The Scope and Responsibilities of Monetary Policy> > >
by Charles I. Plosser, President and Chief Executive Officer
“… we are in danger of assigning to monetary policy a larger role than it can perform, in danger of asking it to accomplish tasks that it cannot achieve, and, as a result, in danger of preventing it from making the contribution that it is capable of making.”
These words are taken from the presidential address of the distinguished economist and Nobel Laureate Milton Friedman to the American Economic Association in 1967.1 Although the message was delivered over 40 years ago, I believe Friedman's caution is one well worth remembering, especially in this world where central banks have taken extraordinary actions in response to a financial crisis and severe recession. I believe policymakers and the public need to step back from our focus on short-term fluctuations in economic conditions and to think more broadly about what monetary policy can and should do and in the process adjust our expectations of what we believe to be the scope and responsibilities of our central bank.2
First, it may help to put Friedman's words into context. His remarks were directed at an economics profession that had gravitated toward believing that there was a stable and exploitable trade-off between inflation and unemployment, otherwise known as the Phillips curve. According to this view, policymakers should pick a point on the Phillips curve that balances the nation's desire for low unemployment and low inflation. Friedman argued that this was a false trade-off, and the experience in the U.S. in the decade that followed his remarks, often referred to as the Great Inflation, was a painful demonstration of Friedman's valuable insight. In particular, that episode illustrated quite dramatically and painfully that there was no stable relationship between inflation and unemployment. We witnessed the dangers inherent in monetary policies that take low inflation for granted in a world of high unemployment or perceived large output gaps.3 Our experiences clearly showed that efforts to manage or stabilize the real economy in the short term were beyond the scope of monetary policy, and if policymakers made aggressive attempts to do so, it would undermine the one contribution monetary policy could and should make to economic stability — price stability.
Of course, monetary theory has advanced over the past four decades as economists have developed more sophisticated models and better empirical methods to test the validity of these models. However, the proper scope of monetary policy remains an important issue of our day. In response to the global financial crisis, central banks have been asked to use monetary policy and other central bank functions to deal with an increasing array of economic challenges. These challenges include high unemployment, asset booms and busts, and the allocation and availability of credit.
I believe we have come to expect too much from monetary policy. Indeed, broadening monetary policy's scope can actually diminish its effectiveness. When monetary policy overreaches and fails to deliver desired, but unattainable, outcomes, its credibility is undermined. That makes it more difficult to deliver on the one goal that monetary policy is actually capable of meeting. Moreover, when the central bank is asked to implement policies more appropriately assigned to fiscal authorities, the independence of monetary policy from the political process is put at risk, which also undercuts the effectiveness of monetary policy.
In this year's annual report essay, I discuss the appropriate scope of monetary policy in dealing with real economic fluctuations, asset-price swings, and the allocation of credit. My views are informed by Friedman's caution that we should be careful not to expect too much of monetary policy. If we recognize the limits to what monetary policy can do effectively, we will be better able to understand what monetary policy should do.
The U.S. Congress has established the broad objectives for monetary policy as promoting “effectively the goals of maximum employment, stable prices and moderate long-term interest rates.” This has typically been characterized as the “dual mandate,” since if prices are stable and the economy is operating at maximum employment, long-term nominal interest rates will generally be moderate.
Most economists now understand that in the long run, monetary policy determines only the level of prices and not the unemployment rate or other real variables.4 In this sense, it is monetary policy that has ultimate responsibility for the purchasing power of a nation's fiat currency. In the long run, employment depends on factors such as demographics, productivity, tax policy, and labor laws. Nevertheless, many economists believe that monetary policy can sometimes temporarily stimulate real economic activity in the short run, albeit with considerable uncertainty as to the timing and magnitude, what economists call the “long and variable lag.” This type of activist monetary policy is actually quite difficult to do successfully for several reasons. First, any boost to the real economy from stimulative monetary policy will eventually fade away as prices rise and the purchasing power of money erodes in response to the policy. Even the temporary benefit can be mitigated, or completely negated, if inflation expectations rise in reaction to the monetary accommodation. Moreover, a variety of shocks can simultaneously buffet the economy. Shocks can occur to specific sectors, such as a sharp drop in housing prices or a sharp rise in the price of oil, or to specific regions. Some may be large and some may be small. Some may be positive and boost economic growth, while others may be detrimental to growth. If monetary policy responds to one shock in an attempt to offset its possible effects, it may aggravate the effects of another shock. Thus, monetary policy's ability to neutralize the impact of shocks is actually quite limited.
