Switzerland (1)

Title of the survey

Statistique de l'Emploi (STATEM) (Employment Statistics) (until 1992).

Organization responsible

Office fédéral de la statistique (OFS), Section des Entreprises et de l'Emploi.

Periodicity of the survey

Quarterly.

Objectives of the survey

To obtain information on employment trends by branch of economic activity and by canton. The results of the survey are used for calculating various indicators (such as averages per capita, productivity, aggregates, etc.). They are used mainly by the public authorities responsible for economic and regional policies.

Main labour topics covered by the survey

Employment.

Reference period

Final pay day of the quarter.

Coverage of the survey

Geographical

Whole country.

Industrial

All branches of economic activity, excluding agriculture, fishing and international organizations.

Establishments

All types and sizes of establishments in the private and public sectors.

Persons

All employed persons, excluding home workers.

Occupations

Data are not collected by occupation.

Concepts and definitions

Employment

The survey covers all employed persons and, of these, the number of production workers All employed persons comprise all persons employed for at least six hours a week in the establishment surveyed. These include proprietors, administrators, directors, company managers, salaried employees, wage earners, apprentices, external agents and volunteers, as well as auxiliary workers, temporary workers (even when remunerated by a third party) and family workers. Also included are persons absent on the reference date because of military service, sickness, injury, vacation, leave or any other reason. Employed persons working outside the establishment (for example, fitters, representatives, craftsmen in construction and civil engineering) are taken into account at the head office of the establishment. Production workers are workers assigned predominantly to production in the strict sense, i.e. in industrial, craft and manual operations (skilled workers, supervisors, shift, workshop or site foremen, checkers, craftsmen, fitters, storekeepers, forwarding agents, packers, caretakers, drivers, kitchen, service and cleaners, etc.). The number of office workers is obtained by subtracting the number of production workers from the total number of employed persons. In each category (total employment and production workers), data are collected separately by sex, and a distinction is made between the following groups:

persons employed on a full-time basis - for at least 90 per cent of the hours normally worked in the establishment,
persons employed on a part-time; these are subdivided into two categories:
- persons employed for 50 to 89 per cent of normal working hours,
- persons employed for less than 50 per cent of those hours.

Short-time work is not considered as part-time work. Apart from the number of persons employed, data are also collected on the shortage or surplus of skilled, semi-skilled and unskilled labour and on the level of and prospects for employment in each establishment covered.

Earnings

Not relevant.

Wage/salary rates

Not relevant.

Hours of work

Not relevant.

Classifications

Industrial

The data are classified in accordance with the Nomenclature générale des activités économiques of the 1985 Federal census of enterprises. There is some correlation between this classification and the International Standard Industrial Classification of All Economic Activities (ISIC), Rev. 2, 1968. There are also correspondence tables between this classification and the classification of economic activities of the European Communities (NACE), Rev.1, which is in turn based on ISIC, Rev. 3, 1990.

Occupational

Not relevant.

Others

The data on employment are classified:

by region (canton);
by category of worker (production worker or office worker);
by sex;
according to type of employment (full-time and two categories of part-time employment).

The data on labour shortage or surplus are classified according to skill level (skilled, semi-skilled or unskilled).

Sample size and design

Statistical unit

The sample unit is either the establishment (local unit) or the enterprise (legal unit).

Survey universe / sample frame

The sample frame comprises the register of enterprises and establishments of the OFS. This register is compiled on the basis of the Federal census of enterprises which is generally conducted every ten years. The last census of enterprises was taken in September 1991 and carried out on the basis of a simplified questionnaire. The 1985 census was more comprehensive and allows more detailed analysis. The register is constantly updated on the basis of existing information: registrations and modifications in the Feuille officielle du Commerce, information received during surveys, telephone lists, etc.

Sample design

The survey is conducted on a non-exhaustive basis using judgement sampling, taking into account the size of the establishments and the industry to which they belong. For this, the survey depends largely on data supplied by economic associations. The sample covers about 47,000 establishments (i.e. 11 per cent of all establishments listed in the register) and 1.9 million persons employed (i.e. about 61 per cent). Representativeness varies according to branch of economic activity. The sample is renewed only occasionally.

Field work

Data collection

The data are collected by means of questionnaires mailed to the selected establishments and enterprises, two weeks before the reference date. Respondents (establishments and enterprises) are asked to sent the questionnaires back up to three weeks after that date. Reminders are sent a few days later to those who have not replied. Certain data are supplied in machine-readable form.

Survey questionnaire

The questionnaire consists of one page, with accompanying explanations, in which data are inserted on the number of persons employed and other information according to the categories mentioned above.

Substitution of sampling units

In the case of unexplained non-response on the part of an establishment or enterprise, the figures from the preceding quarter are used up to three times. In the case of identifiable change (activity terminated, change of address) the data are adapted accordingly.

Data processing and editing

The questionnaires are transmitted for processing to the Centre de calcul électronique de l'administration fédérale. Data entry consists of optical character reading of the identification characteristics of the sampling units and manual keying in of the data concerning persons employed. All data are subjected to plausibility and other basic tests, by comparison with the data of the preceding periods and between total employment and production workers. Consistency ranges are established and, in the case of seasonal activities, plausibility checks are carried out. All data classified as doubtful or wrong are checked and corrected if possible.

Types of estimates

employment:
- employment index,
- variations in relation to the preceding quarter and year,
- percentage share of categories of workers,
labour shortage or surplus: in percentage of establishments and jobs,
employment level and outlook: in percentage of establishments and jobs.

