The Reserve Bank's submission to the Minister of Finance on the Monetary Policy Review conducted by Professor Lars Svensson


1) In my memo to you dated 23 February 2001, I undertook that the Reserve Bank would report back to you before the end of March on our reactions to the conclusions and recommendations in Professor Svensson's Independent Review of the Operation of Monetary Policy in New Zealand (hereinafter referred to as the Review). This memo sets out those reactions.

2) The key consideration which motivates our reactions is the desirability of maintaining and enhancing the credibility and integrity of the structure created by the Reserve Bank of New Zealand Act 1989 (Reserve Bank Act), so that that structure can be durable through both economic and political cycles, and can give us the best chance of securing top quality outcomes. One of the important aspects of that structure is the independence of the Bank to operate monetary policy aimed at the goal established in general terms by legislation, and in more specific terms in the agreement with the Minister of Finance made under that legislation.

Professor Svensson's overall assessment

3) Professor Svensson's key conclusions on the conduct of policy were as follows:

· "The Reserve Bank's current conduct of monetary policy in New Zealand is entirely consistent with the best international practice of flexible inflation targeting, with a medium-term inflation target that avoids unnecessary variability in output, interest rates and the exchange rate." (Review, p.3)
· "During the 1990s, the Reserve Bank has achieved a remarkable stabilization of inflation at a low level and successfully anchored inflation expectations on the inflation target." (p.2)
· "In gathering, processing and analysing information, considering alternative scenarios, constructing projections, and making informed decisions, (the Bank) is second to none and hence compares well with the Bank of England and the (Swedish central bank)." (p.40)
· "The Reserve Bank is currently using the appropriate instrument for the operation of monetary policy". (p.4)
· "The Bank uses available data in the same way and to the same extent as other best practice central banks". (p.4)
· "No changes are required in the current coordination of monetary and fiscal policy." (p.7)
· "The Bank's communication of monetary policy decisions to the public and financial markets is generally exemplary". (p.7)
· Having heard a number of New Zealand critics argue that the Bank was "out of touch with reality", or "an ivory tower filled with boffins", or "mechanistic in its reliance on the output-gap model", he specifically rejected those criticisms. (pp.40-43)

4) Professor Svensson also reviewed quite carefully the operation of monetary policy during the 1990s.

· He agreed with our own submission that the Bank had been somewhat slow to tighten policy in the 1992-93 period, although concluded that this mistake "was not unreasonable considering the circumstances". (p.26)
· Reviewing policy in 1995 and 1996, he concluded that "given the circumstances, policy may have been either appropriate or somewhat on the tight side, but not substantially too tight". (p.32)
· However, he was strongly critical of the Bank's use of a Monetary Conditions Index between mid-1997 and March 1999, and described that period as "a substantial deviation from best-practice inflation targeting." (p.3)
· He concluded his assessment of the 1990s by saying that

"some mistakes have been made in the past, but perfection is an inappropriate bench-mark. The Reserve Bank has been a pioneer in inflation targeting, without the possibility of learning from previous experience by other central banks. The economy has regularly been subject to new kinds of shocks. Mistakes in such a situation are to a large extent unavoidable, and they need to be seen in a broader perspective. The inflation history of New Zealand before the 1990s is dismal. During the 1990s the Reserve Bank achieved a remarkable stabilization of inflation at a low level. With hindsight, the timing of policy could have been somewhat better. However, given the circumstances at the time of decisions, aside from the MCI period there is no evidence that policy has systematically resulted in unnecessary variability in output, interest rates and the exchange rate. The period of using the MCI to implement monetary policy is the exception..." (p.36)

5) Except for his criticism of our use of a Monetary Conditions Index between mid-1997 and March 1999, Professor Svensson's assessment is very positive. We ourselves acknowledged in our own submission to Professor Svensson that our use of the Monetary Conditions Index was inappropriate, and as you know we have not used such an index in implementing policy for two years.

