Central Bank Communication
Address to The Sydney Institute
Sydney - 11 December 2007
This evening I would like to talk about central bank communication. There are plenty of people who probably think that ‘central bank’ and ‘ communication’ are not expressions that are normally thought of as belonging together. Yet communication has become steadily more open, and more important, in the central banking world over the years. At one time, overnight interest rates used to be adjusted without announcement or explanation – ‘ snugging’ was a popular term of the late 1980s. The market would generally take a few days, and the general public considerably longer, to work out what was happening when monetary policy was changing. They usually had to wait longer again for an explanation of why it was changing.
That is no longer the case. These days, central banks say they are moving the interest rate, and say why, very openly. This particular form of transparency was adopted quite early at the RBA, back in 1990, under Bernie Fraser. Not only do they announce policy changes in a forthright fashion, most central banks set out their detailed assessment of economic conditions in published material. In a number of cases, they publish minutes of their policy discussions. Governors and others give speeches, appear before parliamentary committees and are frequently quoted by various commentators in the media.
For central banks, historically quite discreet organisations, this is quite a change. I want to examine the reasons behind that change, and the case for openness. I want also to talk about the limits of openness. Finally, I will elaborate a little on our own recently announced changes in this area
To begin with, it is worth asking the question: why do central banks communicate?
There are at least two reasons. The first arises from the principle of accountability in an open and democratic society. The central bank is charged with some important responsibilities – maintaining the purchasing power of the nation’s currency, pursuing ‘full employment’ and fostering stable and reliable financial and payments systems, among others. It is endowed with substantial powers – in the RBA’s case, to buy and sell financial assets, commodities, real property, to lend money and to make certain regulations – in fact, to do anything that could be thought to come within the remit of ‘central banking business’.
It stands to reason that the central bank should expect to account for the way in which it has used its powers to pursue its statutory goals. The framers of the Reserve Bank Act 1959 nearly 50 years ago very wisely placed the Parliament squarely in the centre of accountability arrangements. In recent times, the Commonwealth Authorities and Companies Act 1997 has required annual reporting of the relevant agencies, of which the Reserve Bank is one, but from the very beginning the RBA was required to issue an annual report to Parliament.
Yet the requirements for more frequent communication have grown over the years. In part this reflects the increased development of capital markets and the speed and force with which they respond to economic developments. Markets crave information, and the central bank’s assessment of the state of the economy is one part of the information set.
In part, the demand for communication accompanies increased operational independence for central banks. When the central bank is making an important policy decision under an authority delegated by the Parliament, as opposed to implementing a decision made by a minister – and that is what we are doing in the case of monetary policy – there will naturally be an expectation for accountability. It also reflects the general development of the community’s expectation to be more informed about important matters. That is a natural concomitant of a more affluent, educated and mature society. As such, it is something to be welcomed.
There is also the role of the media. Our political leaders are expected to answer questions from the media much more frequently, and across a much broader range of issues, than once was the case.1 Some of this spills over onto other institutions, including corporations and central banks. The media is responding to market demand for information here, but as in any other competitive industry, media organisations are also seeking to create new markets by supplying more intensive, more frequent coverage of more issues, including economic policy.
The second rationale for more communication is the desire for more effective policy. Some policy actions have as much effect through conditioning expectations as by constraining current behaviour. This is very much true in the case of monetary policy. A change in the cash rate of modest size has only a pretty small impact on the economy. But expectations about a sequence of possible future interest rate changes can often be more powerful. What a central bank says, as much as what it does, affects those expectations.
Even more important are expectations of future inflation. When people expect prices to rise rapidly, they bring forward purchases, put up their own prices, demand higher wages and so on. That helps to create the very inflation they expect. On the other hand, if people are convinced that inflation will be contained – perhaps because they believe that the central bank will do whatever is required to achieve that – they behave accordingly. In that case, their price‑setting, wage and purchasing behaviour helps keep inflation controlled.
Expectations are more likely to be helpful in fostering economic stability when the public has a clear understanding of what the central bank’s goals are, how the central bank thinks the economy works and how in general terms it is likely to respond to various events, particularly pressure on inflation. That is why most central banks spend a good deal of time talking about their objectives and their policy framework. It is why they publish exhaustive analyses of economic conditions and offer such assessments as they can about likely future developments.
