Independence And Responsibility: Observations From An External Member Of The Bank Of England’s Monetary Policy Committee - Speech By Martin Weale At The Thomas Guggenheim Conference On 28 June 2015

Introduction

Thank you very much for inviting me here today to talk about my experience as a member of the Bank of England’s Monetary Policy Committee. I have now been an external member of the Committee for nearly five years, and my second and final three-year term finishes at the end of July 2016. So, while these are in no sense valedictory remarks I have had the privilege of observing the workings of the Committee for quite a long period.

External members of the Monetary Policy Committee are appointed by the Chancellor of the Exchequer from outside the Bank. They are there to provide an independent voice, and the expectation of independence is set out explicitly from the very beginning: the current advertisement states “candidates must be independently minded... [and] must form their own views and decisions”.1 So the focus of this talk is on independence, and what it has meant in practice.

The independence of the Monetary Policy Committee has two dimensions. The first is that the entire Committee, and process of setting monetary policy, is independent from the government. This is a corner-stone of credible policy making, which ensures that the public have faith that it is free of political interference. Secondly, individual members must themselves be independent, and articulate their own interpretations and assessment of the current economic situation and its implications. They should not limit themselves to expressing a single MPC view.

Taking the Committee as a whole, its collective independence was enshrined in the monetary policy framework set up in 1997 by the newly elected Labour government. In a break from the previous regime, monetary policy is set by a body of nine members, five of whom are drawn from the Bank’s executive – so-called internal members – and four are external members.

The Committee has complete discretion over decisions on monetary policy. But its primary objective – an inflation target of two per cent – is set by the Chancellor of the Exchequer. The Committee’s members are answerable to the Treasury Committee of the House of Commons – and through that, to Parliament – but the decisions on policy themselves are free of political control. The framework therefore combines the benefits of independent decisions, with overall accountability to the public.

The Treasury Committee holds hearings into appointments and re-appointments to the Committee, but it does not have the power to veto them. In contrast to what may be practice elsewhere, it does not enquire into political leanings but expresses a view on whether appointees meet criteria for professional competence and personal independence.

In the report of the hearing after my own appointment the Treasury Committee (2010) stated

“Given that votes on the Monetary Policy Committee are conducted under a one-person,
one-vote system, for the Committee as a whole to be considered independent, its members must be seen to be independent too. Independence is especially important given the high-profile nature of the role, where members of the Monetary Policy Committee are expected, via speeches and interviews with the press, to ensure that the process of monetary policy is accountable to the wider public.”

So I would like to spend much of my time reflecting on independence and its implications as I have seen them. I would also like to say something about transparency. Following a report by Warsh (2014) the Committee is making substantial changes to its practices, with the aim of enhancing transparency. I will offer my observations on these, based on what I have seen so far.

I should remind you that, just as there is no such thing as objective history, so here I am not offering any objective account of the workings of the Committee. I am doing no more than providing my own interpretation and explanation of the way I have seen my role in the light of the expectations created by the quotations above.

Voting and deliberation: the individual and the collective

It is generally accepted that monetary policy decisions reached by committees through voting are, on average, better than decisions reached by individuals (Siebert, 2006, Lombardelli, Proudman and Talbot, 2005). Nevertheless, different committees work in very different ways.2 For example, it is rare for those members of the Federal Reserve Board who are not chairmen of regional reserve banks – whom I might see as the equivalent of their external members – to vote against the Board Chairman. In contrast, members of the Monetary Policy Committee frequently vote in a minority.3

There is of course a substantial body of literature on the characteristics of effective monetary policy committees (summarised by Maier, 2010) and the external membership of the Committee is generally regarded as delivering an important component of independent thought. What I have read misses, however, the importance of the way in which the role is defined, ultimately by Parliament for the conduct of committee members. The extracts in the introduction make clear what is expected in the United Kingdom and the appointments process, handled by the Civil Service since 2008, has been designed to ensure that external members can be appointed from as wide a field as possible in order to deliver a balanced committee. For example, not all of the members have been British subjects.4

In the early days of the Committee there was some press concern that the absence of unanimity would indicate some sort of failure of the policy-making process. The first minority vote came, however, without disaster. And the sky remained resolutely in place when the previous Governor, Mervyn King, was himself out-voted. There are, nevertheless, long periods of unanimity which have, indeed, led some City economists to ask what the point of the Committee is, if there is no dissent. Even the Treasury Committee has been known to ask the Governor why dissent has been limited.

