This Selected Issues paper discusses the issues related to reforms and growth in New Zealand. The paper analyzes the record on growth and productivity outcomes in a comparative perspective. The study provides a brief history of the industrial relations in New Zealand leading up the passage of the employment contracts act. The paper assesses the monetary policy framework, central bank decision-making processes, and also reviews the possible extensions to full funding of the country's future superannuation expenditures.


This Selected Issues paper discusses the issues related to reforms and growth in New Zealand. The paper analyzes the record on growth and productivity outcomes in a comparative perspective. The study provides a brief history of the industrial relations in New Zealand leading up the passage of the employment contracts act. The paper assesses the monetary policy framework, central bank decision-making processes, and also reviews the possible extensions to full funding of the country's future superannuation expenditures.

IV. New Zealand: Monetary Policy Framework and Central Bank Decision-Making Processes

A. Introduction

93. As is well known, New Zealand was a pioneer in the adoption of inflation targeting as an approach to monetary policy. During the 1990s several other countries followed along this path and chose to conduct monetary policy around a publicly announced quantitative target for inflation. In the words of Lars Svensson (1999), therefore, “it is no surprise that Windy Wellington has become something of a Mecca for monetary economists.”

94. The move to formal inflation targeting followed the approval of the Reserve Bank Act in 1989 and aimed at delivering price stability in an economy that had experienced double digit inflation for most of the period since the first oil shock. Although the Reserve Bank of New Zealand (RBNZ) had succeeded in bringing inflation down to around 5 percent in 1990 from almost 17 percent in 1985, inflationary expectations were still deep-rooted in New Zealand and an institutional arrangement that would improve monetary policy credibility was perceived as essential (Archer and Nicholl, 1992).

95. While under inflation targeting New Zealand has managed to achieve and maintain low rates of inflation, monetary policy has sometimes been blamed for contributing to the economic instability experienced during the 1990s. Partly in response to these concerns, the modus operandi of monetary policy has been modified over time to make it more flexible. The latest Policy Target Agreement (PTA)—the contract between the Treasurer and the governor of the RBNZ that defines the operational aspects of inflation targeting in New Zealand—now explicitly states that in pursuing its price stability objective the RBNZ “shall seek to avoid unnecessary instability in output, interest rates and the exchange rate.”

96. Further, a review of the operations of monetary policy has been announced by the government with the aim of investigating “ways of enhancing the Reserve Bank’s ability to implement the PTA with minimal disruption to the economy.”30 A decade after the Reserve Bank Act, this review raises the opportunity to look back and assess New Zealand’s monetary policy, especially in view of the experience of the countries that adopted inflation targeting. As none of them has reproduced the New Zealand institutional monetary policy framework, it may be worthwhile to explore both the arguments in favor and against some possible changes in the current institutional settings. At the same time, it should be noted that’ the government remains strongly committed to maintaining the objective of price stability and the operational independence of the RBNZ, as these areas have been fenced off from the review (see box below).

Terms of Reference for the Review of the Operation of Monetary Policy

The review will consider:

1. The way in which monetary policy is managed in pursuit of the inflation target. The review will examine the way the Reserve Bank interprets and applies the inflation target set out in the Policy Targets Agreement, with a view to ensuring that this approach to achieving medium-term price stability is consistent with avoiding undesirable instability in output, interest rates and the exchange rate.

2. The instruments of monetary policy. The review will assess whether the Reserve Bank has an adequate range of instruments and is using its current instruments effectively in altering monetary conditions in the desired direction.

3. The information used by the Reserve Bank in its decision making. The review will consider the range of sources, availability, type and timeliness of data, and the impact of these variables on forecasting and decision making.

4. The monetary policy decision making process. The review will consider whether the decision making process and accountability structures promote the best outcomes possible.

5. The co-ordination of monetary policy with other elements of the economic policy framework, including an evaluation of the relationship between monetary policy operations and other Reserve Bank functions such as prudential oversight of financial institutions.

6. The communication of monetary policy. The Reserve Bank’s communication of monetary policy decisions will be reviewed to ensure that these decisions are explained to the public and financial markets in the simplest, clearest and most effective way possible.

97. This paper has three objectives. The first is to assess whether there is a problem of excessive economic volatility in New Zealand and, if so, how the RBNZ’s operational approach to inflation targeting deals with economic volatility. The second is to examine the recent change in the PTA and analyze its implications for the conduct of monetary policy and for the accountability and transparency of the current institutional monetary policy setup. The third is to focus on an element of the forthcoming review of monetary policy conditions, namely the decision making process and the accountability structure underlying monetary policy decisions. In particular, the paper looks at the following issue: should the responsibility for monetary policy in New Zealand be shared by a committee, as in most other countries?

98. The structure of the paper is as follows: the next section briefly discusses the current institutional monetary policy framework in New Zealand and presents some data on the volatility of output, inflation, real interest rates and the real exchange rate, before and after the adoption of inflation targeting. Section C analyzes the evolution of the inflation targeting in New Zealand toward a more flexible regime. Section D discusses a number of issues associated with the recent change in the PTA. Section E focuses on the choice between collective and individual responsibility for monetary policy. Section F contains concluding remarks.

B. Inflation Targeting in New Zealand: The Institutional Setup and an Assessment of Recent Experience

99. Although several other countries have followed New Zealand in adopting an inflation targeting framework for the conduct of monetary policy, the RBNZ differs from other central banks in two key respects:

  • the exclusive objective of the RBNZ is to maintain price stability. Of all the other inflation targeters, Sweden (from 1999) is the only other case where the legislation setting out the constitutional and legal basis for the central bank make no mention of secondary objectives such as employment, growth or output (Table IV.1).31

  • while several other countries share the same type of instrument-independence (e.g., England, Canada, and the US), in no other country does the formal responsibility of monetary policy rest solely in the hands of the governor, who can be dismissed for inadequate performance, including if the target is missed or for other breaches of the PTA.

