The issue of central bank independence has generated considerable debate all over the world in recent years. We are all familiar with the much publicised reforms to the Reserve Bank of New Zealand. In Europe, a key element in the European Monetary Union is the formation of an independent supranational central bank. Many of the ‘transitional economies’ of eastern Europe also have adopted reforms aimed at making their central banks more independent. In what is something of a rarity these days, the issue has attracted attention from both practitioners and academic economists.

The issue is as old as central banking itself, having been debated on and off over the past couple of hundred years. The hallmarks of independence – namely, autonomy from the government and non-financing of budgets – were identified clearly by David Ricardo in a paper on the establishment of a national bank in 1824:

It is said that Government could not be safely entrusted with the power of issuing paper money; that it would most certainly abuse it … There would, I confess, be great danger of this if Government – that is to say, the Ministers – were themselves to be entrusted with the power of issuing paper money. But I propose to place this trust in the hands of Commissioners, not removable from their official situation but by a vote of one or both Houses of Parliament. I propose also to prevent all intercourse between these Commissioners and Ministers, by forbidding any species of money transactions between them. The Commissioners should never, on any pretense, lend money to Government, nor in the slightest degree be under its control or influence … If Government wanted money, it should be obliged to raise it in the legitimate way; by taxing the people; by the issue and sale of exchequer bills; by funded loans; or by borrowing from any of the numerous banks which might exist in the country; but in no case should it be allowed to borrow from those who have the power of creating money.

Earlier this century, Keynes expressed his thoughts on central bank independence while testifying before the 1913 Royal Commission into an Indian central bank. The ideal central bank, he said, ‘would combine ultimate government responsibility with a high degree of day-to-day independence for the authorities of the bank’. He added that it would be desirable ‘to preserve unimpaired authority in the executive officers of the bank, whose duty it would be to take a broad and not always commercial view of policy’.

There has always been some kind of a relationship between central banks and governments. Given that central banks are created by government legislation and derive their powers from such legislation, they cannot be completely separate from the government. The debate today is about the appropriate degree of separation.

Is this renewed interest in central bank independence just another passing fashion? I think not. Rather, it is a reflection of history and the changed policy environment. In part it reflects the worldwide surge of inflation in the 1970s and the onset of the new phenomenon of ‘stagflation’. In part also it is a recognition that the anchors which held prices stable in earlier eras – first the gold standard, then the Bretton Woods system – had come adrift and something else had to be put in their place.

Central bank independence has emerged in that search for a new anchor. The earlier arrangements had imposed an international discipline on countries but when they passed into history, the responsibility for maintaining price stability reverted to national authorities. New ways had to be found, in the conflicting priorities of national policy making, to anchor prices against inherent swells in inflationary pressures.

Today I would like to canvass some aspects of this mammoth topic from a practitioner's perspective. To this end, I want to consider five broad questions:

• Why is central bank independence an important issue?
• What is central bank independence?
• Does increased central bank independence lead to lower inflation?
• Does prudential supervision compromise independence? and
• What does accountability mean?

None of these questions has a straightforward answer. Different models of central banks exist and work. This is, I believe, in keeping with the different historical origins of central banks, and the different environments in which they have evolved, sometimes propelled along more by accidental factors than by the deliberate intent of legislators or central bankers. Most central banks these days pursue a number of functions but the emphasis here is on monetary policy; this is the function which, since the 1970s, has been the focus of the debate about the independence of central banks.

Why is central bank independence an important issue?

The brief answer to this question is that price stability is generally considered a good thing, and that an independent central bank can help to achieve it.

I do not need to dwell on the desirability of price stability. Economies work better if investment and wage decisions are not confused and thwarted by high inflation. Some people see price stability as an anchor not only for the economy, but also for society at large. Price stability, however, is not the natural order of things in a modern economy. Apart from higher oil prices and other ‘shocks’, there are two particular threats which bear upon the issue of central bank independence:

• the tendency for policy makers and politicians to push the economy to run faster and further than its capacity limits allow; and
• the temptation that governments have to incur budget deficits and fund these by borrowings from the central bank.