6) The terms of reference under which Professor Svensson was working did not require him to comment on whether the Reserve Bank's single focus on delivering price stability, as mandated by Section 8 of the Reserve Bank Act, was appropriate, but he nevertheless made it clear that he regarded the focus on price stability as being entirely consistent with theory and international experience. He noted that

"people typically ask too much of monetary policy - no less in New Zealand than elsewhere. In the long term, monetary policy can only control nominal variables such as inflation and the nominal exchange rate. It is beyond the capacity of any central bank to increase the average level or the growth rate of real variables such as GDP and employment, or to affect the average level of the real exchange rate. At best monetary policy can reduce the variability of these real variables somewhat. An attempt to increase the average level or growth rate of GDP or employment would trigger ever-rising inflation, at increasing cost to the economy in terms of less efficient resource allocation and arbitrary and inequitable redistribution of incomes and assets." (p.1)

We fully agree with this assessment of the role of monetary policy.

Professor Svensson's recommendations on the Policy Targets Agreement

7) Turning to some of the specifics, Professor Svensson recommends that

"(t)he current formulation of the policy target in the PTA, including the clause `to avoid unnecessary instability in output, interest rates and the exchange rate', should continue to be interpreted as a medium-term point target for 12-monthly increases in the Consumer Price Index (CPI) of 1.5 percent, the midpoint of the target range of 0 to 3 percent. The target should also continue to be interpreted as `soft-edged', in the sense that the edges 0 and 3 percent do not indicate any hard and precise thresholds for policy responses. This is consistent with the best international practice of flexible inflation targeting and with avoiding unnecessary instability in output, interest rates and the exchange rate. It is also entirely consistent with the Reserve Bank's current interpretation of the PTA. At the beginning of the next term of the Governor, section 2b should be modified to read `....the policy target shall be 12-monthly increases in the CPI of 1.5 percent over the medium term'. There is no reason to make any other changes to the PTA." (p.64)

8) There are really two aspects of this recommendation - the first, a recommendation to treat the inflation target as a medium-term target, with 0 and 3 per cent being treated as `soft-edges'; and the second, a recommendation to `anchor' the target on its mid-point of 1.5 per cent. As Professor Svensson says, his recommendation to treat the inflation target as a medium-term target with soft edges is entirely consistent with the way in which the PTA is currently being interpreted, and consistent with the delivery of on-going price stability while avoiding unnecessary instability in output, interest rates and the exchange rate.

9) Specifying the target as 12-monthly increases in the CPI of 1.5 per cent, to be achieved over the medium term, might well have the benefit which Professor Svensson claims, of anchoring inflation expectations more effectively at the centre of the range than the present 0 to 3 per cent band. Certainly it would avoid one aspect of the present situation, where some people assume we should be relaxed about any inflation outcome, irrespective of how close to the edges of the band it is. On the other hand, it might be interpreted by some as a de facto tightening of the inflation target. On balance, we do not feel strongly about the matter, although suspect there might be some net benefit in the move. We note that there is no need to resolve the issue immediately since Professor Svensson recommends that the change not be introduced until "the beginning of the next term of the Governor."

10) Professor Svensson also recommends that

"(t)o maintain (the) stability and credibility of the monetary policy regime, it is desirable that each PTA is maintained for the full five-year duration of the Governor's term. At the formation of a new government, the existing PTA should normally be left in place. Any change to the PTA should be done at the beginning of a new term for the Governor." (p.64)

11) We understand the concern expressed by Professor Svensson on this point, and note that, only 11 years after the introduction of the current legislative framework, we are operating under our sixth PTA. Had each PTA run for the full term of a Governor, we would have been operating under our third. While there are clearly some benefits in having a new government formally specify how it wants the central bank to understand the statutory objective of price stability, there are also some risks in having the PTA changed too frequently. With the PTA having been honed for 11 years now, with refinements made by governments from both sides of the House, there should be a greater prospect that future PTAs can run for the full term of a Governor, or at least that there is no presumption of change with every change in government.

Professor Svensson's recommendation on data availability

12) Professor Svensson concludes that the

"availability and quality of data relevant for monetary policy are largely in place. However, some improvements are required to bring data quality up to international standards. In particular, Statistics New Zealand should collect monthly data on the Consumer Price Index and industrial production." (p.65)

13) Central banks and other users of statistical data will always want data which is more reliable and more timely, and the Reserve Bank is no exception. Our submission to Professor Svensson identified as priority areas for improved statistics:

· better income data (particularly income GDP, which currently is not available at all on a quarterly basis);
· price level data that is more integrated and consistent across the various categories (eg import prices, producer prices, and retail prices); and
· improvements in the quality of some existing data (eg import prices, the terms of trade, and the GDP deflator).