Table IV.1.

Institutional Monetary Policy Framework for Selected Inflation Targeting Countries

article image
article image
article image

The ECB does not follow an explicit inflation target at any given horizon, but tries to keep the year on year increase in the area-wide CPI below 2 percent.

The RBNZ target is fixed also trough a formal agreement between the governor and the Finance Minister but this has a less of a cooperative feel than, for example, in Australia (Bean, 1999) where the Treasury subsequently endorsed the RBA decision of adopting an inflation targeting.

100. Both of these aspects reflect the “managerialist” approach that inspired the comprehensive reform of the public sector that has taken place in New Zealand since the mid-1980s. They establish an employment contract between a principal (the government) and an agent (the governor of the RBNZ), with the inflation rate as the single performance measure and where incentives are shaped by the threat of dismissal of the agent if the inflation target is missed.32

101. This approach prevailed over the more “orthodox” ones that emphasized the advantages of giving the central bank complete autonomy from politicians.33 New Zealand’s path to price stability was rather to recognize the inherently political nature of the goal of monetary policy and, once the objective was chosen, to give the central bank total discretion on how to attain it. Both the adoption of a freely floating exchange rate since 1985 and the reform of government finance (through the decision in 1984 to fund the government’s borrowing requirement directly from the market, at market-determined interest rates) contributed to give the central bank effective independence over the conduct of monetary policy.

102. Under the inflation targeting regime New Zealand has managed to achieve and maintain low rates of inflation. More importantly, in an era of generalized inflation reduction in OECD countries, New Zealand has moved from the bottom to the middle of the inflation ranking among these countries (Table IV.2). Although the achievement of a rapid reduction in inflation could well be the consequence of a different political attitude toward macroeconomic instability (as reflected in the wide-ranging macro and micro changes implemented in New Zealand since the mid 1980s), there can be little doubt that monetary arrangement introduced by the 1989 Act has served its main objective well, namely, to lower inflation and to eradicate inflationary expectations.

Table IV.2.

Average Inflation in New Zealand and Other Selected Countries

article image
Source: International financial Statistics, IMF. Inflation is the annual percentage in the CPI. The ranking is against the following 20 OECD countries: Australia, Austria Belgium, Canada, Denmark, Finland, France, Germany, Ireland, Italy, Japan, Netherlands, New Zealand, Norway, Portugal, Spain, Sweden, Switzerland, United Kingdom, and the United Slates. 1/20 means that a country has she highest average inflation of the 20 countries considered.

103. However, ever since the adoption of an inflation targeting doubts have been periodically raised on whether the current monetary framework achieves price stability only through an excessive output, exchange rate and interest rate volatility.34 In particular, the large appreciation (29 percent from trough to peak) of the real exchange rate and the high real interest rates experienced in the 1990s have sometimes been blamed for putting both the export and the household sector under pressure (Sarel, 1999).

104. A first point to note is that the Policy Targets Agreement (PTA) explicitly recognizes that there are conditions under which keeping inflation within the target may lead to excessive economic instability. In order to lower the probability of such an outcome, the PTA has included a list of “caveats,” that is, of circumstances where inflation could legitimately be allowed to fall outside the target.35

105. The PTA has also undergone some major changes over time as inflation has been brought under control, partly in response to the concern about output and exchange rate volatility. In 1990, based on the perception that the rapid disinflation had already proven too costly in real terms as well as on concerns for future real costs, the newly signed PTA extended the date by which the inflation target (0–2 percent) was to be achieved by one year, from 1992 to 1993. In 1996, the new PTA widened the target band to 0–3 percent, in part as a reaction to the experience of the recent past, when the target was breached twice (in the second quarter of 1995 and during three quarters of 1996) despite the sharp increase in interest rates. The latest version of the PTA, signed in December 1999, while maintaining the 0–3 percent inflation target now states that in pursuing its price stability objective the bank “shall seek to avoid unnecessary instability in output, interest rates and the exchange rate.”36

Has inflation targeting come with excessive output and instrument volatility? A quick look at the data

106. Isolating the impact of monetary policy on the economy from that of shocks and other influences is always a very difficult exercise. This is even more so in a small, open economy like New Zealand which is heavily exposed to terms of trade variability, undertook fundamental structural reforms in the second half of the 1990s and was hit by a sequence of shocks in the 1990s. In particular, the lagged effect on domestic demand of both the structural reforms of the mid 1980s and the net immigration flows lead the RBNZ to underestimate the strength of the inflationary pressures in the early to mid 1990s, while the Asian crises and the drought of 1997/98 exacerbated the downward phase of the cycle.

107. With these qualifications, the evidence in Table IV.3 is used to assess the evolution of output, interest and exchange rate volatility in New Zealand since the adoption of inflation targeting. In particular, Table IV.3 presents a cross-OECD comparison of variability (measured by the standard deviation) in real output growth, inflation, real short term interest rates and the real exchange rate before and after the introduction of the inflation targeting regime. All in all, these data seems to suggest that the shift to inflation targeting did not exact a major price in terms of increased economic volatility.37

Table IV.3.

Relative add Absolute Variability of GDP Growth, Inflation, Real Exchange Rate, and Real Short-Term interest Rate 1/

article image
Source: OECD and International Financial Statistics, IMF.

The rank is against the OECD countries listed in the footnote to Table 2. As in Table 2, 1/20 means that the country has the highest variability of the 20 considered.

Annual growth of gross domestic product, volume, at 1995 PPP (USSPPP 1995).

Changes in quarterly CPI.

Present per annum money market interest rage (90 days commercial bill rate) minus CPI inflation (quarterly data).

