Economic Forecasts and Monetary Policy
Today I would like to share with you some of my thoughts about economic forecasts and monetary policy. I will begin with some comments about the economy's recent performance and the outlook for 2006. Next, I will explain why making sound policy decisions requires me to think about both the demand and supply sides of the economy. Finally, I will describe how the stories behind the forecasts directly relate to the way I think about the appropriate course for monetary policy.
Please note that the views I express today are mine alone. I do not presume to speak for any of my colleagues in the Federal Reserve System.
Economic Performance and the 2006 Outlook
I will set the context for my remarks by first taking a brief look at our recent economic performance and the outlook for 2006. Although the national economy in 2005 was not as strong as it was in 2004, overall we saw solid growth for the year. As you know, the economy endured a series of significant energy-price shocks that, in earlier eras, might have triggered a recession. Instead, third-quarter GDP growth remained strong.
Now, it is true that the advance estimate for fourth-quarter GDP growth was only 1.1 percent. Some may interpret this weak performance as a sign that the energy shocks may have finally taken their toll. However, we should remind ourselves that these are preliminary numbers. Third-quarter GDP growth was substantially revised upward, so we may learn in a few weeks that fourth-quarter growth was not quite as weak as the initial estimate indicates.
We should also remember that it is fairly common to miss on forecasts of quarter-to-quarter economic performance. For instance, it would not surprise me too much to look back and see that some of the growth we thought would occur in the final three months of last year was actually spread across the third quarter of 2005 - when GDP growth was stronger than expected - and the first quarter or two of this year.
Please understand that I am not suggesting we should be complacent about the weak statistic for fourth-quarter growth. I will be watching the incoming data very carefully over the next several months. In fact, the early data for January have been reasonably good.
Assuming the preliminary fourth-quarter report holds up, though, GDP still grew last year by 3.5 percent. For the year as a whole, it is clear that employment growth accelerated, business fixed investment was relatively strong, and core inflation remained subdued.
What about the outlook for 2006? Most forecasters are expecting another solid performance. Forecasts published by Blue Chip Economic Indicators, the Congressional Budget Office, and NABE economists generally call for real GDP to expand by roughly 3-1/4 to 3-1/2 percent this year. Housing investment is generally expected to slow, while business fixed investment is expected to increase. These forecasts call for interest rates, the unemployment rate, and core inflation to remain steady. I know that at your meeting last month, your three forecasters presented views that were largely the same.
No one, of course, is expecting that every detail of these forecasts will prove to be precisely accurate. We have to acknowledge reality: Any forecast is only as good as our ability to look into the future and foresee the unforeseeable.
In fact, the standard deviation of annual real GDP growth is about 2-1/2 percentage points. This means that most of the time we would expect real GDP growth to vary in a range that is quite large - between 1 and 6 percent. Although very good economic forecasting models can narrow that range of uncertainty, a fair amount of uncertainty still remains. Forecasting is a tough business, leading some people to question the value of forecasting altogether.
I find forecasts to be helpful. However, achieving better forecast accuracy is less important to me as a member of the Federal Open Market Committee than understanding the forces that drive the economy. As a result, although I know that forecasts tend to be inaccurate - sometimes very inaccurate - the process of thinking through those forecasts is central to the way I do my job. Within the numbers, I need to look not just at the "what" but also at the "why." In other words, I need to think hard about the demand- and supply-side assumptions that underlie economic forecasts.
Thinking about the Demand and Supply Sides of the Economy
So let me explain why I believe that making sound policy decisions requires me to think about both the demand and supply sides of the economy.
Most often, it seems that forecasts are presented from the demand side - as the sum of growth in different categories of spending. In other words, we gauge how fast consumption spending is likely to grow, add an estimate of how fast investment is likely to expand, throw in a prediction of government spending growth, include some assumptions about net exports and, voila, we have a GDP forecast. But this is really just a description of what the forecast is, not a description of why the forecast is what it is.
As a policymaker, I am less interested in a statement like "GDP will grow by 3-1/4 percent," than I am in answering the question "Why will GDP grow by 3-1/4 percent?" I find great value in the story behind a forecast, even if I suspect that unforeseen events could prove the forecast wrong. Both supply and demand conditions inform the stories behind the forecasts.
Here is an example. Just recently, a major newspaper ran an article that stated: "Largely because consumer spending slowed to a near halt in the fourth quarter last year, overall economic growth fell." On the surface, that statement seems pretty straightforward: If spending is strong, the economy grows. If spending is weak, the economy grows by less. Is that really the best way to think about U.S. economic performance in 2005?
It is certainly true that GDP growth last year was off the pace of 2004. But, as we are all painfully aware, energy prices rose dramatically over the course of the year. Consumers saw energy prices rise about 30 percent by the time they peaked in September. Clearly, energy-market disruptions - a supply condition - could go a long way toward explaining why economic activity, including consumer spending, was more restrained than in the recent past.
So, which interpretation is right? Did economic growth slow in 2005 because spending slowed? Or did spending slow because economic growth was restrained by a series of large energy-price spikes? To put it another way, did growth slow because demand in the economy was too weak, or did it slow because productive capacity in the economy was reduced by adverse supply effects?
For me, from a policymaker's perspective, these are important questions. My job is to help find the course of monetary policy that is consistent with price stability and with the economy operating at maximum sustainable growth or, in other words, growing at its potential. This means that we have to have an idea of where "potential" is, and we have to be able to identify factors that affect potential, as opposed to factors that affect only aggregate demand. In practice, there may be circumstances that affect both demand and potential, but we still must try to disentangle demand effects from supply effects. And the reason we must be able to make these distinctions is that demand and supply effects can have different implications for the appropriate course of monetary policy.
