United Kingdom (4)

Title of the survey

Labour Cost Survey (This description mainly refers to the survey conducted in Great Britain.)

Organization responsible

The Employment Department carries out the survey in Great Britain. In Northern Ireland, the survey is carried out by the Department of Economic Development (DED).

Periodicity of the survey

Every four years. The latest survey was conducted in 1993, with reference to 1992; it is designated hereafter the 1992 survey.

Objectives of the survey

The survey is carried out under the European Community (EC) Council Regulation 1612/88, which requires member States to supply details on the level and trends of employers' total labour costs. At the EC level, the Labour Cost Survey data are used by several Directorates General, in reports on wage cost differentials, social and employment reports, etc. and in the determination of the European Social Fund. At the national level, labour cost data form an essential basis for the calculation of labour costs per unit of output. Such data are used by Government Departments, particularly the Treasury and Central Statistical Office (CSO). The CSO uses the data in the preparation and checking of national accounts. The Labour Cost Survey is also one of the few sources of data on training costs and number of long-term trainees, and the information is used by the Training Enterprise and Education Division of the Department of Employment. Universities and Employers' asssociations also show great interest in labour cost data.

Main labour topics covered by the survey

Employment, earnings, hours of work, labour cost, as well as establishment practices with regard to systems of payment for time not worked, social security and pension schemes, training schemes, etc.

Reference period

A whole year. Normally the calendar year; otherwise, the twelve months' period ending in April of the following year.

Coverage of the survey

Geographical

Figures are available for the whole of the United Kingdom, with the Employment Department carrying out the survey in Great Britain, and the Department of Economic Development (DED) carrying it out in Northern Ireland, on a wholly compatible basis.

Industrial

The 1992 Labour Cost Survey covered the following: manufacturing; energy and water; construction; retail and wholesale distribution, hotels and catering; insurance, banking and finance (including building societies, finance houses and business services); travel agencies and tour operators.

Establishments

Establishments with 10 or more employees.

Persons

All employees on the firm's payroll, who work at the designated establishment(s).

Occupations

Not relevant.

Concepts and definitions

Employment

Employees include all full-time and part-time employees on the company's payroll and Youth Training trainees who have a contract of employment with the establishment. Excluded are working proprietors, directors paid by fee only and self-employed persons working on their own account; those working in Northern Ireland or elsewhere overseas; managers paid mainly by a share of profits; home service agents paid wholly or partly by commission; family workers without an agreed wage or salary; home workers; cleaners employed for only a few hours a week; employees of employment agencies or other companies; and Youth Training trainees who do not have a contract of employment with the establishment. Out of the total number of employees, the following categories are identified separately:

employees who were apprentices or full-time trainees; this includes all students and crafts, and young persons serving a probationary period before apprenticeship;
part-time employees, i.e. employees working 30 hours a week or less.

Labour cost

Defined as employers' total costs for wages and salaries, statutory payments, voluntary payments and subsidised services to employees. Labour cost comprises:

gross wages and salaries paid to all employees during the year (before deductions for tax, National Insurance or superannuation contributions), including pay supplements, pay for overtime, night and weekend work, annual, public and customary holidays, injuries, sickness and maternity absences; pay in lieu of notice; bonuses, productivity and incentive payments, including an estimate of bonuses earned during the year but which will be paid later; commissions and gratuities; payments for time not worked, including all guaranteed payments, statutory or not; other additional allowances such as cost-of-living supplements; and payments to employees while on training courses. Excluded are repayments of expenses incurred by employees while on company business and redundancy payments;
employer's share to National Insurance contributions;
redundancy payments: all payments made under the Redundancy Payments Act; all lump sum or continuing weekly payments made directly to redundant employees and all lump sum compensation payments to employees who have been paid off or kept on short time for a substantial period;
expenditure on vocational training: costs of employee training courses; fees paid to instructors not employed by the company; costs of training materials, power and tools; depreciation of buildings and equipment used solely for training; vocational training and other scholarships; levy paid for Industrial Training Boards (less any grant received or due). Excluded are all capital expenditure on buildings and equipment used for training and wages and salaries of instructors, apprentices and trainees;
employers' premiums under the Employers' Liability (Compulsory Insurance) Act 1969: for the risk of employees suffering injury or disease arising out of their employment or in the course of it;
payments into superannuation and other pension, retirement and death funds, including group life assurance; excluded are the sums paid directly to employees or their survivors;
payments to funds to provide benefits to employees while absent from work through sickness, maternity, industrial injury or disease caused by work;
lump sum ex gratia payments and marriage gratuities, paid to employees, their dependants or other beneficiaries in respect of retirement, sickness, accident or death, long service awards, etc.; less,
contributions from central government towards payments to employees under special employment measures: i.e. in respect of Youth Training trainees with a contract of employment.

