Statistical processing

Source data

KAS is based on the Labour Market Account (LMA). The data sources in LMA are various internal and external registers, e.g.:

• eIncome register
• The central business register
• The register with information about persons receiving public benefits
• The educational register
• The employment classification module
• The income register
• The population register
• The register for persons receiving maternity or sickness benefit

Frequency of data collection

The statistic is annually.

Data collection

During the production of the Labor Market Account (AMR_UN), an initial data processing of the individual data sources takes place, and these are stored in a source database. Subsequently, a cross-sectional data processing (also called overlap processing) is carried out, where information from the various registers is compared and corrected if necessary. Finally, the data are combined with other registers to add background information and to classify the entire population.

Data validation

The data foundation for KAS isThe Labour Market Account (LMA). LMA is produced both with and without an hourly standardization. The non-hourly standardized longitudinal register (LMA-UN) is the data foundation for RAS, and therefore the data validation takes place in LMA-UN.

During the production of AMR the most important data validation are:

1. Employee job data: The data source is the eIncome Register. Information is extracted on the workplace where the job is performed. The workplace information forms the basis for data on industry, sector, and geography. In some cases, the employer’s reporting is incorrect, and corrections are made as needed. The eIncome Register also contains information on the occupational code (DISCO_08) for employees in companies covered by the wage statistics. If the company is not covered by the wage statistics, the occupational information is added from the Occupational Classification Module when available.

2. Self-employed data: The data sources are the Business Statistics Register, the Income Statistics, the eIncome Register, and the Unemployment Statistics. These sources are individually validated when compiling information on the self-employed.

3. Absence due to sickness and maternity/paternity leave: The data come from the Maternity and Sickness Benefits Statistics. The data are processed to provide a preliminary determination of whether the absence is from employment or unemployment.

4. Cross-sectional data processing / data validation

The purpose of the cross-sectional data processing and validation is to delete, correct, or create labor market statuses in cases where the different data sources do not match. This is done through so-called overlap processing. The main overarching areas are:

• Selection of jobs for the self-employed based on a set of criteria
• Determination of whether absence due to maternity/paternity leave or sickness is from employment or unemployment
• Harmonization of information on supported employment

Subsequently, the data are linked to other registers and sources as needed.

Data compilation

Data processing in AMR takes place in several steps. The first step involves checking and processing data from various sources, which are then loaded into a source database. Data are loaded on people receiving public benefits, employees, self-employed, assisting spouses, students, recipients of maternity/paternity and sickness benefits, and pensioners. Additionally, paid hours are imputed for the self-employed and assisting spouses. The second step involves overlap processing of the data, during which “illegal overlaps” between labour market statuses are corrected and links between different statuses are established.

After the overlap processing, the population is classified according to international guidelines from the ILO, as described in more detail under "Groupings and Classifications". The guidelines include a set of priority rules for determining the primary attachment to the labor market. According to the guidelines, employment has higher priority than unemployment, while unemployment has higher priority than statuses outside the labor force. Subsequently, the data are linked to the Business Statistics Register to obtain background information (industry, sector, workplace municipality, etc.) about the workplaces of employed individuals. The data are also linked to the population statistics to construct the resident population in Denmark. In this process, the socio-economic group “other persons outside the labor force” is also imputed; these are individuals residing in Denmark who do not belong to a known socio-economic group.

The AMR-UN register is then formed. This is a non–full-time equivalent longitudinal register containing information on the population’s attachment to the labor market, and it forms the data basis for RAS. In RAS, the population’s attachment to the labor market is measured at the end of November, but based on AMR-UN, it is possible to measure labor market attachment at arbitrary points during the year.

AMR is also produced in a full-time equivalent version (AMR-TN). In this version, the entire population always has a labor market attachment of 37 hours per week. Some of these hours may, for example, be as employment, while other hours may be education or simply hours without attachment to any known socio-economic category. Read more about statistical processing in LMA](https://www.dst.dk/en/Statistik/dokumentation/documentationofstatistics/labour-market-account/statistical-processing).

Adjustment

No corrections of data besides what is described under data validation and data compilation.