Heikensten: Conference on inflation targeting: implementation, communication and effectiveness – opening remarks
More information about the conference you will find in the link below
I would like to begin by welcoming you all to Stockholm and to this conference on inflation targeting. The idea of holding this conference was hatched a couple of years ago, when Lars Svensson and I were talking about how the thinking in this area has evolved in recent years both in theory and in practice. Where do we stand today in research terms as well as in practice? And what can we learn from each other on the basis of our experiences? The fact that the conference was planned to take place this year provided added incentive, as it is ten years since Sweden's inflation targeting regime was fully implemented in accordance with the original decision in 1993. We decided to spread the word in our respective networks: Lars on the academic side and I on the practical side. We are delighted that you, our colleagues and friends, have accepted our invitation, and believe we can look forward to a rewarding discussion of where inflation targeting stands today and how it can be developed further.
I do not think it is an exaggeration to say that monetary policy today is, on the whole, being conducted better and more prudently than ever before. In many countries, the key to this success has been inflation targeting. Although the main reason for introducing inflation targeting in many cases – not least Sweden's – has been frustration over the previous regime and economic performance, it has subsequently proved to work so well that it now stands out as an interesting option also for already well-functioning economies. This is demonstrated in particular by the current debate regarding a possible introduction of inflation targeting in the United States, where the economic performance in recent decades has been remarkably good and monetary policy successful.
But first, before I go into a problem that currently feels significant for us at the Riksbank and that I also think is central to the discussion in the United States and elsewhere – the balance between concerns for output stability and inflation stability – allow me to say a few words about my fundamental views regarding inflation targeting policies. Unsurprisingly, I see considerable advantages of having an explicit, symmetric inflation target. It is also my impression that this view enjoys considerable support in the academic literature, an assertion, however, that no assembly is better placed to question than this one.
The clear framework encompassing an inflation target, published forecasts and a specified target horizon has generally worked well as a guide for policy. There have been communication problems here and there, but it is not obvious that the situation would have been much better under a different regime. Another benefit of a clear framework, which I think has received too little attention in academic circles, but which in my opinion has been important for the favourable outcome, is that it helps bolster both the internal work and the economic discussion. Other considerations than those purely related to inflation have most likely been put to a more rigorous intellectual test than would otherwise have been the case, which I think has been useful.
However, the fact that inflation targeting has worked well does not mean, of course, that all problems have been solved and that there are no more issues worth discussing; if that were so, there would have been no reason to arrange this conference. By way of introduction, I intend to describe some Swedish experiences of a problem area in inflation targeting that I see as one of the more significant, perhaps even the most significant – how to deal with the trade-off between stabilising inflation and stabilising the real economy. I will focus in this regard on two of the areas central to this conference – how policy is implemented and communicated.
The inflation-real economy trade-off
Flexible inflation targeting policy
It is presumably safe to say that all inflation-targeting countries today conduct what Lars Svensson would call a flexible rather than a strict inflation targeting policy, even though the degree of flexibility probably varies somewhat from country to country. In other words, they are not, to use Mervyn King's well-known expression, "inflation nutters", i.e. they do not focus only on stabilising inflation but also take account of real economic developments and possibly other factors too.
However, the present degree of flexibility in inflation-targeting policies has probably not always been regarded as so obvious as it is today. At least as far as Sweden is concerned, I think I am right in saying that it has been a gradual process that has brought us to where we are today. That is probably also true for many other inflation-targeting countries.
Immediately after the announcement of the inflation target in 1993, the Riksbank tended to formulate policy as if the intention was to be anything but flexible. This may have been because the credibility problems in the 1970s and 1980s were considered damaging and were deemed to warrant "tough talk". Whether it was a wise strategy to begin in this way with a one-sided focus on inflation is perhaps open to question. One could argue that it would have been better to choose a more balanced line from the outset. It might then have been possible to counteract some of the criticism levelled at the Riksbank for supposedly "focusing exclusively on inflation and completely disregarding employment".
From the mid-1990s and onwards, though, the Riksbank tried in various ways to clarify that we were not focusing solely on stabilising inflation. I believe it is correct to say that the implementation of policy also began to become more flexible around that time. One of the examples of this is the fact that we did not utilise the scope for rate cuts that – according to a stricter interpretation – deregulation and other supply factors had created during the last few years of the 1990s.
The clearest example of how we attempted to convey that our inflation targeting policy was flexible rather than strict was the clarification of monetary policy that the Executive Board published in 1999. In that, we explained that it may be appropriate in some situations – e.g. in the event of large shocks and deviations from target – to refrain from bringing inflation back to target too quickly in order to mitigate the adverse consequences for output and employment. (1) We expressed this in terms of there being reasons to occasionally depart from our simple policy rule – that the repo rate is normally raised if inflation one to two years ahead is forecast to be above 2 per cent, and vice versa if inflation is forecast to undershoot the target.
Explicit framework as basis
The strategy that we in Sweden have chosen for dealing with the "flexibility problem" could, in simplified terms, be described as that we, when we have found it necessary, have in various ways modified and adapted a framework – mainly represented by an explicit symmetric target, regular publication of detailed decision-support data including forecasts, and a simple policy rule for how the repo rate is normally adjusted against the background of the forecasts. This is probably the key reason why the flexibility on our part can chiefly be said to be reflected in "exceptions" to the policy rule, rather than it being explicitly described in terms of a continual procedure of balancing concerns for output and inflation stability. I do not think this way of handling the flexibility issue is in any way unique to Sweden but instead is fairly typical.
