Annual labour force surveys. Ralf Hussmanns Head, Methodology and Analysis Unit Bureau of Statistics International Labour Office. International recommendations on periodicity of labour force statistics (1).
Annual labour force surveys Ralf Hussmanns Head, Methodology and Analysis Unit Bureau of Statistics International Labour Office
International recommendations on periodicity of labour force statistics (1) • “Current statistics of the economically active population, employment, where relevant unemployment, and where possible visible underemployment, should be compiled at least once a year.” • ILO Recommendation No. 170 (Labour Statistics Recommendation), 1985, Paragraph 1. (1)
International recommendations on periodicity of labour force statistics (2) • “The current statistics programme should encompass statistics of the currently active population and its components in such a way that trends and seasonal variations can be adequately monitored. As a minimum programme, countries should collect and compile statistics on the currently active population twice a year … ” • 13th ICLS (1982), Paragraph 2 (a)
International recommendation on periodicity of statistics on the informal sector • “The data collection programme should provide both for (a) the current monitoring, if possible once a year, of the evolution of employment in the informal sector and (b) the in-depth examination, if possible every five years, of informal sector units with respect to their numbers and characteristics ...” • 15th ICLS (1993), Paragraph 21 (1)
Annual labour force surveys • Periodic data collection (point in time estimates): • once a year • two, four or twelve times a year • Continuous data collection (annual, quarterly or monthly averages): • every week
Continuous data collection (1) • Seasonal and other variations over time are captured and period effects eliminated through division of sample in monthly, fortnightly or weekly sub-samples and continuous data collection during the year (examples: Mauritius, South Africa). • Estimates reflect the average situation during a month, quarter or year, i.e. for many purposes (including national accounts) they are more useful than point-in-time estimates.
Continuous data collection (2) • Flexibility in periodicity of data dissemination (example Colombia: dissemination of monthly, bi-monthly, quarterly, bi-annual and annual averages), • but inverse relationship between (i) level of geographic and other detail of estimates and (ii) frequency of data dissemination. • It becomes unnecessary to use concepts based on long reference periods (e.g. usual activity, annual income), which are prone to recall errors.
Continuous data collection (3) • As data entry and processing can be carried out on a continuous basis, the time lag between data collection and dissemination can be much reduced (South Africa: one month) and the users’ demand for more timely statistics be satisfied. • Data quality is improved because field work is carried out by small teams of permanent interviewers and supervisors (reduced cost & improved quality/intensity of training, recruitment and supervision of field staff facilitated, including re-interview programme). • Additional topics can be included in the survey as modules attached to it from time to time.
Enhanced flexibility to meet demands for additional data • Additional topics included in the survey should be somehow related to the core topics of the survey • to avoid that the survey will become an omnibus multi-purpose survey. • Not all additional topics require to be investigated in using the whole sample. • To reduce response burden and survey costs, sub-samples can be used to investigate various additional topics.
Sub-samples • Defined by rotation groups • or determined by serial number of the interview • or representing all households interviewed during a specific period of time (month, quarter or year)
Determining factors for type of sub-sample • Urgency of user needs for the information, time available for data collection. • Nature of the topic, especially its being subject or not to seasonal or other variations over the year. • Usefulness of inclusion of topic in repeated interviews of the same households. • Precision requirements for estimates. • Response burden of households. • Etc.