In addition, companies with more than 100 employees are asked further details about the components of labour cost, as follows:

amount of periodic bonuses not payable in each pay period,
amount of annual and public holidays and other time off with pay (excluding holiday bonus payments),
amount of payments for absence through sickness or injury,
amount of payments for absence through maternity leave,
other payments for time not worked (marriage, other family reasons, trade union activities, attendance at training courses, etc.),
amount of benefits in kind for employees or their dependants,
assistance with housing for employees,
costs of all other services for employees: canteens, staff restaurants and other refreshment facilities, first aid rooms, rest rooms and other health services, sports grounds, club houses, staff car parks, education services, etc., provision of work's buses, help with fares or season ticket loans, uniforms, protective clothing and footwear, luncheon and other meal vouchers, etc.
cost of recruitment: sums paid to recruitment agencies, expenditure on job advertisements, travel expenses paid to candidates called for interview, special allowances paid to newly recruited staff, etc.
net running costs of company cars and vans for which employees had some private use (annual leasing costs, insurance, maintenance, servicing, fuel, garaging and parking), excluding the proportion of costs attributable to their use for business travel.

Out of the total amount of wages and salaries paid to all employees, the amounts paid to apprentices and full-time trainees are separately identified.

Hours of work

Data are collected on the total number of hours worked by all employees during the year. This includes all normal working hours worked by full and part-time employees; actual overtime hours; and hours during which employees were available for work and for which guaranteed payments were made. Excluded are main meal breaks and absences due to annual and public holidays, sickness or attendance at a training course. If establishments do not have details of paid absences readily available, they are required to estimate the number of hours employees would have been expected to work during the year. Data are collected for all employees together, and out of that total, hours worked by apprentices and full-time trainees, and by part-time employees, are separately identified.

International recommendations

The definition of hours worked used in this survey is in line with the international recommendations on hours actually worked. That of labour cost conforms to the international guidelines.

Classifications

Components of labour cost / compensation of employees

The components of labour cost may be grouped according to the major groups of the International Standard Classification of Labour Cost (ISCLC-1966).

Industrial

The Standard Industrial Classification (SIC-1992) is convertible to the Statistical Classification of Economic Activities of the European Communities (NACE) Rev.1, which itself is based on the International Standard Industrial Classification of All Economic Activities (ISIC), Rev. 3.

Occupational

Not relevant.

Others

Labour cost statistics are classified by size of establishments (six size groups), and for manufacturing only, by region (11 regions are identified for statistical purposes).

Sample size and design

Statistical unit

In manufacturing, the sampling unit is the establishment, so that regional labour costs may be obtained. In the non-manufacturing sector, the sampling unit is the enterprise. Firms can choose to report at the establishment level or at the enterprise level.

Survey universe / sample frame

The 1992 sample was drawn from the 1989 Census of Employment records. These records are updated every two to three years, on the occasion of each Census of Employment. The Census of Employment covers some 300,000 reporting units, i.e. about one fourth of all reporting units in Great Britain.

