Wake a central banker out of a sound sleep and ask him what he should try to achieve with monetary policy. "Stable prices" would be the quick, instinctive answer from most. But the answer merely prompts harder questions.

"Do you mean zero inflation, which is what would be required for stable prices?" "Well," goes the usual response, "it is probably best to shoot for inflation between zero and 2 percent, given measurement error and the danger of overshooting to negative inflation or deflation." Falling prices can be associated with liquidity traps, and those can cause real problems. A liquidity trap defines a situation wherein monetary policy can’t affect aggregate demand. Deflation can get out of hand. The images of contemporary Japan, not to mention the Great Depression in the United States, flicker across the central banker’s mind. But still, he will often add quickly that low inflation is the best target because low inflation is stable inflation, and stable inflation means that the economy will function better because the messages that prices convey to households and businesses, as they allocate demand and resources, are easier to read when inflation is low.

One presses a little harder. "Is there anything else a central banker should do when setting monetary policy?" Now he is fully awake and starting to perspire. "Well, we do have to maintain orderly financial markets." "What do you mean by orderly financial markets?" "I mean financial markets that do not interfere with the functioning of the ‘real’ economy where goods and services are actually produced and exchanged, as distinct from the financial sector that prices claims on the assets of the real economy."

One presses harder still. "So central bankers should avoid asset market bubbles on the one hand and financial panic on the other. Of course, panics often follow bubbles, so avoiding bubbles is a good idea. But that is difficult, since a bubble is usually defined only by looking back after a financial panic has occurred."

The central banker is wide awake and looking especially uncomfortable. It’s time for the kill. "What if stock prices or other asset prices keep going up without any inflation? If your main goal is to keep inflation low and stable, then low and stable inflation shouldn’t trigger any response from the central bank. But if financial market bubbles, or suspected bubbles, can be followed by financial panics, shouldn’t the central bank do something to deflate a bubble, even if there is no inflation?"

The central banker is anxiously searching for a way to end the conversation. The disconcerting fact is that central banks weren’t founded primarily to fight inflation. Rather, they were established to maintain order in the financial sector and to prevent financial panics or booms from harming the real sector of the economy where goods and services are produced. They have just adopted the objective of low inflation as a good way to achieve order in financial markets and, thereby, they hope, to avoid bubbles and/or panics. But things have not always worked out as they hoped.

Central Banks in the Twentieth Century

In the United States, the financial panic of 1907 was barely staunched by the extraordinary efforts of J. P. Morgan. Fear that the country could not always count on such a fortuitous and timely presence launched the movement toward the formation of the Federal Reserve System, which came into being in 1913.

The financial sector, where stocks and bonds are traded and priced, exists to smooth out spending and earnings streams in a way that allows the swapping of purchasing power over time. A producer may wish to undertake a profitable expansion of capacity by making an outlay for capital equipment that cannot be funded out of current cash flow or assets. So he issues a claim to a lender whose current assets or cash flow exceed his current needs. The borrower must pay for the accommodation by the lender, usually with an interest rate on a bond or through some sharing of the prospective returns on capital by issuing to the lender an ownership claim on those returns.

The financial sector is both crucial and fickle. But central banks can’t actually say that their main purpose is to prevent booms or panics in the financial sector that might interfere with the performance of the "real" economy--in part because financial bubbles and panics are quite unusual events. It would be awkward for central banks to spend long periods of time awaiting a boom or a panic--with which they might not be able to deal anyway. The record is not good. Look at the United States in 1929, the United Kingdom in 1989, and Japan in 1989-90.

So central banks adopted the idea that securing low inflation is a good proximate objective; properly undertaken, it should prevent booms or panics. Much of the postwar period has kept central banks busy just achieving the goal of low inflation. In Germany, the Bundesbank was created as a bulwark against a reemergence of the ruinous inflation of the 1920s. Right after World War II, the Federal Reserve maintained orderly functioning of the market for U.S. government bonds, the stock of which had surged to over 100 percent of GDP during the war. When the Korean War threatened to overheat the economy during the 1950s, the Fed tightened monetary policy and ended the postwar policy of always accommodating government borrowing. Low and attendant stable inflation was maintained until the mid-1960s.

