Labour Force SurveyMetadataPeriod: AnnualYear: 2020
|02. Metadata update|
|03. Statistical presentation|
|04. Unit of measure|
|05. Reference Period|
|06. Institutional Mandate|
|08. Release policy|
|09. Frequency of dissemination|
|10. Accessibility and clarity|
|11. Quality management|
|13. Accuracy and reliability|
|14. Timeliness and punctuality|
|15. Coherence and comparability|
|16. Cost and Burden|
|17. Data revision|
|18. Statistical processing|
|01.1. Contact organisation|
Statistical Office of the Republic of Serbia (SORS)
|01.2. Contact organisation unit|
Labour Market Statistics Department, Labour force survey group
|01.5. Contact mail address|
Belgrade, 5 Milana Rakića St.
|01.6. Contact email address|
|02. Metadata update||Top|
|02.1. Metadata last certified|
|02.2. Metadata last posted|
|02.3. Metadata last update|
|03. Statistical presentation||Top|
|03.1. Data description|
Labour Force Survey (LFS) is a research focused on the household and is designed to obtain information about the labour market and labour force characteristics.
Regarding methodology, organization, sampling procedure, questionnaire design and definitions of employment and unemployment, LFS is carried out in line with the standards and recommendations of the International Labour Organisation – ILO as well as with Eurostat regulations, which allows comparing a large number of LFS data with equivalent data from other countries, particularly European.
The main objective of this Survey is to obtain data on three main mutually exclusive population contingents – employed, unemployed and inactive persons.
As far as employed are concerned, LFS registers their professional status in three ways:
In addition, LFS registers also the type of ownership they work in, rights at work, occupation and activity, number of working hours, additional job (if any), etc.
As for the unemployed, LFS scrutinizes the:
As for inactive persons the data are collected on:
Apart from the previously mentioned categories, demographic, socio-economic, educational and other characteristics (age, sex, marital status, education, citizenship, national affiliation, etc.) are also studied.
The results of the Survey are published as aggregates, thus securing full confidentiality of information about households and individuals, according to the provisions of the Law on Official Statistics.
Survey data are published on quarterly and annual basis.
|03.2. Classification system|
The following international classifications are used in Labour Force Survey:
– International Standard Classification of Occupations (ISCO – 08);
– Statistical classification of economic activities (NACE, rev.2), which is comparable to UN classification of activities ISIC – rev.4;
– International Standard Classification of Education, ISCED–2011;
– Classification of fields of education( ISCED-F – 2013), which complements the International Standard Classification of education ISCED-2011)
– International Classification of Status in Employment (ICSE – 93).
|03.3. Coverage - sector|
Labour Force Survey covers all sections defined by NACE rev. 2.
|03.4. Statistical concepts and definitions|
The population is classified as employed, unemployed and inactive based on the information from the questionnaire concerning the activity of the respondents in the reference week. Data obtained in this way differ from the data on employed and unemployed persons obtained through other statistical researches.
The Survey does not take into account the formal status of the respondent, but her/his employment status is determined based on the activity that is carried out in the reference week. Thus, in addition to the persons who have a formal job, employed persons in this Survey can also be individuals who have no formal employment, but they have performed some work (at least one hour) during the reference week in order to earn their living. Similarly, inactive persons are formally unemployed persons (as they are noted in the National Employment Service) if during the four weeks preceding the survey have not undertaken active steps to look for work and was not able to start working within two weeks of the survey. In the second case, formally unemployed or inactive persons (persons without a formal job, retired, student, housewife, etc.) that have performed in the reference week any paid work shall be considered employed, regardless of their formal status.
Employed persons are persons who performed a paid job (for cash or in kind) for at least one hour as well as persons who had an employment, but were absent from it during the reference week (with a guarantee of returning to it).
Employed persons are persons who have a formal employment in an enterprise, institution or in another type of organisation, persons who work by their own (unincorporated private enterprises, individual farmers, persons who performed any job they had found and made agreement by themselves – orally or in written). Apart from the mentioned, employed persons are also contributing family workers in a family business owned by a household member and defined as unpaid work that directly contribute to the functioning of the observed family business.