Real effective exchange rate based on relative CPI (quarterly data).

108. Turning first to output variability, New Zealand’s position in the ranking (based on the variability of real GDP growth) has not changed much in the 1990s as compared to the 1980s. The same is true for the absolute standard deviation. This suggests that output volatility is, to a large degree, the unavoidable result of the structure of New Zealand’s economy and has little to do with the change in monetary policy regime.

109. Likewise, New Zealand’s position in the ranking based on the variability of the real exchange rate has not changed much in the last 10 years as compared to the previous decade. Moreover, the relative and absolute volatility in the real exchange rate have remained broadly unchanged and high.38

110. Finally, the data show that in the last decade New Zealand has significantly improved its relative position in the volatility ranking of the real short-term interest rate. Under inflation targeting, the real short-term interest rate has shown a marked decline in absolute variability and has been among the least volatile in the OECD sample (its level, however, was the highest on average for the 1990s).

C. Inflation Targeting and Economic Stabilization

111. The findings discussed above—that there has been little change in the economic instability since the adoption of inflation targeting—do not mean that there exists no trade off between the variability of inflation and of output. Indeed, a criticism of monetary policy in New Zealand seems to be that the inflation targeting as practiced by the RBNZ has tended to give excessive weight to the inflation stabilization objective.39

112. A first point to make in this respect is that even if output stabilization is not an argument of the central bank loss function there are several channels through which a concern for output stability may affect the actual implementation of monetary policy under inflation targeting. Batini and Haldane (1999) illustrates this point by focusing on a relatively general class of policy reaction functions in an inflation targeting regime such as the following:


where rt is the short term real interest rate (rt = it - Et πt+1) where it is the nominal interest rate, rt is the equilibrium real interest rate, Et is the information set at time t, πt is the inflation rate and π* is the inflation target. According to this rule, monetary authorities control the nominal interest rate so as to hit a path for the short-term real interest rate relative to a weighted average of the lagged and equilibrium real interest rates. The last term shows that a deviation of expected inflation from the target triggers a policy action.

113. According to this policy reaction function, monetary authorities have three parameters to choose in formulating monetary policy, namely, γ, θ and j. The first, γ determines the degree of interest rate smoothing. The second, θ is the policy feedback parameter and defines the aggressiveness of the policy response to a given deviation of the inflation forecasts from target: an increase of the value of θ means a more aggressive policy response to a given deviation of the inflation forecast from the target and thus amounts to shortening the policy horizon (the date at which the target is approximately hit). The third parameter, j, identifies the forecast horizon (the future date at which policy makers react to deviations of inflation from target).

114. Every combination of these parameters identifies a certain speed at which inflation is brought back to target after an inflationary shock, and thus defines the output dynamics associated with the transition path of inflation. In the words of Batini and Haldane (1999): “any degree of output smoothing can be synthetically recreated by judicious choice of the parameters entering an inflation forecast based rule… That is evidence of the output encompassing nature of inflation targeting based around inflation forecasts.”40

115. According to Goodhart (2000), the difficulty in defining an appropriate specification for a loss function with both inflation and output stabilization as objectives is the reason why some countries restrict themselves to give primacy to price stability, while specifying explicitly that there are conditions in which monetary authorities should not aim to return inflation to the target excessively quickly, as this could cause excessive output volatility.41

116. The trade-off between variability of inflation and of output can be illustrated diagrammatically as in Figure IV.1. Increasing the weight that a central bank places on the stability of inflation implies giving up some degree of output stabilization, while the position of the frontier depends on the degree of credibility of monetary policy (gaining monetary stability of inflation implies giving up some degree of output stabilization, while the position of the frontier depends on the degree of credibility of monetary policy (gaining monetary policy credibility and eradicating Inflationary expectations amounts to a downward shift of the frontier)42.

Figure IV.1.
Figure IV.1.

Trade Off Between Output and Inflation Variability

Citation: IMF Staff Country Reports 2000, 140; 10.5089/9781451830262.002.A004

117. Following Svensson’s terminology, a central bank aiming at the top left hand end of the curve (point A) is following a “strict” inflation targeting, under which a central bank tries to correct the divergence of inflation from the target at the shortest horizon possible. Such a policy minimizes the variance of inflation around its target but only at the expense of significant fluctuations in output. The adoption of a more “flexible” inflation targeting regime corresponds to a movement along the curve to the south-east, as the central bank accepts a more gradual return to the target, thereby avoiding sharp fluctuations in activity.

118. New Zealand’s initial approach to inflation target was influenced in large part by the imperative of persuading New Zealanders of the anti-inflation commitment of the RBNZ.43 Over time, the approach has became less strict. Having established the credibility of this commitment (and slashed inflationary expectations) the RBNZ has been able to benefit from a downward shift of the volatility frontier and can now afford more policy flexibility. Simulations run by the RBNZ show that the optimal forecast horizon is now around a period of 6–8 quarters (Drew and Orr, 1999), a much larger time span than the one initially applied (around 4 quarters).44

119. The extension of the forecast horizon (j in the previous policy reaction function) reflected a shift in emphasis from the direct to the indirect inflationary effects of movements in the exchange rate.45 Together with the widening of the target band in 1996 the longer forecast horizon imply that, compared to the initial approach, the RBNZ is likely to react less vigorously once the edges of the band are threatened, and to tolerate some additional short-term inflation variability while focusing on the persistent element of inflation.46

120. The move (in 1999) from a quantity-based (official settlement balances) to an interest rate-based (the Official Cash Rate) implementation regime also goes in the direction of delivering more instrument stability. Under the old regime the main monetary policy instrument was represented by periodic statements from the RBNZ trying to maintain the exchange rate and interest rates consistent with the desired monetary stance (from 1997 this stance was identified with a desired level of the Monetary Condition Index). This created a tendency for interest rates to respond immediately to exchange rate developments, and resulted in high short-term interest rate instability. Under the new regime, the RBNZ no longer attempts to respond to day-to-day fluctuations in the exchange rate and these fluctuations have a smaller impact on short-term interest rate variability (Brookes and Hampton, 2000).

