Statistical processing

Source data

The AMFORA statistics are compiled on the basis of personal data from the municipalities, the State Education Grants and Loan Scheme, Register for Labour Market Statistics (FLEUR-register) on recipients of early retirement pay and The labour market policy register (Bestandsstatistikken). Both registers are operated by the National Labour Market Authority (STAR).

Frequency of data collection

Some of the administrative data is collected monthly and some of the administrative data is collected quarterly. The statistics is only published quarterly.

Data collection

All data necessary for the statistics is collected from administrative registers. The data is collected through SAS-files and FTP-servers.

Data validation

When the data files are collected we make a preliminary validation of each file. Here it is investigated if each file has the expected size, format, covers the right period of time and doesn't has to many invalid personal numbers. These preliminary validations can either cause a rejection or a data load to our database (PSD). Afterwards the data is loaded in the database (PSD) on a uniform format. After that illegal overlapping time periods where the same person is receiving different conflicting public benefits are eliminated before the total data amount is transmitted into a 'production database' (also inside PSD). Here we make a final validation of the data, where dynamics and levels on a macro level are investigated and compared with the previously loaded data transmissions.

Data compilation

All data necessary for the statistics is collected from administrative registers. The data is collected through SAS-files and FTP-servers. When the data files are collected we make a preliminary validation of each file. Afterwards the data is loaded in the database (PSD) on a uniform format. After that illegal overlapping time periods where the same person is receiving different conflicting public benefits are eliminated before the total data amount is transmitted into a 'production database' from where the statistics is published.

Adjustment

It is not normally used to impute data or make other sorts of enumerations before the statistics is published. In 2014 the statistics was seasonal adjusted for the first time. The statistics is seasonal adjusted from the 1. quarter of 2007 and onwards.