Of course the answer is that there is no such thing as too little or too much dissent. The right way for each member to vote is the way that they think appropriate at the time and that in one sense is the end of the matter. While I have voted in a minority at a number of the meetings I have been to, I have not doubted that my colleagues’ votes were as correct as my own. The future is, by its nature, uncertain and it does not surprise me in the least that nine people whose job it is to come to their own conclusions about economic prospects do not always come to the same conclusions.

The fact that we vote individually does not, of course, mean that members lose sight of the need to make the Committee work sensibly as a committee both internally and publicly. The full sentence from the job advertisement I quoted earlier reads “Candidates must be able to form their own views and decisions whilst working closely and constructively with other members of the Committee”. This means more than just listening to each other’s views; it means trying to understand what lies behind different interpretations of the facts. Warsh (2014, p. 22) describes working closely and constructively as deliberation; it is something that a committee can offer in a way that atomised individuals cannot. I agree completely with the point he makes

“Genuine deliberation is therefore the process by which participants not only share information but also learn from and influence each other. It is the crux of good decision-making processes within both the public and private spheres.”

and would add two points. First, I have found this process of deliberation, learning from my colleagues, one of the most fascinating and enjoyable parts of my job. Secondly, while it is helpful to refer to the first phase of the Committee’s meeting as a deliberation meeting deliberation involves much more than a Committee discussion for three hours in each meeting cycle.5 It summarises the whole process whereby we study, discuss, and interpret briefing material kindly provided by the Bank staff and how, each quarter, we produce a forecast which represents the best collective judgement of the Committee. A particularly valuable component of deliberation is that members read and comment on each other’s speeches before they become public. I hardly need add that that process has invariably helped me express my own arguments.

Barabas (2004) argues that deliberation is a route to a deliberative consensus and suggests that if a consensus is not reached deliberators will just talk past each other. My own experience is that, at least in the context of the Monetary Policy Committee, deliberation is extremely valuable even when a clear range of opinion remains.

Perhaps I can illustrate that there is more to working together than just deliberation, with reference to the situation I found myself in in autumn 2013. I had voted against the specific proposals for interest rate forward guidance (Monetary Policy Committee, 2013) on the grounds that, given the recent history of inflation at that time, they ran a substantial risk of weakening the credibility of the Committee’s commitment to delivering the inflation target. Having, however, been out-voted by eight votes to one, it seemed to make no sense for me to continue as though the policy did not exist. Both at the time and with the benefit of hindsight, the right thing was for me to set my subsequent contributions in the context of the guidance framework that everyone else was using. This did, however, lead to some journalists observing, at least privately, that I had voted against the policy because of some technical flaw rather than because I had a material and fundamental difficulty with it.

Could I conclude this section on voting with a stray thought?6 I have mentioned the inevitable uncertainty which surrounds decision-making. I have often thought that if each member could cast a hundred votes instead of just one, I could represent, at least to some degree, the uncertainty I see about the appropriate policy. If the Committee were to vote in this way, with each member representing the situation as they actually see it, the outcome would, on occasion, be different from that emerging from conventional majority voting.

To give an example, in December 2014 I voted for an increase in Bank Rate. With a hundred votes, I might have cast fifty-two for an increase in Bank Rate, forty-six to keep Bank Rate unchanged and two for a reduction in Bank Rate to ¼ per cent. This small vote for a reduction would have reflected the probability I gave to the economy lapsing into a deflationary spiral. By January 2014, the price of oil had fallen further, and I felt both that the case for an increase was less pressing and that the risk of a deflationary spiral had increased somewhat. So I might have cast forty-seven votes for an increase, forty-eight for no change and five for a reduction. This was reflected by the use of the term “finely balanced” in the minutes of the January meeting.