Unemployed persons are persons who have not performed a paid work in the reference week, have sought actively work during four weeks ending with the reference week and were available to start working within two weeks following the reference week.
This category includes also persons who did not seek work during the four weeks ending with the reference week because they had found a job that should start after the end of the reference week, or at the latest within three months and are able to start working within two weeks following the reference week.
Active population – labour force includes all employed and unemployed persons.
Inactive population includes all persons age 15 and over who are not classified into employed and unemployed population. Inactive population covers:
other persons who did not perform any paid work in the reference week, have not actively sought work, and were not available to start working within two weeks after the end of the reference week.
Potential active population includes the persons who take active steps to find a job but cannot start working immediately, i.e. within two weeks, as well as persons who can immediately start working but did not take active steps to find a job.
This contingent includes so-called discouraged persons, i.e. persons who do not undertake steps to find a job because they are convinced they cannot find one although they could start working immediately in the case a job was offered.
Formal employment includes:
Informal employment involves:
Activity rate is the percentage share of active population in the total population aged 15 and over.
Employment rate is the percentage share of employed population in the total population aged 15 and over.
Unemployment rate is the percentage share of unemployed population in the total active population aged 15 and over.
Inactivity rate is the percentage share of inactive population in the total population aged 15 and over.
|03.5. Statistical unit|
The survey unit is every member of a randomly selected household, and the unit of choice is the household selected in the sample.
Basic socio-demographic data are collected for all persons in selected households, while only persons aged 15 and over are interviewed about work activity.
|03.6. Statistical population|
The target population is all persons living in private (non-institutional) households on the territory of the Republic of Serbia, which represents the usual population, excluding the Region Kosovo and Metohija.
Usual population, by the UNECE and Eurostat recommendations for the Census 2011 is made of:
The target group is not composed of persons who reside or intend to reside less than a year in the Republic of Serbia, neither of persons who are absent or intend being continuously absent from the interviewed household one year or longer (live and/or work abroad, permanently moved out from the household, etc.).
|03.7. Reference area|
Reference territory is the territory of the Republic of Serbia without the region of Kosovo and Metohija. Since 1999, SORS has no available data for AP Kosovo and Metohija, therefore they are not included in the data for the Republic of Serbia (total).
|03.8. Coverage - Time|
LFS was first carried out in 1994 in our country as a pilot survey. From 1995 till 2008 it was conducted regularly once a year, then over the period 2008 – 2013 the survey was carried out twice a year (in April and October). In 2014 it was carried out four times (quarterly – in February, May, August and November) and in 2015, it switched to a continuous periodicity which is still applied.
|03.9. Base period|
|04. Unit of measure||Top|
|04. Unit of measure|
Most results are expressed as a number of persons (thousands) or as rates (employment rate, unemployment rate or growth rate). There are also variables that are reported in other units of measurements (age in number of years, working hours in hours, etc.).
|05. Reference Period||Top|
|05. Reference Period|
LFS is carried out as a continuous survey where interviews take place every week during the year.
Reference week is the week preceding the survey and covers the whole calendar week (from Monday to Sunday).
Labour force status is defined in relation to the reference week, which is also called the observed week. The first reference week in a year must cover the first Thursday in a year, and the first quarter encompasses 13 references weeks.
|06. Institutional Mandate||Top|
|06.1. Institutional Mandate - legal acts and other agreements|
The power and responsibilities for collecting, processing and publishing statistical data to the Statistical Office of the Republic of Serbia (SORS) are given and regulated by the Law on Official Statistics (“Official Journal of the RS”, number 104/09).
|06.2. Institutional Mandate - data sharing|
|07.1. Confidentiality - policy|
Protection of individual data is compulsory by Article 3 of the Provision on the Protection of Respondents and Articles, 44, 45, 46, 47, 48 and 49 of the Provision on Confidentiality of the Law on Official Statistics (“Official Journal of RS “, number 104/09).
|07.2. Confidentiality - data treatment|
Protection of individual data is an extremely important issue. The results of the Survey are published as aggregates, thus securing full confidentiality of personal data on households and individuals, according to the provisions of the Law on Official Statistics.
