The Sample Design for the Register-Assisted Census 2011

Abstract

Within the context of the Europe-wide census elicitation in 2011 there will be the first population census in reunified Germany. In contrast to the last German census in 1987, where all households were interviewed, the new census will be conducted by means of a cost- and resource-effective register-assisted census. In addition to the register information, a household sample will be drawn. On the one hand this sample will provide information that is not included in the register, on the other hand it will allow for the estimation of over- and undercounts in the register. Reliable estimates for total values of interest are to be derived from the register and sample data. The aim of the research project, which was initiated by DESTATIS, is to elaborate an efficient sample design as well as to develop estimation strategies which allow accurate estimates for the census 2011. This article presents some important findings from the research project. However, one focus is on the description of an optimal sample design