Current Developments in Monetary and Financial Law, Vol. 3

CHAPTER 20 Relative Autonomy of the Central Bank

International Monetary Fund
Published Date:
April 2005
  • ShareShare
Show Summary Details

We have grown accustomed to the autonomy of the central bank, that mantra of modern monetary policy. An idea that grew out of the traumatic monetary conditions between the two world wars and developed in the United States, it became one of the cornerstones of the German Bundesbank and was ultimately enshrined in the Maastricht Treaty, through which it came to dominate the central banks of the European Union.

In the interest of full disclosure, I must admit that I have an instinctive dislike of mantras. Invariably, the wisdom they try to capture is too complex for their simple words. I also have a healthy distrust of autonomous institutions that are a law (nomos) unto themselves (autos). This is not surprising. I grew up in the polders of the Netherlands, where strong dikes require close cooperation between peers, and where mantras and autonomies are exposed to the elements and their strengths are constantly tested by water and wind.

So let us do some testing of our own. Let us spend some time examining the strength of what we call the autonomy of the central bank. It is logical that we should do so in the conduct of monetary policy, because that is the area of primary responsibility of the modern central bank.

Price Stability as a Primary Objective of the Central Bank

The autonomy of the central bank is established by law. All of our countries have adopted the practice of assigning the conduct of monetary policy to a more or less autonomous statutory agency.

Ideally, the law endows the central bank with financial, administrative, and operational independence from the state and freedom from political interference in making and executing its decisions. The law elevates price stability to the primary objective of the central bank and makes the central bank exclusively responsible for the conduct of monetary policy required to achieve this objective. The law defines the content and scope of the powers of the central bank in the conduct of monetary policy. Thus, it lists the monetary policy instruments that the central bank may use; it prohibits or restricts advances from the central bank to the state (including its subdivisions, agencies, and instrumentalities); and it requires the central bank to report periodically to parliament on its conduct of monetary policy.

Modern central bank laws establish price stability as the primary objective of the central bank and as the single objective of monetary policy.1 In examining the statutory powers of the central bank, we should keep in mind that these have been defined to promote a particular goal, namely, price stability. If the primary objective of the central bank were different and consisted of another goal such as economic prosperity,2 or several competing goals,3 such as domestic price and exchange rate stability, the central bank would require different powers. In practice, however, the superficially simple proposition that the central bank should focus primarily on the single overriding objective of price stability has turned out to be problematic, because monetary policy is subject to limitations that severely restrict the ability of the central bank to achieve that objective. The existence of such limitations is recognized by modern central bank law, where the law expresses the goal of price stability, not in terms of an obligation to be discharged, but rather as an objective to be pursued by the central bank.

Limitations of Monetary Policy

The conduct of monetary policy is notoriously difficult. In pursuing price stability, the central bank is like a battleship riding shifting waves while aiming at a target constantly changing its speed and course. Economic currents are difficult to read and even more difficult to predict. Sizing up the target in current economic time is problematic, because the significance of statistical data is often impaired by subtle changes in the patterns of economic activity or the quality of goods that they measure.4 Moreover, the time needed to collect, process, and analyze statistical data causes their value to be reduced when they are used for monetary policy decisions. To make matters worse, as the instruments of monetary policy available to the central bank take effect only after a time lag of many months, it is the prediction of the future effects of current monetary policy that should determine monetary policy. These factors conspire to render the conduct of monetary policy as much an art as a science.

There are yet additional complications. Monetary policy is intended to change economic behavior. Instability of prices generally results from economic market activity that is based on inflationary or deflationary expectations of market participants. Therefore, the success of a central bank in fighting price volatility is largely dictated by its ability to control or change market psychology and thus market behavior, whereby the credibility of its policy stance plays an important role.

Ultimately, the success of monetary policy depends on its conduct. Monetary policy is there to be carried out and monetary policy instruments are there to be used. What if the central bank fails in its task? What if monetary policy is unsuccessful because it does too little or comes too late to be successful due to inactivity on the part of the central bank? Here, the central bank law presents a dilemma. In order to endow the central bank with autonomy, the grounds specified in the law that permit the dismissal of the management of the central bank often omit dereliction of duty.5 And yet, that is what would be required to get rid of an incompetent or crooked central bank governor. One solution to this difficulty has been to force the central bank into the straitjacket of a currency board.