Statistical micro-databases that are released for use for scientific research purposes contain reduced information about the respondents in order to prevent identification of the surveyed individuals or households. In addition to the fact that publicly available databases do not contain information on the name, surname and address of individuals, additional personal data protection measures apply:
- the age of the respondents is given in groups of five years;
- employees’ earnings are given at intervals of 10 groups;
- the country code (birth, citizenship, work) is given separately for Serbia and the countries of the region (Bosnia and Herzegovina, Croatia, Montenegro, North Macedonia), while the other countries are grouped into three sections: EU countries except Croatia, other European countries, rest of the world;
- occupation is given at the 3-digit level
- activity is given at the sector level (21 groups in total);
- the household identifier changes at each subsequent annual basis to prevent monitoring of household members over time.
The above data groupings are made in accordance with the rules applied by the European Statistical Office - Eurostat aiming to personal data protection. More information can be found at link in the annex.
Annexes : Protection of personal data, Eurostat
|08. Release policy||Top|
|08.1. Release calendar|
LFS data are published quarterly and annually on dates specified by the release calendar.
|08.2. Release calendar access|
LFS data are published in accordance with the release calendar of the Statistical Office of the Republic of Serbia, which can be accessed through the link given in annex:
Annexes : Release calendar
|08.3. Release policy - user access|
LFS publications and data are available on the SORS website on the dates specified in the release calendar. The Dissemination Group also sends SMS notification to specific users on the day of the publishing, accompanied with the information on main indicators. The Group also sends a printed version of the published publication to different government officials, ministries, National Library, etc.Annexes : SORS website
|09. Frequency of dissemination||Top|
|09. Frequency of dissemination|
Quarterly and annually.
|10. Accessibility and clarity||Top|
|10.1. Dissemination format - News release|
LFS data are published quarterly in press releases.
|10.2. Dissemination format - Publications|
The Statistical Office of the Republic of Serbia publishes the following publications:
The publications can be accessed through the link in the anneks.Annexes : Publications
|10.3. Dissemination format - online database|
LFS data are available in an online dissemination database of the Statistical Office of the Republic of Serbia and are located in section Labour market / Labour Force Survey. Link is given in annex.Annexes : Dissemination database
|10.4. Dissemination format - microdata access|
Microdata are available, upon request, to international organization, research institutes, etc.
|10.5. Dissemination format - other|
Data that require additional processing are to be delivered to Dissemination Unit, after which they are to be forwarded to the LFS group that begins the process of production/processing of requested data. Afterwards, the requested data are delivered to users.
|10.6. Documentation on methodology|
Detailed methodology is available on links given in annex.Annexes : Abridbed Labor Force Survey methodology Labor Force Survey methodology for users
|10.7. Quality management - documentation|
|11. Quality management||Top|
|11.1. Quality assurance|
SORS is working on improving the quality of the Labour Force Survey by adopting the following EU regulations:
Since 2010, SORS transmit annual and quarterly LFS quality reports to Eurostat.
Special attention is paid to harmonizing the main quality indicators and main LFS concepts, such as:
In addition to improving the methodology, constant control of the obtained results and the interviewers is carried out, as well as the improvement of the instruments for conducting the survey.
|11.2. Quality management - assessment|
The Labour Force Survey can be considered as a reliable source of data, knowing that Eurostat and ILO standards and definitions have been applied in questionnaire design and methodology development.
The Survey is sample-based and is therefore subject to sampling errors. Depending on sample size, figures with a few observations are marked in LFS publications and warning about unreliability of the figures is given for them.
|12.1. User needs|
The Labour Force Survey is carried out in every European country as well as in many other countries around the world following common concepts and guidelines. This makes the Labour Force Survey the most adequate Serbian survey for international comparisons on labour market statistics.