Of course the world is highly uncertain, so central banks cannot spell out exactly what they would do in response to every conceivable scenario. There will always be the potential for some surprises. But if people understand the framework and the goals of policy, then their own response to that knowledge will usually be helpful in achieving the policy goal. Hence, communication is an important part of the policy process.
How to communicate? And how much?
So much for the rationale for communication. What are the channels central banks use? There are several.
Most central banks have a program of publications. Australia’s central bank has been publishing the Bulletin every month since 1937. 2 In most cases central banks publish the results of research of their staff. The views in these papers are those of the authors alone, but much of the research is part of the knowledge base available to the policy-makers and hence is usually of interest to those seeking an understanding of policy issues. The RBA has been publishing Research Discussion Papers since the late 1960s.
Over the past 15 years or so, many central banks have upgraded their regular economic reporting, giving a more in‑depth account of economic conditions with a more analytical flavour. In some countries with formal inflation targets, these are called ‘inflation reports’, though they are about much more than inflation. These documents set out the factual background in a way which shows how policy‑makers account for economic and financial conditions in their decisions. In the process they usually reveal a good deal about how the central bank thinks the economy works – its ‘model’, if you will.
In many, probably most, cases, these documents contain forward‑looking material. Forecasts for inflation (and often for other key macroeconomic variables), the assumptions on which the forecasts are based and the extent of uncertainty surrounding the forecasts are, to varying degrees, spelled out. A few central banks even publish a future interest rate path – albeit one heavily conditioned by assumptions which are almost certain not to be realised, for one reason or another.
The RBA has progressively upgraded its own regular report. The Statement on Monetary Policy appears in the mid month of each quarter. This is a very comprehensive treatment of the local and global economies, financial markets and considerations for monetary policy. It contains the Bank’s inflation forecast, a sense of the risks surrounding the forecast, and a discussion of the forces conditioning the outlook for the economy. It also offers an account of the policy decisions the Board has made in the preceding period.
Central bankers give public speeches on issues of the day. In our case, we typically will take questions from the floor on such occasions as well. Sometimes the questions are even about the topic of the speech! The speeches, and the answers to questions, are routinely sifted by economists, the media and others for hints about the central bank’s intentions. That is not surprising, though sometimes readers are remarkably inventive in their efforts to read between the lines. I have certainly marvelled at some accounts of what I am supposed to have said. Nonetheless, these occasions do give the central bank the opportunity to talk about the issues it thinks are important, and to signal, if necessary, any changes to its view that might occur between the formal assessments of the economy appearing on the regular timetable.
Central bankers make appearances in front of Parliamentary committees. As you may know, in our case, the Governor has been appearing twice yearly in front of the House of Representatives Economics Committee for about a decade now. Other officials of the Bank appear in front of various other committees as appropriate. These hearings give elected representatives the chance to ask questions at length. They are a key part of accountability and offer a useful opportunity for communication. They can also play, if used well, an educative role, developing a better understanding of policy issues than would be achievable in many other fora. Of course, there is always the risk that in such sessions the questioning can tend to aim more at political point‑scoring.
Public communication is, however, a two‑edged sword. For every occasion when there is something important to say, there is another at which a central banker finds him or herself giving a speech, or releasing a formal document, which has little new to say on the economy, in an environment in which markets and observers already have a good understanding of the central bank’s assessment. On those occasions there is always the risk that further communication will inadvertently dislodge perfectly sensible expectations – which is one reason why speeches and documents are often, quite deliberately, a little on the unexciting side. Colour and movement are not necessarily useful in the central banker’s case.
It is for this reason that the RBA was, for a long time, somewhat ambivalent about the practice adopted by some other central banks of making a statement of reasons for the policy decision even when the decision is to leave rates unchanged. While a decision to change rates has for many years been accompanied by a pretty detailed explanation in our case, a decision not to change rates often meant that we had little new information to impart. We adhered to the old adage “when you have nothing to say, keep your mouth shut”.
These reservations were valid. But an increasing number of other central banks have managed to construct these statements and issue them regularly, without apparently doing much harm. More importantly, while there are plenty of occasions in which no change in interest rates is widely expected, and hence perhaps needs little explanation, there are others in which a no‑change decision does require careful explanation.
Having reflected on this for some time this year, the Reserve Bank Board came to the judgement that the downside risks of such statements no longer outweigh the likely benefits. Accordingly, we have adopted the practice of issuing a short statement after every meeting, explaining the policy decision, whether or not the cash rate is to change. The first ‘no-change’ statement was issued last week.