But by 1967 the Vietnam War and the Great Society in combination elevated total demand too rapidly, and a thirteen-year period of accelerating inflation began, punctuated by several oil crises, rising commodity prices--remember the silver-price bubble?--and a massive boom in real estate. That resurgence of inflation, largely a result of inattention to the basic principles of monetary policy, gave the Federal Reserve and its two great chairmen--Paul Volcker (1979-1987) and Alan Greenspan (1987-the present)--a solid proximate challenge: reduce inflation.

Other major central banks, particularly those in Germany, Japan, and the United Kingdom, fought similar battles with inflation after the mid-1960s, sometimes exceeding, sometimes falling short of the Fed’s performance. But they focused broadly upon the easily defined goal of bringing down inflation--and, broadly, they succeeded. The Bundesbank remained the most consistently successful opponent of inflation, to a degree that inflation faded as a major concern in Germany. By January 1999, the German people--or perhaps one should say the German government--was prepared to let the Bundesbank cease to control German monetary policy and be replaced by a European Central Bank, proximately modeled after the Bundesbank.

The Bank of Japan tackled the first major postwar equity market and real estate bubble in 1989, when there was no inflation in Japan. The resulting collapse of equity and real estate values has erased about $18 trillion of Japanese wealth, or about four years of national income. Japan’s problem in the aftermath of this catastrophe (in fairness, not attributable entirely to monetary policy) is now a tendency toward deflation--and the danger that monetary policy, thanks to a liquidity trap, cannot arrest it. Japan now confronts the reality that central banks are better at eradicating inflation than they are at eradicating deflation, though the latter is arguably more dangerous to the real and financial sectors of the economy.

After Low Inflation, What?

With inflation low and stable (or gone entirely), the major central banks, like parents who have worked hard to raise their now-grown children, are left with the tricky combination of a diverse set of new problems and no clear sense of how to address them. The Fed has little inflation to contend with, yet stock prices are 40 percent above where they normally would be, given current interest rates and earnings prospects. The Bank of Japan has pushed short-term interest rates virtually to zero, yet the only demand growth is coming from massive government spending programs that have lifted Japan’s government debt to well over 100 percent of GDP. The new European Central Bank has encountered no inflation; prices are rising about 1 percent annually in most of Europe. The ECB cut short-term interest rates by half a percent, to 2.5 percent, in April, while grousing, perhaps wistful ly, about the need to be vigilant on inflation. The German economy, about a quarter of the new ECB’s enlarged European responsibility, has yet to resume growth while unemployment remains high and budget deficits threaten to rise.

It is worth asking how central banks have achieved, or overachieved, their goal of largely eliminating inflation. Most central banks operate by targeting an interest rate, usually a short-term interest rate that the central bank can manage by raising and lowering the rate at which it lends overnight to commercial banks in the system. If the inflation rate (or the expected inflation rate) is above the target level, the central bank raises interest rates, hoping to slow the growth of demand and, thereby, to slow inflation. If, simultaneously, the economy is growing faster than the central bank’s estimate of "capacity growth," the interest rate may be raised somewhat further, although the primary response of the central bank is to remove deviations of the inflation rate above or below its target level. That is the so-called Taylor rule, named after the Stanford University economist John Taylor.

