MESSAGE TO INFLATION TARGETING CENTRAL BANKS:
‘SAY WHAT YOU DO AND DO WHAT YOU SAY’

Transparency and commitment are essential to any independent and inflation targeting central bank - and even more so for the European Central Bank (ECB), which, unlike most central banks, will set both its goals and the instruments by which it aims to achieve those goals. That is the message of Professor Mike Artis, Paul Mizen and Zenon Kontolemis, writing in November’s Economic Journal. ‘Say what you do and do what you say’, they urge the ECB and other central banks, whatever the precise objective of your monetary policy and whatever the chosen target.

Choice of operating target: The ECB seems to have opted for a combination of monetary targeting and inflation targeting. This contrasts with pure inflation targeting, the regime that many newly independent central banks have chosen to operate, including the Bank of England. Some interpretations of the Bundesbank’s policies suggest that it too is a covert inflation targeter: not only has it delivered a low inflation outcome, but in justifying uncorrected ‘misses’ of its monetary targets, it invariably points to its low inflation achievements.

Choice of specific inflation target: An important technical issue for the ECB is the choice of price index to be targeted. The usual choice has been some version of the consumer price index (CPI), with the exclusion of lumpy and erratic terms and the deletion of any component directly reflecting the monetary policy instrument. Thus, in the UK, the target is the retail price index excluding mortgage interest payments. Artis et al consider additional exclusions of erratic items and alternative smoothing methods: choosing the median not the mean; focusing on a moving average; and extracting a ‘core’ component. The key is to avoid making monetary policy hostage to an erratic, jumpy inflation series; policy should focus on controlling the underlying rate.

For the ECB, the choice of CPI might not be so obvious. A wholesale price index (WPI) - which corresponds more closely to traded goods where the market should ensure the ‘Law of One Price’ - might be preferred. This is because a large proportion of any CPI consists of the prices of non-traded goods and services, which may evolve in quite different ways across the EMU. However, WPI inflation is notably more volatile than CPI inflation and, on these grounds, makes a bad target.

Policy transmission lags and inflation forecasts: A problem with inflation targeting is that because of the lags in policy transmission as well as the uncertainties of forecasting, monitoring the central bank’s policy is difficult. It has become fashionable to argue that the intermediate variable of inflation targeting is the inflation forecast. If the private sector’s ‘unconditional’ forecasts of inflation (forecasts which allow for the central bank’s future policy responses) lie within the target range, the implication is that the market believes the central bank is ‘doing a good job’. But this means that the bank can learn nothing about future inflation from the market, and the private sector cannot provide a professional forecasting standard against which the central bank’s predictive competence and policy performance can be assessed.

It is important that a central bank’s technical forecasting expertise is maintained by peer group pressure. These researchers advocate that, at a minimum, full information about its forecasting procedures should be made available. Bank of England practice involves the presentation of forecast inflation ‘fan charts’, illustrating the uncertainties surrounding a forecast. The advance that this represents is qualified by some doubts as to whether potential users can handle this type of information efficiently.

Transparency and commitment: The choice of operational strategy exposes some important issues of transparency and commitment. The inherent lack of transparency of inflation targets - due to the lags and forecast uncertainty - has resulted, in practice, in a compensating supply of information in countries using inflation targets: minutes of meetings; details of the forecasting process and the bank’s reaction function, etc. The Bank of England’s experience suggests that too much ‘openness’ is also possible: the revelations of voting behaviour on the MPC have provided an unwelcome diversion.

Inflation targeting central banks typically have independence to use their policy instruments to pursue goals laid down by government: instrument-independence, but not goal-independence. The ECB has, by statute, both goal- and instrument-independence. This is perceived to have two merits: the EMU governments have pre-committed not to meddle in monetary policy; and the central bank is accountable. In the European context, the statutes give the ECB a huge degree of formal independence, but the ‘implicit rules’ of democratic governance require it to give an account of itself.

Moreover, the ECB needs a constituency of support to make a strong and effective bulwark of its formal independence from governments. It must in some way offer to make itself accountable to the ‘people of Europe’ - perhaps in practice to the European Parliament. Fortunately, inflation targeting as a regime requires many of the same accoutrements of transparency with which the ECB will probably need to arm itself.

Note for Editors: ‘Inflation Targeting: What can the ECB Learn from the Recent Experience of the Bank of England? by Mike Artis, Paul Mizen and Zenon Kontolemis is in the November 1998 issue of the Economic Journal. Artis is at the European University Institute; Mizen at the University of Nottingham; and Kontolemis at the IMF.

For Further information: contact Paul Mizen on 0115-951-5479 (fax: 0115-951-4159; email: paul.mizen@nottingham.ac.uk); or RES Media Consultant Rom