§ 12. Sir H. Williamsasked the Secretary to the Treasury what is the presumed degree of error, plus and minus, in the estimated statistics contained in the documents, United Kingdom Balance of Payments, 1949 to 1952 (No. 2), Preliminary Estimates of National Income and Expenditure, 1948 to 1952, and Economic Survey for 1953, respectively.
§ Mr. MaudlingIt is impossible to state any general degree of error applying to all the statistics in the publications concerned. Each figure is the best estimate possible in the light of the information available at the time and many of the estimates can be checked by their relationship with others which are independently calculated. It is recognised that as new information becomes available many figures, especially those for more recent years, will be subject to revision.
§ Sir H. WilliamsAs one of those who, many years ago, used to have something to do with these things, and who knew the extent of error which is involved, is 1377 it not most important that the Government should not give endorsement to figures which everybody knows are substantially inaccurate, which are quoted in the House by hon. Members on both sides, but which do not represent the precise truth? Is it not time that we became more scientific?
§ Mr. MaudlingTo get a mathematical computation of the degree of error, it would be necessary to calculate the degree of both bias and random error in a vast range of original data. On the general question, I need not remind my hon. Friend, who is an expert in these matters, that statistics are a basis for judgment, and not a substitute for it.
§ Mr. Ellis SmithDoes the hon. Gentleman admit that the men and women in this Department are doing a great service to the country, especially when we need to measure the services of people more and more; and, instead of trying to get at them, should we not give credit where it is due?
§ Mr. MaudlingI am sure that the men and women concerned will be most grateful to the hon. Member for his sincere tribute to their work.