Research data centres of the Federal Statistical Office and the statistical offices of the Länder

Access to microdata | Anonymity of microdata

Absolute anonymisation | De facto anonymisation |
Project-related anonymisation | Formally anonymised microdata

As a rule, microdata of official statistics are subject to strict confidentiality. However, two special provisions of the Federal Statistics Law make it possible to pass on microdata for purposes of data analysis if certain requirements are met.

If the microdata cannot be allocated to the respondent or the person concerned, i.e. conclusions cannot be drawn from the data as to the body or person providing the information (absolute anonymity), these microdata may also be used outside the sphere of official statistics (Art.16 Para.1 No. 4 Federal Statistics Law).
For the purpose of scientific research, the Federal Statistics Law permits the provision of microdata which “can be allocated” to the respondents or statistical units only with an "excessive amount of time, expenses and manpower". A de facto anonymity of the data is presupposed here. However, the data may be provided in this form only to domestic "institutions entrusted with tasks of independent scientific research", and the recipients have to be incumbents or specifically sworn in (Art.16 Para. 6 Federal Statistics Law).

The provision of anonymised data in the research data centres belonging to the statistical offices of the Federation and the Länder is based on these provisions. The various degrees of data anonymisation will be described in the following.

Absolute anonymisation

Absolutely anonymised data are modified by coarsening or by the removal of individual variables to such a degree that an identification of the respondents is made impossible. Official statistics offers absolutely anonymised microdata in the form of Public Use Files (PUF). These can be made available to all those who are interested.

So far, such data records have been prepared e.g. for the statistics of public assistance and time use surveys. Efforts are made by the research data centres belonging to the statistical offices of the Federation and the Länder to extend this service.

| top

De facto anonymisation

Microdata are called de facto anonymised if deanonymisation cannot be ruled out completely but the data can be allocated to the respective statistical unit only with an excessive amount of time, expenses and manpower (Art. 16 Para. 6 Federal Statistics Law). Pursuant to that Law, de facto anonymised data may be made available to scientific institutions only and only for the purpose of scientific projects.

De facto anonymisation mainly aims at reducing the possibilities of allocating the values of a variable to the respective statistical units by careful information reduction and information modification while preserving the informational value in statistical terms. The cost and benefit of deanonymisation have to be analysed for each individual survey. Different anonymisation procedures may be used (aggregation, forming classes for values of a variable, etc.).[Literature]

However, de facto anonymity does not result only from the remaining informational value of the data but also from the possibilities existing for deanonymisation. If a microdata record can be regarded as de facto anonymised therefore depends above all on the context in which the data are accessible. It is of decisive importance what additional knowledge of the statistical units can be drawn upon and where the data are used. Depending on whether the microdata are used externally or at the statistical offices, de facto anonymity can be achieved with greater (off-site use, Scientific Use Files) or smaller losses of information (on-site use at safe centres, workplaces for guest scientists at the statistical offices).

De facto anonymised microdata may be used by foreign scientists only on the secure premises of official statistical agencies.

| top

Project-related (de facto) anonymisation

The project-related anonymisation of microdata also results in de facto anonymity within the meaning of Art. 16 Federal Statistics Law. In contrast to the preparation of Scientific Use Files, only the variables required under a concrete research project are supplied and anonymised. As the microdata can only be used on the premises of official statistical agencies at specific workplaces for guest scientists, often more information is retained with project-related (de facto) anonymisation than with de facto anonymisation of standardised Scientific Use Files

| top

Formally anonymised microdata

Although de facto anonymised data are suitable for many areas of scientific research, the possibilities of analysis are limited also by de facto anonymity. This holds especially for evaluations with a detailed regional and subject-matter breakdown. Moreover, not all microdata of official statistics are suited for de facto anonymisation or they can be anonymised only to a certain extent.

Therefore, the research data centres of official statistics provide an opportunity for data users to analyse formally anonymised microdata by way of remote execution. Formal anonymisation includes the removal of the direct identifiers while the scope of the variables and subject-matter and regional breakdowns are largely retained. Evaluation results are checked for cases to be kept confidential before being sent back to the data users.

| top