The Stories Behind Forecasts and Monetary Policy
Now I will turn to how the stories behind the economic forecasts directly relate to the way I think about the appropriate course for monetary policy. I use the term "stories" as a shorthand way of describing the various hypotheses that lurk behind the economic data we receive. Each story that I might consider rests on some hypothesis about potential GDP, full employment, and other concepts such as the neutral real rate of interest. The challenging part of using these concepts to inform policy decisions is that they are not fixed numbers, or even fixed ranges of numbers. Their values themselves are a complicated function of the supply and demand conditions that I have been discussing.
Former Chairman Greenspan has expressed skepticism about the ability of policymakers to translate these concepts into precise numerical benchmarks. But the concepts themselves must be taken into account when the FOMC assesses the appropriate path for monetary policy. When you hear policymakers - or at least this policymaker -- speak of concepts like "potential GDP" or "full employment," you are really hearing an attempt to sort out the supply and demand conditions that frame the economic environment.
Thinking through these concepts is not just an academic exercise. It is a necessary part of creating a framework for monetary policy decision-making. To fulfill its mandate, the Federal Reserve must look not just at the economy's top-line performance. We must try to determine the stories that may be hidden within the numbers. And, as I said earlier, even though it can be frustrating to know that our economic forecasts will likely be inaccurate, we know that the very process of constructing the stories behind the forecasts is invaluable.
Let me suggest some concrete examples to help clarify what I mean. From 1997 through 1999, real GDP growth averaged almost 4.4 percent, which is well above most traditional estimates of how fast the economy can grow without accelerating inflation. But several members of the FOMC - among them, former Chairman Alan Greenspan and Jerry Jordan, my predecessor as president of the Federal Reserve Bank of Cleveland - argued that growth itself is not inflationary if it is driven by productivity gains.
Although they were considering the same projections that others were studying, Chairman Greenspan, Jerry Jordan and the others had a much different picture of what it meant for inflation. They saw favorable supply-side conditions as the cause of the rapid growth, while others saw overheated demand conditions. In my view, history has proven that the supply-side perspective was correct.
Consider a more recent example. Many people expected job growth to bounce back more quickly over the past four years than it did. Job creation was well below expectations as the economy emerged from the 2001 recession. Just how far below was made clear by the dating of the recession itself. As you know, a committee sponsored by the National Bureau of Economic Research determines the precise timing of the beginning and the end of recessions. Although the committee concluded that the recession ended in November 2001, it was not until July 2003 that it was confident enough to make that judgment. The main reason for the delay was the unusually weak behavior of job growth.
In fact, it took almost four years for the economy to reach the point where there were as many people working in the United States as there were at the beginning of the recession in 2001. It was not until last year that we reached the milestone of having created 2 million new jobs in this expansion. A reasonable interpretation of this period is that demand for goods and services was not strong enough to create more robust demand for workers. That view implies that the economy had generally been operating below its potential. If an economy is operating below its potential because of weak demand, then a relatively more accommodative monetary policy is the right medicine - and it can be administered without fear of stoking inflation. Indeed, the FOMC followed this course for a considerable period of time.
It is true that even today, new jobs are still being created more slowly than the roughly 3 million jobs created each year between 1994 and 2000. How can we account for this performance? Does a demand-side story make the most sense? In this case, I think that trends on the supply side of the economy suggest that we might need to interpret sluggish labor markets differently today.
Perhaps the most interesting trend is the pattern of labor-force participation - that is, the fraction of people who either have a job or are actively seeking a job. Since 2001, the labor-force participation rate of all age groups, except those 55 and older, has declined. The change has been especially noticeable among younger workers - 16- to 24-year-olds. Those participation rates have declined by about 5 percentage points. That amounts to 1.9 million young people who, for now, are no longer potential workers.
Has this episode of slower employment growth resulted from demand conditions, supply conditions, or some combination of both?
If it is defined as a demand condition, perhaps poor job prospects have discouraged people from even attempting to find work. Will another year like 2005 reverse the recent trend? And does that mean that monetary policy should be accommodative until the economy is once again generating substantially more than 2 million new jobs per year?
Alternatively, if the lower labor-force participation rate is defined as a supply condition, then it may be driven by younger workers' deferring entry into the labor force - perhaps to obtain more schooling and skills. If that is the correct explanation, then potential employment will be calculated much differently from the number we saw in the 1990s. In that case, attempting to spur more rapid job growth with an accommodative monetary policy is exactly the wrong thing to do. It will not accomplish the goal of maximum sustainable growth in the long run, and it may threaten our goal of price stability.
As economists - whether we work in corporate life, the academic world, or the policymaking realm - we are all in the business of economic forecasting, but each of us may approach our profession from a different perspective.
The mission of the Federal Reserve leads me to look at economic data - and the very process of economic forecasting - with a different purpose, and thus from a somewhat different perspective than simply producing the most accurate GDP forecast. Our motivation for forecasting economic conditions begins with our need to shape monetary policy to the evolution of the economy as we work to fulfill our dual mandate of price stability and maximum sustainable growth. For that reason, we look to the supply side as well as the demand side of the economy when evaluating conditions.
I hope this discussion has helped clarify how I approach my responsibilities as a member - and this year a voting member - of the Federal Open Market Committee. Armed with that knowledge, the next time you hear news reports about the FOMC's decisions, I hope that you will have an even deeper understanding of the process I use: considering not just the "whats" that make it into the headlines, but thinking about the "whys" that I carefully weigh, within policy deliberations, in an effort to maintain price stability and maximum sustainable growth.