One of our experiences is that we have found it quite difficult from time to time to get this kind of message across, at least when the flexibility is reflected in an unchanged or raised policy rate, as opposed to a cut. In such circumstances it seems to be rather hard to debunk the idea that monetary policy should be conducted according to a mechanical rule that says that the policy rate, regardless of what forces are acting on the economy at the time, should always be set to ensure that the inflation forecast at a given time horizon meets the target. On those occasions we have found it appropriate to depart from our policy rule in this regard, our actions have all too seldom been taken in the public domain as a sign that we conduct a flexible inflation targeting policy. Instead, the interpretation at times has been that we have changed our way of conducting monetary policy or that we have decided, for some more or less obscure reason, to maintain a certain line irrespective of what our forecasts tell us.
To some extent, the problems we have experienced probably reflect the fact that most people always prefer lower interest rates to higher ones. But it could also be the case, of course, that the way in which we have dealt with and communicated the flexibility problem has simply not been clear enough, despite our best efforts. One could perhaps see it as a sign that we have been too successful in “anchoring” the simple rule that we introduced at an early stage in an attempt to clearly explain our decisions and to create confidence in the new regime.
Discrepancy between word and deed
Of course, Sweden is not the only country in which the communication of flexible inflation targeting has caused problems. For instance, at the international level there is more general criticism claiming that the central banks in inflation-targeting countries are biased in their monetary policy communication towards stabilising inflation.(2) The criticism is hence not levelled at how inflation targeting is conducted in practice; it recognises that the central banks in inflation-targeting countries do take account of the real economy. It does however claim that there is a discrepancy between word and deed that raises questions regarding what status the stabilisation of the real economy really has in relation to stabilising inflation. I should perhaps underline here that this criticism, as I understand it, does not say that there has been greater clarity in countries that do not have specific targets for inflation.
In my opinion, the assertion that the communication of inflation-targeting central banks generally tends to focus a lot more on the stabilisation of inflation than of the real economy is essentially correct. Inflation forecasts undeniably have the leading role in the reports published by these central banks, and it is usually on the basis of these forecasts that the banks explain their policy decisions.
So why have central banks that pursue a flexible inflation targeting policy chosen to concentrate on inflation in their communication and often been rather vague as regards how they look upon the trade-off between stabilising inflation and stabilising the real economy? One important explanation, which I already have touched upon, is probably that many of the economies that have adopted inflation targeting policies were previously beset with high, sharply fluctuating inflation. The task, accordingly, was to create broad understanding and credibility for a new policy focused on price stability. It therefore was natural to communicate policy in terms of factors that are relatively easy to measure and comprehend, and to do so according to concrete, relatively simple principles that could serve as a guide for monetary policy and that could contribute to stabilising inflation expectations. Moreover, from the point of view of explaining policy it is good if the overall objective of price stability that has been adopted by elected bodies in many countries can also be used in the implementation of policy and be allowed to play a prominent part in the debate.
That the communication still tends to focus so strongly on inflation, despite the considerable confidence in low-inflation policies today, could reflect a perception that it is not all that easy to make substantial changes to the policy rhetoric. Personally, however, I think the problem runs deeper than that. It is quite simply difficult to clearly communicate the kind of varying, subtle deliberations that are involved. Of all inflation-targeting central banks, Norges Bank is perhaps the one that has come furthest in terms of the ambition to clarify the way policy takes the real economy into account, regularly publishing estimates of the output gap parallel with inflation forecasts as well as discussions of how policy takes account of it. It seems an interesting path to take, but at the same time it is evident – on the basis of my ten years' experience – that far from all the considerations we have faced would be captured by this model.
Concluding remarks
By way of conclusion, there are a number of issues on which we need to evolve our thinking. I believe that the contributions to this conference have indeed succeeded in capturing the most important and relevant ones.
One of the most central from my point of view concerns, as I have indicated, the balance between inflation stability and output stability – both in terms of the most appropriate way to accommodate it in the practical policy and the best way to communicate it. Although there is hardly any doubt that "we are all flexible inflation targeters now", how can we progress as regards concretising this flexibility in action and communication?
Another interesting question is whether inflation targeting is better suited than other regimes when it comes to anchoring long-term inflation expectations. I have always thought so myself, but it is of course an issue that requires empirical investigation. One might also ask whether inflation targeting, even if it should have succeeded in anchoring long-term inflation expectations, can become better than it is today at facilitating the formation of expectations in the near and medium term?
This relates to an issue that is addressed in several of the contributions, namely how we should best communicate our expectations and plans regarding the future conduct of policy. Recently, a number of central banks have begun to depart from the assumption of an unchanged policy rate in their forecasts, choosing instead to work with paths for future interest rate developments. What kind of interest rate paths should we use in that case – "exogenous" paths based on implied forward rates or "endogenous" paths that reflect the central bank’s own expectations and plans, and what are the advantages and disadvantages of both options? And could, for example, an excessive degree of openness and detail as regards the central bank's own expectations risk increasing market volatility instead of facilitating economic agents’ planning?
These are just a handful of the issues that are addressed in the contributions to the conference and that I hope will provide the basis for a rewarding discussion.
Once again – you are all very welcome!
Notes:
1) For an overview of what inflation-targeting central banks say about stabilising output, see Kuttner, K.N. (2004), "A Snapshot of Inflation Targeting in its Adolescence" in Kent, C. and S. Guttman, The Future of Inflation Targeting, Conference Volume, Reserve Bank of Australia.
2) See, for example, Fischer, S. (1996), "Why Are Central Banks Pursuing Long-Run Price Stability", in Achieving Price Stability, symposium sponsored by the Federal Reserve Bank of Kansas City, Jackson Hole, Wyoming. A recent reference that has brought the problem to the fore once again is Faust, J. and D.W. Henderson (2004), "Is Inflation Targeting Best-Practice Monetary Policy?", Federal Reserve Bank of St. Louis Review 86(4), 117-143.