Let us examine these two sources of inflationary pressure more closely.

It is understandable that policy makers and politicians (in part in response to public pressure) should wish to squeeze as much growth as possible out of the economy – to run it a bit faster and a bit further than its capacity limits allow. In the jargon, there is a tendency to exploit the short-run trade-off between output and inflation. Even economies which have few spare resources can grow at rates above their long-term capacity limits for a time. Although it may take some time to show up, higher inflation will be the inevitable result.

Cynics blame this inflation bias on the political process, claiming that politicians have short time horizons, stretching only as far as the next election. According to this view, we end up on the treadmill of a political business cycle, where interest rates are eased prior to an election to generate strong growth around election time, and the inflation does not surface until after the election. After the election, interest rates are raised, thereby perpetuating a boom/bust cycle. In this story, even politicians who understand that there is no long-term trade-off between inflation and economic activity are drawn irresistibly to the short-term trade-off. Independent and longsighted central bankers are needed to rescue politicians from this temptation.

I am less cynical about politicians generally, but I nonetheless think there is an inherent bias in policy making which pushes policy makers towards a trade-off point which is soft on inflation. We can warn that economies cannot continue to operate above long-term capacity but no-one knows with any confidence just where those limits are or what the ‘natural’ rate of unemployment is. Nor do they know precisely where the economy is in the cycle, or what in-built momentum it is generating. And they do not know the length of the lags between policy changes and their impact on growth and inflation.

In short, as all practitioners know, policy is determined and implemented under conditions of great uncertainty. It is this uncertainty – rather than any cynical exploitation of the short-run inflation/output trade-off – which can inject an inflationary bias into the policy making process. In conditions of uncertainty, policy makers may believe that the economy can run just a little longer at a fast pace, or go a little further. They may be reluctant to take away the punchbowl just when the party gets going – that unpopular task which the Governor of the US Fed from 1951 to 1970, William McChesney Martin, said was the duty of central bankers.

The other source of inflationary pressure is the obligation some central banks have to fund their governments' budgets. There is a fundamental conflict between independence and an obligation to finance the budget deficit – a conflict which, generally, is resolved at the expense of price stability.

I do not need to labour the potential dangers of this channel. (Here we are entitled to be cynical about governments which survive by printing money when they could not survive by other means.) It has been the major cause of high inflation and hyperinflation in a number of countries over the years. It was responsible for the hyperinflation of Indonesia in the 1960s, before the balanced budget rule was introduced in 1966. It is contributing to current inflationary problems in several former Soviet Union and eastern European countries. The examples are numerous, and the lesson is clear: unless a central bank is protected from the need to fund budget deficits, price stability rests more in the hands of the budgetary authorities, than the central bank. In short, sound money cannot be maintained without a sound fiscal system.

So much for the problem. What can be done? To be able to do their job of keeping inflation under control, central banks have to be able to say ‘no’ to governments when that objective is threatened. This is why the notion of central bank independence is so important.

What is central bank independence?

In principle, the answer to the first threat is relatively straightforward – give central banks a charter which includes a strong commitment to price stability, and the freedom to pursue it. This does involve the government in setting the goals, but that is the way it should be: central banks cannot expect to determine the goals they should pursue, but they should have adequate scope to pursue the goals that have been set. In the jargon, they should not have goal independence but they should have instrument independence.

That is clear enough in principle, but it has been interpreted differently in practice, resulting in different approaches by central banks. The Bundesbank is directed simply to ‘safeguard the currency’. The Reserve Bank of New Zealand has the very specific goal of pursuing an inflation target between zero and 2 per cent. In Australia, the Reserve Bank Act specifies stability of the currency and maintenance of full employment as the central bank's objectives. In the United States, the Humphrey-Hawkins Act requires the Federal Reserve to conduct monetary policy to promote the goals of ‘maximum employment, stable prices, and moderate long-term interest rates’.

