ECB Watchers conference: monetary policy and the role of the price stability definition
Panel speech by Prof. Otmar Issing, Member of the Executive Board of the European Central Bank, Milan, 10 June 2002.
The ECB Watcher's conference, now at its fourth edition, provides a unique forum where academics, central bankers, financial market participants, journalists discuss the single European monetary policy, among themselves and with the ECB. The ECB is, to my knowledge, the only central bank that actively participates in meetings of this kind. We see this as part of our responsibility to maintain an open dialogue with the community of experts and the public opinion in general. But, in addition, such conferences are useful to us as well. Exposure to critical thinking about what we do is for us a challenging but rewarding exercise.
This conference takes place for the first time outside of Germany, signalling an effort to bring the debate about the single monetary policy as close as possible to all segments of the European public. Bocconi, the university linked to the names of Luigi Einaudi, Paolo Baffi, and more recently Mario Monti, is a perfect choice for this first occasion. I thank this university, and profs. Giavazzi and Favero in particular, for hosting this meeting.
In the first edition of the conference in May 1999, the ECB strategy was first presented and discussed in public. Many things have happened and many things have changed since then. The uncertainty we faced after the changeover to the euro in 1999 was much higher than that normally faced by central banks. Not by coincidence, the first academic conference the ECB organised was titled "Monetary policy-making under uncertainty". The ECB faced many critical junctures: deflationary influences from emerging markets, Y2K, significant changes in oil prices, September 11, just to name a few; and, most recently, the introduction of euro notes and coins. The ECB successfully tackled these challenges. The technical infrastructure of the single monetary policy has worked well, and the ECB strategy has provided an effective framework for the Governing Council's debate and decisions. Our measures of inflation expectations tell us that the markets have confidence in the long run stability of the euro. The ECB has gained credibility together with experience. Now we can look back, surely not with complacency, but with some satisfaction for what was accomplished.
The reports that have been presented to us this morning suggest that the understanding of the ECB's monetary policy has improved, and that our policy decisions have proved to be broadly correct. Nonetheless areas of disagreement remain on certain key aspects of the ECB strategy, not only with us, but also among our watchers. I would like to focus my remarks today on two of them: the Governing Council's definition of price stability; and the two-pillar structure of our strategy and its relation to other monetary policy strategies.
Defining Price Stability
The Treaty assigns the ECB the primary and overriding objective of maintaining price stability in the euro area. This was clearly intended from the start to mean neither prolonged inflation nor prolonged deflation. The Treaty indicates that, for purposes of evaluating price stability, inflation shall be measured by means of a consumer price index. Beyond this, no other guidance is given in the Treaty on how to make the notion of price stability operational. But several indications, from both theory and central banking experience, were available to the Governing Council when the price stability definition was discussed in late 1998.
First, the lags of monetary policy transmission on prices (as indicated also by recent research on monetary transmission that the ECB has conducted with the National Central Banks) are long and complex, hence no fine tuning of prices is possible. This suggests that the ECB's monetary policy needs to be firmly geared to maintaining price stability over the medium-term, pinning down trend inflation, rather than trying to counteract short-term fluctuations around the trend. It also suggests that the ECB keeps a forward-looking policy stance, looking through temporary changes in inflation, but acting decisively if signs of persistent deviations from price stability (either in the upward or downward direction) emerge.
Second, the experience of the 1990s shows that European central banks have successfully brought inflation under control by focusing market expectations on a medium term inflation objective below 2 percent. This was therefore a natural goal to adopt for the ECB as well. It is also worth recalling that at the end of 1998, the euro area inflation rate, as measured by the year-on-year increase in the HICP for the euro area, stood at 1 percent.
Third, the experience and analyses of recent years have shown that the accuracy of price indices is not complete, due to the presence of measurement biases, and that inflation and interest rate levels excessively close to zero entail risks of deflation and reduce the scope for monetary policy action. Hence, the Council adopted a price stability definition, based on the harmonised definition of consumer prices for the area, of below 2 percent, to be maintained over a medium term horizon, maintaining flexibility as to the lower bound.
As a side remark, I note that in the one report it is stated that "the mandate of the ECB obliges it to be concerned about output and employment". In fact, the Maastricht Treaty does not give the ECB direct responsibility for any additional objective other than price stability. Therefore, it is not appropriate to refer to a 'double mandate' of the ECB. The letter of the Treaty can be seen as indicating that the overriding objective of maintaining price stability should not be pursued independently of the general economic policies in the Community, which include, inter alia, "sustainable and non-inflationary growth respecting the environment", "a high level of employment and social protection" and "raising the standards of living and quality of life". The ECB's strategy with its commitment to achieve price stability over the medium term can contribute to this wider objectives by creating favourable conditions for growth, employment and welfare.