Thus, we agree with Professor Svensson's general observation that there are areas where New Zealand economic data need improvement to reach international standards. But his suggested priorities do not accord with those we had identified ourselves. We note that Statistics New Zealand is currently proposing to construct a manufacturing production index, which will go part way towards meeting the need for an industrial production index identified by Professor Svensson.

Professor Svensson's recommendations on improving FPS, measuring inflation expectations, communicating uncertainty, and reporting on the financial system

14) Professor Svensson makes four recommendations of a relatively technical nature. First, he recommends that the Bank
"may want to consider some further developments of its Forecasting and Policy System (FPS). Alternative interest rate reaction functions and alternative interest rate paths could be used and presented systematically to the MPC to provide a larger menu of policy choices for discussion and consideration. Although these are not likely to materially affect policy choices, they represent developments that will improve the internal consistency of the FPS." (p.66)

Professor Svensson notes that adopting this recommendation would probably not make a material difference to policy outcomes, and "would push the frontier of best international practice in inflation targeting further out." (p.45) We agree with this, and hope to adopt this recommendation as we further develop our current technology. Having said that, our current emphasis is on diversifying the approaches we take to analysis and forecasting to avoid the dangers in having just a single view of the economy, an emphasis that we believe will help increase the robustness of our policy judgements.

15) Secondly, Professor Svensson recommends that the Bank

"consider reporting and discussing alternative measures of inflation expectations for the medium and long term more extensively than is currently the case, as part of assessing the long-term credibility of the monetary policy regime." (p.66)

This seems entirely sensible. He suggests that some of the data might be commissioned from an arm's-length source, such as Statistics New Zealand, and this also seems sensible. We will follow-up this proposal.

16) Thirdly, Professor Svensson recommends that the Bank

"consider alternative ways of conveying the inherent uncertainty in projections while maintaining transparency, for instance, modelled on those used by the Bank of England and Sveriges Riksbank (the Swedish central bank)." (p.66)

In explaining how he came to this recommendation, Professor Svensson criticised the Bank's move to round the numbers in its projections to the nearest half per cent, rather than to one decimal place. We are certainly very conscious of the importance of communicating uncertainty and conditionality, and indeed the move to round the numbers was an attempt to avoid the appearance of spurious precision in our projections. It is a bit rough and ready, to be sure, and we are looking for alternative ways of dealing with this issue. Those suggested by Professor Svensson are certainly becoming increasingly commonplace internationally, but so far only within the context of projections that presume no change in policy even as inflation surprises are revealed. We will think about how well these and other techniques might translate into our context, where we assume that policy may change over time.

17) Fourthly, Professor Svensson recommends that the Bank

"summarise its information about the financial system, including a number of macro-prudential indicators of financial stability, in a regular report, modelled on those published by the Bank of England and Sveriges Riksbank". (p.66)

We see some merit in this proposal, and note that the Bank already publishes a regular article in our quarterly Bulletin on the state of the financial sector, and has recently established a small group to develop macro-prudential indicators for New Zealand. We envisage publishing the work of this macro-prudential group, and will evaluate the costs and benefits of pulling this material into a stand-alone publication, as compared to publishing in the Bulletin.

Professor Svensson's recommendations on governance and accountability

18) By far the most far-reaching of Professor Svensson's recommendations are designed to deal with what he describes as "weaknesses" in governance and accountability structures. He recommends in particular that monetary policy decision-making be vested not in the Governor, as currently, but in a committee of five, and I will deal with that recommendation shortly. He also makes four other recommendations aimed to strengthen the weaknesses he sees in governance and accountability, specifically that:

