Gill Marcus: The importance of central bank communication

Address by Ms Gill Marcus, Governor of the South African Reserve Bank, to the Central Banks Communicators Conference Dinner, South African Reserve Bank, Pretoria, 13 March 2014.
Central bank speech  | 
14 March 2014
PDF version
 |  6 pages

Good evening everyone. Welcome to the South African Reserve Bank. And to those of you who are from abroad, a warm welcome to South Africa. It is difficult to imagine that twenty years ago it was unlikely that a conference on communication would be held in a central bank. We have come a long way from the days when central banks were arcane institutions, shrouded in secrecy, with the markets left guessing as to what the policies actually were. Karl Brunner, the well-known monetary theorist, once referred to the impossibility of central bankers to articulate their insights in explicit and intelligible words and sentences, instead pretending this was part of an esoteric art. And when a US senator claimed that he understood a comment made by former US Fed Chairman Alan Greenspan, he replied: "If you understood what I said, I must have misspoke". I am sure that twenty years ago, central banks communications departments, if they existed at all, were small, and communication strategies were marginal to the overall strategy of the bank.

Central bank communication became increasingly important with the moves towards increased independence and evolving monetary policy frameworks, as well as because of the pivotal role that central banks played, and are still playing, during global financial crises. Communication is now at the heart of central banks, and communications strategies have, to some extent, become policy instruments. Central banks have to consider carefully not only what they communicate, but how they communicate. Central banks not only need to communicate their policy decisions and views about the future - which is always uncertain - but also have to communicate their expanding mandates, so that the public is aware of what a central bank can do, and what it cannot do.

This is of particular significance since the global financial crisis of 2007/8, when increasing responsibilities were placed on the shoulders of central banks. The apparent success in dealing with the crisis has led to unrealistic expectations about what a central bank can do, and it is important that we ensure that expectations are unreasonable. If not, when we do not live up to these unrealistic expectations, as will no doubt happen at some stage, we will lose credibility even in those areas that are within our appropriate policy purview.

The global trend towards independence means that central banks have to become ever more accountable. Accountability requires increased transparency, and its corollary, effective communication. But this has brought its own challenges of how to communicate and how much to communicate, and, as this conference has highlighted, practices differ around the world. More recently, the issue of accountability has been accentuated with the explicit addition of financial stability to central bank mandates in many countries, including South Africa. Macroprudential policies require more coordination with other agencies, and these policies have strong distributional effects, more so than monetary policy. This requires even more effective communication to gain societal legitimacy. Nothing can undermine independence more quickly than being opaque and unaccountable.

We at the South African Reserve Bank engage on an ongoing basis with the markets, and make ourselves available, where possible, to meet with a constant stream of domestic and foreign investors in the country. But communication is not just about our interaction with the markets. We have made a conscious effort to communicate with different stakeholder groups in the broader civil society through our outreach programme. We meet on a regular basis with political parties, trade unions and business associations from different sectors of the economy. We also convene an economists' roundtable every second month, where we meet with economists and analysts from various institutions, and discuss topical issues, including, but not exclusively, monetary policy. We find it a useful forum not only to express our own views but also to hear the views of market participants. We recognise the importance of knowing what the thinking about the Bank and our policies are.

Although we are not always successful in changing perceptions and misconceptions about the Bank and monetary policy, we have found that these interactions and initiatives have helped create a greater appreciation of the challenges we face, the reasons why we act in a particular way and has reduced antagonism towards the Bank, particularly when unpopular policies have to be implemented. Such meetings are usually led by me and include a deputy governor and senior staff, with counterparties being CEOs and senior executives. The discussions often last up to three hours, which demonstrates the seriousness with which we regard such interactions.