Recently, our definition of price stability has been criticised by some international organisations and academics. Three main arguments have been made:
The definition is asymmetric, focusing on an upper bound and giving no indication of a lower bound.
The definition is excessively ambitious, hence reducing the room for manoeuvre of monetary policy and creating risks of deflation.
It leaves insufficient room for relative price movements, particularly among countries with different economic structures.
Let me review each of these criticisms. As mentioned before, the Governing Council has defined price stability as an increase in harmonised consumer prices of below 2 percent. From the start, the ECB has made it very clear that the word "increase" implies that declines in the HICP are inconsistent with the definition of price stability. The definition is therefore symmetric.
One reason for aiming to achieve low measured rates of inflation (rather than zero inflation) is the possibility that conventional price measures may overstate inflation. If this is the case then a policy which achieves a constant measured price level may actually lead to a decline in the "true" price level. There is a wide range of possible measurement biases affecting consumer prices, each of which can potentially be substantial.
A recent CEPR/ECB Workshop on Measurement Issues in Prices Indices, held in November last year, provided a forum where economists and statisticians could discuss recent research on such measurement issues in the euro area. One conclusion from this workshop is related to the potential bias in the HICP. Somewhat surprisingly there is no clear notion on the sign of the bias, and it should also not be regarded as constant over time, but should rather be understood as changing with the development of the economic environment.
A second argument for small positive inflation rates relates to the lower zero bound on nominal interest rates. This is certainly not a new idea --- for example, in 1954 William Vickrey already pointed to this problem () ---, but the Japanese experience for much of the past decade has revived interest in the problem. If inflation and nominal interest rates are low it will become more likely that in the event of a large deflationary shock, causing prices and output to fall, the reduction in real short-term interest rates needed to stabilise the economy would be constrained by the fact that nominal rates cannot fall below zero. As a consequence, the economy may be subject to greater instability as it experiences higher variability in both inflation and output and its long-term output performance may even worsen.
Research has not clarified yet the extent to which pursuing an inflation objective closer to zero indeed implies substantially larger risks of getting trapped in a deflationary spiral. What is clear is that a small positive inflation rate, say between 1 and 2 percent, substantially reduces such potential risks. For example, some research, discussed at the First ECB Central Banking Conference on Why price stability?, suggests that the zero-bound constraint would in practice only rarely be encountered in the euro area (). On the basis of the shocks experienced in the euro area in the past and given estimates of the equilibrium real interest rate, most of the problems associated with the lower bound on nominal interest rates could be avoided for rates of inflation as low as 1%. It is important to note that also here the credibility of the central bank plays an important stabilizing role by maintaining stable inflationary expectations.
Finally, on the third line of critique. It has been suggested using, for example, estimates of the Balassa-Samuelson effect, that our definition of price stability leaves insufficient room for relative price movements, particularly among countries with different economic structures. This may for instance be the case in countries with low initial productivity levels that are "catching-up" with more advanced economies. In the case of similar wage developments in the tradable and non-tradable sectors, this will result in higher inflation in the countries with larger productivity growth differentials. Several remarks can be made on the underlying conditions for this scenario. For example, it is difficult to measure such effects, mainly due to difficulties in isolating them from other explanatory factors for inflation differentials. The results are sensitive to which measure of productivity that is used, length of sample data, as well as the definition of tradable and non-tradable goods sectors. Given these and other caveats, it still seems that such effects have a small impact in the euro area and that the maintenance of price stability as currently defined would not create excessive rigidities or prevent internal adjustment within the area. These effects may be more important for accession countries. The overall impact on the euro area should be limited, given the relatively small size of their economies. In addition, one should remember that accession countries will have to fulfil the convergence criteria in order to adopt the euro which in particular requires the sustainability of a low inflation environment.
Reviewing the various arguments in favour of a small positive inflation rate and their application to the euro area, it is clear that there is still a lot of uncertainty about what may happen when HICP inflation moves down towards zero percent. In view of these uncertainties, the central bank should be vigilant when inflation falls towards an excessively low level, say below 1%. Still, it needs to be stressed that what matters are not monthly changes in HICP, but projections for the trend of HICP inflation. The implication of all this is that the ECB would act in a timely and pre-emptive way if indications of either a deflationary or a significant inflationary trend would emerge.
The two-pillar strategy
The second topic I would like to address relates to the considerable misunderstanding that sometimes still exists concerning the ECB's two pillar strategy. According to some, the ECB should change its strategy and adopt inflation targeting. Others, on the contrary, believe that the ECB in fact is already following such a strategy without calling it by its proper name.