· "The Board of Directors should only consist of non-executive directors. To ensure sufficient independence of the Board, the chair of the Board should be selected by the non-executive directors among themselves and not by the Treasurer.
· "The Board of Directors should publish an annual report with an evaluation of the Bank's monetary policy. This report could be separate or part of the Bank's Annual Report. The Board may appoint a panel of experts for its assistance.
· "The Reserve Bank should fund an annual conference for the evaluation of monetary policy in New Zealand. The conference should be organised by an independent committee.
· "Parliament's Finance and Expenditure Select Committee should, with the help of appointed experts and other assistance, conduct thorough and detailed hearings of the Governor and other Reserve Bank officials. This may require additional resources for the Committee. These arrangements could be modelled on those undertaken by the UK Parliament and the European Parliament." (p.65)

19) On Professor Svensson's recommendation that the Bank's Board should consist of non-executive directors only, we certainly agree that the Governor should not chair the Board. Having the Governor chair the Board is, of course, almost universal practice in central banks all over the world, but is inappropriate in a situation where the primary functions of the Board of our central bank relate to the appointment of the Governor and to monitoring the Governor's performance. We have dealt with that in the past by creating a Non-Executive Directors' Committee of the Board, to which the full Board has delegated its monitoring function, but, although it would require minor legislative change, we believe it would be better to remove the Governor from the chair of the Board. We also agree with Professor Svensson that, to protect the independence of the Bank, the chair of the Board should be selected by, and be one of, the non-executive directors of the Board.

20) There are clearly pros and cons of having the Governor on the Board. Removing the Governor from the Board would have the advantage of making the monitoring role of the Board "cleaner". Indeed, retaining the Governor on the Board might suggest the need to enshrine the Non-Executive Directors' Committee of the Board in the statute, so that the monitoring role of the Board can be clearly separated from the Governor. But, on balance, we believe that it would be a mistake to remove the Governor from the Board. It would be quite unusual in the corporate world to remove the chief executive from the board, even though corporate boards are normally heavily involved in monitoring the performance of the chief executive, and having the Governor on the Board does, we believe, materially facilitate communication between Board and Governor.

21) It may be marginally less important to have the Deputy Governors on the Board, but I would nevertheless argue strongly for their remaining members of the Board. Certainly, if I were a non-executive director of the Bank, I would regard having the Deputy Governors on the Board as very desirable:

· Having Deputy Governors on the Board is a very important protection against a Governor who might be tempted to "filter" the information provided to the Board. In other words, having the Deputy Governors on the Board provides an opportunity for the provision of contestable advice and comment from two other insiders.
· One of the issues in assessing the Governor's performance is an assessment of his relationship with his immediate colleagues, and that is materially assisted by observing the interaction of Governor and Deputy Governors at Board meetings.
· Non-executive directors are responsible for recommending to the Minister the person to be appointed Governor. Among the potential appointees will often be a Deputy Governor, and in that regard having the ability to observe the performance of a Deputy Governor at Board meetings is useful.
· At the margin, being able to offer a potential appointee to the position of Deputy Governor a position on the Board may make it a little easier to attract top quality candidates.

The literature on corporate governance usually argues that it is important that non-executive directors are in the majority on company boards, but since the Reserve Bank Act already provides for between four and seven non-executive directors, retaining the Governor and up to two Deputy Governors on the Board would not compromise that majority.

22) We can see considerable merit in the non-executive directors of the Board publishing an annual assessment of the conduct of monetary policy, as recommended by Professor Svensson. Of course, the non-executive directors regularly make such an assessment now, because the Reserve Bank Act directs the Board to "keep under constant review the performance of the Governor in ensuring that the Bank achieves the policy targets agreed to with the Minister" (Sec. 53(c)). But having that assessment published, probably as part of the Bank's Annual Report as suggested by Professor Svensson, would have the advantage that it would be seen by the general public and the markets. Indeed, if Professor Svensson's recommendation that the Chairman of the Board should be a non-executive director rather than the Governor is adopted, the inclusion of a Chairman's report in the Bank's Annual Report would make eminently good sense, probably covering not only an assessment of the way in which monetary policy has been conducted but also an assessment of other aspects of the Bank's operations, all of which the Board is charged with keeping under "constant review".

23) We believe there is also merit in Professor Svensson's suggestion that the Bank fund a conference from time to time to assess the Bank's monetary policy performance. Professor Svensson suggests that this be done annually. This may be too frequent, but the principle is a good one. Perhaps funding a conference every second year would be sufficiently frequent.