Meanwhile, not much thought has been given, until recently, to the problem of inflation that is too low--deflation. That may be so because the process of bringing the inflation rate down from a high level, say 10 percent or more, to 1 or 2 percent can take a long time. When inflation had reached double-digit levels in the United States by the end of the 1970s, the Federal Reserve raised interest rates sharply and simultaneously restricted the flow of liquidity available to the economy. By the end of 1982 the Fed had brought inflation down from a frightening 13.3 percent at the end of 1979 to a tame 3.8 percent, an incredible accomplishment. There was no stock market crash because the stock market had already been depressed by the disruptions associated with rising and more volatile inflation. The real estate market stopped rising, and fell in some areas, although the associated balance sheet problems were left to be addressed a decade later. After 1982, CPI inflation in the United States drifted irregularly around 4 percent throughout the 1980s; it jumped briefly to 6 percent with the disruptions from the Gulf War in 1990 and then fell gradually from 3.1 percent at the end of 1991 to 1.6 percent at the end of 1998. In 1999, inflation has risen to 2.3 percent (in August).

In general, central banks have been quite successful at reining in double-digit inflation insofar as they have been able to get it back into single digits, and then low single digits, without bursting financial bubbles or causing financial panics. The problems arise after the banks have achieved the low-inflation target, and especially as financial markets celebrate the success of eliminating inflation.

When Central Bankers Meet

At the annual meeting of the central banks at Jackson Hole, Wyoming, in late August, the matters of principal concern were, first, the experience of the Federal Reserve and other central banks in meeting their inflation targets, and, second, determining what to do about financial market bubbles, panics, and deflation. The conference was aptly titled "New Challenges for Monetary Policy," and its nominal focus was on how to maintain the low inflation and stable prices that are now being enjoyed by the U.S. and European economies. The discussion summarized a good deal of what we currently know about how central banks operate (or should operate) to achieve the proximate target of price stability; there was discussion as well of how to deal with deflation and a liquidity trap in Japan.

Despite the great proximate success in bringing down inflation rates, the persistent subtext at the conference was how responsive the Federal Reserve should be to the behavior of the U.S. stock market. The formal papers argued that the Fed should concentrate on targeting expected inflation, not on responding directly to movements in the stock market. But two nagging questions remained: Is the U.S. stock market a bubble? And what should the Fed do if the bubble bursts and stock prices drop sharply?

Bank of Japan deputy governor Yutaka Yamaguchi gave an eerie account of conditions in Japan during the late 1980s, just before the bursting of the Japanese equity and real estate bubbles that began in 1990. The conditions he described, such as zero inflation, rapid investment growth, and talk of a new era of unlimited prosperity, reminded the audience of current conditions in the U.S. economy.

One of the papers at the conference suggested through modeling that, by 1989 to 1990, the Bank of Japan ought to have raised interest rates to 8 percent--well above the 4 to 6 percent rates then prevailing. The increase was warranted by the roaring expansion that was underway and the rapid rise in stock prices that had swelled the balance sheets of households and businesses and led to a borrowing boom. But Mr. Yamaguchi reminded a somewhat discomforted audience that it would have been very difficult for the Bank of Japan to set short-term rates at 8 percent, given the zero inflation at the time and the primary stated goal of achieving low inflation.

Mr. Greenspan Speaks

In a discussion of the current possible "bubble" situation facing the Fed, one prominent economist at the conference suggested that, while targeting inflation was a good general policy rule for the Fed and other central banks to follow, the rule did not address the question of what the Fed should do if the stock market begins to drop rapidly with no inflation in sight. Federal Reserve chairman Greenspan responded to this comment with a remarkably candid statement about Fed policy and asset markets. The chairman maintained that the Fed would react to offset rapidly falling or rapidly rising asset market prices, but that if changes in asset prices were gradual, the Fed would not react and would continue to focus on changes in expected inflation when setting interest rates and monetary policy. In his statement, Chairman Greenspan was perhaps rationalizing the Fed’s passive stance on the U.S. stock market since his "irrational exuberance" comments in December 1996. But since then, the U.S. stock market, measured by the S&P Index, has risen at an average annual rate of 25 percent--well above the average 14 percent annual growth rate since the start of the great U.S. bull market in August 1982.