Sample design

The survey is based on a stratified cut-off random sample. Stratification is by industry and size of establishments. In energy, manufacturing and construction, all establishments with 500 and more employees are included with certainty. In wholesale and retail distribution as well as in hotels, catering, travel agencies and tour operators, all establishments with 200 or more employees are included with certainty. In insurance, banking and finance, all firms with 25 or more employees are surveyed. Below these cut-off points, the sampling fraction varies by industry and employment size, from 1/3 to 1/30 (or 1/50 in construction). The stratification into size ranges follows the EC requirements, except that the smallest size range in services was split into two so as to reduce the burden on the smallest firms. The survey is conducted on the minimum sample size acceptable for statistical reliability of the results. The 1992 survey covered about 15,000 reporting units, i.e. 6,000 in manufacturing and 9,000 in non-manufacturing. This represent about 11 per cent of the total population in terms of establishments and enterprises, and 28.5 per cent in terms of employees. The banking, insurance and finance sector was almost wholly covered by approaching the relevant employers' associations for information on member companies. Similarly, data for most of the energy sector - coal, coke, nuclear fuel, gas, electricity and water - were obtained by central returns sent in by the head offices of the national organizations involved. Firms with fewer than 200 employees, which take part in one survey (e.g. in 1988), are deleted from the sampling frame of the next survey (e.g. in 1992). In 1992, completed returns with usable data were received from almost 85 per cent of the establishments surveyed. These completed returns covered over 5 million employees, i.e. about 40 per cent of employees in scope of the survey.

Field work

Data collection

The survey is conducted by postal questionnaires. Employers in the sample are first sent a specimen questionnaire in March of the reference year, so that they are informed of the forthcoming survey and prepared to set up their records in order to be able to give the required information at the time of the actual survey. Questionnaires are posted in December of the reference year for completion in respect of that year or the nearest complete financial year, and return by 1 March of the following year if using the calendar year, or by 1 June if using a financial year. Large firms receive questionnaires for more than one establishment. Reminder letters are sent to all non-respondents in March, June and September and contacts are made by telephone with large firms who did not respond.

Survey questionnaire

Two questionnaires are used: one for manufacturing, the other for non-manufacturing. The two questionnaires are identical, except that one seeks information on the particular establishment (in manufacturing) and the other information for the whole firm or organization. The 1992 questionnaires contained 16 questions on:

details of the establishment
details of employees
details of hours actually worked and
details of labour costs

The questionnaires contain information about the objectives, uses and requirements of the survey, definitions of terms used and details on inclusions and exclusions from the various components. Certain questions on the components of labour cost are not asked of small firms with less than 100 employees (see also under Estimates.

Substitution of sampling units

The sample is sufficiently large to allow for non-response due to firms going out of business. However, if coverage of an industry were to fall to an unacceptably low level, facilities exist for topping-up the sample.

Data processing and editing

Completed returns are computer-checked against credibility limits. Coding is done by specially recruited and trained staff who also look at the questionnaires rejected by the computer validation programmes. A large number of consistency and range checks are made. Any data significantly outside the credibility limits are queried with the firm, usually by phone. Missing data are imputed by computerized programmes.

Types of estimates

Labour costs are calculated on a pence per hour basis, and some on an annual total basis. Summaries are made by industrial sector, region (manufacturing) and size of establishment. Expenditure by each category of labour cost are expressed as percentages of total labour costs. In the case of small firms which are not asked to provide data on benefits in kind or subsidised services, estimates for such items are made on the basis that expenditure in proportion to wages match those in larger companies.

Construction of indices

Not relevant.

Weighting of sample results

Grossing up is carried out using data from the Census of Employment.

Adjustments

Non-response

Not relevant.

Other bias

Not relevant.

Use of benchmark data

Not relevant.

Use of other surveys

Not relevant.

Indicators of reliability of the estimates

Coverage of the sampling frame

Since 1984, the Census of Employment is in fact conducted on a sample basis. Its objective is to have about 300,000 reporting units contributing to its results, i.e. about one fourth of all reporting units in Great Britain.

Sampling error / sampling variance

Not calculated.

Non-response rate

It is about 15 per cent in terms of reporting units. It is not calculated for employment.

Non-sampling errors

Not available.

Conformity with other sources

Earnings figures are compared with those available from the New Earnings Survey. No other sources of labour cost data exist.

