"The Role of Monetary Policy Rules in Inflation Targeting Regimes - Theory meets Practice"

Opening address by Governor Svein Gjedrem, Norges Bank

Norges Bank, 5 - 6 May 2003

Ladies and gentlemen,

I am very pleased to welcome you all to Norges Bank.

One of the main purposes of this workshop has been to gather a small group of people with one interest in common - the conduct of monetary policy under an inflation targeting regime. The group should be small enough to allow for an open and stimulating discussion around the table, but at the same time broad enough to include both people that are interested in monetary policy from a theoretical point of view and those that have practical experience with inflation targeting.

Inflation targeting in practice involves several demanding steps. The stance of monetary policy is designed to attain a target some periods ahead. But as we all know, before taking the first step towards reaching our target, we need to know where we are. The signals that indicate where we are at present are many, and they might be confusing. And even with a perfect knowledge of where we are, there are different tracks that we can follow. In addition, our knowledge of how the economic system works is not at all perfect.

We have chosen to focus on three different topics.

In today's first session we want to draw attention to the interaction between theoretical and practical monetary policy under the title "Evaluation of policy rules for practical use". Both approaches are well represented as we will have two main presentations in this session. The first will be given by Lars Svensson from Princeton, representing the theorists. After that, David Archer from the Reserve Bank of New Zealand will give a presentation more from a practician's point of view.

We all know there is a trade-off between output and inflation stability. We have to make a choice as to how to achieve a balance between these two objectives. Flexible inflation targeting implies that we place some weight on both of them. However, central banks are not free to choose whatever weights they want. Our mandates clearly constrain our choices. From a theoretical point of view, this may be described as minimising a loss function. I think this is a good description of how we think. And how we think is certainly influenced by the academic literature. The theoretical frameworks with which this literature presents us, although they are rather rough approximations of the reality we have to face, help us to clarify, discipline and elaborate our own practice.

Lars Svensson and other academicians with him urge us to be open about our loss function. Few central banks, if any, publish their loss function with a specific figure for the "lambda" - the relative weight on output stability. Norges Bank follows a specific targeting rule. The interest rate is set on the basis of the inflation forecast two years ahead. The forecast is our intermediate target. Behind our choice of the two years horizon lies an implicit loss function. It should be possible to reveal our "lambda" through our actions. Late in his life, the Norwegian economist Ragnar Frisch1 , winner of the Alfred Nobel Memorial Prize in Economic Sciences, tried to chart politicians' utility function through interviews. The project was certainly too ambitious, and it led to a dead end. Describing the central bank's objectives through a loss function and quantifying the relative weight put on output stability is a more clearly delimited task.

The specific targeting rule may, however, be easier to communicate than a loss function. In some respects, the horizon chosen and the manner in which we explain our targeting rule to the public is a translation of our loss function. This translation, especially with a fixed horizon, may be seen as a rather rough approximation of the more elaborate theoretical framework. However, in our communications we have emphasised that it may be appropriate to depart from the two-year horizon in certain situations. We are of the opinion that we have done a good job of converting Svenssons framework to an explicit targeting rule with a fixed horizon and the inflation forecast as our intermediate target. We should, however, always be receptive to other suggestions on how to communicate.

Most inflation targeting central banks base the setting of interest rates on forecasts of the inflation rate some periods ahead. But in order to make these forecasts and take the necessary action, we ought to have a qualified view of the most recent developments. Before we can move on, we need to know where we stand. The different signals we get do not always tell the same story. The signals may even point in different directions. National accounts data form our main point of reference when we describe economic developments. But producing good national accounts data is a costly and time-consuming process. It takes time before all the data that make up the picture are collected and processed. New sources of information that are made available also compel the statistical authorities to revise the national accounts system now and then in order to make use of these new sources. It may take several periods before the final data is published. So we have to search for data that are available to us as soon as possible and that may serve as good indicators of the overall pressure in the economy. We often need to judge these indicators against some sort of benchmark - often called natural or neutral rates. These rates are unobservable and they may change over time.

The workshop will deal with these matters in the first session tomorrow. We are very pleased that Athanasios Orphanides from the Federal Reserve is going to share some of his thoughts on these topics with us.

As central bankers, we do not face competition in our own countries that may encourage us to improve our efficiency in attaining our targets. As an alternative, we should aim at improving in relation to an international standard. We all want to achieve the label "best international practice" for our performance. But what are best practice criteria? That is the subject for the last session tomorrow. And we want to focus on one special part of our activity, namely our inflation reports. These reports are the corner stone of an inflation targeting regime.

A year ago, Norges Bank invited Hans Genberg and Charles Wyplosch to evaluate our Inflation Report. To be honest, evaluating one volume of Norges Bank's Inflation Report did not appear the most stimulating task from an intellectual viewpoint. So we challenged them to study the inflation reports of several inflation targeting countries, and come up with a set of "best practice" criteria. The focus of their report should not be Norges Bank's Inflation Report, but inflation reports in inflation targeting countries. We are very pleased with the result of their work, although it showed that there is room for improvement when it comes to our own Report. But due to their report, we know what we should aim at, and we have in fact already made some improvements. Inflation reports are certainly only one part of central bankers' communication strategy. Some of you may have chosen to focus on other information channels. These channels are, however, outside the scope of Genberg and his co-authors' present work. I am looking forward to hearing your views on the set of criteria that Genberg and his co-authors, Andrea Fracasso and Charles Wyplosch, have come up with. You may not all agree with their choice. In any case, I am sure that tomorrow's discussion will help to clarify what should be the criteria. And I hope we can learn from each other.

I hope you are all looking forward to one and a half days of presentations from distinguished members of the academic and central banking world. But most of all, I hope you will take the opportunity to engage yourselves in the discussions in order to strengthen the bridge between the theory and practice of monetary policy.


1Ragnar Frisch: Samarbeid mellom politikere og økonometrikere om formuleringen av politiske preferanser (Cooperation between politicians and econometricians on the formulation of political preferences), Sosialøkonomen nr. 6, 1971.

Published 29 April 2005 15:06