The UK MPC process in the light of the Warsh Review

Charles Goodhart

02 March 2015

a

A

The Warsh Review (Warsh 2014) is leading to some sweeping changes in the procedures of the UK’s Monetary Policy Committee (MPC). The number of meetings will be reduced in future to eight per year. In my view, there might have been a case for a more radical reduction to six per year. The timing of the MPC meetings has been altered in order to allow for a simultaneous provision of the decision, the supporting minutes and, in the relevant months, the Inflation Report, plus associated press conference, all on the same single day. While this change purports to have been driven by concerns for ‘transparency’, I shall argue that its purpose is otherwise, both to reduce the number of annual monetary ‘news’ events and to lay the ground for a possible future change in the format for setting out the future policy rate path in the Inflation Forecast. Finally, and for current market analytical purposes the least important, transcripts will be made of the second session (Day 2), in which MPC members indicate their individual policy preferences, for release in eight years’ time, with the actual unvarnished recording to be released in 20 years’ time.

I shall review these changes in ascending order of importance:

  1. Recording of Day 2.
  2. Frequency of meetings.
  3. Timing of meetings.

Recording of meetings

An equivalent of the Heisenberg effect in the social sciences is that the manner in which a Committee meeting is recorded affects the way in which that Committee operates. The subsequent provision of a (nearly exact) transcript of what was said will make proceedings more formal and careful, often with members reading a prepared statement into the Record, and less interactive, inquisitive, and open-minded.

Clearly, if there were to be only one meeting, the conflict between full and wide-ranging deliberation, which must require an open-minded search for the truth, and decision, where the rationale for that decision must be made clear, would be acute. Fortunately the MPC has two meetings: a deliberation meeting, Day 1, and a decisive, proposal meeting, Day 2, on which each member sets out his/her preferences for the policy rate. So Warsh can suggest, and the Bank agree, that Day 1 (the deliberation meeting) should remain, as it is already, unrecorded, though the arguments and analysis will be contained in the Minutes. In contrast Warsh, and the Bank, propose that Day 2 (for advocacy and decision) will be fully recorded, with the transcripts, and ultimately the basic recording itself, released.

Warsh is right to note that the character and purpose of these two meetings differ. But, in my view, he underestimates what may be lost by such future publication of the records. With the exception of much of the recent period at which the policy rate has been stuck at the zero lower bound (ZLB), the decision whether to alter policy rates by a small amount (e.g. 25 basis points), or not, was usually finely balanced. Quite a number of MPC members did in previous years enter the Day 2 meeting without having their mind made up in advance, and, if called upon early to speak, would ask to delay their own vote until they could hear others’ arguments. Could such apparent lack of conviction survive future publication? Would not publication force inherent uncertainty into polarised positions? One prior criticism of reporting the names of voters was that it made changing one’s mind difficult. Will not full subsequent reporting make that even more difficult?

But MPC members are rarely shrinking violets, and the eight-year lag is long enough to enable members to focus more on the job at hand, and less on their own personal future image. There is a valid argument that more could, and should, be done to reveal how monetary decisions were taken. Naturally the objective and aim of those outside the MPC room is to get more information about what happened inside to be revealed. In this compromise recommendation, Warsh, and the Bank, will be providing (eventually) relatively more information at the expense of relatively little cost (but not zero cost) in the shape of a less interactive, less interesting, and less balanced Day 2 meeting.

Frequency of meetings

The Bank of England Act 1998 states that the MPC shall meet every month, and they have done so ever since they started in 1997. Perhaps one, or two, say in September and January, of such 12 meetings could have been perfunctory, ‘hello and goodbye’, but in practice it was not played that way. The resulting burden, especially on the staff, and the high frequency of such meetings, have been widely felt to be excessive. Warsh and the Bank have taken the opportunity to reduce the number of meetings to eight: four Inflation Report meetings, as at present, and four mid-course opportunities for a correction. This puts the MPC onto the same timetable as the FOMC.