121. To sum up the discussion thus far, there is no convincing empirical evidence that the adoption of price stability as the sole objective of monetary policy and of an inflation targeting regime have exacerbated economic instability in New Zealand. To a certain extent, economic volatility is the unavoidable consequence of the small, open nature of New Zealand’s economy. Moreover, consistent with the view that inflation targeting does not preclude significant attention to conventional stabilization objectives, the RBNZ has progressively changed its approach to inflation targeting in a way that should imply less output and instrument instability.

D. The New PTA

122. As noted above, the new PTA directs the RBNZ to seek to avoid unnecessary instability in output, interest rate and the exchange rate in the pursuit of its price stability objective. This section asks a series of questions in relation to this change, such as; does the new PTA really help in improving the conduct of monetary policy? The answer to this question depends on the existence of a trade-off between the short run variability of interest rates and of output. In case such a trade-off exists, is there an inherent inconsistency in the PTA? Does the new PTA have any implications for accountability and transparency of monetary policy?

Is short terra interest rate smoothing consistent with low economic variability?

123. One way of looking at the new PTA is that it merely formalizes what the RBNZ is already doing in terms of the choice of the parameters of the policy reaction function [1] that is, the feedback and policy horizons (the parameters θ and j, respectively) and the degree of interest rate smoothing (γ).47

124. However, even if they allow a better understanding of central banks behavior, policy reaction functions are unlikely to drive day-to-day monetary policy making. The identification of the optimal values of the parameters is too sensitive to the model of the economy and to the nature of shocks to ease the task of policy makers. In other words, the uncertainty surrounding monetary policy is such that it is often difficult to exploit the volatility frontier, as it is unclear where the economy is located at the time of decision and how it will evolve in the future.

125. One implication of this uncertainty is that it is not clear to what extent a gradual adjustments of interest rate is consistent with smaller output and inflation variability over the cycle, as suggested by Woodford (1999). In his model (that aims to describe the behavior of the US Federal Reserve Bank) a central bank achieves stabilization only by affecting long-term interest rates. As these are determined by market expectations of future short-term rates, a central bank is effective only in so far as it communicates a credible commitment to a future path of short rates. A straightforward way of establishing this credibility is to maintain interest rates at a higher level for a period of time after they are raised, or to follow initial small changes with further changes in the same direction (in a word, to smooth interest rates adjustments).

126. The relevance for New Zealand of the transmission mechanism between short and long-term interest rates is weakened by two factors. First, long-term interest rates may be less relevant for economic agents in New Zealand than in the States, as most debt contracts in New Zealand are linked to short-term rates. Second, given that long-term interest rates in New Zealand are influenced by US long-term rates, there is some doubt on the actual ability of the RBNZ to affect long-term interest rates by changing the OCR (Figure IV.2). Overall, it could be that in order to reduce short-term interest rate volatility the RBNZ may have to accept more variability in output and inflation over the cycle.

Figure IV.2.
Figure IV.2.

New Zealand and US long term interest rates

Citation: IMF Staff Country Reports 2000, 140; 10.5089/9781451830262.002.A004

127. The upshot of this discussion is that, even if the central bank has a rough idea of the optimal degree of instrument smoothing, it may not always be possible to operate monetary policy accordingly. Some times a sharper than planned change in interest rates is necessary in order to hit the inflation target within the desired policy horizon and to avoid a more severe adjustment down the road, indeed, in the recent history of New Zealand’s monetary policy there have been cases in which the RBNZ realized that had it moved earlier and/or faster it would have probably reduced the amplitude of the economic cycle (thereby avoiding a “too little, too late” type of mistake).

Implications of the new PTA on accountability and transparency

128. The new PTA may have some implications on both the accountability structure and the clarity and effectiveness of the RBNZ’s communication of monetary policy decisions to the public. The main argument for delegating monetary policy to a single individual is that this arrangement is the strongest and, at the same time, simplest (in terms of operational complexity and enforceability) way to enforce monetary policy accountability.48 The new PTA parts with this simplicity, as the governor’s performance will not only be judged on whether he did enough to maintain the inflation rate within the target, but also on whether he did enough to avoid unnecessary output and instrument volatility. This may have at least three drawbacks.

129. First, in cases where the desired degree of instrument smoothing is precluded by the risk of increasing inflation and output variability over the cycle, it is possible that the governor may be called to justify his behavior and persuade both the principal (the government) and the public that his “aggressive” move was justified by the need to restrain inflationary pressures and prevent further economic volatility. This may once again increase the attention paid to discussing the implementation of monetary policy, a feature that was seen as one of the major drawbacks of the MCI-based implementation regime (Brookes and Hampton, 2000).

130. Second, it is very difficult to determine what “unnecessary volatility” really means. A typical measure of volatility is the standard deviation of economic variables over a certain period of time, but this can be only calculated ex post. Any assessment of the appropriateness of monetary policy should rather be conducted ex ante, taking into account both the information set available and the uncertainty prevailing at the time of the decision. Even if such a careful (and costly) review is carried out, it is quite unlikely to produce any compelling evidence about the specific responsibility of monetary policy during the cycle.

131. Third, although the objective difficulty in evaluating whether the RBNZ has delivered an adequate performance as defined by the new clause may weaken the threat posed by the dismissal procedure by blurring monetary policy accountability, the same difficulty is a double-hedged sword that can be used to put more pressure on the operation and implementation of monetary policy, in order to influence its implicit degree of output stabilization.

