I am delighted to be invited to address the Annual Financial Services Forum of the New York State Bankers Association. Your program today provides a superb opportunity to discuss some of the important public policy issues currently confronting the banking industry, and I am pleased to be able to participate once again. In my remarks this morning, I would like to share with you my thoughts on some of the achievements of monetary policy to date and where I believe we may want to consider heading as we move toward the twenty-first century.

At the outset, let me note that I am quite encouraged by the recent performance of the U.S. economy. Monetary restraint beginning in early 1994 brought about a much needed slowing in aggregate demand, which had started to run well ahead of the economy's potential. One of the most heartening features of the current moderate expansion is the continuing good news on inflation. In fact, inflation has been at or below 3 percent for the last five years, the lowest level in a generation. With modest increases in labor costs, I do not see conditions in place for an acceleration of inflation in the near future.

From a personal perspective, I am convinced that much of our success in containing inflation in recent years reflects monetary policy actions that pre-empted inflationary pressures before they actually showed up in general prices. As you recall, the Federal Reserve adopted a pre-emptive policy approach starting in February 1994, when it began firming monetary conditions because of the potential it saw for inflation re-emerging. The main reason we need a pre-emptive approach, in my view, is because monetary policy works with uncertain and long time lags. Although most of its effects on output take place within one to two years, its effects on inflation take even longer--over a three-year time frame, which is the appropriate horizon for monetary policymakers.

When one stands back and considers monetary policy over the past several decades, the case is strengthened for a pre-emptive approach to squeeze off incipient inflation before it shows through in broader price increases. Economic analysis has shown not only that an overheating economy has a strong effect in raising inflation but also that reducing inflation is a very painful process. We learned these lessons during the long and costly disinflation of the early 1980s, following the explosion of inflation in the 1970s. Thus, both analysis and experience reinforce the need for pre-emptive monetary policy actions. Failure to contain inflationary pressures at an early stage makes it much costlier to deal with inflation later.

I believe that there is broad support within the United States today for a rigorous and consistent anti-inflation policy. Most elected officials, economists, and the general public have become much more aware of the substantial economic costs of even moderate rates of inflation. These costs, as we all know, stem from a variety of sources, including:

• the increased uncertainty about the outcome of business decisions and profitability;
• the negative effects on the cost of capital resulting from the interaction of inflation with the tax system;
• the distortions that create perverse incentives to engage in nonproductive activities to cope with inflation; and, more generally,
• the reduced effectiveness of the price and market systems that inflation imposes on society.

Inflation also imposes serious social costs because it tends to fall particularly hard on the less fortunate in our society, often the last to get employment and the first to lose it. These people do not possess the economic clout to keep their income streams steady, or even buy necessities, when a bout of inflation leads to a boom-bust scenario for the economy. As a double blow, they suffer disproportionately when the bust comes.

Against the background of generally low inflation and broad public support for policies that rigorously contain inflation, the recurring congressional debate on the appropriate objectives of monetary policy has recently been revived. One initiative is a bill Senator Mack has introduced to make price stability the primary long-term goal of monetary policy. Congress is likely to hold hearings on this bill in coming months and it is certainly worth careful consideration.

There is no question in my mind that the primary long-term goal of monetary policy should be price stability. Stable prices have always been the prism I have used for making monetary policy decisions. As I look back, I am pleased that monetary policy has been successful these past several years in controlling prices while encouraging both growth of the real economy and financial system stability. At the same time, I am well aware that the gains of some earlier monetary policy successes were nonetheless allowed to dissipate. For example, in the late 1950s and 1960s, inflation levels were below 3 percent for ten years, similar to the situation we have today. Unfortunately, as we all know, inflation began rising in the late 1960s, reaching double-digit levels by the end of the 1970s.

In my view, therefore, the challenge to monetary policy in today's environment is to consider how we may most effectively build on our current low inflation by making its permanence a credible policy goal. This goal raises a host of important questions. For one, even if we agree--as I believe we already do--that price stability must be the primary long-term goal of monetary policy, what exactly does price stability mean in practical terms over both the intermediate and long term? Second, what kind of institutional structure would enable the Federal Reserve to convey to the markets and the public an explicit commitment to price stability? A related question is how should such a policy be articulated to the public to make the central bank accountable and to foster a political consensus in support of this commitment? Finally, how can an explicit policy commitment to price stability be implemented in practice without pushing the economy too hard in one direction or another? These are a few of the questions we at the Federal Reserve Bank of New York are asking as we consider the merits of our country's taking a step further in its conduct of monetary policy.