24) We believe there would be merit in assisting the Finance and Expenditure Select Committee of Parliament to critique the Bank's performance in a more thorough way, as Professor Svensson suggests. That might require the provision of some additional resources to the Committee, perhaps enabling the Committee to commission an occasional assessment of the Bank's performance by an independent expert chosen by the Committee for that purpose, but that is a matter for the Government rather than for the Bank.

Single decision-maker or committee?

25) Arguably the most important of Professor Svensson's recommendations is to change the decision-making structure from the end of my current term as Governor from one where all monetary policy decisions are vested in the Governor to one where those decisions are vested in a committee, preferably one made up of the Governor, two Deputy Governors, and two other senior Reserve Bank staff. His argument is essentially that vesting the power to make all monetary policy decisions in a single individual runs a very considerable risk that monetary policy might "run off the rails".

26) He notes that the present decision-making structure in the Reserve Bank of New Zealand is unusual internationally in vesting all monetary policy decision-making in a single individual. While he says that the current arrangement "works very well", this "is to a large extent because of the exceptional qualities of the current Governor". (p.4) He feels that the structure is risky because "future Governors may not be of the same standing". (pp.4-5) In particular,

"another Governor may not, to the same extent, encourage open and comprehensive discussion and advice within the Bank and support the Board in its monitoring of the Bank. Another Governor may not cope as well with the pressure, criticism and even abuse that seem to go with the territory, and may, in difficult times and under high pressure, lose confidence and let policy go awry". (p.5) He feels that the present system "relies on the ability to quickly dismiss the Governor on such occasions. However, the lags in the effects of monetary policy, the difficulties in objectively identifying whether outcomes are the result of policy or luck, and the actual mechanics of removing a Governor together imply that removal of a Governor will never be easy and never be quick." (p. 5)

27) To deal with this problem, Professor Svensson recommends that

"a formal Monetary Policy Committee of the Reserve Bank, responsible for monetary policy decisions, should be formed. A suitable time for this is at the beginning of the next term of the Governor. The committee should have five members, each with one vote: the Governor as chair (with the casting vote), the two Deputy Governors, and two other senior Bank staff. The Governor and the Deputy Governors should be appointed in the same way as now. The two senior staff should be appointed in the same way as the Deputy Governors, that is, by the Board of Directors on the Governor's recommendation. The members of the committee should have overlapping five-year terms, so only one member is appointed or re-appointed each year. Named votes and non-attributed minutes of the MPC should be published with a lag of about two weeks. The MPC should only be responsible for decisions related to monetary policy. In all other respects, the Governor should continue to be the single decision-maker of the Bank." (pp.64-65)

Professor Svensson notes that such a committee would essentially institutionalise the present "well-tested" practice in the Bank, and so would involve the minimum change to present structures.

28) Professor Svensson gave careful consideration to other committee structures, particularly those involving members from outside the Bank. After rejecting the various arguments which are sometimes advanced for including external members on monetary policy decision-making committees, he concluded that, on balance,

"there are some quite practical reasons for not including external members. One great disadvantage is the limited supply of competent experts without serious conflict of interest. This may be a problem even in a country as large as the UK, with a population some 15 times that of New Zealand. For New Zealand this problem may be overwhelming. Even if initial appointments would be good - some have suggested to me that there are a number of suitable candidates - with some turnover a drop over time in quality and competence is unavoidable. Another problem with external members is that there may be a conflict between responsibility to the committee and personal interests. For instance, temptations may arise to take advantage of the increased visibility to raise a personal profile, for instance, by publicizing dissenting views from the rest of the committee, perhaps as a way of furthering one's future career after a stint on the committee. Finally, having external committee members raises the issue of what amount of resources in terms of staff and assistance is required for the external members to be effective. Experience from the Bank of England also points to a tendency for part-time external members to become full-time internal members rather than external ones. Altogether, I believe it would be quite unsuitable for New Zealand to appoint external committee members." (pp.52-53)

29) If Government nevertheless felt obliged to have external members on the decision-making committee, Professor Svensson argued strongly that those members should be people with a strong understanding of monetary policy, or of macroeconomics more generally, rather than "non-experts", because