132. The system of “checks and balances” set up by the Reserve Bank Act and, especially, the high degree of transparency that must be given to any conflict between the RBNZ and the political principal, strongly mitigates the risk that the new clause will be used to unduly politicize the implementation of monetary policy. At the same time, even if the costs (in terms of loss of monetary policy credibility, higher inflation, larger risk premium on foreign debt etc.) implied by such a conflict are so high, to make it unlikely to occur, any perceived attempt to affect the instrument-independence of the RBNZ may damage the credibility of its policy and, therefore, its ability to influence inflation expectations.

133. All in all, while the new clause in the PTA is not expected to alter the current RBNZ modus operandi, it is difficult to consider it as a completely neutral and irrelevant addition. As discussed above, it is not impossible to envisage scenarios in which the introduction of the new clause might end up exposing monetary policy to tensions from which it was previously immune. In order to contain these risks the RBNZ needs, first, an additional investment in clarity of the communication of monetary policy decisions and, second, to make sure that the system of “checks and balances” is internally consistent and effective.

134. As for the transparency of monetary policy, the RBNZ is known to be one of the most transparent central banks of the world (Table IV.1). However, as shown by the debate over the degree of economic stabilization that is implicit in its approach to inflation targeting, the main communication challenge for the RBNZ is to make monetary policy “clear” in the sense specified by Winkler (2000). This amounts to establish a coherent frame of reasoning through which every individual can interpret the subset of relevant information in the same way as everybody else, an objective that is not automatically achieved by making the maximum amount of information available (Issing, 1999).49

135. As for the system of “checks and balances” a possible step in the direction of efficiency and internal consistency would be to remove the governor from the membership to the Board of Directors (the body which assesses his performance), Indeed, the governor himself has signaled in the past that there is some “awkwardness” because he is both the chief executive and the chairman of the board that reviews his performance.50

E. Accountability: Collective or Individual?

136. This section turns to what is likely to be a key element in the forthcoming review of monetary policy operations, namely the issue of decision making process and the accountability structure. In particular, the following-question is addressed: if consideration has to be given to moving to a collective decision-making structure, what are the issues at stake in the formal delegation of monetary policy to a committee?51.

137. Under the “managerialist” approach, the main arguments for centering monetary policy accountability on the sole figure of the governor are:

  • the operational complexity associated with any attempt to maintain such individual accountability within a committee. If one admits that individual accountability cannot be reduced to the mere act of voting, distinguishing individual responsibilities would require very detailed minutes and, in principle, individual forecasts (Issing, 1999).52

  • the risk that providing individual incentives within a committee may jeopardize the effectiveness of collective decision-making, because free-riding and, in general, strategic behaviors can dilute individual incentives within a committee.53

  • the negative effect of decision-making by a committee on transparency and clarity, insofar as it makes the communication problem (and the public perception of the central bank decision-making process) more difficult.

  • the need to solve a series of practical issues, such as how to choose the number of its members, whether to select some of them outside the central bank, how to employ them so as to avoid any potential conflict of interests, etc. While some suggestions can be derived from the arrangements adopted in other countries (as showed in Table IV.1.) these problems are probably more difficult to solve in a country with a low population density as New Zealand.

138. The desire to save on the organizational costs deriving from a committee-based decision making process (i.e., the lengthening of the decision making process which may result in an increase in the likelihood of failure to make decisions on time) is also an argument for a centralized structure. However, one should not overstate this argument because, although the governor of the RBNZ is personally responsible for the formulation of monetary policy, in practice he makes decisions with the advice of an internal body, the monetary policy committee. However, the fact that there is no imperative to achieve consensus may speed up the whole decision making process.

139. On the other hand, key arguments in favor of collective responsibility are:

  • concentration of responsibility can lead to great pressure being exerted on the governor when the government or the public believe that the policy is inappropriate. As this conflict could affect monetary policy credibility, it may be preferable to take the spotlight off the governor and delegate decision-making to a group of people that may better be able to resist external pressures (Siklos, 1996).

  • a committee could allow a candid exchange of information between the committee members prior to voting, thereby leading to a better informed decision-making process (Cukierman, 1999). Such an argument could be based on an intuitive reluctance to delegate decision making authority to any single individual (“there is an implicit belief that the wisdom of a committee might be greater than that of any single member, that collective decision making avoid some of the worst errors that might otherwise occur” Sah and Stiglitz, 1988), but also on an application of the Condorcet (1785) “jury theorem,” stating that majorities are more likely than any single individual to select the best of two alternatives when there exist uncertainty about which of the two alternatives is in fact the best.54 Along the same lines, Nitzan and Parush (1985) show that in the presence of uncertainty appropriate aggregation of information possessed by different experts leads to a better decision.

140. These arguments suggest that having the members of the committee sign different contracts (so that only one, or some of them, could be fired in case of inappropriate policy) would probably lead to a suboptimal institutional framework. Such an arrangement would, in fact, boost the costs associated with both the identification of individual responsibilities and the strategic interaction among the members of the committee.

141. This leaves two possibilities. The first one is to have a single employment contract signed by more than one person. If monetary policy was considered to be inconsistent with the principles set by the PTA, all members of the committee would be fired (independently of the individual responsibilities). This framework would benefit from the advantages of a committee-based decision making process and, at the same time, could avoid the costs associated with enforcing multiple contracts, but would amplify the difficulty in finding a number of potential Board members.

142. The second possible arrangement would be to delegate decision making to a committee and to abandon the employment contract. A weaker form of individual accountability could be still maintained by having monetary policy decisions taken through a majority-rule and revealing how members voted through the publication of minutes (like in the UK).