Estimates for non-survey years

Estimates of labour costs for the main industry groups are calculated on the basis of the Average Earnings Index, the New Earnings Survey and statistics from other sources. They use the latest information on changes between years in wages and salaries, national insurance contributions and redundancy payments, as well as the results of the latest Labour Cost Survey. Earnings and national insurance contributions make up between 83 and 95 per cent of total labour costs in the UK. However, there is no way other than by survey of estimating such costs as benefits paid in kind, welfare payments and assistance with housing.

Available series

The following standard tables are published regularly:

Summary by industrial sector and category of labour costs
Components of labour costs as percentages of labour costs
Labour cost in manufacturing: average expenditure per employee as a percentage of total labour costs
Labour cost in service industries: average expenditure per employee as a percentage of total labour costs
Labour cost by size of establishment in manufacturing: average expenditure per employee as a percentage of total labour costs
Summary by industrial sector and main components of labour cost

Additional tables are regularly prepared and available on request, such as:

Labour costs expressed as a percentage addition to wages and salaries, in production and construction industries
labour costs per employee by size of establishment in manufacturing (pence per hour)
Labour costs per employee by size of company in non-manufacturing industries
Annual labour costs in manufacturing and non-manufacturing industries: average expenditure per employee
Annual labour costs in manufacturing and non-manufacturing industries: average expenditure per employee, part-time employees converted to full-time equivalents
Annual hours worked per employee; all industries
Regional labour costs in manufacturing: annual average expenditure per employee

Other tables are prepared by the Statistical Office of the European Communities (EUROSTAT), which include average monthly and hourly labour costs, in national currency and in ECU and average number of hours actually worked during the year.

History of the survey

The United Kingdom first took part in the EC labour cost survey in 1973, although two pilot surveys had been held earlier, in 1964 and 1968. Up to 1984, these surveys were carried out every three years, but the frequency has been reduced to every four years. The most recent survey was conducted in 1993, with reference to 1992. The 1973 survey covered manufacturing, mining, construction, gas, electricity and water. The coverage was expanded in 1988 to cover the banking, insurance and finance sector, as well as wholesale and retail distribution. In 1992, it was further expanded to cover hotels and catering, travel agencies and tour operators, and business services. The 1973 survey contained 36 questions but employers were also asked to provide information separately for manual and non-manual workers, thus doubling the requirement. This distinction was dropped in 1984 (even though it still formed part of the EC specification), along with a number of further questions, reducing the total to 20. In order to progressively minimise the burden on businesses, the survey has been conducted on the minimum sample size acceptable for statistical reliability of the results, and the level of detail requested has been reduced, either by dropping questions (strictly below the level of detail prescribed by the EC) or not asking certain questions of smaller firms. In the 1988 survey, one further question about rebates under the Redundancy Payments Act was dropped so that the survey contained 19 questions. In the 1992 survey, questions have been added on recruitment costs and company cars, and questions on payments for time not worked, which had been dropped, were reinstated. The distinction between manual and non-manual workers was altogether dropped from the EC regulation.

Documentation

Employment Department: Employment Gazette. The results of the 1988 Labour Cost Survey were published in the September 1990 issue, and those of the 1992 survey were published in the September 1994 issue. Both contain methodological information on the survey. Estimates of labour costs in non-survey years were published in the August 1991 issue of the Employment Gazette. A note giving details on the make-up of labour cost estimates and their basis is available from the Employment Department. Preliminary and final results of the EC Labour Cost Survey are also published in: EUROSTAT: Labour Costs 1988, initial results (Luxembourg, 1991). idem: Labour Costs 1988, volume 1: principal results, and volume 2: results by size classes and by regions (ibid., 1992).

Confidentiality / Reliability criteria

Data for individual firms are collected under the Statistics of Trade Act and remain confidential. Individual data cannot be disclosed to a third party.

Other information

Data supplied to the ILO for publication

Statistics of average hourly labour cost in manufacturing are published in Tables 22A and 22B of the Yearbook of Labour Statistics.