Letter: Quantifying the cost of statistics

Sir: The articles and letters about UK government statistics published in the Independent in December had one striking omission: no one seemed to be much concerned about cost and the inevitable trade-off which must be faced between the cost and q uality of statistics.

To give one example, the availability of monthly results of acceptable reliability and accuracy from the Labour Force Survey might require a threefold increase in the present cost to the taxpayer, which is already several million pounds a year. Even if this assessment proved to be on the high side, substantial sums of additional public spending would be involved and the issue of whether this would be justified would have to be faced.

It is reassuring to know that these particular statistics are currently the subject of a study by the president, and two past presidents, of the Royal Statistical Society who, as good statisticians, will no doubt have this point very much in mind.

Yours faithfully, JACK HIBBERT Beckenham, Kent 5 January The writer was director, Central Statistical office, and head of Government Statistical Service, 1985-92.