Comment regarding Statistics Sweden’s changed methods for computing the consumer price index and inflation rate

Date 05/05/2004

Statistics Sweden’s (SCB) committee for the consumer price index has decided to make a number of methodological changes in the calculation of the consumer price index (CPI) and inflation rate. The changes will come into force as from January 2005. In view of this decision, Deputy Governor Irma Rosenberg, who is responsible for preparing matters relating to monetary policy, has given the following comment:

”The changes are of a technical nature and include partly a new index construction for computing the CPI figure, and partly a change in the way that the inflation rate is calculated. The Riksbank has no objections to the announced changes in method, and the same methodological changes will also be made to the measures of underlying inflation that SCB calculates on behalf of the Riksbank. It is uncertain how large the difference in the measured inflation rate will be between the current and new measurement method, but SCB’s estimates indicate that measured inflation can be expected on average to be around 0.2 percentage points lower per year using the new method. The difference between the current and new method of calculating the inflation rate may vary substantially, however, between different years, and the new method may also result in a higher measured inflation rate in certain years,” says Mrs Rosenberg.

”The methodological changes in themselves have no effects on prices or production in the Swedish economy, but they nevertheless raise questions as to whether the Riksbank, as a result of the changes, will adjust its definition of the inflation target or in some other way allow the changes in the measured inflation rate to influence monetary policy. The Riksbank’s experiences of the CPI as a target variable for monetary policy have been positive, and the inflation target has won credibility with the public. The reason that the target is defined in terms of the CPI is, among other things, that it is the best known and most used measure of inflation. The changes that are now being implemented do not alter the assessment that inflation, measured in terms of the CPI, is an appropriate target variable,” says Mrs Rosenberg.

”That the target should be 2 per cent inflation per year is not as self-evident, but many other central banks have chosen targets around this level. After a decade with this target definition, the public’s inflation expectations are presently anchored firmly around the target. Therefore, despite the changes in method, the Riksbank judges that 2 per cent remains an appropriate target. The target for monetary policy continues thereby to be that inflation, measured in terms of the CPI, should be 2 per cent, with a tolerance for deviations around this target of plus/minus 1 percentage point. The formulation of the target thus remains intact, even if the way in which the inflation rate is measured will change somewhat due to the changes made by SCB," says Mrs Rosenberg.

”The changes in method do indeed give reason to expect a somewhat lower measured inflation rate on average compared with present inflation measures. Under otherwise unchanged conditions it could advocate temporarily somewhat looser monetary policy. To avoid irregularity in monetary policy, however, it may be appropriate to allow a purely technical change of this kind to filter through gradually. When making monetary policy decisions in the future, we will as usual consider all the other new information that could affect the inflation forecast. What the consequences will be for current monetary policy will be decided by the Executive Board of the Riksbank at its next monetary policy meeting on 27 May,” concludes Mrs Rosenberg.

For details of the changes in method and their effects on the measured inflation rate, see SCB’s website.