Part of this controversy is, I believe, of semantic nature. If having price stability as a primary monetary policy objective, being forward-looking, and using all relevant information to achieve it, is what is meant by inflation targeting, then the ECB may be considered an inflation targeter and any central bank in the world should be one. But in my view inflation targeting means something else, and here more substantive differences arise.
To start with, inflation targeting central banks tend to place the inflation forecast at the centre of their analytical and communication frameworks, and use such forecast ideally as if it was an intermediate policy target. This implies that the inflation forecast is deemed to incorporate all policy relevant information, and must be endorsed by the policy making body. The economic projections that the ECB publishes do not have these characteristics or this dominant status in the policy process. First, they are called projection to stress their conditional nature, i.e. dependence on a number of technical and conditional and therefore "unrealistic" assumptions. Second, they are the work of the staff alone without input or necessary endorsement by the ECB Council. As a consequence, they do not incorporate a host of other considerations, of strategic or tactical nature, that the Council may deem important in its decisions. In fact, the Council in general and some members in particular may well disagree with certain elements of the projection supplied by the staff. This is what we mean when we say that the projection is but one of the elements used by the Council in reaching its decisions.
One question arises here. How do inflation targeting central banks incorporate in the inflation forecasts, that are by nature the result of technical analyses, carried out with the help of formal econometric tools, all the elements that are relevant for policy decisions? I should maybe not be the one to answer this question. But it seems to me that there is a risk that in trying to establish a full mapping between inflation forecasts and decisions the temptation may arise to introduce arbitrary adjustments in the forecasts, thereby blurring its technical nature. This is likely to happen particularly when the policy making body is involved. Hence, from this standpoint, a clear distinction between forecasters and policy makers is beneficial, in my view.
A second important issue concerns the role of money in the strategy. This is linked to the optimal horizon within which forward-looking monetary policy decisions should be made. Today's mainstream macroeconomic models often do not accord any explicit role to money. Much of the empirical evidence, on the contrary, suggests that money has a substantial influence on medium and long term price dynamics. Thus, macroeconomic projections produced on these basis of those models cannot claim to encompass all relevant evidence. The ECB strategy takes this into account by providing a robust framework, organised into "two pillars".
The first pillar assigns a prominent role for money. In recognition of the fundamentally monetary origins of inflation in the medium to long term and on account of the stable relationship between money and prices observed in the euro area, the analysis conducted under the first pillar ensures that monetary and credit developments are appropriately taken into account. In parallel, the second pillar ensures that other analytical frameworks, such as those emphasising the interplay between supply and demand as well as cost-push dynamics, that normally dominate in the short to medium run, are incorporated into the policy analysis of price developments. Jointly, the two pillars ensure that information and analyses are always cross checked and give rise to robust policy decisions.
Besides contributing to our internal analysis and to external communication, the two pillar strategy has also helped establishing an orderly debate in the Governing Council, by providing a framework within which different views on the monetary transmission process can be encompassed. Looking back to almost 3 ½ years since our monetary policy strategy was first implemented I find that it has served well.
Let me conclude by saying that differences in strategy among central banks should not be exaggerated. In practice, all central banks analyse a wide range of information, including money and credit developments. For example, also the Bank of England's regular Inflation Report starts with a chapter on the analysis of money and credit. All central banks make intensive use of forecasting tools, but at the same time devote great attention to the analysis of individual indicators. In no case are policy decisions driven exclusively by a single forecast number or indicator, whether it is an inflation forecast or M3 growth. In the end, what counts is that the central bank achieves its objectives by making the correct decisions and by thereby maintaining or (if necessary) building up its credibility. We believe that in the face of the unique historical challenge of European monetary unification, the ECB's strategy with its focus on maintaining annual HICP inflation in the euro area below 2 percent and with its two-pillar structure provides an efficient framework to achieve these goals. In any case, our reflection on the ECB strategy continues and we will keep paying due attention to the experts' debate, such as the one taking place today in this conference.
 See, e.g., Vickrey, W.S. (1954), reprinted in Arnott, R. and Dreze, J. (eds), Public Economics: Selected Papers by William Vickrey, Cambridge University Press: Cambridge.
 See, e.g. Vinals, J. (2000), "Monetary Policy Issues in a low Inflation Environment", in Garcia Herrero, A., Gaspar, V., Hoogduin, L., Morgan, J. and Winkler, B (eds), Why price stability?, First ECB Central banking Conference: European Central Bank, Frankfurt.