Economically Active Population Survey (Encuesta de Población Activa - EPA).
National Institute of Statistics (Instituto Nacional de Estadística - INE).
The survey is based entirely on the "national" as opposed to the "internal" concept, and therefore includes:
and, by agreement :
Excluded are :
The survey excludes persons living in collective households (hospitals, hotels, orphanages, barracks, convents, etc.) but includes families forming an independent group who live in those establishments, such as the families of the directors of the centres, or of custodians and porters. Furthermore, not all the population living in collective households is excluded: young men doing their compulsory military service, pupils in boarding schools, patients in sanatoria and detainees in prisons are included in the households from which they come, as their family give information on them. However, these categories are sometimes underestimated, either because their family ties are no longer very strong or because they are not declared to the interviewer.
The economically active population comprises persons aged 16 years or over forming:
- the armed forces, i.e., the army (career members) and the Civil Guard;
- the civilian economically active population, comprising the rest of the labour force.
Since the third quarter 1975 the survey has been continuous and the reference period of the final information obtained is the quarter. From 1964 until the first half-year of 1975 it was carried out half-yearly.
The week (Monday to Sunday) before the interview week.
The survey provides information on employment, unemployment, underemployment, hours of work, monthly household income, duration of employment and duration of unemployment, discouraged and occasional workers, industry, occupation, status in employment (occupational category) and level of education/qualifications.
Persons absent from their work because of illness or accident, labour-management dispute, suspension from work and payment of wages as a disciplinary measure, holidays, study leave, or maternity or any kind of leave, voluntary absence, temporary work disorganisation for reasons such as bad weather or mechanical breakdown, are considered as employees.
Apprentices who have been paid in cash or kind, and students who have worked full time for pay are considered as employees and classed as "at work" or "not at work" on the same basis as other persons in the service of an employer.
Persons suspended or separated from their employment as a result of an employment regulation measure are not considered as employees unless they expect to be able to resume employment with their enterprise.
Active members of the armed forces are classified as employees.
In accordance with this definition self-employed workers are employers, own-account workers, members of producers' co-operatives working in such co-operatives, and unpaid family workers."
Also considered as employed are :
Excluded from the employed and from the economically active population are:
Population counted separately: Young men doing their compulsory military service (or compulsory civilian service in lieu) are considered as a population counted separately; that is, they are not counted either as members of the economically active population or as inactive, whether or not they worked in the reference week.
"Specific steps to find work" include trying to set up on their own account and making arrangements for that purpose; registering at a private or official employment office; using the press or personal relations, or preparing or sitting for competitive examinations, etc.
Also considered as unemployed are :
Unemployed persons are divided into unemployed first-time jobseekers and unemployed persons with previous work experience.
Excluded from the unemployed, and considered as inactive but as 'potential members of the economically active population', are all persons without work and available for work, who are not looking for work for any of the following reasons :
Hours actually worked include the hours worked during normal periods of work, overtime, time spent at the place of work waiting or standing by, as well as short rest periods at the workplace including snack breaks of less than one hour. They exclude paid annual vacation, paid public holidays, paid sick leave or other paid leave. They also exclude time spent in travelling to and from the place of work.
Employed persons are classified as full-time or part-time workers, according to the duration of the working day. The basis for this classification is the interviewee's own statement, with the proviso that any person who usually works more than 35 hours a week cannot be regarded as working part time, and that persons who do not work as many as 30 hours a week cannot be considered as working full-time.
Employed persons, and persons who have lost their job less than three years before, are classified by industry, occupation and status in employment. The whole population aged 16 years and over is classified by level of education.
Employees are classified as: public sector employees (Public Administrations and Public Enterprises with State majority participation), and private sector employees. The first group includes both officials and contract employees, whether permanent or temporary.
Self-employed workers are divided into the following groups: employers, entrepreneurs without employees and self-employed workers, family workers (persons working without regulated payment in a business belonging to a member of their family with whom they live) and members of co-operatives (partners working in the co-operatives, but excluding persons working in workers' limited companies, who are regarded as employees).
This classification is convertible to the International Classification of Status in Employment (ICSE).
The areas frame is updated at each census; the lists frame is updated every 18 months, by rotation groups.
The first-stage units (census sections) are stratified in two ways: (i) by geographical stratification - sections are grouped into strata according to the province and type of municipality (by size of population) to which they belong; and (ii) by socio-economic substratification - the census sections in each geographical stratum are grouped into substrata reflecting the socio-economic category of the households in the section.
The survey uses 3,000 census sections, and each quarter investigates on average 20 dwelling units per section. This gives a total sample of 60,000 dwelling units (approximately 230,000 persons); primary units are selected with probability proportional to size (reckoned by the number of family dwellings in each), whereas the ultimate units are selected with equal probability by applying systematic sampling to the list frame.
The sample is uniformly distributed throughout the quarter, each of its sections being visited in one of the 12 weeks of the quarter. The whole sample is divided into three independent monthly subsamples, each of which representing the whole population.
The EPA is continuous, being interrupted only during August, when the interviewers are on holiday. For 1989 the data collection schedule is the following:
|Period||Dates of interview|
|1st quarter||2 January to 25 March|
|2nd quarter||27 March to 17 June|
|3rd quarter||19 June to 29 July
28 August to 7 October
|4th quarter||9 October to 30 December|
At first interview "not-at-home" and inaccessible households are revisited until information is obtained. Households refusing to reply are replaced by random selection of other households in the same section.
In case of total lack of response in interviews subsequent to the first interview, replacement consists in repeating the data obtained for the household in the immediately previous quarter.
The quality control programme consists in an evaluation survey in which the interview is repeated in a subsample of dwelling units from the original survey (the size of the subsample is approximately 4 per cent of the original sample). The data obtained at the first or "original" interview are compared with those obtained at the second or "repeat" interview. This procedure provides various indicators of quality.