"a lack of expertise in monetary policy means reduced competence for independent assessment and capacity to participate in the high-level discussion required in efficient forward-looking inflation targeting. Non-experts therefore easily become dominated by committee members with more expertise." (p.51)

30) We accept that there are some arguments in favour of monetary policy decisions being made by a committee rather than by a single individual. Committees can potentially bring a wider range of perspectives to a decision than can a single individual, and may, as Professor Svensson suggests, have more "legitimacy", particularly in a democracy, than any single individual. There is the inherent risk, which Professor Svensson stresses, of placing considerable power in the hands of a single individual.

31) But we nevertheless believe that Professor Svensson substantially over-states the risk in the present structure. There are a number of powerful mechanisms that currently protect the Bank and the economy from a Governor "running off the rails".

· First, by statute the Governor is appointed by the Minister on the basis of a recommendation from the Board. In other words, the appointment process itself is one which lends itself to the careful selection of a suitable person, and protects against technically incompetent or politically-motivated appointees.
· Secondly, the statute under which the Bank operates mandates that monetary policy must be used to "achieve and maintain stability in the general level of prices", and that general objective is given considerable specificity by the statutory requirement for the Minister and Governor to agree on a Policy Targets Agreement. In other words, the Governor is not free to invent his or her own policy objectives.
· Thirdly, the primary function of the Board is to monitor the Governor's performance on behalf of the Minister, and by statute must meet at least 10 times annually.
· Fourthly, and arguably most important of all, the statute mandates a high level of transparency for the operation of monetary policy, including the publication of at least two Monetary Policy Statements annually. This not only assists the Board in its monitoring role, but also makes it possible for financial markets, financial commentators, and the general media to monitor the Governor closely. Were the Governor to depart in any material way from the intentions of the Policy Targets Agreement, this would very quickly be evident.

32) The establishment of any formal committee to make monetary policy decisions would, moreover, have several disadvantages.

· First, it would somewhat complicate the task of communicating the Bank's monetary policy message. There would be a need to resolve who "owned" the Bank's projections: the Governor, the majority of the committee, or perhaps just the staff of the Bank. It might be difficult to continue presenting projections incorporating monetary conditions which change over the course of the projection because of the difficulty of agreeing such a course among a committee of five. Market attention might focus on the balance of the votes on the committee, rather than the analysis justifying the monetary policy decision.
· Secondly, it might make it more difficult to hold anybody accountable for the inflation outcome. To be fair, it is not easy to hold anybody accountable now - there are always a lot of unpredictable and extraneous factors which explain a particular inflation outcome. But having a committee of five would clearly make accountability more complicated: if removing an errant Governor is seen as difficult, removing an errant committee would be almost impossible.
· And thirdly, it is perhaps worth noting that any move to a formal decision-making committee would involve quite a substantial re-write of the present legislation.

33) There are also particular problems with having external members on a decision-making committee, all of them mentioned by Professor Svensson - the difficulty of finding skilled people without conflicts of interest, the problem that part-time external people often tend to become "internal" over time, the danger that external people use membership of the committee to build their careers by dissenting from the majority decision, and the problem of allocating research resources among competing interests. It is surely significant that the great majority of central banks, in developed countries at least, do not include external members on their decision-making committees, although a number appoint external people "to the inside" as full-time committee members for varying terms of years.1

34) What about the possibility of adding some external members to the currently-internal committee which advises the Governor on monetary policy decisions? We believe that this option too would have some significant difficulties. If the part-time external members of a purely advisory committee were to be monetary policy specialists, their involvement would make most sense if they participated throughout the decision-making process, extending over several weeks each quarter. This would give rise to the same conflict of interest concerns as would be the case if the committee was a decision-making one.

35) If the part-time external members of a purely advisory committee were to be non-specialists - perhaps business people, trade unionists or farmers without formal training in macro-economics - there would be three significant problems:

· First, the external members would almost inevitably be dominated by the insider members of the advisory committee in what is, in part at least, a highly technical task.
· Secondly, most potential external members would still have conflict of interest problems unless their involvement in the advisory process were carefully circumscribed.
· Thirdly, unless appointment procedures were carefully prescribed in the statute to avoid it, there would be a significant risk that the appointment of external advisers would be perceived in financial markets as an attempt to reduce the Bank's operational independence and compromise its goal of price stability.

