally further in the period since the previous meeting.

Another dimension of the 2001 recession, typical of past cyclical behavior, is that substantial inventory accumulation occurred as the economy's growth rate slowed in 2000. Because inventories had accumulated beyond the levels firms wanted, they began to reduce production. In the first quarter of 2001, this process had reached the point that firms cut production below the level of final sales, so that inventories were actually reduced. In the second and third quarters, firms continued to liquidate inventories, and by increasingly large amounts. Although we do not have official data yet, forecasters believe that the rate of inventory liquidation was larger yet in the fourth quarter.

The importance of the inventory cycle can be shown with a few numbers. Using quarterly data, the peak quarter was 2001:Q1. A reasonable guess for 2001:Q4 puts the decline in real GDP from the first quarter at only 0.7 percent. However, that decline was more than accounted for by inventory liquidation. Over this period, final sales likely rose slightly, by 0.2 percent. Expressed at an annual rate over these three quarters, inventory liquidation knocked about 1.2 percentage points off the economy's growth rate.

The inventory cycle is a standard feature of almost all recessions. At some point, firms get inventories down to desired levels, and production then catches up with final sales. Assuming that final sales growth occurs this year at a modest pace, as most forecasters expect, the recession will come to a natural end when inventory liquidation is complete.

In many ways though, the current recession is atypical of U.S. economic history. First, industrial production peaked in September 2000, well in advance of the cycle peak date. Second, housing investment, which historically has been a leading indicator of a cycle peak, has remained on a high plateau throughout 2001. By way of contrast, consider the 1990-91 recession; housing starts peaked in January 1989, well before the cycle peak, and by January 1991 had dropped by 51 percent.

Third, consumer expenditures on light vehicles have continued at near record levels, albeit with considerable support through price reductions in the form of zero interest rate financing and/or substantial rebates. Data released last week indicated that sales of light vehicles during 2001 were second only to the record sales in 2000. More generally, durable goods consumption expenditures, which typically decline substantially during a recession, have held up pretty well this time.

Fourth, productivity growth has remained strong despite the slowdown in real growth. There has been debate in recent years about how much of the strong productivity growth has been cyclical. The typical cyclical pattern is that productivity growth falls, or even becomes negative, as the economy enters a recession and then recovers as an expansion takes hold. Productivity growth slowed in the middle of 2001, but continued at an annual rate of 1.5 percent for the nonfarm business sector and 2.5 for manufacturing in 2001:Q3. These are healthy rates of productivity growth for a period of recession.

Fifth, in contrast to the typical cyclical pattern, but in common with the 1990-91 recession, market rates of interest reached a peak in advance of the cycle peak. Historically, market rates of interest turn within a month of two of NBER reference peaks. Ahead of the most recent cycle peak in March of last year, the three-month Treasury bill rate reached its peak in early November 2000 and the ten-year government bond rate actually reached its peak in January 2000, or 14 months before the cycle peak. Using monthly average data, by March 2001 both these rates were down by about 175 basis points from their peaks.

Sixth, the FOMC acted aggressively in advance of the cycle peak to reduce the intended federal funds rate. At the December 2000 FOMC meeting, at which time two months of data indicating negative growth of industrial production were available, the FOMC altered its "balance of risks" statement from one with equal weight on the risk of inflation and the risk of slower growth to one indicating that the risk was weighted toward slower economic growth. At that time, the most recently released data on the unemployment rate (for November 2000) had increased only 0.1 percent from the cycle low level of 3.9 percent. On January 3, 2001 the FOMC lowered the intended federal funds rate by 50 basis points to 6.00 percent. By the August 2001 FOMC meeting, 250 basis points of additional reductions in the intended federal funds rate were implemented. Over the course of these eight months the intended rate was reduced from 6.50 percent to 3.50 percent.

Let me now link the decline in market interest rates well in advance of the cycle peak to Fed policy actions last year. I believe that the market understands quite well the Fed's mode of setting a target for the federal funds rate. That is, the market understands the policy within which individual policy actions fit in a consistent and coherent way. Given this understanding, the market brought down interest rates in advance of Fed policy actions, because the market sensed the economy's slowing and was confident that the Fed would take appropriate steps. Of course, no one-Fed or market-accurately anticipated the economy's evolution over the course of the year.

The effect of Fed policy actions was to supply the economy with significant amounts of liquidity. From December 2000 through August 2001 the M2 measure of the money stock grew at an annualized rate of 9.5 percent. During the same period, a narrow measure of the money stock known as MZM, which incorporates all cash and assets that can be converted to cash quickly at no cost, grew at an annualized rate of 17.8 percent. Thus, monetary policy was already quite expansionary when prospects for the economy changed with the attacks of September 11.

As news of the attacks arrived, the attention of the Federal Reserve became totally focused on sustaining the smooth functioning of the payments mechanism and preventing a liquidity crisis. The "lender-of-last-resort" function rose to the forefront. Lessons learned from previous financial crises including the Penn Central default, the stock market crash of 1987, and the Asian crisis and Russian default in 1998 provided valuable insights. On the morning of September 11, even before the extent of the terrorist attacks was fully certain, Vice Chairman Ferguson announced that "the Federal Reserve System is open and operating. The discount window is available to meet liquidity needs." Over the next several days, about $100 billion of short-term liquidity was injected through discount window lending, open market repos, float and "swap" agreements with four major foreign central banks.

These actions were sufficient for the financial system to weather the first days of the crisis and continue to function smoothly. Within a week most of the short-term liquidity injections had matured and trading in the securities markets had resumed.