Very broadly, three approaches to goal setting can be identified. One is to give the central bank a single goal of price stability. A second is to give the bank an intermediate target – such as an exchange rate fixed to the currency of a strong anti-inflation country, or specific targets for the rate of growth of a particular money aggregate. A third approach is for the government to set multiple goals, which include price stability.

Each approach has its drawbacks. A single goal of price stability could prove unduly constraining if unforeseen circumstances were to warrant the central bank giving more weight to things other than the simple inflation target. (In those circumstances, of course, the government could always change the goal.)

Intermediate targets (sometimes embellished in ‘rules’) imply a somewhat mechanical and automatic approach to monetary policy: if the exchange rate is under downward pressure, tighten policy; or if M3 or some other category of money is growing above the target, again the response would be to tighten. (This is, of course, to oversimplify; even if a central bank were to respond within the dictates of the rules, it has still to exercise its judgment as to how fast it should seek to get back on target.)

For some countries, intermediate targets have provided a valuable compass for monetary policy settings, adding to the credibility of the central bank in the process. The role of the fixed exchange rate in Hong Kong in recent years is one example. Many European countries also have found it useful to commit themselves to maintaining a more or less fixed exchange rate with the German mark.

In the years following the breakdown of the Bretton Woods arrangements, most central banks experimented with intermediate targets of one kind or another. Some, as just noted, still have them. But for many countries (including my own), they proved unsatisfactory, either because any relationship that might have existed between monetary aggregates and nominal GDP broke down with the onset of financial deregulation, or because maintaining a fixed exchange rate was not a sensible policy option.

A potential danger with all intermediate targets is that they can require policy adjustments inconsistent with ultimate objectives. A slavishly pursued monetary target could result in either inflation or deflation; a monetary policy directed at a fixed exchange rate could prove too tough or too soft (and may ultimately prove unsustainable, as several European countries found with the ERM in 1992). In short, having a target other than the ultimate objective might help to anchor monetary policy, but the authorities have to be sure that the intermediate target is the correct one.

The third approach is to require the central bank to give a high priority to price stability while also having regard to other objectives, such as growth and employment. In this case, the priority attached to fighting inflation at any point in time will depend a good deal on the make-up of the Board of the bank and its Governor, and some people (particularly those who rank price stability above all else) will say that this leaves too much to chance.

The charter of the Reserve Bank of Australia comes within this third category. It is one which I am comfortable with. We have a very clear commitment to keeping inflation under control, but that is not formulated in a way which restricts flexibility in the exercise of monetary policy. To my way of thinking, the flexibility of this approach suits the complexity and uncertainty of the real world better than simple, single goals.

In one sense, central banks with multiple goals have more independence, because they have extra dimensions on which they must make decisions. Some would say that this gives the central bank more scope for making mistakes and for downgrading the priority of price stability. But this need not be the outcome. An independent central bank has no inherent reason to misuse the short-term trade-off, provided the bank has a clear commitment to pursuing price stability. In Graph 1 Australia's case, this commitment is reinforced by the self-imposed discipline of keeping underlying inflation to around 2 to 3 per cent through the cycle. Independence is vital to good policy making where the central bank has multiple goals but, given that independence, there is nothing inherently inconsistent in pursuing multiple goals.

As to the threat of inflation which arises when central banks are required to fund government budgets, the simple answer is to remove that requirement – which is easier said than done in the countries concerned. Although Ricardo drew attention to this problem 170 years ago, it is only in recent decades that central banks have moved towards positions where they can say ‘no’ to funding budget deficits (and other development objectives) which are at odds with good macro management. The longer term answer is to be found in building domestic bond markets and financial intermediaries which will be able to shoulder these funding tasks.

Does increased central bank independence lead to lower inflation?