In addition, successfully implementing such an economic stabilization policy requires predicting the state of the economy more than a year in advance and anticipating the nature, timing, and likely impact of future shocks. The truth is that economists simply do not possess the knowledge to make such forecasts with the degree of precision that would be needed to offset the economic shocks. Attempts to stabilize the economy will, more likely than not, end up providing stimulus when none is needed, or vice versa. It also risks distorting price signals and thus resource allocations, adding to instability. In most cases, the effects of shocks to the economy simply have to play out over time as markets adjust to a new equilibrium. Monetary policy is likely to have little ability to hasten that adjustment. For example, monetary policy cannot retrain a workforce or help reallocate jobs to lower unemployment. It cannot help keep gasoline prices at low levels when the price of crude oil rises to high levels. And monetary policy cannot reverse the sharp decline in house prices when the economy has significantly over-invested in housing. In all of these cases, monetary policy cannot eliminate the need for households or businesses to make the necessary real adjustments when such shocks occur. Asking monetary policy to do what it cannot do with aggressive attempts at stabilization can actually increase economic instability rather than reduce it.
Let me be clear that this does not mean that monetary policy should be unresponsive to changes in broad economic conditions. Monetary policymakers should set their policy instrument — the federal funds rate in the U.S. — consistent with controlling inflation over the intermediate term. So the target federal funds rate will vary with economic conditions. But the goal in changing the funds rate target is to maintain low and stable inflation. This will foster the conditions that enable households and businesses to make the necessary adjustments to return the economy to its sustainable growth path and to long-run maximum employment. Monetary policy itself does not determine this path, nor should it attempt to do so.
For example, if an adverse productivity shock results in a substantial reduction in the outlook for economic growth, then real interest rates tend to fall. As long as inflation is at an acceptable level, the appropriate monetary policy is to reduce the federal funds rate to facilitate the adjustment to lower real interest rates. Failure to do so could result in a misallocation of resources, a steadily declining rate of inflation, and perhaps even deflation.
Conversely, when the outlook for economic growth is revised upward, real market interest rates will tend to rise. Provided that inflation is at an acceptable level, appropriate policy would be to raise the federal funds rate. Failure to do so would result in a misallocation of resources and, in this case, a rising inflation rate.
In both cases, changes in the federal funds target are responding to economic conditions in order to keep inflation low and stable and doing so in a systematic manner. Monetary policy is not trading off more inflation for less unemployment or vice versa. As I have already argued, the empirical and theoretical case for such a trade-off is tenuous at best. And the data to support the view that central banks can favorably exploit such a potential trade-off are even more dubious.
So what should monetary policy do? To strengthen the central bank's commitment to price stability, I have long advocated that the Federal Reserve adopt and clearly communicate an explicit numerical inflation objective and publicly commit to achieving that objective over some specified time period through a systematic approach to policy. It is one of the messages of economic research over the last 40 years that policy is best conducted in a rule-like manner. This systematic approach helps promote more effective communication so that the public and the markets will understand and better predict how policy will evolve as economic conditions change. This, in turn, helps reduce economic volatility and makes policy more effective in achieving its long-run goals.
Indeed, the Federal Reserve is one of the few central banks among the major industrialized countries that have not made such a public commitment to a numerical inflation objective. I believe it is time we did. Such a commitment will help the public form its expectations about monetary policy, which would enhance macroeconomic stability.
Let me now turn to the role of monetary policy in the evolution of asset prices. Some argue that monetary policy can be a source of distorted asset prices. But a systematic approach to achieving price stability would help monetary policymakers avoid exacerbating the effects of asset-price swings on the economy. I think it is fair to say that no one takes issue with the view that asset prices are important in assessing the outlook for the economy and inflation. Movements in asset prices can provide useful information about the current and future state of the economy. Even when a central bank is operating under an inflation target, asset prices are informative. Put another way, judgments about the inflationary stance of monetary policy should be informed by a wide array of market signals, including asset-price movements.5
The broad view among many monetary policymakers is that asset prices should not be a direct focus of monetary policy. While asset prices may be relevant in the normal course of monetary policymaking, the presumption is that such prices are responding efficiently and correctly to the underlying state of the economy, including the stance of monetary or fiscal policy. The bottom line of this view was that monetary policy should not seek to actively burst perceived asset bubbles. Instead, various forms of prudential regulation or supervision of financial institutions are likely better suited to addressing asset-price swings, should such intervention be called for.