The Labour Force Survey is used for monitoring the labour market.
The data that are most frequently asked for but are not available from other sources are those on:
LFS data are used by national and international users.
National users are the Government organisations of the Republic of Serbia such as: the Ministry of Labour and Social Affairs, Ministry of Science and Technology, Ministry of Youth and Sports, National Bank, National Employment Service, local authorities, etc. Other national users are various academic and research institutions (universities and institutes), enterprises, journalists and individual users such as students, scientists, researchers and others.
The National Accounts use LFS data to produce the adequate estimations of the labour input. For this purpose, the variables used are: actual hours of work, section of activity, professional status, type of labour contract etc.
The international users are: Eurostat (for main LFS indicators), UNESCO and UNICEF (for educational data), ILO (for employment and unemployment), World Bank, International Monetary Fund (IMF) and others. Some countries from the region need LFS data for various regional projects.
|12.2. User satisfaction|
SORS does not carry out a satisfaction survey targeted at users of labour markets statistics.
The Labour Force Survey is harmonized with the Commission Regulation 430/2005, which makes the data complete.
|13. Accuracy and reliability||Top|
|13.1. Overall accuracy|
The LFS is the most reliable source of data when comparing labour market internationally. Constant harmonization with international recommendations and standards make the Labour Force Survey a reliable research. As with any sample-based survey, this survey has sampling errors which are associated with sample selection and non-response patterns.
|13.2. Sampling error|
Sampling errors are a measure of the variability between the estimates from all possible samples. The extent of variability is not known exactly, but can be estimated statistically from the survey data.
In the dissemination phase, it is not possible to publish each estimate with its sampling error. In order to allow users to use sampling errors (for selected data) when doing analyses, approximate values of sampling errors have been calculated based on regression model.
Tables with approximate values of sampling errors for totals and means, method of using said tables and calculation of the CVs for rates and changes are available in the available in the Methodology for Users – “Labour Force Survey, 2017 (pages 25-26). Link of the publication is given in the annex.Annexes : Labor Force Survey methodology for users
|13.3. Non-sampling error|
а) Coverage errors
Demolished dwellings, evicted households and changed purpose of the apartment are the main causes of over-coverage. Under-coverage problems result from non-updated enumeration areas.
Errors related to coverage (over-coverage, under-coverage, as well as wrong classification) are documented in the Quality Report of the Labour Force Survey.
b) Measurement errors
No measurement error estimates are available.
c) Processing errors
Data can be modified and edited in the course of the survey/interview. If an interviewer wants to modify the initial respondent’s answer when she/he sees there has been an error, she/he can do it on site. Upon the arrival of data in the head office, additional control is to be done, as well as the coding of occupations, economic activities and fields of education. Editing and processing phases refer to detecting logically incompatible information, as well as to their modification. Logical control in data entry is incorporated in the application, thus disabling the interviewer to enter illogical data and as such limiting such cases to the minimum.
There are no available estimates on the number of processing errors in the LFS.
d) Non-response errors
Evaluation of non-response errors is not available.
|14. Timeliness and punctuality||Top|
Annual data are published once a year, more specifically at the last working day of March. The quarterly data are published four times a year (I quarter of the year – May of the same year, II quarter of the year – August of the same year, III quarter of the year – November of the same year and IV quarter of the year – February of the following year). The data are published approximately 6 to 7 weeks after the end of the fieldwork or more specifically on the last working day of the following month.
The data are generally published on time and in accordance with Statistical Calendar.
|15. Coherence and comparability||Top|
|15.1. Comparability - geographical|
The data are comparable on the level of regions and districts with exception of the region Kosovo and Metohija for which data are not collected. The data are also comparable with other countries of the Balkans and EU.