The monetary policy instruments that are available to a central bank are limited in scope and effect. They are usually designed to affect short-term interest rates charged by financial institutions. As was said before, the effects of monetary policy measures of the central bank on spending patterns, and thus on the rate of inflation, show up only after a considerable time lag. Because in a market-based economy monetary policy instruments typically target economic behavior, the effect of monetary policy instruments depends on the perceptions and expectations of the marketplace. Sometimes, the perception of the central bank is out of sync with the perceptions of the financial markets. For example, if the financial markets are concerned about inflation, a lowering of short-term interest rates by the central bank may fuel these concerns, cause the yield curve to steepen, and even perversely lead to higher long-term interest rates.6

The objective of price stability serves to prevent and combat not only price inflation but also price deflation. Although this objective appears to be evenly balanced, the monetary policy instruments available to the central bank to pursue this objective are not. In particular, the effects of the interest rate instrument in the conduct of monetary policy on inflation and deflation are qualitatively different. On the inflationary side of price stability, higher interest rates are quite effective in increasing the cost of borrowing by the public to pay for the accelerated buying that characterizes inflation. However, on the deflationary side of price stability, lower interest rates often fail to stimulate increases in buying that are required to break a deflationary spiral. Among the reasons for this asymmetry are the following: the interest rate instrument, though unlimited on the upside, is limited on the downside, as interest rates cannot go below zero. When deflation accelerates, lower interest rates that at first stimulate buying lose their intended effect, because falling prices impede capital investments by the corporate sector, while widespread unemployment accompanying a deflationary cycle causes the public to shun the assumption of new debt even at minimal cost. Japan is an unfortunate example of a country where structural economic weaknesses combined with limitations of monetary policy have presented sizable obstacles in the fight against deflation.

Monetary policy instruments are designed to conduct monetary policy. They are not designed for the conduct of other macroeconomic policies. There is ample evidence that price stability depends not only on successful monetary policy of the central bank but also on other macroeconomic policies affecting price stability, such as fiscal and wage policies, that are pursued by the government outside the control of the central bank. Price stability depends on a proper macroeconomic policy framework of which monetary policy is a part.

The dependence of price stability on a proper macroeconomic framework that stretches beyond monetary policy has been recognized for many years. The limitations of monetary policy in this respect are among the reasons why the first report on European monetary union concluded that to be successful the introduction of a single currency required not only a single European central bank but also a European center of macroeconomic decision making.7 Nearly two decades later, the Report on Economic and Monetary Union in the European Community prepared by the Delors Committee in 1989 emphasized that economic and monetary union are integral parts of a whole and must proceed in tandem.

In particular, the dangers of inconsistencies between monetary and other macroeconomic policies are well understood. There are many examples of countries in which, at one time or another, the efforts made by more or less autonomous central banks to maintain stable prices were undermined or brought to naught by excessive deficits in government budgets.

Based on this experience, the Maastricht Treaty requires as one of the principal elements of a proper macroeconomic framework that member countries avoid excessive budget deficits. Although the Treaty defines and caps such deficits for purposes of joining the single currency,8 it is vague on the standards and procedures to be observed by countries thereafter. To remedy this situation, the European Council adopted in June 1997 its so-called Stability and Growth Pact. The pact builds on the progress achieved by countries in convergence, as evidenced by their ability to meet the convergence criteria, and especially in reducing budget deficits and curtailing public debt. For the future, the pact requires countries in the euro area to have medium-term budgetary positions that are close to balance or in surplus, and attaches sanctions to the failure of countries to do so; exceptions are made for excessive deficits resulting from exogenous conditions or a severe economic downturn.9 The pact establishes procedures for surveillance of progress made by countries in achieving this goal. It has been observed that a principal weakness of this system is that the numerical thresholds do not vary in line with cyclical changes in economic conditions and therefore fail to require policy adjustments according to phases in the cycle.10

Toward a Model of Cooperation

Based on the foregoing, it is fair to conclude that the limitations in the power of the central bank to pursue price stability through monetary policy and the influence on price stability of other macroeconomic policies that are controlled by the government raise questions about the model that makes an autonomous central bank solely responsible for maintaining price stability. Perhaps the most obvious of these questions is why the duty to maintain price stability is a unilateral duty that rests only on the central bank and is not a joint responsibility of central bank, government, and parliament.

Autonomy of the central bank is often supported by the argument that countries with autonomous central banks have been more successful in fighting inflation than other countries. Thus, it is said that the public is more likely to place its trust in a central bank that is independent from political influence in the formulation and execution of monetary policy than in the political establishment, and that an autonomous central bank is therefore more efficient and successful in achieving the objective of price stability than a central bank without political independence.