If we are to shed light on some of these questions--and if we are to set a goal for monetary policy--we must be clear as to what we can expect monetary policy to do and what we know it cannot do. What monetary policy can do is to anchor inflation at low levels over the long term and thereby lock in inflation expectations. In addition, monetary policy can help offset the effects of financial crises as well as prevent extreme downturns in the economy. One point is worth emphasizing: Allowing even a low level of inflation to persist without a commitment to bring that level downward toward price stability permits--and may even encourage--expectations for still sharper price rises in the future. Such expectations provide an opening for a demand-driven burst of inflation.

But what monetary policy cannot do, in and of itself, is produce economic growth. Economic growth stems from increases in the supply of capital and labor and from the productivity with which labor and capital are used, neither of which is directly influenced by monetary policy. However, without doubt, monetary policy can help foster economic growth by ensuring a stable price environment.

There are some people who think that establishing price stability as the primary goal of monetary policy means that the Federal Reserve would no longer be concerned about output or job growth. I would like to make clear for the record that I believe this view to be simply wrong. A stable price and financial environment will almost certainly enhance the capacity of monetary policy to fight occasions of cyclical weakness in the economy.

Some commentators argue as if Federal Reserve policy aimed at containing inflationary pressures is tantamount to pursuing targets for real economic growth limited to 2 to 2 1/2 percent. The Federal Reserve, of course, has no such targets, as was made clear by Chairman Greenspan in his recent congressional testimony. In trying to determine the extent of future inflation, we at the Federal Reserve must look at a broad array of economic indicators that reflect demand pressures and supply developments in the economy. Unfortunately, there is no single summary measure that provides a reliable overall assessment of the many complex and diverse influences on inflation--a fact that makes it more difficult to reach a national consensus on policy at any given point. In short, while its one explicit goal must be price stability, monetary policy can and must also maintain the broad environment for sustainable economic growth.

If, then, we are to take monetary policy one step further and formalize a commitment to price stability, how should price stability be defined? And, what do we mean by price stability in practice? Let me offer two possible basic definitions, each relating to a different time frame. One definition would apply over the long term. In this time frame, I would agree with Chairman Greenspan that price stability exists when inflation is not a consideration in household and business decisions.

What does this mean in practice? We know that, as currently measured, a zero inflation rate is not the same thing as price stability. This is because of well-known errors in measuring inflation that stem from many factors, including how quality improvements and new products are valued in the Consumer Price Index. Although there is much research on this topic, economists and policymakers cannot agree upon a single number for the magnitude of this measurement error. In most studies, the error has been estimated to range from 0.5 percent to 2.0 percent. Therefore, as a practical matter, price stability may best be thought of as an inflation rate falling somewhere within this range.

Were we to move to a monetary policy strategy that has a numerical inflation goal, given the problems with measurement error, how might this goal be set? If the inflation goal is set too high, we run the risk of allowing the start of an upward spiral in inflation expectations and inflation. Indeed, this is why I do not believe that price stability is consistent with the 3 percent inflation rate we currently have.

If, on the other hand, the inflation goal is set too low, we run the risk of tipping the economy into a deflation in which the true price level is actually falling. History has shown that deflation can be extremely harmful to the economy in general, and to financial markets in particular. The worst financial crises in our history have been associated with deflationary periods.

Therefore, were we to set a numerical inflation goal for monetary policy, I believe that an appropriate number for this goal should be within the reasonable range of measurement error-- but in the upper end of the range because of the dangers of deflation. Such a numerical goal could be understood as the premium needed to prevent the economy from being tipped toward deflation or needlessly forgoing output.

In the long term, a numerical definition for price stability would provide a framework for the discussion and evaluation of monetary policy. In practical terms, this would mean that the Federal Reserve would be held accountable to--and when successful, judged credible by--an explicit inflation performance standard that would ensure stable inflation expectations.