All definitions and procedures are laid down by the European Union, which makes the Survey suitable for both overall and very specific international comparisons of labour market issues.
|15.2. Comparability - over time|
The LFS was first introduced and carried out in 1994 as a pilot survey. From 1995 till 2007 it was conducted regularly once a year, in October. In 2004, the questionnaire was harmonised with the EU Regulation No. 577/98 when data series were interrupted for the first time. Over 2008 – 2013, the survey was carried out twice a year (in April and October). In 2008, it was additionally aligned with the EU Regulation No. 377/2008. Supplementary questions covering the contingent of employed persons were introduced as well as variables concerning education; this harmonisation led to a second break in data series. In 2013, the sampling frame was changed (2011 Census), causing the next break in data series. In 2014, the survey was carried out four times a year (once in every quarter – in February, May, August and November). Then in 2015, it switched to a continuous periodicity. During this period modifications were made not only to methodology and periodicity but also to sample size.
Due to the change in the rating system in 2015 (post stratification according to NUTS-3, gender and five-year age categories), the first revision for data from 2014 was made. Due to the introduction of uniform weights at the level of individuals and households in 2016, a revision of the data for 2014 was redone (rev.2), as well as a revision of the data for 2015.
Although the LFS has been continuously carried out since 1994, all the aforesaid changes led to lower data comparability between certain periods. Therefore, data collected over 2004 – 2007 are comparable between each other, and so are those gathered from 2008 to 2013. Owing to the revisions data from 2014 onwards are also comparable between each other.
Although improvements in time have brought some time series break the comparability of the main indicators for the period starting from 2008 is still high.
|15.3. Coherence - cross domain|
Population statistics are not fully comparable with socio-demographic data collected by the Labour Force Survey. There are many conceptual differences between them, for example:
The LFS includes all employed persons being residents (permanent citizens) of the Republic of Serbia (national concept), particularly the persons working on the territory of the country. National accounts applies the so-called “domestic concept” that includes all employed persons, whether residents or not, who belong to the production activity performed in the country. This implies that National Accounts cover overall employment in agriculture and overall employment not being monitored by the official economy.
National accounts estimate the number of employed persons based on different sources, such as:
In the Republic of Serbia, National Accounts currently estimate only experimentally the number of employed, so the procedure of estimation of employed has not been published yet.
Data on registered employment differ from data from the Labour Force Survey for reasons cited below:
There is a difference between the definition of unemployed persons in the LFS and registered unemployment:
|15.4. Coherence - internal|
LFS estimates for a given reference period have full internal coherence, as they are all based on the same corpus of microdata and they are calculated using the same estimation methods.
Annual data are calculated as the average of quarterly data, which ensures coherence.
|16. Cost and Burden||Top|
|16. Cost and Burden|
|17. Data revision||Top|
|17.1. Data revision - policy|
The main principles and standards of the revisions of released statistical data are defined by the General Revision Policy of the Statistical Office of the Republic of Serbia.Annexes : General Revision policy
|17.2. Data revision - practice|
Since 2016 the estimation system has been changed. The Statistical Office of the Republic of Serbia has undertaken revision in order to overcome the effect of those changes and secure data comparability for the following period with data for 2014 and 2015.
Data for 2014 are available both by old and new estimation technique, which allows measuring their impact on different estimation systems and applying the calculated index to previous years.
|18. Statistical processing||Top|
|18.1. Source data|
The Labour Force Survey is the most comprehensive continuous research in Serbia. Since 2015, when there was the change in periodicity, interviews have been conducted every week (from Monday to Sunday), where in one week only one subsample is interviewed, and in the following one another subsample, and so on. The respondents provide answers about their employment activity in the week which is the reference week for them.
Type of sample
The LFS uses a two-stage stratified cluster sampling approach. The sample of enumeration areas (cluster of households) was selected at the first stage. A sample of households was selected in each enumeration area at the second stage. Enumeration areas as primary sampling units are stratified according to the type of settlements (town and other) and territory at NSTJ 3 level.
Rotation scheme and sample allocation
The LFS is based on the rotation panel design, by which each household participate in the sample for four times. The rotating scheme 2-2-2 was applied.