Other communication initiatives include the biennial release of our Monetary Policy Review and the Financial Stability Review; the Quarterly Bulletin; and for the past two years, we have conducted a monetary policy challenge, which is a competition for high school pupils. Coinciding with the release of the biannual Monetary Policy Review, the Bank organises Monetary Policy Forums at nine centres around the country, where members of the MPC and senior staff of the Bank engage with various stakeholders and interested members of the public. A press conference is held after each Monetary Policy Committee meeting, and is broadcast live on a number of TV stations, as well as a webcast. After the statement is read, the full MPC responds to questions from the press as well as other interested parties, who are welcome to send in questions electronically. The Governors and senior officials of the Bank are regularly invited to deliver speeches or participate in panel discussions. Over the past two years, the number of speaking engagements accepted by myself and the deputy governors has exceeded 50! Part of the challenge we face is striking the right balance between communicating effectively and over-exposure.

However, despite these communication initiatives, we have a long way to go in terms of reaching the broader public. In 2011 we commissioned a survey of the attitude of the public towards the Bank. The results of the survey were gratifying in that they indicated that the Bank had a high degree of credibility in the eyes of the markets. However, 45 per cent of South Africans were unaware of our very existence!

Perhaps the biggest communications challenge confronting central banks currently is the issue of forward guidance, or how to communicate monetary policy decisions. Twenty years ago, the FOMC did not announce its monetary policy decisions, or give any justification for it, and it was left to the markets to infer from the Fed's actions whether or not there had been a change of monetary policy. Gone are the days when central banks believed it was appropriate to surprise the markets. Today's conventional wisdom is that monetary policy predictability will reduce trade-offs, and reduce volatility in both inflation and real economic activity. In other words, communication has become a policy instrument in itself.

The more recent trend has been towards explicit forward guidance. This approach was started by the Reserve Bank of New Zealand, and later adopted by Norway and Sweden, where the expected path of policy interest rates were published. During the financial crisis, forward guidance became more widespread, particularly in those countries where the zero bound had been reached, and where conventional policy options became limited. Different types of forward guidance strategies were tried, ranging from providing a bias; a vague or explicit time path for future policy changes; to indicating state-contingent indicators for future monetary policy actions. For example, both the US Fed and the Bank of England specified unemployment thresholds, but generally subject to other objectives, such as inflation and financial stability considerations.

While there are good arguments for forward guidance, particularly during times of crisis, providing forward guidance to the markets has not been plain sailing, and has been challenged on both theoretical and practical grounds. As argued by Prof Marcel Fratzscher, formerly of the ECB and currently President of the German Institute for Economic Research in Berlin, there are major risks and costs from over-reliance on communication strategies. His concerns are three-fold: first, the dominance of the voice of central banks in financial markets has resulted in price movements becoming a reflection of responses to their statements and action, rather than to changing economic and financial realities. Second, central bank communications dominance could "crowd out" private sources of information or analysis, depriving the monetary authorities of an independent view of trends that sound policy-making would find helpful. And third, central banks risk losing credibility, when over-commitment to a specified path of action turns out to be a misleading assurance. He argues against central banks giving forward guidance, including announcements about when they will begin to tighten monetary policy and by how much. Rather, "they need to reintroduce true two-way risk, so that asset prices again reflect underlying fundamentals - Central bankers need to communicate more clearly how they think about risks and opportunities in the economy and financial markets, and then let private investors decide the balance of risks and reward for themselves."

Other critics have focused on the usefulness of forward guidance. Charles Goodhart, for example has been skeptical of its usefulness, particularly over longer horizons, where guidance is most needed. Specifying paths is only as good as the forecasts upon which they are based. His research shows that while central bank forecasts are very good up to two quarters ahead, they are "dire" at longer time horizons. Furthermore, he finds that existing empirical studies have failed to show that markets coordinate on the published forward guidance of central banks. His own research in fact found that in Sweden and Norway, the official path adjusts to the market rather than vice versa, except on short horizons in Sweden where a two-way relationship exists. Similarly, his review of state-contingent forward guidance concludes that while it has been largely successful in influencing the short end of the yield curve, they have no effect over longer horizons.