Prior to Mr. Greenspan’s "irrational exuberance" comment, the U.S. stock market had been rising at a 15 percent annual rate since the start of the current expansion early in 1991, quite close to the post-1982 average of 14 percent, though still robust. The acceleration thereafter to a 25 percent annual growth rate apparently has not struck the chairman as too rapid.

Clearly what Chairman Greenspan has in mind when he speaks of the need for the central bank to react to a rapid drop in equity prices is the impact of a sharp drop in the stock market--over, say, several days--that seriously damages the balance sheets of households and businesses and thereby threatens to cause liquidity to dry up as everyone rushes for the safety of short-term Treasury securities. (The events of October 1987 come to mind.) The chairman is harkening back to the fundamental reasons behind the founding of the Federal Reserve System in 1913: disorderly financial markets are to be avoided, especially if their condition threatens the performance of the "real" economy.

But even the chairman’s unusually explicit discussion of the Fed’s possible reaction to changes in the U.S. equity market left conference participants uneasy. If the central bank is prepared to allow asset prices to rise steadily at a brisk 25 percent pace, yet is still prepared to support the stock market if it falls (more) rapidly, it is hardly surprising that a "buy on the dips" mentality has emerged over the past several years and has resulted, in turn, in higher and higher equity prices.

The Fed’s Looming Choices

The Fed, like the Bank of Japan a decade ago, may be forced by circumstances to increase interest rates steadily--and, in so doing, to test severely the resiliency of the stock market. Part of the reason for the rapid growth without inflation in the United States over the past eighteen months has been an unusual desynchronization of the global business cycle. The Asian crisis, the Russian collapse and the associated Long-Term Capital Management crisis, and the emergence of deflation in Japan after the ill-timed tax increase in spring 1997 have all produced a remarkably benign atmosphere within which the U.S. investment-led recovery has been able to flourish without placing strains on real resources or generating inflation. Further, the run-up to European Monetary Union produced unusually tight monetary and fiscal policies in Europe during 1997 and 1998, which increased the external accommodation offered to the noninflationary U.S. expansion.

Now, with the emergence of the European Monetary Union--and lower interest rates in Europe accompanied by somewhat higher growth--the reemergence of expansion in most of Asia, perhaps even in Japan (though this remains in question), and the unwinding of the Asian debt crisis, U.S. inflation may not rise spectacularly but simply by enough to put heavy pressure on current valuation in U.S. equity markets.

Based on pricing in the interest rate futures market, U.S. financial markets expect that, at most, the Fed will raise short-term interest rates by another 25 basis points before the end of the year, to bring the total rate increase this year to 75 basis points. That increase would mean a complete reversal of the rate reductions seen last year in response to the Russian/Long-Term Capital Management crisis. Indeed, based on the policy rules that relate interest rate changes to Fed inflation targets, the Fed is about on track with its current rate increases--provided that it does raise rates another 25 basis points during the next several months. U.S. inflation has risen by about half a percent, from 1.5 percent to over 2 percent, during the first half of 1999. Based on a policy rule estimated for the U.S. economy, the inflation increase would require an interest rate increase by the Fed of about 80 basis points, or roughly three 25 basis point increases in succession, which markets now expect.

However, the conditions whereby U.S. inflation would rise another half percent, from 2 percent to 2.5 or 2.75 percent, are not very stringent (we are already at 2.3 percent), and those conditions would require still another 75 to 100 basis points of rate increases, again based on policy rules estimated for the Fed. For example, if CPI inflation rises at two-tenths of a percent per month between now and December, year-over-year CPI inflation will have reached the 2.50 percent to 2.75 percent level that calls for another 100 basis point increase in the Federal funds rate. U.S. markets are currently not contemplating the possibility of such a rate increase, and the stresses that it would place on the U.S. stock market might give the Fed pause about proceeding.

The Fed faces two other technical problems over the next year. The first, and less significant, is the Y2K problem, which threatens possibly to disrupt liquidity conditions over the end of the year. Specifically, lenders may be unwilling to commit to supply funds over that period because they fear possible glitches with the computers that keep track of fund balances between banks and other financial institutions. However, the Federal Reserve System has indicated clearly that it will be a residual supplier of credit over the end of the year, and that action, along with steps by other central banks, should be sufficient to deal with the Y2K financial sector problem.