The questionnaires filled in by the interviewers are exhaustively revised by the heads of groups, and partially (a sample) by the head of the survey in each province. The questionnaires are again exhaustively analysed by a specialised group in the Institute's national head office. Incomplete answers (i.e., when a few data on the person surveyed are lacking) result in automatic imputation.
When the data are computerised, errors in the persons' identification code and in age and composition of the household are corrected by referring back to the original questionnaires and/or the interviewer. The rest of the variables are automatically corrected by using the DIA (Detection and Automatic Imputation) computer kit devised by the INE for the treatment of surveys with qualitative variables.
First of all the results from each household are weighted by the inverse of the probability of selection, so obtaining unbiased estimators. Later on, ratio estimators are applied, taking as auxiliary variable the INE's quarterly population projections.
For information purposes, the following figures for sampling errors have been obtained, based on most significative values:
|Relative standard error (%)||0.50||0.58||0.97|
|Relative standard error (%)||1.20||1.40||1.90|
The sampling error of a given characteristic is calculated by the replicated semi-sample method. Each replication is a subsample formed by a number of sections equivalent to 50 per cent of the total sample. The EPA uses 40 replications formed by pairing the sections in each stratum and subjecting at random the first section in each pair to 20 replications and the second section to the remaining 20 replications. In this way each section appears in half the replications.
There are quarterly estimates of the population in collective households, made from the data of the last Population Census (a decennial statistical census) or Municipal Register of Inhabitants (a quinquennial statistical census). and by hypothesis on the evolution of the characteristics of this population group, based on the evolution observed in the EPA of these characteristics.
No adjustment is made for for seasonal variations.
The EPA started in 1964 as a half-yearly survey and became quarterly from the third quarter 1975 onwards. It has gone through various stages and updatings. The definitions, classifications and questionnaire were modified in the first half of 1972. Methodological modifications were made in the third quarter of 1976, and from then until the first quarter of 1987 there were not changes in the questionnaire or the methodology of the survey.
Until the first quarter 1980 data referred to the population aged 14 years and over, and from the second quarter 1980 to the population aged 16 years and over. To produce homogeneous series, the data of persons aged 14 and 15 in the period from the first quarter 1977 to the first quarter 1980 were therefore eliminated.
As from the second quarter 1986 the estimates given by the EPA were obtained by raising the sample results in accordance with the new demographic projections prepared by the INE from the 1981 Population Census, whereas in the period from the second quarter 1976 to the first quarter 1986 estimates were obtained by using the demographic projections based on the 1975 census.
In the second quarter 1987 a new methodology incorporating modifications in the definitions of the survey's main concepts was introduced together with a new questionnaire. The result has been a break in the historical series. However, a link has been established between the data obtained by the two methodologies.
Until that date family workers working as such in the reference week were not considered as employed unless they had worked more than one-third of the normal duration of work in the three months before the interview; if they had not, they were considered as inactive. Similarly family workers who had worked during the reference week in any other activity than that of family worker, and occasional workers, were considered as employed, unemployed or inactive, depending on the duration of the work they had done in the three months before the interview, whether they were available for work, and whether or not they looked for work.
Persons who as disciplinary measure had been suspended from their employment with loss of pay were previously considered as unemployed. From the second quarter 1987 they were classified as employed.
The distinction between employed persons in the strict sense of the term and marginally employed was eliminated; precarity of employment is analysed with reference to underemployment and the number of hours usually worked.
Regarding unemployed persons, the new EPA extended the reference period of job search from one to four weeks and fixed the period for verifying availability as the 15 days following the interview.
In the old methodology, unemployed persons who had worked before becoming unemployed were classified by the economic sector of their last employment, irrespective of the time they had been unemployed. Accordingly the classification "unclassifiable" included first-time jobseekers and those unemployed persons who had previously worked but had not declared the industry engaged in during their last employment, or had declared an industry that for various reasons could not be codified. In the new methodology the heading "unclassifiable" comprises not only the two groups just mentioned but unemployed persons who have previously worked and have been unemployed for three years or more.
A subclassification of inactive persons by causes of inactivity has been introduced, in addition to the previously existing subclassification by reasons for inactivity (retirement, studies, etc.). It consists in distinguishing between "potentially active" and "other inactive persons" and, in the first group, between "discouraged workers" and "remaining potential members of the economically active population".
The changes in the questionnaire have not been numerous and some of them have simplified it; thus the questions on activities in the last three months have been abolished; they are now unnecessary because all concepts now refer to the week previous to the interview week.
The EPA's results are published in the following volumes:
Instituto Nacional de Estadística : "Encuesta de Población Activa, Principales Resultados (Economically Active Population Survey, Main Results) (quarterly) (Madrid).
idem: "Encuesta de Población Activa, Resultados Detallados" (Economically Active Population Survey, Detailed Results) (quarterly) (ibid.).
The quarterly results are published approximately six weeks after the reference quarter of the survey.
idem: "Encuesta de Población Activa, Tablas anuales" (annual tables ) (ibid.); issued three months after the end of the year in question.
idem: "Encuesta de Población Activa, Series Revisadas (Revised series), 1976-1988" (ibid., January 1989).
idem: "Encuesta de Población Activa, Informe Técnico: Diseño de la Encuesta y Evaluación de la Calidad de los Datos (Economically Active Population Survey, Technical Report: Survey design and data quality evaluation) (ibid., 1987).
Another publication is being printed, which will present the results of the questionnaire's retrospective questions asked only in the second quarter of each year, and flow statistics.
The EPA results are also published in the INE's "Boletín mensual de estadística" and "Anuario estadístico".
The Institute's data bank also contains series of results of the Encuesta de Población Activa.