143. Abandoning the employment contract would amount to recognizing that the RBNZ is accountable to the market and the general public, rather than to the political principal.55 In a way, this is already the case for the current framework: the first line of defense against a “wrong” monetary policy is the reaction of markets, and the exercise of the dismissal clause is conceivable only in the context of extreme scenarios. However, the need to preserve uniformity of treatment between the central bank and the public sector accountability frameworks may make it politically infeasible to pursue this option.

144. To conclude, adopting a committee-based decision making process presents both gains and drawbacks. In principle, it is possible to carefully weight all of them so as to identify the optimal institutional design for monetary policy in New Zealand, that could well be different from the actual one. Such an exercise, on the other hand, should take into account that every institutional change in monetary policy involves a transitional period in which credibility must be re-established by the new institution. To the extent that the volatility frontier in Figure IV.1 shifts upward, this may imply some temporary additional costs in terms of economic instability. In the absence 6f a compelling reason to change, it may thus be worthwhile to remain with the current decision-making process.

F. Conclusions

145. After the adoption of an inflation targeting regime New Zealand has experienced a decade of low and stable inflation. Although it is difficult to assess the specific contribution of inflation targeting to this result, the new monetary framework appears to have served well the main objective of the reform, namely to communicate a credible change in attitude toward inflation and eradicate inflationary expectations. During the last decade, however, monetary policy in New Zealand has been occasionally blamed for amplifying output, exchange rate and instrument rate volatility.

146. In part as a result of these concerns, a number of changes have been implemented in the PTA in the second half of the 1990s, with the latest change being a new clause added to the last Policy Target Agreement, stating that in pursuing price stability the RBNZ should avoid unnecessary economic instability. Also, with 10 years of experience with inflation targeting in place, the new government has launched a review of the operation of monetary policy. This review has been designed so as not to question the exclusive focus on price stability, but will “examine the way the Reserve Bank interprets and applies the inflation target set out in the Policy Targets Agreement, with a view to ensuring that this approach to achieving medium-term price stability is consistent with avoiding undesirable instability in output, interest rates and the exchange rate.”

147. This paper argues that there are two reasons why inflation targeting might not be responsible for an excessive degree of output and instrument volatility in New Zealand. First, New Zealand’s relative economic instability has not worsened in the last decade compared to the pre-Reserve Bank Act period. Second, even if the only goal of monetary policy is price stability, this does not imply that inflation targeting is inconsistent with other conventional stabilization objectives.

148. In the implementation of inflation targeting a central bank has several ways to “factor in” a concern for output and instrument stability. After the initial relatively “strict” approach to inflation targeting, the RBNZ has been moving in the direction of a more flexible approach to inflation targeting by extending its forecast and policy horizon and widening the target band. Abandoning the MCI-based implementation regime has also reduced the volatility of short-term nominal interest rates.

149. It can therefore be said that the change in PTA formalizes the existing modus operandi by making explicit the RBNZ’s concern for economic stabilization. However, this explicit formalization may in principle expose monetary policy to tensions from which it was previously safe. First, in order for the governor of the RBNZ not to be exposed to undue criticisms, it is important that everyone understand the framework, objectives and transmission mechanism for monetary policy. This requires an additional effort in making monetary policy clear about how it is avoiding unnecessary volatility in output, interest rates and the exchange rate without elevating this discussion to the same level as that of the price stability objective, In this way, it risks directing excessive market and public attention towards implementation issues. Second, given the objective difficulty in assessing ex post the appropriateness of monetary policy, there is scope for more political pressure on the RBNZ and this may damage monetary policy credibility, even if the system of “checks and balances” set up by the 1989 Reserve Bank Act is sufficient to ensure that the RBNZ’s instrument independence is maintained in principle.

150. The paper next focuses on a possible modification of the current monetary policy institutional framework, namely the move towards a committee-based decision making process. This idea has some advantages, especially as it moves the spotlight in the discussion of the conduct of monetary policy off a single individual. But it also raises a series of additional problems, not least its consistency with the principal-agent framework underlying public sector management in New Zealand. Considering that at least part of the credibility so far established risks being lost following such an important institutional change, it may be worthwhile following the principle “if it is not broken don’t fix it.”


  • Archer, David and Peter Nicholl, 1992, “An Announced Downward Path for Inflation,” Reserve Bank of New Zealand Bulletin, Vol. 55, No. 4.

    • Search Google Scholar
    • Export Citation
  • Austen-Smith, David and Jeffrey Banks, 1996, “Information Aggregation, Rationality and the Condorcet Jury Theorem,” American Political Science Review, Vol. 90. No. 1.

    • Search Google Scholar
    • Export Citation
  • Batini, Nicoletta and Andrew Haldane, 1999, “Forward looking rules for monetary policy,” Bank of England Working Paper series, No. 91.

    • Search Google Scholar
    • Export Citation
  • Bean, Charles, 1999, “Australasian Monetary Policy: A Comparative Perspective,” The Australian Economic Review, Vol. 32, No. 1.

  • Bernanke, Ben, Thomas Laubach, Frederic Mishkin and Adam Posen, 1999, Inflation Targeting: Lessons from the International Experience, Princeton University Press, Princeton New Jersey.

    • Search Google Scholar
    • Export Citation
  • Brash, Donald, 2000, “Inflation targeting in New Zealand: 1988–2000,” Reserve Bank of New Zealand Bulletin, Vol. 63, No. 1.

  • Brash, Donald, 2000, “Monetary Policy into the New Millennium,” Reserve Bank of New Zealand Bulletin, Vol. 62, No. 2.

  • Briault, Clive, Andrew Haldane and Mervyn King, 1996, “Independence and Accountability,” IMES Discussion Paper, No. 17.

  • Brookes Andy and Tim Hampton, 2000, “The Official Cash Rate one Year On,” Reserve Bank of New Zealand Bullettin, Vol. 63, No. 2.

