The Labour Force Survey is the most comprehensive continuous survey in Denmark. The survey is conducted quarterly and is based on a sample of the population. Each year 85,000 Danes aged 15-74 years participate. The Labour Force Survey in Denmark, called Arbejdskraftundersøgelsen (AKU), is the Danish contribution to the European Labour Force Survey and the survey is included in the unemployment statistics of Eurostat and the OECD. Data is collected and processed according to the same principles in all EU member states. The survey has been conducted continuously since 1994.
The main purpose of the Labour Force Survey is to cast light upon the attachment of the population to the labour market. The population aged 15-74 years is generally divided into employed persons, LFS unemployed persons and persons outside the labour force.
Furthermore the survey covers a range of other matters, and it is especially suited for:
In addition to the fixed questions, the questionnaire is regularly supplemented with additional questions with regard to a variety of subjects, e.g. disabled persons in the labour marker, work and family life or the transition to retirement.
The results from the Labour Force Survey are published quarterly in the publication Nyt fra Danmarks Statistik (News from Statistics Denmark).
The quality declaration provides information about the Labour Force Survey. You can find information about the purpose of the survey, the possible applications and who the typical users of the survey are. Furthermore there is information about the basic definitions and the methodological prerequisites that the LFS is built upon. In addition the quality declaration contains contact information on the administration of the Danish LFS.
The documentation contains information on the methodological basis for LFS. Here you can find papers on the most central statistical and methodological changes influencing the development of the LFS. Also more in-depth explanation to external users of the LFS that wish to publish figures from the LFS can be found. Furthermore it is possible to get the survey questionnaire and the codifications applied in the LFS. At the same time you can get information on the special modules that are often linked to the LFS (the so-called Ad hoc modules).
Here is a range of information about the Labour Force Survey central to the documentation of the survey. For example how the questionnaire is constructed, which codifications are used and furthermore papers describing the conceptual and methodological conditions and developments in the Labour Force Survey. These papers are supplementary to the declaration of content of the LFS. In addition, guidelines for the use of data from the LFS are described.
Statistics Denmark has described some guidelines for the use of data from the LFS. The purpose is to assure quality in the analysis based on the LFS and furthermore inform external users of the LFS on e.g. sampling errors. It is possible to achieve knowledge about publishing limits on yearly and quarterly basis. For further information:
The incoming results from the Labour Force Survey are weighted before publishing the results for the entire population. The method of weighting has been revised several times over the years, which can influence the development in the figures of employment and unemployment at the aggregate level, as well as the developments of figures for subgroups.
One should be aware of four revisions in the method of weighting: 2003, 2007, 2011 and 2015. In connection with the method of weighting in 2011, data going back to 2007 were revised. The present method of weighting was implemented in Q3 2015 and the method now includes a weighting method based on the panels. The new weighting method led to marginal changes in the data, and therefore the data was not revised back in time.
The actual effect of the revision in the method of weighting in 2011 is described in the paper below. Here you can read about in which way the revision in the method of weighting has influenced the level of employment in general and additional the size of subgroups (i.e. age groups, part-time and fulltime employees, educational groups and employed/self-employed persons).
Paper on the 2011 weighting-method (practically) (Danish version)
The theoretical considerations behind the changes in the method of weighting as well as the gains the revisions have led to on the LFS, are described in the following document. Here one can read about the background and the motivation for the revisions. Among other things, the revisions has led to a more precise age distinction and improved use of help information from several registers.
Paper on the 2011 weighting-method (theoretically) (Danish version)
Prior to 2007 a method implemented in 2003 was used. This revision implemented a correction where sex is corrected according to secondary age groups. At the same time it is described how the register of unemployment (CRAM) was used to divide the LFS-unemployed in the survey.
Paper on the 2003 weighting-method (Danish version)
The Labour Force Survey can be supplemented with additional questions, so-called “Ad hoc modules”. The main focus of the ad hoc modules changes yearly, but at the same time the construction “rotates” meaning that the same subjects are repeated regularly. In Denmark, all respondents participating in the survey during the second quarter are being asked the questions from the ad hoc modules.
How can one use the LFS ad hoc modules?
If your organization or firm is interested in the subjects from the ad hoc modules, it is possible to:
• Gain access to data from already completed ad hoc modules
• Link additional questions to forthcoming ad hoc modules
• Get involved in the development of ad hoc modules
For an overview of the ad hoc modules: Ad hoc modules 1999-2015 (Danish version)
Here you can find definitions of the central concepts in the LFS e.g. who is counted as employed, LFS unemployed and outside the labour force. In addition there are a range of papers that deepens the LFS concepts on unemployment, employment and working time, and compare them with the concepts of the register-based labour force statistics.