In the intermediate term, by contrast, over a period of, say, three years--the time horizon over which monetary policy affects inflation--the goal of monetary policy is to put the economy on the path that moves it toward long-term price stability, taking into account the economic and financial pressures on the economy. At low levels of inflation, there are substantial risks to the economy from driving out the remaining inflation too quickly. In the current environment, therefore, the path for monetary policy in the intermediate term would have to be gradual.

A key principle for monetary policy is that price stability is a means to an end--to promote sustainable economic growth. Therefore, the introduction of a numerical goal for price stability in no way implies that the health of the economy should be sacrificed in single-minded pursuit of this goal.

Moreover, it is important to stress that, even if we were to agree on a numerical goal for price stability and even if price expectations were as a result contained, we still would not have eliminated all sources of potential inflationary shocks. The reality is that monetary policy can never put the economy exactly where we want it to be. For example, supply shocks that drive prices up sharply and suddenly--such as the two oil shocks of the 1970s--are always possible. In such an eventuality, the appropriate monetary policy in the presence of a price stability goal would not be to tighten precipitously, but rather to bring the inflation rate down over time once the economy adjusts to the shift in relative prices. In the event of a shock to the financial system, the appropriate monetary policy might require a temporary reflation.

Such policy responses might require the numerical inflation goal to move above the long-term goal for a period of time, but then to trend downward toward the long-term goal. In practice, this means that even though the intermediate policy goal would change, the underlying strategy and the long-term goal of price stability would remain the same.

This gradual and forward-looking strategy is essentially the course that the Federal Reserve has been following over the past several years. Integral to this course have been increased efforts toward greater transparency in the conduct of monetary policy. The announcement of changes in policy at the conclusion of FOMC meetings is evidence of these efforts.

What, then, might be some of the advantages of further increasing transparency by committing the Federal Reserve to an explicit inflation goal? For one, were the Federal Reserve to formalize its strategy by announcing specific intermediate and long-term goals for price stability, it might reduce uncertainty about policy. Moreover, the Federal Reserve could clarify why specific policy moves were made at specific times, with reference to its numerical intermediate-term goal. In addition, an explicit commitment to price stability and specific numerical goals for inflation could help lock in low inflation expectations, making future inflations and disinflations less likely. Lastly, I believe that, were the Federal Reserve to move to the articulation of such a strategy, public discussion and evaluation of monetary policy would be directed to a tighter, less contentious framework than that which currently exists. This is because the performance of the Federal Reserve in fulfilling its monetary responsibilities would be the issue, while the goals would be unambiguous and well established.

The institutional framework to implement such a strategy is, of course, a question. I believe that the mandate for price stability is of sufficient importance to society that it should be set by the legislative process. Were such an approach to be formalized, the Federal Reserve could articulate its strategy as it currently does under the Humphrey-Hawkins law, or Congress might choose to replace the Humphrey-Hawkins law. The fundamental point is that once numerical inflation goals were set, it would be logical and useful to create some kind of an institutional framework for the Federal Reserve to report its progress in meeting its monetary policy goals.

I am pleased to share these thoughts with you this morning, encouraged as I am by the favorable developments in monetary policy and the credibility I believe the Federal Reserve has earned in controlling inflation. The discussion of the appropriate goals for monetary policy and what these goals might mean in practice is currently an intellectual one, although, I hasten to add, one not confined to ivory towers. We at the Federal Reserve Bank of New York are, in fact, currently engaged in a very active research program designed to address some of these questions.

Public debate about these issues has begun, and certainly there are many points of view to listen to and evaluate. My remarks today are intended to contribute to and help stimulate such discussion.

The most important asset a central bank possesses is public confidence. I believe firmly that achieving and maintaining public confidence are squarely related to the success the central bank has in the discharge of its monetary policy responsibilities. The benefit of a declared inflation goal is that it puts the focus of public oversight where it belongs--on how well the central bank is doing its job--and settles the issue of what that job should be.

In sum, I am increasingly convinced that in a democracy, a central bank can maintain price stability over the intermediate and long term only when it has public support for the necessary policies. Numerical goals for monetary policy may help a credible monetary authority maintain and build upon that support.

This is a course worthy of further consideration.

Thank you.