Would this work as a general approach to voting? If some members cast most or all their votes behind one decision, perhaps out of a belief that being decisive was more important than trying to come to a sensible collective decision, then other members would find it difficult to follow the spirit of this approach. If, however, everyone voted to represent what they actually thought, my hunch is that it would yield, on occasion, different and better decisions.7

Who does what? The scope and function of monetary policy

Let me now turn to an aspect of independence which, I think it is fair to say, could not have been anticipated. In 1997, there was a straightforward distinction between the monetary policy decisions made by the MPC, and their subsequent implementation by Bank of England staff. There was only one policy instrument in active use, Bank Rate, and the operational side of policy – collectively called the Sterling Market Framework – was the responsibility of the Bank’s executive.

The financial crisis of 2008-09 shook things up, as further monetary stimulus was warranted even once Bank Rate reached ½ per cent, an effective floor at the time. The response of the MPC – before I joined – was to provide ‘unconventional’ monetary stimulus, mainly in the form of asset purchases. This preserved the same distinction between decisions and implementation as before: the MPC voted on the total size of asset purchases, and the Bank oversaw the purchases of government stock that the Committee had voted for.

Where I felt there was tension, though, was in the provision of further monetary stimulus in 2012. In an attempt to encourage banks to increase their net lending to the economy, the Bank and HM Treasury set up the Funding for Lending Scheme (FLS) in 2012. This guaranteed funding on particularly favourable terms for banks that did increase their net lending.

It was, in my view, a means of easing the monetary stance, and therefore regular use of such a tool was a decision about which the Monetary Policy Committee had a role (Weale, 2012); at least it seemed to me that Parliament would have asked for this had it anticipated the circumstances. This highlights the uncertainty around a suitable demarcation of monetary policy since the crisis. Should the FLS, or similar schemes, be thought of as monetary policy-making, or simply part of the central bank’s function in looking after the banking system?

We have been fortunate that another independent review, led by Bill Winters, has since set out the problem clearly. Winters (2012) commented that “it would seem appropriate that the MPC should not only be informed, but should also have an opportunity to give a view” on any liquidity operations undertaken by the Bank.8 This has been implemented in a subsequent Concordat between the Bank and the MPC (Bank of England, 2013), in which any prospective changes to the Sterling Monetary Framework must be discussed with the whole MPC. The Concordat states:

“In the event that some key parameters of the Sterling Monetary Framework are varied with the primary intent of affecting monetary conditions, then their control will pass to the MPC.”

Steps like these, which ensure that the Committee is included in discussions and decisions about the monetary stance, help enshrine the overall independence of the monetary policy arrangements.

Independence and communication

The role of communication as a means of making monetary policy effective has been widely discussed (see e.g. Woodford, 2005). External members are expected to “make a strong contribution to the MPC’s communications, explaining policy decisions to the public”. I, and I think all members, have interpreted this as meaning that we have to explain our own positions and our own individual votes by means of newspaper articles, interviews and speeches, statements for which texts are publicly available. The need is particularly acute when policy changes. It is at least as strong when a member starts voting in a minority or, equally reverts to voting with the majority.

Members, and particularly those in a minority, have a special obligation to explain their positions, not as certainties, which they never can be, but as their own balanced judgements put in the context of the Committee’s best collective judgement. The duty of members to express their own views clearly in public while at the same time maintaining the collegiate nature of the Committee creates a challenge which I have not faced on other decision-making bodies I have contributed to. My sense is that the Treasury Committee would be concerned if we did not seek to explain changes in our votes, and thus in our underlying analysis in this way.

Obviously a framework in which individual members are required to express their own views means that there cannot always be a single message emerging from the Committee. Blinder (2007) drew attention to the risk that this would mean that there was no clear message; one of the purposes of communication would be frustrated by a multiplicity of views. Dale, Orphanides and Österholm (2008) highlighted the costs of uncertain information; conflicting views cannot contribute to overall clarity. Siklos (2003), on the other hand, found that volatility in interest rates was lower when the minutes of the Committee indicated disagreement than when they showed unanimity. That said, this seems to me only a weak indicator of whether a multiplicity of voices sows confusion.

In any case, there is a separate point which Blinder (2007) does not discuss. Communication of the range of views on the Committee serves a separate purpose of explaining to the public some of the uncertainties which affect the economy, and reassuring them that the Committee is doing its job thoroughly. In turn this can sustain public confidence that inflation will be kept close to its target.