  • Condorcet, Marquis de, [1785] 1994, Essai sur l’application de l’analyse a la probabilite’ des decisions rendues a la pluralite’ des voix, Trans. Iain McLean and Fiona Hewitt, Paris.

    • Search Google Scholar
    • Export Citation
  • Cukierman, Alex, 2000, “Accountability, Credibility, Transparency and Stabilization Policy in the Eurosystem,” unpublished manuscript.

    • Search Google Scholar
    • Export Citation
  • Drew, Aaron and Adrian Orr, 1999, “The Reserve Bank’s Role in the Recent Business Cycle: Actions and Evolution,” Reserve Bank of New Zealand Bulletin, Vol. 61, No. 1.

    • Search Google Scholar
    • Export Citation
  • Fischer, Stanley, 1995, “Modern Approaches to Central Banking,” NBER Working Paper No. 5064.

  • Goodhart, Charles, 2000, “Monetary transmission lags and the formulation of the policy decision on interest rates,” LSE Financial Markets Group, Special Paper No. 124.

    • Search Google Scholar
    • Export Citation
  • Issing, Ottmar, 1999, “The Eurosystem: Transparent and Accountable,” CEPR Policy Papers, No. 2.

  • Minot, Dean and Dominick Stephens, 2000, “International Monetary Policy Decision Making and Transparency Arrangements,” mimeo.

  • Nitzan, Shmuel and Jacob Parush, 1985, Collective Decision Making: An Economic Outlook, Cambridge University Press, Cambridge London and NY.

    • Search Google Scholar
    • Export Citation
  • Orr, Adrian, Alistair Scott and Bruce White, 1998, “The Exchange Rate and Inflation Targeting,” Reserve Bank of New Zealand Bullettin, Vol. 61, No. 3.

    • Search Google Scholar
    • Export Citation
  • Sah, Raaj K. and Joseph E. Stiglitz, 1988, “Committees, Hierarchies and Polyarchies,” The Economic Journal, Vol. 98.

  • Sherwin, Murray, 1999, “Strategic Choices in Inflation Targeting: the New Zealand Experience,” Reserve Bank of New Zealand Bulletin, Vol. 62, No. 3

    • Search Google Scholar
    • Export Citation
  • Sibert, Anne, 1999, “Monetary Policy Committees: Individual and Collective Reputations,” CEPR Discussion Paper, No. 2328.

  • Siklos, Pierre, 1996, “Charting a Future for the Bank of Canada: Inflation Target and the Balance Between Autonomy and Accountability,” unpublished manuscript.

    • Search Google Scholar
    • Export Citation
  • Svensson, Lars, 1997, “Inflation Forecast Targeting: Implementing and Monitoring Inflation Targets,” European Economic Review, Vol. 41.

    • Search Google Scholar
    • Export Citation
  • Vickers, John, 1999, “Economic Models and Monetary Policy,” Bank of England Quarterly Bulletin, May.

  • Winkler, Bernard, 1999, “On the Need for Clarity in Monetary Policy Making,” paper presented at the European Central Bank conference on “Monetary Policy-Making under Uncertainty,” Frankfurt, December 3/4, 1999.

    • Search Google Scholar
    • Export Citation
  • Woodford, Michael, 1999, “Optimal Monetary Policy Inertia,” NBER Working Paper, No. 7261.


See Finance Minister Cullen’s speech announcing the monetary policy review, May 9, 2000.


In Brazil (where an inflation targeting regime was adopted in 1999), the legislation charges the Central Bank with the main task of promoting the stability of the purchasing power of the currency but also refer to secondary objectives, such as providing the economy with adequate liquidity, maintaining the international reserves of the country at adequate levels and promoting savings mobilization at adequate levels to finance domestic investment.


As noted above, the contract is embodied in the PTA, which both the governor and Minister have to sign. It should be stressed that the institutional framework established by the Reserve Bank Act commit the Minister as much as the Governor to the target fixed by the PTA. Should the governor regard the proposed PTA as inconsistent with price stability he or she can decline to sign it, requiring the government to use the override facility and thus signaling to the market the existence of such inconsistency.


According to Fischer (1995), given the difficulty in finding a mechanism that makes sure that central bankers behave optimally from a welfare point of view, it is accountability (and thus instrument-independence) rather than goal-independence that really matters for inflation performance.


To quote the Governor of the Reserve Bank of New Zealand, Don Brash: “even those who concede that New Zealand’s performance in keeping inflation low and stable has been good often argue that the cost of achieving this, in terms of economic growth and employment foregone, has been too high and that perhaps something more moderate, or “less obsessive” in the words of our critics, would have been desirable” (Brash, 2000).


These include large swings in the terms of trade, significant changes in indirect taxes and in other government policies that directly affect prices, and natural disasters. Initially, these caveats were taken into account through the estimation of “underlying” inflation, which excluded the impact of the shocks. Estimates of underlying inflation were reported alongside headline inflation until 1997. Since then, the RBNZ no longer estimates underlying inflation, and has started following a more qualitative approach to taking account of shocks.


The initiative for these changes came from the newly-elected government. It is interesting to note that in commenting the possibility of a change in the PTA after the 1999 election, the Governor of the RBNZ drew attention to the risk of excessively frequent changes in the operational aspects of inflation targeting, by noting that “consistency and continuity are attractive in monetary policy as elsewhere, and I shall certainly not be promoting a change” (Brash, 2000).


These results are roughly consistent with those presented by Drew and Orr (1999). Drew and Orr, however, use overlapping periods (e.g., 1979–1999 compared to 1989–1999), which do not clearly distinguish between the pre- and post-inflation targeting periods.