This question has been the focus of a number of studies of the relationship between central bank independence and inflation. The usual approach has been to create an index of central bank independence (based largely on elements in the relevant legislation which the authors consider to be proxies for independence), and to compare that with the inflation performances of the countries sampled. These studies show that countries with (legally) more independent central banks tend to have lower inflation (see graph for one example). They show also that greater (legal) central bank independence is not associated with a lower rate of economic growth. At face value, this suggests that central bank independence is in the nature of a ‘free lunch’: increasing the independence of the central bank delivers lower inflation which, in the long run, is not at the expense of lower economic growth.

As with all ‘free lunches’, I think this is too good to be true. Most rankings of central banks by legal independence which I have seen do not capture the full complexity of the issue, and some are downright misleading. To list a couple of points:

• legal independence, which is what the empirical studies seek to measure, can be very different from practical independence. In Australia, the legislation provides that in the event of a dispute over monetary policy, the government can override the Reserve Bank by tabling its objections before both houses of parliament. While this is the legal position (which scores a negative on the index of independence), such a situation has never arisen in practice; and
• indexes of legal independence cannot capture the role of personalities (such as the Governor and the Treasurer/Finance Minister) who fill in the operational gaps in the relevant legislation. Nor can they capture the changing policy-making environment. For example, the greater reliance on market-based policies (rather than controls) in Australia over the past decade has significantly enhanced the degree of independence of the Reserve Bank, without any change in the Bank's charter.

An important point here is that the correlations reflect a presumption – wrongly – that central banks and monetary policies are the only influences on inflation. Fiscal, wages and other policies are important too – which is why, incidentally, I put a considerable premium on effective consultation and policy coordination between the central bank and the government. Monetary policy needs to be properly co-ordinated with other economic policies; the prospects for price stability are best where the Governor and the Minister(s) are close allies in that battle.

Other ‘third factors’ might also lie behind some correlations. The low inflation record of Germany and the independence of the Bundesbank, for example, are both related to the inflation aversion of the German people, following the experience of hyperinflation in the 1920s.

We should, then, be wary about drawing too firm conclusions from the findings of these sorts of studies. It is a big jump from their findings to the conclusion that every country needs or can have a Bundesbank. The Bank of Japan is usually rated as having a low level of independence, but Japan has an excellent inflation record. There are no universal rules. We can all learn from what other central banks have done, but each country must establish the legal framework for its central bank, and allow for its evolution, in ways which best fit that country's own history and institutions.

Does prudential supervision compromise independence?

Prudential supervision of the banking system is a formal core function of most central banks, but not all. (In Germany, for example, the central bank does not have formal supervisory responsibility.) This is another illustration of the point that there are different approaches to tasks.

It is sometimes argued that central banks with responsibilities for bank supervision will be inhibited in conducting an independent monetary policy because of those responsibilities. In circumstances where monetary policy should be tightened, the central bank might hesitate, so the argument goes, because of a concern about the effects on one or more banks.

I think this and other arguments for separation of the two functions are overdrawn. During several years when conditions have been difficult for some banks, it does not seem to me that there has been a conflict between monetary policy and prudential supervision. In practice, it seems natural that a central bank will always consider the effects of its actions on the financial system as a whole. Moreover, if an individual bank should find itself in distress it will almost certainly require remedial action which is different in kind from that which could be afforded through deferment of monetary policy action. Tensions will sometimes arise whether the supervisory agency is within the central bank or external to it; in my mind, there is something to be said for internalising these tensions.

More positively, coming from a central bank which performs both functions, I am aware of some ‘synergies’. Much of the information we gather on the financial system in our role as supervisor assists us in our consideration of monetary policy, given that the banking system is a major transmission mechanism of central bank actions. We get a good feel, for example, for the extent to which any slackness in credit growth is concentrated on the demand or supply sides, as well as for the strengths of the different sectoral (e.g. housing and business) demands for credit. This and other information about the system could always be obtained from another agency – as it is by non-supervising central banks – but its internalisation can be both convenient and efficient.

What does accountability mean?

As interest in central bank independence has increased, so too has interest in central bank accountability – which is understandable, given that they are closely related. As a central bank becomes more independent, it needs to be more accountable for its actions. Accountability, however, is a broad concept: to whom should a central bank be accountable, and for what?