Construction of indices

The employment index is a chain index obtained by applying the rate of change in employment for a quarter to the index of the previous quarter. Only reporting establishments whose data are comparable from one quarter to the next are used. An index is thus calculated for each basic stratum of the sample (type, group or economic division of the Nomenclature des activités économiques). The aggregated indices are subsequently calculated as weighted averages of the basic indices. Since the third quarter of 1985, the quarterly employment index has used data from the 1985 census of enterprises as weights, with the third quarter of 1985 as 100. The employment index is also used to determine the total number of employed persons in absolute figures. Evaluative indices relating to employment level and outlook are obtained by weighting the shares of the different responses in terms of percentages of employment. The data on all establishments that have replied to these questions are used. The share of good responses is weighted by the coefficient 1.5, that of satisfactory or uncertain responses by 1, and that of bad responses by 0.5. The index can thus take values ranging from 50 to 150. The higher the value, the better the situation and outlook are judged.

Weighting of sample results

Since the second quarter of 1987, and for all indices after the fourth quarter of 1985 which have been recalculated, the weights for economic activities in Switzerland, in the cantons and the major towns have been based on the employment figures (number of persons employed at least 50 per cent in establishments offering a working week of 28 hours or more) contained in the 1985 census on enterprises. As regards the evaluative indices on employment levels and outlook, the lowest levels (groups and classes of economic activity) are calculated directly on the basis of the data reported by establishments, whereas the higher aggregates are calculated on the basis of the figures for the lower levels and their weights based on the 1985 census of enterprises. Data on labour shortages or surplus are not weighted and concern only the establishments participating in the survey.

Adjustments

Non-response

None.

Other bias

Not relevant.

Use of benchmark data

Not relevant.

Seasonal variations

Not relevant.

Indicators of reliability of the estimates

Coverage of the sampling frame

The register of enterprises and establishments is constantly updated, so as to reflect almost all establishments and enterprises. However, units where the employees do not work a total of at least 20 hours a week are not considered as establishments and are therefore covered neither by the register nor the employment statistics.

Sampling error / sampling variance

Not relevant.

Non-response rate

The non-response rate is about 10 per cent, representing 4,000 to 5,000 establishments.

Non-sampling errors

A certain lack of precision in definitions (production and office workers, skilled, semi-skilled, etc.) may give rise to errors in the estimation of employment levels.

Conformity with other sources

Not relevant.

Available series

employment index for the whole of Switzerland by category of worker (production, office), sex, economic activity, and by canton and major town,
variations in employment levels in relation to the preceding quarter and year, for Switzerland, by economic activity, category of worker and by canton and major town,
labour shortage and surplus, by skill level, economic activity and by canton and major town,
employment levels in percentage of establishments and workers, by economic activity,
employment prospects as a percentage of establishments and workers, by economic activity,
evaluative indices by economic activity and by canton and major town,
percentage distribution of production workers and office workers, by economic activity, canton and major town,
percentage distribution of full-time and part-time workers, by economic activity and sex, and by canton and major town.

History of the survey

The employment survey dates back to 1921. At that time, it was conducted by the Office fédéral de l'industrie, des arts et metiers et du travail (OFIAMT). Since 1925, regular quarterly surveys have been carried out on the employment situation in industry and construction. These statistics were revised for the first time in 1995, and have undergone other modifications since 1969:

1969: inclusion of the service sector, on the basis of the results of the 1966 census of enterprises;
1971-1981: computerization of data analysis and handling, inclusion of part-time employment, removal (in 1977) of questions relating to hours of work and wage rates (replaced by statistics on the earnings of workers who have suffered accidents) and adjustment of the classification and weighting of the series to the results of the 1975 census of enterprises;
1984: regionalization of results (by canton and major town),
1987, second quarter: the OFS became responsible for employment statistics,
1987: adoption of a new classification of economic activities and of weights based on the results of the 1985 census of enterprises.

A complete revision of employment statistics took place in 1992, which allowed, in particular:

improved updating of the survey universe, the enterprise census, as a result of more frequent complete surveys (every three years) and of an initial inquiry on new entries in the register;
recourse to a rotating panel resulting from random sampling with stratification by establishment size and economic activity;
adoption of a method of estimation based on absolute figures and using the results of both the most recent three-yearly census and the initial inquiries;
removal of the distinction between production workers and office workers;
addition of a question on job vacancies.

The first survey conducted in accordance with these concepts took place in the third quarter of 1992, based on the results of the September 1991 census of enterprises. In this connection, see the following description Statistique de l'emploi, révisée.

Documentation

Office fédéral de la statistique: Statistiques de l'emploi et de la population active occupée (quarterly; Berne); published about two months after the end of the quarter. This publication also contains methodological notes. idem: Annuaire statistique de la Suisse (annual; ibid.). Département fédéral de l'économie publique: La Vie économique (monthly; ibid.).

The employment indices are stored in the OFS database STATINF, accessible on line, and certain more detailed data are available on diskette. For more comprehensive methodological information , see: Office fédéral de la statistique: Off-print of La Vie économique, No. 8/87: Les nouvelles bases de la statistique de l'emploi. OFIAMT: Off-print of La Vie économique, No. 2/82: Les bases de la statistique révisée de l'emploi.

Confidentiality / Reliability criteria

No personal data from this survey are communicated to third parties.

Other information

Data supplied to the ILO for publication

The official employment estimates published in Tables 3A, 3B and 3C of the Yearbook of Labour Statistics are derived from several sources, including the employment statistics (the base is the five-yearly Federal population census and the enterprise census). The employment indices for all persons employed, by sex, in non-agricultural activities, manufacturing, construction, and transport, storage and communication are published in Tables 4, 5A, 7 and 8 of the Yearbook of Labour Statistics.