Retail Prices Index (Hansard, 11 December 1987)

§ Mr. Fowler

The November monthly figure for the retail prices index shows that there has been a welcome fall in the annual rate of inflation to 4.1 per cent. It has, however, been discovered that from February 1986 to October 1987 a computer programme error affected the monthly index. As a result, the annual inflation rate will have been understated by one tenth of 1 per cent. in most months during that period.

Each month 130,000 prices, covering many categories of expenditure, are collected and analysed in a complex set of calculations. Following a decision in March 1983 the index has since February 1986 been calculated by computerised methods. Monitoring of the system has revealed that during the programming work in 1985 a mistake was made in the detailed calculations for certain items.

From February 1986 until January 1987 there was a regular understatement of about 0.06 per cent. in the price change in each month relative to the base month, which was January 1986. From February 1987, as the faulty programme was applied to a wider range of items, the understatement relative to January 1987 was about 0.09 per cent.

The programming fault has now been rectified, and the annual inflation rate of 4.1 per cent. for November is correct. The figure for the November level of the index is also correct at 103.4. It has not, however, been possible in 284W the time available to calculate the figures for the many detailed components. These will be published in the course of next week.

The RPI will have been understated, albeit by a numerically small margin, for up to 21 months. I wish, however, to make it plain now that special extra payments will be made to national insurance retirement and supplementary pensioners and severely disabled people, and that the Exchequer will not benefit from the effects on social security expenditure. My right hon. Friend the Secretary of State for Social Services will make an announcement on this as soon as possible.

We believe that the computerised calculation system is now wholly correct. In view of the RPI's importance, however, I am asking the head of the Government statistical service to carry out a review of it.