There is a problem with this latter change. The date of the MPC meeting within each month was carefully chosen at the outset to coincide with a contemporaneous dearth of major UK news events, so that the MPC’s deliberations would be less likely to be shocked by some sudden new statistical revelation. But, if the mid-term MPC is to be halfway between Inflation Reports, it will come in a news-heavy period. For example, on the Bank of England’s illustrative timetable for 2016, some of the MPC decision days will now be mid-month, in fact most of those that are not Inflation Report meetings will be mid-month. In this case, the decision meeting is more likely to take place in the same week as key releases such as CPI inflation and the labour market report.

However, this is less of an issue now than it would have been at the outset of the MPC. The contrast between news-light weeks at the start of the month and news-heavy weeks in the middle of the month is now less marked. This reflects the rise in importance of the PMI surveys in assessing the contemporaneous health of the UK economy.

There is no mention in the Warsh Review of the desirable frequency of forecasts and of their associated Inflation Reports. Since the underlying data are quarterly, the implicit assumption is that the forecasts (IRs) should be also, and again done at a common time within each quarter. But it does not have to be that way. One could have three forecasts each year, say February, June, and October, with mid-term corrections in April, August, and January.

If the intention is to reduce the number of meetings (and of monetary news events) then is not six better than eight? If we move down from 12 per year, then either (some of) the monthly meetings or (some of) the forecast (IR) meetings have to move away from their preferred calendar synchronisation. Which might be better? There was no discussion in either the Warsh Review or in the Bank’s response of this, or of any other, alternative to the eight-MPC, four-IR calendar framework (as practiced in the US). Did they miss a trick?

The timing of meetings

It takes a considerable time to provide an agreed and accurate report of the wide-ranging deliberations of a sizeable, vocal, and deeply intelligent Committee, and of the varied reasons subsequently put forward for the resultant decision. Each member will be jealous to ensure that the nuances of his/her position will be accurately and fully recorded. The difficulties of composing a brief paragraph after the current Thursday meeting have been such that when no change has been made, the absence of a requirement to make any statement on the day is greeted with enormous relief by the staff and by (some) MPC members. Outsiders, notably journalists who often have to react instantaneously to news, do not generally recognise the time-consuming and difficult exercise of putting together official minutes.

This means that there are two alternatives. Option 1, the present procedure, is to hold the deliberations and the decision-making together, and then immediately to announce that decision. This implies that (full) minutes cannot be released until after an elapse of time, of nearly two weeks. Option 2 is to hold the deliberations and (most of) the decision-making well before announcing that decision. The differences between the two timings are set out in Table A2 of the Bank of England’s response to Warsh (Bank of England 2014), reproduced as Table 1 below. Note that the deliberation meeting has been pushed back by over a week and the ‘decision’ meeting by three days, Thursday till Monday. Of course, the final so-called decision meeting will not be until Wednesday, but with all MPC member Monday positions known, recorded, and subsequently to be published, the outcome is effectively settled for almost all practical purposes. The Wednesday meeting will generally be pro forma and perfunctory.

Table 1. Current and new MPC meeting schedules

Source: Bank of England.

In my view this has almost nothing to do with transparency. Which do you regard as more transparent?

  1. Take a decision, and announce it immediately, leaving a (full) explanation for later;
  2. Take a decision, but leave the decision unannounced until it can be accompanied by a full explanation.

People may take different views on this, but some of us would regard (1) as more transparent than (2). If you should accept that view, the outcome of such an exercise about transparency would have been to shift to a less transparent process.

Be that as it may, the prospective longer-drawn out process has two inherent dangers. The first, and less worrisome, is of a leak. All those closely involved will know, as of the end of the Monday meeting, what the Thursday announcement will (or is likely to) be. But central bank members exhibit loyalty and care, and central banks that have run such an Option 2 system (such as the Riksbank) have never been troubled by leaks.