However, in light of the recent housing boom, its subsequent collapse, and the financial crisis that followed, some people have begun to rethink this position concerning the scope of monetary policy and advocate an active role for monetary policy to restrain asset-price booms. They tend to believe that asset prices are not always tied to market fundamentals. They worry that when asset values rise above their fundamental value for extended periods — that is, when a so-called bubble forms — the result will be an over-investment in the over-valued asset. When the market corrects such a misalignment — as it always does — the resulting reallocation of resources may depress economic activity in that sector and possibly the overall economy. Such boom-bust cycles are, by definition, inefficient and disruptive. So, the argument goes, policy should endeavor to prevent or temper such patterns.
This argument for monetary policy to respond directly to a perceived mispricing of specific assets is controversial and in my view not persuasive. It requires that policymakers know when an asset is over-priced relative to market fundamentals, which is no easy task. For example, equity values might appear high relative to current profits, but if market participants expect profit growth to rise in the future, then high equity values may be justified.
Another challenge in addressing asset-price bubbles is that contrary to most of the models used to justify intervention, there are many assets, not just one. And these assets have different characteristics. For example, equities are very different from real estate. Misalignments or bubble-like behavior may appear in one asset class and not others and may vary even within a specific asset class. But monetary policy is a blunt instrument. How would policymakers have gone about pricking a bubble in technology stocks in 1998 and 1999 without wreaking havoc on investments in other asset classes? After all, while the NASDAQ grew at an annual rate of 81 percent in 1999, the NYSE composite index grew just 11 percent. What damage would have been done to other stocks and other asset classes had monetary policy aggressively raised rates to dampen the tech boom. During the housing boom, some parts of the U.S. housing market were experiencing rapid price appreciation while others were not. How do you use monetary policy to burst a bubble in Las Vegas real estate, where house prices were appreciating at an annual rate of 45 percent by the end of 2004, without damaging the Detroit market, where prices were increasing at an annual rate of less than 3 percent?
Ultimately, sound policymaking requires us to understand the limits of what we know. I doubt we could find enough agreement among policymakers or economists about the interpretation of asset-price movements to allow for stable, rule-based policymaking. In the absence of such a clearly stated rule, we risk uncertainty about central bank policy itself as well as its effect on the economy. That could become a source of volatility in asset markets and, ultimately, in real activity and inflation. Put more bluntly, asset prices are often volatile, and creating expectations that monetary policy will intervene directly to influence the price-setting mechanism seems more dangerous for the orderly functioning of markets than helpful even in the rare instances when a true and significant distortion may in fact exist. Moreover, the moral hazard created by the belief that the central bank would intervene if prices of a certain class of assets became “misaligned” might, in fact, cause more inefficient pricing and more instability, not less. Humility in policymaking requires that we respect the limits of our knowledge and not overreach, particularly when it involves overriding market signals with policy actions.
Finally, let me address another issue that has loomed large during the financial crisis and where great caution is required going forward: the role of monetary policy in the allocation of credit. At various times during the crisis, the Federal Reserve and many other central banks around the world intervened in various markets to facilitate intermediation. In many cases, these efforts were targeted to specific sectors of the economy — e.g., the Fed's purchases of mortgage-backed securities issued by the federal housing agencies — to specific types of firms or financial markets — e.g., the primary dealer credit facility and the term asset-backed securities loan facility, or in some cases, to specific firms — e.g., the lending to Bear Stearns and to AIG.
Many of these efforts, especially earlier in the crisis, were justified on the grounds that central banks should act as “lender of last resort” in order to preserve financial stability. The specific criteria for undertaking these actions could not help but be somewhat arbitrary as policymakers had little experience with such a crisis, and little theory to guide them beyond Walter Bagehot's dictum from the 1873 classic Lombard Street to limit systemic risk by “lending freely at a penalty rate against good collateral.”6 In general, these actions, especially in the U.S., involved extensive use of the central bank's balance sheet and likely went far beyond what Bagehot would have imagined.
Even when it is appropriate for a central bank to function as a lender of last resort, in my view, such policy should follow a rule-like or systematic approach. This suggests announcing in advance the criteria that will be used to lend and who will be eligible to participate. Economic and financial stability would be best served by establishing such guidelines in advance and committing to following them in a crisis. That commitment is hard to deliver on, but institutional constraints can help tie the hands of policymakers in ways that limit their discretion. Most central banks, including the Fed, have not developed such systematic plans and thus, during the crisis, behaved in a highly discretionary manner that generated moral hazard and volatility.
My purpose here is not to critique the myriad programs that were put in place or the varying degrees of moral hazard they created but to make a more general point: that these actions, for the most part, are better thought of as forms of fiscal policy, not monetary policy, because they involved allocating credit to specific firms or industries and putting taxpayer dollars at risk. Moreover, asking monetary policy to do something that it should not do — engage in fiscal policy — can be detrimental to the economy by undermining monetary policy's effectiveness at fulfilling its ultimate responsibility: price stability.