In general, where state contingent guidance has been given, it is clear that the conditionality relates to inflation and inflation expectations remaining under control. It is more difficult to give state contingent guidance if inflation is close to the upper end of the target range and the risks are to the upside (as is the case in South Africa currently). Furthermore, many of the criticisms of the Fed and the Bank of England in particular are that the conditionalities or "knockouts" are far too general, and are easily subject to change. It would appear that at times the markets will only be satisfied if a firm commitment is given, and this can clearly not be done. Forward guidance is always conditional, and is only valid as long as nothing unexpected happens. But it always will!

Fratzscher's view is consistent with our own approach to forward guidance: that we need to communicate the risks and opportunities in the economy. That means that we as central bankers have to act in a consistent way, in line with our analysis, but bearing in mind that our assessment of risks facing the economy may change, as circumstances change.

Our approach has been to give some form of qualitative guidance, where we try to indicate the nature of our reaction function, and the factors that are likely to cause us to change the policy stance. We do this through our monetary policy statements and speeches, where we identify and discuss the balance of risks around our central forecasts in such a way that market participants and other stakeholders are in a reasonable position to form a view about our possible reaction to changing circumstances.

In trying to give guidance to the market, we give our view of the outlook as well as our perceptions of the risks to the outlook. But this is not always straightforward. For example, any given inflation forecast could elicit different policy responses depending on the current and expected state of the economy, our perceived risks to the forecast, the degree to which inflation expectations are anchored, and how close we are to breaching the target band. So a forecast close to the upper end of the target range could result in a preemptive tightening if the economy is operating close to or above potential and the output gap is positive. But if the output gap is negative and growth is below potential, we are likely to be more tolerant of inflation being close to the upper end of the target.

However, periods of heightened uncertainty (such as we are currently experiencing) make our own forecasts more uncertain and subject to greater risk, and therefore the markets tend to be more uncertain and volatile. Under such circumstances the markets tend to demand more guidance, but this is precisely when it is more difficult to give. Our recent experience is a good illustration. During 2013, the rand depreciated significantly, in part in response to expectations of tapering by the US Fed. Given the importance of the exchange rate in the inflation process, our inflation forecasts deteriorated progressively, and the risks to these forecasts were seen to be on the upside. Our dilemma was that the domestic output gap was still negative, and other indicators of demand were relatively subdued. However, we increasingly signaled our concerns in the monetary policy statements, first by indicating explicitly that there was no room for further monetary accommodation; and then, in November, by stating that, should there be further significant deterioration in the inflation outlook, we would have to take appropriate action.

At that stage, given the turbulence in global financial markets, we were uncertain about what the actual outcomes would be. But the rand depreciated from around R10 against the US dollar in November, to over R11 against the US dollar by the January meeting, in response to Fed tapering which adversely affected capital flows to emerging markets. This, together with a reassessment of the possible path of food prices, led to a marked upward revision of our inflation forecast. The risks were assessed to be on the upside and inflation was expected to breach the upper end of the target range for an extended period.

This deterioration in the outlook led to an increase in the policy rate, an action that surprised the markets, despite our prior warnings. However, we felt that the move was appropriate, despite the fact that the markets in general were not expecting it. This raises the question of whether or not central banks should delay acting if circumstances change, in order to placate the markets. I would suggest not. While we do not want to surprise the markets, if we believe action to be appropriate, we need to act.

It is always more difficult to signal the beginning of a cycle. But, understandably, the markets now want clear guidance about the future. Our most recent statement in January stressed the fact that further action will be highly data dependent. This reflects the increasingly volatile and uncertain nature of the current global financial market environment. The initial reaction to our decision was quite extreme, with some expectations of an interest rate cycle of in excess of 400 basis points, although that has since moderated somewhat.

