§ 1. I was asked by the Home Secretary to enquire into the error which resulted in double counting of some passengers embarking from London Airport (Heathrow) in 1973 and part of 1974. This report (Report I) was submitted on 28th January 1976. I was then asked by the Home Secretary to enquire into the significance that should be attached in general to the figures of admissions at and embarkations from UK air and sea ports. This report (Report II) was submitted on 29th March 1976. The two reports are separate though clearly related.
§ 2. The enquiries were carried out rapidly and with few resources which must mean that their conclusions are not absolutely firm. In particular it was only possible to examine the procedures in use at three ports and even that could not be done at the busiest time of year when the procedures for compiling statistics are under the greatest pressure.
§ 3. The statistics being enquired into are of admissions and, more particularly, embarkations and not of migration.
§ 4. As to Report I, the error dealt with affected the count of those non-patrial Old and New Commonwealth citizens leaving Heathrow Airport from January 1973 to July 1974 who had previously been admitted subject to special conditions about their period of stay. On 1st January 1973 the cards used in order to check their embarkation against the records of their previous entry were changed; these cards should not have been counted for compiling the statistics. The change in the procedure was not mentioned in new instructions dealing with statistics. At ports other than Heathrow, cards were counted at the port concerned and Immigration Officers either 637W counted the cards themselves or worked alongside clerical staff doing so. Heathrow cards were counted at Harmondsworth, but the responsible Executive Officer remained a t the airport. At Harmondsworth the mistake made was that both the cards used to check the embarkation of controlled Commonwealth citizens against the records of their previous entry and the statistical cards were counted. The error was discover at Harmondsworth in July 1974. Double counting stopped on 1st August. When the error was found, incorrect annual data for 1973 had already been published and priority was given to estimating the number of Commonwealth citizens who embarked in 1974. Adjusted estimates were published in the 1974 White Paper, but they were not rounded nor was any mention made of the facts that the 1974 figures had had to be adjusted and that the published 1973 figures were wrong. Adjusted 1973 estimates—relating to New (but not Old) Commonwealth citizens —were made in the second half of 1975 and first announced in a Written Answer on 7th November.
§ 5. It seems to me clear that the error itself was the result of accidentally defective procedures and not of any desire deliberately to distort the statistics, and that the delay in announcing it was not the result of any desire to conceal. It was clearly the intention of the officials in the Statistical Department and in the Immigration and Nationality Department to find out what had gone wrong, to adjust the 1974 figures as well as possible and after that to estimate and publish an adjusted figure for 1973. It is plain from the files I have examined that these officials made no attempt to conceal the existence of the error while they were investigating its extent; it was referred to in a file seen by Ministers in February 1975, though it was not specifically drawn to the attention of the Home Secretary until nine months later.
§ 6. Although responsibility for preparing the immigration statistics from the material provided by the Immigration Service was transferred from the Immigration and Nationality Department to the Statistical Department in 1974, there remained some ambiguity about the responsibility for interpreting the statistics and for drawing the attention of Ministers to significant trends. Consultation between the two departments is of course essential, but it is now clearly understood that the Statistical Department is primarily responsible not only for the accuracy of the statistics, within the possible bounds set by the available resources and procedures for the control of immigration, but also for their interpretation and presentation to Ministers and the public. Ministerial authority for the publication of any figures of admissions or embarkations will in future be sought by the Statistical Department after consultation with the Immigration and Nationality Department.
§ 7. The main conclusion of Report II is that future figures of net balances between admissions and embarkations cannot provide a clear guide to migration unless additional and expensive information is also prepared. Some of those extensions were suggested by Home Office officials in 1966 but the procedures at 638W the ports then in use for Commonwealth citizens made them impracticable. Since March 1975 Commonwealth citizens have been dealt with in the same way as foreign nationals and those extensions would now be less difficult. They would, however, be very expensive. Many extra staff would be needed for the Immigration Service and some for the Statistical Department. Political dangers might be equally important. They could stem from the delays that would have to be imposed on travellers and from the development of something like a National Register in order to match individual embarkations with admissions.
§ 8. The other conclusions of Report II concern past figures, and their general import is that there is no way in which past net balance figures could provide a guide to migration. For 1973 and 1974 the figures were subject to double counting and I feel that the estimated adjusted figures are not reliable. Moreover, the effort to prepare wholly acceptable adjusted figures for those two years would not be justified because they would still not provide a reliable guide to migration. For those two years and for earlier years, about which the evidence of unreliability is equivocal, the additional information and analysis that the essential if net balances are to be validly compared with migration are not available and cannot now be provided.
§ CLAUS MOSER, Central Statistical Office. 5th May 1976.