The definitions in the LFS comply with the guidelines as laid down by the ILO, the international labour market organization of the UN. The operationalization of the concepts is made by Eurostat, who coordinates the common European Labour Force Survey. The operationalization is hereby recommended by ILO and Eurostat.
The general definition of the attachment to the labour market is built up hierarchically implying that:
For further information about the concept of employment in the LFS and other relevant statistics read Paper on Employment concepts in the Labour Force Survey, National Accounts and Working Time Accounts and for a brief overview read Employment concepts (Danish version).
- Net unemployed: Persons who are actively seeking for and are ready to start a job (so-called match group 1), who receive unemployment benefits or social assistance and furthermore fulfill the criteria of LFS unemployment.
- Activated persons: Persons in activation who receive unemployment benefits or social assistance and who are ready to start a job (match group 1) and furthermore fulfill the criteria of LFS unemployment. The net unemployed together with the activated persons ready to start a job form the group of gross unemployment.
Other groups that will not be included in the registered unemployment but will be included in the LFS are:
- Students: This group will typically not be included in the registered unemployment, while this group receives educational support (SU) which is not an unemployment benefit. If students, as well as other persons, fulfill the definition of unemployment in the LFS, they are included in the LFS unemployment.
- Other LFS-unemployed: Other LFS-unemployed is a group that can contain all the persons who fulfill the definition of unemployment in the LFS, but are not gross unemployed or students. This group is very inhomogeneous. Some of the typical groups will be unemployed who are not able to receive unemployment benefits or social assistance (i.e. because of their spouse’s income or because of a missing membership of an unemployment insurance or unemployed who are not ready to start a job (match group 2 and 3), but have still not surpassed to early retirement etc. and hereby left the workforce)).
Read more about the concept of unemployment in LFS and other statistics Paper on Unemployment in LFS and other Danish statistics (Danish version) and for an overview Unemployment concepts (Danish version)• Outside the labour force are: er: Persons, who neither comply with the employment definition, nor the unemployment definition. This could be minors, pensioners and students, who either do not have a job or have actively been seeking for a job.
Other central concepts for the LFS are:
• Reference week: The specific week that the respondent is asked about. Whether you are employed or LFS unemployed, how many hours you have worked during the week etc., is related to the specific reference week. The date of the interview can be up to four weeks after the reference week, typically 1-2 weeks after. There are 13 reference weeks per quarter.
• Weighting method: The way in which the sample is weighted to the entire population, in order to make the results as representative as possible. It is always weighted figures that are being published. The method of weighting practically means that each person participating in the LFS gets his or her own weight and hereby represents a specific sample of the population with regards to sex and age. The method of weighting has been revised several times over the years (read under “Documentation/Method” for more information).
• Sample bias (uncertainty): This covers the general uncertainty connected to sample based surveys as the Labour Force Survey. The uncertainty is approximately +/- 10,000 persons on the general quarterly unemployment figure and approximately +/- 20,000 persons on the general quarterly employment figure.
• Seasonal adjustment: A method that removes yearly recurrent patterns in the three time series: employment, unemployment and people outside the labour force. There is not introduced seasonal adjustment on more disaggregated levels. The purpose of seasonal adjustment is to take into account i.e. unemployment or employment caused by seasons that can affect the development.
• Working time: The Labour Force Survey asks about three different types of working time in a specific reference week; usual, contractual and actual working hours. For more information about working time read Paper on Working time in LFS and other Danish statistics (Danish version)
Statistics Denmark is also publishing register based statistics compiling the population’s labour market status. For further details see the subject pages and the papers present on this site.
For more information on the Register-based labour force, employment (RAS) Labour force participation
For more information on the registered unemployment Unemployment
For more information on the statement of employment in RAS, the Working Time Accounts (ATR) and Public Sector Employment (BFL) Employment
Here it is possible to get a description of the developments and breaks in the time series that have been analyzed more in-depth. There will be general information about the time series and short presentations of i.e. the employment series, unemployment series and the working time series.
The LFS has been conducted since 1994 and in general it is possible to compile consistent time series based on LFS data going back to 2000. Continuous LFS tables from 1996 and onwards are available from Statistics Denmark’s database StatBank Denmark. However, following adjustments in the method of weighting, changes have been made in some data series. As a consequence, these cannot be used in monitoring developments in specific quarters or years. Developments in long durations of time can however be applied.
It is furthermore possible to find yearly data for some selected estimates back to 1983 at Eurostat’s Stat bank. However the yearly data between 1983-1993 are based on results from the 2nd quarter and should therefore be treated with caution.
The employment series measure how many persons are employed having minimum 1 hour of paid work or are temporary absence in the reference week. As figure 1 shows the total employment in 2000 initially was 2,700,000. The following two years until 2003/2004, employment is relatively stable. In 2004 the employment raises with 23,000 persons compared to 2003, which indicates a raise on 0.9 percentages. It is important to notice that the employment series was affected in 2003 when a new method of weighting was implemented, and the stratification of the sample was changed. Both these conditions made the number of employed persons rise and are the reason for the development from 2003-2004.