The focus is mainly on policy accountability. In general, central banks should be accountable for achieving the goals specified for them in their charters, and they should be accountable to the parliament, as representatives of the public. Other bodies – such as the media and the financial markets – will also take it upon themselves to pass judgments upon monetary policy; they are entitled to do that, but we must remember that their judgments will usually reflect narrower perspectives, and shorter time horizons, than those of the central bank.

Central banks should be accountable in terms of their charters, but they can express their accountability in different ways. In New Zealand, the Governor reports on progress in achieving the government's very specific inflation target. In the United States, the Chairman of the Federal Reserve is obliged by the Humphrey-Hawkins Act to testify before Congress several times a year. In the United Kingdom, the Bank of England now publishes a Quarterly Inflation Report as part of its endeavours to be more accountable.

In Australia, the Reserve Bank engages in the usual practices of regular public speeches, quarterly articles and annual reports, and testimony before parliament. In addition, and unlike some other central banks, it issues relatively detailed press statements at the time of each change in interest rates, both to announce the change and to explain the reasons for it. This serves to increase the transparency of the monetary policy process and helps to avoid confusion in the market place. More generally, by reducing the mystique surrounding the process and clarifying the central bank's role in it, this transparency serves not only to increase accountability but also independence.

Central banks should be accountable also for the sometimes substantial staff and financial resources which are entrusted to them. In Australia, the Reserve Bank is responsible for its own budget, which gives the Bank financial independence from the government. It also heightens the need for the Bank to account publicly for its budgetary and financial operations. To this end, our reporting includes data on trends in staff numbers and productivity performances, as well as on the returns the Bank earns from its management of the country's international reserves and from its intervention in the foreign exchange market.

What can we conclude about central bank independence?

A number of points can be drawn from the foregoing discussion:

1. Central bank independence is a major policy issue today, largely because of the on-going search for an institutional framework that will help monetary policy to deliver low inflation over the medium term. It can be seen as part of the lineage of the gold standard of the late 19th century, the Bretton Woods system of the early post-war era, and the monetary targeting of the 1970s.
2. ‘Independence’ in this context means the freedom of central banks to pursue monetary policies which are not dictated by political considerations. It does not preclude Ministers from commenting on monetary policies, and it does not preclude central banks from consulting with the government on monetary and other policies. In practice, varying degrees of independence have been exercised through a variety of approaches. Some of these constrain the central bank's room for manoeuvre by providing a single final objective or a single intermediate target, while others provide more flexibility to central banks to respond to uncertainty. Different approaches can and do work.
3. Increased central bank independence does not necessarily lead to lower inflation. This is because monetary policies, on their own, cannot guarantee to deliver lower inflation without unacceptable costs in terms of lost output and jobs. Fiscal and wages policies have an important bearing on inflation outcomes, and these need to be compatible with an anti-inflation monetary policy.
4. Legislated independence will not necessarily deliver practical independence, but any action that limits the ability of governments to finance their budgets from central banks is likely to enhance effective independence.
5. ‘Credibility’ is helpful to central banks in implementing monetary policy and a pre-condition for this is that the central bank be perceived to be independent and free from political interference. Beyond that, however, credibility has to be earned, essentially through the consistent demonstration over a long period of the bank's determination to achieve its goals.
6. If central banks are to be independent of the government, then they must be accountable for their actions. Not only is this proper in a well run society, but public accountability can help to preserve the independence of central banks. Provided the decisions of central banks are competent to begin with, and are transparent and understood by the public, there will be less opportunity for political interference.
7. The competence of central banks and the personalities of their Governors, and of Treasurers and other Ministers, are obviously important, whatever the precise legislative framework. In this regard, central bank independence can be likened to a game of cricket. The legal framework that central banks operate in is important, as the rules of the game are important in cricket. What matters most to the outcome of the game, however, is the performance of the players on the field.