Immediately before trading resumed in the stock market on Monday, September 17, the FOMC met by conference call and implemented an additional 50 basis point reduction in the intended federal funds rate. At the regularly scheduled FOMC meetings in October, November and December, additional reductions in the intended funds rate were implemented, bringing the total policy actions since September 11 to 175 basis points. Ample liquidity has been provided to the economy as measured by the high growth rates of monetary aggregates from August through the end of last year.

The Longer-run Economic Outlook

Where do we go from here? I'm not going to offer a specific forecast, but instead will emphasize my conviction that the U.S. economy contains very powerful forces promoting growth and full employment. Our strengths include a resilient people, efficient markets and low inflation. The Federal Reserve has made clear for many years its commitment to maintaining low inflation.

Our culture and institutions reward entrepreneurial activity. They are intact, completely undiminished by the tragedy of September 11. People are motivated by the intellectual and financial rewards of building companies and serving markets. They will be looking forward for opportunities to move the U.S. economy forward. Government policies and the structure of our labor and capital markets enable entrepreneurs to be successful. Those conditions are in place, undepreciated. For these reasons, all of us have every reason to be optimistic about the course of the U.S. economy in the years ahead.

Recent Economic "News"

In recent weeks, as usual, the short-run outlook is pushed first one way and then the other way by the arrival of economic news. I will take just a few minutes to review some of this news, but want to emphasize that, just as economists are not particularly good at forecasting cyclical peaks, they are also not particularly good at forecasting the cyclical troughs that mark the end of recessions.

Consensus forecasts at the present time are quite optimistic. The Blue Chip consensus forecast released yesterday projects real growth to be slightly positive in 2002:Q1, at an annual rate of 0.7 percent, and then for the economy to reach a growth path of about 3.5 percent in the second half of 2002 and continuing in 2003.

Data that have become available in the last several weeks give mixed signals. On the positive side, industrial production for November fell 0.3 percent, a considerably smaller decline than had been experienced in September and October and smaller than had been anticipated just the day before the data were released. Both the Michigan Consumer Sentiment index and the Conference Board index of consumer confidence increased in December, the latter quite sharply. Personally, I don't place much weight on these measures, because I believe that at best their contribution as an indicator of future consumption demand is marginal. Recall that late in 2000, the Michigan Consumer Sentiment Index dropped sharply and was the source of considerable commentary that consumption demand would weaken substantially. The weakness in consumption demand never was realized.

The December reading on the Manufacturing Report published by the Institute for Supply Management (formerly, the Purchasing Managers Index) also came in much stronger than expected, and close to the level consistent with the end of contraction in output of the manufacturing sector. Hours worked per week during December in the manufacturing sector jumped to 40.7 from 40.3 in November.

In recent weeks, prices of computer chips have stabilized and even risen somewhat. At that same time orders for chips have increased. This may signal that the sharp contraction in the computer segment of the "high-tech" sector is coming to an end. There are even some signs of stabilization in the beleaguered telecom industry. Earlier this week Corning announced that it will resume production at four plants that were idled for three months because of excess inventories of fiber.

On the other hand, incoming information continues to signal weakness in labor markets. In December, according to the data released at the end of last week, the unemployment rate increased to 5.8 percent. Nonfarm payroll employment decreased by 124,000 workers, with employment in the manufacturing sector decreasing by 133,000 workers. These signals from labor markets are not necessarily in conflict with the other income and production data, since historically the unemployment rate has been a lagging indicator that continues to rise after the economy has started to recover.

I could go on in this vein for the rest of the morning-indeed, for long after all of you had walked away. In summary, it is too early to pick a precise date for the recession trough, but there is a bottoming out feel to the data.

Forecasts and Fed Policy

I'm going to finish with a few comments on the relevance of the economic forecast for Fed policy. Consider this analogy: you are travelling to Seattle the day after tomorrow for meetings on Monday and Tuesday. You have gone to several web sites and discovered that weather forecasters are predicting rain for the two days of about half an inch. You ask me for my expert opinion and I tell you that I expect 0.1 inch of rain-just a little drizzle for a short time. Do you leave your umbrella home? I don't think so.

The fact is that weather forecasts and economic forecasts have a considerable range of error. If it makes no sense to base my behavior in carrying an umbrella to Seattle on the details of the weather forecast, why should I spend much time trying to guess whether the rainfall will be 0.1 or 0.5 inches?

The right way to look at monetary policy, in my view, is that the primary responsibility of the central bank is to maintain the purchasing power of the currency-to be successful in a policy of maintaining low and stable inflation. While achieving that primary responsibility, the Fed has a great deal of room to make policy adjustments to help stabilize employment and output. As short-run conditions change, often in unpredictable ways, we do our best to adjust the target federal funds rate in the direction conducive to maintaining economic equilibrium. All we need is a general sense of where the economy is going, and a willingness to act decisively when something happens that calls for such a response. We did not predict the terrorist attacks, but we did act decisively when confronted by them. The markets understand that we will act decisively when required, and that understanding yields better market results.

Let me reiterate: what is important about the strategy of monetary policy is not that the Fed has a superb crystal ball, which it doesn't, but that its long-run goals are clear and that it is ready to act when required. The Fed is also ready to do nothing, when required. My detailed parsing of the data-and I've done a little of that this morning-is not for the purpose of coming up with a better forecast but rather to make sure that there is not something important going on that forecasters in general are missing.

Given all the data I've studied, I don't think there is any mystery to the current situation. The patient is recovering from recession in the normal way. We have a pretty good idea what is going on; our diagnostic tests are coming up negative. You can walk out of the economic doctor's office this morning still knowing that you don't feel just right, but that nothing serious is wrong with you looking out to the years ahead. I know that I am reassured when I walk out of a physician's office with that message; I hope you are reassured this morning with regard to the economy.

Back to top