On the other hand, if the European countries that were previously in the European Monetary System and are now in the European Monetary Union were excluded (basically all European countries in the list, with the exception of Norway and Switzerland), New Zealand would be in middle of the ranking (4 out of 7 in the first period and 3 out of 7 in the second one).


For example, Bean (1999) compares monetary policy in New Zealand, U.K., and Australia and argues that the sole concentration on price stability in the 1989 Reserve Bank Act may result in excessive volatility in activity.


Svensson (1997) also shows that inflation targeting is completely consistent with a conventional quadratic central bank loss function that places arbitrary weights on the output-gap and inflation.


In the UK, the letter of intent sent by the Chancellor to the Governor of the Bank of England and specifying the inflation target contains a similar clause, which states that: “The framework is based on the recognition that the actual inflation rate will on occasions depart from its target as a result of shocks and disturbances. Attempts to keep inflation at the target in these circumstances may cause undesirable volatility in output.” In the case of New Zealand, these conditions are explicitly identified in the “caveats” in the PTA.


It is useful to think to these curves as if they were obtained by subjecting to random shocks a model of the economy that incorporates a monetary policy reaction function like [1]. Using the same battery of shocks but alternative values of the parameters of choice in [1] gives the combination of output variability and inflation variability shown in the figure (for example, for given values of γ and θ point A would be obtained by fixing a relatively short policy horizon, that is, a relatively low value of j).


To quote Murray Sherwin, Deputy Director of the RBNZ, “the target was sometimes described as being bounded by electric fences—approach if you want, but not touch!” (Sherwin, 1999).


It should be noted that the forecast horizon of 6–8 quarters is optimal not because it is the result of a loss function minimization problem, but rather because it efficiently exploits the convexity of the trade-off (“if is evident that as the policy horizon is extended much beyond the 6–8 quarters not much is to be gained in terms of reduced output and instrument volatility, while inflation volatility increases quite markedly” Drew and Orr, 1999). This is the same result obtained by Batini and Haldane in the context of the UK (1999).


The higher emphasis on the indirect inflationary impact (via the aggregate demand) of movements in the exchange rate partly reveals the intention of smoothing interest rates responses to these movements, and partly reflects recent evidence that the direct impact of changes of the exchange rates on tradable prices has become more muted (Orr, Scott and White, 1998).


To quote the Deputy governor Sherwin again, “the upshot has been a shift in emphasis away from avoiding a breach of the target at all costs and toward a firmer focus on having inflation always reverting to somewhere near the midpoint of the target range in the medium term” (Sherwin, 1999).


The reference to “unnecessary instability” may not appear that unusual, prima facie. For example, the letter of intent between the Chancellor and the Bank of England (footnote 41) mentions the possibility of “undesirable volatility in output” in the context of shocks and disturbances. In this sense, it plays the same role of the caveats identified in the “Unusual Events” subsection of the PTA. The latest PTA, however, includes the clause concerning the unnecessary instability in output, interest rates and the exchange rate in the subsection entitled “Implementation and Accountability”. The need to make explicit the concern for such volatility not only in the context of the caveats suggests that the new clause goes somewhat further than what may already have been implicit in the conduct of monetary policy.


Referring to the governance arrangement of the RBNZ, Minot and Stephens (2000) state that: “none of the other central banks surveyed had any comparable accountability measure in their legislation. This could be because designing such clearly defined consequences for poor performance is more difficult when a committee formulates monetary policy.”


It should be stressed here that the objective difficulty in making monetary policy “clear” should not be used as an alibi to maintain a larger degree of secrecy. As pointed out by Winkler (2000), recognizing the “limits” of transparency makes the task of being transparent not less desirable, only much harder.


In practice, although the governor chairs the Board’s meetings, the Board has delegated the responsibility for the formal monitoring and evaluation of the governor’s performance to the non executive directors committee of the Board. Thus, the proposed change would bring the de jure framework to assess the governor in line with the de facto institutional setup,


According to the JPMorgan’s “Guide to Central Bank Watching,” out of the 39 countries considered, only in New Zealand and Norway is monetary policy decision-making formally in the sole hands of the governor. In the case of Norway, however, the central bank is not even instrument-independent as it has to consult with the government on interest rate changes (the Reserve Bank follows an implicit inflation targeting), so the governor is responsible for the day-to-day management of monetary policy. His performance in this task is monitored by an independent body, the Supervisory Council, whose 15 members are elected by the Parliament.


This point is also raised by a member of the Bank of England Monetary Policy Committee (MPC), John Vickers, who entertains the possibility of a tension between individual accountability of MPC members and the apparently collective nature of the published forecasts by asking “how is individual freedom in voting consistent with collective ownership of the means of projections?” (Vickers, 1999).


As RBNZ governor Don Brash puts it: “7 think you may get people who are very aware of their reputation, either currently or historically, who start playing to the gallery”. However, comparing monetary policy made by a committee and by a single individual within a reputation-building model, Sibert (1999) shows that the strategic interaction among the members of a committee could ultimately work both ways and the inflation outcome ultimately depends on the degree of patience of policy-makers.


This result, however, depends on each member of the committee behaving in exactly the same manner as when he alone selects the outcome (that is, it depends on individuals voting “sincerely”). If “sincerity voting” is not rational (it does not lead to a Nash equilibrium) and strategic behavior is introduced (with the opportunity for the members of a committee to manipulate the collective decision to their particular advantage) the result may change (Austen-Smith and Banks, 1996).


Briault, Haldane and King (1996) distinguish between accountability de jure and accountability de facto (or transparency). Accountability de jure is based on a formal, “legal” mandate to carry out a duty, i.e. through a fully specified contract between the government and the central bank. But making the central bank’s actions, intentions and analysis transparent exposes central bank’s reputation to an external check, for which it will suffer loss or failure if it is found wanting.