The other technical difficulty facing the Fed, of course, is the 2000 election year. Some observers suggest that the Fed is reluctant to undertake aggressive policy measures, especially those that might risk creating disorder in financial markets, during an election year. But here the Fed faces a dilemma. It achieved a low and stable inflation rate (its stated objective) by predictably following policy rules; to abandon those rules would be highly counterproductive. The Fed must tighten if inflation does actually go from 2 to 2.5 percent or higher while growth remains strong and/or the dollar weakens. The dollar may weaken because of the rising U.S. current account deficit, which is more difficult to finance in a growing world economy. If the Fed is passive in the face of rising inflation, the dollar could weaken still further, with consequent positive feedback on expected inflation in the United States. Then the Fed would be even more behind the curve in attempting to control inflation.

In short, having achieved its goals by following a policy rule, the Fed can’t very well stop following that rule because of Y2K or an election year. The disorder in markets that would result from the Fed’s failure to follow its stated rule could be just as great as, or greater than, the disorder from raising interest rates according to the inflation rule. That is the Fed’s dilemma, and it arises directly from the potential conflict between the Fed’s proximate and explicitly stated target of keeping inflation low and stable and its implicit target of avoiding disorder in financial markets.

A Scenario for the Banks

The Fed probably faces a "pay-me-now-or-pay-me-later" choice with respect to monetary policy. The economic recoveries in other major economies and in most emerging markets in Asia needn’t be very dramatic to undo the extreme accommodation that contributed mightily toward U.S. noninflationary growth over the past two years. Further, U.S. demand growth isn’t going to slow until U.S. asset values either stabilize or fall somewhat. During the year between the first quarters of 1998 and 1999, the net worth of U.S. households rose by $1.66 trillion, thanks largely to higher stock prices and higher real estate prices. The rule of thumb that spending rises by about 4 percent of the increase in net wealth means that the windfall should add about $67 billion to consumption during the current year. That would add as well about three-quarters of a percentage point to the growth of aggregate demand, and probably be associated with a growth rate between 3.5 and 4 percent of the overall economy. That growth rate lies somewhere between 0.5 and 1 percentage point above potential U.S. growth, depending on whose estimates you believe, and it would require, in turn, between 100 and 150 basis points of tightening to bring it back down to the long-run potential growth rate of the U.S. economy, which is somewhere between 2.5 and 3 percent.

It won’t be easy for the Fed to stick to the proximate rules without breaking the underlying rule of not disturbing financial markets too much. The uncomfortable fact is that stock holdings have become a large part of household balance sheets, eclipsing even the value of the real estate holdings that were formerly the dominant asset for most Americans. Simultaneously, Fed chairman Greenspan has become a familiar figure to virtually every American and is associated with America’s great prosperity over the past decade. His high profile may be something of a burden if his actions precipitate, or are even thought to precipitate, a sharp drop in the stock market. Mr. Greenspan has made it clear that the Fed has no mandate to kill the stock market unless inflation rises. And even then he doesn’t want the stock market to fall too rapidly and leave blood on his hands. As the Bank of Japan’s experience a decade ago reminds us, that is a tall order.

What, then, should central banks do, now that inflation is low but the United States may be facing a stock market bubble and Japan may be facing more deflation? Probably, the Fed should keep inflation low by responding to the recent modest uptick in inflation with appropriate modest tightening. The Bank of Japan, simultaneously, needs more than ever to be easing aggressively by printing more money. That might stem the recent deflationary appreciation of the yen, while helping to offset the negative global liquidity effects of a modest Fed tightening.

In short, the Fed should tighten a little, and the Bank of Japan should ease a lot. And both should pray.

John H. Makin is a resident scholar at the American Enterprise Institute.