For each quarter four subsamples (rotating groups) are allocated. Each quarter (Q) in the sample is included: one new rotating group, two rotating groups from the previous quarter (Q-1) and one rotating group from the same quarter of previous year (Q-3). Based on the applied rotating scheme, each household and person selected in the sample will be interviewed four times within 18 months, i.e. each household and person is in the sample for two consecutive quarters, then for two quarters out of the sample and again for two consecutive quarters in the sample.
The applied rotation scheme based on the overlapping of the sample for two consecutive quarters, as well as for the same quarter in two consecutive years enables an efficient sampling plan for measuring the changes to the previous quarter and for measuring changes to the same quarter of the previous year.
Proportional sample allocation according to the number of persons aged 15 and over, within the type of settlement and territory, has been slightly corrected in order to obtain more precise estimates on the level of areas.
The transition to the continuous periodicity of the survey required the sample distribution over time. Every subsample allocated for a quarter is uniformly and randomly distributed into 13 weeks, and exceptionally, which happens once in four years when there are 53 observed weeks, the sample for one quarter (first or fourth) is distributed to 14 weeks. This was the case in the fourth quarter of 2015.
|18.2. Frequency of data collection|
Every respondent is interviewed about his/her activity during a specific reference week and the interviews are conducted daily all year round. Each household and person selected in the sample will be interviewed four times within 18 months, i.e. each household and person is in the sample for two consecutive quarters, then for two quarters out of the sample and again for two consecutive quarters in the sample.
|18.3. Data collection|
The tool for data collection in LFS is the electronic questionnaire, which is both used by interviewers doing the telephone interview (CATI) and those working on the field (CAPI).
Field and telephone interviewers are engaged in the collection of data. They are the main stakeholders in LFS conducting and the success of the survey depends on their work. The realization of LFS is dependent on how they make interviews, are acquainted with the essence of the questions in the questionnaire and on how they interpret them. Their diligence and efforts are also important.
The first interview round is made through personal contact, i.e. interviewer’s visit to the household. The interviewers have laptops with an installed electronic questionnaire. During the first visit, the interviewers record household’s contact telephone in order to carry out the next three interviews through the call-centre. The interviewer collects first socio-demographic information for each household member, and then information about employment status of members aged 15 and over. In the next round, the interviewer checks the socio-economic information collected in the first interview, and then she/he gathers information about the employment status for the current quarter.
Until the fourth quarter of 2014 there were two modes of data collection PAPI (personal visits with paper questionnaire) and CATI (computer-assisted telephone interviewing). In the fourth quarter of 2014, PAPI was replaced with CAPI mode of data collection (computer-assisted personal interview).
Participation in the Labour Force Survey is on voluntary basis.
|18.4. Data validation|
The tool for data collection in the LFS is the electronic questionnaire, which is both used by interviewers doing the telephone interview (CATI) and those working on the field (CAPI). Utilising the electronic questionnaire saves time and money needed for data entry and processing. All the interviewers use the same questionnaire, but logical control (implemented to reduce as much as possible the number of errors in data entry) is not the same for telephone interview and fieldwork. Logical control for CAPI is lighter because it is reckoned that an interviewer should stay on in the household as little time as possible, thus they do not enter the codes of occupations, economic activities and field of education, but only describe the requested information. Coding and complete logical control of fieldwork materials are performed after data entry in the central database located on the server of the Statistical Office of the Republic of Serbia in the Computer Centre by statisticians and engaged supervisors.
The data collected by telephone interview are directly registered in the central database, and then coded. Full logical control is applied for them during data entry.
Logical control recognizes obvious incompatibilities (e.g. a respondent aged 15 who worked for the last time in 1940). If the value is suspect but reasonable, the programme will alert the interviewer to check the entered data and confirm them if they are correct. This way the probability of making this type of errors in data entry is prevented.
In the Labour Force Survey 'depending interviewing' is used, which means that previous answers are used to reduce the response burden and to limit the amount of time used to manually seek out errors. The interviewed persons validate previous answers.
|18.5. Data compilation|
Annual results of quarterly variables are produced as simple averages of the quarterly results.