This argument is somewhat suspect, however, at least where it omits showing a relationship of cause and effect between price stability and central bank autonomy. Often the defenders of central bank autonomy fail to demonstrate that price stability is a product of central bank autonomy, and not (also) of some other political or socioeconomic condition that produces or contributes to price stability. It is conceivable that central banks with the greatest autonomy are found in those countries that have the strongest sociopolitical commitment to price stability; an example would be Germany. There are other countries where a strong sociopolitical consensus elevates price stability to one of the primary macroeconomic concerns of the nation, regardless of the autonomy granted to the central bank by law; examples of these would be Australia, Canada, New Zealand, and the Netherlands before the Maastricht Treaty, all countries with a strong commitment to price stability notwithstanding provisions in their central bank law permitting the government to give monetary policy directives to the central bank.11 Another example would be France, which in the mid-1980s made a strong political commitment to domestic price stability, even though at that time its central bank failed to meet the strict standards of autonomy that were subsequently introduced by the Maastricht Treaty. There are indications that in these countries price stability resulted not so much from central bank autonomy as from effective policy cooperation between the central bank and the government.

The importance of a governmental commitment to price stability can also be concluded from the experience of countries in which such a commitment on the part of the government was lacking. A recent example of a country where a strict monetary policy alone proved inadequate to avoid economic disaster is Argentina. To impose discipline on the monetary authorities, the central bank had been turned into a currency board and its monetary policy role had been reduced to that of a foreign exchange bureau. Economic conditions deteriorated when the peg to the strong U.S. dollar produced an equally strong peso in the foreign exchange markets and made it difficult for Argentine exporters to compete in Europe and Brazil, Argentina’s principal export markets. When, consequently, the risk increased that Argentina would have to break the peg, foreign lenders commanded higher interest rates to fund Argentina’s external debt. In the end, the financial system of Argentina collapsed because the government lacked the discipline to act.

These examples point to the need for some form of agreement between the central bank and the government in the pursuit of price stability. There are countries where monetary policy agreements are concluded between the government and the central bank. In New Zealand, for example, the central bank law requires that every time a Governor of the Reserve Bank is appointed or reappointed, a new policy targets agreement be negotiated between the Treasurer and the Governor.12 However, these policy targets agreements create only a superficia1 appearance of cooperation between the government and the central bank; a closer look reveals that the cooperation is rather one-sided, as only the central bank is required to observe the targets. There is no equivalent commitment on the part of the government. Actually, the government is by law authorized to direct the Reserve Bank by Order in Council to formulate and implement monetary policy for an economic objective other than price stability.13 It has been observed that, because the law requires that policy target agreements be published, and that any change to such agreement or any order directing the Reserve Bank to disregard the target also be published, the financial markets will keep unjustified departures from the target by the government in check. However, a major weakness of this system is that the Act fails to address the possibility that other macroeconomic policies of the government might be or become inconsistent with the agreed-upon monetary policy targets.

It is clear from the foregoing that, in and of itself, the model of an autonomous central bank does not suffice, at least not for all countries or for all seasons. Even the best central bank cannot successfully use its monetary policy tools to sterilize all conceivable inappropriate macroeconomic policies of the government. And there are monetary conditions—a deflationary spiral is one of them—under which monetary policy is inherently inadequate to do the job.

The autonomous central bank appears to be the product of a fragmented view of economic policymaking. By setting monetary policy apart and assigning its conduct solely to the central bank, the state has made it harder for itself to adopt a comprehensive framework for all relevant macroeconomic policies. To be effective, monetary policy cannot operate successfully in a vacuum but must be incorporated into a good and complete national macroeconomic policy strategy that is supported by a strong sociopolitical commitment and combines all elements of macroeconomic policy in a manner that does justice to the relative weight assigned to each element and the interaction among them.

In addition, there are constitutional reasons to question the model of an autonomous central bank exclusively assigned to conduct monetary policy. As such, the central bank is a product of the peculiar practice whereby society relinquishes some of its political controls to delegate an important public task to a technocratic agency that, after it has been created, is more or less independent from the state.

The delegation of state power to an autonomous central bank raises questions concerning its position in a democracy. The hallmark of a parliamentary democracy is that the power of decision making in matters concerning the state rests with the elected representatives of the people. At its most autonomous where it has become apolitical, the central bank represents a departure from this principle. There is something inherently undemocratic about an apolitical central bank. In a democracy, state responsibility is political responsibility. The more autonomous the central bank becomes, the closer the delegation of state responsibility reflected by its autonomy moves toward an abdication of political responsibility.

There are not only good technical reasons for cooperation between the central bank and the political establishment. There are also good political reasons for such cooperation: monetary policy cannot succeed without firm commitment at the political level. Such a political commitment is generally lacking, in part because the model of an autonomous central bank does not explicitly require the central bank to take political concerns into account in the conduct of monetary policy. For example, monetary policy often carries social and economic costs, such as slower economic growth, that are of political concern because they are borne by society as a whole. One would expect that this would require sensitivity analyses between the costs and benefits associated with various monetary policy choices. However, most central banks are not required to perform such analyses. More important, the model of the autonomous central bank does not provide for a mechanism that would ensure that monetary policy decisions would take the results of such analyses into consideration in the conduct of monetary policy.

