Page 14 - Invited Paper Session (IPS) - Volume 1

IPS30 Stefan D. et al.
administrative registers (such as the population registers maintained by the
municipalities, and the employment statistics register of the Federal
Employment Agency). Together with a complete enumeration of buildings and
dwellings (as no sufficient register information was available on such units)
and a supplementary sample survey (for variables on persons not available
from registers), a ‘census-typical’ data set was to be constructed. A census test
carried out in 2001 revealed that it was also necessary to use the
supplementary sample survey to correct for the errors detected in the registers
(see Statistische Ämter des Bundes und der Länder (2004)).
The demographic information available from the decentralised population
registers were completed – and where necessary, corrected by merging it with
information from other registers and mandatory primary surveys. By
combining different data sources and methods of automatic data generation,
a distinct data set containing all required census information could be created
for each person, each household and each building with dwellings. 1
In order to merge the data of the different parts of the census data
collection, first a basic register was established, containing a list of all
addresses where buildings with residential space existed at the census
reference day. This address and building register was the key link for all data
collections during the census. It was also used as the statistical population for
the sampling procedure of private households and for the housing census.
The main data sources used in the combined model were the following
(see figure 2):

• The population registers provided the main demographic data as well as
information on family relationships for all individuals that belong to the
target population (about 86 million data records). The data from the
municipal population registers were collected at the census reference day
(9 May 2011) and were updated three months later in order to take into
account delayed register entries and delayed deregistrations. The register
data were merged in a nationwide data set and it was subsequently tested
whether people were registered at a secondary place of residence only.

1 The following presentation is based on the English summaries provided by Bechtold (2013,
2016). A more detailed (German) description of the methodology applied in 2011 can be found
in Statistische Ämter des Bundes und der Länder (2015).
3 | I S I W S C 2 0 1 9