A body of empirical research indicates that when central banks have a degree of independence in conducting monetary policy, more desirable economic outcomes usually result. But such independence can be threatened when a central bank ventures into conducting fiscal policy, which, in the U.S., rightly belongs with Congress and the executive branch of government. Having crossed the Rubicon into fiscal policy and engaged in actions to use its balance sheet to support specific markets and firms, the Fed, I believe, is likely to come under pressure in the future to use its powers as a substitute for other fiscal decisions. This is a dangerous precedent, and we should seek means to prevent such future actions.7
I have long argued for a clear, bright line to restore the boundaries between monetary and fiscal policy, leaving the latter to Congress and not the central bank. For example, I have advocated the elimination of Section 13(3) of the Federal Reserve Act, which allowed the Fed to lend directly to “corporations, partnerships and individuals” under “unusual and exigent circumstances.” The Dodd-Frank Wall Street Reform and Consumer Protection Act sets limits on the Fed's use of Section 13(3), allowing the Board, in consultation with the Treasury, to provide liquidity to the financial system, but not to aid a failing financial firm or company.8 But I think more is needed. I have suggested that the System Open Market Account (SOMA) portfolio, which is used to implement monetary policy in the U.S., be restricted to short-term U.S. government securities. Before the financial crisis, U.S. Treasury securities constituted about 90 percent of the Fed's balance-sheet assets. Given that as of the end of 2010, the Fed holds almost $1 trillion in agency mortgage-backed securities (MBS) and agency debt securities intended to support the housing sector, that number is 41 percent. The sheer magnitude of the mortgage-related securities demonstrates the degree to which monetary policy has engaged in supporting a particular sector of the economy through its allocation of credit. It also points to the potential challenges the Fed faces as we remove our direct support of the housing sector. Decisions to grant subsidies to specific industries or firms must rest with Congress, not the central bank.
I have also advocated that the Fed and the Treasury reach an agreement whereby the Treasury takes the non-discount-window loans and other non-Treasury assets from the Fed's balance sheet in exchange for Treasury securities. I have further advocated that if, in the future, the fiscal authority wanted the central bank to engage in lending outside its normal operations and, importantly, should the Fed determine “unusual and exigent circumstances” warranted such action, then any accumulation of nontraditional assets by the Fed would be exchanged for government securities. Such an accord would offer two major benefits.9 First, it would transfer funding for the credit programs to the Treasury — which would issue Treasury securities to fund the programs — thus ensuring that credit policies that place taxpayer funds at risk are under the oversight of the fiscal authority. Second, it would preserve the Fed's independence to control its balance sheet and ensure that the full authority and responsibility for fiscal matters remained with the Treasury and Congress, where it rightfully belongs.
There is a historical precedent for such an accord. In 1951, the Treasury and the Fed struck an accord that freed the Fed from pegging the interest rate on long-term Treasury debt below 2.5 percent, which the Fed had done during and after World War II.10 By pegging long rates below 2.5 percent, the Fed was committing to add reserves to the banking system when market interest rates began to rise without regard to its inflation goal. This inability to control its own balance sheet was a fundamental problem for the credibility of the Fed in achieving its dual mandate. After considerable negotiations, the Treasury and the Fed reached an accord that freed the Fed to set interest rates consistent with its long-term goals. This allowed the Fed to re-establish its independence and to conduct monetary policy in accordance with its dual mandate.
Today, an accord to substitute Treasuries for non-Treasury debt on our balance sheet would similarly help ensure that the Fed will be able to implement its policy decisions. After all, the time will soon come when the Fed will need to begin exiting from the extraordinarily accommodative monetary policy in order to achieve its goals. With Treasuries back on the balance sheet, the Fed will be able to drain reserves in a timely fashion with minimal concerns about disrupting particular credit allocations or the pressures from special interests.
Like Milton Friedman in an earlier time, I too am concerned that we are in the process of assigning to monetary policy goals that it cannot hope to achieve. Monetary policy is not going to be able to speed up the adjustments in labor markets or effectively burst perceived asset bubbles, and attempts to do so may create more instability, not less. Nor should monetary policy be asked to perform credit allocation in support of particular sectors or firms. Expecting too much of monetary policy will undermine its ability to achieve the one thing that it is well-designed to do: ensuring long-term price stability. It is by achieving this goal that monetary policy is best able to support full employment and sustainable growth over the longer term, which benefits all in society. Although Friedman's words were spoken more than four decades ago, policymakers would do well to heed his insights, which remain particularly relevant for our times.