170 likes | 185 Views
This report highlights the advantages, levels of anonymization, and possibilities for accessing German microdata for research purposes. It includes information on the German microdata as public use files, census data from the former GDR, and the microcensus.
E N D
Country report Germany III. Workshop ‘Integration European Census Microdata’Barcelona, July 9, 2008 Markus Zwick, Research Data CentreFederal Statistical Office Germany
Research Data Centers (RDC) in Germany • RDC of the Federal Statistical Office • RDC of the Federal Employment Agency at the Institute for Employment Research • RDC of the German Federal Pension Insurance Federal Statistical Office Germany, Dr. Markus Zwick www.forschungsdatenzentrum.de
RDC of official statistics • interface organisation between data producers and empirical science • consulting and service for the use of official microdata • possibility for access to microdata with low anonymisation level Federal Statistical Office Germany, Dr. Markus Zwick www.forschungsdatenzentrum.de
RDC: Advantages for the data producer • More research on our own data • Higher data quality • Greater network on researchers • More competence on data and research knowledge • International acceptance of RDC researcher • Better reputation of German research Federal Statistical Office Germany, Dr. Markus Zwick www.forschungsdatenzentrum.de
Level of Anonymisation Degree of confidentiality stronger anonymisationmethod delete direct identifier anonymisationmethod de-facto anonymised microdata fully anonymised microdata completemicrodata confidential microdata Degree of analysis potential Federal Statistical Office Germany, Dr. Markus Zwick www.forschungsdatenzentrum.de
Possibilities for microdata access - de-facto anonymised microdata (Scientific Use Files) - fully anonymised microdata (Public Use Files and Campus Files) - Visiting Researcher‘s Desktop - Controlled Remote Data Processing (Remote Execution) - Special Data Processing Federal Statistical Office Germany, Dr. Markus Zwick www.forschungsdatenzentrum.de
CAMPUS-Files as special Public Use File TargetApplication for university teaching, training in use microdata and simultaneous qualification by use a statistical programm like SAS, Stata or SPSS ConceptionSmall sample size and less variables primary not for scientific use Federal Statistical Office Germany, Dr. Markus Zwick www.forschungsdatenzentrum.de
German Microdata as Public Use File for the IECM project nine anonymised microdata files - census 1970 and 1987 for the Federal Republic of Germany - census 1971 and 1981 for the former German Democratic Republic - five micro censuses for the Federal Republic of Germany; 1973, 1982, 1987, 1991, 2001 Federal Statistical Office Germany, Dr. Markus Zwick www.forschungsdatenzentrum.de 8
1990´s 1980´s 1970´s 1970 . 1973 1982 1987 1991 2001 1987 1981 1971 German Microdata for the IECM project Federal Statistical Office Germany, Dr. Markus Zwick www.forschungsdatenzentrum.de 9
Census of the former GDR 1971 Characteristics and metadata two data files - Person file (demography, income, education, employment etc.) - Dwelling and building file (state of repair, occupancy, etc.) 16,4 mio. persons, 6,2 mio. households, 6 mio. dwellings metadata no codebooks at FSO and Federal Archive Archives of regional statistical offices in the former GDR states (Field of study, occupation codes) Federal Statistical Office Germany, Dr. Markus Zwick www.forschungsdatenzentrum.de 10
- matching by unique combination of variables in both files (person file and dwelling file) - anonymisation by local suppression and by top and bottom coding - systematic random draw to create a 25% sample - available Public Use File include 4,1 Mio person in 1,6 Mio household with 104 Variables Census of the former GDR 1971Collecting Data Federal Statistical Office Germany, Dr. Markus Zwick www.forschungsdatenzentrum.de 11
The political discussion about the census in the eighties in Germany - great social opposition against the census in 1983 - refusal after the decree of the Federal Constitutional Court - implementation of an ‘informational self-determination‘ as an civil right in the german law - census execution in 1987 Federal Statistical Office Germany, Dr. Markus Zwick www.forschungsdatenzentrum.de 12
Census of the FRG in 1987 - 63 Mio. person, 27 Mio. households with 210 variables - the census include a occupation, a building and a dwelling census - the Public Use File is planned as an 1% systematic random sample - anonymisation by local suppression and by top and bottom coding Federal Statistical Office Germany, Dr. Markus Zwick www.forschungsdatenzentrum.de 13
Microcensus for the FRG - annual 1% revolving household sample with obligation to give information by law - 800.000 person, 380.000 households with nearly 750 variables - Scientific Use File as 70% subsample for researcher in Germany Federal Statistical Office Germany, Dr. Markus Zwick www.forschungsdatenzentrum.de 14
Microcensus as Public Use File - microcensus Public Use File for 1973, 1982, 1987, 1991 and 2001 - 35% subsample - ca. 300 variables - anonymisation by local suppression and by top and bottom coding Federal Statistical Office Germany, Dr. Markus Zwick www.forschungsdatenzentrum.de
Multiple source mixed mode model Federal Statistical Office Germany, Dr. Markus Zwick www.forschungsdatenzentrum.de
Country report Germany III. Workshop ‘Integration European Census Microdata’Barcelona, July 9, 2008 Markus Zwick, Research Data CentreFederal Statistical Office Germany