EMPLOYMENT STATISTICS (Hansard, 5 March 1986)

§ The Secretary of State for Employment (Lord Young of Graffham)

Employees in employment, the self-employed and HM Forces are estimated separately, and together comprise the employed labour force. Since 1979 there have been changes in the methodology for estimating the first two elements of the employed labour force. In 1979 the estimates for employees in employment were based on the census of employment, updated by applying proportionate changes in the numbers of employees as estimated from sample surveys of employers. Estimates for the self-employed were obtained from the census of population, updated to 1975 by applying proportionate changes in self-employment from counts of national insurance cards; self-employment was assumed unchanged since 1975.

The first change in methodology since 1979 was the use of labour force survey (LFS) results to produce new estimates of self-employment for 1975 to 1979. These were published in the January 1982 Employment Gazette. The LFS data were used in place of the discontinued information from the national insurance card count.

When provisional results from the 1981 census of employment became available at the end of 1982 they showed that the application of changes estimated from sample survey data to the previous, 1978, census figure was producing substantial underestimates of the number of employees in employment. The methodology was reviewed and a supplementary set of estimates, which included an undercounting allowance based on the average rate of shortfall which had developed between 1978 and 1981, was introduced. At the same time the conventional assumption that the level of self-employment had remained constant since the date of the latest LFS data was reviewed. As there were reasons for expecting some continuation of the upward movement in self-employment, a supplementary series which assumed a continuation of the rate of growth observed between the latest two surveys was introduced. The reasons for, and full details of, these changes in methodology were set out in the June 1983 Employment Gazette.

Data from the 1983 labour force survey confirmed that the supplementary figures provided more accurate estimates than the basic series. The estimates for employees in employment were revised, using LFS data for 1981 and 1983 to assess the current extent of underestimation from the sample survey of employers. The self-employment series was updated at the same time. As the department could now produce estimates in industrial and regional detail incorporating the adjustment for underestimation, estimates not incorporating the adjustment were no longer published and the term "supplementary" was no longer used. An article explaining the basis of the new 293WA estimates was published in the July 1984 Employment Gazette.

When the estimates of both employees in employment and the self-employed were updated to take account of results from the.1984 LFS and revised data from the 1983 LFS the figures for self-employment showed exceptional growth between 1983 and 1984 and the department's statisticians considered it inappropriate to assume that this rate had continued. The estimates of self-employment for dates after June 1984, which will be reviewed when the 1985 LFS data become available next year, now incorporate the assumption that average rate of increase between 1981 and 1984 is continuing. An article describing these estimates was published in the March 1985 Employment Gazette.