§ Mr. DonohoeTo ask the Secretary of State for Scotland what was the interval between holding and publishing the results of the census in(a) 1981 and (b) 1991; and if he will make a statement on the validity of the results of the 1991 census.
§ Mr. Stewart[holding answer 17 June 1993]: Census results are published as a series of monitors and fuller reports for each region and for Scotland as a whole, followed by a series of reports for specific topics—mainly for Great Britain but with some Scottish reports. Comparison with the 1981 census publication dates is given in the table.
757WOwing to some new questions and more detailed analyses, the 1991 Census publications contain roughly twice as many statistics as their 1981 equivalents.
1Region reports were published in four parts per region for 1981 and two parts for 1991.
2Planned dates.
Computer processing of 1991 data was delayed as a result of a problem discovered during a routine data quality audit. Some 400,000 people in Great Britain, who were apparently not students, had completed the question on term time address as though they were students, causing them to be wrongly allocated in initial processing. More information about this problem is given in "Census Newsletter No. 18", a copy of which is in the Library. The full publication programme is still expected to be completed more quickly than was the case for the 1981 census.
On the validity of the results, the Registrar General included a statement entitled "1991 Census under-enumeration: guidance for users in Scotland" as Annex A to the 1991 "Census Monitor for Scotland", a copy of which is in the Library. This indicated that the main 1991 census tables are in fact more complete than the equivalent 1981 tables and that the effects of under-enumeration in the 1991 census are likely to be unimportant for the majority of users.
Further information on the coverage and quality of the census results will be published in a series of OPCS reports based on analysis of the census validation survey for Great Britain.