Against these negatives, it is hard to see what external "non-specialist" members of an advisory committee could bring to the process that we can not acquire in other ways. If appointed to ensure that the Bank is fully aware of what is going on in the wider economy, it would seem that they could add little in a situation where the Bank is, as Professor Svensson found, already very well informed. If there were a need to become still better informed, there would seem to be much more efficient ways of achieving that, with fewer risks involved.

36) Without any decision-making responsibility or accountability, moreover, external members, especially if non-specialists, would not have a strong incentive to tailor their advice to the achievement of the inflation target. Rather they would face some incentive to court popularity by routinely advising the need for easier policy, and announcing that that had been their advice. Precluding part-time external advisors from making public their advice seems unlikely to be a sustainable policy over time.

37) Accordingly, we believe that having an advisory committee with part-time external members on it has no clear advantages and a number of disadvantages. If it were decided that there should nevertheless be external members on the committee which advises the Governor on monetary policy decisions, however, it would be vitally important that appointment procedures were established which would avoid any perception that appointments could be used to undermine the Bank's operational independence. A perception in financial markets that the external advisors were appointed to try to sway the Governor to a less inflation-averse monetary policy would raise the risk premium on all New Zealand financial assets, not necessarily very obviously but with widespread and insidious effect. The best approach would be to have members of any committee, whether "internal members" or "part-time external members", appointed by the Board of the Bank on the recommendation of the Governor, as is now the case with the appointment of Deputy Governors. This "double veto arrangement", which is also similar to that now applying to the appointment of the Governor (who is appointed by the Minister on the recommendation of the Board), would protect the Bank from a perception of political interference on the one hand and from domination by the Governor on the other.

38) On balance, and in the light of the reduced accountability and loss of transparency in any committee decision-making structure, we strongly recommend the retention of the single-decision-maker structure. But we also propose that this decision-making structure should be supplemented in a number of ways, most of them recommended by Professor Svensson, to assist in monitoring the Governor's performance and to increase both the perception and the reality of the Bank's openness to the wider community. Specifically:

· The Bank will make the extent of its existing consultation with the community more visible by including a list of companies and organisations consulted prior to each Monetary Policy Statement in those Statements. (This was done for the first time in the Statement published on 14 March.) It may also be desirable to find other ways of making the extent of two-way communication between the Bank and the community more visible because, as Professor Svensson found,

"the Bank makes a huge effort in collecting information, data and informal views from a large number of businesses and organisations from all sectors of the economy, for instance by having its economists travelling and talking to people all over New Zealand. I doubt that there is a single business, organisation or authority that has better and more detailed knowledge about the state of the New Zealand economy." (p.40)

If among some people the Bank is seen as "aloof" and "out of touch with the real New Zealand", then this may be because the extent of this consultation is not widely known. As you know, we call on 40 to 50 businesses and business organisations during the course of preparing each of our Monetary Policy Statements, while my senior colleagues and I speak to literally scores of public meetings each year. Most of those speaking engagements involve the opportunity for audience feedback.

· The Bank will increase even further the extent of that consultation with the public and, perhaps more important, involve more of its senior management in the consultation.
· The (non-executive) Chairman of the Board should publish an annual assessment of the Bank's monetary policy on behalf of the non-executive directors of the Board.
· The Bank will fund a conference to evaluate monetary policy, to be held every second year.
· One additional possibility might be the adoption of a practice recently adopted by the Reserve Bank of South Africa, involving the convening of public monetary policy forums two or three times a year to provide a platform for open discussions on monetary policy with business, labour, and community groups. Although there are some risks in this approach (it is difficult to see many groups arguing for tighter monetary policy under almost any circumstances), the idea might at least warrant further exploration.