Last week’s decision was a good example, in fact, of one where no change to the cash rate needed explanation. The statement noted the concern the Board had about the outlook for inflation, given the recent price data and the apparent strength in demand. It also noted that the outlook for the world economy looked a little weaker, and that trends in financial market pricing, over the preceding month, were likely to result in a rise in borrowing rates for Australians. In other words, financial conditions were shifting in the right direction for containing inflation, even without a further adjustment in the cash rate. Hence the Board decided to leave the cash rate unchanged, for the time being.
In a further change to communication arrangements, commencing at the February 2008 meeting, our statement will be made on the day of the meeting, at 2.30 pm, rather than the next morning. Any change in the cash rate will still take effect the following day. This schedule will inform the markets during the Australian day, but will limit the time period between the decision being taken and it being publicly announced. The previous practice of delaying the announcement until 9.30 am the following day was originally adopted for logistical reasons, but they have long since disappeared. It is much better practice, and less risky, to announce the decision as promptly as possible once it has been made.
Minutes and the Limits to Transparency
That brings me to the one remaining area of central bank accountability and communication that I would like to talk about, which is the treatment of the minutes of policy‑making meetings.
Practices vary in this area. The United States Federal Open Market Committee has been publishing minutes for many years. It was prompted initially by Congressional pressure for more openness, but before long moved beyond minimum requirements to a fairly full set of minutes including voting records of individual members. The Bank of England’s Monetary Policy Committee (MPC) is required by statute to publish its minutes. The MPC’s culture is expressly, by intention of its creators, one of individual accountability. Hence the minutes include voting records of the members, and it is not uncommon for a significant proportion of members to differ from the majority – including two cases where the Governor was in the minority. MPC members hold a second meeting, subsequent to the policy meeting at which the interest rate decision is taken, for the purpose of approving the minutes, which are then released prior to the next policy meeting.
The European Central Bank, in contrast, does not publish minutes. One of the key reasons is that the presidents/governors of national central banks sit on the Governing Council of the ECB but are required to make decisions for the euro area as a whole, as opposed to decisions that might suit the particular circumstances of their own countries. It is argued, not unreasonably, that publication of minutes and voting might prejudice the capacity of the national governors to take a euro area, rather than national, perspective. The ECB does make additional efforts at communication of other kinds, including a regular press conference.
Clearly there are views on both sides of this question, which reflect the different institutional arrangements across countries. This is why we have argued over time that, in the pursuit of the optimal degree of transparency, it would not make sense to ‘cherry pick’ the high transparency aspects of every other system and assume that they should simply be grafted onto the Australian system. The nature of the Reserve Bank Board – a majority of whom are part-time members, drawn from various parts of the Australian community, but seeking to make decisions in the national interest as opposed to any industry, geographical or sectional interest – needs to be considered when thinking about disclosure practices.
It is also important, I think, to articulate the point that there are reasonable limits to transparency in any system. It is not the case that releasing more material is always, by definition, going to lead to better‑informed public debate or better policy decisions.
In contemplating the release of minutes of meetings in particular, and the form any such release might take, we need to balance a legitimate desire for information and accountability against the need to maintain a frank, open discussion at the meeting. At many meetings I have taken part in over the years, people have considered various arguments that ultimately were found to be unconvincing, but which did need to be examined in the interests of reaching a balanced decision. People also change their minds in the course of a meeting – and one would hope that that happens occasionally, since one of the key benefits of having a meeting is to learn from and respond to the views of the other participants. It is unlikely that these dynamics of a meeting could be sustained if every utterance were disclosed. The incentive could easily arise for people to be much more cautious in what they said, and to come to the meeting with a pre-written statement, rather than to engage in a genuinely interactive discussion. That would reduce the quality of the discussion and, ultimately, of the decision. And it is the quality of the decision, after all, which we should be seeking to maximise.
What all of this means is that a decision to release minutes should not be taken lightly. Such documents have to be written carefully, taking into account the institutional structure, including the nature of the Board, and the need to preserve an environment of candid discussion.