That process is, of course, helped if members whose views diverge from the collective judgement of the Committee take care to explain, not just what their views are, but how and why they differ from that judgement. Let me give one example of this. In early 2011 I was voting with two colleagues for an interest rate rise because I was concerned that inflation was likely to remain above the target for more than the next two years; I feared that persistent high inflation would raise inflation expectations appreciably. While this never became the majority view, it is possible that the credibility of the inflation target in the public mind was enhanced by public awareness of the debate on the Committee. If true, this would have helped bring inflation back towards its target.

Independence, transparency and accountability

If communication is important as a tool of effective policy-making, transparency is even more important. Without transparency the Committee’s processes and therefore the Committee itself would not maintain, or indeed deserve, the confidence of Parliament.

The Committee also has to maintain the confidence of the public; if there were strong and sustained public dissatisfaction with its decisions, then it seems likely that this would result in Parliament taking back control of monetary policy. Public confidence requires a degree of public scrutiny of the way in which we go about our business.

The quid pro quo of independence is accountability and thus transparency. The public is entitled to be assured that the Committee is going about its work properly. The individual members have to be accountable for the decisions that they take. I have already explained that public explanation of the factors behind our votes is an important part of this accountability. It also means, however, that the processes of the Committee must be open to scrutiny and review.

In its early days the Committee was one of the most transparent bodies making monetary policy. It was one of the first to publish minutes, which it did initially after about five weeks and, from October 1998, onwards, less than two weeks after the decision. The quarterly Inflation Report contained the Committee’s forecast, represented as a best collective judgement of the different views of the members. The only variables for which it contained explicit projections were, however, inflation and GDP growth. This led to frequent criticism from readers that they could not judge what the MPC expected of other macro-economic variables and therefore how it expected the economy to evolve.

The Stockton Report (Stockton, 2012), a review of the Monetary Policy Committee’s forecasting capability, suggested that transparency could be enhanced if more information were provided on what lay behind the Committee’s forecasts. We therefore have started to publish, in the Inflation Report a table of key judgements, which describe the shape of the forecast in a bit more detail. The Bank also publishes, on its web site, indicative projections of the components of demand together with projections for household income, employment, and average weekly earnings. These are consistent with the MPC’s modal forecast but are not themselves agreed by the Committee. They nevertheless provide a useful reference point not only for outsiders but also for MPC members who wish to discuss, say, prospects for household spending.

At the same time the Committee has started to provide more information about how it sees the recent past and very near future. Thus we show our nowcasts and nearcasts of unemployment, GDP growth and inflation, together with indicators of their accuracy. My sense is that these changes have been a success, and users of our forecasts have welcomed the extra information.

There have recently been further changes to the working of the Committee. When I joined the Committee I was told that meetings were recorded but that the recording was destroyed once the minutes had been prepared. The point of the recording was to assist in producing accurate minutes and there had been a small number of occasions when the secretariat had listened to the recording to this end. The Treasury Committee expressed the view in early 2014 that a transcript of the meetings should be made public after some suitable interval.

The Governor then commissioned Kevin Warsh, a former Governor (i.e. member) of the Federal Reserve Board, to review the matter. He took the view that to release a transcript of the deliberation meeting, at which we test out each other’s arguments, would be likely to inhibit debate. On the other hand there were good reasons to release transcripts of the discussion meetings at which we explain our views and why we vote the way we do. The Treasury Committee accepted this and, since March of this year transcripts of the deliberation meeting have been kept to be released after a delay of eight years.

One consequence of releasing transcripts of the discussion meeting is that I now stick to a prepared text, which summarises the key points influencing my decision. My own contributions have become more stilted as a result. But I hope I am adjusting to the new procedure. More importantly, I find it hard to believe that this makes our monetary policy decisions worse. The broader benefits of releasing the transcripts in terms of sustaining public confidence in the process far outweigh any theoretical cost. Given that the transcript is a summary of what is likely to become public fairly soon, I myself favoured release much earlier than eight years. However, it is also the case that much prompter release might influence the way in which members present their views.