Some analysts have justified the initial market reaction on the basis of the previous tightening cycle in 2006-08 of 500 basis points in two years. However, the past is not always a good guide to the future. The previous tightening cycle was conducted in very different circumstances to what we are seeing today: at that stage, the South African economy was growing in excess of 5 per cent per annum, well ahead of potential; credit extension was growing at around 25 per cent; and household consumption expenditure growth was around 9 per cent. With rising inflation risks, this called for a strong monetary policy response.

Today we have a very different situation, with the output gap being negative and subdued growth in both household consumption expenditure and credit extension. We would therefore expect the monetary policy response to be more moderate, given our concerns about the slow growth in the economy. I therefore indicated recently that I thought that the market expectation of a further 200 basis points increase over the coming year sounded overdone, and I still believe this to be the case. However, we also stressed that future moves are highly data dependent in the current circumstances of heightened uncertainty. We have no history of tapering or normalisation to guide us. Of course, this moderate base case is highly conditional on developments and we cannot give any unconditional commitments of future policy moves. At best we can indicate our assessment of the risks, but it is for the markets to do their own assessments. It also does not mean that there will be adjustments at every meeting or that if rates are increased, that they will be increased by the same amount. And if circumstances change, we will have to change our views.

Other elements of communication of monetary policy relate to the publication of minutes and voting records. In contrast to a number of other central banks, we do not keep formal minutes. However, even the practice of releasing minutes differs quite widely. Some central banks, for example the Riksbank, release verbatim minutes about two weeks after the meeting. Others release unattributed minutes, where the flavour of the debate is given. This includes the Bank of England and the Reserve Bank of Australia. Our practice is similar to that of the ECB. We release a detailed statement on the same day of the conclusion of the MPC meeting at a press conference. The statement attempts to capture the main issues considered by the MPC, and the final paragraphs attempt to capture our perception of the risks to the outlook and conveys the stance of monetary policy.

Central banks also take differing approaches to the publication of voting records. In our case, we do not publicise the individual votes, as we do not want to personalise the debate. This gives individual members greater freedom to speak their minds and change their views during discussions, and it avoids the labeling of "hawks" and "doves". However, we do indicate if there were dissenting views, but we do not reveal who the dissenters were. It does reduce the transparency of the decision-making process, but this has to be weighed up against the advantages of allowing for more rigorous debate in the Committee.

While it could be argued that revealing differences of views could reduce the clarity of the policy message, we believe that there are clear advantages to doing so. In particular it shows that there is vigorous debate among committee members; that the MPC is not subject to group-think; and that differences of views are welcomed. However, to ensure clarity and consistency of our message, we have adopted the principle of committee responsibility. So even if the decision is not unanimous, once the decision is taken, all members of the committee take joint responsibility. This reduces the possibility of mixing the messages, as all members communicate with one voice.

In conclusion, communication of monetary policy decisions is still evolving. There is by no means universal acceptance of some of the recent developments globally, and there are criticisms that at times, forward guidance has introduced more volatility into the markets, and that too much guidance can encourage excessive risk-taking and delay monetary policy responses. It may also be the case that forward guidance is more difficult in emerging market economies that are more prone to volatility arising from exogenous shocks and spillover effects. Charles Goodhart has also argued that "pre-commitment, even if desirable, is not even really feasible, especially in a system of decision-making by committees, with changing membership and political oversight." We will continue to learn from what others are doing, from their successes and their mistakes. Our communication strategy has come a long way in the past few years, but it is evolving and we need to tread carefully.

Conferences such as these are important mechanisms for learning from each other, and are critical for the evolution and dissemination of ideas and experiences about what works and what does not work. As communication practitioners, I am sure that you have no shortage of topics to discuss tonight and tomorrow.

Thank you again for taking the time to travel to Pretoria to participate in this conference, and at the conclusion of your deliberations travel safely back home.

Thank you.


Fratzcher, M (2013): The Drawbacks of Forward Guidance.

Goodhart, C. 2013. Debating the Merits of Forward Guidance. In Forward Guidance. Perspectives from Central Bankers, Scholars and Market Participants, ed W. den Haan, eBook, Centre for Economic Policy Research.