That said, there is no reason why, with careful drafting, a set of minutes that strikes the right balance cannot be compiled. Indeed, for a while now, we have been writing the minutes in just that way. After discussion among the Board members, we recently decided that there is no longer a strong case for not being prepared to release minutes of the monetary policy discussion. Accordingly, as announced last week, we will in future release the minutes of the monetary policy meetings with a lag of two weeks. The minutes from the November meeting were released last Wednesday. The minutes from the meeting held last week will be released on December 18. In addition, with the agreement of the Board, I am releasing at this time on the website the minutes from all meetings since I have been the chairman of the Board (that is, commencing with the October 2006 meeting)
Those who are interested to read these documents – which will, outside the media and economic and financial professionals, be relatively few, I expect – will find the following features.
First, there is an account of the main factual material available to the Board and the issues arising from that material that came up during their discussion. There is nothing particularly startling there – the information available to the Bank is pretty much known by everyone else. The material does not cover every single indicator the Bank tracks – there are too many.So if your favourite statistic is not mentioned, that doesn’t mean we are ignoring it. Equally, those statistics that have been prominent in recent sets of minutes should not necessarily be seen as all-important for future decisions. The Board will always strive to form a comprehensive picture of the whole economy in assessing the economic outlook and the prospects for achieving the inflation objective, and in coming to its decision.
Second, there is an account of the policy discussion that occurs towards the end of the meeting, which outlines the key considerations involved in the decision. Sometimes these considerations are quite straightforward. On other occasions, they may point in different directions in terms of their implications for interest rates, in which case the Board has to make an ‘on balance decision’. The minutes will set out as clearly as we can the logic behind the decision.
Third, there is a record of what the decision was – that is, what target cash rate the Bank is to maintain in the period until the next meeting
Readers will note that comments are not attributed to individuals. The material is not a transcript – indeed we do not keep a transcript – and it is not an edited version of some other set of more detailed minutes. No other record of the monetary policy discussion exists. The minutes do not attempt to provide a ‘blow-by-blow’ description of every comment made. But nor do the minutes released by other central banks, and it would not be sensible to do so, for the reasons I articulated a few moments ago.
Readers will also observe that the pattern of votes of individuals is not recorded, only the outcome. That is a point of difference with other central banks which publish minutes. But in those cases the decision-makers are full-time appointees, in some cases in systems with expressly individual, as opposed to collective, responsibility for their decisions. That is not the system Australia operates, and our pattern of disclosure reflects the institutional arrangements.
In the interests of clarity, let me also state that these minutes do not cover issues other than monetary policy. Other matters that the Board considers from time to time – such as the Bank’s accounts, audit processes, issues concerning subsidiaries and other governance questions – are not recorded in these published minutes, because there is no public policy case to do so. There is proper disclosure on these matters, but through the appropriate vehicle, which is mainly the annual report. This is in line with practice in other central banks, and with common sense.
Our view is that minutes of the type we are now releasing, in combination with the regular statement after each meeting, and the large volume of other material released by the Bank, meets the legitimate claim of observers to know the basis of the Board’s policy decisions. It is consistent with arrangements that prevail in the countries to which we would look for examples of good practice. At the same time, this approach should preserve the candour of discussion at the meeting and recognises and respects the basis on which the non‑executive members of the Board serve. As such, it strikes the right balance.
Communication has become a more important part of the central banker’s tool kit. While we will rarely be found courting publicity, neither will we shirk the responsibility to explain what we are doing and why. That is a requirement of good governance, but also it will usually make policy decisions more effective.
There are limits to transparency. More is not always better, and because the decision to disclose additional information is hard to reverse once made, it should be made with care. Nonetheless, after reflecting on our own experience and evaluating experience around the world, we recently judged that the time had come to move Australia’s arrangements to conform with normal practice elsewhere. I was very pleased to learn when I met the new Treasurer a couple of weeks ago that he supported the changes.
This material will not compete too well with the best‑seller lists, and almost everyone will (I hope) have better things to read while on the beach over the summer than the Reserve Bank Board’s minutes. But for the professionals, this will hopefully make a modest further contribution to their understanding of the Board’s decisions. And for any of the rest of you who are having trouble getting to sleep …
Whatever your preferred reading material, I wish you a pleasant festive season and a happy and prosperous New Year.
- They are also expected to respond to more media questioning than their counterparts in other democracies, as far as I can see.
- In Australia, it was a recommendation of the 1935–1937 Royal Commission that the Commonwealth Bank publish a regular statistical bulletin. The Reserve Bank took over this obligation after ‘separation’ in 1960.