Employment and commuting
Appendix C: Data sources
Annual Business Inquiry (ABI)
The Annual Business Inquiry, conducted by the Office for National Statistics. This superseded the Annual Employment Survey. It provides very high, often 100%, coverage of large employers and large samples of smaller employers. It provides information about employment workplaces, and about employment at these places – gender, part-time and full-time, occupations, industrial sector.
Its major problems appear to derive from errors in completion, and errors in coding of industry and place. These particularly affect large companies with several major offices, or whose ownership / head offices have changed. Also, there appears to have been some discontinuity between the Census of Employment and the Annual Business Inquiry, with the latter showing higher employment totals than the Census of Employment.
The core purpose of the Census of Population is to gather data on residents on Census day. It provides the most detailed and comprehensive data available on people’s economic activity, qualifications, employment, occupation, hours worked, travel to work, etc.. This data is not comparable with the Annual Business Inquiry data, which is based on workplace as distinct from residence. However, Census forms are also processed to provide data on the workplace of people, from information they have provided. This provides the only near-100% data available on patterns of travel to work, including modes, purposes and distances. The key limitation is that this data is only available every 10 years, and that the residence based data is not available until 2/3 years after the Census date, and the workplace data is even later.
Census data suffers from some of the same problems as the Annual Business Inquiry, in that it depends for its accuracy and completeness on people filling it in fully and carefully. In 1991, where someone’s workplace / destination was missing or could not be deciphered, the data was excluded from the Census Travel to Work figures, with the result that of about 85,000 Calderdale residents in employment, around 7,000 did not have an identifiable destination. By contrast, the 2001 Census has imputed missing records to ensure that it provides data for all people employed.
Additional problems have arisen in the 2001 Census due to the adjustment method used to ensure confidentiality and that no-one can be identified through the Census data. The method used has repeatedly led to somewhat different totals being available, according to how they have been derived in the Census analysis. Although this can lead to some confusion and raise questions about whether errors have been made, the differences are not of a magnitude to alter the message emerging from Census data.
To allow comparability between 1991 and 2001, it has been necessary to make assumptions about the 1991 missing records – see Appendix D.
Learning & Skills Council (LSC) Household Survey
The LSC West Yorkshire commissions large household surveys that provide detailed information on occupation, qualifications, in-work training, etc.. Recent surveys covering Calderdale have achieved responses from about 1,000 people, giving a good level of accuracy.
Labour Force Survey (LFS)
The Labour Force Survey is a rolling survey conducted by a combination of initial visit and regular phone calls, with a rolling panel. It enables Government to produce estimates of economic activity levels, full-time and part-time employment, and ILO-compliant (non-claimant) estimates of unemployment levels and rates.
The major problem of the LFS is that, for a district of Calderdale’s size, the sample is too small to provide robust data.
Household income data
Data on household income is taken from a commercial database entitled Paycheck, produced by CACI Ltd, a leading market analysis consultancy. Household income figures are gross figures inclusive of income from all sources (employment, benefits, pensions etc.).
The data produced by CACI are based upon market research statistics generated from a wide range of “lifestyle” surveys (by mail shot to households / individuals, or with purchases / guarantees). These are combined with the latest data from the Census and population estimates, and with geodemographic classifications, and then weighted in line with the Family Expenditure Survey. Modelling the results enables derivation of estimated income ranges, as well as mean, median and modal incomes, down to unit postcode level.
For each postcode, Paycheck provides a median income within a £ 5,000 range (the median income is the income of the middle ranking record). For the purpose of this report, more precise figures have been estimated by aggregating the postcodes and household numbers in different median ranges. The distribution of households in the £5,000 range below and above the median was then analysed, to enable a rough estimate of the actual median income.