Online Program

Methods for using administrative data to estimate survey variables not directly available from administrative sources

Keywords: Administrative Data, Efficiency, Variables

Statistical offices across the world are increasingly looking to make better use of administrative data when producing their outputs. This leads to increased efficiency for the statistics office and a reduction in burden on respondents, both highly desirable outcomes. For some variables, administrative sources can provide a direct alternative to a sample survey. However, there are many survey variables that statistics offices are required to produce which are not directly available from administrative sources. In order to fully replace a survey by an administrative source, it is often necessary to find methods to estimate such variables. This paper describes approaches to estimate a range of commonly required survey variables which are not directly available from administrative data, but can be estimated in some way through a combination of administrative and historical survey data. The work is part of the European Statistical System network project on the use of administrative and accounts data for business statistics. The paper describes results from a wide range of variables considered by the project, focusing in detail on the most useful examples.