The foregoing does not mean that there would be no role in the conduct of monetary policy for an autonomous central bank. It does, however, point to the need to integrate this role into a broader setting where all macroeconomic policies are considered together, first at the technical level and then, conclusively, at the political level. If one thing is clear in this, it must be that price stability demands a firm commitment, not only at the technical level of a central bank, but also and especially at the political level, where decisions can be reached that reflect a broad national consensus and commitment that price stability is a great good that is well worth the cost.


See, e.g., Article 2 of the Statute of the European System of Central Banks and of the European Central Bank set forth in the Protocol attached to the Maastricht Treaty of December 1991.

As a macroeconomic objective, price stability is not the same as economic prosperity. However, price stability is generally understood to serve the broader goal of economic prosperity. For an example of a central bank law where this relationship has found expression, see Article 2 of the Bank of Japan Law of 1997.

In the United States, the Board of Governors of the Federal Reserve System and the Federal Open Market Committee are required “to promote effectively the goals of maximum employment, stable prices, and moderate long-term interest rates.” The Federal Reserve Act, 12 U.S.C. § 225a. The standard argument against multiple policy objectives is that, eventually, they tend to lead to policy conflicts.

Textbook examples of cases where the consumer price index fails to measure such changes adequately are a shift from beef consumption to chicken consumption when beef prices rise, or an increase in output of electronic equipment without price increase (which is tantamount to a price reduction).

An exception is Section 49 (2) of the Reserve Bank of New Zealand Act 1989, which provides for the removal of the Governor from office on the ground (a) that the Reserve Bank is not adequately carrying out its functions; or (b) that the Governor has not adequately discharged the responsibilities of office; or (c) that the Governor has obstructed, hindered, or prevented the Board from discharging its responsibilities under the Act; or (d) that the performance of the Governor in ensuring that the Bank achieves the monetary policy targets agreed upon with the Minister of Finance has been inadequate; or (e) that the Minister of Finance and the Governor have not been able, within the time prescribed by the law, to agree on new monetary policy targets.

Monetary policy instruments are designed to have a direct effect on short-term interest rates only. Inflationary expectations, however, are often reflected primarily at the long end of the maturity spectrum in the form of an inflation risk premium that increases long-term rates. Long-term rates do not always move in tandem with the short-term rates that are controlled by the central bank. A lowering of short-term rates by the central bank may have the unintended effect of producing higher long-term rates, namely, if the capital markets judge the rate cut unjustified in view of their inflation expectations and interpret the action of the central bank as a weakening of its anti-inflationary stance. To avoid misinterpretations by the financial markets of monetary policy, and especially surprises that may trigger damaging increases in price volatility, the central bank must constantly announce and explain changes in its monetary policy stance to the public; when done correctly, public statements on monetary policy may cause the market correctly to anticipate subsequent changes in monetary policy, doing some of the work of the central bank by raising or lowering market prices of debt securities, translating into lower or higher interest rates.

The Werner Report of 1970 (Report to the Council and the Commission on the realization by stages of economic and monetary union in the community).

Article 109j(l) regards as excessive a general government deficit in excess of 3 percent of gross domestic product (GDP) accompanied by a general government debt of more than 60 percent of GDP.

Sanctions initially take the form of nonremunerated deposits with the European monetary authorities of between 0.2 percent and 0.5 percent of GDP; if the deficit is not corrected within two years, the deposit turns into a fine and is forfeited. Additional fines may be levied if excessive budget deficits continue thereafter. See generally, Council Regulation (EC) No. 1467/97 of July 7, 1997, on speeding up and clarifying the implementation of the excessive deficit procedure; Official Journal No. L 209, 02/08/1997, at 6–11.

European Commission Communication on strengthening economic policy coordination within the euro area, 7/2/2001, COM(2001) 82.

Australia: Section 11 of the Reserve Bank Act 1959; Canada: Section 14(2) of the Bank of Canada Act; New Zealand: Section 12 of the Reserve Bank of New Zealand Act 1989; Netherlands: Article 26 (old) of the Bank Act.

Article 9 of the Reserve Bank of New Zealand Act 1989. The Policy Targets Agreement of December 1999, which is currently in force, specifies a policy target of 12-month increases in the All Groups Consumers Price Index of between 0 and 3 percent.

Article 12(1) of the Reserve Bank of New Zealand Act 1989.

    Other Resources Citing This Publication