Warsh’s review was, however, more wide ranging. In fact it, together with Stockton’s earlier report, amounted in large part to a board efficiency review. Warsh also suggested that the Committee met too frequently and that we release a fuller statement immediately after our meeting.

The Bank’s response was to propose moving to eight MPC meetings a year, from the current twelve, a change in which I can see considerable merit. The Government announced last month that it will be proposing legislation to put this into place. Given the relatively slow pace with which the macroeconomy usually evolves, this change will give both staff and Committee members more time to work on underlying issues.

In addition, the MPC agreed to publish the minutes of its meetings alongside the announcement of the decisions themselves. There are a number of gains from this new procedure. One is that there is only one news event rather than a ‘drip feed’ of releases, between the announcement of the decision, the subsequent publication of minutes, and on four occasions a year, the additional publication of the Inflation Report. Some commentators have expressed concern about a deluge of information, but I doubt that the public wants to hear from the Bank every day, and there is nothing wrong with journalists taking their time to absorb and interpret the documents that we produce. Overall, I think the change will give our communications more impact.

It does, of course, necessitate some changes to the timing of MPC meetings, to accommodate the process of writing and agreeing the minutes. The process starts, as it always has done, with a briefing from Bank staff on a range of indicators about the state of the economy in the United Kingdom and the rest of the world. This now usually takes place on a Wednesday. The next day the Committee has its deliberation meeting, which is not recorded. On the following Monday, the first of the transcribed meetings takes place, in which each member explains how they expect to vote, and why. The actual vote, however, does not take place until Wednesday, with the results and minutes of the meeting to be published the day after that.9

This does carry the risk that “events” or at least material new data may emerge in between meetings, although I do not expect that to happen very often. If that does occur, then the drafting skills of the Committee will be put to the test because the minutes will have to be prepared more quickly. In any case, what cost this does bring should be set against the benefits of transparency and immediacy. And once we move to eight meetings the risk of them coinciding with the release of key data will be lower.

Conclusions

Although I have covered only some of the issues related to the work of the MPC, I hope I have offered some personal insights into its workings, and have given some indication of what it is trying to achieve.

I mentioned early on in this text that none of the literature I have read on the importance of the Committee structure and the role of external members draws attention to the way in which their function is defined and the implications of this for the effectiveness of policy. For me this, as summarised by the advertisement for the post, sets the standard for the way in which I have attempted to fulfil the role.

It will be some time before the arrangements that I have described here match the Gold Standard in terms of durability. That, after all, lasted from 1717 to 1931.10 The runner-up, the fixed exchange rate system, lasted from 1944 to 1972, so that post-War record will soon be in the Committee’s sights. I believe that the policy of targeting inflation with the Monetary Policy Committee having operational independence has, however, proved its worth in the very difficult circumstances of the last years. As I hope I have demonstrated, while monetary policy-making has been subject to necessary innovation, the fundamental principles of independence and individual responsibility have been maintained, and I am confident that these will continue to serve the country well.

¹ See Cabinet Office (2015).² Warsh (2014, p. 26) summarises the structures of decision-making at ten central banks. In six of these policy is the outcome of a vote, although, as a comparison between the Federal Reserve Board and the Bank of England shows, voting means different things in different places.
³ There have been minority votes at about half of the meetings.
⁴ Kristin Forbes and Adam Posen are American, while the current Governor, Mark Carney, is Canadian. There has not, as yet, been a member who has spent most of their career in continental Europe.
⁵ The original format of MPC meetings comprised two parts: first, what, following Warsh, I refer to as a deliberation meeting. Second, a discussion and voting meeting at which members of the Committee explained the rationale behind their votes, and voted. These meetings took place on a Wednesday afternoon and a Thursday morning. A pre-MPC meeting at which staff presented their views and analysis took place on the Friday before. I discuss the current format later.
⁶ Derived from observations by Charlie Bean.
⁷ The approach would mean that the Committee based its decision on a collective view of the uncertainty, rather than as a result of individual members coming to a decision based on their own assessments of uncertainty.
⁸ Winters (2012), paragraph 417.
⁹ In June and July 2015 the publication of the minutes has followed its previous schedule. That will change in August.
¹⁰ Although it was suspended from 1797 to 1821, and gold exports were banned from 1919 to 1925.