The second concern is that the longer lags before the announcement both of the deliberation, Day 1 (seven days rather than one), and of the policy discussion, Day 2 (four days rather than one hour), leave much room for a news shock, which could make the prior deliberation/policy discussion look inappropriate. We can all think of examples, e.g. Lehman Bros, 9/11, October 1987. Then would the most sensible thing be to change the decision and scrap the accompanying minutes? Possibly a more difficult problem would be when there is a large, but less obviously, dramatic news intervention that puts the prior deliberations/policy discussion in some question, but without giving a patent hook for a complete review.

So, Option 2, the proposed new schedule is not self-evidently more transparent, and has some inherent drawbacks, though many would see them as minor. So what are its benefits? The first, with which I would agree, is that the present system has too many monetary news events, 28 in all: 12 announcements, 12 minutes, and 4 IRs plus press conference. Financial markets spend too much of their time in a tizzy about the next such news event. Option 2 cuts down such events to eight per year. That is surely a much better balance for monetary policy within the wider scheme of things. This is the main current benefit of the proposed change.

Warsh offers as a subsidiary benefit that it shortens the ‘purdah’ period in which MPC members are constrained from expressing views on policy. Thus he writes, p. 35,

“Under current practice, members often find themselves in the uncomfortable position of necessarily obfuscating their views in public in the two-week gap between the policy meeting and the publication of the minutes. This also leads to a blizzard of communications from MPC members in the short period between the release of the minutes and the ‘purdah’ period for the next policy meeting that begins after the Friday ‘pre-MPC’ briefing for the Committee by Bank Staff. This period is frequently as compressed as ten days or so. The interests of effective communication are not well served by a flurry of speeches and op-ed articles from different MPC members with different perspectives in such a compressed period.”

While I see the point, in my view this is an over-statement. MPC members give speeches rather rarely; not so much a ‘blizzard’ as an ‘occasional snow shower’. Moreover, while the ‘purdah’ period may be shortened, its intensity will be enhanced. Between Monday (Day 2) and the Thursday announcement, all MPC members, and accompanying staff, will be sitting on market-moving information. Under the old Option 1, until the Minutes were published, they were only sitting on their own idiosyncratic views, which is not the same thing at all.

Finally, the old procedure would have made it very hard, if not impossible, to incorporate a fully worked-out path for future policy rates, beyond the immediate decision, determined by the MPC. Since such a path would have to be incorporated into the IR, it would seemingly have to involve a wider interaction between the staff forecast team and the MPC than occurs at present. Currently the staff extract a path from the money market yield curve, and an (unforeseen by the market) change to the current rate, decided by the MPC, can be super-imposed if need be on the market path by a simple and quick rule of thumb.

Option 2, however, with a rather longer prior process could be extended further, if so desired, quite easily to shift the policy rate future path to one selected by the MPC, in place of one derived from market expectations. There are many, though not myself, who would welcome such a second step. But no doubt the Bank will seek to try out how this first step is working before deciding how, and whether, to move further to this second step.

Disclaimer: Charles Goodhart is an Emeritus Professor at LSE and Senior Economic Consultant to Morgan Stanley. This article is not an offer to buy or sell any security/instruments or to participate in a trading strategy. For important current disclosures that pertain to Morgan Stanley, please refer to the disclosures regarding the issuer(s) that are the subject of this article on Morgan Stanley’s disclosure website. https://www.morganstanley.com/researchdisclosures.

References

Bank of England (2014), “Transparency and Accountability at the Bank of England”, 11 December.

Warsh, K (2014), “Transparency and the Bank of England’s Monetary Policy Committee”, December.

a

A

Topics:  Monetary policy

Tags:  monetary policy, Bank of England, transparency, Central Banks, central bank transparency, Monetary Policy Committee, inflation, news shocks, leaks, noise

Emeritus Professor in the Financial Markets Group, London School of Economics

Events