39) It is recommended that you:

a) Note that Professor Svensson found that the way in which monetary policy is now being implemented in New Zealand is "entirely consistent with the best international practice of flexible inflation targeting".

b) Note that Professor Svensson had no substantial criticism of the way in which monetary policy was formulated during the 1990s, with the exception of the Bank's use of a Monetary Conditions Index between the middle of 1997 and March 1999.

c) Note that the Reserve Bank agrees with Professor Svensson's suggestion to amend the PTA after my current term as Governor ends to make explicit what is now implicit, namely that the policy target should be treated as a medium-term target, with 0 and 3 per cent treated as `soft edges'.

d) Note that the Reserve Bank accepts that there might be some net benefit in specifying the PTA target as 12-monthly increases in the CPI of 1.5 per cent, but note also that there is no need to resolve this issue until the next PTA is signed.

e) Note that there are costs as well as benefits in having the Policy Targets Agreement changed whenever the government changes, but that, after 11 years of honing, there may now be a greater prospect that in future the PTA can run for the full term of a Governor.

f) Note that the Reserve Bank agrees with Professor Svensson that some improvements in New Zealand economic data are required to reach international standards, but does not regard his specific suggestion that there should be a monthly (rather than a quarterly) CPI as amongst the highest priorities.

g) Note that the Reserve Bank plans to adopt Professor Svensson's suggestions for improving its Forecasting and Policy System (FPS) model as our current forecasting technology develops.

h) Note that the Reserve Bank intends to implement Professor Svensson's recommendation to make greater use of measures of medium and long term inflation expectations.

i) Note that the Reserve Bank will carefully consider how to improve the way in which it communicates uncertainty and conditionality in its published projections, as recommended by Professor Svensson.

j) Note that the Reserve Bank intends to continue publishing material on the financial system, and will publish material on macro-prudential indicators, though whether in a stand-alone report as recommended by Professor Svensson or in its quarterly Bulletin has not yet been decided.

k) Note that the Reserve Bank agrees with the suggestion that the (non-executive) Chairman of the Bank's Board should publish, on behalf of the non-executive directors, an assessment of the Bank's monetary policy performance each year, probably in the Bank's Annual Report.

l) Note that the Reserve Bank agrees with Professor Svensson's suggestion that it should fund a conference to assist in the evaluation of monetary policy in New Zealand, although believes that holding such a conference every second year may be preferable to holding a conference every year.

m) Note that the Reserve Bank is supportive of the Finance and Expenditure Committee having additional resources to assist it to conduct a thorough examination of the Bank's monetary policy performance from time to time.

n) Agree that the Governor of the Reserve Bank should no longer chair the Bank's Board, with the chair instead elected by, and being one of, the non-executive directors.

o) Note that, if recommendation (n) is agreed, change would be required to Section 60 of the Reserve Bank Act.

p) Agree that the Governor and Deputy Governors of the Bank remain on the Board of the Bank, but that the appointment and monitoring responsibilities of the Board be formally vested in the non-executive directors.

q) Agree that the single decision-maker structure of the present Reserve Bank legislation be retained beyond the end of my current term as Governor.

r) Note that, if recommendation (q) is agreed, the Bank intends to improve the way in which the single decision-maker structure works as indicated in paragraph 38 above.

s) Note that, with the exception of recommendations (n) and (p), all other proposed actions can be undertaken within the existing provisions of the Reserve Bank Act.

t) Note that the Bank sees no clear advantages and a number of disadvantages in having external members on the Bank committee which advises the Governor on monetary policy decisions.

u) Note that, if recommendation (q) is not accepted and a move to a decision-making committee is preferred, or if instead Government determines that the committee which advises the Governor on monetary policy decisions should have some external advisors on it, the Bank believes that the crucial issue is that appointments to the committee be made in such a way that the Bank's operational independence is protected, preferably by providing that appointments to the committee be made by the Bank's Board on the recommendation of the Governor, as recommended by Professor Svensson.

v) Note that, if recommendation (q) is not accepted, substantial amendments to the Reserve Bank Act would be required.

1 The Federal Reserve System in the United States, the European Central Bank, the Swiss National Bank, the Swedish central bank and the Bank of Japan are among central banks which have only full-time members on their decision-making committees. Decision-making in the Bank of Canada is formally vested in the Governor, as in New Zealand. Only in the United Kingdom and Australia, among the central banks we have considered, do the decision-making committees include part-time external members.