In 2007 the sample was expanded and the panel structure changed, and furthermore data from 2007 and forthcoming have been revised on the basis of the method of weighting from 2011. This has led to a decrease in the estimated number of persons employed. The decline in the employment in the period of 2006/2007 is untypical, since the employment rises from 2005-2006 and again from 2007-2008. Consequently, caution should be made with respect to drawing conclusions on the basis of the development between the years 2006 and 2007, especially with regards to subgroups, while longer time series can be concluded upon.
More details about the development in employment Paper on Employment concepts in the Labour Force Survey, National Accounts and Working Time Accounts
The unemployment series in the LFS is built upon the ILO definition of unemployment, where a person is unemployed if he or she is not employed in the specific reference week. Besides fulfilling this criterion, furthermore it is demanded that the person is actively seeking a job and is able to start working within 14 days. If the person does not fulfill these criteria, one is categorized as being outside the labour force and not LFS unemployed.
At the same time one should take notice of the revisions in the method of weighting in 2003, 2007 and 2011. The latest revisions from 2011 revise the figures back to 2007.
Consequently, caution should be made with respect to conducting analyses directly on the basis of the developments over the years 2003/2004 and 2006/2007, especially with regards to sub groups. In longer time series the before mentioned periods are not so problematic.
More details about the development in unemployment Paper on Unemployment in LFS and other Danish statistics (Danish version)
The education series is also affected by the discontinuity in data series from 2006 to 2007. The percentage of the population aged 30-34 years with a higher education fell from 2006 to 2007 from 43 to 38 percentages. The percentage has risen since then and in 2009, 2010 and 2011 the percentage was once again above the 2020-goal (40 percentage). This development is a consequence of the 2011-weigthing method that has a better correction with regards to the bias on education caused by age.
This remarkable shift is first of all caused by the change in method of weighting in 2007. Before 2007 the LFS tends to have an overrepresentation of persons with higher education. With the implementation of the latest method of weighting from 2011 (that is revised back to 2007), the educational level falls. From here on the level of education in the LFS is more in compliance with the level of education in the register of education. The fall from 2006-2007 is thereby not a real decline in the educational level.
For the educational series, the changes caused by the method of weighting are so significant, that one cannot make long-term analysis over the years 2006-2007. Therefore it is recommended to initiate time series in 2007.
More details about the development in the educational series Paper on the development in the educational level in the LFS (Danish version)
The development in the average working time in the LFS (in hours)
In the Danish LFS there are three types of working time; the contractual, usual and actual working hours. The working time series do not contain any substantial changes in connection with the adjustment in 2006/2007 in general. However, this is not the case for the working time for persons aged 15-24 years. The average estimated working time has increased by one hour for persons aged 15-24 years from 2006 to 2007, which probably is due to the method of weighting. This must be taken into account in the case of long time series for the working time of young people. In line with this change, there has also been a shift from the group of persons working “1-15 hours” to the group of persons working “38-48 hours” among young people. However, this is primarily the case between the 4th quarter of 2006 and the 1st quarter of 2007. Subsequently, there is again a fall that is instrumental in reducing the general change in the level, which implies that the shift is problematic only to a minor degree.
More details about the development in the working time series Paper on Working time in LFS and other Danish statistics (Danish version)
Here it is possible to gain information on tailor-made analyses coupled to the LFS. One can include questions in the Danish LFS and furthermore it is possible to ask for special analysis on existing data.
If your organization or firm is in need of new or updated knowledge about the labour market e.g. social engagement, working time or undeclared work, the LFS is a straightforward alternative to initialize a survey from point zero.
The Labour Force Survey makes it possible to carry out special analyses that are based on the regular analyses of the LFS i.e. by introducing questions or tailor-made analyses.
In this regard tailor-made analyses are specific operations of data from the LFS that are not published elsewhere (i.e. in Statbank). An example of this could be the number of part-time employees that works nighttime in the capital area or how many people with the highest educational level that are currently working on temporary basis etc.
Special analyses cover variables defined by the user that are included in the questionnaire. These variables can be measured against all other variables from the LFS as well as variables from registers. An example of this could be the number of working accidents or the satisfaction one has with the membership of a union etc. This can be linked to the additional results from the LFS. Special analyses will be developed in close connection with the staff of the Danish LFS.
By using the LFS to get your specific subjects examined, you will have access to the staff of the LFS’ expertise, infrastructure, our respondents and other data.
We can promise a flexible solution that suits you well. We are able to deliver data three months after the quarter terminates.
The prize of tailor-analyses is depended on the numbers of interviews, the audience, the panel structure, which quarter is chosen, as well as the preparations and the complexity of the questions (simple or complex).