The ECB's quantitative definition of price stability and its comparison with such definitions or inflation targets applied in other large economic areas
Letter of Dr. W. F. Duisenberg, President of the ECB to the Chairperson of the Committee on Economic and Monetary Affairs, Mrs. Christa Randzio-Plath
Dr. W. F. Duisenberg, President
Mrs. Christa Randzio-Plath
Chairperson of the Committee on Economic and Monetary Affairs
16 October 2001
Testimony before the Committee on Economic and Monetary Affairs of the European Parliament on 12 September 2001
Dear Mrs. Randzio-Plath,
Further to my testimony before the Committee on Economic and Monetary Affairs on 12 September 2001, please allow me to come back to your enquiry on an important element of the ECB's monetary policy strategy, namely the quantitative definition of price stability, and its comparison with such definitions or inflation targets applied in other large economic areas.
The ECB's definition of price stability operationalises the Treaty mandate attributed to monetary policy, which is to maintain price stability. By setting this definition, the ECB quantified its understanding of the primary objective set by the Treaty. The ECB's definition followed on from, and is, in fact, fully consistent with, definitions adopted by national central banks in the euro area prior to Stage Three of EMU. For example, the Deutsche Bundesbank, in setting its monetary target for 1997- 98, adopted a medium-term price assumption of 1.5% to 2%. The Banque de France specified 1998 price stability as a CPI inflation rate not exceeding 2% over the medium term. A similar objective was adopted by the Banca d'Italia for 1998. Moreover, when adopting the Broad Economic Policy Guidelines in July 1995 ECOFIN indicated that a value of 2% would be the maximum rate of inflation compatible with price stability. This was implicitly reconfirmed in the 1998 Guidelines.
All this demonstrates that the ECB's definition of price stability matches a clear demand for low inflation by citizens in the euro area, as embodied in the framework of national central banks prior to the start of Stage Three. By maintaining continuity with past definitions of price stability in the euro area, the ECB respected existing public expectations of long-term price stability. This enhanced the ECB's reputation and credibility and public support for EMU more generally.
I would also like to stress that the numerical value set for the inflation target in the United Kingdom is not very different from the definition of price stability of the ECB. The target in the United Kingdom is set at 2.5% for the retail price index excluding mortgage payments (RPIX). When comparing this target with that for the euro area, it must be taken into account that the RPIX has historically increased at a significantly faster pace than the Harmonised Index of Consumer Prices in the United Kingdom. The difference amounted to ½ of a percentage point over the 1990s and it has increased more recently to levels of around ¾ of a percentage point. Thus, if the inflation target for the United Kingdom were to be specified for the HICP, one would, ceteris paribus, expect it to be in the order of 2% or below.
Central banks of the largest economic areas outside Europe have neither set an inflation target nor provided a quantitative definition of price stability for their own country. In both the United States and Japan, so far only qualitative definitions of price stability have been provided to the public.
The ECB's quantitative definition of price stability reflects sound and well-established economic criteria. By allowing for only low rates of increase in the price level, it permits the minimisation of the costs of inflation, which are well-known to the general public and which have been extensively documented in the economic literature. At the same time, by providing a numerical value, our quantitative definition of price stability helps to anchor inflation expectations, thereby reducing long- run inflation uncertainty and the associated risk premia and distortions. The effectiveness of this approach is very visible in measures of long-term inflation expectations such as the spread between yields on long-term index-linked bonds and yields on nominal bonds in the euro area. In fact, such measures have been relatively stable since the start of Stage Three within the range of levels of price increases falling within our definition of price stability. Furthermore, having a quantitative definition of the objective of monetary policy is important for the ECB's accountability and enhances the credibility and effectiveness of monetary policy.
In setting the numerical value for our definition of price stability, three considerations also had to be taken into account. First, the Treaty mandate of maintaining price stability, for sound economic reasons, does not allow us to exclude, in an arbitrary manner, low positive numbers for increases in the price level from the definition of price stability. In this sense, it would not have been appropriate to exclude inflation rates below a certain value, say 1.5%, from the definition of price stability. At the same time, in view of the potential existence of a measurement bias in the HICP inflation, we did not deem it appropriate to specify explicitly 0% as the lower boundary for the definition of price stability. Taking this into account, we have decided not to exclude low positive inflation rates from the definition, but deliberately to leave the lower (positive) boundary of the definition of price stability open.
Secondly, there are strong arguments for setting a relatively narrow range of price increases admissible over the medium term. Only then is the definition effective in anchoring inflation expectations - thus fully reaping the benefits of price stability.
Thirdly, setting an upper boundary of 2%, the definition should allow for a sufficiently large safety margin to ensure that the potential risks associated with deflation are avoided. As I have emphasised at previous meetings with your committee, the ECB's definition of price stability precludes inflation as well as deflation.
In assessing our approach, it is also important to recall that the Governing Council has explicitly adopted a medium-term orientation for the conduct of the single monetary policy. This, inter alia, acknowledges the existence of shocks to inflation which monetary policy cannot control, implying that it is not always possible to keep inflation consistent with the definition of price stability every month or quarter. A medium-term orientation also reflects the need for the ECB to pursue its objective by responding differently to different economic shocks, in particular avoiding too harsh a response to some types of shock in order not to introduce unnecessarily high variability in real activity and interest rates.
Finally, it is crucial to bear in mind that, over the longer run, monetary policy cannot directly affect real variables, such as the level of real interest rates or the rate of growth of output. Allowing for a higher rate of inflation would not imply a lower level of the real interest rate and/or help the economy to grow faster. On the contrary, higher levels of inflation are typically associated with higher inflation variability and the related economic distortions and risk premia in real interest rates, which tend to have a negative impact on real activity. With our definition of price stability the ECB should reduce such distortions and risk premia to a large extent and thereby favour the efficient allocation of resources and create an environment conducive to enhancing welfare and employment.
With best regards,
Willem F. Duisenberg