Extrapolating the information from the household survey
The 2011 Census was conducted according to a new and largely register-based method whose methodology differs from previous population censuses in Germany. The statistical offices used mainly data from administrative registers. In addition, a household survey was conducted among just under ten percent of the population, whose purpose is not only to determine the numbers of inhabitants but also to provide information on issues such as education and employment.
But how can reliable information be obtained for the whole population if just a small part of the inhabitants were questioned in the household sample survey?
Determining the official numbers of inhabitants
A major basis in determining the numbers of inhabitants in the 2011 Census is the data sets transmitted by the residents' registration offices from the population registers to the statistical offices for census reference date 9 May 2011. As the population registers can contain errors - because not all people are really registered at their current place of residence (undercoverage) and others are still recorded in the register although they moved away a long time ago (overcoverage) - the stock data of the population registers have to be corrected.
By means of the household survey, such undercoverage and overcoverage was ascertained for large municipalities with at least 10,000 inhabitants and was extrapolated for the relevant municipality. Then the register stock data were corrected statistically. However, undercoverage or overcoverage was not reported back to the population registers (statistical confidentiality).
Information on the population structure
The household survey is used not only to determine the number of inhabitants. It also collected data not contained in the administrative registers. This includes variables such as school education, activity status and migrant background. Such data are also extrapolated to the whole population.
