Letter: Statistical reliability

Sir: Gavyn Davies's article 'The costly consequences of dodgy data' (11 October) covers a number of issues concerning the reliability of economic statistics, including the accuracy of recent trade figures. In releasing early estimates from the new Intrastat sytem, which has been introduced this year throughout the European Community, we have stressed that the figures are especially prone to revision while the new system settles down. However, we consider it important to let users have the results as quickly as possible rather than wait until all the effects of the change can be fully evaluated.

Gavyn Davies acknowledges that a number of improvements have been made in the estimates of gross domestic product. He also identified several factors which particularly complicate the task of estimating growth in the economy, including the rapidly changing structure of output and differing rates of inflation. Steps we have taken recently are intended to help in dealing effectively with these issues, and we will continue to search for the best ways of ensuring that our estimates are both reliable and timely.

We are grateful to Gavyn Davies for drawing attention to how economic data are interpreted, and we are willing to listen to the views of all users in industry and commerce, and elsewhere, on how best to improve our service to them.

Yours sincerely,

D. C. L. WROE

Deputy Director of National

Accounts,

Central Statistical Office

London, SW1

12 October